Compensating Injured Railroad Workers Under the Federal Employer's Liability Act: Special Report 241 (1994)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Special Report 241 Compensating Injured Railroad Workers Underthe S Federal Employers' Liability Act

1994 TRANSPORTATION RESEARCH BOARD EXECUTIVE COMMITTEE Chairman: Joseph M. Sussman, JR East Professor and Professor of Civil and Environmental Engineering, Massachusetts Institute of Technology, Cambridge Vice Chairman: Lillian C. Liburdi, Director, Port Department, The Port Authority of New York and New Jersey, New YorkCity Executive Director: Thomas B. Deen, Transportation Research Board Brian J. L. Berry, Lloyd Viel Berkner Regental Professor and Chair, Bruton Center for Development Studies, University of Texas at Dallas John E. Breen, The Nasser I. Al-Rashid Chair in Civil Engineering, Department of Civil Engineering, The University of Texas at Austin Kirk Brown, Secretary, Illinois Department of Transportation, Springfield David Burwell, President, Rails-to-Trails Conservancy, Washington, D.C. L. G. (Gary) Byrd, Consultant, Alexandria, Virginia A. Ray Chamberlain, Vice President of Freight Policy, American Trucking Associations, Alexandria, Virginia (Past Chairman, 1993) Ray W. Clough (Nishkian Professor of Structural Engineering Emeritus, University of California, Berkeley), Structures Consultant, Sunriver, Oregon Richard K. Davidson, Chairman and CEO, Union Pacific Railroad, Omaha, Nebraska James C. DeLong, Director of Aviation, Stapleton International Airport, Denver, Colorado Jerry L. DePoy, former Vice President, Properties and Facilities, USAir, Arlington, Virginia Delon Hampton, Chairman and CEO, Delon Hampton & Associates, Chartered, Washington, D.C. Don C. Kelly, Secretary and Commissioner of Highways, Transportation Cabinet, Frankfort, Kentucky Robert Kochanowski, Executive Director, Southwestern Pennsylvania Regional Planning Commission, Pittsburgh James L. Lammie, President and CEO, Parsons Brinckerhoff, Inc., New York City William W. Millar, Executive Director, Port Authority of Allegheny County, Pittsburgh, Pennsylvania (Past Chairman, 1992) Charles P. O'Leary, Jr., Commissioner, New Hampshire Department of Transportation, Concord Brig. Gen. Jude W. P. Patin (retired), Secretary, Louisiana Department of Transportation and Development, Baton Rouge Neil Peterson, former Executive Director, Los Angeles County Metropolitan Transportation Commission, Long Beach, California Darrel Rensink, Director, Iowa Department of Transportation, Ames James W. van Loben Sels, Director, California Department of Transportation, Sacramento C. Michael Walton, Ernest H. Cockrell Centennial Chair in Engineering and Chairman, Department of Civil Engineering, University of Texas at Austin (Past Chairman, 1991) David N. Wormley, Dean of Engineering, Pennsylvania State University, University Park Howard Yerusalim, Secretary of Transportation, Pennsylvania Department of Transportation, Harrisburg Robert A. Young III, President, ABF Freight Systems, Inc., Fort Smith, Arkansas Mike Acott, President, National Asphalt Pavement Association, Lanham, Maryland (cx officio) Roy A. Allen, Vice President, Research and Test Department, Association of American Railroads, Washington, D.C. (cx officio) Andrew H. Card, Jr., President and CEO, American Automobile Manufacturers Association, Washington, D.C. (ex officio) Thomas J. Donohue, President and CEO, American Trucking Associations, Inc., Alexandria, Virginia (cx officio) Francis B. Francois, Executive Director, American Association of State Highway and Transportation Officials, Washington, D.C. (cx officio) Jack R. Gilstrap, Executive Vice President, American Public Transit Association, Washington, D.C. (cx officio) Christopher Hart, Acting Administrator, National Highway Traffic Safety Administration, U.S. Department of Transportation (cx officio) Vice Adm. Aihert J. Herberger, Administrator, Maritime Administration, U.S. Department of Transportation (cx officio) David R. Hinson, Administrator, Federal Aviation Administration, U.S. Department of Transportation (cx officio) Gordon J. Linton, Administrator, Federal Transit Administration, U.S. Department of Transportation (cx officio) Rose A. McMurray, Acting Administrator, Research and Special Programs Administration, U.S. Department of Transportation (cx officio) Jolene M. Molitoris, Administrator, Federal Railroad Administration, U.S. Department of Transportation (cx officio) Rodney E. Slater, Administrator, Federal Highway Administration, U.S. Department of Transportation (cx officio) Lt. Gen. Arthur E. Williams, Chief of Engineers and Commander, U.S. Army Corps of Engineers, Washington, D.C. (cx officio)

Special Report 241 Compensating Injuted Railroad Workers Underthe Federal Employers' Liability Act TRANSPORTATION RESEARCH BOARD National Research Council NATIONAL ACADEMY PRESS WASHINGTON, D.C. 1994

Transportation Research Board Special Report 241 Subscriber Categories IA planning and administration VI public transit VII rail Transportation Research Board publications are available by ordering directly from TRB. They may also be obtained on a regular basis through organizational or individual affiliation with TRB; affiliates or library subscribers are eligible for substantial discounts. For further information, write to the Transportation Research Board, National Research Council, 2101 Constitution Avenue, N.W., Washington, D.C. 20418. Copyright 1994 by the National Academy of Sciences. All rights reserved Printed in the United States of America NOTICE: The project that is the subject of this report was approved by the Governing Board of the National Research Council, whose members are drawn from the councils of the National Academy of Sciences, the National Academy of Engineering, and the Institute of Medicine. The members of the committee responsible for the report were chosen for their special competencies and with regard for appropriate balance. This report has been reviewed by a group other than the authors according to the procedures approved by a Report Review Committee consisting of the members of the National Academy of Sciences, the National Academy of Engineering, and the Institute of Medicine. This study was sponsored by the Federal Railroad Administration and the Federal Transit Administration, U.S. Department of Transportation. Library of Congress Cataloging-in-Publication Data Compensating injured railroad workers under the Federal Employers' Liability Act / Transportation Research Board, National Research Council. p. cm.—(Special report LSSN 0360-859X; 241) ISBN 0-309-05561-X 1. Railroads—Employees—Legal status, laws, etc.—United States. 2. Workers' compensation—Law and legislation—United States. National Research Council (U.S.). Transportation Research Board. Series: Special report (National Research Council (U.S.). Transportation Research Board) ; 241. KF3629.C66 1994 344.73'021—dc2O • 93-46860 [347.304211 CIP Cover design: Karen L. White Cover photographs courtesy CSX Transportation

Committee for Study of the Federal Employers' Liability Act CLINTON V. OSTER, JR., Chairman, Indiana University, Bloomington W. BRUCE ALLEN, University of Pennsylvania, Philadelphia JOHN ARCUDI, Workers' Compensation Commission, Hamden, Connecticut RICHARD J. BUTLER, Brigham Young University, Provo, Utah JAMES R. CHELIUS, Rutgers University, New Brunswick, New Jersey FRANK J. DOOLEY, North Dakota State University, Fargo JOHN T. DUNLOP, Harvard University, Cambridge, Massachusetts EVELYN S. FERRIS, Department of Insurance and Finance, Salem, Oregon DEBORAH R. HENSLER, RAND, Santa Monica, California JOHN L. MELVIN, MossRehab Hospital, Philadelphia, Pennsylvania HERBERT R. NORTHRUP, University of Pennsylvania, Philadelphia GARY T. SCHWARTZ, University of California, Los Angeles DOUGLAS F. STEVENSON, Stevenson, Rusin and Friedman, Ltd., Chicago, Illinois WAYNE K. TALLEY, Old Dominion University, Norfolk, Virginia EDWARD M. WELCH, Michigan State University, East Lansing Liaison Representatives THOMAS H. HARTMAN, Federal Railroad Administration, U.S. Department of Transportation ELAINE KING, Transportation Research Board CHARLES T. MORISON, JR., Federal Transit Administration, U.S. Department of Transportation Transportation Research Board Staff ROBERT E. SKINNER, Jr., Director, Studies and Information Services MARK R. DAYTON, Study Director STEPHEN R. GODWIN, Senior Program Officer JOSEPH R. MORRIs, Senior Program Officer NANCY A. ACKERMAN, Director, Reports and Editorial Services NAOMI C. KASSABIAN, Associate Editor

Preface Acting through its report on the 1991 Department of Transportation and Related Agencies Appropriations Bill, the House Appropriations Committee requested a comprehensive analysis of the Federal Em- ployers' Liability Act of 1908 (FELA). FELA prescribes a tort-based compensation process for injured railroad workers that has been a long-time source of controversy within the railroad industry. In response to this request, the Transportation Research Board, with partial funding support from the Federal Transit Administration and the Federal Railroad Administration, established a committee of 15 experts to assess the injury compensation system that has evolved under FELA and compare it with the no-fault compensation systems that cover most U.S. workers. The committee, under the chairmanship of-Clinton V. Oster,Jr., Professor of Economics in the School of Public and Environmental Affairs at Indiana University, included individuals with expertise in freight transportation and costing, workers' compen- sation economics and policy, tort law, safety, medical disability and rehabilitation, labor law and economics, and workers' compensation administration. The committee's task was, first and foremost, to take an unbiased look at FELA and other workers' compensation systems to measure their strengths and weaknesses and ultimately to compare them for V

vi • Preface their effectiveness in successfully addressing the problems created by workplace injuries. The committee addressed the chief points of con- tention between supporters and opponents of FELA including the level and cost of benefits, the administrative costs and delays in providing compensation, the amount of litigation and legal costs, the effect on employee relations, and FELA's safety incentives. The resulting report describes the operation of injury compensation under both FELA and workers' compensation systems and details the committee's findings and conclusions regarding both approaches. In accordance with its charge from the National Research Council, the study committee limited its analysis to technical matters and did not recommend any specific changes in federal policy. As part of the effort to understand the positions of both the sup- porters and opponents of FELA, the study committee met with repre- sentatives of railroad labor and railroad management. To develop the empirical information necessary to describe injury compensation un- der FELA and compare it with that under workers' compensation systems, railroad industry data were gathered from the Association of American Railroads, Burlington Northern, CSX Transportation, Norfolk Southern, Southern Pacific, Union Pacific, the Metropolitan Transporta- tion Authority of New York, and the Railroad Retirement Board. Infor- mation on workers' compensation systems was sought from numerous sources including the Workers' Compensation Research Institute, the National Council on Compensation Insurance, the state of Washing- ton, and the U.S. Department of Labor. One railroad provided detailed data on injuries and compensation that are used in Chapter 6; the railroad is not identified to maintain the confidentiality of the data. Because FELA and workers' compensation systems have different objectives and are parts of different benefit systems, there are, not unexpectedly, differences in the levels of benefits, the sources of costs, and the employee benefits that complement and substitute for em- ployer injury compensation. Because of differences in approach be- tween the tort-based compensation of FELA and the no-fault, strict liability systems of workers' compensation, few data are available that are directly comparable among the systems. Therefore, in reaching its conclusions, the committee relied not only on the analysis of available data, but also on its collectivejudgment and experience with compensa- tion systems for injured workers. ACKNOWLEDGMENTS The study committee's efforts were aided by the cooperation and assistance of many individuals within and outside the railroad industry.

Preface • vii The views of railroad management regarding FELA were provided by Edmund W. Burke, Executive Vice President—Law and Government Affairs at Burlington Northern; Robert W. Blanchette, Vice Presi- dent—Law and General Counsel, Association of American Railroads; Daniel Saphire, Assistant General Solicitor, Association of American Railroads; and H. Russell Smouse, Whiteford, Taylor, & Preston. The views of railroad labor were presented to the committee by Donald C. Buchanan, Director of the Railroad Division, Sheet Metal Workers International Association; G. Thomas DuBose, President, United Transportation Union; Mac A. Fleming, President, Brotherhood of Maintenance of Way Employees; Ronald P. McLaughlin, President, Brotherhood of Locomotive Engineers; W. D. Pickett, President, Brotherhood of Railroad Signalmen; Michael W. Babcock, Kansas State University; Robert Hayden, Representative, Railway Labor Exec- utives Association; Jim Ellenberger, Assistant Director for Occupa- tional Safety and Health, AFL-CIO; and Edward Wytkind, Executive Director of the Transportation Trades Department, AFL-CIO. In order to gain additional insight into the problems addressed, the committee invited presentations from industry experts, whose volun- tary contributions to the study committee's effort are gratefully ac- knowledged. Ed Codd, Assistant Vice President—Risk Management, CSX Transportation, presented in detail CSX's system for handling FELA claims and provided claims data. Randy Regula, District Director in the Office of Workers' Compensation Programs, U.S. Department of Labor, detailed the operation of the Longshore and Harbor Workers' Compensation Act and the Federal Employees' Compensation Act. Gary Dellaverson, Director of Labor Relations, Metropolitan Transpor- tation Authority of New York, discussed the effects of FELA on transit agencies. Information on the operation of FELA and empirical data for the study were provided by Dan Saphire of the Association of American Railroads, Kenneth Coy and Don Wind of Burlington Northern, Mag- gie Connor and Nicholas Anyasodor of the Metropolitan Transporta- tion Authority of New York, Henry Light and Sarah Corey of Norfolk Southern, Floyd Parker and Orvil Pilcher of Southern Pacific, and Dave Harbert of Union Pacific. Information on workers' compensation and federal compensation systems was provided by David Durbin, National Council on Compensation Insurance; Richard Victor, Executive Direc- tor, Workers Compensation Research Institute; and John Ruser, U.S. Department of Labor. At the request of the study committee, a number of individuals estimated the benefits that injured railroad workers would receive under the regulations and benefits of alternative compensation systems

viii • Preface for a sample of one railroad's FELA claims. Robert McCallister, Wash- ington State Board of Industrial Insurance Appeals, arranged for the estimates under Washington's system, and the estimates were prepared by Sandy Dziedzic, Quality Assurance Manager, Washington State Department of Labor and Industries. The Michigan estimates were produced jointly by John Meskus, Ervin Vahratian, and study commit- tee member Edward Welch. The contributions of these individuals are gratefully acknowledged. Their assistance expanded the understanding of the study committee and staff. However, the findings and conclusions expressed in this report, along with any errors or omissions, are solely those of the study committee. The study was performed under the overall supervision of Robert E. Skinner, Jr., Director, Studies and Information Services, Transporta- tion Research Board. Mark R. Dayton served as study director and, under the guidance of the committee, prepared the Executive Sum- mary, Chapters 1 through 4 and 6 through 8. Stephen R. Godwin drafted Chapter 5 and Joseph R. Morris drafted the modal competi- tion section of Chapter 7. The final report was edited and prepared for publication under the supervision of Nancy A. Ackerman, Director, Reports and Editorial Services, Transportation Research Board. Naomi Kassabian was the editor for the report. Frances E. Holland and Marguerite E. Schneider typed the manuscript and throughout the study provided assistance in meeting logistics and committee correspondence.

ContentS Executive Summary 1 Introduction 11 Railroads and the Transportation Industry, 14 Historical Development of Compensation Systems for Injured Workers, 15 Criticisms and Defenses of FELA, 19 2 Overall Context for Compensation of Injured Workers 26 FELA Benefits within the Overall Compensation Context, 28 Workers' Compensation Benefits within the Overall Compensation Context, 34 3 Criteria for Analysis of Compensation of Injured Workers 38 Disability and the Basis for Compensation, 38 Goals for Compensation Systems, 44 Legal Concepts of Injury Compensation, 49 Assessment Criteria for Injury Compensation Systems, 53

4 Railroad Injury Compensation Process 59 Overview of FELA Process, 60 Industrywide Data on FELA Process and Costs, 73 Summary, 78 5 State and Federal Workers' Compensation Programs 81 Program Operation, 83 Compensation for Loss, 85 Medical Coverage, 86 Compensation for Lost Wages, 87 Rehabilitation, 99 Compensation for Survivors, 102 Dispute Resolution, 103 Coverage of Workers in Interstate Commerce, 111 Summary, 112 6 Comparison of FELA Process and Workers' Compensation Systems 119 Data Problems and Sources, 120 Levels of Benefits, 121 Comparison of Railroad and Workers' Compensation Benefits, 124 Comparison of FELA and NCCI Data, 129 Case Study Comparison, 131 Comparison Between Commuter Rail Under FELA and Rail Transit, 137 Transactions Costs, 138 Summary, 146 7 Modal Competition and Federal Appropriations 149 Modal Competition, 150 Federal Appropriations, 154 8 Findings and Conclusions 159 General Findings, 159 Specific Findings, 160 Opportunities for Improving the FELA Process, 166 Appendix A: Railroad Safety 168 Appendix B: Terminology 181 Study Committee Biographical Information 184

Executive Summary Railroad workers who are injured on the job seek compensation for their injuries under the provisions of the Federal Employers' Liability Act of 1908 (FELA). This act prescribes a tort-based approach to compensation for on-the-job injury as opposed to the no-fault ap- proach of workers' compensation systems applicable to most U.S. workers. For many years railroad management has argued that the FELA process imposes higher costs on the industry than workers' compensation systems do and thus places railroads at a competitive disadvantage with other transportation modes. Railroad labor leaders, on the other hand, argue that FELA is no more costly to administer than workers' compensation systems and that the higher benefits pro- vided are fair to workers and provide incentives for the railroads to improve workplace safety. In response to this debate, the House Appropriations Committee, in its report on the 1991 U.S. Department of Transportation appropria- tions bill, requested that the Transportation Research Board conduct a comprehensive study of FELA. The main study objectives were an analysis of FELA and other approaches to compensating injured workers and a comparison among them. In accordance with its charge, the study committee did not recommend specific changes in federal policy.

2 • Compensating Injured Railroad Workers Under FELA THE FELA PROCESS Many employment and compensation practices for railroad workers date from an era in which railroads were the premier industry in the United States and working for a railroad was especially hazardous. The enactment of FELA strengthened railroad workers' rights to sue their employers for all damages related to on-the-job injuries without limit or restriction, including compensation for medical expenses, lost wages, disabilities, future earnings losses, and pain and suffering. To collect, the injured worker has to demonstrate negligence on the part of the employer, and awards can be reduced depending on the degree of employee negligence through a "comparative negligence" standard. An injured worker can receive no compensation at all if the injury is judged to be entirely the worker's fault. Although the essential characteristics of FELA law have changed little since 1908, the process has evolved over the years into a more administrative system. Medical expenses for on-the-job injuries are now handled administratively through a health insurance plan (though they remain elements of damages that can be recovered, if necessary, at trial or settlement). Railroads and their employees have developed procedures for dealing with routine claims so that 70 percent are handled without litigation or employee legal representation. In addi- tion, the courts have effectively reduced the standards of proof re- quired to demonstrate negligence on the part of the employer. Currently, FELA applies to approximately 273,000 workers, most of whom are employed by major freight railroads. The total includes roughly 45,000 National Railroad Passenger Corporation (Amtrak) and commuter railroad employees. The total also includes employees of regional railroads, who may not have the same benefits found in the major labor collective bargaining agreements. WORKERS' COMPENSATION SYSTEMS In the same year that FELA was enacted, Congress established a work- ers' compensation system for federal employees whose underlying approach was markedly different from that of FELA. Instead of re- quiring that an injured worker prove employer negligence to receive compensation, this early workers' compensation system provided com- pensation regardless of who was at fault and benefits that were not reduced for the employee's negligence. The no-fault approach proved popular, and today more than 90 million U.S. workers are covered by workers' compensation systems that incorporate the no-fault approach. Each state has established its own workers' compensation system, and the federal government oversees two systems, one for federal govern-

Executive Summary • 3 ment employees and one that covers private employees in longshoring operations. Workers' compensation is a liability system that mandates insurance. Governments set benefit standards and provide oversight; employers are legally required to provide coverage through self-insurance, pri- vate carriers, or state funds; and insurance carriers deal directly with injured workers, typically with little involvement by state agencies. The details of workers' compensation systems can vary considerably from state to state, but all systems feature limitations on types of damages covered and limits on the benefits paid. In particular, workers' compen- sation systems do not provide compensation for pain and suffering. COMPARISON OF FELA PROCESS WITH WORKERS' COMPENSATION SYSTEMS No injury compensation system for workers exists in isolation from other workplace benefits. The value of either the FELA process or workers' compensation systems to injured workers depends in part on the availability of other worker benefits that may complement or substi- tute for the benefits provided through the injury compensation system (see Chapter 2). Thus, comparisons between the FELA process and state workers' compensation systems must be viewed in the context of overall workplace benefits. Moreover, because these other benefits may be provided through collective bargaining agreements, changes in the injury compensation system that result in reduced levels of statutory benefits are likely to induce a response in the next round of collective bargaining that might partially offset the changes to workers' benefits and employers' costs. Over the years, the FELA process and workers' compensation sys- tems have evolved in ways that make them increasingly similar. As the standards for proving employer negligence have lessened, FELA has functioned less and less as a negligence-based, trial-driven tort system, and the majority of FELA claims are now settled without litigation or employee legal representation. At the same time, litigation within workers' compensation systems has been on the rise, although disputes are usually handled through an administrative proceeding that is less costly and generally faster than court actions. Despite this evolution, the two approaches to compensating injured workers remain quite different. Compared with state workers' compensation programs, the FELA process generally provides higher, more extensive benefits to injured workers but can result in delays in payments, involves some- what higher transactions costs, and results in higher costs to railroad employers.

4 Compensating Injured Railroad Workers Under FELA More specific study findings are presented in the following discus- sion for each of the critical areas of evaluation relating to injury com- pensation systems: benefits to injured workers, the costs of providing these benefits, and incentives (see Chapter 3). Before these specific findings, a review is presented of the underlying approach and the basis on which the comparisons were made. Overall goals of injury compensation involve equity, efficiency, and incentives. Ideally, an injury compensation system should be equitable to the injured worker, should provide benefits in an efficient manner, and should be structured so that each party has incentives to reduce both injuries and the costs of those injuries that occur. A system's efficiency and incentive structure can be assessed objec- tively, but the fairness of any particular system depends on more subjective perspectives or social philosophies of individuals or groups. The criteria that may be considered injudging the fairness of a particu- lar injury compensation system, however, can be defined and investi- gated. They include the extent of coverage, including who and what is compensated; the level of the compensation for losses; the speed with which the losses are compensated; the certainty with which they are compensated; and who bears the costs of compensation. No attempt was made in this study to define an equitable injury compensation system. Instead, to the extenCdata are available, these criteria were compared for both FELA and alternative compensation systems. FELA and workers' compensation systems have different ben- efits, in part because they set out to do different things and so make different trade-offs between and among the criteria. A final caveat should be stated regarding the following discussion. The comparisons that were made in this study had to use incomplete and rarely comparable data. Although the results should therefore be viewed as rough comparisons and rough estimates of magnitudes, they are the best possible at this time. Future efforts to analyze FELA and compare its delivery of benefits to other systems would be greatly improved if the railroads were to begin collecting and classifying their injury compensation data within categories comparable with those used in workers' compensation systems. FINDINGS Benefits to Injured Workers Benefits to injured workers involve questions of the extent of coverage, the levels of compensation provided, the degree of certainty in the system, and the relative cost burden for losses.

Executive Summary • 5 Do All Injured Receive Compensation? Neither FELA nor workers' compensation compensates all injured individuals. FELA does not compensate individuals who are solely negligent in causing their injury, whereas workers' compensation does because it is a no-fault system. In practice, however, FELA fails to compensate injured workers in only a small percentage of cases. Workers' compensation systems do not provide benefits for individuals who do not meet minimum lost-work-time requirements and may not recog- nize certain impairments or occupational illnesses as compensable, whereas all of these losses are reimbursable under FELA. Do Like Injuries Receive Like Compensation? Under FELA, there is a potential for different settlements for the same injury. Identically injured workers may seek and receive different compensation if they incur different consequences from the injury. Furthermore, for claims that go to trial, different juries and jurisdic- tions may provide varying results. Railroads, however, use tacit for- mulas or rules of thumb in their negotiations with injured employees (and plaintiffs' attorneys); therefore, more uniformity exists than might be expected. Workers' compensation systems, on the other hand, tend to have formulas and schedules for compensation that make benefit payments more certain and comparable among workers with similar injuries. Nevertheless, these formulas and schedules can vary signifi- cantly among states and some states do not use them. Moreover, some workers' compensation claims are resolved through negotiation rather than according to administrative schedules. What Are the Medical Benefits? Medical coverage is the same under both FELA and workers' compen- sation: 100 percent of medical costs associated with the injury is cov- ered. Some data suggest that medical costs are lower in the FELA process than in workers' compensation. This may be because FELA medical benefits are provided under the regular railroad health insur- ance plan and therefore are perhaps subject to tighter scrutiny and cost control. Steady, large increases in medical costs are a continuing and growing problem in workers' compensation systems. Both systems' approach to medical benefits is likely to be affected by national health care reform, but it is not possible to determine at this time what changes will take place.

6 • Compensating Injured Railroad Workers Under FELA What Are the Indemnity Benefits? FELA provides a higher level of benefits than workers' compensation systems because all losses are compensable without preset limitations. Unlike workers' compensation, FELA has no cap on the level of wage- loss recovery or on recovery for impairment, and the process allows compensation for pain and suffering. FELA benefits are, however, subject to reduction for comparative negligence. Although compari- sons are complicated by inconsistent data, available information indi- cates that FELA indemnity payments (payments for all nonmedical losses) to injured workers are about two to four times greater than they might be under workers' compensation systems for comparable inju- ries. Part of this difference is accounted for by the higher wage levels of railroad workers compared with average state levels and benefit caps in workers' compensation systems. The remainder can be attributed mainly to the broader coverage under FELA for all out-of-pocket costs and for nonmonetary losses such as pain and suffering. The scope and levels of injury compensation are functions of each system's compensation philosophy. Because of the no-fault trade-off made in workers' compensation, all workers receive some award, but not all losses are reimbursed. On the other hand, in the negligence- based approach of FELA, all losses are potentially compensable, but not all workers are compensated. Lower compensation benefits, how- ever, are not inherent in a no-fault system, just as a tort-based system such as FELA does not have to compensate for nonmonetary losses. Ultimately, federal and state legislatures control what is compensated. How Certain Is Compensation? More uncertainty exists about receiving benefits and their levels under FELA than under workers' compensation systems. However, railroads have voluntarily adopted a routinized system of benefit delivery for most claims that reduces this uncertainty. Under workers' compensa- tion there is little uncertainty concerning the receipt of benefits, but the level of benefits in some states may be less certain, especially for injuries with lasting impairment. What Is the Relative Burden Share of Losses? The efficiency and incentives of a negligence-based compensation system depend in part on the relative burden for the costs of injury that the system places on each party. The relative burden share between the railroads and their workers is not known because the comparative

Executive Summaiy • 7 negligence and the employee's estimated total losses for each claim (for which compensation was sought) are not known. There is a fixed burden in workers' compensation that forms part of the trade-off for no-fault coverage; the employer bears the monetary losses for the most part and the employee bears the nonmonetary losses. Costs of Running the System The costs for an injury, compensation system include the direct cost of the benefits and the indirect or overhead costs of the transactions necessary to deliver those benefits to the injured worker and support his or her rehabilitation. What Are the Direct Costs to Employers? FELA compensation to injured workers has grown moderately since 1987, with continuous increases in real compensation per injury claim. There was a short-term expansion in occupational illness claims that was responsible for part of the increase in overall costs, but that may have run its course. Nevertheless, claims for occupational illnesses such as silicosis, carpal tunnel syndrome, and mental stress have the poten- tial for increasing costs in the future. Clearly, the higher and more extensive benefits in FELA create higher direct costs for railroads than the same injury would incur under alternative workers' compensation approaches. What Are the Transactions Costs for Both Employees and Employers? Transactions costs appear to be lower in workers' compensation systems than under FELA. FELA has become a routinized system of compensa- tion in which settlements are achieved primarily through negotiation and claims rarely go to trial. Delays in FELA settlements, however, create costs for both railroads and workers that are less likely in workers' compensation, which is more successful in rapid replacement of wages. Litigation occurs in both the FELA process and workers' compensa- tion systems. Overall, it appears that there is more litigation in the FELA process and that cases are more protracted. Lump-sum settle- ments cannot be made until all actual and potential losses are known or estimated, and given the overall higher level of benefits in FELA, there is more. at stake and more to negotiate. There is also considerable litigation in many workers' compensation systems, and although litiga- tion increases are not universal, many systems are experiencing them.

8 • Compensating Injured Railroad Workers Under FELA Workers' compensation litigation takes place mostly before administra- tive bodies rather than courts and juries and is likely to be less expensive than FELA litigation. Finally, there is some reason to believe that FELA is more efficient in transactions costs than other tort systems. Incentives The incentives that are of primary concern in an injury compensation system are those that promote workplace safety, rehabilitation and the return to work of the injured worker, and efficient dispute resolution. What Are the Safety Incentives? Both FELA and workers' compensation provide incentives to improve safety, and there is no clear empirical evidence to support FELA's claims to superiority. The costs of both systems are likely high enough to provide safety incentives. An injury compensation system can promote workplace safety through incentives that deter unsafe behavior and aid the discovery of unsafe conditions. If there were an allocation of losses under FELA solely in accordance with each party's negligence in the injury, FELA might strongly discourage unsafe behavior, more so than workers' compensa- tion systems do, because the injured worker could potentially be ex- posed not only to uncompensated nonmonetary losses (as in workers' compensation) but also to uncompensated monetary losses arising from the injury. The lack of evidence concerning the relative burden sharing for compensation between the railroads and their workers prevents a firm conclusion on this point. The discovery of unsafe conditions may be hampered more under FELA than workers' compensation. Because the investigation of an accident or injury may bear on the question of the railroad's and the employee's negligence, the willingness of all parties to cooperate in such an investigation may be reduced. No empirical evidence was available on this point, however. The data show that there are substantial differences among the railroads in the frequency of injuries and fatalities. This would suggest that the railroads with higher injury rates might have opportunities to improve their situation by more emphasis on safety, regardless of what injury compensation law is in effect. What Are the Return-to-Work Incentives? Workers' compensation may provide greater overall incentives to re- turn to work than FELA, but there are no good data to make definitive

Executive Summary • 9 findings on this issue. Workers' compensation systems deal more di- rectly with rehabilitation and return-to-work costs than does the FELA process. Nevertheless, the delays in settlement under FELA do not necessarily result in delayed return to work for the injured worker. Furthermore, the lump-sum settlements in FELA may provide an incentive for the worker to return to some work, whether in the original job or a new one. What Are the Dispute Resolution Incentives? Under FELA, the dispute resolution process is more adversarial, and the incentive for quick resolution of the dispute may be less than in workers' compensation systems. As noted previously, this finding should be expected in a tort-based system. Just how great the disincentive is under FELA is not clear from the data available. Modal Competitiveness and Federal Appropriations The effect that FELA may have on railroad modal competitiveness and federal appropriations is indeterminate. If cost-saving changes can be made to the railroad injury compensation system, the freight railroad market share or profitability may increase and federal appropriations for both Amtrak and railroad transit operations may decrease. Modal Competitiveness A system like workers' compensation would likely be less expensive for railroads, perhaps resulting in either higher profits or a larger market share, or both, but the effect is likely to be modest. Some savings from reductions in injury benefits may eventually be put into satisfying higher wage demands as well as to be shared by stockholders and customers, and thus the actual savings could be reduced. Whatever the initial effects appear to be, they would almost certainly be less after the system had had time to adjust to the change. Federal Appropriations There is perhaps some potential for savings in outlays if Amtrak or commuter railroad agencies move to an alternative system or make improvements in FELA. Outlay savings depend on whether reduced injury compensation costs are translated into reduced appropriations, are used to finance additional passenger and transit services, or are mitigated by changes in wages or benefits to transit and Amtrak workers.

10 • Compensating Injured Railroad Workers Under FELA OPPORTUNITIES FOR IMPROVING THE FELA PROCESS Changes to FELA depend on a willingness of the railroads and their employees to discuss solutions. Indeed, nearly all labor and social legislation affecting the railroad industry has resulted from negotia- tions between the carriers and their employees. Because of the com- plexity of the issues, links to negotiated compensation packages, and the lack of clear advantages to any one approach, joint labor and management consultation is probably the most promising approach to reducing the costs of injury compensation in the railroad industry. Although the study committee was asked not to make specific recom- mendations about improvements to FELA, the consensus was that it would be wise if all changes to FELA were developed through negotiation. Regardless of what changes are made to FELA or what alternatives are proposed, reductions in the costs of providing injury compensation would come mainly from reduced benefits or from reduced transac- tions costs. Any decrease in injury benefits would result in reduced overall worker income. Employees would likely resist such an effective wage cut and might seek a compensating change elsewhere in their overall wage and benefit package. Savings from reductions in transac- tions costs through decreases in delays, litigation, and administrative costs are likely to be more feasible because such reductions would benefit both employees and the railroads. Improving the dispute resolution process is clearly one area in which it is possible to make constructive changes to FELA to the advantage of both sides. The parties might well explore the range of alternative dispute resolution (ADR) methods that are increasingly being used in the public and private sectors. If the level of litigiousness and delay in all settlements could be reduced, both injured workers and railroads could benefit.

I ............................ Introduction The Federal Employers' Liability Act of 1908 (FELA) established the process through which railroad workers may seek compensation for on-the-job injuries. Unlike most U.S. workers, railroad workers—in- cluding those of freight railroads, the National Railroad Passenger corporation (Amtrak), and commuter railroads—are not covered by no-fault workers' compensation systems. Instead, a tort-based process prescribed by FELA compensates injured workers. Under this process, injured railroad workers can sue their employer, but they must prove employer negligence to collect. Injury compensation can be reduced or eliminated to the extent that employees are judged to have negligently contributed to their injuries. On the other hand, if employer negli- gence is proved, individual awards may exceed, sometimes substan- tially, the levels typically provided by no-fault workers' compensation systems. U.S. railroad management has become increasingly critical of FELA in recent years. They note that because of FELA's unique provisions, railroad companies bear a higher cost for compensating on-the-job injuries than do their competitors and other employers. They point out that FELA allows unlimited damage awards and claim that FELA imposes higher administrative costs than other injury compensation systems. The system's reliance on lawyers and litigation is said to result 11

12 • Compensating Injured Railroad Workers Under FELA in high legal fees for both employees and employers. Protracted delays in resolving cases are said to not only increase costs for employers but also delay benefits for employees. Furthermore, railroad management believes that the tort approach creates an adversarial relationship with their employees. Perhaps most important to the railroad industry, they argue that FELA places them at a competitive disadvantage with other transportation modes, particularly trucking. The leaders of railroad unions, however, are strong supporters of FELA. They argue that FELA is no more expensive to operate than state workers' compensation programs and that it provides injured employees with better benefits. They point out that most FELA claims are handled routinely and without a lawyer and that only a tiny per- centage of cases proceed through a full trial to verdict. FELA suppor- ters also argue that railroad work is much more hazardous than most other jobs. Thus, the unlimited damage awards possible under FELA act as a strong incentive to the railroads to provide a safe workplace. Congress began reexamining the issues concerning FELA during the 1980s. A 1986 study by the General Accounting Office (GAO) of Amtrak's FELA costs for 1984 estimated that FELA costs Amtrak (and the taxpayers who subsidize it) $2.7 million more per year than the company would have paid under a high-benefit state workers' compen- sation system and $14.4 million more than it would have paid under a low-benefit system (GAO 1986). Subsequently, Amtrak requested a 3-year exemption from FELA, and hearings were held in the Senate in 1988 and the House in 1989 (the first hearings on FELA held in 50 years). The exemption, however, was not approved. Nevertheless, both Amtrak and the freight railroads have continued to advocate changes in the current law. In response to this policy debate, the House Appropriations Committee report on the 1991 Department of Transportation appropriations bill (U.S. Congress. House. 1990, 115-116) requested that the National Research Council's Transportation Research Board "conduct a comprehensive study of the issues related to retention, modification, or repeal of the Federal Employers' Liability Act, including an analysis of appropriations uti- lized for this purpose." (See the accompanying box for the full text of the relevant section of the House report.) The findings and conclusions reached by the Transportation Research Board study committee are given in this Special Report. The remainder of this chapter consists of background information about the nature and size of the railroad industry; the historical devel- opment of compensation systems for injured workers, including the FELA process and workers' compensation systems; and the specific criticisms and defenses that have been raised concerning FELA by the

Introduction • 13 EXCERPT FROM CONGRESSIONAL REPORT CALLING FOR TRB STUDY The National Transportation Policy recommends the repeal of the Federal Employers' Liability Act (FELA). According to the policy, FELA imposes "a variety of outmoded and archaic requirements" on the railroads, which keep them "from playing the most efficient and effective role in the transportation system and competing on an equal basis with other modes." Administration officials believe that FELA results in unnecessary federal costs and postpones needed medical rehabilitation. However, the Committee has also received testimony from labor organization officials, who believe that FELA provides the most appropriate form of compensation in a hazardous industry. The merits of both sides have been debated for several years with no comprehensive study. The Committee notes that the continuation or repeal of FELA is a legislative matter outside the jurisdiction of the Committee. However, oversight of federal appropriations utilized for FELA expenses are within the jurisdiction of the Appropriations Com- mittee, as stated in clause 2(B)(3) of Rule X of the House of Representatives. While taking no position on the merits of the FELA issue at this time, the Committee requests that the Transportation Research Board of the National Academy of Sciences conduct a compre- hensive study of the issues relating to retention, modification, or repeal of the Federal Employers' Liability Act, including an anal- ysis of appropriations utilized for this purpose. This is intended to be an independent study which considers equally the views of proponents and opponents of FELA. (U.S. Congress. House. 1990, 115-116) affected parties. In Chapter 2 the FELA process and workers' compen- sation systems are put in context with other employee benefits that can assist injured workers. In Chapter 3 a framework for evaluating the issues involved in compensating injured workers is laid out. The FELA process for compensating injured railroad workers is described in detail in Chapter 4, and in Chapter 5 comparable information is pre- sented about state and federal workers' compensation systems that represent alternative approaches to compensating work injuries. The purpose of Chapter 6 is to present more-detailed information about

14 • Compensating Injured Railroad Workers Under FELA the cost and performance of the FELA process using railroad industry data assembled as part of this study and where possible to compare the cost and performance of FELA with those of state and federal workers' compensation systems. Finally, the estimated effect, under various assumptions, of FELA on modal competitiveness and federal appro- priations is analyzed in Chapter 7. Committee findings and conclusions about FELA are presented in Chapter 8. RAILROADS AND THE TRANSPORTATION INDUSTRY Transportation accounts for about 17 percent of the gross national product in the United States (Smith 1993).' Within the overall trans- portation sector of the economy, railroads no longer play the dominant role that they once did. Nevertheless, they continue to be a significant and vital component of the nation's transportation system. Railroads are especially important to the intercity freight market for which they typically haul bulk, lower-valued commodities over long distances. They also are still an important factor in intercity and commuter passenger travel in selected corridors and markets. Railroad workers from all three of these segments—freight, intercity passenger, and commuter rail—are covered by FELA. Freight railroads handle 37 percent of the freight transported in the United States on the basis of ton-miles, account for 8 percent of the nation's freight bill, and employ roughly 226,000 workers (Smith 1993; AAR 1992). Of these railroads, 13 Class! companies account for about 75 percent of the track mileage, 89 percent of the employees, and 91 percent of the freight revenue (Table 11).2 Of the Class I railroads, seven are considered the major carriers: Consolidated Rail Corpora- TABLE 1-1 Freight Railroad Characteristics, 1991 (AAR 1992) TYPE OF RAILROAD NUMBER OF CARRIERS ROUTE- MILES OPERATED EMPLOYEES AT YEAR END FREIGHT REVENUE ($ MILLIONS) Class I (revenues 13 129,839 200,958 26,949 > $96.1 million) Regional (revenues 32 19,662 11,382 1,362 $40 million to $96.1 million) Local (revenues 490 24,307 13,657 1,305 <$40 million) Total 535 173,808 225,997 29,617 NOTE: 1 mi = 1.6 km.

Introduction • 15 tion (Conrail), CSX Transportation, and Norfolk Southern in the East and Burlington Northern, Santa Fe, Southern Pacific, and Union Pa- cific in the West. The other six, Chicago and North Western, Florida East Coast, Grand Trunk Western, Illinois Central, Kansas City South- ern, and Soo Line, are smaller and more regional in nature.3 The trucking industry is the principal competitor of the railroads in the area of freight transportation. It carries 26 percent of the nation's domestic freight on a ton-mile basis, accounts for 77 percent of the domestic freight bill, and employs, in its for-hire component, roughly 1.5 million workers4 (Smith 1993). In contrast to the railroad industry, trucking is highly decentralized with many small firms. Total Class I, II, and III for-hire carriers regulated by the ICC numbered more than 46,000 in 1991.5 Furthermore, the top 15 Class I trucking firms ac- counted for only 50 percent of the 1991 revenues of regulated for-hire carriers; 200 firms reported gross revenues over $35 million (Commer- cial Carrier 1992; ICC 1992). Except for selected markets, railroads do not retain a significant share in passenger transportation. All intercity rail passenger service is provided by Amtrak, which operates equipment on rights-of-way that are for the most part owned by the private railroads. Nationwide, Amtrak serves about 22 million passengers each year, which is equiva- lent to about 5 percent of the air passenger market, and it employs 24,000 workers (Smith 1993; NRPC 1993). Many more passengers, about 325 million each year, are carried by commuter rail services (Smith 1993). Commuter railroads provide vital capacity to Northeast Corridor commuters in the Baltimore, Boston, New Jersey—New York City, Philadelphia, and Washington, D.C., markets. Other cities with established commuter railroad operations include Chicago, Miami, and San Francisco (Middleton 1991). Approx- imately 21,000 workers are employed in commuter rail operations in the United States (APTA 1992). HISTORICAL DEVELOPMENT OF COMPENSATION SYSTEMS FOR INJURED WORKERS FELA and Refinement of Tort Approach Many procedures and requirements governing the employment and compensation of railroad workers date from an era in which railroads were the premier U.S. industry, employed the largest work force in the United States, and faced little competition from other modes. During this era railroad work was extremely hazardous. At the turn of the century, 1 in every 8 train workers was injured on the job and 1 in every

16 • Compensating Injured Railroad Workers Under FELA 125 was killed every year; the average life expectancy of switchmen after starting work was 7 years (Elkind 1981). The federal government had been wrestling with the railroad safety problem for some time before FELA was enacted. As early as 1893 Congress passed the Federal Safety Appliance Act, which mandated certain safety equipment on engines and cars. As the hazards of rail- road work continued unabated, Congress, with pressure from orga- nized labor, began reforming the legal system under which employees sought compensation. At that time, there was no systematic corporate or public provision for those injured, nor was there support for the families of those killed on the job.6 The concept of guaranteeing compensation to injured employees, though adopted in Europe during the 1800s, had not yet gained widespread acceptance in the United States. Injured workers at the time could seek compensation from their employer by suing for dam- ages under the common-law doctrine of negligence. Under this negli- gence standard, employers had a number of legal responsibilities to their employees, including to provide and maintain a reasonably safe place to work, and safe appliances, tools and equipment; to provide a sufficient number of suitable and competent fellow employees to permit safe performance of the work; to warn employees of unusual hazards; and to establish and enforce proper safety rules. (NCSWCL 1973, 11) The employer could be held negligent for failure to perform these duties, though the burden of proof was on the employee. "As the test of the performance of the employer's duty extended only to proper diligence, breach of this duty was not easy to prove in court" (NCSWCL 1973, 12). Moreover, failing one of the due care tests, the employer still could invoke three defenses to liability: the doctrine of contributory negligence, the fellow-servant doctrine, and the doctrine of assump- tion of risk. Through the use of the principle of contributory negligence, employers were able to avoid liability if they could demonstrate that any negli- gence lay with the employee. Under the fellow-servant doctrine, the injured employee could not establish employer negligence solely based on the negligence of a fellow employee.' Employers could also use a defense based on the assumption-of-risk doctrine, which held that the employee in continuing to work after knowing of a particular danger assumed the risks of injury or death entailed by that particular danger.

Introduction • 17 Because of the difficulty experienced by injured workers in recover- ing under the common law, states passed a variety of employer liability laws that sought to limit the application or eliminate one or more of the common-law defenses to liability or to invalidate employment contracts that relieved employers of any negligence (Fishback 1987). Between 1855 and 1911, 25 states adopted some form of employer liability law (NCSWCL 1973, 51). No state passed a law to eliminate all of the defenses, and most sought to limit the scope of the fellow-servant doctrine. In legislation passed in 1906, Congress attempted to make it easier for injured railroad employees to prevail in court. The Supreme Court, however, ruled the law unconstitutional because it attempted to regu- late intrastate commerce. In 1908 FELA, which was restricted to inter- state commerce, passed constitutional muster. FELA "was the high point in this phase of employee protection" because it "contained all of the most advanced features of state laws up to that time" (NCSWCL 1973, 14). FELA made it easier for injured workers to seek compen- sation by eliminating the fellow-servant defense and restricting the assumption-of-risk doctrine. The act also eliminated the defense of contributory negligence and replaced it with the concept of compar- ative negligence. This change allowed the employee to recover propor- tionate damages when part of the responsibility for the accident was the employee's own negligence. Around 1900, states began examining and adopting in limited forms alternative, no-fault approaches to compensating injured workers. In 1911, Wisconsin became the first state to adopt and judicially sustain the first workers' compensation statute that replaced the principle of negligence with a shared strict liability standard. The strict liability was shared in the sense that both the employee and employer bore some of the costs because compensation was substantially less than the em- ployee's full common-law damages (Chelius 1976). A congressional commission created in 1910 recommended aban- doning FELA and replacing it with a no-fault system. Proposals to establish a no-fault system actually passed both the Senate and House in 1912, but minor differences could not be reconciled in conference and the bills expired (Fox 1990). Further efforts to replace FELA in the next several years "were defeated largely because of the opposition of the railroad brotherhoods" (Weiss 1935, 594). Though efforts to re- place FELA continued through the 1930s, Congress made incremental changes to FELA rather than replacing it. The last significant amend- ments to FELA, including the elimination of the assumption of risk defense, were enacted in 1939.

18 • Compensating Injured Railroad Workers Under FELA Since enacted in 1908, FELA has been modified considerably by judicial interpretation in addition to the foregoing congressional amend- ments. Most important, over the years, Supreme Court decisions have substantially reduced the standard for employers' negligence that em- ployees have had to show under FELA. This reduced standard has made it much easier for injured employees to gain standing in court and to win compensation. Some claim that this has made FELA a no- fault system. However, unlike in no-fault programs, employees covered by FELA can still secure full common-law damages in court (Schwartz and Mahshigian 1986; Tidd and Saphire 1988). FELA currently applies to approximately 273,000 workers.8 The Class I railroads, which account for about three-fourths of this total, spent $911 million in FELA damage claims in 1991. Workers' Compensation and Refinement of No-Fault Approach In 1908, the same year that Congress passed the tort-reform approach of FELA, it enacted a no-fault workers' compensation system for fed- eral employees.9 Actually, the very first workers' compensation pro- gram in the United States was enacted in 1882, but it was limited to federal workers in the most hazardous "life-saving" work, such as rescue workers in the Coast Guard. The 1908 legislation, though a landmark in many ways, applied only to one-fourth of federal workers, again mainly those in hazardous occupations, and compensation was available only for 2 years, was limited to traumatic injuries, and re- quired a 15-day waiting period. Even many of its supporters were unsatisfied with the final bill, and efforts for additional reforms began almost immediately (Nordlund 1991). The debates leading to the 1908 federal workers' compensation legislation and subsequent efforts to improve it ultimately forged the consensus for much-expanded programs for both federal workers and state-covered privately employed workers. In 1916 Congress reformed the federal workers' compensation program to include all federal em- ployees, to compensate occupational disease as well as injury, to extend duration limits, and to cover medical costs (Nordlund 1991). Even before 1916, state legislatures had passed a flurry of laws applying to privately employed individuals. Although the constitu- tionality of state laws was not resolved until 1917, between 1911 and 1921 all but six states enacted some form of workers' compensation coverage (Nelson 1988). The state programs shared the common prin- ciple that employers should provide benefits for income maintenance and medical coverage as a result of injuries on the job, regardless of

Introduction • 19 fault. However, the state programs differed markedly from each other in terms of the extent of coverage and the benefits provided. The last state program was not in place until 1949. In 1992 employers paid $62 billion to support workers' compensa- tion (Burton 1993). The programs cover over 90 million employees, or about 85 percent of all workers. Agricultural workers and state and local government employees represent the largest categories of exempt employees, though they are not exempt in all states. Almost all other private employees are covered by one of the state programs. Two major federal workers' compensation programs exist—one that covers em- ployees of the federal government (established by the Federal Em- ployees' Compensation Act of 1908 and expanded by the Act of 1916) and another that covers private employees in longshoring operations (established by the Longshore and Harbor Workers' Compensation Act of 1972).10 CRITICISMS AND DEFENSES OF FELA Railroad management and labor, as noted earlier, have opposing views on FELA. Railroad management would like to change the law and move toward a national no-fault system similar to state workers' com- pensation systems. Railroad labor supports the current FELA statute and believes that it has established a very beneficial system for railroad workers. The arguments for and against FELA provided to the study committee by railroad labor and management are summarized in this section. Those arguments in favor of FELA are based largely on a presentation to the study committee by a number of railroad union presidents.1 ' Management arguments against FELA were presented by the Association of American Railroads (AAR). The major areas of disagreement concern worker coverage and benefits; the cost of the systems, including the direct costs of benefits and the time and litigation involved in settling claims; and the safety incentives with the FELA process. On the most general level, manage- ment claims that the FELA process is inordinately costly and that costs are increasing. These costs come from both the actual settlements with injured workers and the time and expenses involved in producing those settlements within a tort system. Labor, on the other hand, claims that although FELA awards may be greater than those received through workers' compensation systems, they are fairer to injured workers; labor maintains that workers' compensation payments are too low, not that FELA payments are too high. Furthermore, labor argues that litigation and delay are no worse than the current experience in workers' compensation systems.

20 Compensating Injured Railroad Workers Under FELA Worker Coverage and Benefits The degree of worker coverage and the level of benefits are key sources of the dispute over FELA. Labor makes the case that FELA provides better coverage by permitting full economic recovery, including full replacement of actual wage losses and future earning potential, full medical costs and rehabilitation payments, and compensation for pain and suffering. Although management contends that similar cases lead to widely different results and that some injured workers receive noth- ing, labor sees the variability as a strength in that it allows settlements to be designed to match the particular circumstances of each case. With regard to the level of benefits and the resulting direct costs of settlements, management believes that settlements are higher than they would be under alternative workers' compensation systems, in part because of the evolution of the negligence standard over the years by Supreme Court decisions. The standard for demonstrating causation has been reduced, and on the question of negligence, the amount of information necessary to take the case to the jury has been reduced. Tidd and Saphire (1988) claim that this has made FELA nearly a no- fault system but with much higher awards available through jury trials. By contrast, in workers' compensation systems, award levels are gener- ally set by formula and depend on earnings but are limited by pre- scribed maximum monthly payments. The workers' compensation process leads to more predictable results that would let railroads gauge their potential liability compared with the uncertainty associated with jury awards for pain and suffering. Labor argues that the higher awards under FELA are more an indication of the weakness of workers' compensation systems than a failing of FELA's. Limits on the level and duration of wage recovery under many state workers' compensation systems result in less than full wage compensation for injured workers. These caps compel much lower wage-loss benefits than the average railroad worker's earnings. FELA allows workers to seek full wage recovery. 12 Furthermore, workers' compensation systems are viewed as being unfair to workers because they bar the recovery of damages for pain and suffering. Costs from Delay and Litigation One of the main criticisms of FELA by railroad management is that the tort-based system increases the time required for resolution of claims and the legal costs to both railroads and workers. They contend that negotiation and litigation over the comparable negligence of the parties and the level of settlement beyond wage recovery create long delays in settling cases. Such delays raise the cost to the railroads of administer-

Introduction • 21 ing the claims process and legal costs accrue in the negotiation and settlement process and for litigation when negotiation is not successful. The AAR reports that $288 million was spent for administrative costs and legal fees in 1991. The railroads also note that the extended negotiation process, especially for serious claims, forces employees to hire lawyers to pursue their claims and may result in a net award that is less than the railroad's last offer. Labor counters that in the majority of cases there are no delays, because FELA is a de facto administrative system rather than strictly a litigious one. Labor argues that the level of litigation is no more exten- sive than in workers' compensation systems. Furthermore, labor notes that some workers' compensation programs have been experiencing rapid increases in costs, and in some states, the number of disputed claims is growing rapidly. There is also the question of the incentive that each program pro- vides for rehabilitation and the return to work. Some workers' compen- sation programs have specific incentives in this regard. Management claims that although workers may eventually be rehabilitated under FELA, there is a strong incentive for them to delay rehabilitation so that juries will be more sympathetic to their claims (RAILS 1992). This delay increases the cost to the employer and prevents the workers' return to health and work. Labor believes that FELA does not delay return to work as much as workers' compensation systems force premature return to work. That is, the ability to recover full wage loss prevents a railroad worker from having to return to work before complete recovery has occurred and while rehabilitation is continuing, whereas limits on wage replacement during convalescence under workers' compensation force workers to resume their jobs too soon. Finally, management believes that the higher cost associated with FELA benefits raises their cost of operations relative to that of their competitors in the trucking and barge industries. This affects their ability to compete with other transportation modes that have less costly injury compensation systems. Labor argues that if this is indeed the case, the solution is not to force railroad labor into an inferior no-fault compensation system but to extend the better benefits of FELA to the other modes. Safety Railroad labor contends that the railroad environment is an inherently risky one with a high rate of accidents and severe injuries and that

22 Compensating Injured Railroad Workers Under FELA FELA provides a strong incentive to hold railroads accountable for the safety of their operations. They argue that the large awards available to workers under FELA act as strong inducements to eliminate unsafe conditions, to avoid violations of federal safety laws, and to invest in safety. Under FELA, if the employee shows that a railroad violation of one of the federal safety statutes was involved in the injury, negligence is automatically presumed. Railroad management counters that industry safety has improved dramatically since FELA was enacted, especially during the past decade with the capital investment that has been made since partial deregula- tion under the Staggers Rail Act of 1980. The industry claims that railroad work is not more dangerous than work in comparable indus- tries and that no special provisions are necessary for rail workers. Management also contends that the cost of damaged equipment and freight is a strong inducement in and of itself to avoid accidents. Railroad management further contends that the fault-based nature of the FELA system fosters an adversarial relationship between man- agement and labor. This impedes the ability of both parties to fully understand and correct unsafe workplace practices. Full investigation of accidents becomes difficult because the incentives for each party are not to cooperate in revealing any facts or circumstances that would indicate their fault. This undermines the working relationship with employees and hinders efforts to improve the safety of the workplace environment. NOTES Unless noted otherwise, all data cited are for 1991, the last year for which complete data are available. Class I is a designation of the Interstate Commerce Commission (ICC) based on annual revenues as shown in Table 1-1. The Association of American Railroads (AAR) uses different standards for dividing the Class II and Class III carriers. The ICC standard is based strictly on an- nual revenue, and in 1991 it was as follows: Class II, revenue between $19.1 million and $96.1 million; Class III, revenue less than or equal to $19.1 million. The AAR defines regional railroads as line-haul railroads with less than 350 mi of track or revenue between $40 million and the Class I threshold ($96.1 million in 1991), or both. Local carriers are the short- line carriers that do not meet the regional criteria and include switching and terminal railroads (AAR 1992). Southern Pacific lines include the Denver and Rio Grande Western Rail- road and Southern Pacific Transportation. Grand Trunk Western is a part of CN North America.

Introduction • 23 The 1.5 million excludes drivers and delivery persons for private concerns, which total roughly 2.1 million workers (Smith 1993, 61). Class I motor carriers are those with gross operating revenues in 1991 equal to or greater than $6.5 million. Class II carriers are those with revenue between $1.3 million and $6.5 million. Class III motor carriers are those with revenue less than $1.3 million (ICC 1992). Some railroads did have individual provisions for injured workers, but there was no industrywide approach. The general tort rule is that the plaintiff can show the negligence of the employer based on the negligent action of the employee. Under the fellow- servant doctrine, this showing does not suffice to establish employer lia- bility. Rather, the injured employee must establish the employer's own managerial negligence (Schwartz 1989). Not all employees of public agencies with railroad operations are covered by FELA. The number of current railroad workers participating in the Railroad Retirement Board Program as of July 1993 was 273,000, which should match the total number covered by FELA. The original, limited federal employees' compensation program was passed in 1908, but the more comprehensive program that became the Federal Employees' Compensation Act was passed in 1916. Maritime workers covered by the Jones Act are compensated under the tort-based approach of FELA. The remarks, "Rail Labor Presentation to the FELA Study Committee of the Transportation Research Board," constitute the primary source for this section. The presentation was made on behalf of the American Train Dispatchers Association, Brotherhood of Locomotive Engineers, Brother- hood of Maintenance of Way Employees, Brotherhood of Railroad Signal- men, Hotel Employees and Restaurant Employees International Union, International Association of Machinists, International Brotherhood of Boilermakers and Blacksmiths, International Brotherhood of Electrical Workers, International Brotherhood of Firemen and Oilers, International Longshoremen's Association, National Marine Engineers' Beneficial As- sociation, Railway Labor Executives Association, Sheet Metal Workers International Association, Transportation Communications International Union, Transportation Trades Department of the American Federation of Labor-Congress of Industrial Organizations (AFL-CIO), Transport Workers Union of America, and the United Transportation Union. FELA awards are calculated on an after-tax basis, but are not subject to taxes. Workers' compensation awards are, generally, two-thirds of before- tax income as an average approximation of after-tax income and are not taxable. Conventional tort awards, on the other hand, are calculated on a before-tax basis and, yet, are not taxable either. FELA is, therefore, more like workers' compensation in this respect than other tort actions. However, there are no caps on potential wage recovery as there are in all workers' compensation systems.

24 • Compensating Injured Railroad Workers Under FELA REFERENCES ABBREVIATIONS AAR Association of American Railroads APTA American Public Transit Association GAO General Accounting Office ICC Interstate Commerce Commission NCSWCL National Commission on State Workmen's Compensation Laws NRPC National Railroad Passenger Corporation AAR. 1992. Railroad Facts, 1992 ed. Washington, D.C., Sept. APTA. 1992. 1992 Transit Fact Book. Washington, D.C., Oct. Burton, J., Jr. 1993. Workers' Compensation 1960-1990, the Increases, the Causes, and the Consequences. John Burton's Workers' Compensation Monitor, March-April, p. 1. Chelius,J. R. 1976. Liability for Industrial Accidents: A Comparison of Negli- gence and Strict Liability Systems. Journal of Legal Studies, Vol. 5, June, pp. 293-309. Commercial Carrier. 1992. Top 100. Commercial CarrierJournal, Vol. 149, No. 7, July, pp. 55-73. Elkind, A. 1981. Should the Federal Employers' Liability Act Be Abolished? The Forum, Vol. 17, No. 2, pp. 415-422. Fishback, P. V. 1987. Liability Rules and Accident Prevention in the Workplace: Empirical Evidence from the Early Twentieth Century. Journal of Legal Studies, Vol. 16, June, pp. 305-328. Fox, B. 1990. Has FELA Run Off the Tracks? The Brief, Vol. 19, No. 2, pp. 22-27. GAO. 1986. Amtrak—Comparison of Employee Injury Claims under Federal and State Programs. Washington, D.C., Aug. ICC. 1992. Interstate Commerce Commission 1992 Annual Report. U.S. Department of Commerce. Middleton, W. D. 1991. Commuter Railroad Planner's Guide: A North Ameri- can Overview. Railway Age, Nov. NCSWCL. 1973. Compendium on Workmen's Compensation. Washington, D.C. Nelson, W., Jr. 1988. Workers' Compensation: 1980-1984 Benchmark Revi- sions. Social Security Bulletin, Vol. 51, No. 7, July, pp. 4-17. Nordlund, W. 1991. The Federal Employees' Compensation Act. Monthly Labor Review, Sept., pp. 3-14. NRPC. 1993. 1992 Annual Report. Washington, D.C. RAILS. 1992. The Federal Employers' Liability Act: The Case for Repeal. Washing- ton, D.C., April. Schwartz, G. 1989. The Character of Early American Tort Law. UCLA Law Review, Vol. 36, No. 4. Schwartz, V., and L. Mahshigian. 1986. The Federal Employers' Liability Act, a Bane for Workers, a Bust for Railroads, a Boon for Lawyers (Torts Sym- posium). San Diego Law Review, Vol. 23, No. 1, Jan.-Feb., pp. 1-15. Smith, F. A. 1993. Transportation in America: A Stati.sticalAnalysis of Transportation in the United States, 11th ed. Eno Foundation for Transportation, Inc., Landsdowne, Va.

Introduction • 25 Tidd, J., and D. Saphire. 1988. The Case for Repeal of the Federal Employers' Liability Act. Working Paper Series No. 23: Critical Legal Issues. Washington Legal Foundation, April. U.S. Congress. House. 1990. Committee on Appropriations. Report 10 1-584. 101st Congress, Second Session, July 10. Weiss, H. 1935. Employers' Liability and Workmen's Compensation. In History of Labor in the United States, 1896-1 932, Vol. 3, Reprints of Economic Classics, Augustus M. Kelley Publishers, New York, 1966, pp. 564-610.

2 Overall Context for Compensation of Injured Workers Neither the benefits attainable under the Federal Employers' Liability Act of 1908 (FELA) nor those of no-fault injury compensation systems exist in isolation from other workplace benefits. The value of any particular injury compensation system to the injured worker depends in part on the other worker benefits outside that system that comple- ment or substitute for the benefits within the system. As a result of injury, a worker may suddenly face costs for medical treatment and rehabilitation, wages lost because of the inability to work during reha- bilitation for an acute injury, potential income lost from a long-term decline in income-earning ability, and nonmonetary loss of physical function that affects nonworkplace activities. The injured worker could look to either general health and sickness benefits for help in meeting these costs or to a specific injury compensation system, or perhaps to both. Table 2-1 gives the benefits available to injured workers in the railroad industry through FELA or other railroad benefit programs and compares these with benefits that may be available to injured workers covered by workers' compensation. Railroad benefits may be provided by employers, employees, a third party such as the Railroad Retirement Board (RRB), or some combination of these. Under FELA an injured worker can sue to recover all of his or her incurred costs, both monetary and nonmonetary. If the worker is not 26

TABLE 2-1 Potential Sources of Benefits for Injured Workers: FELA Compared with Workers' Compensation BENEFIT FELA WORKERS' COMPENSATION (WC) Medical treatment Travelers' Medical Insurance Plan Workers' compensation medical (100 percent coverage) - benefits (100 percent coverage) Compensation for lost wages, Sickness benefit Lost time < WC minimum short term (Railroad Retirement Board) Sick leave/salary continuation policy Short-term disability insurance (employer) (employee) Short-term disability insurance Wage continuation (employing (employer, employee) railroad) Lost time > WC minimum Cash advances (employing railroad) WC indemnity payment Compensation for lost wages. FELA settlement WC indemnity payment long term Long-term disability insurance Long-term disability insurance (Railroad Retirement Board) (social security disability) Long-term disability insurance Long-term disability insurance (employer, employee) (employer, employee) Compensation for pain and suffering FELA settlement None Medical rehabilitation Travelers' medical plan WC medical coverage FELA settlement Employer health plan Vocational rehabilitation FELA settlement WC vocational rehabilitation payments

28 • Compensating Injured Railroad Workers Under FELA successful in negotiation or trial, there are still other benefits available to compensate some of the losses. These benefits are also available to cover losses while the injured worker is waiting for a FELA settlement. Similarly, employers and government agencies often provide benefits that complement workers' compensation benefits and some that may substitute for them if an injured worker does not qualify for workers' compensation. The specific interactions between FELA benefits and other benefits available to railroad workers are described in the next section, followed by a discussion of how workers' compensation benefits may be comple- mented or substituted for by other benefits available to injured workers. FELA BENEFITS WITHIN THE OVERALL COMPENSATION CONTEXT Benefits for the injured railroad worker are not restricted to specific tort remedies through FELA. Instead, various programs and policies have evolved to complement FELA that either were designed specifi- cally for that purpose or were driven by considerations outside work- place injury and compensation. In either case, the benefits and costs of FELA are best understood in this broader context of social and private insurance and benefit programs. The various programs within the railroad industry are outlined here with the goal of establishing a clear distinction between benefits that arise from FELA and those that are independent of it. No benefit programs, however, exist in a vacuum. The impetus for development of benefits and their eventual shape are affected by the universe of exist- ing benefits that may substitute for or complement them. In that sense, no benefits are completely independent, though some are more di- rectly affected by the existence of FELA than others. Medical, Sickness, and Disability Benefits Medical, sickness, and disability benefits are available to the injured railroad worker that are independent of the worker's ability to recover damages under the FELA statute. The coverage of on-the-job injuries by these benefits and their levels, however, are without doubt influ- enced by FELA. Medical Benefits Health insurance coverage to pay the medical cost of all sickness and injuries is a standard part of many industries' benefits for their workers.

Overall Context for Compensation of Injured Workers • 29 However, for most workers' compensation systems, the costs of on-the- job injuries are covered in the premiums for workers' compensation, which are separate from regular health insurance premiums. For the railroad industry, the cost of on-the-job injury has been incorporated into the basic health insurance coverage, which is not a requirement of the FELA statute. Because medical treatment and rehabilitation for on-the-job injuries are included in their basic health insurance coverage, railroad workers do not need to show employer negligence to be eligible for these benefits. The cost of medical treatment for on-the-job injuries is cur- rently covered fully by the railroads and the level of medical treatment provided to injured railroad workers is comprehensive and flexible. The railroad industry has a group health insurance policy with the Travelers Insurance Company.1 Premiums for the policy are paid fully by each railroad without any employee matching requirement. Pre- miums are calculated separately fo'r on-the-job injuries and off-duty illnesses and injuries. Basic copayments under the policy are 20 percent by the employees and 80 percent by the insurance company. For on- the-job injuries, however, the employee's 20 percent is paid by the railroad, so 100 percent of the cost of medical treatment is covered. It is likely that national health care reform will affect how medical care is provided under both FELA and workers' compensation. As this is being written, it is impossible to predict the eventual system adopted, but it seems likely that it will narrow the differences between FELA and workers' compensation. It is possible the health care portions of both systems will eventually be taken over by a national health care system. FELA benefits for the coverage of some injuries (e.g., heart attacks) or some occupational illness claims (e.g., mental stress) may be con- tested by the railroads if the connection to on-the-job factors is not well established. Medical coverage under nonwork provisions of the health benefits would be provided in such cases but indemnity payments would not. Whether mental stress claims are compensable under FELA is an issue that is currently before the courts. If the work-related showing standard established by court decisions is weak, mental stress claims could be wide-spread and involve significant costs for indemnity benefits. At the same time, a more stringent standard could lead to lower indemnity costs but higher litigation costs for establishing eligibility.2 Sickness Benefits Sick-leave benefits are another nonwage benefit of employment often offered by companies to their employees. Whether sick leave takes the

30 Compensating Injured Railroad Workers Under FELA form of a set number of days per year or a general wage continuation policy, the goal is to maintain wage income during the period of sickness. Some employers do not offer any sick-leave benefits; if em- ployees do not work, they do not get paid. Railroads do not provide sick-leave benefits directly to employees, but do so indirectly by paying a payroll tax. The railroad tax and sickness benefits are administered and provided to employees by the RRB. The railroad worker may be entitled to these benefits when injured and forced to miss time from work.3 The RRB is an independent agency of the federal government that administers retirement, survivors', unemployment, and sickness bene- fit programs for railroad workers. The sickness benefit is specified by provisions of the Railroad Unemployment Insurance Act (RU IA) that were added to the general unemployment benefits in 1946. The worker need show no negligence to receive these benefits; they are the em- ployee's right by virtue of employment in the industry. Unemployment and sickness benefits are funded by a wage tax paid entirely by the railroads. The tax rate has been experience based since 1991; that is, it is set for each railroad individually on the basis of the use of the benefits by its employees in prior years. The tax rates in 1992 ranged from 3.1 percent to 12 percent of wages. Eligibility, benefits, and duration are the same for both sickness and unemployment benefits. Eligibility is based on service in the previous year; generally, workers with at least 2.5 months of full-time railroad employment are eligible. The benefit is calculated as 60 percent of the worker's daily wage rate subject to a maximum amount. The maximum daily benefit for the benefit year July 1, 1992, to June 30, 1993, is $33, or $330 maximum biweekly. The maximum benefit is indexed to national wage levels and adjusted each year. Except for sickness benefits paid for on-the-job injuries, sickness and unemployment benefits are subject to federal taxes and a portion is subject to railroad retirement taxes. No sickness benefits are subject to state income taxes. A 2-week waiting period exists before an injured worker is eligible for sickness benefits. Normally these benefits extend for 26 weeks, but workers with 10 to 14 years of service are eligible for an additional 13 weeks of benefits. Those with 15 years or more of service are eligible for a full year of sickness benefits. An injured employee is not eligible for sickness benefits while receiv- ing any pay, including "wages, salary, pay for time lost, dismissal al- lowances," and so forth [RRB 1992, 112 (emphasis added)]. Therefore, railroad wage continuation programs (examined in the next section) preclude both the need and the eligibility for sickness benefits. Fur-

Overall Context for Compensation of Injured Workers • 31 thermore, the sickness benefits that an employee receives, at least those provided by the RRB, must be paid back if a settlement is received from the railroad that covers the lost wages. "The amount recoverable or withheld in such cases cannot exceed the net amount of the damages or settlement after medical and legal expenses have been deducted" (RRB 1992, 117). The sickness benefit replaces only a small portion of the average railroad worker's income, which for all Class I railroad employees in 1991 was about $42,000 a year (AAR 1992, 56) or about $800 a week. Although the sickness benefit is only $165 per week, it is not subject to taxes if the worker is injured on the job. Comparing the sickness benefit with weekly wages requires calculating the after-tax weekly income for the average worker. This in turn depends on the worker's marital status and number of dependents. Two cases can be used to illustrate the possible effective level of income replacement by the sickness benefit: a married worker with two children and a single worker with no depen- dents. For the married worker, the average total federal, state, and local tax rate (including railroad retirement taxes) would be about 25 per- cent, whereas the same tax rate for the single worker would be about 33 percent.' This implies an after-tax weekly income of $600 for the married couple and $540 for the single worker. The sickness benefit would therefore replace only about 27 percent of the married worker's and 31 percent of the single worker's weekly after-tax income. Short-Term Disability Benefits To supplement the RRB sickness benefits, some workers purchase short-term disability insurance either on their own accord or through programs sponsored by their union. These benefits are considered supplementary by the RRB and do not result in any reduction of the sickness benefits described above. However, the insurance policy is likely to require, just as the RRB does, that the benefits provided be repaid from any recovery the worker receives from the railroad (assum- ing that the worker nets at least full wage recovery in the settlement). For example, one disability insurance policy that was reported to the study committee costs $15/month. It provides a $500/month benefit, so the total of the sickness benefit and the disability benefit would be $1,160/month or $290/week. Using the previous example, the short- term disability coverage could raise weekly income to 48 percent of the married worker's and 54 percent of the single worker's after-tax in- come. Although this is a definite improvement, the injured worker's net income during the recovery period would still be well below the worker's net income while on the job.

32 • Compensating Injured Railroad Workers Under FELA Long-Term Disability Benefits Long-term disability benefits are provided to railroad workers by the RRB for on-the-job injuries without regard to negligence just as they are for non-work-related disability. These RRB long-term disability benefits are similar to social security disability benefits provided to workers covered by state workers' compensation systems. Disability retirement benefits are available to railroad workers only with serious long-term disabilities and are not a replacement for temporary wage losses. RRB disability has two features that are different from social secu- rity. First, whereas social security benefits are only available if one is disabled for all work (total disability), RRB provides for occupational disabilities (defined below) as well. Second, RRB benefits are higher than social security benefits because they include a benefit that substi- tutes for a private pension plan in the railroad industry. Eligibility for RRB disability benefits depends on the level of disabil- ity and the worker's age and years of service with the railroad industry. Total disability benefits are payable "at any age, if an employee is permanently disabled for all regular work and has at least 10 years (120 months) of creditable railroad service" [RRB 1992 (emphasis added)]. Occupational disability is payable "at age 60, if an employee has at least 10 years of railroad service or at any age if the employee has at least 20 years (240 months) of service, when the employee is permanently disabled for his or her regular railroad occupation" [RRB 1992 (emphasis added)].5 To qualify for occupational disability, the employee must have a current connection to the railroad industry, which is defined as having worked "at least 12 months in the 30 months immediately preceding the month his or her railroad retirement [disability] annuity begins" (RRB 1992, 29). There is a 5-month waiting period after the onset of the disability before an employee can receive annuity payments. The amount of disability retirement benefits depends on formulas that take into account factors such as years of service, earnings, and various tax bases. The calculation of the benefit is divided into two components, Tier I and Tier II. Tier I benefits are equivalent to those payable under social security but with railroad service and age require- ments. Tier II benefits are equivalent to a private pension supplement to social security. Because of the number of variables involved, it is difficult to devise general cases for purposes of illustration or compari- son. Nevertheless, the RRB indicates that for 1991, the average monthly award for an employee retiring because of disability was $1,250 and that the average years of service was 23.7 (RRB 1992, 33). This is not

Overall Context for Compensation of Injured Workers • 33 much higher than the amount for combined sickness and short-term disability described earlier. Furthermore, "the Tier I portion may, under certain circumstances, be reduced for receipt of workers' com- pensation or public disability benefits" (RRB 1992, 36). FELA-Related Benefits Benefits available to workers that are independent of FELA but more directly tied to the process of securing FELA recoveries are cash ad- vances and wage continuation. These benefits will be described only briefly in this section. They are more thoroughly explored in Chapter 4 because they are so closely coupled with the FELA process. Cash Advances Because sickness benefits and private insurance cover only about half the average employee's after-tax income at best, employees may be faced with a significant loss of income during periods of recovery from injury. A widely used approach to helping railroad employees maintain income has been through the use of advances by the railroad on the value of the potential settlement. The railroad employer voluntarily provides a portion of the estimated eventual settlement to the employee to help maintain income during the recovery stage. One large railroad indicated that it provides about 25 percent of the expected settlement. Advances are negotiated between the claims agent and the employee, and any future settlement is reduced by the amount of the advance. Advances, or additional advances after an initial one, may be condi- tional on the employees' cooperating in seeking medical and rehabilita- tive treatment. Wage Continuation Some railroads have instituted wage continuation programs for their employees. Wage continuation, as the name implies, is simply the continued payment of a worker's normal wages from the time of injury until the worker's return to the job. The purpose of wage continuation is to reduce the inherent conflict between labor and management that occurs under the FELA process after injuries occur. Wage continuation is a voluntary program of the railroads and can be altered or withdrawn at any time. One railroad, for example, initially offered wage continuation for up to 12 months. It subsequently de- cided that this lengthy benefit period was a disincentive for employees to return to work and that it did not dissuade enough lawsuits by

34 • Compensating Injured Railroad Workers Under FELA encouraging negotiations. Therefore, the railroad employer concluded that the program was too costly and reduced the eligible period for wage continuation to 3 months. Availability of Benefits All the medical, sickness, and disability benefits just discussed are available to the injured worker whether the worker recovers through FELA or not. However, if no recovery occurs because the claim is fraudulent, these benefits are likely to be denied as well. On the other hand, the employee's access to cash advances or wage continuation will depend on the likelihood of the employee's winning a FELA settlement. A railroad is unlikely to offer a cash advance for a claim that it does not believe it will have to pay. Wage continuation may be denied as well, though it is more variable. Even if the railroad does not foresee a FELA settlement because of clear-cut employee negli- gence, it may still provide wage continuation during the injury recovery period. WORKERS' COMPENSATION BENEFITS WITHIN THE OVERALL COMPENSATION CONTEXT As with FELA, some benefits available to injured workers covered by workers' compensation systems are outside these systems and may complement or substitute for workers' compensation benefits. Because these benefits vary between industries and individual employers, the overall package of benefits complementary to workers' compensation cannot be generalized as readily as the benefit package complementary to FELA, which applies to a single industry. Nevertheless, it is possible to identify the types of benefits that may be available, some of which have already been introduced in the description of the complementary and supplementary benefits of FELA. Medical Benefits Under workers' compensation systems, the medical costs of on-the-job injuries are included in the premiums for workers' compensation, and there is no parallel or dual coverage through the employer's regular health insurance program. This difference may be transparent to the employee, since claims may be submitted through the regular health insurance program that are then submitted by the insurer for workers' compensation.

Overall Context for Compensation of Injured Workers • 35 Sickness Benefits Many employers offer sick-leave or salary continuation programs through which the employees continue to receive their normal wages, up to some limit, while they are sick or recovering from an injury. The limit may be based on the number of sick-leave days accrued by the employee and unused or, in the case of salary continuation, there may be a limit to the total number of days per occurrence. SOCIAL SECURITY DISABILITY Disability benefits to individuals and families are provided under the Social Security Act and related legislation (Social Security Administration 1988) and administered by the Social Security Administration. Unlike the most familiar social security pro- grams, which provide for cash benefits after retirement, disability benefits can be provided before age 65 (or an earlier retirement age). Funding is provided through the payroll taxes paid by em- ployers, employees, and self-employed individuals who are cov- ered by social security (about 90 percent of workers). The main exceptions are workers in the railroad industry, federal civilian employees hired before 1984, some workers in state and local government, and certain domestic and agricultural workers. The eligibility standards for workers' disability benefits are quite stringent. They require that an individual be insured under social security, classified as completely disabled, less than 65 years old, and meet a 5-month waiting period. The disability need not result from injury on the job. Benefits are provided during the period of disability, but they end once the individual reaches age 65. After age 65, regular retirement benefits begin and disability benefits end. Disability benefits are equivalent to 80 percent of the disabled worker's average wages subject to the current social security maximums, but these benefits are reduced by coverage pro- vided by workers' compensation or other public programs. Med- icare medical benefits are available for those receiving social security disability. In cases where vocational rehabilitation is an option, individuals are referred to state vocational rehabilitation agencies, and as an incentive to return to work, benefits are pro- vided for up to 9 months after the return to work.

36 • Compensating Injured Railroad Workers Under FELA Short-Term and Long-Term Disability Benefits Short-term disability insurance may be available to employees through their employer. The costs may be borne by the employer, the employee, or both. For employees who are injured and completely disabled, workers' compensation benefits are supplemented by disability benefits through social security (see accompanying box on social security dis- ability). In addition, long-term disability insurance may be available through the employer with the costs borne by the employer, the em- ployee, or both. Long-term disability can also be purchased by an employee through an insurance company. Medical and Vocational Rehabilitation Workers' compensation systems often provide payments for medical and vocational rehabilitation, subject to eligibility requirements. Medi- cal rehabilitation costs not covered through workers' compensation may be covered through the employer's regular health insurance program. NOTES The railroad industry health insurance policy is industrywide rather than on a company-by-company basis. Work-related stress claims hold the potential for significant costs regardless of the injury compensation system. In the short run, for workers' compensa- tion systems, administrative and judicial bodies will determine whether mental stress claims are compensable; whether they are compensable under FELA is currently at issue before the Supreme Court. In the long run, the compensation for any particular occupational illness is ultimately controlled by legislatures at the state or federal level regardless of the injury compensa- tion system in place. That is, mental stress claims are not a problem because of the inherent approach of any particular system. Stress claims are difficult to resolve irrespective of the venue, whether administrative or judicial. The following description of the sickness benefits is largely taken from the Railroad Retirement Handbook 1992 published by the Railroad Retirement Board (RRB 1992). For a married couple with two children and an income of $42,000 who take the standard deduction, the 1991 average federal tax rate would be about 10 percent. Assuming a state and local tax rate of 4 percent, the average tax rate with the same deductions would be about 2.5 percent. The railroad retirement tax rate of 12.55 percent would apply to the entire amount and so the average rate would be 12.55 percent. The total average tax rate for this worker would therefore be about 25 percent. Doing the same calcula- tions for a single worker earning $42,000 gives average tax rates of

Overall Context for Compensation of Injured Workers • 37 17.5 percent, 3.4 percent, and 12.55 percent, respectively, with a total average rate of about 33 percent. 5. The employee is eligible for occupational disability even if the employee can do other work. REFERENCES ABBREVIATIONS AAR Association of American Railroads RRB Railroad Retirement Board AAR. 1992. Railroad Facts, 1992 ed. Washington, D.C., Sept. RRB. 1992. Railroad Retirement Handbook 1992. Chicago, Ill. Social Security Administration. 1988. Social Security Handbook. U.S. Depart- ment of Health and Human Services.

3 Criteria for Analysis of Compensation of Injured Workers The purpose of this chapter is to present the assessment criteria used to guide the study committee's examination of the compensation process under the Federal Employers' Liability Act of 1908 (FELA) and its comparison with workers' compensation systems (as reported in Chap- ters 4 through 6). These criteria draw from models of the disabling process that have been developed by researchers on health care and rehabilitation, an examination of goals for injury compensation sys- tems from the perspectives of the principal affected parties, and the legal concepts of liability that underlie specific injury compensation systems. DISABILITY AND THE BASIS FOR COMPENSATION Workplace injury is the starting point that brings into motion the compensation systems for injured workers. It also provides the focus for society's efforts to reduce accidents and the negative effects they have on individuals, families, businesses, and society as a whole. Injury on the job is for the most part a temporary event that passes quickly, with a short time lost to recuperation. The acute effects of some injuries, however, may require a lengthy recovery and involve lingering effects of the injury after recovery.' The worker's ability to return to his

Criteria for Analysis of Compensation of Injured Workers • 39 or her old job or to work at all may be affected. Any injury compensa- tion system must address both temporary and permanent effects of injury. In the first part of this section, the disabling process is examined from the injury to disability. Next, these concepts are applied to injury compensation systems, and last, their implication for the design and evaluation of those systems is explored. IOM Model of Disability Various models have been developed by researchers on health care and rehabilitation to analyze disability and its consequences. For example, a committee of the Institute of Medicine (IOM) developed a model for use in its study Disability in America: Toward a National Agenda for Pre- vention (IOM 1991). The model provides a useful perspective for exam- ining injury and disability with regard to FELA and other injury compensation systems. It was designed to put disability in context, showing it as the potential outcome of a process that begins with injury or illness and that may or may not progress to disability. The IOM model is addressed to all disabilities and their causes, not just those related to on-the-job injury. The IOM model was chosen for its emphasis on process and the relational definition of disability. That is, disability is not an absolute condition, but arises only in context with social, environmental, or other demands placed on the individual and that may be peculiar to the individual. This is particularly appropriate for work disability in which work environments and worker behavior are important factors in both accident and disability prevention. The IOM model is itself an outgrowth of two models of disability: one by Nagi that is based on functional limitation (Nagi 1965) and one developed by the World Health Organization (WHO) called the Inter- national Classification of Impairments, Disabilities, and Handicaps (WHO 1980). The IOM model adopts the Nagi approach to functional limitation and incorporates a four-step progression from pathology to impairment to functional limitation and finally to disability. Each of these steps is considered in the following discussion, which draws heavily from Chapter 3 of the IOM report (IOM 1991,76-108). Figure 3-1 shows the IOM model and is reproduced from that report (IOM 1991, 85).2 Pathology Pathology involves the actual physical changes in the body at the cellu- lar or tissue level, whether caused by trauma, disease, toxins, infections,

40 Compensating Injured Railroad Workers Under FELA FIGURE 3-1 IOM model of disability (IOM 1991). congenital conditions, or other problems. These changes to the body may be temporary and accompanied by temporary impairment or temporary with some persistent impairment, or both the pathology and the impairment may persist for an indefinite period. Impairment Impairment involves the effect of the pathology on the organ or organ system but not on the entire organism and is defined as "a discrete loss or abnormality of mental, physical, or biochemical function.... Exam- ples of impairments are the absence or displacement of body parts, reduced blood flow, mechanical problems ofjoints, paralysis, stiffness, and numbness" (IOM 1991, 80). All pathologies are accompanied by some level of impairment, and specific impairments may stem from many different pathologies. Functional Limitation The term functional limitation comes from the work of Nagi and as proposed by him describes

Criteria for Analysis of Compensation of Injured Workers • 41 effects manifested in the performance or performance capacity of the person as a whole. An example of a functional limitation is the inability to lift a 25-pound box and carry it 25 feet. This type of limitation may be caused by impairment of any one of several body systems, including reduction of pulmonary function (emphysema), denervation of mus- cle tissue (amyotrophic lateral sclerosis), or restriction in range ofjoint motion. (IOM 1991, 80) Functional limitations always arise from some impairment, but an impairment does not necessarily result in a functional limitation. Of two individuals with the same impairment, one may be capable of an action such as dressing, whereas another may not. The difference depends on the state of the individual apart from the impairment, including physical condition and stamina or tolerance for pain. Disability "Disability is the expression of a physical or mental limitation in a social context—the gap between a person's capabilities and the demands of the environment" (IOM 1991, 81). Disability arises from the interaction of functional limitations with the expectations and requirements of the various social, work, and familial roles of the injured individual. As a result, disability is not inherent in the functional limitation, but is an expression of its effects on the individual's ability to carry out normal activities. As expressed in the IOM report: Thus one way in which disability differs from pathology, impairment, and functional limitation is in the role of factors external to the individ- ual. Pathology, impairment, and functional limitation can be deter- mined by examination and testing of the individual, but disability is a relational attribute—the interaction of an individual's functional lim- itation with the demands of expected tasks and roles and with the environmental conditions under which roles and tasks are to be per- formed. Referring to specific pathologies or impairments as disabil- ities ignores the interactive nature of the process that can lead to disability. (IOM 1991, 82) Risk Factors The disabling process is affected by factors peculiar to the worker and to the worker's environment that increase or decrease the risk of an injury's leading to a disability. These factors may act to accentuate the process or mitigate it; that is, they may increase the risk that a particular

42 • Compensating Injured Railroad Workers Under FELA pathology will lead to an impairment, a functional limitation, or a disability or decrease that risk, independent of the severity of the injury itself. The IOM model shows these factors—environment, biology, and behavior—interacting with each step in the disabling process. Of par- ticular importance in workplace injury are environment and behavior. Environment The work environment affects the level of risk that injuries will occur. Further, both workplace and nonworkplace environments affect the reaction of individuals to injuries and their associated impairments. Whether impairments lead to functional limitations and disabilities is also influenced by both these environments. Injury prevention efforts are aimed at changing the workplace environment to reduce the likeli- hood and consequence of on-the-job injury. Behavior and Biology Worker behavior may be a key influence on the safety of the workplace and the aftermath of injury. Worker behavior and physical condition (biology) can produce a wide range of individual responses to the same pathology and impairment. How the individual reacts to injury helps to determine the progression of an injury in the disabling process. Disability Model and Injury Compensation Systems Injury compensation systems can be designed to ameliorate the prob- lems caused by injury at each point on the disability model pathway. Rapid medical intervention and rehabilitation can prevent temporary impairments from leading to permanent impairments and from result- ing in functional limitations. Vocational rehabilitation may prevent a functional limitation from so disabling an injured employee that he or she is unable to return to work or to any employment. Pathology that is serious enough to lead to temporary impairment and short-term func- tional limitation can generate the need for both medical benefits and wage compensation for the period of medical convalescence. By fund- ing full medical treatment and medical rehabilitation and compensat- ing wage loss during the recovery period, injury compensation systems can reduce the period of impairment and its severity and reduce the risk that functional limitation will persist after medical recovery. In addition, the systems can be structured to provide incentives for em- ployers to improve the safety of the work environment and for em- ployees to engage in safety-conscious behavior.

Criteria for Analysis of Compensation of Injured Workers • 43 The role of behavior is particularly important in a fault-based system such as FELA in which negligence by the employee or the employer can affect the level of injury compensation. To the extent that workers engage in unsafe behavior, they may be found to be partially responsi- ble for their subsequent injury and therefore entitled to a lesser benefit from their employer. Theoretically, an effective standard for assessing negligence under such a system would discourage risky behavior on the part of the worker. Similarly, it would provide an incentive to the employer to reduce dangerous operating environments and avoid liability. Impairments and their associated functional limitations that remain after maximum medical improvement may result in a disabling condi- tion that hampers the worker's productivity and may require perma- nent wage supplementation. If the functional limitations are severe and persistent, their interaction with the skills required for the job may disable the worker for his or her prior occupation. If the limitations are severe enough, they may disable the worker for any occupation. Fur- thermore, work-related injuries may result in functional limitations that do not lead to workplace disabilities but may result in nonwork- place disabilities. Bases for Permanent Disability Compensation An important distinction must be made in an injury compensation system between temporary and permanent effects. Berkowitz and Bur- ton examined the underlying rationale for compensating permanent disabilities in workers' compensation systems and used a model similar to the IOM model.3 As used by Berkowitz and Burton, temporary is defined as "the period from the onset of injury or disease until maximum medical improve- ment (MM!) has been achieved" (Berkowitz and Burton 1987, 5). They define permanent as the period following maximum medical improve- ment. Both terms are applicable to the description of impairments, functional limitations, and work and nonwork disabilities. For permanent injuries, workers' compensation systems generally apply one of three approaches for compensating ongoing losses. One is based on the physical impairment or actual damage caused to the body from the injury and its associated functional limitations. The second approach uses impairment or functional limitation as a basis for deter- mining the loss of wage-earning capacity. A third approach bases compensation on actual wage loss over time resulting from the injury or illness.

44 • Compensating Injured Railroad Workers Under FELA Impairment evaluations are critical in the first two approaches. These evaluations attempt to measure the remaining impairment to the body after maximum medical improvement is attained. For exam- ple, the loss of a hand through amputation represents an ongoing impairment after the remainder of the arm has healed. Impairment evaluation is generally based on the effect of the loss on the individual as a whole, that is, what percentage of the whole person is affected by the loss. For example, according to the American Medical Association Guides to the Evaluation ofPermanent Impairment (AMA 1990), the loss of a hand is considered 54 percent impairment of the whole person. The hand is considered to be 90 percent of the arm and the arm to be 60 percent of the whole person; 90 percent of 60 percent is 54 per- cent. The AMA Guides go beyond the simple listing of injury to the ex- tremities, sight, and hearing and cover all the organs and systems of the body. Under the first compensation approach, which is based strictly on physical impairment, some state workers' compensation systems set up a limited schedule, often confined to impairments to the extremities, sight, and hearing, that lists specific impairments and their compensa- tion. Often this schedule is not concerned with whether a functional limitation or disability is associated with the injury. The second ap- proach requires more subjective judgments about how impairments and the resulting functional limitations will affect wage-earning capacity.4 Injury compensation for railroad workers does not depend on mak- ing these distinctions and establishing eligibility for particular sched- ules or levels of compensation as most workers' compensation systems do. Certainly the establishment of the underlying impairment and functional limitation conditions is important in assessing the potential settlement level because of their effects on potential wage loss, future medical or vocational rehabilitation expenses, nonworkplace activities, and compensation for pain and suffering. GOALS FOR COMPENSATION SYSTEMS Which goals of injury compensation systems should be emphasized and what trade-offs should be made? The answers to these questions will depend on the priorities of the parties involved. Four parties are involved with compensation for on-the-job injuries: employees (and their families), employers, agents, and society or the public at large. Agents include attorneys, state government agencies, and private in- surance companies, all of whom play various roles in injury compensa- tion systems.

Criteria for Analysis of Compensation of Injured Workers • 45 Employees' Perspective In the broadest sense, workers are concerned with the net benefits that they receive from an injury compensation system. That is, they would like to maximize the medical, wage, and rehabilitation benefits they receive while minimizing the costs they bear for securing those benefits. These costs include direct costs such as outlays for legal assistance and indirect costs from delays in receiving benefits and from the time required to prepare and pursue their claims. Specific workers' benefit goals include (a) medical coverage and rehabilitation, (b) compensation for lost wages, (c) compensation for pain and suffering and other nonmonetary losses, (d) compensation for survivors and dependents, and (e) vocational rehabilitation. Medical Coverage and Rehabilitation From an employee's standpoint, an important goal of an injury com- pensation system is to ensure medical coverage and treatment for the injury itself. Aspects of medical coverage include access to adequate emergency and short-term health care and to medical rehabilitation or long-term care and eventually full recovery from the injury. Access has both a time and a financial element. The injured em- ployee would like ample time for full recovery from injury before returning to work and compensation for the full range of medical care required. In order to have access, the employee may .require both continual coverage of the costs of the medical care and wage replace- ment during the recovery and rehabilitation period. Typically, employer-sponsored health plans require workers to pay a portion of their medical costs for non-work-related injuries. A compen- sation system for injured workers could offer similar provisions or provide full coverage. Regardless of who pays, the goal of medical coverage is full treatment and rehabilitation of the injured worker. Compensation for Lost Wages Another primary goal of injury compensation systems from the em- ployee's standpoint is the coverage of wages for work days lost because of the injury. For the majority of injuries that are temporary and short in duration, this is likely to be the largest out-of-pocket expense. Whether covered by a sick-leave policy, an injury compensation program, or some general disability insurance policy, the worker would like to ensure that the wage loss sustained during recovery is replaced. For injuries that are not temporary, but involve a lingering impair- ment, wage compensation may take various forms and the need for

46 • Compensating Injured Railroad Workers Under FELA wage compensation may vary. As discussed earlier, permanent disabil- ities may partially limit the worker's wage-earning capability or totally eliminate it. Workers would like these benefits to be immediately available with as little administrative delay as possible and at no cost to them. However, because of the difficulty in assessing permanent partial disabilities and the subjective nature of calculating potential wage loss or loss of wage- earning capacity, replacement of lost wages may be delayed. Compensation for Pain and Suffering and Other Nonmonetary Losses Worker injuries may inflict more than just out-of-pocket losses from lost wages and medical expenses. The suffering that accompanies the injury and the effect it has on the worker's life may be significant and may, from the employee's standpoint, exceed the monetary losses. In addition to this loss from pain and suffering, the employee may experience lost nonmarket productivity. Just as an injury may render a worker less productive in the market sector, leading to lower wages, it may make the worker less productive in the nonmarket sector (e.g., unable to perform household chores). This loss is distinct from pain and suffering since an individual may be in pain but still able to perform tasks. Compensation for Survivors and Dependents Lost wages and pain and suffering also affect the worker's dependents. A worker would like the compensation system to have the flexibility to take account of the number of individuals that depend on his or her income. In the case of injuries that lead to death, a worker would like compensation for surviving dependents. Vocational Rehabilitation The injured employee may also need or expect vocational rehabilita- tion as part of the compensation for an on-the-job injury. When an injured worker remains disabled for his or her normal occupation after maximum medical improvement, vocational rehabilitation may be able to return the worker to some productive employment. This may be accomplished by two approaches. One is to attack the interaction be- tween the worker's functional limitations and the job skill requirements that is leading to the occupational disability in the current occupation. This may involve improving some functional skills that might substitute

Criteria for Analysis of Compensation of Injured Workers • 47 for those lost or revising job procedures to match the worker's capa- bilities that are unaffected by the functional limitation. The second approach is to retrain the worker for alternative employ- ment. This involves, as above, improving functions that can substitute for those limited by continuing impairment. It also may involve learn- ing new skills and occupations that do not require the functions that have been limited by impairment. Employers' Perspective As might be expected, employers would like to minimize the overall costs they bear from worker injuries and the resulting compensation; this includes the costs of accident and injury prevention, direct pay- ments to workers, transactions costs, and other business costs associated with accidents. Employers would like to minimize the additional cost they bear for direct payments to workers (for medical, wage, rehabilita- tion, and other benefits) and for the administrative cost of running their injury compensation systems. Administrative costs include those for claims processing, investigation, and benefit monitoring, and the costs of the appeals process and litigation for disputed claims including those for claims staff, in-house attorneys, and outside counsel. In addition, employers would like their costs to be as predictable as possible. From the employer's perspective, minimizing this overall cost or loss from injury is likely to be the primary goal. Cost reduction may take various forms, including reducing accidents to avoid injuries alto- gether; reducing the types, extent, and level of benefit coverage; shar- ing the cost of coverage with the employee; and reducing the incentives for disputes and administrative delays in the system. Certainly the goals of many employers include the desire to protect their employees from harm or to provide generous benefits for injured workers. However, most employers, as a result of perceived business needs, probably incur only those costs that are necessary to meet legal requirements, retain workers, and maintain a productive level of employee morale. Full medical treatment and employee cooperation with medical re- habilitation may be goals of the employer as well. Full and rapid recov- ery may limit the cost of short-term wage compensation, the demands for compensation for pain and suffering, and, by avoiding any perma- nent disability, long-term wage supplementation or replacement. Agents' Perspective State agencies monitor injury compensation to ensure that the interests of various groups are protected, with a likely emphasis on promoting

48 • Compensating Injured Railroad Workers Under FELA the public interest. Furthermore, many states offer competitive or exclusive state funds to insure against workplace injuries. In this role, states have an interest similar to private insurance firms and employers in ensuring the validity of claims and in minimizing the costs of provid- ing benefits that are deserved. Insurance companies' goals may match those of employers in some areas, but they may also be at odds with them in others. Insurance companies would like to minimize the costs of benefits paid relative to the premium income they charge and to minimize claims through the promotion of safer operations. Experience rating of firms for insur- ance shifts some of the safety incentive to the firm as well as incentives to reduce fraud and the costs of providing benefits. Other stakeholders include attorneys involved in injury compensa- tion. They would like to protect their clients' rights to review of their cases and a fair adjudication of them. Lawyers for both plaintiff and defense also have a financial interest in protecting their role in the system and the income that it generates for them. System trade-offs that reduce the variability of benefits and define fixed and invariable benefits may be viewed by attorneys as incompatible with both of these goals. Public Interest Perspective Public interest goals are less specific than those of the private interests represented by employers, employees, and their agents. Broad goals from society's view may be defined as achieving equity and efficiency in compensating injured workers. The goal of efficiency is perhaps best expressed as minimizing the overall cost to society from workplace injuries. This may include efforts to avoid all injuries when it is cost-effective to do so, to minimize the cost of compensating injured workers for any given level of benefits deliv- ered, and to return workers to productive employment as quickly as possible. In other words, the goals include achieving optimum levels of safety, transactions costs, and rehabilitation of injured workers. From the disability model discussed in the previous section (Figure 3-1), minimizing the cost of an injury involves preventing injuries from progressing along the pathology-to-disability pathway, that is, prevent- ing a pathology from becoming an impairment, an impairment from becoming a functional limitation, and a functional limitation from becoming a disability. This leftward restriction is efficient only to the point at which the cost of additional actions needed to reduce disabil- ities, functional limitations, or impairments just equals the benefits of those actions. In making such assessments, the relevant costs include

Criteria for Analysis of Compensation of Injured Workers • 49 not only actual monetary or out-of-pocket costs, but also the loss from the foregone output and societal contributions of disabled workers and such subjective costs as the pain and suffering of injured individuals and the losses to their families. Trade-Offs Among Goals Various types of trade-offs are likely to be made in the development of an injury compensation system. Elements that might be varied include (a) what is compensated—the types of coverage (e.g., medical rehabili- tation, vocational rehabilitation, compensation for wages, or compensa- tion for pain and suffering) and the levels, the length, and the breadth of coverage; (b) who is compensated—all injured, only those injured by employer negligence, and so on; and (c) how workers are compen- sated—similar compensation for similar injuries, the same proportion of income lost awarded to all, different types of compensation mecha- nisms and administration, and so forth. In designing a compensation system, choices must be made among the goals for delivered services, allowable levels of transactions cost, and the incentives that the agents wish to embody in the system. Berkowitz and Burton point Out that there is a direct relationship between the level of benefits provided and the administrative costs of providing them. Therefore, some trade-off must be made in maximizing benefits to control the transactions cost. Berkowitz and Burton apply the term "myopic efficiency" to efforts that simply reduce transactions cost with- Out regard for its effect on the benefits delivered, in other words, that take the trade-off too far, giving the impression of efficiency while not delivering the services needed by injured workers (Berkowitz and Burton 1987, 26-28). LEGAL CONCEPTS OF INJURY COMPENSATION The basic choice in designing an injury compensation system is that between a tort (fault-based) system and a no-fault system. With the tort approach, only injuries that are due to the wrongful actions of another are actionable and permit recovery for damages, and there is no as- sumption that all injured workers should be compensated. A no-fault system, on the other hand, provides compensation regardless of who might have been negligent. Tort and no-fault systems represent different concepts of liability. These concepts and their implications for injury compensation systems are first reviewed; their reflection in FELA and workers' compensation systems follows.

50 • Compensating Injured Railroad Workers Under FELA Concepts of Liability Who is liable when a worker is injured? That is, who bears the respon- sibility for making the worker whole? Two alternative standards of liability are strict liability and negligence liability.-' In assessing the outcome when there is negligence on the part of both the employer and employee, alternative approaches are traditional contributory negli- gence and comparative negligence. Strict Liability Versus Negligence Liability A strict liability standard imposes liability on the employer whether or not the employer is at fault (negligent) for a worker's injury. Liability arises from the fact that the employee's injury is caused by the employ- ment. Strict liability could also be imposed on the employee, which would be the equivalent of no liability for the employer. A negligence liability standard, on the other hand, requires fault on the part of the employer before liability accrues. Under its conventional definition, negligence is conduct whose expected risk in terms of both the probability and severity of harm exceeds its benefits. A defendant's negligence is generally assessed by considering what precautions the defendant could have adopted; the failure to incorporate a precaution is negligent if the risk-reducing advantages of the precaution exceed the precaution's cost. From a deterrence standpoint, the goal of a negligence standard is to induce defendants to adopt all cost-justified precautions. Both equity and economic efficiency arguments are put forth in favor of a negligence standard. From an equity standpoint, intuitive ideas of fairness support the notion that there is justice in requiring a negligent defendant to compensate the victim of its negligence. Con- versely, where there is no negligence on the part of any defendant in a victim's injury, a negligence standard would give the victim no compen- sation and neither the victim nor any defendant would be considered to have been treated unfairly. From an economic perspective, a negligence standard requires that the defendant adopt only those precautions whose costs do not exceed the expected benefits from those precautions. A defendant's conduct can be risky yet not negligent if the costs of risk prevention exceed the benefits. The application of a negligence standard gives the defendant incen- tives to determine which precautions are available to avoid future liability and to adopt all precautions whose costs do not exceed the benefits. From an efficiency standpoint, there is no obvious deterrence

Criteria for Analysis of Compensation of Injured Workers • 51 advantage in imposing liability if the defendant is not negligent and has no cost-beneficial precautions available. Under a strict liability regime, the primary economic incentive for adopting precautions is likely to be the same as under a negligence liability standard. Consider the defendant whose conduct creates an expected risk of $100 when the only available safety precaution costs $140. Under a negligence liability standard, this defendant is not negligent and hence would not be held liable for the failure to adopt this precaution. Even under strict liability, the defendant would proba- bly not adopt this precaution. It is cheaper for the defendant to bear the expected liability of $100 than to accept the safety-precaution burden of $140. Under strict liability, however, the victim would receive compensation. Strict liability standards may have negative implications for the care- taking behavior of both employers and employees. Strict liability on the employee would reduce the employer's incentive to adopt safety pre- cautions, both in investments and operating practices. Strict liability on the employer may reduce the care-taking of employees in their work operations, especially if liability is imposed for both out-of-pocket and nonmonetary losses. Other economic arguments favor strict liability. First, the overhead from litigation is reduced through the removal of the contentious issue of negligence, though this reduction is offset somewhat by an increase in the number of claims litigated. Second, strict liability may induce more efficient safety investment behavior. Courts and juries may have difficulty in accurately determin- ing what cost-effective precautions were available to the employer. In that case and if employers are aware of it, strict liability may provide a greater incentive to adopt all cost-justified precautions than a negli- gence liability standard. Facing a rule of negligence, the employer can expect that failure to adopt the precautions will not result in liability. Facing a strict liability standard and knowing that liability will accrue regardless of what the plaintiff may prove, the employer may decide that the precautions are well worth adopting. A third economic argument, which is similar to the second, is that a rule of strict liability gives employers the incentive to invest in safety research and development (R&D). Employers are more apt to know what cost-justified R&D is feasible in their business, and they have the incentive to pursue it. Employers' obligation to engage in reasonable- cost R&D is very difficult to enforce in negligence actions. Finally, strict liability involves loss distribution. Absent first-party insurance, the employee can be financially devastated by a serious accident. Strict liability shifts the loss to the employer, who is likely

52 • Compensating Injured Railroad Workers Under FELA better able to purchase insurance than the individual. Depending on assumptions about the incidence of the loss, the employer may redis- tribute it among customers, shareholders, or all employees. Comparative Negligence Versus Contributory Negligence Under a strict liability regime, the worker need only show on-the-job injury to recover compensation. This is the case for most no-fault workers' compensation systems. Negligence liability, on the other hand, requires the demonstration of negligence on the part of the employer to give the employee a cause of action. Furthermore, the outcome may depend on whether the employee is negligent as well. Two approaches to determining the degree of liability are compara- tive negligence and contributory negligence. Under a contributory negligence standard, the employer may escape liability 'if it can show that the accident would not have occurred but for the negligence of the employee. A comparative negligence standard allocates liability among the parties involved on the basis of their degree of negligence. In this instance, if the plaintiff bore part of the responsibility, compensation would be in proportion to negligence. Comparative negligence is often defended on efficiency grounds because it provides incentives for both defendants and plaintiffs to adopt all cost-justified precautions given the potential for each bearing a portion of any loss due to their own negligence. Whether comparative negligence is in fact effective in influencing plaintiffs' behavior is disputed by some. Nevertheless, it is also defended on equity grounds. If the plaintiff would not have suffered the injury in the absence of his or her own conduct, it would be unfair to expect the defendant to pay full compensation for the injury. Workers' Compensation Approach Workers' compensation systems generally employ a strict liability stan- dard that incorporates limits to liability. This approach is often charac- terized as a bargain struck between employers and employees. In accepting strict liability, employers give up freedom from liability for nonnegligent injuries but are protected from paying full common-law damages for those injuries that result from their own negligence. The reverse is true for the injured employee, who gives up the right to full common-law damages in return for the benefits of strict liability. The question remains why this particularbargáin might make sense. One explanation is that injuries in the aggregate are usually caused by some combination of employee and employer negligence. Workers'

Criteria for Analysis of Compen.sation of Injured Workers • 53 compensation programs acknowledge this and recognize the appro- priateness of dividing liability. Rather than allocate and divide negli- gence on a.case-by-case basis, workers' compensation divides liability at a more general level. The employer bears strict liability for most out-of- pocket costs of the injury, whereas the employee bears strict liability for his or her own pain and suffering and other nonmonetary losses. This arrangement economizes on litigation, gives both employers and em- ployees significant incentive to avoid unduly risky conduct, and serves to protect the basic compensation needs of the injured employee for out-of-pocket losses. FELA Approach The FELA process incorporates the negligence approach instead of the strict liability approach used in workers' compensation. FELA, however, departs from a conventional negligence system by diluting the requirement that the plaintiff demonstrate that the employer's negligence was the cause of his or her injury.6 The current FELA approach has some support in early English common law. Weaver versus Ward, a 1606 English case, states that if a defendant "directly" causes a plaintiff's injury, the defendant should bear liability unless he or she can show that the accident was "inevitable" and that it happened "utterly without its own fault." Although the plaintiff in a FELA action is in no way required to show "direct" causation, the FELA approach to the issue of negligence is similar to this English precedent. ASSESSMENT CRITERIA FOR INJURY COMPENSATION SYSTEMS The preceding sections have focused on medical and legal concepts that have shaped, or helped explain, injury compensation systems, particularly the FELA process and workers' compensation. In addition, a review was given of the goals for these systems from the perspective of the principal parties involved. Assessment criteria used to guide the study committee's examination of the FELA process and the compari- son of FELA with workers' compensation systems are presented in this section. The criteria more or less follow the interests of the involved parties: for the worker, the key criterion is the benefits delivered; for the employer, it is the cost of injuries and compensation; and for the public, it is the prevention of loss of societal output from accidents and injuries and the development of incentives for equity and efficiency. The overall assessment of an injury compensation system turns on the concepts of equity and efficiency. Whether a system is equitable or

54 Compensating Injured Railroad Workers Under FELA not depends on the perspectives of the parties involved. The concept of efficiency, however, is much more uniformly accepted. An efficient system is one that tends to minimize transactions costs for given levels of benefits and provides cost-effective incentives. Assessment of Benefits Delivered The employee is concerned with the net benefits delivered. In assessing how well a compensation system meets the needs of workers and is equitable to them, important factors are the types of coverage, the extent of coverage, the level of coverage, and the speed of daim resolution. Types of Coverage The more types of coverage that are available, presumably the better it is from the worker's perspective. FELA and workers' compensation systems differ in types of coverage they offer, and among workers' compensation systems, the coverage varies. Types of coverage include medical treatment and rehabilitation, compensation for lost wages and wage-earning capacity, compensation for pain and suffering, compen- sation for survivors and dependents, and vocational rehabilitation (see Table 2-1). Extent of Coverage Similarly, the greater the extent of coverage within each type of benefit listed, the more beneficial it is to the employee. Extent of coverage governs whether all losses within the type are compensable or whether there are restrictions. For example, does occupational illness cover on- the-job heart attacks that may be related to job stress? Does pain and suffering include compensation for nonwork disabilities that result from on-the-job injuries? Level of Coverage Of obvious importance is the actual level of the benefit. For example, wage replacement may include all wages lost or there may be maxi- mums or duration limits that leave the worker with less than complete compensation. Speed of Resolution Particularly from the employee's standpoint, the speed of resolving the status of a claim and the start of benefit payments are important. Long

Criteria for Analysis of Compensation of Injured Workers • 55 delays in establishing the right to or the amount of a benefit may force the worker to borrow or use savings to cover lost income and other injury expenses. Assessment of Costs Direct and Indirect Costs The costs of providing workers' compensation can be divided between the direct cost of the benefits provided and the indirect cost of adminis- tering the system. The direct cost of the system matches the type and extent of benefits, as discussed in the previous section. The more generous the benefits received by workers, the more costly the pro- gram is to the employer for any given level of injury.7 Therefore, there is an inherent tension between the workers' goal of enhanced benefits and the employers' interest in minimizing cost.8 To the extent that employers "win" by succeeding in reducing direct benefits, workers "lose." Indirect costs accrue from all the transactions required to supply the benefits to injured workers. These transactions costs include the ad- ministrative costs of processing and investigating claims, litigation costs for appealing and litigating disputed (or controverted) claims, financial costs from paying fraudulent claims, and the employee-relations cost from a contentious system. Administrative costs involve the staff to assist employees in filing claims, to investigate the merits of each claim, to determine the benefits to which a claimant is entitled, and to monitor the case during medical and vocational rehabilitation. Administrative costs may be borne by the employer directly or included in the premiums paid for workers' com- pensation coverage when these services are provided by insurance carriers or state systems. As used here, litigation costs include all appeals, reviews, and court actions associated with denied claims. All systems have either adminis- trative or judicial rights of appeal (or both) for claims that are denied. Early steps in the appeal process may be handled directly between claims staff and the employee. After the initial steps, however, both the employee and employer are likely to seek legal counsel. Trade-Offs Among Costs There are trade-offs between and among direct and indirect costs. For example, if the benefits provided to injured workers are kept at very

56 • Compensating Injured Railroad Workers Under FELA low levels, the likelihood of fraudulent claims will be low and workers are likely to return to work as quickly as possible to avoid financial losses. In this instance, the cost of investigating, processing, and litigating claims may be kept relatively low through simple forms and pro- cedures, near automatic approval of claims, and standard, if low, benefit payments. Although this approach may be efficient from a transactions-cost standpoint, it may be inequitable for employees with serious injuries that result in a lengthy absence from work. Therefore, transactions costs must be considered in context with the benefits provided. Assessment of Incentives Ideally injury compensation systems should provide incentives for maintaining a safe work environment, minimizing transactions costs, and promoting rehabilitation to prevent impairments and functional limitations from becoming workplace disabilities. Because transactions costs have been examined in the preceding section, only rehabilitation and safety incentives are discussed here. Rehabilitation and Return to Work Injury compensation systems may provide incentives for rehabilitating and returning the injured employee to productive employment, whether in the employee's original job or in some alternative employment. Incentives for return to work may also be crafted to inhibit malingering and induce workers to return to work at the earliest time they are able. Less than full compensation for monetary losses is one method that has been used to provide this early-return-to-work incentive. Once there has been maximum medical improvement, the employee may not be capable of returning to his or her original job without some vocational training. In some instances, vocational rehabilitation may permit the employee to learn the original job in a way that compensates for the lingering functional loss, thereby preventing the loss from becoming a disability. Society and the individual would both benefit by avoiding the loss of the employee's output and the cost of the social services the employee might have required if the functional loss had remained or evolved into a disability. A system that reduces the speed of return to work may reinforce the idea of disability in the mind of the worker, which could lead to secondary disabilities, or "the additional disabilities that are sometimes caused by another disability" (IOM 1991, 332).

Criteria for Analysis of Compensation of Injured Workers • 57 Safety Safety incentives promote the societal goals of avoiding losses both to individuals and to the output of goods and services. All costs associated with an injury can be avoided if the injury itself is avoided. Avoiding injuries, however, may impose investment or operational costs on em- ployers and employees. Those safety improvements that reduce system net costs should be encouraged by the system adopted, whereas those that are not justified should be foregone. Safety incentives depend on the level and incidence of the cost from the injury. Other things being equal, as costs to the employer for worker benefits and transactions increase, the rewards to the employer from adopting safety precautions also increase. This is not to imply, however, that very expensive systems are the most efficient from an overall societal viewpoint. In terms of the incidence of cost, employers and employees theo- retically incur costs to increase safety up to the point at which the savings produced are greater than or equal to the costs incurred. If, however, one party is successful in shifting the burden of payment regardless of negligence, this safety incentive will be reduced. The cost burden may also be passed to third parties if experience rating is not used for devising injury compensation insurance pre- miums. Because railroads self-insure under FELA and would likely self-insure under some alternative workers' compensation system, they would fully sustain all compensation liabilities in both cases and there- fore experience the full safety incentives as well. NOTES Chapters 3, 4, and 5 present data showing the distribution between tempo- rary and permanent effects of injury. For a comparative analysis of disability models, see Chapter 3 of the IOM study (IOM 1991). This section is based largely on the work of Berkowitz and Burton (1987, Chapters 1 and 2). The underlying rationale for schedules may be to compensate the individ- ual for the expected, average future wage loss implied by the scheduled injury. Other principles of compensation include social insurance, such as social security disability insurance and Medicare, and assistance to the poor through social transfer programs such as Medicaid. Another approach is no employer liability at all, that is, injured workers bear the full cost of their own injuries (Rolph 1992, 2012-2014). This reduced showing requirement is likely a result of somejudicial dissat- isfaction with FELA's basic approach. For example, in Reed v. Philadelphia,

58 • Compensating Injured Railroad Workers Under FELA Bethlehem & New England R., 939 F 2d 128 (3 Cir. 1991), Judge Weis, with references to statements by Justices Douglas and Frankfurter, complained that "recompense for industrial injuries such as those suffered by plaintiff here should not have to depend on the vagaries of a statute such as the Safety Appliance Act or FELA or on the whims of juries or the delays of litigation." Furthermore, a number of studies have indicated that there may be strong incentives to make fraudulent claims and strong disincentives to return to work if the benefits provided approach or exceed full replacement of the worker's wages. This moral hazard may result in a greater than one-to-one increase in costs as benefits increase because the number of reported inju- ries and their duration increase. Employers' efforts to reduce or minimize costs should not be taken to mean simply reducing benefits and hurting workers. Employers are well aware of the overall costs of accidents and injuries in their businesses including the suffering of their employees, the effects on morale, and the cost of goods and machinery damaged. Employers will minimize their costs by improving safety and limiting injuries, by making efforts to reduce the cost of adminis- tering the injury compensation system, by working to limit wasteful medical costs, and, only in part, by considering limits to benefits. REFERENCES ABBREVIATIONS AMA American Medical Association IOM Institute of Medicine WHO World Health Organization AMA. 1990. Guides to the Evaluation of Permanent Impairment, 3rd ed. (revised). Chicago, Ill. Berkowitz, M., and J. Burton. 1987. Permanent Disability Benefits in Workers' Compensation. W.E. Upjohn Institute for Employment Research, Kalama- zoo, Mich. IOM. 1991. Disability in America: Toward a National Agenda for Prevention. Na- tional Academy Press, Washington, D.C. Nagi, A. Z. 1965. Some Conceptual Issues in Disability and Rehabilitation. In Sociology and Rehabilitation (M.B. Sussman, ed.), American Sociological Asso- ciation, Washington, D.C. Rolph, E. 1992. Framing the Compensation Inquiry. Rand Reprint RP- 115. RAND Corporation, Santa Monica, Calif. WHO. 1980. International Classfication of Impairments, Disabilities and Handicaps: A Manual of Classfication Relating to the Consequences of Disea.se. Geneva.

........................... Railroad Injury Compensation Process Some elements of the railroad injury compensation process are inde- pendent of the provisions of the Federal Employers' Liability Act of 1908 (FELA), as noted in Chapter 2, and would exist without it, whereas others are the result of the FELA statute. However, in reality all ele- ments of the railroad injury compensation process are shaped and influenced by the existence of FELA. Some of the benefits and pro- cedures that are similar to those of workers' compensation systems are the result of collective bargaining agreements between railroad labor and management. Others have been instituted by the railroads uni- .s- laterally. In most instances, these elements have as goals the reduction of time and cost to resolve claims and of the loss of good will between employees and management that results from the adversarial relation- ship created in extended disagreements and negotiation. Figure 4-1 outlines the basic FELA injury compensation process, which includes recoveries from FELA settlements and from sources for income main- tenance outside of FELA. The intent of this chapter is to present a description of the FELA compensation process that highlights its elements, beginning with an overview of the basic process and followed by a more detailed descrip- tion of FELA benefits and costs in the context of the FELA claims resolution process. The concluding section is an examination of the 59

60 Compensating Injured Railroad Workers Under FELA INJURY File Report/Start Claim No 4~ MEDICAL TREATMENT Lost I ___rTim ________________ Medical Benefit (AR) RETURN TO WORK Lost Time INCOME MAINTENANCE Sickness Benefits (RRB) MEDICAL Short-Term Disability Insurance (Emptoyee) REHABILITATION Wage Continuation (AR) Medical Benefit (AR) Cash Advances (AR) SETtLEMENT NEGOTIATION Direcl RETURN TO WORK Represented FELA Indemnity Settlement Litigated NO RETURN TO WORK VOCATIONAL FELA Indemnity Settlement H REHABILtTATION Long-Term Disability Benefits (RRB, Emptoyee) Employee FIGURE 4-1 FELA injury compensation process. NOTE: Though they appear as discrete steps in this flowchart, medical treatment, rehabilitation, and negotiation are ongoing activities from the time of injury. (RRB = Railroad Retirement board; RR = employing railroad.) data available on an industrywide basis on the FELA compensation process and costs. More detailed data on individual railroads are pre- sented in Chapter 6, in which FELA is compared with workers' com- pensation systems. OVERVIEW OF FELA PROCESS Basic Process In the basic FELA process, a worker has the right to sue the railroad employer for full compensation for the costs incurred from on-the-job injuries. These costs include medical treatment, extended medical rehabilitation, replacement of wages lost during recovery and rehabili-

Railroad Injury Compensation Process • 61 tation, compensation for future earning potential lost because of a lingering impairment, and compensation for pain and suffering. As noted in Chapter 3, the courts have created a liberal liability standard for the FELA statute. A worker need only establish some negligence, however small, on the part of the employer to recover some damages. The portion of the total costs that each party, the worker or the railroad, must bear is determined by a comparative negligence standard wherein the worker's recovery is reduced in proportion to his or her own negligence in the injury. However, if the injury is the result of a railroad violation of any safety rules or regulations, the worker may recover 100 percent of costs regardless of his or her own negligence. The right to sue does not imply the need to sue in every case. On the most basic level, the right to seek recovery conferred by the statute merely ensures that a judicial remedy is available if the employer refuses to settle or negotiate when asked by the injured worker to pay that portion of cost for which the employer is responsible. The likely results at trial strongly influence the process of negotiating settlements. If used for all claims, the judicial remedy would create exorbitant costs for transacting claims for both railroads and employees, especially since most claims involve relatively minor injuries. (The judicial rem- edy would also create costs for society for building courtrooms, hiring judges and administrative staff, etc.) An incentive for both the rail- roads and their employees to develop routine processes and alterna- tives to litigation has been the avoidance of such costs. Nevertheless, administrative resolution of claims may be difficult in a number of situations, particularly those in which the potential liability is great. Cases likely to result in extended negotiation include those in which doubt is present about negligence, the validity of the claim, or the actual and potential level of losses. Determining negligence and compensation for pain and suffering may require lengthy negotiations, especially for serious injuries. The extended time to resolve claims is a common criticism of the current FELA process by both management and labor. As the time to resolve claims increases, railroad costs for administration and legal representa- tion increase as well as railroad workers' costs for lost income and legal representation. Benefits All railroad workers are covered by FELA. Injuries may run the gamut from first aid to hospitalization. They may include cases in which no time or wages are lost but there are medical expenses. Most of these claims are settled quickly and only occasionally involve an attorney.

62 • Compensating Injured Railroad Workers Under FELA Although the elements of the FELA process follow essentially the same basic pattern among all the railroads, there is some variation. The following description of the FELA process is based on discussions concerning their approaches with a number of railroads, including the Burlington Northern, CSX Transportation, Norfolk Southern, and Southern Pacific. The goal here is not to describe in detail the approach being used by every railroad, but to establish the basic approach and note the areas that vary. Medical Treatment and Rehabilitation The FELA statute permits injured workers to sue for the full costs of their medical expenses and rehabilitation. All potential medical cost reimbursement can be sought and there are no exclusions. For this type of cost, FELA provides the fullest extent and level of coverage. How- ever, whether an injured worker will be compensated for all medical costs could ultimately depend on the finding of negligence and the decision on whether the treatment was necessary as a result of the injury. As described in Chapter 2, the railroads and their employee unions have collectively bargained an administrative solution to the provision of medical coverage for injured workers.1 Instead of having to litigate medical costs for each case, workers are provided 100 percent coverage under the collective bargaining agreement. Beyond coverage and reim- bursement, however, medical treatment under FELA is conducted in the following manner. Medical treatment can be divided into the initial care for the injury and the longer-term treatment for anyconditions requiring extended convalescence and rehabilitation. Initial medical treatment may involve only first aid, a visit to a physician's office, or for serious injuries, emergency medical treatment at a hospital. For nonemergency and continuing care, the employee generally has the right to choose a physician. However, many railroads request that the employee be examined by a physician designated by the railroad, sometimes in consultation with the employee's union, when the worker is being treated for an on-the-job injury. The company's designated physician does not provide treatment, but the examination provides a second opinion of the diagnosis and prognosis for the case for railroad use in negotiations with the employee over the severity and effects of the injury. Long-term care and rehabilitation are provided by the employee's physician as well, though, as with many medical plans, the railroad insurance carrier may require proof of need for medical procedures

Railroad Injury Compensation Process • 63 and care. From the railroads' standpoint, a speedy rehabilitation and return of the employee to work is desirable. Therefore, any voluntary (not collectively bargained) income maintenance assistance to the employee until settlement is usually conditioned on the employee's meeting requirements for rehabilitation. This involves the employee's remaining in rehabilitative treatment until he or she has recovered or achieved maximum medical improvement. Railroads may also require that their nurse consultants monitor the worker's case, and will cer- tainly require that the employee agree to periodic examinations by the railroad's designated physician. When that physician determines that the employee is fit to return to work, the railroad's voluntary assistance generally ceases, whether the employee returns to work or not. One railroad noted that it emphasizes a rapid medical intervention by its nurse consultants after the initial care and recovery. The aim is to determine with the employee and the employee's physician the next steps in treatment and the projected date of the employee's return to work. If the employee is still not fit to return to his or her former job after maximum medical improvement has been attained, vocational evaluations begin to determine other job possibilities. Guard and clerk positions are often the most appropriate, but many of these jobs are covered by seniority rights. Given the reductions in rail employment over the last decade, alternative positions are usually not available. Several railroads reported that they tried assigning injured em- ployees to light-duty work during rehabilitation, but that most such efforts have been discontinued for various reasons, including problems with reinjury, resentment from the injured worker's fellow workers, antagonism from the unions, and disagreements with the worker about when he or she is ready for regular work. Income Maintenance FELA permits workers to recover all losses connected with an on-the- job injury, including lost wages, pain and suffering, and survivor bene- fits. Reimbursement for lost wages is not limited to the temporary period of work lost during medical recovery and rehabilitation. If the worker remains impaired after maximum medical improvement or is unable to return to work, he or she may seek compensation for future earning loss, whether total or partial. The benefit of compensation for pain and suffering may be particularly important when a lingering impairment does not disable the worker's income-earning capability. As discussed in Chapter 2, there are several sources of compensation other than FELA that help an injured worker maintain income during

64 • Compensating Injured Railroad Workers Under FELA the out-of-work period and while a claim is being negotiated. These include sickness benefits from the Railroad Retirement Board, private short-term disability policies that some railroads and unions offer, long-term disability benefits, cash advances, and railroad wage contin- uation programs. Cash advances and wage continuation are examined below since they are closely coupled with the FELA process. Although cash advances and wage continuation programs are not requirements of the FELA statute, they have arisen because of and in concert with the process of recovery that the statute engenders. These programs are meant to influence the FELA settlement process and therefore are not completely independent of it. Because the delay in negotiating a settlement leaves the worker without a normal income, which is a fundamental problem with the tort approach, the railroads have developed the income maintenance programs of cash advances on FELA settlements and wage continuation. Cash Advances Because sickness benefits and private insurance at best cover only about half the average employee's after-tax earnings, employees may be faced with a significant loss of income during periods of recovery from injury. The traditional method of helping employees maintain income has been through voluntary advances by the railroad on the value of the potential settlement. Railroad cash advances are negotiated between the claims agent and the employee. As noted previously, initial or subsequent advances may be conditional on the employee's cooperating in seeking medical and rehabilitative treatment. One large railroad company indicated that their cash advances provide about 25 percent of the expected settlement. Railroad management and labor have strongly differing views of the motives for cash advances and their effect on workers. Some railroad employees have expressed the view that railroads use advances to try to force workers to settle their claims early. By providing some money but not enough to maintain the worker's normal income, the employer "starves out" the employee, who returns to work too soon or settles for less than he or she is due. Railroads, on the other hand, maintain that advances help the em- ployee and maintain good relations, but argue that as advances ap- proach the full value of the expected settlement, the disincentive to return to work increases. Management believes that replacing only a portion of wages provides the incentive to the worker to negotiate a settlement and return to work as quickly as possible.

Railroad Injury Compensation Process • 65 Wage Continuation One union official noted that one of the biggest problems in on-the-job injury compensation from the worker's standpoint is the length of time for recovery and the often-resulting need to borrow money. To remedy that problem, some railroads, including the Burlington Northern, CSX Transportation, and Southern Pacific, have voluntarily instituted wage continuation programs for their employees instead of providing cash advances. Wage continuation is the continued payment of a worker's normal wages from the time the worker is injured until his or her return to work. Because these are not collectively bargained programs, the benefits can be altered or withdrawn at any time. Union officials have welcomed the effect of these programs for maintaining workers' income during recovery; however, they have criticized them for bein unilaterally imposed by the railroads without negotiations or bargaining. Railroad management apparently wishes to retain the flexibility of adjusting the programs as they gain experience with them. One company, for example, initially offered wage continuation for up to 12 months. Subsequently, it was decided that this lengthy benefit period was a disincentive for employees to return to work and that it did not dis- suade enough lawsuits by encouraging negotiations. Therefore, the company concluded that the program was too costly and reduced the eligible period for wage continuation to 3 months. If the employee is close to returning to work at that point and settlement is near, cash advances may be used to tide the employee over until the negotiations have been completed. This company noted that their only condition for wage continuation is that the employee cooperate with the company and maintain medical treatment. Another railroad has a significantly more extensive wage continua- tion program, which provides up to 3 years of benefits if four criteria are met. First, the injury must not be due to drugs, alcohol, or an illegal act; second, the employee must cooperate with the company concern- ing medical treatment; third, the employee must also cooperate by giving the railroad a statement about the incident; and fourth, after negotiations for a settlement have concluded, the employee must not reject what the railroad considers a reasonable final settlement offer. If he or she does not meet the fourth criterion, wage continuation stops. The employee would then file suit to recover his or her losses. As noted in Chapter 1, one of railroad management's criticisms of the tort process under FELA is that it affects safety because it reduces labor-management cooperation in investigating accidents. The lack of cooperation is viewed as stemming from the need to avoid admitting any fault during the investigation that would directly affect the relative

66 • Compensating Injured Railroad Workers Under FELA levels of negligence of each party. The requirement of this railroad's wage continuation program for cooperation in the investigation of the accident and injury reflects this concern. From the railroads' perspective, wage continuation programs can mitigate some of the problems that they view as resulting from the tort process. First, they may reduce the animosity created by the employee's lack of income and the adversarial relationship necessarily created in the negotiations over levels of recovery and negligence. Second, by requiring cooperation in medical treatment, these programs offset incentives to slow recovery as leverage in the settlement negotiations or litigation. Third, by eliminating wage loss, these programs may reduce the employee's need for legal representation. Finally, without wage loss, the settlement negotiations or litigation may be reduced to questions of pain and suffering, though cases that reach litigation often involve disputes over the extent of the worker's injury and therefore the level of disability and potential loss of income. Negotiation and Settlement After being injured, the employee is required to file an accident and injury report with the railroad that explains the nature of the accident and describes the circumstances of the injury.3 This filing is generally enough to start the claim process, because accident reports are usually passed on to the claims departments of the railroads. There is much concern among the workers and their supervisors over the description of the accident because of the implications it may have for fault and potential benefits. Workers have been advised to be cautious in describing the extent and severity of their injuries, how the accident happened, whether it could have been prevented, and whether any defective equipment was involved (Blount and Dvorak 1987). One union advises its members to have a union representative present when the accident report form is filled out, especially when serious injuries are involved. Railroad management has expressed concern about this disincentive to openness because it leads to difficulties in determining unsafe operational practices. The facts of an accident that might point to a needed change in procedure may be obfuscated by all the parties involved, both management and labor, in an effort to absolve them- selves of responsibility. Furthermore, one railroad company indicated to the study committee that this disincentive may lead to a reluctance to change practices immediately after an accident because of the negli- gence implication that the prior practice was unsafe. For most claims, the employee and the railroad agent negotiate directly and settle the employee's compensation. In these direct cases,

Railroad Injury Compensation Process • 67 one railroad estimates that it makes the first offer to the employee in about 95 percent of the cases because it knows from experience what such cases are worth. For minor injuries and little lost time, FELA functions administratively almost as workers' compensation does. Com- pensation is higher, however. Wage loss is covered and, so long as no major infraction by the employee or railroad is alleged, some compen- sation beyond lost wages is included, typically two to three times the lost wages. For more serious claims, direct negotiations may fail, especially if the railroad believes that the employee should bear a large share of the negligence burden. One railroad noted that many employees treat FELA like a no-fault system and believe that they should be compen- sated regardless of fault. This attitude makes it difficult to negotiate in those circumstances in which the railroad's settlement offer is greatly reduced to reflect this negligence judgment. For these cases and for most major claims, employees almost always retain a lawyer. Each party's degree of negligence is perhaps the most crucial factor in the negotiations for settlement before or after a lawsuit is filed and during litigation. One railroad company indicated that there usually are no significant disputes over the nature and severity of the injury because the company's coordinating physician works with the em- ployee and the employee's physician to seek common ground on this issue. For cases in which the railroad suspects a false claim, however, there is less likelihood of settlement, and the case goes to trial. The impetus to settle minor suits is the same as that which governs negotiations for more significant claims. As the management of one railroad put it, there is rarely an accident in which there is no negli- gence on anybody's part. The railroad can nearly always argue that the employee contributed to the injury by misjudgment or a rule violation. At the same time, it is difficult to find cases in which the railroad is clearly not negligent, because at the very least, the employee can argue that the railroad's supervision or training was inadequate or that its nondelegable duty to provide a reasonably safe place to work contrib- uted to the accident. The same railroad management indicated that they rarely fail to make some offer, even in cases where their share of negligence is slight. In negotiation, both sides tend to use rules of thumb to decide what they are seeking and for what they will settle. These include such factors as their view of the comparative negligence, the likelihood of a favorable jury outcome given the venue, the nature of the injury, the character of the employee, and the pattern of settlements in similar cases.

68 • Compensating Injured Railroad Workers Under FELA Litigation The majority of claims (an average of 70 percent from 1987 through 1991) are settled directly with the employee; another 10 percent are settled after the employee has retained an attorney but before a suit is filed. Therefore, about 20 percent of claims each year are litigated. Of these claims, only 5 percent (1 percent of all claims) reach the point of a jury verdict. Much of the settlement occurs just before the start of the trial, and many cases that begin trial are settled before they go to the jury. The settlement patterns for the industry as a whole are examined in the following section on industrywide data and on a more detailed basis in Chapter 6. Death and catastrophic loss claims such as those for loss of an arm or leg are types of represented cases that tend to be settled quickly. Lawyers take FELA cases on a contingent-fee basis, and the em- ployee must pay the fee from the settlement or award. The typical lawyer's fee is 25 percent of the employee's recovery, though it can be as high as 33 percent. The actual percentage is a matter of negotiation between the employee and the lawyer. There are law firms that special- ize in FELA defense for the railroads and ones that specialize in FELA plaintiff actions for employees. Typically, when the employee decides that a lawyer is needed, the employee's union will provide a list of FELA lawyers recommended by the union. Employees do not have to use attorneys suggested by the union, but typically they do. When cases cannot be settled directly, most railroads use outside counsel to repre- sent them in the negotiations and litigation. Employees have the right to sue in either federal or state courts. The decision depends in part on the location of the injured employee and on the likely delays in each court. According to the Southern Pacific Railroad, most of its cases in southern California are tried in federal courts because of delays in the state courts, but in other jurisdictions in which the railroad operates, employees prefer state courts. CSX Trans- portation and Norfolk Southern indicate that most of their cases are tried in state courts as well. The reasons cited include histories of higher awards in some jurisdictions. Costs The direct and indirect costs of providing compensation benefits were identified in Chapter 3. The direct costs of the benefits provided through FELA can be substantial because the statute permits full recovery of all losses and damages (except punitive damages) from the injury. This recovery is mitigated by reductions for the employee's negligence.

Railroad Injury Compensation Process • 69 Whether there actually is any mitigation of the employer's liability is an important point in assessing the FELA process. Full recovery may result in an unfair burden on employers if the sharing of the burden for losses is not allocated by findings of comparative negligence. Rail- road employers believe that because of the weakening of the compara- tive negligence standard, FELA operates like a no-fault system in which the employer bears most of the cost. This violates the basic efficiency and equity arguments in favor of a negligence liability standard based on comparative negligence. Direct costs represent a zero-sum trade-off between the employer and employee. Reducing out-of-pocket direct costs for employers in- volves reducing benefits received by the employee on a one-for-one basis. These reduced benefits may occur either from transferring responsibility for monetary costs to the employee or from not compen- sating the employee for nonmonetary losses such as pain and suffering. On the other hand, the employee may raise the employer's cost by transferring costs that are the employee's responsibility (due to his or her negligence). Information concerning direct costs on an industrywide basis is fairly limited. The number of claims and payouts for them give some indication of the use of these benefits and their levels. These data are examined in the next section. It is not possible to determine whether shifting of the employee's negligence burden to the employer occurs or not because evidence is not available. The majority of cases are handled administratively (which is used here to mean all cases that do not go to trial), and no explicit negligence finding is made. Nor is any effort made by the railroads to compile closed claims and classify them on the basis of negligence. High benefits received by employees, even when compared with those in other systems (see Chapter 6), are not necessarily indicative of the shifting of the compensation burden. Indirect costs arise from the administrative and transactions costs involved in delivering benefits. As noted in Chapter 3, both employees and employers would like to minimize their own indirect costs. The method of achieving that reduction, however, may involve losses for one or the other, or they may both benefit. Evidence on indirect costs is also limited. Individuals railroads were unable to provide precise estimates of administrative costs for their processing of FELA claims nor their costs for in-house counsel. Typ- ically, FELA claims are processed within departments that handle loss and damage and other claims; costs associated with FELA claims are not compiled separately. Attorneys work on more than just FELA claims and their time and charges were not tracked separately for

70 • Compensating Injured Railroad Workers Under FELA FELA claims by the railroads contacted. The Association of American Railroads (AAR) annually surveys Class I railroads for these costs, and these aggregate data for 1990 to 1992 are available. These data are examined in the final section of this chapter. Some of the largest transactions costs originate from the delays involved in negotiating or litigating settlements. The longer the settle- ment period, the more staff and staff time are required to handle each claim for the railroad and the greater the cost to the employee from the delay in receiving benefits. The more litigation that occurs, the greater the cost of legal counsel for both employees and employers. Indus- trywide data on the level of representation and litigation and on settle- ment delays provide a proxy for these indirect transactions costs. These data are examined in the next section. Settlement patterns and delays are examined in Chapter 6. Incentives From the public interest perspective, FELA should promote equitable treatment of all parties and minimize the resources used for injury compensation. Regarding the latter goal, two principal approaches were noted in Chapter 3: minimizing the transactions cost involved in delivering compensation and minimizing the need for that compensa- tion by cost-effective safety investment to reduce injury. Transactions Costs Because the benefit coverage of FELA provides the opportunity for full cost reimbursement, there are strong incentives on the part of em- ployers to minimize their direct costs by negotiating reduced liability for losses, eliminating fraudulent claims, supporting rehabilitation and the employee's return to work, and reducing injuries. Many of these efforts require the railroad to incur indirect costs. Investment of staff resources in negotiation, oversight, rehabilitation, investigation, and litigation may reduce the direct benefit expenses but at a cost to the company and society. Employees have similar trade-off possibilities among investment in transactions costs such as negotiation and litiga- tion and the resulting potential for improvements in the benefits they receive (direct costs to employers). Information concerning the level of transactions cost on an industrywide basis is examined in the following section. The question of whether the costs are excessive (that is, whether the same benefits could be delivered at a lower transactions cost by an alternative approach) is assessed in Chapter 6.

Railroad Injury Compensation Process • 71 Safety Supporters of FELA maintain that railroad jobs are relatively dan- gerous and that FELA is necessary to provide strong safety incentives. Accident and injury statistics indicate that although railroading is more dangerous than many industries, it is by no means the most dangerous industry. (See Appendix A on railroad industry safety.) Whether rail- road industry safety statistics present an accurate picture of safety within the industry is currently in question, however. The General Accounting Office has criticized the accuracy of these statistics in a number of reports and has recommended improvements (GAO 1989, 1990, 1991). Arguments concerning whether the railroad industry is a dangerous one or more dangerous than other industries are somewhat irrelevant, however, to the question of the design of a compensation system.4 The more relevant question is whether the compensation system acts to reinforce employers' and employees' incentives to adopt all cost- effective safety precautions. On the face of it, FELA is likely to provide employers with strong incen- tives for investing in safety. Because workers may recover the full cost of their injuries, employers are likely to invest in precautions whose costs equal the costs of potential liabilities avoided. Because these costs include those that are normally shared by the employee in no-fault workers' compensation systems, there may be an incentive to adopt more invest- ments in safety than would be the case under workers' compensation. The question then arises, "Does this increased employer burden lessen the employees' incentive to operate safely?" A true comparative negligence system would give employees the incentive not to reduce their vigilance in avoiding injury because their recovery for losses would be reduced in proportion to their negligence in the injury. If employees are successful in shifting the burden of injury costs to the employer with some certainty, however, an inefficient adoption of safety precautions between employers and employees may result.5 As pointed out in Chapter 3, however, strict liability systems may provide incentives for the employer to invest in precautions that are not readily apparent to other parties but that are cost-effective. Whether FELA leads to efficient or inefficient safety incentives cannot be as- sessed without information on the nature of burden sharing between the railroads and their employees. Equity and Efficiency FELA, being a comparative negligence statute, proceeds on the basis that each party will bear the costs for which it is responsible. In this

72 • Compensating Injured Railroad Workers Under FELA sense, FELA is potentially quite equitable for all parties. As noted earlier, whether FELA is actually equitable is difficult to determine. Railroad labor believes that FELA is equitable for railroad employees; railroad management believes that the railroads bear an undue share of the compensation costs. On the efficiency side, the FELA approach inherently increases transactions costs. The ability to recover full cost and the necessity of allocating the cost burden require negotiation among the parties. The subjective nature of some losses (e.g., pain and suffering) and the uncertainty of others (e.g., earnings potential) may cause those nego- tiations to be prolonged. The unlimited liability of the employer forces the railroad t&commit the resources necessary to avoid unjustified outcomes. Whether high transactions costs are a negative result from society's perspective depends again on one's viewpoint. For those who believe that FELA yields a more equitable result for employees (and em- ployers), the extra transactions costs involved may be justified: they are the normal and unavoidable costs of securing equitable outcomes. On the other hand, for those who view current outcomes as inequitable, these transactions costs add more costs on top of an already unfair burden. Rehabilitation and Incentives To Return to Work An efficient injury compensation system should provide incentives for employees to return to work once their injuries have healed because employees who return to productive employment contribute to eco- nomic output and reduce demands for disability-related social services. FELA provides both incentives and disincentives. On the negative side, there may be an incentive until the settlement, especially for serious injuries, for the worker to remain out of work to improve the negotiat- ing position in settlement discussions. This same incentive may be at work in workers' compensation as well. On the positive side, however, the fact that nearly all FELA settlements are lump sums encourages an injured worker to seek some type of employment once that settlement is reached, whether with the railroad or elsewhere, since the employee need not remain "injured" or "disabled" to receive compensation. Evidence on rehabilitation and return to work is limited because the railroads do not track workers who do not return to work after they leave employment, nor do their records indicate reasons why workers do not return, such as permanent total disability or early retirement. The claims files for one railroad indicate that the percentage of workers who do not return to work averaged 18 percent over the 1986-1991

Railroad Injury Compensation Process • 73 period for claims that were closed with an indemnity payment. Because the railroads were reducing their work force throughout the 1980s, some of these workers' jobs were likely eliminated, and the settlements therefore may have included elements of severance inducements as well as injury compensation. INDUSTRYWIDE DATA ON FELA PROCESS AND COSTS Some of the cost issues involved in FELA can be illustrated by the industrywide data available. To make detailed comparisons with other systems requires more detailed railroad and claim-specific informa- tion, which will be presented in Chapter 6 along with available infor- mation from workers' compensation systems. The AAR collects industrywide data from the Class I railroads on FELA claims, payments, legal representation, litigation, and adminis- trative costs. Claims and payment data indicate the level of direct costs, and administrative costs, the extent of legal representation, and the level of litigation are key indicators of transactions costs. Claims The increased total nurpber of FELA claims closed in the last 5 years gives the appearance of a growing problem because railroad employ- ment has been declining at the same time. This increase, however, is not caused by more deaths and injuries but rather by more occupational illness claims, the majority of them for hearing loss. The number of hearing-loss claims peaked in 1990, and the increase is apparently over, according to railroad claims officials. Data on claims closed on an annual basis from 1987 through 1991 are presented in Table 41.6 Normalizing the claims figures by dividing them by employment levels in each year gives a truer picture of the trends. The frequency of claims among railroad employees has decreased for death and injury and increased in the last 3 years for occupational illness.7 The entire increase in occupational illness claims per 100 employees is accounted for by the temporary bulge in hearing-loss claims (Table 4-1). Claims settled directly with the injured employee without the in- - volvement of counsel account for 70 percent of closed claims.8 Further- - more, during the 1987-199 1 period, only 20 percent were litigated. claims (ones in which a suit was filed). Although FELA is a judicial remedy for railroad injury compensation, the majority of cases are handled administratively.

74 • Compensating Injured Railroad Workers Under FELA TABLE 4-1 FELA Claims Closed for All Class I Railroads, 1987-1991 1987 1988 1989 1990 1991 No. of Claims Injury NA NA 18,074 17,398 15,324 Illness 6,283 16,544 16,949 Asbestos NA NA 734 651 751 Hearing loss NA NA 5,549 15,893 16,190 Total 20,722 20,280 24,357 33,942 32,273 Claims per 100 Employees Injury 6.9 6.8 6.3 Illness 2.4 6.5 7.0 Asbestos 0.3 0.3 0.3 Hearing loss 2.1 6.2 6.7 All 7.6 7.6 9.3 13.3 13.3 Railroad employment 273,429 268,208 261,063 254,661 241,860 Claims by Settlement Status Direct 14,213 13,717 17,774 24,845 20,668 Represented 1,868 2,084 1,830 3,368 5,153 Litigated 4,641 4,479 4,753 5,729 6,452 Total 20,722 20,280 24,357 33,942 32,273 NOTE: Injury claims include death claims. NA = not available. SOURCE: Data supplied by the Association of American Railroads and the Railroad Retirement Board. - Payments Total payments for FELA claims closed have risen steadily from $686 million in 1987 to $911 million in 1991 (Table 4-2). Part of this increase is accounted for by the bulge in occupational injury claims and by inflation, but part is also due to a steady increase in injury payments per claim on a constant-dollar basis. The payment per claim for all claims closed reached a peak in 1988 and has declined since. However, this peak is misleading because the number of occupational illness claims settled has increased greatly since 1988. Occupational illness settlements were about 20 to 25 per- cent of the value of injury settlements from 1989 to 1991 (Table 4-2). Because of these smaller payments per claim, the increase in the pro- portion of occupational illness claims has reduced overall payments per claim for all closed claims. Assuming that the occupational illness claims trend carries back- ward to 1987 and 1988 and that therefore there are few of these claims in those years, the rise in the average payment for injury claims on a

Railroad Injury Compensation Process • 75 TABLE 4-2 FELA Payments for Closed Claims for All Class I Railroads, 1987-1991 1987 1988 1989 1990 1991 Payments ($ millions) Injury 647 NA 728 761 758 Illness 40 NA 61 116 153 Total 686 768 789 877 911 Actual and Constant-Dollar Payments per Claim Actual ($) 33,120 37,852 32,403 25,851 28,213 Injury claims NA NA 40,279 43,741 49,465 Illness claims NA NA 9,693 7,030 9,003 Constant-dollar ($) 29,125 31,983 26,125 19,772 20,703 Injury claims NA NA 32,474 33,456 36,298 Illness claims NA NA 7,815 5,377 6,607 Actual and Constant-Dollar Payments per Employee Actual ($) 2,510 2,862 3,023 3,445 3,765 Injury payments 2,366 NA 2,789 2,988 3,134 Illness payments 146 NA 233 457 631 Constant-dollar ($) 2,207 2,418 2,437 2,635 2,763 Injury payments 2,081 NA 2,248 2,286 2,300 Illness payments 129 NA 188 349 463 CPI-U (1982-1984 = 100) 1.14 1.18 1.24 1.31 1.36 Railroad employment 273,429 268,208 261,063 254,661 241,860 NOTE: Payments are in millions of dollars except per claim and per employee amounts. Total payout was actually $811 million in 1988, but the number of claims closed for one railroad was not available for this year, so that railroad's payment amount was deleted. The deflator used is CPI-U, the consumer price index for urban consumers. NA = not available. SOURCE: Data supplied by the Association of American Railroads and the Railroad Retirement Board and Economic Report of the President, 1992. constant-dollar basis is continuous over this period, increasing between $500 and $3,000 per year from about $29,000 to $36,000.9 This is a total increase of 24 percent or about 5.5 percent per year. Payments per employee on a constant-dollar basis reveal the real change over time in railroad labor costs that is attributable to injury compensation. As with payments per claim, payments per employee in constant dollars rise continuously during this period from about $2,200 to $2,800, or about 27 percent. For injury claims alone, however, the increase is a more moderate 10 percent or about 2.5 percent a year.

76 • Compensating Injured Railroad Workers Under FELA Tort settlements, like workers' compensation payments, are not sub- ject to income taxes or other taxes.'0 These payments are not affected, therefore, by changes in tax laws and need not be adjusted for them. Administrative Costs Medical and administrative cost data are collected by the AAR during an annual survey of its member railroads, and these data are available for 1990 to 1992. The reported values for these costs were $287 million in 1990, $288 million in 1991, and $281 million in 1992. The inability of individual railroads to provide administrative cost data to the study committee suggests that these survey data are crude estimates, and this concern is reinforced by the breakdown of these totals between medical and administrative costs. Medical costs are shown to have declined throughout the period from $129 million in 1990, to $115 million in 1991, and to $91 million in 1992, whereas administrative costs in- creased continuously from $157 million in 1990, to $173 million in 1991, and to $190 million in 1992. Such large and rapid increases in administrative costs and decreases in medical costs seem anomalous given the slight decline in total claims. Furthermore, they are at odds with the other direct and indirect evidence concerning administrative and medical costs. (See Chapter 6.) Litigation The number of FELA claims actually tried in each year is very small in proportion to the number of suits filed (litigated claims) and even smaller in comparison with the total claims closed each year. Table 4-3 shows the number of litigated claims tried in each year and the out- comes of the verdicts for the period 1987 through 1991. Over the period 1987-1991, litigated claims closed constitute only about 20 percent of all claims closed. The filing of a suit, however, is not indicative of the level of trial activity because 95 percent of litigated claims are settled before being tried to a verdict. Claims tried to conclu- sion, therefore, represent less than 1 percent of all claims closed over the 1987-1991 period. Defense verdicts, those in which the employee receives no injury compensation, have averaged about 26 percent of all verdicts in 1987 through 1991, and this equals less than one-fourth of 1 percent of the claims closed during the period. Verdicts in which the award is less than or equal to the railroad's final settlement have averaged about one-sixth of all verdicts, and verdicts favorable to the plaintiff, as defined by the railroads, have exceeded 50 percent."

TABLE 4-3 FELA Litigated Claims Experience for All Class I Railroads, 1987-1991 1987 1988 1989 1990 1991 TOTAL Claims Total claims closed 20,722 20,280 24,357 33,942 32,273 131,574 Litigated claims closed 4,641 4,479 4,753 5,729 6,452 26,054 Litigated claims tried 223 196 200 208 307 1,134 Litigated claims as percent of total claims 22.4 22.1 19.5 16.9 20.0 19.8 Claims tried as percent of litigated claims 4.8 4.4 4.2 3.6 4.8 4.4 Claims tried as percent of total claims 1.1 0.97 0.82 0.61 0.95 0.86 Verdicts Litigated claims tried 223 196 200 208 307 1,134 Defense verdicts 67 59 46 65 62 299 Verdicts in railroad's favor 40 38 42 35 31 186 Plaintiffs' verdicts 116 99 112 108 214 649 Defense verdicts as percent of total 30 30 23 31 20 26 Verdicts in railroad's favor as percent of total 18 19 21 17 10 16 Plaintiffs' verdicts as percent of total 52 51 56 52 70 57 Defense verdicts as percent of total claims closed 0.32 0.29 0.19 0.19 0.19 0.23 NOTE: Data are for all Class I railroads except as follows: one major railroad did not participate in 1987, three did not in 1988, and one did not in 1989. "Verdicts in railroad's favor" are those that are equal to or less than the railroad's final settlment offer. SouRcE: Data supplied by the Association of American Railroads.

78 • Compensating Injured Railroad Workers Under FELA SUMMARY FELA has led to the development of a mixed administrative and judi- cial process for compensating workers in the railroad industry. Cover- age for the injured worker is broad, including all types of losses and providing the potential for full recovery of all losses, both monetary and nonmonetary. The level of actual recovery may be reduced de- pending on the worker's proportion of responsibility for his or her own injuries. How accuratejuries are in finding fault and therefore whether appropriate burden-sharing of losses occurs is not known. Direct costs of FELA flow from use of the system and the level of claims paid. The number of injury claims per employee has remained stable over the last 5 years, but occupational illness claims, mainly for hearing loss, increased sharply in the late 1980s and are now declining. In terms of the level of benefits, constant-dollar payments per em- ployee for injury claims alone have been rising at about 2.5 percent per year in the 1987-199 1 period. Therefore, the direct cost of benefits to the railroads has been increasing. Indirect costs are reflected in costs for negotiation and litigation. The proportion of injury claims in which the employee is represented by an attorney has remained stable at 30 percent. Evidence on litigation indicates that only 4 to 5 percent of litigated claims are tried to conclu- sion, representing less than 1 percent of all claims closed in each year. Litigated claims in which the injured employee receives no injury compensation represent less than one-fourth of 1 percent of all claims closed from 1987 through 1991. Although it is not possible to draw conclusions concerning the relative costs and efficiency of FELA from industrywide data, direct costs of injuries have been increasing because of moderate growth in constant- dollar injury payments. There has been a short-term expansion in occupational illness claims that appears to have run its course, but which may recur for other, emerging occupational illness issues. NOTES Not all railroad employees are covered by collective bargaining agreements with unions. Of the total 201,000 Class I employees, 79 percent are union employees. Nevertheless, nonunion employees receive the same medical benefits as union employees do. The source of this information is the remarks presented to the study committee by the unions (see Chapter 1, Note 6). Accidents and injuries must be reported to the Federal Railroad Adminis- tration as well if they pass reporting requirement thresholds. Although the level of safety is less relevant to the design of a system, it does indicate the need for one. Clearly, the level of accidents and injury in the

Railroad Injury Compensation Process • 79 railroad industry forcefully argues for an effective injury compensation system. Besides burden shifting, another concern about high benefits to injured workers is that benefit increases may increase the number of injuries, or at least claims of injuries, and their duration. This "moral hazard" argument that workers with a smaller financial incentive to avoid injury are more likely to be injured, although sound in principle, has always been contro- versial in practice. Some research has found moderate moral hazard effects for minor injuries, whereas other research has found the effects to be small, particularly for more severe injuries. For a discussion of this research, see work by Moore and Viscussi (1990). Although the bulge in occupational illness claims appears to have passed, other types of sickness claims may arise in the future, including claims for silicosis, carpal tunnel syndrome, and emotional distress. Both FELA and workers' compensation are likely to struggle during the next decade with the expansion of the definition of injury that is represented by these new occupational illness claims. To simplify the discussion, "injury claims" will be used to denote "death and injury claims." A direct settlement is one that is agreed on directly by the railroad and the employee without assistance of counsel. If the employee and the railroad cannot reach agreement through direct negotiations, the employee may retain an attorney to pursue the case. A represented claim is one that is settled through negotiation at this stage, before a suit is filed. A litigated claim is one in which a suit is filed. This should not be interpreted to mean that the claim goes to trial or even that a court date is set. Most litigated claims are settled through negotiation and resolved before trial. The deflator used in this analysis is the consumer price index for urban consumers (CPI-U). An alternative might be to deflate the figures by the wage increase percentages given to railroad workers. The CPI was used because it gives a good indication of purchasing power or the goods that FELA compensation can command; this represents to the injured worker the value of injury payments. Whether wage increases occur or not, the worker will seek in settlement the purchasing power required to maintain his or her real income. Workers' compensation awards are generally two-thirds of before-tax in- come as an approximation, on average, of after-tax income. Similarly, FELA awards are calculated on an after-tax basis and are not subject to taxes. Conventional tort settlements, on the other hand, are calculated on a before-tax basis and yet are not taxable either. FELA is therefore more like workers' compensation than other tort actions for the valuation of settle- ments. The AAR defines a plaintiff's verdict as one that is greater than the railroad's final settlement offer. Clearly, any award less than the plaintiff's final demand, however, would be considered favorable to the railroad by plaintiffs. Therefore, from the plaintiff's standpoint, the category of plain- tiff's verdict would be smaller.

80 • Compensating Injured Railroad Workers Under FELA REFERENCES ABBREVIATION GAO General Accounting Office Blount, J. H., and J. E. Dvorak III. 1987.Jv.stice for the Injured Railroad Worker (And How To Achieve It). Justice Publishing Company, Inc., Griffith, md. GAO. 1989. Railroad Safety: FRA Needs to Correct Deficiencies in Reporting Injuries and Accidents. RCED-89-109. Washington, D.C., Apr. GAO. 1990. Railroad Safety: New Approach Needed for Effective FRA Safety In.spec- tion Program. RCED-90-194. Washington, D.C., July. GAO. 1991. Railroad Safety: Weaknesses Exist in FRA'S Enforcement Program. RCED-9 1-72. Washington, D.C., Mar. Moore, M. J., and W. K. Viscussi. 1990. Compensation Mechani.smsforJob Risks. Princeton University Press, Princeton, N.J.

5 ............................ State and Federal Workers' Compensation Programs Workers' compensation programs, though quite varied in practice, share certain goals. Injured workers receive payments that cover their medical expenses and a portion of their lost wages depending on the extent of their disability and its duration. They receive these benefits whether the injury was their own fault or due in some part to the negligence of their employer. Workers' compensation has operated without much fanfare over the course of its nearly 90-year history in the United States, but during the last two decades costs have been rising sharply. These cost increases are partly attributable to inflation and more expansive coverage as recommended by a national commission in 1972. There is increasing controversy within workers' compensation programs about how to control rapidly rising medical costs, limit cover- age for certain kinds of illnesses, and manage increased litigiousness in some state programs. The purpose of this chapter is to describe the main features of workers' compensation programs and highlight some of the controver- sial areas as backgroundjor Chapter 6, in which the benefit levels that injured workers receive under the Federal Employers' Liability Act of 1908 (FELA) are compared and contrasted with those they might receive under alternative workers' compensation programs. The many differences among the state and federal programs make it difficult to [31

82 • Compensating Injured Railroad Workers Under FELA portray the general features that they share without misrepresenting the unique aspects of each program; the discussion in this chapter errs on the side of generality. Figure 5-1 is a flowchart of the basic injury compensation process under workers' compensation, which includes benefits arising from workers' compensation systems and those com- plementary to them. INJURY File Report/Stan Claim I I Lost Time > WC Minimum Lost Time <WC Minimum No Lost Time MEDICAL TREATMENT MEDICAL TREATMENT I I MINOR INJURY Medical Benefit (WC) Medical Benefit (WC) Medical Benefit (WC) No Indemnity Benefit No Indemnity Benefit INCOME MAINTENANCE Sickness Benefits (Employer) -1 RETURN TO WORK MEDICAL REHABILITATION Medical Benefit (WC) INCOME MAINTENANCE Indemnity Benefit (WC) VOCATIONAL REHABILITATION Indemnity Benefit (WC) RETURN TO WORK Scheduled Payments (WC) NO RETURN TO WORK Indemnity Benefit (WC) Scheduled Payments (WC) Compromise and Release (WC) Long-Term Disability Benefits (SSD, Employee) FIGURE 5-1 Workers' compensation injury compensation process. NOTE: Though they appear as discrete steps in this flowchart, medical treatment, rehabilitation, and income maintenance are ongoing activities from the time of injury. (WC = workers' compensation; SSD = social security disability.)

State and Federal Workers' Compensation Programs • 83 After a brief overview of program administration, an explanation is given of how workers' compensation provided for direct employers' liability for medical coverage and a substantial portion of lost wages. Subsequently, the benefits available under workers' compensation for medical expenses, lost wages, rehabilitation, and survivors of those who die on the job are discussed. Workers' compensation programs were meant to replace the liti- giousness of the tort-based system; nonetheless, disputes over benefits arise and appear to be increasing. An overview of how disputes are resolved is given next, and available information on the frequency and cost of disputed cases is summarized. Before a summary of the key points in this chapter, a special issue is addressed that arises in the consideration of how railroad workers, who are often involved in interstate commerce, might be covered by individ- ual state workers' compensation programs. PROGRAM OPERATION Because workers' compensation represents such a different approach to compensating injured workers than FELA, it is worth reviewing briefly how workers' compensation operates.' The different workers' compensation programs have unique features that make it difficult to describe the characteristics that they share without becoming over- whelmed with specifics about the exceptions. The important points to derive from this overview are that workers' compensation is largely an insurance system in which (a) state governments mandate coverage, set standards, and provide oversight; (b) employers provide coverage through self-insurance, private carriers, or state insurance funds; and (c) carriers deal directly with injured workers, typically with little in- volvement of state programs unless a dispute arises. State Programs State programs cover about 90 million workers and in 1990 paid out benefits of about $35 billion at a cost to employers of about $62 billion (Nelson 1992; Burton 1993). For the most part, the public is not very knowledgeable about workers' compensation programs. Few people understand exactly what the programs do or how they operate, and few appreciate the large sums of money spent each year to compensate the victims of workplace injuries and diseases. Roughly 2 percent of the total payroll in the country is devoted to supporting these programs. Most employees in the country are covered, with the major excep- tions of railroad workers, some farm and casual laborers, and some

84 Compensating Injured Railroad Workers Under FELA employees of state and local government. (Police and fire fighters receive special treatment in many jurisdictions.) In virtually all jurisdic- tions employers are required to purchase workers' compensation insur- ance for their employees. In the isolated cases in which coverage is voluntary, most employers elect to do so. In almost all states, employers purchase insurance from private carriers. If a worker is injured, a claim is filed with the employer's carrier. The private carrier then works directly with the employee in the determination and payment of benefits. In contrast to states that work through private carriers, six states have an exclusive fund to which all employers in the state must contribute. In these jurisdictions, the state is the sole insurer and the administrator of the system. Several other states also operate a state fund for insurance, but it is not exclu- sive; private carriers provide coverage as well. In almost alljurisdictions, large employers are allowed to self-insure; that is, the employer does not purchase insurance from a private carrier or a state fund but pays workers' compensation claims directly as they arise. Given the size of the largest freight and passenger railroads, they would most likely choose to self-insure if they operated under workers' compensation. Funding for most state workers' compensation agencies is provided through a tax on the premiums charged to private employers, which is roughly 1 percent of premiums collected. A few state agencies are funded in part or entirely through the general fund of those states. Federal Programs The federal government administers two major workers' compensation insurance programs.2 Employees of the federal government are cov- ered by the Federal Employees' Compensation Act of 1916 (FECA). Private employees in longshoring and related operations are covered by the Longshore and Harbor Workers' Compensation Act of 1972 (LHWCA). The FECA and LHWCA programs are similar to state programs in many respects, but they are national in scope and the ben- efits are higher. FECA is currently administered by an office of the U.S. Department of Labor (DOL) and covers about 3 million government employees. This program pays workers' compensation benefits for injured em- ployees and subsequently assesses the employing agency for the cost. DOL takes a very active role in administering the program, with the determination and payment of benefits based on the decisions of FECA claims examiners. The U.S. government is essentially self-insured un- der FECA, as opposed to allowing federal agencies to buy workers'

State and Federal Workers' Compensation Programs • 85 compensation insurance from private carriers. FECA made payments on about 255,000 cases in 1991 for a total cost of about $1.6 billion. An employee eligible for coverage under LHWCA is generally de- fined as anyone engaged in maritime employment, including any long- shore worker or person engaged in longshoring operations, and any harbor worker, including those engaged in ship repair, ship building, and ship-breaking activities on the navigable waters of the United States or adjoining pier and dock areas (DOL 1986). Under LHWCA, employers can purchase insurance or self-insure. The program is administered by DOL's Office of Workers' Compensa- tion Programs, but unlike FECA the payments are handled by private insurance carriers. LHWCA staff become involved only if a claim is contested. In 1990, LHWCA payments totalled $596 million for 18,889 cases. COMPENSATION FOR LOSS Workers' compensation offers employees and employers a compact. The employer is liable for complete medical coverage and the replace- ment of a substantial portion of the injured worker's wages regardless of fault. The employee, in exchange for guaranteed benefits, gives up the right to sue for recovery of damages based on the employer's negligence. Employees also give up the right to recover for pain and suffering. The intent of this approach was to offer advantages to both sides. Employers and employees would no longer have to fight in court. Employers could operate without fear of large damage claims. Em- ployees would receive their benefits promptly because they would be handled administratively rather than being delayed and possibly de- nied by court proceedings. (The extent to which these goals of workers' compensation have been reached will be described in subsequent sec- tions.) With the adoption of workers' compensation in all jurisdictions, the workers' compensation benefit became the "exclusive remedy" for workers' injuries. Although a few states have statutes that allow workers to sue in cases of egregious negligence, the exclusive remedy remains a cornerstone of workers' compensation.4 A major exception to the exclusive remedy is the emergence of third- party product liability Suits as a way for workers covered under workers' compensation to seek additional benefits for workplace-related injuries (Weiler 1989). In these suits a tort case can be pursued if the injury is alleged to have been caused by the negligence or the tort on the part of a third party. For example, an injured employee can sue the supplier of equipment owned or operated by the employer on the grounds that

86 • Compensating Injured Railroad Workers Under FELA some defect contributed to the injury or because of exposure to a hazardous material used in the construction of the equipment. All states except New York protect the employer from being a party to this suit, even in the case of employer negligence in the purchase or use of the equipment. Because of the relatively modest recovery possible under workers' compensation in the case of extreme injuries, however, injured workers and their attorneys are increasingly bringing third- party suits (Weiler 1989). MEDICAL COVERAGE As part of the compact between employers and employees that eliminated negligence suits, employers assumed liability for complete medical cover- age for injured workers. In almost all state and federal programs, medical benefits are provided without time or monetary limitations for injuries and diseases covered under the program.5 (Medical and vocational rehabilitation are discussed in later sections.) Most state and federal statutes were written initially to cover physical injuries only, but the majority of programs have been expanded by court interpretation and legislative action to include certain diseases and occupational illnesses, albeit with considerable variation across jurisdictions. The extent of coverage for occupational illnesses, such as work-related stress, is one of the most controversial areas in workers' compensation. Most jurisdictions will compensate for mental stress caused by physical injuries suffered on the job or for physical injuries that result from mental or emotional stress caused by the job (e.g., a heart attack), but few compensate for mental stress when that condition is not the outcome of a tangible injury (Schwartz 1993). California has been more expansive in the coverage of stress than other jurisdictions but has been saddled with an extraordinary escalation in such cases. Between 1981 and 1990 the number of reported stress claims in Cali- fornia increased by an average of 21 percent per year (Barth and Telles 1992). The controversy about the sharply rising cost to employers, along with a widespread perception of fraud and abuse in the filing of stress-related claims, has resulted in new legislation in recent years, and even at the time of this writing both the legislature and Governor were calling for additional reform (Schwartz 1993). Coverage of medical expenses in workers' compensation depends on the worker's having been injured on the job. In the case of a broken limb sustained in a fall, this determination is fairly straightforward; if the accident is witnessed or reported shortly after it occurs, coverage of medical expenses is not an issue. Determinations are more complex, however, for certain kinds of injuries and diseases, such as lower-back

State and Federal Workers' Compen.sation Programs • 87 injuries and heart disease, which are more complex and may have occupational and nonoccupational components. The occupational component of a back injury, for example, may be difficult to prove if the employee had a preexisting condition and the injury was not witnessed or promptly reported. The expansive coverage of mental stress claims in California has opened a debate over what share of the stress has to be caused on the job and to what extent this can be documented (Schwartz 1993). (The delays in receiving benefits because of contested or liti- gated cases are covered in a subsequent section.) Although the extent of coverage for typical injuries sustained at work has not been a matter of controversy in workers' compensation, the cost of medical coverage has become a major concern. As fast as medical costs have been rising generally, they have risen faster in workers' compensation programs (Morrison 1990; Boden and Fleish- man 1989). From 1980 to 1989, general medical expenses increased at about 10 percent per year, but in workers' compensation they increased by about 14.9 percent per year. Medical costs accounted for about one- third of total workers' compensation costs in 1980 compared with about 40 percent by the end of the decade. State programs have been slower generally than health insurance programs in imposing procedures to control rising medical costs (Morrison 1990). The debate in California to restrict coverage for occupational stress reflects concerns shared in many states about the escalating cost to employers of workers' compensation. COMPENSATION FOR LOST WAGES Since the earliest workers' compensation statutes in the United States were enacted, state and federal programs have been guided by the assumption that a substantial portion of lost wages, often interpreted as two-thirds of the worker's wage, should be replaced when workers are injured. (Because workers' compensation benefits are not subject to federal or state income tax, actual take-home pay is greater than two- thirds of net preinjury wages unless affected by other provisions, as discussed below.) During the design of these programs in the early 1900s, it was believed that replacing most of the employee's wages would provide adequate coverage during recovery from an injury or if the employee was permanently disabled. At the same time, it was thought that pay- ments in the full amount of lost income would reduce the employee's incentive to return to work, seek rehabilitation, or participate in re- training for another occupation. The landmark 1972 report of the National Commission on State Workmen's Compensation Laws reiter- ated the goal of providing two-thirds of the worker's wage for cases of total disability.

88 • Compensating Injured Railroad Workers Under FELA Although the two-thirds rule of thumb is now in place in most programs, each jurisdiction has other provisions that affect total pay- ments. Each program has a maximum and often a minimum weekly payment that can be made. Many limit the duration of payments and the total amount that can be compensated. Despite' the prevalence of the two-thirds rule of thumb, there has always been a lack of clarity regarding the goal of some nonmedical cash payments made to injured workers. As part of the exclusive remedy of workers' compensation, injured workers are no longer able to sue for employer negligence and are therefore unable to recover for pain and suffering. Yet many workers' compensation programs make disability payments for injuries that may result in a lifetime impair- ment or disability but may not entail lost earnings. The prevalence of payments, albeit often quite modest, for impairments in workers' com- pensation that are unrelated to wage loss indicates that workers' compensation incorporates some social welfare features into a system originally conceived as a replacement for lost wages (Berkowitz and Burton 1987). State Programs Nonmedical cash payments received in workers' compensation are referred to as indemnity payments, which may be classified in three broad categories: (a) temporary total disability, (b) permanent total disability, and (c) permanent partial disability. A few jurisdictions also make payments for temporary partial disability. The payments made by most, but not all, jurisdictions can be categorized in this way, but these terms are applied somewhat differently by each jurisdiction. There- fore, the following discussion should be regarded as only an overview. Each jurisdiction named could raise any number of qualifications on or exceptions to how the specific provisions described apply within its program. In addition, the states are constantly modifying the laws and regulations governing their programs. Temporary Disability In the period following an injury, workers are paid temporary disability benefits. All jurisdictions require some waiting period before the worker becomes eligible, which ranges in length from 3 to 7 days.6 In some jurisdictions the benefits are retroactive to the date of the injury, but this usually depends on the extent of the injury and the duration of disability. Temporary disability cases represent the majority of claims filed in workers' compensation programs (about two-thirds), but they are the least controversial. "These are usually the cases where there is

State and Federal Workers' Compensation Program.s • 89 an obvious injury, a clear period of disability, and an early return to work" (Welch 1989, 40). Temporary Total Injured workers unable to return to work begin by receiving temporary disability benefits. For temporary disability, most jurisdictions in the United States (41 in 1991) allow the payment of two-thirds of the worker's wage (DOL 1992, Table 6). Because disability payments are not subject to income taxes, the actual payment, unless limited by the state maximum payment, is more generous than two-thirds of the average take-home wages, and some workers may receive more take- home pay when injured than when working. Three jurisdictions cover 70 percent of workers' wages, and three cover 80 percent of spendable (after-tax) earnings. One state (Massachusetts) covers 60 percent of the workers' wages. Three jurisdictions have slightly more complex provi- sions; in Washington State, for example, the coverage varies from 60 to 75 percent depending on marital status and the number of dependents. Regardless of the share of earnings covered, the actual weekly pay- ment that a worker might receive, particularly higher-income workers, is limited in all jurisdictions.8 The maximum payments vary across states, but are often (in 22 jurisdictions in 1991) equal to the state's average weekly wage (DOL 1992, Table 6). In these jurisdictions, a worker on temporary total disability would receive two-thirds of his or her wage or the maximum, whichever was lower. Thirteen jurisdictions set the payment below the state average weekly wage, ranging from 66 percent in Delaware to 91 percent in Colorado. Seven jurisdictions set the maximum at some percentage over the state average weekly wage, ranging from 105 percent in Mississippi to 200 percent in Iowa. Eightjurisdictions have a mandated maximum weekly payment rather than one based on the state average wage or some other basis; in these states, the maximum payment ranges from a low of $265 in Nebraska to a high of $700 in Alaska. Most states stopped using a fixed maximum payment because these limits made no provision for inflation adjust- ments. By using a percentage of the state average wage as a capping device, inflation is incorporated into the legislated maximum. The maximum weekly payments for temporary disability across jurisdictions are quite varied because of the different provisions that limit them (Figure 5-2). At.the lower end, eight jurisdictions limit the payment to between $200 and $300 per week. At the higher end, only three jurisdictions (Alaska, Connecticut, and Iowa) allow a weekly pay- ment of $700 to $799.

90 • Compensating Injured Railroad Workers Under FELA 20 a) -a E 35 z Dollars FIGURE 5-2 Maximum weekly payments for temporary total disability in state workers' compensation programs, 1991 (DOL 1992) (includes 50 states and the District of Columbia). Benefits for temporary disability are paid for the duration of the temporary disability in most jurisdictions (35 in 1991) (DOL 1992, Table 6). Fourteen jurisdictions limit the payments to a fixed period of weeks. Presumably at or before the end of this period, maximum medical improvement has been reached. It is then decided whether the claimant has a temporary or permanent disability, and if it is the latter, the claimant is then reclassified. Among states with limits on the number of weeks of payment, Texas allows up to 104 weeks or the point of maximum medical improvement for temporary benefits, whichever is the shortest. At the other extreme, New Mexico allows up to 700 weeks and three other states allow up to 500 weeks. Alaska and Minnesota use a different approach for temporary disability. Payments are made until the injured worker reaches the point of maximum medical improve- ment or completes retraining. Temporary Partial Somejurisdictions provide temporary partial benefits for workers who have returned to work but at reduced wages because of the lingering effects of their injuries. Workers of this type are usually provided a

State and Federal Workers' Compensation Programs • 91 benefit based on the difference between his or her wages before and after the injury. Permanent Disability Permanent disability payments are paid for worker injuries that result in a lifetime disability. If the disability is total and the worker is unable to return to work, the worker's condition is classified as permanent total disability. If the impairment is partial ind the worker is able to return to some form of work, the condition is classified as permanent partial disability. Permanent Total Claims for extreme injuries that result in total disability are among the least controversial. The benefits for these cases are provided more or less along the same lines as those for temporary total disability. Usually the same proportion of wages is covered, with the same maximum weekly benefits. In most cases the benefits are paid for the duration of the injury or for life. Eightjurisdictions, however, provide benefits for a fixed number of weeks. Among them, Massachusetts has the lowest at 156 weeks and New Mexico the highest at 700 weeks. Five jurisdictions also set the maximum payment for total disability. Of these, Mississippi has the lowest at $96,000 and New Mexico the highest at about $210,000. In 1988 the average benefit per compen- sable case for permanent total disability in all jurisdictions was about $200,000 (Nelson 1992). In some cases of extreme injuries, whether the worker is able to work in the previous area of employment or able to work at all is open to interpretation. If the worker could be retrained for another occupa- tion, his or her condition would be classified as permanent partial disability. Permanent Partial Permanent partial disability is the most complex, diverse, and contro- versial workers' compensation category.'° It takes the largest share of total cash benefits (60 percent) even though it represents only about one-fourth of total cases paying cash benefits (Berkowitz and Burton 1987). Most of the dispute and litigation in workers' compensation occurs over claims in this category. "Workers argue that benefits paid are not adequate, employers claim that they are too costly, and scholars point out that they are not distributed equitably" (Welch 1989,41). The

92 • Compensating Injured Railroad Workers Under FELA average payment for "major" cases of permanent partial disability was about $65,000 in 1988; the average benefit for "minor" cases was $10,242 (Nelson 1992). In most jurisdictions, if a worker has an injury that results in the loss of a finger or hand, compensation is based on a schedule of benefits for this kind of injury. These kinds of permanent partial disability claims are less complex. Most of the complexity, and therefore most of the dispute, in permanent partial disability claims involves cases that do not fit within the schedule. Scheduled Benefits. Almost all jurisdictions have a schedule of benefits for certain kinds of injuries, such as the amputation or loss of a limb, loss of vision in one eye, loss of hearing, and so forth (see Table 5-1). In most jurisdictions, individuals with an injury that has a scheduled benefit can receive temporary total disability benefits con- currently if eligible, but if they receive a scheduled benefit, they are not eligible for any other permanent disability benefits. Benefit schedules date back to the earliest state legislation. In 1911 New Jersey was the fourth state to enact legislation and the first to include a schedule of benefits for certain injuries. The schedule was apparently derived from the benefits paid for accidental dismember- ment by private insurance companies; the schedule was justified as an administrative expediency [Reede (1947); cited by Berkowitz and Bur- ton (1987)]. Indeed, New Jersey legislators expected that the schedule would be self-explanatory and not require staff to administer it (Ber- kowitz and Berkowitz 1991). Over time, the schedules in some states have become quite lengthy and complex, which has reduced the in- tended convenience of this approach. In a few states the scheduled benefit bears no relationship to the worker's previous wages. In these states, everyone receives the same cash award for the same impairment. Many states, however, base the award on the worker's prior income so that higher-income workers receive a higher award. Although this practice incorporates some ele- ments of potential wage loss, it still does not relate directly to actual wage loss. Because of the many differences across jurisdictions, the variation in the maximum payment across states is quite large. For example, in 1991 the maximum compensation for the complete loss of use of an arm at the shoulder was $22,167 in Massachusetts, whereas in neighboring Connecticut the maximum payment was $153,192 for the same injury (Table 51)11 A severe injury resulting in the complete loss of use of the arm would be eligible for the maximum benefit. In some jurisdictions, however, a

State and Federal Workers' Compensation Progranu • 93 judgment is made about the degree of economic loss for some injuries if any motor function remains after the injury has healed. A severe injury that leaves some use of the arm might be treated as a scheduled injury with a payment of some percentage of the maximum or as an un- scheduled injury. Unscheduled Benefits. jurisdictions determine payments for un- scheduled injuries on three bases: (a) the degree of impairment that results from the injury, (b) the loss of wage-earning capacity, or (c) the loss of actual wages. In jurisdictions where benefits are based on the degree of impairment, physicians judge how much function the em- ployee has lost as a result of the injury. In jurisdictions where benefits are based on the loss of wage-earning capacity, the tendency is to combine an impairment rating with the employee's previous earnings, age, and sometimes occupational disability. Thus, the estimated effect of an impairment would be modified on the basis of age and occupa- tion, and benefits would be based on a percentage of previous earnings. In a few jurisdictions benefits are based more on the extent of the workers' wage loss on return to work; this method makes no attempt to estimate impairment. Benefits are paid retroactively and are adjusted from time to time as the worker's wages change. For the few jurisdictions that rely on impairment ratings alone to assess compensation for injuries, there is no direct link between the injury and income loss. In Washington State, for example, benefits for unscheduled injuries are based on a medical assessment of the percent- age loss in "total bodily impairment," with a maximum payment of $90,000 (Pease 1989; DOL 1992, Table 8). A worker rated as having lost 20 percent of bodily function would receive a flat payment of 20 percent of $90,000. This payment may be far less than the actual wage loss suffered. In contrast, the worker may have been able to return to work without significant wage loss but would still be eligible for a benefit. States that use loss of wage-earning capacity make a somewhat more direct link between occupational illness or injury and lost income. Indeed, in California one of the occupational factors to be considered includes an assessment of the employee's diminished ability to compete in an open labor market. A disadvantage of using loss of wage-earning capacity to determine benefits is the complexity of the method and its openness to dispute and litigation. The injured party and the employer must first reach some agreement about the degree of impairment, which is usually ajudgmental decision on which experts can disagree.'2 Even after agreeing on an impairment rating, the parties must then decide how much the impairment will affect future earnings, which is

TABLE 5-1 Maximum Benefits and Number of Weeks for Selected Permanent Partial Disabilities, 1991 (DOL 1992, Table 9a) SCHEDULED INJURIES HEARING LOSS ARM AT SHOULDER HAND LEG AT HIP FooT EYE BOTH EARS ONE EAR JURISDIcTION BENEFIT ($) WEEKS BENEFIT (8) WEEKS BENEFIT ($) WEEKS BENEFIT (8) WEEKS BENEFIT (8) WEEKS BENEFIT (8) WEEKS BENEFIT ($) WEEKS Alabama 48,840 222 37,400 170 44,000 200 30,580 139 27,280 124 35,860 163 11,660 53 Alaska" - - - - - - - - - - - - - - Arizona 84,001 260 70,108 217 70,108 217 55,893 173 42,000 130 84,001 260 28,108 87 Arkansas 38,105 210 28,669 158 33,586 184 23,770 131 19,052 105 28,669 158 7,620 42 California" - - - - - - - - - - - - - - Colorado 31,200 208 15,600 104 31,200 208 15,600 104 20,850 139 20,850 139 5,250 35 Connecticut 153,192 312 123,732 252 116,858 238 92,308 188 115,385 235 76,596 156 25,532 52 Delaware 78,097 250 68,725 220 78,097 250 48,982 160 62,478 200 54,668 175 23,429 75 Dist. of Col. 191,284 312 149,593 244 176,569 288 125,683 205 98,094 160 122,618 200 31,880 52 Florida' - - - - - - - - - - - - - - Georgia 50,625 225 36,000 160 50,625 225 30,375 135 33,750 150 33,750 150 16,875 75 Hawaii 136,344 312 106,628 244 125,856 288 89,585 205 69,920 160 87,400 200 22,724 52 Idaho 59,400 300 53,460 270 39,600 200 27,720 140 34,650 175 34,650 175 - - Illinois 196,719 500 124,588 190 180,325 275 101,638 155 104,916 160 70,776 200 17,694 50 Indiana" 31,000 - 22,000 - 26,500 - 17,500 - 17,500 - 22,000 - 7,500 - Iowa 168,500 250 128,060 190 148,280 220 101,100 150 94,360 140 117,950 175 33,700 50 Kansas 60,690 210 43,350 150 57,800 200 36,125 125 34,680 120 31,790 110 8,670 30 entucky" - - - - - - - - - - - - - - Louisiana 59,000 200 44,250 150 51625 175 36,875 125 29,500 100 None - None - Maine" - - - - - - - - - - - - - - Marylandd 142,800 400 118,881 333 142,800 400 118,881 383 118,881 333 118,881 333 19,875 125 Massachusetts' 22,167 (43) 17,528 (34) 20,105 (39) 14,950 (29) 20,105 (39) 39,695 (77) 14,950 (29) Michigan 118,629 269 94,815 215 94,815 215 71,442 162 71,442 162 None - None - innesota" - - - - - - - - - - - - - - Mississippi 45,436 200 34,077 150 39,757 175 28,398 125 22,718 100 34,077 150 11,087 40 Missouri' 49,548 232 37,375 175 44,209 207 32,036 150 29,900 140 35,880 168 9,397 44 Montana" - - - - - - - - - - - - - - Nebraska 59,625 225 46,375 175 56,975 215 39,750 150 33,125 125 - 13,250 50 Nevada" - - - - - - - - - - - - - - New 132,930 210 119,637 189 88,620 140 62,034 98 53,172 84 77,859 123 18,990 30 Hampshire

NewJerseyt 128700 330 78,032 245 122,850 315 65,782 230 49660 225 38,200 200 6,540 60 New Mexico 61,460 200 38,413 125 61,460 200 35,340 115 39,949 130 46,095 150 12,292 40 New York 109,200 312 85,400 244 100,800 288 71,750 205 56,000 160 52,500 150 21,000 60 North 102,960 240 85,800 200 85,800 200 61,776 144 51,480 120 64,350 150 30,030 70 Carolina North Dakota' 83,500 250 66,800 200 78,156 234 50,100 150 50,100 150 66,800 200 16,700 50 Ohio 99,675 225 77,525 175 88,600 200 66,450 150 55,375 125 55,375 125 11,075 25 Oklahoma 46,250 250 37,000 200 46,250 250 37,000 200 37,000 200 55,500 300 18,500 100 Oregon' - - - - - - - - - - - - - - Pennsylvania 186,550 410 161,525 355 186,550 410 113,750 250 125,125 275 118,300 260 27,300 60 Puerto Rico 12,000 300 12,000 200 12,000 300 11,375 175 h - 12,000 200 3,250 50 Rhode Island 28,080 312 21,960 244 28,080 312 18,450 205 14,400 160 18,000 200 5,400 60 South 83,560 220 70.267 185 74,065 195 53,175 140 53,175 140 62,670 165 30,386 80 Carolina South Dakota 61,600 200 46,200 150 49,280 160 38,500 125 46,200 150 46,200 150 -' - Tennessee 58,800 200 44,100 150 58,800 200 36,750 125 29,400 100 44,100 150 -, - Texas 61,400 200 46,050 150 61,400 200 38,375 125 30.700 100 46,050 150 -' - Utah 47,124 187 42,336 168 31,500 125 22,176 88 30,240 120 25,200 100 -' - Vermont 127,280 215 103,600 175 127,280 215 103,600 175 74,000 125 106,560 180 30,784 52 Virgin Islands 59,620 220 48,780 180 48,780 180 32,520 120 52,845 195 48,780 180 32,520 120 Virginia 83,600 200 62,700 150 73,150 175 52,250 125 41,800 100 41,800 100 20,900 50 Washington" 54,000 - 48,600 - 54,000 - 37,800 - 21,600 - 43,200 - 7,200 - West Virginia" - - - - - - - - - - - - - - Wisconsin 72,000 500 57,600 400 72,000 500 36.000 250 39,600 275 31,104 216 5,184 36 Wyoming 39,200 150 31,882 122 35,280 135 26,133 100 24,565 94 20,906 80 10,453 40 "Ratings for compensation purposes are determined as a percentage of permanent total disability (Alaska, California, Idaho, Kentucky, Maine, Minnesota, Montana, Nevada, and West Virginia). tFlorida: Benefits are paid according to a wage loss formula rather than a statutory schedule, subject to a maximum of 364 weeks of compensation for disability greater than 24 percent. "Indiana: Benefits are paid according to the degree of permanent impairment suffered by the employee. dMaryland: The number of weeks of benefits is increased by 331/3 percent if the number is at least 250. "Massachusetts: Provides for lump sum payments determined by multiplying the state average weekly wage by a certain number (given in parentheses). 'Missouri: If the injury is total by reason of severance or complete loss of use thereof, the number of weeks of compensation allowed in the schedule for such disability is increased by 10 percent. Nebraska: Loss of hearing in both ears constitutes permanent total disability. New Jersey: Where members are amputated, an additional 30 percent is added to the award. North Dakota: Maximum weekly benefits increased to 100 percent from 331/ percent of the state average weekly wage. Benefits are increased by 25 percent if loss is to master arm or hand. Oregon: Law provides for a payment of $305 for each degree of scheduled injury and $100 for each degree of unscheduled injury subject to a maximum of 320 degrees, in monthly payments. 5 Puerto Rico: The manager of the state insurance fund determines the extent, of an eye disability on the basis of an expert report of an oculist. Monaural loss is determined as a percentage of binaural loss (South Dakota, Tennessee, Texas, and Utah). "Washington: Law provides for payment of fixed sums for specified injuries in weekly, monthly, or lump sum payments, under certain circumstances.

96 • Compensating Injured Railroad Workers Under FELA an inherently difficult assessment. Wisconsin, for example, whose pro- gram is often cited as a model with a low rate of litigation for its overall program, uses the loss of wage-earning capacity in determining the benefits for unscheduled permanent partial disability. Nevertheless, the amount of dispute and litigation in this component of the program is comparable with that of other states (Boden 1988). A few states (Florida and, to a lesser degree, New York, Pennsylva- nia, and Michigan) rely on actual wage loss to determine benefits. Wage loss is assessed after the employee has returned to work; typically, the worker receives two-thirds of the difference between his or her wages before and after the injury. This method, in concept, most closely relates compensation for a workplace injury to the actual lost earnings caused by the injury. In practice, no state has a pure wage-loss system, though several states incorporate some wage-loss assessment with their other methods. Although the wage-loss method has some advantages in terms of administrative simplicity compared with the other methods— it reduces disputes over impairment ratings—it also has disadvantages. Pegging benefits to a substantial proportion of preinjury wages may reduce the employee's incentive to return to work. On the other hand, employers would like to show that the employee is capable of returning to full work without wage loss. Therefore, establishing the right to a particular benefit level may be as difficult as in the loss of wage-earning- capacity approaches. Regardless of the approach used in determining permanent partial disability benefits, most jurisdictions place limits on the amount and duration of weekly payments. In many jurisdictions (33), weekly pay- ments for permanent partial disability are limited to two-thirds of the preinjury average wage. Two jurisdictions have lower limits, that is, 50 to 60 percent of preinjury wages. In jurisdictions that rely on impair- ment ratings rather than preinjury wages to determine permanent partial disability benefits, there is no wage-related maximum payment but a tendency to limit total payments. The distribution of maximum weekly payments across jurisdictions is shown in Figure 5-3. Some states place limits on the duration of permanent partial disabil- ity benefits, whereas others place limits on the total amount of pay- ments for those benefits. The duration of benefits (as of 1991) is most often based on a fixed number of weeks but with considerable differ- ence across jurisdictions (Figure 5-4). Nine jurisdictions provide bene- fits for the duration of disability. In eight otherjurisdictions, there is no maximum duration; total payments are limited instead. There are no specific limitations on the total amount of permanent partial disability payments for unscheduled injuries in 15jurisdictions. In 14 jurisdictions the limit is less than $100,000. The highest awards

20 U) C .215 0 0 U) a) .0 E5 J 2 Dollars FIGURE 5-3 Maximum weekly payments for permanent partial disabil- ity in state workers' compensation programs, 1991 (DOL 1992) (includes 47 states and the District of Columbia; provisions in three states are too complex to be simplified for this figure). 12 w4 .0 E J z Weeks FIGURE 5-4 Duration of payments for permanent partial disability in state workers' compensation programs, 1991 (DOL 1992) (includes 50 states and the District of Columbia).

98 • Compensating Injured Railroad Workers Under FELA 16 14 C 0 . 10 - 8 Cb S b ' Dollars FIGURE 5-5 Maximum payments for permanent partial disability for unscheduled injuries, 1991 (DOL 1992) (includes 50 states and the District of Columbia). possible in states with limits on total permanent partial disability are less than $400,000 (Figure 5-5). Federal Programs For government workers with temporary disability, FECA provides employees with 100 percent wage coverage for the first 45 days out of work. After that time, 66 percent of the worker's salary is covered for the duration of the disability. The maximum weekly payment ($1,555.84 in 1991) is set at three-fourths of the salary of a GS- 15, Step 10, which is roughly equivalent to $62,600 a year (9/4 x. $85,500). If a worker is later assigned to permanent disability, three-fourths of wages are paid if the employee has one or more dependents. If an injury is judged to be partial, fixed payments are made based on FECA's schedule of benefits. FECA's maximum benefits for sched- uled injuries are well above those of the average state and even higher than those in the states with the highest benefits for individual injuries. For example, in 1991 the maximum compensation for the loss of an arm at the shoulder under FECA was $375,760, well above the $196,719 paid in the state with the highest benefit for this injury (Illinois). Unscheduled injuries are compensated on the basis of wage loss.

State and Federal Workers' Compensation Progranas • 99 That is, injured employees who return to work will be provided with two-thirds to three-fourths of the difference between their wages be- fore and after injury (the larger payment depends on whether the injured worker has dependents). Under LWHCA, temporary total and permanent total disability benefits equal two-thirds of the injured employee's average wage, sub- ject to a minimum of 50 percent and a maximum of 200 percent of the national average weekly wage ($699 in 1991). The act sets a waiting period of 3 days, with benefits paid within 2 weeks retroactive to the occurrence of injury. For permanent partial disability, a schedule of partial disability benefits is set forth in the act, as amended in 1984. The current schedule is more generous than that in most states and is roughly equivalent to the highest state benefit levels. Unscheduled injuries are compensated as they are under FECA; the payment is based on 662/3 percent of the wage loss. Cost Increases The benefits provided under state and federal programs have been expanded since the 1972 report of the influential National Commission on State Workmen's Compensation Laws. The commission focused national attention on the paltry benefits provided in some states and suggested that federal action may be required if states did not improve their benefits. Over time most states have expanded their benefits as recommended by the commission, but this has affected the cost to employers, which has grown from $2 billion in 1960 to $62 billion in 1993 and has more than doubled since 1985 (Nelson 1988; Burton 1993). Only part of this increase is attributable to inflation and ex- panded benefits. In addition, utilization of the program by workers is increasing, medical expenses in workers' compensation are growing faster than medical expenses generally, and, as discussed in the next section, in some states litigation costs are increasing sharply (R. Victor, unpublished data, 1992). REHABILITATION Although the earliest laws enacting workers' compensation programs were not explicit about rehabilitation, the provision of medical services and income support to allow employees to return to work was soon extended to include the periods of medical and vocational rehabilita- tion (Berkowitz 1991). At one point in the development of workers' compensation programs, rehabilitation was viewed as an effort to re-

100 • Compensating Injured Railroad Workers Under FELA store injured workers to full function or to the highest level of function possible. The goal of rehabilitation has more recently become the return of the worker to his or her job at minimum cost (Berkowitz 1991). Studies of rehabilitation efforts suggest that getting the worker back to the job, familiar surroundings, and coworkers is not just less costly, but also more successful for the individual (Berkowitz and Berkowitz 1991). Although workers' compensation programs clearly facilitate rehabil- itation and offer injured workers options and benefits to regain self- sufficiency, the provision of these benefits is not without controversy. There is considerable debate over whether the provision of disability indemnity benefits for an extended or unlimited period of time re- duces the incentives of injured employees to participate in rehabilita- tion, regain self-sufficiency, and return to work. This is one of the more controversial issues in the treatment of lower-back injury, which is "the most common and costly musculoskeletal problem affecting the work- ing population" (Walsh and Dumitru 1988). Studies comparing injury recovery rates in cases with workers' compensation and those without workers' compensation have been hampered by methodological flaws, but, taking the studies as a whole, the consistency in findings suggests that extended compensation benefits prolong recovery periods for workers with lower-back injuries and reduce the frequency with which injured workers return to work (Walsh and Dumitru 1988). In situa- tions where benefits are in dispute or being litigated, it is sometimes argued that injured employees have an incentive to prolong their recovery to maximize the benefits they might achieve; at least some research supports this claim (Talo et al. 1989). State Programs Most state statutes require coverage for both medical and vocational rehabilitation. A small number ofjurisdictions do not pay for vocational rehabilitation directly but continue disability payments while workers are enrolled. Many state agencies have their own rehabilitation units, which typ- ically do referrals or monitoring; a small number of these provide rehabilitation services directly. In the typical state, the agency monitors the rehabilitation services provided by either the insurer or employer and has the option of intervening should the examiner have a concern about the case. Most jurisdictions require an injured employee to participate in rehabilitation if it is recommended by a rehabilitation specialist. If the claimant refuses to participate, disability benefits are '3 reduced or suspended.

State and Federal Workers' Compensation Programs • 101 Treating physicians stress that medical rehabilitation should begin as early as possible in order to achieve the maximum possible gains (Nelson 1988). Although insurers are required to provide coverage for rehabili- tation, there is some belief that they are reluctant to pay for medical rehabilitation early enough (Nelson 1988). During the mid-1970s to mid-1980s, a few states enacted'statutes mandating that insurers screen selected claims after a set period of 30, 60, or 120 days; determine whether rehabilitation is needed; and prepare a rehabilitation plan. Lack of clear success and higher costs have caused some states to retreat from mandating rehabilitation (Berkowitz 1990). Medical rehabilitation is typically provided if recommended by a specialist as part of the goal of obtaining maximum medical improve- ment. If it becomes apparent that the worker's injury will not allow a return to the original job, many programs offer vocational rehabilita- tion benefits. During the vocational rehabilitation period, the employee usually receives temporary total disability (TTD) benefits. The cover- age of vocational training costs or tuition, board, and travel during the training varies considerably. Some states pay TTD benefits plus tuition, others pay TTD benefits plus board and travel, but some pay only TTD benefits. Some jurisdictions place a 1- or 2-year limit on the duration of vocational rehabilitation benefits, but in many states the duration is more open ended. Summary data on the rehabilitation benefits provided injured workers are difficult to obtain. Illustrative data, however, are available from studies of individual state programs. In 1985, for example, the average worker in Florida's vocational rehabilitation program received about $5,000 in benefits (Gardner 1988). For workers eligible for education and on-the-job training, the benefits were slightly more than $14,000 (for both indemnity payments and tuition). Those eligible for only counseling or evaluation received benefits worth $1,000 to $3,000. California's vocational rehabilitation services are similar to those in Florida, but they cost more to administer and are having less success over time in returning injured workers to the workplace (Barth and Telles 1992). Federal Programs FECA operates with the philosophy of returning workers to the job as soon as practicable. FECA emphasizes the importance of rehabilitation, the initiative for which depends on the claims examiner because of the central role the examiner plays in FECA. The examiner will typically refer a client to a rehabilitation specialist, who will in turn make a

102 Compensating Injured Railroad Workers Under FELA determination and refer the client for services, usually with a private provider. All rehabilitation is paid for by FECA and the employee receives full TTD payments during medical rehabilitation plus $200 per month during vocational rehabilitation. If vocational rehabilitation is offered, the employee must accept it or disability benefits are reduced. FECA is believed to have a model rehabilitation program because it offers the most generous disability benefits of any program and be- cause the injured employee receives full pay for the first 45 days of injury; however, there is some question whether the high benefits are a disincentive for enrolling in rehabilitation and returning to work (Berkowitz 1991). FECA's emphasis on rehabilitation is believed to en- courage a timely return to the employee's previous job, but the cost- effectiveness of this approach has not been demonstrated empirically (Berkowitz 1991). The FECA program collects very little information about the injured workers handled by the program; Berkowitz (1991) points out that systematic data collection is sorely needed. LHWCA also provides both medical and vocational rehabilitation. The benefits are somewhat less generous than those under FECA. Because LHWCA is administered by the insurers rather than directly by DOL, the claims examiner has little role in rehabilitation. As with most state programs, a rehabilitation plan, if needed, is developed between the insurance provider and the employee. Benefits are re- duced if the employee refuses medical rehabilitation, but not if the employee refuses vocational rehabilitation. The LHWCA program does not keep information on the duration and cost of rehabilitation. COMPENSATION FOR SURVIVORS All workers' compensation programs provide some level of benefit for survivors. Such payments total about 3 percent of all workers' compen- sation. In 1988 the average benefit was about $121,000 (Nelson 1992). State Programs When an employee is killed on thejob or dies from injuries suffered, all jurisdictions provide some level of income support for surviving de- pendents. If a spouse is the only surviving dependent, many jurisdic- tions provide 50 to 662/3 percent of the employee's wages but apply the same caps on weekly payments as those used in disability payments. Some jurisdictions provide 50 percent of wages to a surviving spouse and increase this to 66 percent if there are surviving dependent chil- dren. The benefits usually continue until the spouse remarries. For

State and Federal Workers' Compensation Program.s • 103 dependent children, benefits typically continue until age 18. Many jurisdictions extend benefits to children until they reach age 22 or 23 if they are enrolled in school full time. A few jurisdictions limit the time period to a number of weeks, typically in the range of 400 to 500 weeks. Most state laws also provide burial allowances. The maximum al- lowed ranges from a low of $700 in Delaware to $5,000 in California, Georgia, Missouri, Nevada, New Hampshire, and Rhode Island. Federal Programs FECA provides 50 percent of the worker's wage if he or she is survived by only a spouse and 75 percent if survived by a spouse and a depen- dent child. The same maximums apply as in disability payments. The benefits continue throughout the spouse's life. Benefits for children are provided until age 18 (and beyond if they are disabled) or until age 23 if the child is a full-time student. The maximum burial allowance allowed under FECA is $800 unless the death occurred away from home, in which case provision is made for embalming and transportation. Under LHWCA, a widow or widower receives half of the deceased's average wages for life or until remarriage. If there are dependent children, the amount increases to 66 of the deceased's weekly wages. Benefits to children continue until age 18 or 23 if the child is a student (and indefinitely if the child is incapable of self-support). Death bene- fits are also paid to eligible surviving dependents. Reasonable funeral expenses up to $3,000 are covered. DISPUTE RESOLUTION One of the major goals of the workers' compensation reform movement of the early 1900s was to ensure that workers would receive medical and income support payments when they needed them rather than having them delayed by protracted court proceedings. In addition, by adopt- ing an administrative approach for handling claims it was assumed that less of the worker's final settlement would go to legal costs. For relatively minor injuries, which represent the bulk of injury claims, the goals of timely payment and reduced legal costs have been achieved. Workers' compensation, however, is in no sense free of dispute and litigation. Disputes arise over issues such as worker eligibility for benefits, the level of benefits, payments of certain medical bills, and the readiness of workers to return to work or take light-duty work and the consequent reduction or termination of benefits.

104 • Compensating Injured Railroad Workers Under FELA State Programs Almost all disputed cases are resolved through an administrative pro- cess. Appeals beyond the administrative level are possible but rare. Most cases are handled initially by an informal hearing or review. A few jurisdictions place considerable emphasis on mediation when the dis- puted claims are initiated. If resolution is not achieved in this informal process, an appeal can be made to the next level, which might be an administrative lawjudge or a commissioner of a workers' compensation board. Usually one other level of administrative appeal is possible. (Tennessee is the most notable exception; all hearings and appeals are handled through the state courts.) Case Studies In part because of the wide differences in state administrative pro- cedures, aggregate information on the frequency of disputes, formal appeals, and appeals to civil courts is not available. Detailed case studies have been conducted of several state processes, however, and these illustrate the range of procedures. Pennsylvania The dispute resolution procedures in Pennsylvania illustrate a tradi- tional, legalistic approach (Ballantyne and Telles 1991a). When parties are unable to reach a voluntary settlement, one or more will petition for a formal hearing, after which the state makes no effort at mediation. A petition is filed for a hearing in about 17 percent of lost-time claims (Table 5-2). The average case requires two hearings for resolution. By the hearing stage, both parties are usually represented by attorneys. The process is fairly slow: the average case requires 9.6 months from the petition until a hearing decision is filed, but 29 percent require more than 1 year and 9 percent require more than 2 years. Virtually all disputes are resolved administratively; only 0.5 percent of lost-time claims is ultimately appealed to civil court. The costs of the process are fairly high because of the reliance on attorneys in all formal hearings and the frequent reliance on formal depositions of medical witnesses rather than medical reports. Litigation is believed to be increasing. Wisconsin In contrast to Pennsylvania, Wisconsin has a workers' compensation program that makes a concerted effort to mediate and resolve disputes

TABLE 5-2 Measures of Dispute Resolution from Selected States STATE YEAR INDEMNITY CLAIMS HEARING REQUESTS (%) FORMAL HEARINGS (%) ADMINISTRATIVE APPEALS (%) APPEALS TO STATE COURT (%) Pennsylvania 1989 115,719 17.0 13.0 2.4 0.5 Wisconsin 1991 63,718 9.6 3.7 1.0 0.1 Minnesota 1989 46,589 29.2 13.2 1.3 * Washington 1988 55,963 13.6 2.0 1.0 0.5 Georgia 1990 53,898 23.3 5.6 2.0 0.8 New York 1991 146,945 NA NA 12.0 0.3 NOTE: The asterisk indicates less than 0.1 percent. For Wisconsin and New York, the indemnity claims are claims with payments as opposed to all claims for indemnity payments. NA = not available. SOURCE: Pennsylvania (Ballantyne and Telles 1991a), Wisconsin (Ballantyne and Telles 1992a), Minnesota (Ballantyne and Telles 1991b), Washington (Pease 1989), Georgia (Ballantyne and Eccleson 1992), and New York (Ballantyne and Telles 1992b).

106 • Compensating Injured Railroad Workers Under FELA before hearings are required (Ballantyne and Telles 1992a). Claims adjusters monitor cases closely to ensure that benefits are paid accu- rately and promptly. Program staff and administrative law judges encourage and facilitate mediation. As a result, only 3.7 percent of cases require a formal hearing (Table 5-2), which typically occurs in permanent partial disability cases because of disputes over the injured workers' earning capacity. Workers are almost always represented by an attorney in formal hearings. Only 1 percent of the decisions in these formal hearings is appealed to the highest administrative level, and only 0.1 percent of cases is appealed to the civil court system. Litigation is not increasing in Wisconsin. New York As in Wisconsin, New York has a workers' compensation program that actively seeks to resolve disputes, but the New York approach is both more extensive and more formal (Ballantyne and Telles 1992b). State claims adjusters are involved in every facet of a case. At least one meeting was held in 77 percent of all new cases in 1991 and an average of 2.8 meetings was required per case. The formal dispute resolution process takes a good deal of time because of formal procedural rules, numerous filings, and the involvement of administrative law judges in almost every step. A high proportion of decisions (12 percent) from formal hearings are appealed to the highest administrative level, but less than 1 percent is appealed to the state courts (Table 5-2). Data are not available regarding the duration of time to resolve cases, but Ballan- tyne and Telles (1992b) found that delay in resolving cases is a major concern to all parties. Washington In Washington State about 13 percent of lost-time cases are protested (Pease 1989), but only 8 percent resort to the state's resolution process and only 6 percent of appeals are accepted. A considerable effort is made to resolve disputes before formal appeal. A series of informal meetings is held to bring the parties together. For the 2 percent of cases that ultimately require formal hearings, the process is time-consuming and costly. Workers are represented by attorneys at this level. Hearings follow formal rules for civil procedure, evidence, and discovery, and rely heavily on live testimony of expert witnesses. It takes 9 to 10 months for a case to be accepted for a hearing and another year for a decision to be reached. About 0.5 percent of lost-time claims is ultimately appealed to a civil court. The frequency of litigation is increasing.

State and Federal Workers' Compensation Programs • 107 Minnesota Formal hearings are requested in about 13 percent of lost-time claims in Minnesota, but few are required (Ballantyne and Telles 1991b). A heavy reliance on informal resolution processes and mediation efforts settles most disputes. Settlement conferences are held with an adminis- trative law judge present, and three-fourths of disputed cases are resolved at this level. Decisions on formal cases are rendered in about 3 percent of lost-time claims and less than 0.1 percent is appealed to the state courts. Despite the informal dispute resolution process, attor- ney involvement is common. Workers are represented in three-fourths of conferences. Georgia Hearings are requested in about 23 percent of lost-time claims in Georgia (Table 5-2), but most of these cases are withdrawn or resolved before a hearing is held (Ballantyne and Eccleston 1992). Settlement claims arrived at before hearings, usually involving lump sums, are fairly common. The duration of cases in Georgia is relatively short. Cases involving hearings are resolved in 7 months or less. About half of judges' decisions are appealed to state court, but two-thirds of these are resolved before being heard. Those that are heard in superior court represent less than 1 percent of lost-time claims. Requests for hearings and attorney involvement in cases has increased considerably in recent years. California In California a claim for workers' compensation is considered a matter to be resolved between the parties unless there is a dispute (which is occurring with increasing frequency) (Barth and Telles 1992). Ap- proximately 45 percent of indemnity claims are litigated. (Data from California are not available in a form comparabl&with the data from other states and are therefore not summarized in Table 5-2.) If the claim is denied, the employee applies to the state for adjudication. Following this, a mandated settlement conference is held, and more than one may be required. If the dispute is not settled, a formal hearing is scheduled. The judge's decision is accepted or the claim is settled in 98 percent of formal hearings. Data are not available on the frequency of appeal to state courts.

108 • Compensating Injured Railroad Workers Under FELA Degree of Litigiousness These case studies suggest that most claims are uncontested, but there is considerable variation across states. It appears from the data pre- sented above that about 90 percent of lost-time cases in Wisconsin proceed without controversy, but this proportion drops to 70 percent in Minnesota and to 77 percent in Georgia. California appears to be the most litigious of the states reviewed here. The Workers' Compensation Research Institute (WCRI) measures litigiousness by the frequency of disputed cases and the percentage of attorney involvement. Although a consistent definition of attorney involvement is not available so that California can be compared with the other states, the California Workers' Compensation Institute estimates that about 45 percent of indemnity cases involve some degree of litigation (Barth and Telles 1992). For those cases that require formal hearings-2 to 13 percent (excluding California)—almost all are resolved in the administrative hearing pro- cess. Although administrative, and usually somewhat less formal than civil procedures, these hearings typically involve attorneys for both parties. The duration from hearing request to decision is fairly long and the process is fairly expensive in some jurisdictions. Although formal administrative hearings make up a relatively small proportion of lost-time claims, the frequency of hearings and the involvement of attorneys appear to be increasing. The amount of litigation and dispute in workers' compensation is perceived by some to be greater than it really is, but this may be attributable to the frequency of dispute in certain kinds of cases. Permanent partial disability cases are much more likely to be litigated than others. Systematic information on disputes across programs by level of injury is not readily available, but some studies have been conducted on the amount of litigation in lower-back permanent partial disability (PPD) cases, which are the most common and costly PPD cases (Boden 1988; Boden et al. 1991). These studies examined the litigious- ness of lower-back PPD claims in four states. In states with low levels of litigation for lower-back PPD claims (Oregon and Wisconsin), half or less of cases are disputed. In Oregon, half of injured workers hire attorneys and 45 percent of lowerback PPD cases are disputed. In one of Wisconsin's PPD systems, only about one-third of injured workers hire attorneys and about one-fourth of lower-back PPD cases are dis- puted.14 In the Maryland and NewJersey programs, which are believed to be typical of most state programs, virtually all lower-back PPD claims are disputed, and in almost all of these cases the injured worker retains

State and Federal Workers' Compensation Program.s • 109 an attorney. In California 98 percent of mental stress claims are liti- gated (Barth and Telles 1992). Although the cases that routinely require litigation represent a small share of total cases, the litigiousness of workers' compensation appears to be growing. The frequency with which requests are made for hear- ings is increasing faster than the increase in indemnity claims in five of the six states studied by WCRI (excluding California), and the fre- quency with which hearings are held is increasing faster than indem- nity claims in three of the six states (Table 5-3). Information about attorney involvement is not generally available. WCRI studies found that the percentage of closed indemnity cases with plaintiff attorneys in New York was 22 percent in 1987 and in Georgia was 14 percent in 1991. Information is not readily available about plaintiff attorney in- volvement over time. The percentage of cases with requests for a hearing shown in Table 5-2 probably represents an upper bound for an estimate of plaintiff attorney involvement (plaintiff attorneys are usually involved in formal hearings). In a separate estimate, about 45 percent of indemnity cases in California involved some degree of litigation in 1991, up from 29 percent in 1981 (Barth and Telles 1992). Cost of Litigation Workers' compensation was designed to be a more efficient and expe- ditious means of compensating injured workers than reliance on the legal system. The apparent growing involvement of plaintiff attorneys TABLE 5-3 Indicators of Change in Litigiousness AVERAGE ANNUAL CHANGE (%) INDEMNITY HEARING STATE YEARS CLAIMS REQUESTS HEARINGS APPEALS Pennsylvania 1981-1990 3.2 8.1 5.8 2.3 Wisconsin 1986-1991 3.1 4.0 8.3 —5 Minnesota 1981-1989 3.5 1.4 3.0 NA Washington 1985-1988 1.8 15.0 1.8 5.8 Georgia 1982-1991 3.9 5.9 0.9 1.1 New York 1982-1991 3.2 4.8 5.2 1.1 NOTE: NA = not available. SOURCE: Pennsylvania (Ballantyne and Telles 1991a), Wisconsin (Ballantyne and Telles 1992a), Minnesota (Ballantyne and Telles 1991b), Washington (Pease 1989), Georgia (Ballantyne and Eccleston 1992), and New York (Ballantyne and Telles 1992b).

110 • Compensating Injured Railroad Workers Under FELA raises a concern about the share of total compensation payments that is taken up in attorney costs. Almost all jurisdictions regulate attorney fees for claimants in some way, but the regulations vary considerably. (Attorney fees for insurance companies or private companies are not regulated.) A few jurisdictions (14) settle fees on an individual basis at the discretion of the agency. Most set fees at some percentage of the settlement, but the percentages themselves vary from 15 in Alabama to between 25 and 40 in Montana. Somejurisdictions limit the percentage as the size of the settlement grows. For example, Kentucky allows 20 per- cent of the first $25,000, 15 percent of the next $10,000, and 5 percent of the balance. A few jurisdictions also place caps on the total. For exam- ple, Delaware sets attorney fees at 30 percent or $2,240, whichever is smaller. A few jurisdictions provide for fees to be settled by the court, but even in these cases the range is set at some percentage—usually 20 to 25 percent—by regulation or statute. Georgia, Mississippi, and Mon- tana set attorney fees in the range that usually applies in FELA cases (25 to 33 percent in Georgia, 33 percent in Mississippi, and between 25 and 40 percent in Montana). In most jurisdictions the attorney's fee comes from the claimant's settlement, but a few states require the employer or insurer to split the attorney's fee. Specific information about the actual cost of litigation in workers' compensation programs—both the legal fees paid by claimants and those paid by employers and insurers—is not available. However, some general information is available from individual states. In Wisconsin's system, for lower-back injury cases that resulted in permanent partial disability and involved a typical degree of litigation, the average indem- nity benefits totaled $29,045 in 1985-1986 (Boden 1988, Table 1.1). Workers with back injuries were represented by attorneys in 89 percent of disputed cases, with an average attorney's fee of $3,576. Estimates of total plaintiff legal costs for Wisconsin are not available. In Georgia, where plaintiff attorneys are involved in 14 percent of indemnity claims, attorneys receive about 20 percent of the compensation awarded to represented workers (Ballantyne and Eccleston 1992). The $34 million in plaintiff attorney fees paid in 1990 in Georgia represented 6 percent of the total payout for indemnity claims that year. In New York, attorney involvement is higher, but total attorney costs are lower than those in Georgia. Attorneys represented 22 percent of paid in- demnity claims in 1987 in New York but with an average fee of only $630 (Ballantyne and Telles 1992b). This represented less than 1.5 percent of total benefit payments. Data are available for total legal costs in California (plaintiff and defense) per case and over time. Attorney's fees in 1990 averaged $3,145 in disputed cases; these costs were exceeded by the average cost of medical depositions ($3,317). Total average

State and Federal Workers' Compensation Programs • 111 litigation costs per disputed case were $7,030 in 1991, a 188 percent increase over the $2,500 cost in 1981 (Barth and Telles 1992). Total litigation costs equal about one-fourth of total benefits paid in Califor- nia that year (Barth and Telles 1992). Federal Programs Unlike in most other workers' compensation programs, dispute resolu- tion in FECA is a totally administrative process. The Secretary of Labor has exclusive jurisdiction over the program and any appeal (Nordlund 1991). If the employer does not contest the claim made by an injured employee, the claim is processed and the benefits are determined by a claims examiner, who reviews the claim, may request medical opinion, and collects any additional information required to arrive at a decision. If the claimant disagrees with the claims examiner's judgment, he or she can appeal to a senior claims examiner in the same district office, request a hearing from the Branch of Hearings and Review, or proceed directly to the Employee's Compensation Appeals Board, which is the final level of appeal. Although the lack of court review of FECA cases has long been controversial, the Supreme Court has upheld this feature of the program in past cases (Nordlund 1991). Almost no information is available about disputed cases in FECA. Disputed cases in LWHCA are handled administratively, but unlike in FECA, they can be appealed to a civil court. Most disputed cases (about 90 percent) are resolved in an informal process before a formal hearing is required (DOL 1986). If the issues in contention cannot be resolved informally, the case will be referred to an administrative law judge. The next level of appeal is the DOL Benefits Review Board. The case can then be referred to the U.S. Court of Appeals and ultimately to the Supreme Court. Very little information is collected about disputed cases in LHWCA. COVERAGE OF WORKERS IN INTERSTATE COMMERCE One of the issues that arises in comparing FELA with the state pro- grams is how to treat workers of companies such as railroads that operate in multiple states. Railroad labor opposes workers' compensa- tion programs in part because of the wide variation in benefits from one state to another, which would cause an employee injured in a high- benefit state like Connecticut to receive generous benefits whereas the same employee with the same injury would receive much less generous benefits in the neighboring state of Massachusetts. The answer to the

112 • Compensating Injured Railroad Workers Under FELA question of which jurisdiction provides coverage is somewhat compli- cated, but in interstate trucking a worker is usually covered in the state in which he or she is domiciled. The carrier, however, must have insurance in all jurisdictions in which it operates, because in most states the injured worker has the right to file for benefits in the state where the injury occurred. Some states refuse to acceptjurisdiction for out-of- state workers. In these cases, employees do not have an option about the state in which to file and must file in their home states. In providing coverage for workers, a carrier operating in multiple states will have to comply with the insurance requirements of that state. Hence, a nation- wide carrier might be self-insured in states that allow such coverage, would purchase private coverage in states that allow that option but do not allow self-insurance, and would participate in the state insurance fund where that approach is applicable. SUMMARY A major goal in creating workers' compensation programs was the replacement of the unreliability and delay of seeking compensation for injuries through the courts with the assurance that injured workers would receive complete medical coverage and cash payments to replace a substantial portion of lost wages. They would receive these benefits regardless of who caused the accident and would receive them in a timely way. For employers, the requirement that they compensate injured workers was softened by restrictions on their tort liability. Despite some minor incursions, states have protected employers against expanded liability over time. The growth in third-party product liability suits represents a potential threat to this protection, but, except for New York, state legislatures have not exposed employers to increased liability. There is relatively little doubt that injured workers have adequate medical coverage under workers' compensation. Claims for medical coverage for routine injuries are fully covered and rarely disputed. Nevertheless, workers' compensation programs have been slow to con- trol escalating medical costs, which is a major concern. Programs are growing in complexity—and cost—from expansion of coverage to oc- cupational illnesses and disease such as heart disease and (in a few states) mental stress. The political reforms to limit coverage of mental stress in California, which has provided the most expansive coverage, suggest a retrenchment in this once-growing area. Workers' compensation programs have met the goal of replacing a substantial portion of lost wages in varying degrees. Since the earliest workers' compensation statutes, a substantial proportion of lost wages

State and Federal Workers' Compensation Programs • 113 has been interpreted as two-thirds. In practice, however, before the 1980s, many state programs did not meet the two-thirds goal. Since the 1972 report of the National Commission on State Workmen's Compen- sation Laws, mostjurisdictions have done so, but the payments are also subject to caps for workers with incomes above the state average. For example, 1 5jurisdictions have maximum payments for temporary total disability between $200 and $300 per week; temporary total disability benefits in these states (which are exempt from federal income tax) are equivalent to an annual take-home pay of about $13,000 to $18,000. If covered by state workers' compensation programs, railroad workers, who had an average annual wage of about $42,000 in 1991, could expect to receive disability benefits equivalent to only 30 to 40 percent of wages in these jurisdictions. They would receive two-thirds of their wages for temporary total disability in only seven states (see Figure 5-6). Payments for permanent total disability correspond fairly closely to those for temporary total disability, but there are important exceptions. A few jurisdictions place limits on the duration or the amount of total disability payments, sometimes on both. Thus, not only would perma- nently and totally disabled workers with higher-than-average wages receive less than two-thirds of their wage in most jurisdictions, but these benefits would also be limited in amount or duration in a fewjurisdictions. Payment for permanent partial disability is more complex. In some jurisdictions the benefits are not related to wage loss; instead the worker receives a benefit based on the degree of impairment. Predict- ing how an injured worker would fare under this system depends in large part on how much the worker's impairment reduces earning potential. If able to return to work at close to the preinjury wage, the employee may receive benefits that exceed two-thirds of lost wages. If unable to approach preinjury wages after a return to work, however, the worker may receive considerably less than two-thirds of preinjury wages, particularly in states that limit the amount or duration of bene- fits. In jurisdictions that base permanent partial disability benefits on methods that weigh wage loss or potential wage loss to varying degrees, workers tend to receive two-thirds of wages, but again subject to max- imum payments that disadvantage workers who earn higher-than- average wages. The federal programs, FECA and LHWCA, provide temporary and permanent disability benefits that are similar to those of the state programs, but the potential payments are much higher and the programs are also nationwide in scope. If railroad workers were covered by these programs, they would be more likely to have two- thirds of lost wages replaced. These federal programs, however, are also more costly because of the higher benefit levels offered.

114 • Compensating Injured Railroad Workers Under FELA Georgia Mississippi Arkansas Oklahoma Nebraska Kansas Tennessee Louisiana New Mexico South Dakota Delaware Arizona Idaho Indiana North Dakota California Montana New York Utah South Carolina Kentucky Alabama Wyoming West Virginia Colorado Florida New Jersey Virginia Nevada Rhode Island North Carolina Oregon Missouri Washington H awn ii Texas Michigan Ohio Minnesota Wisconsin Pennsylvania Maryland Massachusetts Maine Vermont D.C. New Hampshire Illinois Alaska Iowa Connecticut Two Thirds / 20 40 60 80 100 Percentage FIGURE 5-6 Percentage of average railroad worker weekly wage cov- ered under temporary total disability, 1991. Unlike FELA, workers' compensation programs make explicit al- lowance for medical and vocational rehabilitation. The benefits pro- vided, however, are fairly modest. Little systematic information is available across programs, but data from Florida are illustrative. In 1985, vocational benefits (including both tuition and wage loss) for on- the-job training and classroom education averaged $14,000.

State and Federal Workers' Compensation Programs • 115 Workers' compensation was also meant to reduce the litigiousness of the tort system by an administrative delivery of needed benefits. The administrative system appears to work well for the majority of workers' compensation claims, which tend to be for relatively minor injuries that involve only medical expenses or a relatively short absence from work. Information available from six states indicates that between 70 and 90 percent of lost-time claims, depending on the state, proceed through the system without controversy. Many of the disputed cases are re- solved before any formal administrative hearing is held, but between 2 and 13 percent of indemnity claims require a formal hearing at which workers are usually represented by attorneys. Formal administrative hearings and appeals typically operate like civil court proceedings, though with somewhat less formality. Nonetheless, this process is cum- bersome and expensive. Although almost all disputed claims are resolved in the administra- tive processes of workers' compensation programs, in all programs except FECA, a dissatisfied party can appeal to a civil court. Although such appeals occur with regularity in most states, they represent less than 1 percent (in some states considerably less) of all lost-time claims. The litigiousness of workers' compensation appears to be growing but is far from rampant. Attorney involvement, for example, occurs in 14 percent of indemnity cases in Georgia and in 22 percent of indem- nity cases in New York, but it is believed to occur far less frequently in Wisconsin. Although data are sparse, attorney fees do not appear to consume a large share of total compensation in most states. Plaintiff attorney fees represented 6 percent of total 1990 indemnity claims in Georgia and less than 1.5 percent of total 1987 claims in New York. Workers' compensation in California appears to be much more af- fected by litigiousness than programs in other states: roughly 45 per- cent of indemnity cases involved litigation in 1991 and total litigation costs equaled about one-fourth of total benefits. The wide disparity in state program benefits has led supporters of FELA to argue that it would be inequitable for a nationwide industry like the railroads to be covered by state programs. Such coverage could lead to widely divergent benefits for employees with the same injury depending on the state in which they were domiciled. Trucking firms, major competitors of the railroads, operate within a state-by- state system, albeit not without complaint about the high costs it im- poses on their operations. An alternative to the state-by-state approach would be the national scope of the federally administered programs or a new nationwide program specifically for the railroad industry.'5

116 • Compensating Injured Railroad Workers Under FELA NOTES This section draws heavily from the work by Welch (1989). The federal government also administers a program to pay benefits for miners with black lung disease. Because of the special nature of this program, it is excluded from the discussion in this chapter. In addition to harbor, longshore, and other maritime workers, coverage under the act was extended to include employees in the District of Colum- bia (until the District established its own workers' compensation program in 1982) and was later extended to additional employees as provided in the Defense Base Act of 1941, Nonappropriated Fund Instrumentalities Act of 1952, and the Outer Continental Act of 1953. This is not true of many other countries' workers' compensation systems, for example, those of the United Kingdom and Australia. Three states have special provisions that limit coverage. For example, in Arkansas employer liability ceases 6 months after the injury when no time has been lost, or 6 months after the employee returns to work, or after a maximum of $10,000 (DOL 1992, Table 5a). By contrast, FELA may provide coverage from the first day of lost income. The data compiled by DOL include Puerto Rico and the Virgin Islands; the tabulations presented in this report include only the 50 states and the District of Columbia. Manyjurisdictions also set minimum payments, but these benefits are not relevant in considering how railroad workers might be affected by a state workers' compensation program. This figure includes payments by all state and federal programs. The discussion of permanent partial disability draws heavily on Berkowitz and Burton's 1987 study of 10 state programs. Workers in Massachusetts would also be eligible for a wage loss benefit, which reduces the disparity in benefits between the two states but does not eliminate it. Connecticut has reduced its scheduled benefits since these state data were published. There is some movement toward relying more heavily on standard, medi- cally based rating systems, such as the American Medical Association (AMA) Guides to the Evaluation of Permanent Impairment (AMA 1990). Re- liance on AMA Guides appears to narrow the range of dispute about the degree of impairment (Boden 1992). Under FELA, the injured employee may need to cooperate with rehabilita- tion to qualify for income maintenance programs of the railroad, but vocational rehabilitation payments in the FELA settlement are not contin- gent on pursuing treatment or training. Wisconsin has two permanent partial disability systems. If an injured worker returns to his or her previous job at 85 percent of prior wages, the level of disability is determined by a physician. In this system there is relatively little dispute or litigation. If an injured worker returns to work at less than 85 percent of his or her prior wages, disability is determined by loss of wage-earning capacity. Litigation in this latter system is similar to that in Maryland and New Jersey (Boden 1988).

Stale and Federal Workers' Compensation Programs • 117 15. A current problem in workers' compensation is one concerning the "choice of law," wherein claims can be filed in several jurisdictions. Disputes over the choice of law lead to litigation. REFERENCES ABBREVIATIONS AMA American Medical Association DOL U.S. Department of Labor AMA. 1990. Guides to the Evaluation of Permanent Impairment, 3rd ed. (revised). Chicago, Ill. Ballantyne, D., and S. Eccleston. 1992. Workers' Compensation in Georgia: Admin- utrative Inventory. Workers' Compensation Research Institute, Cambridge, Mass. Ballantyne, D., and C. Telles. 1991a. Workers' Compensation in Pennsylvania: Admini.strative Inventory. Workers' Compensation Research Institute, Cam- bridge, Mass. Ballantyne, D., and C. Telles. 1991b. Workers' Compensation inMinnesota:Admin- i.strative Inventory. Workers' Compensation Research Institute, Cambridge, Mass. Ballantyne, D., and C. Telles. 1992a. Workers' Compensation in Wisconsin: Admin- istrative Inventory. Workers' Compensation Research Institute, Cambridge, Mass. Ballantyne, D., and C. Telles. 1992b. Workers' Compensation inNew York:Admini.s- trative Inventory. Workers' Compensation Research Institute, Cambridge, Mass. Barth, P., and C. Telles. 1992. Workers' Compensation in California: Administrative Inventory. Workers' Compensation Research Institute, Cambridge, Mass. Berkowitz, M. 1990. Should Rehabilitation Be Mandatory in Workers' Com- pensation Programs? Journal of Disability Policy Studies, Vol. 1, No. 1, pp. 64-80. Berkowitz, M. 1991. The U.S. Worker Rehabilitation Program. Monthly Labor Review, Vol. 114, No. 9. pp. 15-21. Berkowitz, M., and E. Berkowitz. 1991. Rehabilitation in the Work Injury Program. Rehabilitation Counseling Bulletin, Vol. 34, No. 3, pp. 182-196. Berkowitz, M., and J. Burton. 1987. Permanent Disability Benefits in Workers' Compensation. Upjohn Institute for Employment Research, Kalamazoo, Mich. Boden, L. 1988. Reducing Litigation: Evidence from Wisconsin. Workers' Compen- sation Research Institute, Cambridge, Mass. Boden, L. 1992. The AMA Guides in Maryland: An Assessment. Workers' Compen- sation Research Institute, Cambridge, Mass. Boden, L., et al. 1991. Reducing Litigation: Using Disability Guidelines and State Evaluators in Oregon. Workers' Compensation Research Institute, Cambridge, Mass.

118 • Compensating Injured Railroad Workers Under FELA Boden, L., and C. Fleishman. 1989. Medical Costs in Workers' Compensation: Trends and Interstate Comparisons. Workers' Compensation Research Institute, Cambridge, Mass. Burton, J. 1993. Workers' Compensation Costs, 1960-1992. John Burton's Workers' Compensation Monitor, MarchlApril. DOL. 1986. Longshore and Harbor Workers' Compensation Act: Annual Report of the Act during Fiscal Year 1985. DOL. 1992. Longshore and Harbor Workers' Compensation Act: Annual Report of the Act during Fiscal Year 1991. Gardner, J. 1988. Appropriateness and Effectiveness of Vocational Rehabilitation in Florida: Costs, Referrals, Services and Outcomes. Workers' Compensation Re- search Institute, Cambridge, Mass. Morrison, J. 1990. Medical Cost Containment for Workers' Compensation. Journal of Risk and Insurance, Vol. 57, No. 4, p. 646-653. Nelson, D. 1988. A Practical Approach to Workers' Compensation. American Family Physician, Vol. 37, No. 4, pp. 233-238. Nelson, W. 1992. Workers' Compensation: 1984-1988 Benchmark Revisions. Social Security Bulletin, Vol. 55, No. 3, Fall, pp. 4 1-58. Nordlund, W. 1991. The Federal Employees' Compensation Act. Monthly Labor Review, Sept., pp. 3-14. Pease, S. 1989. Workers' Compensation in Washington: Administrative Inventory. Workers' Compensation Research Institute, Cambridge, Mass. Schwartz, G. 1993. Waste, Fraud, and Abuse in Workers' Compensation: The Recent California Experience. Maryland Law Review, Vol. 52, No.4. Tab, S., et al. 1989. Effects of Active and Completed Litigation on Treatment Results: Workers' Compensation Patients Compared with Other Litigation Patients.Journal of Occupational Medicine, Vol. 31, No.3, March, pp.265-269. Walsh, N.,and D. Dumitru. 1988. The Influence of Compensation on Recovery from Low Back Pain. Occupational Medicine: State of the Art Reviews, Vol. 31, No. 1, pp. 109-121. Weiler, P. 1989. Workers' Compensation and Product Liability: The Interaction of a Tort and Non-Tort Regime. Ohio State Law Journal, Vol. 50, No. 4, pp. 825-854. Welch, E. 1989. Fundamentals of Workers' Compensation. Edward Welch, East Lansing, Mich.

Comparison of FELA Process and Workers' Compensation Systems Injury compensation under the Federal Employer's Liability Act of 1908 (FELA) results in higher benefits than those of the alternative workers' compensation systems. From comparisons made in this chap- ter, indemnity payments to injured railroad workers can be 2 to 10 times the level of benefits from state workers' compensation systems, depending on the injury and the state.' Medical payments under FELA, however, are comparable to or below those in state workers' compensa- tion systems. In this chapter information is developed on the costs of railroad benefits based on data provided by four Class I carriers, and a compari- son is made with data for several state workers' compensation systems and some national summary data on workers' compensation. Directly comparable data are not available for the railroads and workers' com- pensation systems because of differences in data collection and categor- ization. The study committee employed a multipronged approach by developing a number of comparisons using what data were available or could be produced. Because railroads self-insure, the main source of data on compensation is the railroads themselves. After a review of data issues and sources, the discussion turns to levels of medical and indemnity payments for railroad workers. (See Appendix B for definitions of terminology used in this chapter.) In the 119

120 • Compensating Injured Railroad Workers Under FELA first of four comparisons, these payments are assessed against those under workers' compensation. Next, detailed data from one railroad on benefit payments by injured body part are compared with similar data from a national sample provided by the National Council on Compen- sation Insurance (NCCI) that covers 13 state workers' compensation systems. A third comparison is based on case studies that estimated the benefits that a sample of actual FELA claimants might receive if they were compensated under alternative workers' compensation systems. The final benefit comparison is based on data supplied by the New York Metropolitan Transportation Authority (MTA). MTA payments to commuter rail employees covered by FELA are compared with payments to transit workers covered by workers' compensation. The chapter concludes with a consideration of transactions costs. DATA PROBLEMS AND SOURCES Assessing the relative costs and benefits provided by FELA and how they compare with those of other systems is difficult because of the lack of comparable data. Few of the comparisons in this chapter provide definitive answers concerning the relative costs of FELA and workers' compensaton systems. Part of the difficulty can be attributed to the fact that railroads have no reason to define and track injuries and payments according to the same categories used in workers' compensation (temporary total dis- ability, permanent partial disability, and permanent total disability). Nearly all FELA settlements are lump sums, and thus continued mon- itoring of claims after settlement is unnecessary. Whether the injury is temporary or permanent or partial or total is important in the settle- ment negotiations, but there is no need for an official classification of each injury. One of the railroads that provided data for this study only began to make these designations in its claims files 2 years ago. Another difficulty involves deriving workers' compensation data on a per-claim basis. Workers' compensation data are for the most part based on yearly payouts for benefits for all claims active in the benefit year. Since the cost of a claim can be spread over a number of years under workers' compensation, there is no one-to-one match between the number of claims and their payments and the ultimate payout per claim. Deriving the total benefits an individual receives for a claim requires waiting until the claim has been closed or ultimate costs have been estimated. As lump-sum payouts, FELA data reflect the total cost for each claim closed; thus, the. cost per claim can be readily calculated. Four railroads agreed to share detailed FELA data with the study committee: Burlington Northern, CSX Transportation, Norfolk South-

Comparison of FELA and Workers' Compensation • 121 em, and Union Pacific. Although data from only four railroads are used, these railroads represent 60 percent of Class I railroad em- ployees and over half of all freight railroad employees. Information on indemnity payments and the time to resolve claims was provided by all four railroads, three were able to supply information on the cost of outside legal counsel, and two had data on the medical payments for their claims. Workers' compensation data came from a number of sources. The Workers' Compensation Research Institute (WCRI) has conducted a number of state studies from which information was obtained about New York, Wisconsin, Pennsylvania, Georgia, Minnesota, and Wash- ington. NCCI provided sample data for payouts on an injury basis and its Annual StatisticalBulletin (NCCI 1992) provided additional data. The New York MTA provided comparison data for its subway operations and its commuter railroad operations on the Metro North Commuter Railroad and the Long Island Railroad. LEVELS OF BENEFITS The level of medical and indemnity benefits for injured railroad workers is a primary concern for the workers, and the cost of those benefits is a major concern of the railroads. In this section the levels of these benefits for the railroads are examined. In the next section a compari- son is made with similar data from the Longshore and Harbor Workers Compensation Act of 1972 (LHWCA) program and state workers' compensation systems to determine whether and to what extent rail- road benefits paid per claim are higher (or lower) than those of alterna- tive workers' compensation systems. Data Definitions Benefit data were supplied by the railroads on an annual basis and based on "close year" statistics; that is, the injury data reported for the year are based on the claims settled that year regardless of when the injury occurred. Injury payments for 1991 include injuries that oc- curred in 1991, 1990, 1989, 1988, and so on. The data were broken out by the settlement status of the claims and the length of time between the injury and the settlement. This information permitted an examination of the delays and litigation connected with FELA settlements, which are considered in the section on transactions costs at the end of this chapter. The settlement status of claims is categorized as direct, represented, or litigated. Benefits include both medical payments and indemnity payments, and claims are divided between those that are settled with

122 • Compensating Injured Railroad Workers Under FELA medical payments only and those for which an indemnity payment is also made. Medical Payments Payments per claim for claims settled with medical payments only have trended upward from 1987 to 1991, with a downturn in 1991. On a constant-dollar basis, they have remained fairly steady except for a spike in 1990 (Table 61).2 These claims are dominated by low-cost direct claims, which constitute 96 percent during the 5-year period. Medical payments for claims settled with indemnity payments have risen steadily and significantly during this period, even on a constant- dollar basis. This steady increase is reflected in all categories as well as in the total per claim. By 1991, medical costs per claim had increased 98 percent over 1987 levels. On a constant-dollar basis, the increase is 46 percent during this period or about 10 percent per year. There is a direct and positive relationship between the level of medical payments and the time to resolve the claim for claims settled with indemnity payments (see Figure 6-1). There is not much differ- ence among direct, represented, and litigated claims until the 2-year mark. Presumably, less-severe cases (if medical expenses are a good indicator of severity) tend to be settled early regardless of whether the employee is represented by an attorney. Indemnity Payments The cost of FELA indemnity payments per claim has risen steadily since 1987, even after accounting for inflation (Table 6-2). The increase has occurred for the most part because of larger average settlements for direct and litigated claims, whereas represented claims have re- mained fairly steady. The increase during this period in payments per claim was 42 percent, though over half of this can be accounted for by inflation. On a constant-dollar basis, the increase was about 18 percent or about 4.2 percent per year. From the railroads' standpoint, real increases in the average cost of both medical and indemnity benefits may be of particular concern. The increase in the real medical payments per claim in combination with this increase in real indemnity payments might be explained by an increase in the severity of injuries being compensated, but there is no other evidence to indicate whether this is occurring. Up to a point, there is a direct and positive relationship between the level of indemnity payments and the time to resolve the claim, just as there is for medical payments (Figure 6-2). For each settlement cate-

TABLE 6-1 Medical Payments per Claim for Two Railroads TYPE OF CLAIM 1987 1988 1989 1990 1991 Claims settled with medical payments only ($) Direct 118 128 129 145 147 Represented 1,145 741 1,468 639 135 Litigated 1,188 1,355 2,663 2,347 1,571 Total 142 160 180 233 187 Total (constant dollars) 109 115 121 143 106 Claims settled with indemnity payments ($) Direct 1,619 1,882 2,358 2,932 3,496 Represented 3,771 4,894 6,886 7,430 8,204 Litigated 6,642 8,013 9,024 9,542 11,674 Total 2,782 3,348 3,828 4,501 5,522 Total (constant dollars) 2,138 2,416 2,564 2,765 3,120 All claims ($) Direct 708 905 1,155 1,464 1,747 Represented 3,513 4,279 6,282 6,080 6,124 Litigated - 6,198 7,474 8,560 8,529 10,604 Total 1,377 1,803 2,090 2,513 3,092 Total (constant dollars) 1,058 1,301 1,400 1,544 1,747 CP1-U-MC (1982-1984 = 100) 1.30 1.39 1.49 1.63 1.77 NoTE: The constant-dollar deflator is the medical care component of the consumer price index for urban consumers, abbreviated here as CPI-U-MC. SouRcE: Economic Report of the President, 1992, and data supplied by the railroads.

124 • Compensating Injured Railroad Workers Under FELA 8,000 E co 6,000 ci) a- Cl) - 4,000 0 0 C .2 2,000 Cl) C 0 0 1 2 3 4 5 6 8 10 12 14 16 18 20 22 24 26 28 31 34 37 41 45>49 Months Between Injury and Settlement Direct ± Represented Litigated FIGURE 6-1 Constant-dollar medical payments per claim by time to settlement for claims with indemnity payments. gory, however, the real settlement per claim reaches a plateau at about 16 months for direct and represented claims and at about 20 months for litigated claims. COMPARISON OF RAILROAD AND WORKERS' COMPENSATION BENEFITS As noted earlier, WCRI has conducted a number of state studies that provide some data for both medical and indemnity payments under, workers' compensation that can be compared with FELA data. NCCI also publishes state data on medical payments and indemnity payments based on disability level that can be used for selected comparisons. WCRI Data Railroad medical payments under FELA are very comparable with the workers' compensation payments of the three states examined in Table 6-3. For claims settled with medical payments only, the railroad pay- ments are in fact lower than those shown in the WCRI state data.3 For claims settled with indemnity payments, the railroad medical payments

TABLE 6-2 Indemnity Payments per Claim for Four Railroads TYPE OF CLAIM 1987 1988 1989 1990 1991 Claims settled with indemnity payments ($) Direct 15,176 18,210 20,243 22,887 24,594 Represented 81,215 72,306 85,410 81,122 83,379 Litigated 128,771 134,220 145,002 156,742 166,577 Total 37,900 42,055 43,941 49,008 53,770 Total (constant dollars) 33,328 35,534 35,427 37,485 39,457 All claims ($) Direct 7,689 9,847 11,329 12,671 13,150 Represented 78,392 65,671 78,950 71,047 70,848 Litigated 121,506 127,377 138,732 146,057 156,321 Total 21,733 25,321 26,990 29,830 31,992 Total (constant dollars) 19,112 21,395 21,761 22,816 23,476 CPI-U (1982-1984 = 100) 1.14 1.18 1.24 1.31 1.36 NOTE: The constant-dollar deflator is the consumer price index for urban consumers, abbreviated here as CPI-U. SOURCE: Economic Report of the President, 1992, and data supplied by the Burlington Northern, CSX Transportation, Norfolk Southern, and Union Pacific.

126 • Compensating Injured Railroad Workers Under FELA 140,000 E 120,000 0 100,000 a- 22 80,000 Cz CO 40,000 20,000 0 1 2 3 4 5 6 8 10 12 14 16 18 20 22 24 26 28 31 34 37 41 45>49 Months Between Injury and Settlement Direct ± Represented ' Litigated FIGURE 6-2 Constant-dollar indemnity payments per claim by time to settlement for claims with indemnity payments. are comparable with those of the New York and Wisconsin workers' compensation programs and significantly lower than those of Pennsyl- vania. Although state data are available for only a few years, the in- crease in state payments is about the same as that for the two railroads.4 Both the railroads and the state systems appear to be similarly affected by the forces driving these increases. The indemnity payment data, however, provide a different story. The railroad payments per claim for claims settled with indemnity payments range from nearly 12 times the Wisconsin average in 1987 to about 4 times the state averages in several years. These differences reflect a variety of factors: broader coverage in the FELA payments, including pain and suffering; the higher wage levels of railroad workers compared with average state levels; benefit caps in workers' compensa- tion systems; lower benefits in Wisconsin than in the other two states; and possibly a difference in the average severity of railroad injuries. NCCI Data NCCI reports information on the average benefit payment per claim by injury type in its Annual StatisticalBulletin (NCCI 1992). The 1992 data

Comparison of FELA and Workers' Compensation • 127 TABLE 6-3 Payments per Claim to Injured Workers: Comparison Between Selected Railroads and Selected State Workers' Compensation TYPE OF CLAIM 1986 1987 1988 Claims settled with medical payments only ($) Two railroads 152 142 160 New York 184 204 228 Wisconsin 167 185 NA Pennsylvania 223 249 NA Claims settled with indemnity payments ($) Medical payments Two railroads 2,350 2,782 3,348 New York 2,267 2,246 3,169 Wisconsin 2,333 2,485 NA Pennsylvania 4,370 5,049 NA Indemnity payments Two railroads 34,231 42,230 50,341 Four railroad? 29,509 37,900 42,055 New York 5,416 5,255 7,131 Wisconsin 3,459 3,291 NA Pennsylvania 8,187 9,080 NA Total Two railroads 36,581 45,012 53,689 New York 7,683 7,501 10,300 Wisconsin 5,792 5,776 NA Pennsylvania 12,557 14,129 NA NOTE: NA = not available. OnIy three railroads rather than four for indemnity payments in 1986. SOURCE: Railroad data supplied by the Burlington Northern, CSX Transportation, Norfolk Southern, and Union Pacific. State data from Ballantyne and Telles (1991b); Ballantyne and Telles (1992a, 1992b). cover policy periods ending in 1987, 1988, or 1989, depending on the state, and are based on estimated total cost of the claims arising in those years. Therefore, they are comparable with railroad "close-year" data. The medical payments per claim by the railroads from Table 6-3 nearly match the lowest state levels in the NCCI data (Table 6-4). For claims settled with medical payments only, the railroad payment per claim of $160 in 1988 is less than that of the state with the lowest amount, Rhode Island; it is well below the state average of $270 and is less than 50 percent of the LHWCA amount. For claims settled with

TABLE 6-4 Payment per Claim by Injury Severity for Selected Workers' Compensation Systems Low HIGH LHWCA AVERAGE PAYMENT ($) PAYMENT ($) PAYMENT ($) PAYMENT ($) Claims Settled with Medical Payments 170 (Rhode Island) 435 (Hawaii) 340 270 Claims Settled with Indemnity Payments Medical payments 3,077 (Missouri) 11,450 (New Mexico) 8,655 5,568 Temporary total disability 876 (California) 5,188 (Massachusetts) 2,272 1,858 Permanent partial disability 9,461 (Missouri) 104,798 (Maine) 59,034 31,325 Permanent total disability 50,417 (Utah) 734,251 (District of Columbia) 247,852 210,527 Fatalities 34,845 (Mississippi) 355,076 (Massachusetts) 510,134 147,608 NoTE: LHWCA = Longshore and Harbor Workers' Compensation Act. The low and high average benefits paid are based on closed claims for policy periods that vary among states and that end between 1987 and 1989. SOURCE: Data from NCCI (1992, Exhibit XI).

Comparison of FELA and Workers' Compensation • 129 indemnity payments, the railroad medical payments per claim of $3,348 v are again closer to the lowest state amount than the average and are only about 40 percent and 30 percent of the LHWCA and high state amounts, respectively. It is not possible to make exact comparisons for indemnity payments. The NCCI data on indemnity payments are based on three injury categories: temporary total disability (TTD), permanent partial dis- ability (PPD), permanent total disability (PTD). As noted earlier, the railroads do not classify their claims into these categories. Although railroad direct claims tend to be mostly TTD claims and litigated claims to be more PPD and PTD claims, there is no one-to-one match. The railroad indemnity payments per claim in 1989 were $20,243 for direct claims, $85,410 for represented claims, and $145,002 for litigated claims (Table 6-2). Although direct payments are much greater than TTD payments, even for the high-average-benefit state of Massa- chusetts, represented payments fall well within the state ranges for PPD payments and litigated payments fall within the range for PTD payments. Because of this dearth of comparable data, the study committee worked with one railroad to develop more-detailed information on FELA injury cost for comparison with information on workers' com- pensation costs. This information took two forms. One was a break- down of all injury claims for 5 years based on the body part affected for comparison with similar data available from NCCI. The second was a sample of claims from 1991 used in three case studies that estimated the payments the claimants in the sample might have received under alternative compensation systems. These comparisons are described in the following sections. COMPARISON OF FELA AND NCCI DATA One railroad provided a breakdown of its claims by body part for 1987 through 1991 that permits comparison with a sample of workers' com- pensation claims provided by NCCI. The NCCI sample, from the Call for Detailed Claim Information (DCI), is a random sample of claims with lost-time payments based on the part of the body injured from a consistent group of 13 states; it includes injuries to the head, neck, back, trunk, upper extremities, and lower extremities. Both medical and injury payments are included. Finally, the NCCI sample includes claims closed in the year, which means that all payments have been completed. Thus, the claims data cover injuries that may have occurred over a number of previous years and are comparable with the railroad close-year data.

130 • Compensating Injured Railroad Workers Under FELA On the basis of these data, FELA and workers' compensation pay- ments appear more comparable than in the previous section. Total payments per claim average only about 25 percent more for FELA than for the NCCI sample (Table 6-5). The total per claim masks a signifi- cant disparity between FELA and NCCI in the composition of the benefits, however. FELA indemnity payments average about 70 per- cent higher than those in the NCCI data over the 1987-199 1 period, whereas FELA medical benefits are only about 25 percent of those in the NCCI data. The data suggest that the NCCI sample includes injuries that are more severe than those included in the state averages of Table 6-4. As TABLE 6-5 FELA Benefits per Claim as Percentage of NCCI Benefits per Claim by Injured Body Part INJURED BODY PART 1987 1988 1989 1990 1991 AVERAGE Head Indemnity payment 80 76 53 99 67 75 Medical payment 12 9 10 9 10 10 Total 54 52 37 68 48 52 Neck Indemnity payment 131 149 244 90 132 149 Medical payment 21 20 40 16 12 22 Total 96 104 169 64 88 104 Back Indemnity payment 214 241 208 217 154 207 Medical payment 34 40 34 39 26 35 Total 157 179 157 162 113 154 Trunk Indemnity payment 135 191 206 164 160 171 Medical payment 15 18 25 25 29 22 Total 91 128 139 113 107 115 Upper extremities Indemnity payment 100 127 140 84 79 106 Medical payment 11 18 15 21 16 16 Total 70 88 96 62 56 74 Lower extremities Indemnity payment 194 176 183 155 110 164 Medical payment 20 19 29 25 23 23 Total 132 118 126 108 86 114 All injuries Indemnity payment 171 195 192 171 133 172 Medical payment 22 26 28 29 23 26 Total 120 137 136 123 98 123 SOURCE: Data supplied by NCCI and one railroad.

Comparison of FELA and Workers' Compensation • 131 noted in Chapter 5, the cost share of medical benefit payments in workers' compensation systems is typically in the range of 30 to 40 percent,5 and the medical shares for the NCCI sample do range be- tween 31 and 36 percent of the benefit payments. However, the medical payment per claim for 1989 is nearly $18,000 (Table 6-6), which is about 3 times the average of the medical payments in the state data (reported in Table 6-4), whereas the railroad's medical payment per claim in 1989, $5,024, is about equal to the state average medical payment in Table 6-4. For indemnity payments, the average FELA benefits range from a low of 75 percent of workers' compensation for head injuries to 207 percent for back injuries (Table 6-5). The actual percentages for each category have tended to fluctuate significantly over the years, but in the last 2 years, the benefits per claim under FELA ranged from 67 percent to 217 percent of the workers' compensation benefit. In summary, the NCCI DCI sample supports the conclusion that FELA benefits exceed workers' compensation benefits. Because the difference in injury severity, if any, is unknown, definitive percentage comparisons between the two are uncertain. CASE STUDY COMPARISON The lack of comparable data for FELA and workers' compensation systems led the study committee to commission comparisons of a sam- ple of one railroad's FELA claims and those of various workers' com- pensation systems. The approach was to estimate the benefits the injured worker would receive under the regulations and benefits of the alternative system, given the injury and wage loss, and compare them with those the worker actually received under FELA. Two state workers' compensation systems, those of Washington and Michigan, and a na- tional workers' compensation system, LHWCA, were chosen for the case studies. Officials with the Washington State workers' compensation system, which operates a state fund, agreed to estimate the sample for the study committee under their benefit regulations. The Michigan estimate was producedjointly by a retired director of Michigan's Bureau of Workers' Disability Compensation, a retired district claims manager of an insur- ance company providing coverage in Michigan, and a study committee member familiar with Michigan's system. The estimate of costs under the provisions of LHWCA was performed by a retired insurance com- pany manager who had been responsible for the company's LHWCA policies and claims.

TABLE 6-6 Comparison of NCCI Payment Data with Data from One Class I Railroad: Claims Settled with Indemnity TYPE OF CLAIM 1987 1988 1989 1990 1991 Railroad Number of claims 1,428 1,604 1,322 1,148 1,038 Indemnity payment per claim ($) 42,083 58,216 66,591 75,990 85,613 Medical payment per claim ($) 2,862 41011 5,024 6,516 7,049 NCCI Number of claims 38,307 39,822 41,481 40,130 22,535 Indemnity payment per claim ($) 24,668 29,917 34,754 44,462 64,431 Medical payment per claim ($) 12,757 15,538 17,733 22,558 30,427 SOURCE: Data supplied by NCCI and one railroad.

Comparison of FELA and Workers' Compensation • 133 The benefit levels by type of injury reported by NCCI show the relative levels of Michigan and LHWCA benefits among workers' com- pensation systems (NCCI 1992, 292). Michigan's TTD, PPD, PTD, and fatality benefits are 135 percent, 193 percent, 79 percent, and 55 per- cent, respectively, of the average for the systems reported. The same benefits under LHWCA are 122 percent, 188 percent, 118 percent, and 346 percent. Sample and Assumptions The sample universe consisted of all claims closed in 1991 for which indemnity payments were made for a single Class I railroad. These claims numbered 1,073 and a stratified, random sample was taken from them. The total claims were divided into four groups, specifically, return-to-work claims with fewer than 90 work days lost and with more than 90 work days lost, no-return-to-work claims, and fatalities. These groups were created to approximate the workers' compensation cate- gories of TTD, PPD, PTD, and fatalities. The first group is assumed to include mostly TTD claims, the second to have a higher probability of picking up the PPD claims, and the third as being more likely to include the PTD claims. As noted previously, the railroads have no reason to define and track injuries according to these categories. The sample consisted of 38 claims: 16 in the group return-to-work with fewer than 90 work days lost; 8 in the group return-to-work with more than 90 work days lost; 10 in the no-return-to-work group; and 4 fatality claims, which is all of the fatal injuries that occurred in 1991. The other samples represent approximately 3 percent of each group. Limitations of Approach The fundamental limitation with this approach was the lack of detailed information concerning the extent and severity of the injuries and information on any lingering impairment. The estimators had to rely on broad categories of injuries such as bruise/contusion, strain, lacera- tion, and so on, with a short description of how the injury occurred, such as "from pulling a switch," "slipped on ballast," "lifting a non- railroad object," and with accident severity information limited to "treated and released," "disabled-TTD," and "hospitalized." The esti- mators had to use additional indicators of severity such as the number of days lost from work, the cost of medical treatment, and occasionally an indication of bodily impairment or disability.6 Unfortunately, these secondary indicators sometimes conflict and do not give a clear indication of the result. To maintain consistency, the

134 • Compensating Injured Railroad Workers Under FELA assumptions used in the Washington State estimates concerning the severity of the injuries were made available to the estimators for the other two systems, who followed the first estimator's assumptions fairly closely and only differed where the compensation approach of their system required it. Nevertheless, considerable uncertainty exists concerning the seri- ousness of the injuries and the resulting implications for scheduled payments for impairment, potential effects on wage-earning capacity, and likelihood of being disabled for work. Therefore, it should be understood that the estimates of benefits for these claims under workers' compensation systems are rough. Other problems with making estimates based on this sample in- volved a lack of information in the files about the injured employee. Rarely were there entries for marital status and number of dependents. (These data are not always necessary to settle FELA claims and their absence is not an indication that the individuals are single.) The omis- sion of this information was particularly problematic for Washington estimates because in that state, the percentage of wage replacement depends on the injured worker's marital status and number of depend- ents, ranging from 60 percent for a single worker with no dependents to 75 percent for a married worker with five or more dependents. The estimates were based on the assumption of no dependents, so Washing- ton estimates may be somewhat lower than they actually would be if this information had been known for each claim. Where there were entries for spouse and dependents in the sample, no information was provided about ages. The estimators therefore relied on their systems' formulas for computing pension reserves and made the most generous assumptions for the victim's compensation so as not to underestimate the benefit. The result is that the pension reserves might be somewhat high for the fatality claims. Finally, the indemnity payments are reported as the total estimate of all benefits that the worker would receive (wage loss, scheduled pay- ments, vocational benefits, etc.) since the railroad payments include all such benefits. Actual wage loss in each case is also reported for compar- ison with the injury payments received. Results Payments for FELA claims exceeded the estimates of benefits for the three alternative systems by a wide margin in almost every case. The closest estimate to FELA payment for any of the nonfatal claims is the Michigan estimate for the no-return-to-work claims, in which the aver- age settlement is 61 percent of the FELA payment. For the remainder

Companson of FELA and Workers' Compensation • 135 of the categories, the percentage of the FELA payment covered by any system is less than 40 percent. Overall, the FELA indemnity benefits ranged from about 1.5 to 11 times the workers' compensation estimates (Table 6-7). The greatest disparity between the FELA payments and the state estimates was for the return-to-work claims with fewer than 90 work days lost. Estimates of the percentage of the FELA settlement that would be paid in each of the systems are 9 for Michigan, 11 for Washington State, and 23 for LHWCA. This category of claims, how- ever, is greatly affected by the wide variation in how some of the claims would be compensated. For Washington, Michigan, and LHWCA, the compensation in this subsample is principally for wage loss, because the injuries are mostly minor, and for six of the claims in the subsample there was no wage loss. FELA payments for these no-wage-loss claims are clearly compensation for nonwage losses. It is not surprising, there- fore, that there might be the greatest disparity in compensation in this category. For the return-to-work claims with more than 90 work days lost, the results are more uniform. Injury compensation under FELA is, for the most part, in the range of 3 to 6 times the workers' compensation estimates for each claim, and the average of the subsamples is 2.5 to 4 times the estimates. Wage replacement for the other systems, on average, is about 100 percent of wages. For the Michigan claims, however, the level of wage replacement exceeds 60 percent in only one instance. For Washington and LHWCA, the results are more in line with the average levels. FELA benefit payments for the no-return--to-work claims are on average 1.5 to 3 times the level of the estimates for the workers' com- pensation systems. The no-return-to-work claims include more serious injuries, and their likely permanent effects lead to uniformly higher levels of wage replacement in all the claims. FELA payments average about 6 times wage loss, LHWCA and Washington benefits are about twice the wage loss, and Michigan benefits are nearly 4 times the wage loss. These premiums over wage loss in the workers' compensation estimates represent compensation for impairments and expected fu- ture wage loss in most claims. The low number of fatalities in the year makes it difficult to general- ize about the level of benefits for these claims. The benefits for fatality claims depend on the circumstances of the death and the regulations of the system. One claim is for a work-related fall by a young individual. The other three claims are heart attacks on the job. Two of the heart attacks do not appear from the records to be associated with work activity. They are not compensable under the state systems or FELA

TABLE 6-7 Case Study Comparisons of Actual FELA Indemnity Settlements with Estimates for Selected Workers' Compensation Systems FELA LHWCA MICHIGAN WASHINGTON Return-to-work (RTW) claims with fewer than 90 work days lost (N = 16) Indemnity payments per claim ($) 11,830 2,755 1,062 1,282 Percent of wage loss 950 221 85 103 Percent of FELA payment 100 23 9 11 RTW claims with more than 90 work days lost (N = 8) Indemnity payments per claim ($) 68,100 26,380 19,557 15,922 Percent of wage loss 370 143 106 87 Percent of FELA payment 100 39 29 23 No RTW claims (N = 10) Indemnity payments per claim ($) 229,140 73,501 139,083 87,224 Percent of wage loss 620 199 376 236 Percent of FELA payment 100 32 61 38 Fatalities (N = 2) Indemnity payments per claim ($) 350,000 942,385 196,338 232,085 Percent of FELA payment 100 269 56 66 NOTE: The wage loss per claim for each of the samples is $1,245 for RTW < 90, $18,398 for RTW > 90, and $36,981 for no-RTW. 'Paid claims only. SOURCE: Data are from one railroad and the Transportation Research Board.

Comparison of FELA and Workers' Compensation • 137 without some connection to job activity, such as unusual strain. Under LHWCA, proof of this type is not necessary and the claims would be compensable. One of the two claims is for a 32-year-old man, so the pension reserve would be quite large ($1.25 million). The third heart attack occurred while the employee was pulling an object. The implica- tion of unusual exertion is presumed to have led to compensation by the railroad and therefore the claim was also assumed eligible in the state systems. The benefits estimated are the pension reserves for the survivors. Because the wage replacement levels of LHWCA are high, the pension reflects this benefit, plus cost-of-living adjustments, for the extended benefits. The state estimates for the death payments are about one-half to two-thirds the FELA payments, and the LHWCA benefits are 2.5 times the FELA settlement. These pension reserves are based on in- complete information on survivors and thus should be considered maximum levels of potential payments. Although these comparisons are limited because of the small sample size and the lack of full information on each claim, the results are nevertheless striking in their consistency and in their correspondence with the other comparisons. FELA benefits are consistently higher than the estimates for the other compensation systems. Although the numbers derived are rough approximations, the range by which FELA exceeds the estimates for the other systems (0.5 to 11 times) is the same as that observed in the other comparisons. COMPARISON BETWEEN COMMUTER RAIL UNDER FELA AND RAIL TRANSIT The New York MTA conducts both commuter rail operations and heavy rail transit operations. Commuter railroad employees of the Long Island Railroad (LIRR) and the Metro North Commuter Rail- road (MNCR) are covered by FELA, whereas rail transit employees such as those of the New York City Transit Authority (NYCTA) (the subway) are covered by workers' compensation. Although an examina- tion of the two groups might appear to offer an ideal comparison between the benefits that each group receives for on-the-job injuries, data problems are present in this comparison as well. The data prob- lems arise from the unavailability of close-year data. Data are only available from 1988 on and only on an accident-year basis for all three operations—the LIRR, the MNCR, and the subway operations of NYCTA. Accident-year data are based on the accidents that occurred in the cited year, and the closing of those claims may require many years

138 • Compensating Injured Railroad Workers Under FELA from the time of accident. For the years after 1988, many claims and lawsuits are still unsettled. This reduces the amount of the calculated payment per claim since the most serious claims and lawsuits take more time to settle. This comparison is therefore limited to the claims for 1988. Given these difficulties, indemnity payments per claim in 1988 for the LIRR were nearly four times the level for the subway system and indemnity payments per claim for the MNCR were nearly seven times the level for the subway. The lower multiple of FELA on the LIRR is explained in part by the fact that wage continuation payments are not included in the settlement figure, whereas they are for the MNCR. Furthermore, the NYCTA figures include a workers' compensation differential payment by MTA that supplements workers' compensation payments to avoid any loss in take-home pay by injured subway workers. In spite of the data availabil- ity problems, the MTA numbers tend to correspond well with the findings in the previous comparisons. TRANSACTIONS COSTS Transactions costs of providing workers' compensation benefits include administration of claims and settlement of disputes for contested (con- troverted) claims. Individual railroads were not able to provide esti- mates of their administrative costs. Administrative costs will tend to rise, however, with the time to resolve complaints. The study committee examined the settlement pattern for FELA claims on the four railroads that provided detailed data as a way of assessing the level of delay that exists in settling FELA claims and the transactions costs these delays imply. Three of the four railroads provided their costs of outside legal counsel, which provide a good indication of the legal costs of dispute resolution. Future efforts to analyze FELA transactions costs would be improved if injury claims administrative costs and all litigation costs (including in-house counsel and court costs) were compiled by individ- ual railroads. A tort system by its nature requires some minimum amount of time for claims to be settled. Because FELA settlements are lump sums, determination of the total losses from an injury may require waiting until maximum medical improvement is achieved and an assessment of potential future losses can be made. For serious injuries, this may take a significant length of time, especially if long-term medical or vocational rehabilitation is required. In workers' compensation, claims may re-

Comparison of FELA and Workers' Compensation • 139 main open for an extended time as well. The difference between the systems is that with workers' compensation, benefits are paid in the interim, whereas with FELA they are not.7 An extended time to settle or close a claim is therefore not unique to FELA, nor is it necessarily undesirable. What it may indicate, how- ever, is the amount of ongoing activity required to settle a claim, includ- ing resources devoted to negotiating claims and preparing for liti- gated ones. Settlement Patterns and Delays The settlement patterns for the four railroads were examined on the basis of direct, represented, and litigated claims for the period 1987 through 1991 for claims settled with indemnity payments. Most delays in settlement are concentrated in the represented and litigated claims, though they represent only 24 percent of indemnity claims over the 5-year period. Furthermore, there is a very stable settlement pattern for each claims category, though each pattern is different. Over half of all direct clains are settled within 6 months, and 90 percent within 2 years. About half of the represented claims remain unsettled after 20 months, and 90 percent settlement is not reached until the 3-year point. Only half of the litigated claims are settled in the first 3 years, and about 30 percent take more than 4 years to settle. Direct Claims For the railroads studied, there is a remarkably stable pattern of direct- claim settlement over time (Figure 6-3). The number of claims settled in 1990-199 1 declined in the early part of the period but without a shift to the right in the settlement curves, indicating that there was no increase in delay. A large number of direct claims are settled in the first 90 days, another push to close claims occurs at the 9-month point, and at the end of 3 years, there is another increase in settlements. Under FELA, the employee has 3 years to file suit; this presumably adds pressure to settle both direct and represented claims before this point, when the decision to file suit must be made. A slight decline in the number of claims closed quickly can be seen in the rightward shift in the cumulative settlement curve in Figure 6-4. During this 5-year period, 50 percent of the claims were closed in about 4 to 6 months. At the end of 1 year, 70 to 75 percent of the direct claims had been closed, and 90 percent were closed by the 2-year point.

1,400 1,200 1,000 800 3 600 400 200 0 1 2 3 4 5 6 8 10 12 14 16 18 20 22 24 26 28 31 34 37 41 45>49 Months Between Injury and Settlement ± 1988 1989 199O 1991 1987 FIGURE 6-3 Settlement pattern for direct claims: Burlington Northern, CSX Transportation, Norfolk Southern, and Union Pcific. HM (/) E 80 CO 0 60 0) 40 0) CO C 0) C-) 20 1 2 3 4 5 6 8 10 12 14 16 18 20 22 24 26 28 31 34 37 41 45>49 Months Between Injury and Settlement 1987 1988 1989 1990 1991 FIGURE 6-4 Cumulative distribution for direct claims: Burlington Northern, CSX Transportation, Norfolk Southern, and Union Pacific.

Comparison of FELA and Workers' Compensation • 141 Represented Claims Represented claims for the four railroads also reveal a fairly stable rela- tionship during this period except for the higher number of extended- settlement claims in 1987 (Figure 6-5). Few represented claims, approximately 15 percent, are settled in the first year (Figure 6-6). Representation may not be sought by injured employees until direct negotiations during that first year fail. By the end of 2 years from the injury date, however, 60 percent of represented claims are settled. A similar push to settle represented claims by the end of the third year raises settlements to over 95 percent. Litigated Claims Litigated claims also show a remarkably consistent pattern of settle- ment (Figure 6-7). The cumulative distribution shows a slight increase in 1990-1991 of about 4 months to reach the same level of claims closure as that in 1987 (Figure 6-8). At the 2-year point, only about 20 percent of litigated claims have been settled. Fifty percent settle- ment is reached at about the 3-year point and about 70 percent at 180- 160 - 140 - 1 2 3 4 5 6 8 10 12 14 16 18 20 22 24 26 28 31 34 37 41 45>49 Months Between Injury and Settlement 1987 1988 1989 1990 1991 FIGURE 6-5 Settlement pattern for represented claims: Burlington Northern, CSX Transportation, Norfolk Southern, and Union Pacific.

100 Cl) E 80 CO 0 60 ci) 0) .40 c 0) 0 20 @1 1 2 3 4 5 6 8 10 12 14 16 18 20 22 24 26 28 31 34 37 41 45>49 Months Between Injury and Settlement 1987 ±1988 1989 1990 1991 FIGURE 6-6 Cumulative distribution for represented claims: Burling- ton Northern, CSX Transportation, Norfolk Southern, and Union Pacific. 600 400 Cl) E CO 300 0 200 100 0 1 2 3 4 5 6 8 10 12 14 16 18 20 22 24 26 28 31 34 37 41 45>49 Months Between Injury and Settlement 1987 1988 1989 1990 1991 FIGURE 6-7 Settlement pattern for litigated claims: Burlington North- ern, CSX Transportation, Norfolk Southern, and Union Pacific.

Comparison of FELA and Workers' Compensation • 143 100 Cl) E 80 c:i 0 60 a) 40 c a) C) 20 CL 1 2 3 4 5 6 8 10 12 14 16 18 20 22 24 26 28 31 34 37 41 45>49 Months Between Injury and Settlement 1987 1988 1989 1990 1991 FIGURE 6-8 Cumulative distribution for litigated claims: Burlington Northern, CSX Transportation, Norfolk Southern, and Union Pacific. 4 years. In this period, 25 to 30 percent of litigated claims took longer than 4 years to reach settlement. Implications of Settlement Delay The settlement patterns reveal that there may be significant delays in receiving FELA settlements for a number of workers. In some mea- sure, this is to be expected, since settlement is unlikely before maxi- mum medical improvenlerit or before a determination of the effect that a permanent impairment may have on the worker. The necessary compensation cannot be assessed until these points have been reached, and for this reason alone, delays in FELA settlements may be inevitable. These delays also confirm the need for the alternative income mainte- nance programs for injured railroad workers that have been developed by employees and the railroads. Whether these delays are too long is difficult to assess. They have not been increasing greatly since 1987. They do represent a clear cost to the injured worker, who must finance the income loss until settlement unless the railroad provides wage continuation benefits or cash ad- vances. Furthermore, Figure 6-2 appears to indicate that from the

144 • Compensating Injured Railroad Workers Under FELA employee's standpoint, there may not be large returns to protracted settlement negotiations. On the other hand, many of the settlements might have been lower absent the extended negotiation. Nevertheless, there appears to be some indication that reductions in settlement lags would benefit both the railroads and their employees. How these delays affect the level of transactions cost depends on the level of negotiation required to achieve them. If little negotiation or claim monitoring is required in the majority of cases, delays in and of themselves may not add greatly to transactions costs. Direct Cost of Litigation In addition to the administrative costs associated with litigation, there is the direct cost borne by both the railroad and the employee. The railroad's cost involves its expenditures for both in-house and outside counsel to negotiate with plaintiffs' attorneys. Three railroads pro- vided their cost for outside counsel to the study committee but data on in-house counsel were not available. For the employee, the direct costs include attorneys' fees and the financing cost of securing wage replace- ment until the settlement, as noted in the previous section. Railroad legal costs are concentrated on the claims settled through litigation (Table 6-8). In 1991 the payment per litigated claim to outside TABLE 6-8 Legal Expenses per Claim for Three Railroads TYPE OF CLAIM 1987 1988 1989 1990 1991 Claims settled with medical payments only ($) Direct 1 3 4 3 9 Represented 12 1 0 10 83 Litigated 5,189 3,656 6,566 7,439 5,635 Total 56 55 80 147 107 Claims settled with indemnity payments ($) Direct 30 56 117 240 267 Represented 81 151 172 319 428 Litigated 5,979 6,866 8,810 9,687 11,189 Total 1,079 1,259 1,457 1,708 2,013 All claims ($) Direct 13 27 58 120 135 Represented 74 131 154 269 363 Litigated 5,928 6,643 8,671 9,474 10,727 Total 535 680 815 994 1,142 SouRcE: Data supplied by three railroads.

Comparison of FELA and Workers' Compensation • 145 counsel was $10,727, which is nearly twice its level in 1987. This is a faster increase than has occurred in the indemnity payment per claim of litigated settlements. Using those figures from Table 6-2, legal costs as a percentage of the value of litigated indemnity payments have increased from just 4.6 percent in 1987 to nearly 6.7 percent in 1991. There are no data on plaintiffs' payments to their attorneys. The railroad labor presentation to the study committee indicated that a condition of endorsement of FELA plaintiffs' attorneys is that they limit their fees to 25 percent of the settlement on a contingency basis. Assuming this 25 percent and combining it with 7 percent for the railroad's counsel, the legal cost of litigated claims reached nearly one- third of the settlement value in 1991.8 The total legal cost for all settlements for the four railroads may be estimated similarly by applying the 25 percent plaintiffs attorney fee to the total represented and litigated claims and scaling up the legal costs of the three railroads to estimate the legal cost for the fourth railroad. Estimated total legal cost (plaintiffs counsel and defense's outside legal counsel) per indemnity claim in 1991 would be nearly $11,000, or about 20 percent of the average indemnity payment of $53,770. These are low estimates for the direct cost of litigation because the railroads' internal legal costs are not included, nor are associated court costs. The direct litigation costs of FELA compare favorably with those of tort litigation in general. A RAND Institute for CivilJustice study of the costs of tort litigation for 1985 found that plaintiffs' legal fees and expenses were around 30 to 31 percent of their total compensation, about what it is for railroad plaintiffs (Kakalik and Pace 1986). On the basis of total expenditures, net compensation to the plaintiffs was 56 percent, plaintiffs' legal fees and expenses were 24 percent, defen- dants' legal fees and expenses were 18 percent, and court expenditures were 2 percent. FELA appears to have an edge over other tort action in defendants' legal fees and expenses, which were 6.2 percent (6.7 per- cent adjusted to the total expenditure level) in 1991. Although this percentage for FELA does not include the railroads' administrative cost of tracking litigated claims nor their cost of in-house counsel, it seems unlikely that these costs would equal 10 percent of total expendi- tures given that the four railroads use outside counsel as their primary attorneys in all litigated claims. Level of Litigation Comparing the levels of litigiousness between FELA and workers' com- pensation systems is complicated by the many differences in hearing and appeal procedures among workers' compensation systems and the basic differences between workers' compensation systems and the FELA

146 • Compensating Injured Railroad Workers Under FELA process. Because FELA is a tort process, one would expect a certain, perhaps fairly consistent, number of FELA claims to be litigated, whether a trial occurs or not. If all claims were settled directly, railroads might conclude that their settlement standards were too generous. In Chapter 5 a number of case studies are described that examine the types of hearing and appeal procedures available in those states and their level of use. In spite of the administrative nature of workers' compensation, a significant number of claims are controverted in those systems. As an addendum to Table 5-2, the percentage of railroad indemnity claims that are represented and litigated is as follows (in- formation from Burlington Northern, CSX Transportation, Norfolk Southern, and Union Pacific): Average Percentage Type of Claim (1988-1 991) Represented 6:2 Litigated 12J Total 23.3 Represented and litigated claims are roughly the equivalent of ap- pealed claims in FELA because a claim moves to this point after the employee has rejected the railroad's final direct offer and, in a sense, appeals through an attorney for further negotiation or trial. The extent of this "appeal" can be quite variable, with many claims and suits being settled quickly even in these categories. Combined represented and litigated claims as a percentage of total indemnity claims have consistently varied around 23 percent in the 4-year period from 1988 to 1991. On a total claims basis, including claims settled with medical payments only, the percentage has been 12 to 14. SUMMARY Across the spectrum of injury compensation benefits, the level of benefit payments provided to injured railroad workers under FELA is higher than that provided to employees under workers' compensation systems. The benefits for railroad workers, however, can be delayed for a significant period. Because of the limitations of the data, it is not possible to state with great certainty how much higher FELA payments are. Reasonable estimates of the FELA injury premium are between 2 and 4 times workers' compensation injury payments. From this amount, one-fourth may go to an attorney if the employee has retained one. The remaining premium can be attributed to the relatively high average wages in the

Compari.son of FELA and Workers' Compensation • 147 railroad industry, compensation for the full value of wage loss without caps, recoveries for impairments without caps, and compensation for pain and suffering. Significant transactions costs accompany the process for securing these benefits, including substantial attorneys' costs for both railroads and employees and significant delays in receiving benefits. The fact that most injury claims are already handled administratively may re- duce the potential savings from changes to FELA from what might otherwise be presumed. Nevertheless, workers' net benefits received are affected by both attorney and delay costs. Thus, it would appear that both employees and the railroads could benefit from efforts to reduce these transactions costs; both could be made better off without a reduction in the net benefits received by railroad workers. NOTES The meaning of the term "indemnity payments" in the workers' compensa- tion context is often confined to compensation for wage losses and other out- of-pocket expenses only. Within the FELA context, however, the term is expanded to encompass all expenses other than medical expenses, includ- ing pain and suffering. Data on medical payments were provided by two of the railroads, one an eastern and the other a western carrier, and include the full cost of medical care for the injury. Medical payment data are separated into two claims categories: claims settled with medical payments only and claims settled with indemnity payments (see Appendix B on terminology). Railroad medical payments include the total cost of the medical payment, including both the railroad's portion and the insurance portion from the Travelers' policy. For the railroads, the increase from 1986 to 1988 in medical payments for claims settled with indemnity payments was 42 percent and for New York it was 40 percent. The increase from 1986 to 1987 for the railroads was 18 percent; for Pennsylvania, 16 percent; and for Wisconsin, 7 percent. If a state has particularly low indemnity benefits, then, inherently, the medical share will be in the higher end of that range. The railroad is just moving to include more definition of the injury along the lines of the workers' compensation categories, but this effort has begun only recently and few records indicated such information. Wage continuation, cash advances, and sickness benefits have developed over time to compensate for the lack of interim benefits until the claim is closed (see Chapter 2). In contrast to these figures, the direct expense of litigation, including medical-legal cost and attorney and court costs for all parties, in California's workers' compensation averaged $8,649 per case for disputed claims in 1992, and the average net award to the injured employee was $10,868.

148 • Compensating Injured Railroad Workers Under FELA "Thus it cost $8,649 to provide $10,868 in benefits—or 80 cents to deliver one dollar in benefits awarded in a litigated claim in 1992" (CWCI 1993). REFERENCES ABBREVIATIONS CWCI California Workers' Compensation Institute NCCI National Council on Compensation Insurance Ballantyne, D., and C. Telles. 1991a. Workers' Compensation inMinnesota:Admin- i.strative Inventory. Workers' Compensation Research Institute, Cambridge, Mass. Ballantyne, D., and C. Telles. 1991b. Workers' Compensation in Pennsylvania: Administrative Inventory. Workers' Compensation Research Institute, Cam- bridge, Mass. Ballantyne, D., and C. Telles. 1992a. Workers' CompensationinNew York:Adminis- trative Inventory. Workers' Compensation Research Institute, Cambridge, Mass. Ballantyne, D., and C. Telles. 1992b. Workers' Compensation in Wisconsin: Admin- istrative Inventory. Workers' Compensation Research Institute, Cambridge, Mass. CWCI. 1993. California Workers' Compensation Institute Bulletin, No. 93-8, May. Kakalik,J. S., and N.M. Pace. 1986. Costs and Compensation Paid in TortLitigation. RAND Institute for Civil Justice, Santa Monica, Calif. NCCI. 1992. Annual Statistical Bulletin. Boca Raton, Fla.

7 . S SSS•S••• SSSSS••IS•S•SS••• Modal Competition and Federal Appropriations If the cost to railroads and transit agencies for injury compensation under the Federal Employers' Liability Act of 1908 (FELA) is higher than it might be under alternative systems, adopting one of those systems could lower the railroads' cost of providing transportation. The effect of these lower costs on modal freight competition and federal appropriations is examined in this chapter. A strong caveat must be introduced regarding this discussion. As noted in earlier chapters, the benefits that FELA provides are an integral part of the overall level of benefits and wage compensation provided to railroad workers. An assessment of FELA and its effects cannot be made in isolation from these benefits. Where the delivery of a benefit under FELA may be complex and administratively inefficient, non-FELA agreements can and do substitute for them (e.g., health insurance). Likewise, FELA benefits may substitute for other forms of compensation. If FELA were changed or replaced, cost savings might result from reductions in either transactions costs or benefits provided to workers. Reductions in the former might result in lasting operating cost savings to the railroads. Costs saved from reduced benefits, however, could be partially offset by demands for higher wages or other types of benefits to compensate for the FELA reductions, thereby (to the extent that 149

150 • Compensating Injured Railroad Workers Under FELA such offsetting wage and benefit demands were effective) reducing or eliminating the savings. MODAL COMPETITION If railroad costs related to compensation for injured workers were reduced, freight rates, rail traffic volumes, and profits could be af- fected. The translation of these effects to market share is examined in this section. Market Effects If labor costs were reduced, a railroad would likely reduce rates in certain market segments, resulting in higher freight volumes that would increase profits. These segments probably would include those in which there now is close competition between railroads and trucks or barges or in which two railroads compete directly. The competitors in these market segments are continually seeking cost savings, and the competition would work to keep rates close to costs. In market niches in which a single railroad dominates and there is little possibility that price cuts will stimulate new freight demand, the railroad might maximize profits by making little or no change in rates, increasing its net income by the amount of the reduction in labor costs on the traffic it already has. These circumstances may exist in a few market segments—for example, for some long-haul bulk commodi- ties—but they probably are not common, though these segments may have heavy traffic. Even a monopolist generally will be able to increase profit by cutting prices when costs fall. Therefore, in most rail market segments, some part of the savings from a reduction in labor costs might be passed on sooner or later to customers through reduced freight rates. The lower rates would in- crease rail traffic, or at least allow rail to retain traffic it otherwise might have lost. The additional rail freight would be primarily freight diver- ted from truck or barge, although lower rates would generate some net increase in the total quantity of freight traffic. Price Elasticity The effect of rail rate reductions on rail traffic and revenues depends on the price elasticity of demand for rail freight services: the percent- age change in the consumption of rail services in a market segment (measured in ton-miles) per 1 percent change in the price of rail transport. Elasticity estimates vary widely by commodity and from

Modal Competition and Federal Appropriations • 151 study to study, although most estimates appear to fall somewhere between about —0.25 and —2.5. Table 7-1 shows estimates from four representative studies. Table 7-2 is a synthesis of estimates from a survey of the most recent major studies prepared by Oum (1992). It is difficult to interpret relative elasticity estimates among com- modities in terms of conventional wisdom about captive and competi- tive rail market segments. The estimates generally indicate that rail elasticities with respect to rail prices are between 0 and —1 in truck- competitive markets; that is, the demand is inelastic in these markets. In these contested markets, rail traffic volumes would be expected to be more sensitive to changes in rates than this; the low elasticity estimates may reflect the aggregation of competitive and noncompetitive market niches within broadly defined market segments. Also, relative rankings of elasticities by commodity do not appear very consistent from study to study. A shortcoming of many of the published models is reliance on data from the 1970s, when the federal government conducted several freight transportation surveys and rate information could be obtained easily from published tariffs. Structural changes in the industry have been so great since that time that the validity of prederegulation studies is suspect. If an elasticity of —1 is assumed, a labor cost savings equivalent to 1 percent of prior revenues and fully passed on as a rate decrease would yield a 1 percent increase in traffic volume, or 10 billion ton-mi annu- ally, and no increase in revenues. Because of the uncertainties evident in the literature on rail price elasticity and the likely small size of the effect, more refined elasticity estimates do not appear to be necessary. Level of Cost and Modal Share Changes Efforts to estimate the actual savings to railroads from amending FELA or moving to an alternative injury compensation system are tenuous at best. Because railroad workers are likely to seek compensation through other benefits equal to lost injury compensation benefits, lasting cost savings are most likely to flow from reductions in the costs of adminis- tration and litigation.' As discussed in Chapter 4, the railroads' estimate of administrative and legal costs was $173 million in 1991, though the study committee did not receive any data to confirm this estimate from individual railroads. Medical and indemnity costs of FELA were $115 million and $911 million, respectively, in the same year. Assuming that the rail- roads' transactions cost estimate is accurate, the universe of potential

TABLE 7-1 Freight Rate Elasticity Estimates (Oum 1979; Winston 1981; Inaba and Wallace 1989; Wilson 1984) E&sTIcrn' AUTHOR DATA COMMODITY TRUCK RAIL Oum Canadian waybill sample, 1970 Lumber —0.5 —0.5 (aggregate) Chemicals —0.9 —0.6 Petroleum products —0.4 - 1.0 Metallic products —0.3 - 1.2 Nonmetallic products —0.5 - 1.0 Winston ICC survey data, 1973-1977 Lumber —0.1 —0.1 (disaggregate) Chemicals —2.3, - 1.9" —2.2 Petroleum and petroleum products —0.5 —0.7 Primary and fabricated metals —0.2, —0.3 —0.0 Stone, clay, and glass products —2.0, —2.2 —0.8 Inaba and Wallace Sample of shipments from grain Wheat —0.3 to —0.0 to elevators, 1984 (disaggregate) —0.9 —1.0 Wilson Monthly shipments from North Wheat Dakota to two terminals, Barley —0.7, - 1.2' —0.3, —0.4 1973-1982 (aggregate) . —0.2, —0.8 —0.1, —0.6 'Percentage change in the share of the mode with respect to a 1 percent change in that mode's price. Elasticities from study by Oum are evaluated at the mean values of the independent variables in the model (shipment and service characteristics and prices). Oum also reports various point estimates of elasticities; for example, truck elasticity for fruits and vegetables rises from —0.2 to —0.4 as distance increases from 80 to 350 mi. Elasticities from study by Winston are weighted averages of computed elasticities at the observation values of the variables. The range of truck elasticities given by Inaba and Wallace is for various origin-destination pairs; the rail elasticities are for various origin-destination pairs and shipment sizes (single car, multicar, unit train). two truck elasticities are for private carrier and common carrier, respectively. 'The two elasticities given by Wilson are for shipments to Minneapolis and Duluth, respectively.

TABLE 7-2 Demand Elasticities Synthesized from 11 Studies (Oum et al. 1992) LOG- DISCRETE-CHOICE COMMODITY LINEAR AGGREGATE L0GIT TRANSLOG MODELa Aggregate commodities 1.52 0.25-0.35, 0.83, 0.09-0.29, 0.60 NA 0.34-1.06 Chemicals NA 0.66 0.69 2.25 Fabricated metal products NA 1.57 2.16 NA Food products 0.02, 1.18 1.36 2.58, 1.04 NA Iron and steel products NA NA 2.54, 1.20 0.02 Machinery NA 0.16-1.73 2.27-3.50 0.61 Paper, plastic, and rubber products 0.67 0.87 1.85 0.17-1.09 Petroleum products NA NA 0.99 0.53 Stone, clay, and glass products NA 0.69 1.68 0.82 Textiles NA 2.03 NA 0.56 Transport equipment NA NA 0.92-1.08 2.68 Wood and wood products 0.05 0.76 1.97, 0.58 0.08 NOTE: All elasticity estimates are in negative values. NA = not available. 'The estimates in this column are mode-choice elasticities.

154 • Compensating Injured Railroad Workers Under FELA savings from changes to FELA was about $1.2 billion in 1991. Total freight revenue in 1991 was $26.9 billion. Table 7-3 shows what the effect might be on rail market share of various percentage savings in FELA costs when combined with the assumptions that all labor cost savings are passed on as rate price reductions and an own-price elasticity of —1.0. For example, savings of 10 percent would result in $120 million savings in labor costs, which is 0.45 percent of 1991 freight revenues. Using the elasticity estimate above and applying it to 1991 traffic levels, the potential increase in rail traffic would be about 4.8 billion ton-mi. On the basis of 1991 total truck, rail, and barge ton-miles of 2,121 billion, this increase in traffic would amount to an increase in market share of 0.23 percent (Smith 1993). If some of the reduction in cost were passed on to shareholders and railroad workers, the increase in market share would be proportion- ately smaller. For example, if half of the savings were passed on to shareholders and railroad workers, the estimated changes in traffic levels and market share shown in Table 7-3 would be halved. FEDERAL APPROPRIATIONS Two federal programs are potentially affected by FELA costs; these are the expenditures by the Federal Transit Administration (FTA) for transit operating grants and expenditures by the Federal Railroad Administration (FRA) for National Railroad Passenger Corporation (Amtrak) operating grants. Table 7-4 shows the spending authority approved for each of these programs for the last 5 years. Since federal appropriations are used to close the gap between revenues and ex- penses for both Amtrak and transit operations, reductions in expenses for these operators could be reflected directly in reduced future appropriations. Federal Transit Administration Programs FTA spending is concentrated in two programs: the formula grant and discretionary grant programs. Discretionary grants are for capital spending on existing rail systems, new rail starts, and buses. Formula grants can be used for any purpose by transit agencies, though there is a legislative capon the amount that can be used for operating costs. The cap has remained at about $800 million for the last 5 years. This $800 million represents the total universe of potential cost to the federal budget for FELA-induced injury compensation costs for commuter railroads. If FELA does in fact create expenses for corn-

TABLE 7-3 Potential Changes in Rail Market Share Using All Cost Savings To Reduce Rates CHANGE IN 1991 RAIL COSTS PERCENTAGE SAVINGS MILLIONS PERCENTAGE IN FELA COSTS OF DOLLARS OF REVENUE CHANGE IN 1991 RAIL TRAFFIC LEVELS (TON-MI BILLIONS) CHANGE IN 1991 RAIL MARKET SHARE (%) 10 120 0.45 4.8 0.23 15 180 0.67 7.2 0.34 20 240 0.89 9.6 0.45 25 300 1.11 12.0 0.57 30 360 1.34 14.4 0.68 35 420 1.56 16.8 0.79 40 480 1.78 19.2 0.90 45 540 2.00 21.6 1.02 50 600 2.23 24.0 1.13 1991 Estimated total FELA cost $1.2 billion 1991 Rail freight revenue $26.9 billion 1991 Rail ton-miles 1,077 billion 1991 Total rail, truck, and barge ton-miles 2,121 billion NOTE: Rail price elasticity is assumed to be - 1.0. SOURCE: Data are from the Association of American Railroads (1992) and Smith (1993)

156 • Compensating Injured Railroad Workers Under FELA TABLE 7-4 Spending Authority for Relevant Federal Transit Administration and Federal Railroad Administration Accounts MILLIONS OF CURRENT D0LLARsa 1993 1992 1991 1990 1989 Federal Transit Administration Formula grants 1,700 1,520 1,605 1,625 1,605 Operating cap 802 802 802 805 805 Discretionary capital grants 1,725 1,910 1,400 1,282 1,250 Total FTAb 3,800 3,775 3,259 3,193 3,265 Federal Railroad Administration (Amtrak) Operating grants 331 331 343 NA NA Retirement and 146 150 150 NA NA unemployment Total operating 477 481 493 520 554 Capital' 165 175 132 85 30 Total Amtrak 642 656 625 605 584 NECIPd (capital) 204 205 179 25 20 NOTE: Spending authority is generally from appropriations except for FTA discre- tionary grants, which are the obligation limits on contract authority. NA = not applicable. Values for 1993 are estimated. 'Total FTA includes the listed accounts plus capital spending for Washington Metropolitan Area Transit Authority and interstate transfer grants and spending for research and FTA administrative expenses. 'Value for 1989 include labor protection payments, which are included in operating grants thereafter. dNECIp is the Northeast Corridor Improvement Program that rebuilt the Northeast Corridor for Amtrak and hasjust begun the process of electrifying the lines north of New Haven to Boston. SOURCE: Appendix to the budget of the U.S. government for various fiscal years. muter railroads that could be avoided by switching to an alternative injury compensation system, theoretically, if everything else is held the same, Congress could modify FELA and the resulting savings could be used to reduce the formula grant appropriation. For example, if re- placing FELA resulted in savings of $40 million for commuter rail- roads, formula grants could be reduced to $1,660 million and the operating cap could therefore be lowered to $762 million because the $40 million would come from the savings in operations. What the actual sa"ings to transit operators would be is difficult to estimate. Transit railroad workers are likely to seek increases in wages for any perceived loss of FELA benefits. To the extent that they are successful, the reduction in operating loss from changes to FELA is

Modal Competition and Federal Appropriations • 157 likely to be confined to savings in transactions cost. The study commit- tee did not have estimates of these costs for transit systems, but it is suspected that the savings would be small compared with overall oper- ating expenses. To the extent that these savings are produced, however, there could be a dollar-for-dollar reduction in the federal appropria- tions for transit operating subsidies. Federal Railroad Administration Programs For Amtrak, operating grants are split between those required to pay retirement and unemployment costs of the carrier and general operat- ing subsidies. The universe of expenditures that could be affected by Amtrak's coverage under FELA is therefore the $331 million in operat- ing subsidies estimated for 1992. The effect that changes to FELA would have on Amtrak is similar to thatjust described for transit. To the extent that savings are achievable and not partially offset by changes in other benefits, the federal operating subsidy for Amtrak could be reduced by the same amount. Amtrak's expenditures on FELA amounted to $56.5 million in 1992 including indemnity payments of $34.9 million, administrative costs of $12 million, and $9.7 million in medical costs. A 10 percent savings in these costs would yield a 1.7 percent reduction in federal appropria- tions for Amtrak's operating subsidy. Table 7-5 gives other percentages TABLE 7-5 Potential Changes in Amtrak's FELA Costs and Federal Appropriations PERCENTAGE CHANGE IN CHANGE IN FEDERAL SAVINGS IN AMTRAK COSTS APPROPRIATIONS FOR AMTRAK'S FELA COSTS ($ MILLIONS) OPERATING SUBSIDY (%) 10 5.7 1.7 15 8.5 2.6 20 11.3 3.4 25 14.1 4.3 30 17.0 5.1 35 19.8 6.0 40 22.6 6.8 45 25.4 7.7 50 28.3 8.5 1992 Amtrak total FELA cost $57 million 1992 Amtrak operating subsidy $331 million SouRcE: Data are from Amtrak and Appendix 1 of the U.S. budget for fiscal year 1993.

158 • Compensating Injured Railroad Workers Under FELA of FELA cost savings and the resulting percentage savings in appro- priations that might result. NOTE I. Whether railroad workers are likely to be successful depends on many factors, including the supply-and-demand conditions of railroad labor markets. REFERENCES ABBREVIATION AAR Association of American Railroads AAR. 1992. Railroad Facts, 1992 ed. Washington, D.C., Sept. Inaba, F. S., and N. E. Wallace. 1989. Spatial Price Competition and the Demand for Freight Transportation. Review of Economics and Stati.stics, Vol. 71, No. 4, Nov. Oum, T. H. 1979. A Cross Sectional Study of Freight Transport Demand and Rail.Truck'Competition in Canada. BellJournal of Economics, Vol. 10, No. 2, Autumn. Oum, T. H., W. G. Waters II, andJ.-S. Yong. 1992. Concepts of Price Elasticities of Transport Demand and Recent Empirical Estimates.Journal of Transport Economics and Policy, Vol. 26, No. 2, May. Smith, F. A. 1993. Transportation in America: A Stati.sticalAnalysis of Transportation in the United States, 11th ed. Eno Foundation for Transportation, Inc., Landsdowne, Va. Wilson, W. W. 1984. Estimation of Modal Demand Elasticities in Grain Trans- portation. Western Journal of Agricultural Economics, Vol. 9, No. 2. Winston, C. 1981. A Disaggregate Model of the Demand for Intercity Freight Transportation. Econometrica, Vol. 49, No. 4, July.

. .. .. . . S S S S 55 •S•••5••55IS•5 Findings and Conclusions The study committee's findings and conclusions concerning FELA are summarized in this chapter. No specific recommendations are made because the committee was asked not to recommend changes to FELA. GENERAL FINDINGS Although the essential characteristics of FELA law have changed little since 1908, the process has evolved over the years into a more adminis- trative system. Medical expenses for on-the-job injuries are now han- dled administratively through a health insurance plan (though they remain elements of damages that can be recovered, if necessary, at trial or settlement). Railroads and their employees have developed pro- cedures for dealing with routine claims so that 70 percent are handled without litigation or employee legal representation. In addition, the courts have effectively reduced the standards of proof required to demonstrate negligence on the part of the employer. No injury compensation system for workers exists in isolation from other workplace benefits. The value of either the FELA process or workers' compensation systems to injured workers depends in part on the availability of other worker benefits that may complement or substi- tute for the benefits provided through the injury compensation system. 159

160 • Compensating Injured Railroad Workers Under FELA Thus, comparisons between the FELA process and state workers' com- pensation systems must be viewed in the context of overall workplace benefits. Moreover, because these other benefits may be provided through collective bargaining agreements, changes in the injury com- pensation system that result in reduced levels of statutory benefits are likely to induce a response in the next round of collective bargaining that might partially offset the changes to workers' benefits and em- ployers' costs. Over the years, the FELA process and workers' compensation sys- tems have evolved in ways that make them increasingly similar. As the standards for proving employer negligence have lessened, FELA has functioned less and less as a negligence-based, trial-driven tort system, and the majority of FELA claims are now settled without litigation or employee legal representation. At the same time, litigation within workers' compensation systems has been on the rise, although disputes are usually handled through an administrative proceeding that is less costly and generally faster than court actions. Despite this evolution, the two approaches to compensating injured workers remain quite different. Compared with state workers' compensation programs, the FELA process generally provides higher, more extensive benefits to injured workers but can result in delays in payments, involves some- what higher transactions costs, and results in higher costs to railroad employers. More specific study findings are presented in the following discus- sion for each of the critical areas of evaluation relating to injury com- pensation systems: benefits to injured workers, the costs of providing these benefits, and incentives. SPECIFIC FINDINGS Overall goals of injury compensation involve equity, efficiency, and incentives. Ideally, an injury compensation system should be equitable to the injured worker, should provide benefits in an efficient manner, and should be structured so that each party has incentives to reduce both injuries and the costs of those injuries that occur. A system's efficiency and incentive structure can be assessed objec- tively, but the fairness of any particular system depends on more subjective perspectives or social philosophies of individuals or groups. No attempt was made in this study to define an equitable injury com- pensation system. Instead, the criteria that may be considered in judging the fairness of a particular injury compensation system were compared for both FELA and alternative compensation systems. These criteria include the extent of coverage, the level of the compensation for

Findings and Conclusions • 161 losses, the speed with which the losses are compensated, the certainty with which they are compensated, and who bears the costs of compensation. The specific findings concerning these criteria are presented next, but a final caveat should be stated regarding these findings. The comparisons that were made in this study had to use incomplete and rarely comparable data. Although the results should therefore be viewed as rough comparisons and rough estimates of magnitudes, they are the best possible at this time. Future efforts to analyze FELA and compare its delivery of benefits to other systems would be greatly improved if the railroads were to begin collecting and classifying their injury compensation data within categories comparable with those used in workers' compensation systems. Benefits to Injured Workers Benefits to injured workers involve questions of the extent of coverage, the levels of compensation provided, the degree of certainty in the system, and the relative cost burden for losses. Do All Injured Receive Compensation? Neither FELA nor workers' compensation compensates all injured individuals. FELA does not compensate individuals who are solely negligent in causing their injury, whereas workers' compensation does because it is a no-fault system. In practice, however, FELA fails to compensate injured workers in only a small percentage of cases. Workers' compensation systems do not provide benefits for individuals who do not meet minimum lost-work-time requirements and may not recog- nize certain impairments or occupational illnesses as compensable, whereas all of these losses are reimbursable under FELA. Do Like Injuries Receive Like Compensation? Under FELA, there is a potential for different settlements for the same injury. Identically injured workers may seek and receive different compensation if they incur different consequences from the injury. Furthermore, for claims that go to trial, different juries and jurisdic- tions may provide varying results. Railroads, however, use tacit for- mulas or rules of thumb in their negotiations with injured employees (and plaintiffs' attorneys); therefore, more uniformity exists than might be expected. Workers' compensation systems, on the other hand, tend to have formulas and schedules for compensation that make benefit payments more certain and comparable among workers with similar

162 • Compensating Injured Railroad Workers Under FELA injuries. Nevertheless, these formulas and schedules can vary signifi- cantly among states and some states do not use them. Moreover, some workers' compensation claims are resolved through negotiation rather than according to administrative schedules. What Are the Medical Benefits? Medical coverage is the same under both FELA and workers' compen- sation: 100 percent of medical costs associated with the injury is cov- ered. Some data suggest that medical costs are lower in the FELA process than in workers' compensation. This may be because FELA medical benefits are provided under the regular railroad health insur- ance plan and therefore are perhaps subject to tighter scrutiny and cost control. Steady, large increases in medical costs are a continuing and growing problem in workers' compensation systems. Both systems' approach to medical benefits is likely to be affected by national health care reform, but it is not possible to determine at this time what changes will take place. What Are the Indemnity Benefits? FELA provides a higher level of benefits than workers' compensation systems because all losses are compensable without preset limitations. Unlike workers' compensation, FELA has no cap on the level of wage- loss recovery or on recovery for impairment, and the process allows compensation for pain and suffering. FELA benefits are, however, subject to reduction for comparative negligence. Although compari- sons are complicated by inconsistent data, available information indi- cates that FELA indemnity payments (payments for all nonmedical losses) to injured workers are about two to four times greater than they might be under workers' compensation systems for comparable inju- ries. Part of this difference is accounted for by the higher wage levels of railroad workers compared with average state levels and benefit caps in workers' compensation systems. The remainder can be attributed mainly to the broader coverage under FELA for all out-of-pocket costs and for nonmonetary losses such as pain and suffering. The scope and levels of injury compensation are functions of each system's compensation philosophy. Because of the no-fault trade-off made in workers' compensation, all workers receive some award, but not all losses are reimbursed. On the other hand, in the negligence- based approach of FELA, all losses are potentially compensable, but not all workers are compensated. Lower compensation benefits, how- ever, are not inherent in a no-fault system, just as a tort-based system

Findings and Conciwions • 163 such as FELA does not have to'compensate for nonmonetary losses. Ultimately, federal and state legislatures control what is compensated. How Certain Is Compensation? More uncertainty exists about receiving benefits and their levels under FELA than under workers' compensation systems. However, railroads have voluntarily adopted a routinized system of benefit delivery for most claims that reduces this uncertainty. Under workers' compensa- tion there is little uncertainty concerning the receipt of benefits, but the level of benefits in some states may be less certain, especially for injuries with lasting impairment. What Is the Relative Burden Share of Losses? The efficiency and incentives of a negligence-based compensation system depend in part on the relative burden for the costs of injury that the system places on each party. The relative burden share between the railroads and their workers is not known because the comparative negligence and the employee's estimated total losses for each claim (for which compensation was sought) are not known. There is a fixed burden in workers' compensation that forms part of the trade-off for no-fault coverage; the employer bears the monetary losses for the most part and the employee bears the nonmonetary losses. Costs of Running the System The costs for an injury compensation system include the direct cost of the benefits and the indirect or overhead costs of the transactions necessary to deliver those benefits to the injured worker and support his or her rehabilitation. What Are the Direct Costs to Employers? FELA compensation to injured workers has grown moderately since 1987, with continuous increases in real compensation per injury claim. There was a short-term expansion in occupational illness claims that was responsible for part of the increase in overall costs, but that may have run its course. Nevertheless, claims for occupational illnesses such as silicosis, carpal tunnel syndrome, and mental stress have the poten- tial for increasing costs in the future. Clearly, the higher and more extensive benefits in FELA create higher direct costs for railroads than the same injury would incur under alternative workers' compensation

164 • Compensating Injured Railroad Workers Under FELA approaches. These FELA benefits, however, may be substituting for other benefits or higher wages that would have been demanded by railroad workers in the absence of this injury compensation benefit. What Are the Transactions Costs for Both Employees and Employers? Transactions costs appear to be lower in workers' compensation systems than under FELA. FELA has become a routinized system of compensa- tion in which settlements are achieved primarily through negotiation and claims rarely go to trial. Delays in FELA settlements, however, create costs for both railroads and workers that are less likely in workers' compensation, which is more successful in rapid replacement of wages. Litigation occurs in both the FELA process and workers' compensa- tion systems. Overall, it appears that there is more litigation in the FELA process and that cases are more protracted. Lump-sum settle- ments cannot be made until all actual and potential losses are known or estimated, and given the overall higher level of benefits in FELA, there is more at stake and more to negotiate. There is also considerable litigation in many workers' compensation systems, and although litiga- tion increases are not universal, many systems are experiencing them. Workers' compensation litigation takes place mostly before administra- tive bodies rather than courts and juries and is likely to be less expensive than FELA litigation. Finally, there is some reason to believe that FELA is more efficient in transactions costs than other tort systems. Incentives The incentives that are of primary concern in an injury compensation system are those that promote workplace safety, rehabilitation and the return to work of the injured worker, and efficient dispute resolution. What Are the Safety Incentives? Both FELA and workers' compensation provide incentives to improve safety, and there is no clear empirical evidence to support FELA's claims to superiority. The costs of both systems are likely high enough to provide safety incentives. An injury compensation system can promote workplace safety through incentives that deter unsafe behavior and aid the discovery of unsafe conditions. If there were an allocation of losses under FELA solely in accordance with each party's negligence in the injury, FELA might strongly discourage unsafe behavior, more so than workers' compensa-

Findings and Conclusions • 165 tion systems do, because the injured worker could potentially be ex- posed not only to uncompensated nonmonetary losses (as in workers' compensation) but also to uncompensated monetary losses arising from the injury. The lack of evidence concerning the relative burden sharing for compensation between the railroads and their workers prevents a firm conclusion on this point. The discovery of unsafe conditions may be hampered more under FELA than workers' compensation. Because the investigation of an accident or injury may bear on the question of the railroad's and the employee's negligence, the willingness of all parties to cooperate in such an investigation may be reduced. No empirical evidence was available on this point, however. The data show that there are substantial differences among the railroads in the frequency of injuries and fatalities. This would suggest that the railroads with higher injury rates might have opportunities to improve their situation by more emphasis on safety, regardless of what injury compensation law is in effect. What Are the Return-to-Work Incentives? Workers' compensation may provide greater overall incentives to re- turn to work than FELA, but there are no good data to make definitive findings on this issue. Workers' compensation systems deal more di- rectly with rehabilitation and return-to-work costs than does the FELA process. Nevertheless, the delays in settlement under FELA do not necessarily result in delayed return to work for the injured worker. Furthermore, the lump-sum settlements in FELA may provide an incentive for the worker to return to some work, whether in the original job or a new one. What Are the Dispute Resolution Incentives? Under FELA, the dispute resolution process is more adversarial, and the incentive for quick resolution of the dispute may be less than in workers' compensation systems. As noted previously, this finding should be expected in a tort-based system. Just how great the disincentive is under FELA is not clear from the data available. Modal Competitiveness and Federal Appropriations The effect that FELA may have on railroad modal competitiveness and federal appropriations is indeterminate. If cost-saving changes can be made to the railroad injury compensation system, the freight railroad

166 • Compensating Injured Railroad Workers Under FELA market share or profitability may increase and federal appropriations for both Amtrak and railroad transit operations may decrease. To the extent that these savings are made through reductions in the net benefits received by injured railroad workers, however, changes in market share and federal appropriations may be reduced or may not persist. Modal Competitiveness A system like workers' compensation would likely be less expensive for railroads, perhaps resulting in either higher profits or a larger market share, or both, but the effect is likely to be modest. Some savings from reductions in injury benefits may be put into satisfying higher wage demands (as well as to be shared by stockholders and customers) and thus the actual savings could be reduced. Because FELA is tied to a series of other compensation programs, some of which are collectively bargained, a change in injury benefits, even if less costly, might well result in a series of changes in other programs, which would mitigate the cost savings. Thus, whatever the initial effects appear to be, they would almost certainly be less after the system had had time to adjust to the change. Federal Appropriations There is perhaps some potential for savings in outlays if Amtrak or commuter railroad agencies move to an alternative system or make improvements in FELA. Outlay savings depend on whether reduced injury compensation costs are translated into reduced appropriations, are used to finance additional passenger and transit services, or are mitigated by changes in wages or benefits to transit and Amtrak workers. OPPORTUNITIES FOR IMPROVING THE FELA PROCESS Changes to FELA depend on a willingness of the railroads and their employees to discuss solutions. Indeed, nearly all labor and social legislation affecting the railroad industry has resulted from negotia- tions between the carriers and their employees. Because of the com- plexity of the issues, links to negotiated compensation packages, and the lack of clear advantages to any one approach, joint labor and management consultation is probably the most promising approach to reducing the costs of injury compensation in the railroad industry. Although the study committee was asked not to make specific rec-

Findings and Conclusions • 167 ommendations about improvements to FELA, the consensus was that it would be wise if all changes to FELA were developed through negotiation. Regardless of what changes are made to FELA or what alternatives are proposed, reductions in the costs of providing injury compensation would come mainly from reduced benefits or from reduced transac- tions costs. Any decrease in injury benefits would result in reduced overall worker income. Employees would likely resist such an effective wage cut and might seek a compensating change elsewhere in their overall wage and benefit package. Savings from reductions in transac- tions costs through decreases in delays, litigation, and administrative costs are likely to be more feasible because such reductions would benefit both employees and the railroads. Improving the dispute resolution process is clearly one area in which it is possible to make constructive changes to FELA to the advantage of both sides. The parties might well explore the range of alternative dispute resolution (ADR) methods that are increasingly being used in the public and private sectors. If the level of litigiousness and delay in all settlements could be reduced, both injured workers and railroads could benefit.

APPENDIX A ......•• .............S........ Railroad Safety Major concerns of both proponents and opponents of FELA are the safety of the railroad industry and the effect that FELA has on safety. Obviously, the costs to railroads and the suffering of workers from workplace injury can be reduced to the extent that more effective safety programs can be implemented. A summary, of the trends in railroad safety is presented and the railroad industry is compared in terms of safety with other modes and industries. FELA's potential effects on the level of safety are discussed in Chapter 4. TRENDS IN RAILROAD ACCIDENT AND INJURY LEVELS Overall, the level of safety in the railroad industry has clearly improved during the last 15 years. There are problems with railroad data that make exact comparisons with other modes and industries problematic. Nevertheless, the railroad industry appears to be no more dangerous than other transportation modes and less dangerous than other indus- tries and industrial segments of the economy. The railroad data presented in this appendix come from various editions of Accident/Incident Bulletin produced by the Office of Safety of WGR

Railroad Safety • 169 the Federal Railroad Administration (FRA), U.S. Department of Transportation. Data on other modes and industries were provided by the Bureau of Labor Statistics of the U.S. Department of Labor. Accidents and Incidents After a dramatic decline from peak levels in the late 1970s, accident and incident levels in the railroad industry have remained stable since the mid-1980s.' Figures A-i and A-2 show the levels of train accidents and incidents and nontrain incidents for 1976 to 1991 [data from FRA (1982, 1987-1992)]. Part of the decline in accident and incident levels is due to a de- cline in the number of train-miles and employee-hours in the railroad industry and part is due to an improvement in safety. After the acci- dent and incident numbers are normalized for changes in railroad activity, however, safety improvement trends remain clear, and recent levels are fairly stable. Figure A-3 shows train accidents and incidents adjusted for the number of train-miles in each year, and Figure A-4 shows the number of nontrain incidents per million employee-hours in each year. 16,000 14,000 1 12,000 0 - 10,000 8,000 Ci) 6,000 4,000 2,000 0 -h- 1976 1978 1980 1982 1984 1986 1988 1990 — Train Accidents Train Incidents FIGURE A-i Train accidents and incidents, all railroads.

50,000 40,000 U) 4-. C 30,000 0 C 20,000 10,000 0 —I 1 I 1 I I 1 1 I 1976 1978 1980 1982 1984 1986 1988 1990 FIGURE A-2 Nontrain incidents, all railroads. 25 Cl) ci) 20 CO C 15 0 _10 ci) a- ci) 05 E z AP 1976 1978 1980 1982 1984 1986 1988 1990 — Train Accidents ± Train Incidents FIGURE A-3 Train accidents and incidents per million train-miles.

Railroad Safety • 171 I- 0 50 ci) ci) > 0 040 E uJ 030 ci) ii 1976 1978 1980 1982 1984 1986 1988 1990 FIGURE A-4 Nontrain incidents per million employee-hours. Fatalities In line with the decline in railroad accidents and incidents, employee fatalities have trended downward during the last 16 years as well, so that in 1991 there were fewer than one-third the number in 1978. Figure A-5 shows railroad fatalities adjusted for employment levels. The pattern is more erratic than that for accidents, but it clearly shows a trend toward alevelof 1.4 to 1.5 fatalities each year per 10,000 workers. Injuries and Illnesses Injuries and occupational illnesses have declined significantly over the 16-year period 1976 to 1991, both in total numbers and on a per- employee basis.2 Total injuries fell from over 58,000 in 1976 to about 20,000 in 1991. In the period 1986 to 1991, injuries stabilized at 7 to 8 per 100 full-time workers compared with the range of 11 to 13 injuries per 100 workers in the late 1970s. (See Figure A-6.) Comparison Among Major Railroads Among the major railroads, safety performance is by no means uni- form. Table A-i presents average fatality and injury data for the period

L 0.005 0.030 2 0.025 G) 0 0.020 0 0 0.015 a) (L 0.0001 1 1 1 I I I I I 1976 1978 1980 1982 1984 1986 1988 1990 FIGURE A-5 Railroad employee fatalities per 100 full-time workers. Cn 14 a) 12 10 a) U) () U) U) (D 6 D4 C co U) W2 C 0 1976 1978 1980 1982 1984 1986 1988 1990 FIGURE A-6 Employee injuries and illnesses per 100 full-time workers.

Railroad Safety • 173 TABLE A-i Average Employee Fatality and Injury Rates for Selected Railroads, 1987-1991 RAILROAD RATE RAILROAD RATE Employee Fatalities per Employee Fatalities per 100 Full-Time Workers Million Train-Miles Santa Fe 0.022 Norfolk Southern 0.102 Burlington Northern 0.019 CSX Transportation 0.091 Norfolk Southern 0.018 Southern Pacific 0.077 CSX Transportation 0.017 Santa Fe 0.077 Total without Amtrak 0.016 All railroads 0.075 All railroads 0.016 Total without Amtrak 0.075 Total with Amtrak 0.015 Total with Amtrak 0.073 Southern Pacific 0.015 Conrail 0.069 Union Pacific 0.012 Burlington Northern 0.068 Conrail 0.012 Amtrak 0.060 Amtrak 0.010 Union Pacific 0.049 Employee Injuries per Employee Injuries per 100 Full-Time Workers Million Train-Miles Burlington Northern 10.9 Conrail 39.3 Union Pacific 8.1 Amtrak 38.9 All railroads 7.6 Burlington Northern 38.3 Total without Amtrak 7.1 All railroads 36.4 Total with Amtrak 7.0 Union Pacific 33.3 Conrail 6.7 Total with Amtrak 33.1 Santa Fe 6.3 CSX Transportation 32.7 Amtrak 6.3 Total without Amtrak 32.6 CSX Transportation 6.1 Southern Pacific 27.9 Southern Pacific 5.5 Norfolk Southern 27.8 Norfolk Southern 4.9 Santa Fe 22.3 SOURCE: Data are from the Accident/Incident Bulletin published by the Office of Safety of the Federal Railroad Administration, U.S. Department of Transportation, for calendar years 1987 through 1991 (No. 156 through 160). 1987 through 1991 for the major freight railroads and the National Railroad Passenger Corporation (Amtrak). The totals with and without Amtrak are for the seven freight railroads with and without Amtrak. Data for "all railroads" are those for all railroads reported by the FRA and used for Figures A-i through A-b. There is a significant disparity in employee fatalities and injuries among the carriers, with the worst rates twice or more than the best rates for three of the four comparisons. The average for all railroads tends to fall in the middle of the ranges for the large carriers except for injuries per million train-miles, for which the major railroads tend to

174 Compensating Injured Railroad Workers Under FELA have a somewhat better record than the industry as a whole. This is likely an artifact of averaging non-train-related injuries over the high levels of train-miles for the long-haul major carriers. Presumably, the railroads with the lower injury rates are experienc- ing lower costs and their workers less suffering. Railroads with higher injury rates may have cost-effective opportunities to improve safety, regardless of what injury compensation law is in effect. COMPARISON WITH OTHER MODES AND INDUSTRIES The railroad industry compares well with other modes and industries on a safety basis. There are indications, however, that the data are unreliable and that better reporting practices in the future may provide more meaningful comparisons. The issue of data reliability is exam- ined in the next section. Comparison of Injury Frequency Railroads compare very favorably with other transportation modes on the incidence of both total injury and illness cases and lost-workday cases.3 Table A-2 gives the average total cases and lost-workday cases per 100 full-time workers for 1981 to 1990. On a total-case basis, only the pipeline mode has lower injury rates than do railroads, with truck- ing, air, water, and transit all higher. For lost-workday cases, trucking, air, and water have higher rates, whereas transit, the transportation sector as a whole, and pipelines have lower rates. Figure A-7 shows lost- workday cases per 100 full-time workers for 1981 to 1990 for the various transportation modes. Railroads also compare very favorably with other industries on the basis of both total and lost-workday cases. For total cases per 100 workers, construction, agriculture, manufacturing, mining, and the private sector as a whole have higher injury rates than the railroads. The only sectors with lower rates are the service sectors: retail and wholesale trade, services, and finance, insurance, and real estate. For lost-workday cases, the data are similar, though mining and manufac- turing have slightly lower rates than railroads. Figure A-8 gives lost- workday cases per 100 full-time workers for 1981 to 1990 for railroads and the nonservice sectors. Comparison of Injury Duration The duration of injuries in the railroad industry compares positively with that for other modes and industries until 1989. At that time, the

TABLE A-2 Injury and Illness Cases and Lost-Workday Rates for Selected Industries and Transportation Modes, 1981-1990 AVERAGE AVERAGE INDUSTRY OR MODE NUMBER INDUSTRY OR MODE NUMBER Total Cases per 100 Lost-Workday Cases per Full-Time Workers 100 Full-Time Workers Construction 14.8 Trucking 8.3 Trucking 13.8 Air 7.8 Air 13.6 Water 7.1 Water 12.3 Construction 6.6 Agriculture 11.5 Agriculture 5.8 Manufacturing 11.5 Railroad 5.6 Local transit 9.4 Local transit 5.4 Mining 9.0 Transportation 5.1 Transportation 8.7 Manufacturing 5.0 Private industry 8.2 Mining 5.0 Railroad 8.0 Private industry 3.8 Retail trade 7.6 Wholesale trade 3.6 Wholesale trade 7.4 Retail trade 3.2 Services 5.3 Services 2.5 Pipeline? 3.6 Pipeline? 1.4 Finance, insurance, 2.0 Finance, insurance, 0.9 real estate real estate Lost-Workdays per 100 Lost-Workdays per Full-Time Workers Lost-Workday Case Water 260.4 Water 36.4 Trucking 201.1 Mining 28.0 Mining 139.3 Trucking 24.5 Construction 130.8 Local transit 22.8 Local transit 124.1 Transportation 21.4 Air 120.3 Construction 19.7 Transportation 108.9 Pipeline? 19.7 Railroad 106.0 Railroad 19.1 Agriculture 94.4 Private industry 18.0 Manufacturing 91.0 Manufacturing 17.9 Private industry 68.2 Finance, insurance, 17.2 Wholesale trade 61.2 real estate Retail trade 51.0 Services 17.1 Services 43.9 Wholesale trade 16.8 Pipelines" 29.9 Agriculture 16.2 Finance, insurance, 16.0 Retail trade 16.0 real estate Air 15.1 "Except natural gas pipelines. SouRcE: Data are from the Bureau of Labor Statistics of the U.S. Department of Labor.

U) 0 Q 6 0 0)4 0 U) (i) CO CT 0 0 i I I I I 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 Modes Transit ± Truck .'( Water Air " All Transportation Rail FIGURE A-7 Lost-workday injury and illness cases per 100 full-time workers by transportation mode. 8 CO 0)2 U) Cci 01 OH I 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 Industries Agriculture ± Mining ,':r Construction Manufacturing ,( Transportation • Rail FIGURE A-8 Lost-workday injury and illness cases per 100 full-time workers for selected industries.

Railroad Safety • 177 lost workdays per lost-workday case increase dramatically. The General Accounting Office (GAO) identified weaknesses in railroad reporting procedures in 1989 that may be responsible for the change in reported lost workdays after that time. The average data in Table A-2 for 1981 to 1990 for both lost workdays per 100 workers and per lost-workday case show railroads with lower rates than those for other transportation modes, but higher rates than those for other industries except mining and construction. As noted, these averages hide a dramatic shift in 1989 and 1990 in the lost workdays per lost-workday case. (See Figures A-9 and A-b.) Rail- roads generally have the lowest rates among the modes (Figure A-9) and among the nonservice industries (Figure A-b) until 1989. By 1990, railroads had the highest rate of any industry. However, the uniformity of the lost workdays per case in Figure A-10 suggests that other industries may have systematic inaccuracies among their re- ported data as well. Assuming that the recent railroad data are the most accurate, there are at least two possible interpretations, one favorable to proponents of FELA and one favorable to opponents. In the first instance, the data could indicate that injuries in the railroad industry, although not more frequent than in other industries (Figures A-7 and A-8), are more severe and therefore result in more days lost from work. On the other 50 ci) CO ( 40 Ca 20 0 . 10 0 —J 0 I 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 Modes Transit ± Truck ,'( Water Air "( All Transportation Rail FIGURE A-9 Lost workdays per lost-workday case by transportation mode: injury and illness cases.

178 • Compensating Injured Railroad Workers Under FELA 40 a) U) CU ( 3 U) >, ca 20 I.- 0 0 I 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 Industries Agriculture ± Mining ';. Construction Manufacturing Transportation Rail FIGURE A- 10 Lost workdays per lost-workday case for selected industries: injury and illness cases. hand, injury severity might be the same as that for other nonservice industries, but the FELA negotiation and settlement process leads to a slower return to work for injured railroad workers, resulting in more lost workdays per case. Because there are no independent data on the severity of injuries in the railroad industry, a definitive finding on this point cannot be made. QUALITY OF THE DATA Questions have arisen concerning the reporting practices of individual railroads, specifically whether they meet FRA standards for reporting accidents and incidents and whether there is enough oversight at the individual railroad level by FRA inspectors. GAO investigated the accident and injury reporting process within the railroad industry for the 1987 calendar year and found that several of the railroads examined had serious deficiencies in their reports (GAO 1989). GAO examined a sample of five railroads and found underreporting of the number of accidents, the number of injuries, and the number of workdays lost to injuries.

Railroad Safety • 179 Underreporting of accidents ranged from 8.3 to 42.8 percent at three of the five railroads, and accident damage was also underreported in many cases. According to GAO, the level of damage in many of the reports was based on initial field estimates that were later revised upon closer inspection of the equipment in repair shops. This explains some of the underreporting of accidents since the decision to report an accident is based on field estimates that were later revised upward above the accident threshold ($5,200 in 1987). Although such revision requires a revised report to the FRA, apparently such revisions were rarely submitted. Of the five railroads examined, GAO found that three had under- reported the number of injuries that should have been reported by FRA standards and that one did not have the records to determine whether underreporting had occurred. The percentage of unreported cases ranged from 4.3 to 14.3 percent of reportable cases. Concerning lost workdays due to injuries, GAO found significant underreporting of these data. For the five railroads, the actual number of workdays lost was 269 percent greater than the number reported for a sample of 156 injuries. Reported days lost were 2,176, whereas the actual number of days lost was 8,023. Failure to update initial estimates and coordinate reports among claims and safety personnel was cited by GAO as the cause of the underreporting (GAO 1989, 17-18). The Association of American Railroads (AAR) provided the study committee with an unpublished rebuttal to some of the GAO findings. The AAR takes issue with the methods of devising the underreporting percentages and of selecting the sample of railroads, and therefore questions the representativeness of the results concerning the fre- quency of accidents and injuries for the entire industry. With regard to lost workdays, the AAR points out that although underreporting may occur in monthly reports because of uncertainties over prospective lost workdays, year-end summaries correct for any omissions. NOTES Train accidents are defined by the dollar amount of damage done in the accident. The accident threshold, which was $6,000 in 1991, is adjusted each year for inflation. If an accident does not exceed the threshold but causes a reportable injury, it is an incident. Wherever the term "injury" is used in this appendix, it is meant to include occupational illnesses unless it is specifically indicated otherwise. The term "cases" is used here purposefully rather than "claims." The Bureau of Labor Statistics data measure the incidence of all reportable injuries and illnesses. Some of these injuries may not result in the filing of a claim for compensation.

180 • Compensating Injured Railroad Workers Under FELA REFERENCES ABBREVIATIONS FRA Federal Railroad Administration GAO General Accounting Office FRA. 1982. Accident/Incident Bulletin No. 150. U.S. Department of Transporta- tion, June. FRA. 1987. Accident/Incident Bulletin No. 155. U.S. Department of Transporta- tion, June. FRA. 1988. Accident/Incident Bulletin No. 156. U.S. Department of Transporta- tion, July. FRA. 1989. Accident/Incident Bulletin No. 157. U.S. Department of Transporta- tion, June. FRA. 1990. Accident/Incident Bulletin No. 158. U.S. Department of Transporta- tion, June. FRA. 1991. Accident/Incident Bulletin No. 159. U.S. Department of Transporta- tion, July. FRA. 1992. Accident/Incident Bulletin No. 160. U.S. Department of Transporta- tion, July. GAO. 1989. Railroad Safety: FRA Needs to Correct Deficiencies in Reporting Injuries and Accidents. RCED-89-109. Washington, D.C., April.

APPENDIX B . .. S 55•S• •IS•SS •S• S SSS••I••S Terminology Terminology to describe various aspects of injury compensation varies among state workers' compensation systems and between those systems and the railroad industry. The following definitions have been adopted to clarify the meaning of the terms used in this report. Medical payments are payments for the medical cost of treating the injury. This may involve longer-term medical rehabilitation costs as well as acute care for the initial injury. Indemnity payments are payments for all other losses associated with the injury. These may include wage loss, prospective wage loss, compensation under a schedule for a permanent impairment (e.g., loss of a hand), vocational rehabilitation costs, and pain and suffering. All of these losses are compensable under FELA, but some are not under state workers' compensation systems. FELA may also include payments for future medical rehabilitation costs if the worker does not expect to be eligible for continued coverage under the railroad health plan (e.g., the worker does not return to the job). Claims settled with medical payments only are claims that are settled without any indemnity payments, though medical costs of the injury are covered. In workers' compensation, these are often 181

182 • Compensating Injured Railroad Workers Under FELA called "medical-only claims." FELA claims in this category differ from workers' compensation claims. Under workers' compensa- tion, claims in this category are minor injuries that result in no lost time or for which the lost time is less than the state minimum. Under FELA, the same injuries may be included in this category, or they may result in an indemnity payment because there is no minimum lost-time requirement in FELA and because losses other than wages are compensable (e.g., pain and suffering). Furthermore, serious injuries for which the worker receives no indemnity settlement are included in this category (e.g., claims in which there is a defense verdict in a trial). These would be covered by workers' compensation assuming they are valid, work-related injuries. Claims settled with indemnity payments are claims for which an indemnity settlement is received by the injured worker. As noted above, any injury covered by workers' compensation that meets the minimum work time lost would fall in this category. The same injury in the railroad industry, however, may fall in the previous category if for some reason the worker does not receive an indem- nity settlement. Similarly, injuries with no lost time or less than the minimum that would be "medical-only" claims under workers' compensation might receive indemnity payments under FELA and therefore would be listed in this category. These claims will also include medical payments for the injury, which are calculated separately from those in the "medical payments only" category. All claims is the sum of the prior two categories, claims settled with medical payments only and claims settled with indemnity pay- ments. When per-claim figures are presented in an "all claims" category of a table, the denominator is this total of all the claims settled in the year. When medical payments are at issue, the numerator will include both payments in the medical-only cate- gory and medical payments for claims settled with an indemnity payment. When indemnity payments are at issue, the numerator will not change between indemnity claims figures and those for all claims since by definition no indemnity payments are associated with medical-payments-only claims. Only the denominator will change. Direct claim settlement status means that the claim is settled directly between the railroad and the employee without intervention by counsel. Represented claim settlement status means that the claim is settled after an employee has retained an attorney to pursue the claim,

Terminology • 183 generally because the employee and the railroad cannot reach agreement through direct negotiations. Litigated claim settlement status means that the claim is settled after a suit has been filed. This does not mean that the case has gone to trial since most litigated cases are settled through negotiation before trial.

Study Committee Biographical Information Clinton V. Oster, Jr., Chairman, is Professor at the School of Public and Environmental Affairs at Indiana University. Dr. Oster received a B.S.E. from Princeton University, an M.S. from Carnegie-Mellon University, and a Ph.D. from Harvard University. His current research focuses on airline economics, international aviation, airport and airway infrastructure, and the environmental impacts of airline and airport operations. Dr. Oster served on the Transportation Research Board (TRB) Study Committee that in 1991 produced Special Report 230: Winds of Change: Domestic Air Transport Since Deregulation. He also served on the Office of Technology Assessment Advisory Panel on Federal Aviation Research and Technology. Dr. Oster served as Research Director for the President's Aviation Safety Commission and is former Associate Dean of the School of Public and Environmental Affairs, former Director of Indiana University's Transpor- tation Research Center, and has been a Research Fellow at the Harvard Business School. W. Bruce Allen is Professor of Public Policy and Management and Trans- portation at the University of Pennsylvania. He received his bachelor's degree from Brown University and his doctorate from Northwestern Uni- versity. Dr. Allen was Lecturer of Economics at Roosevelt University from 1966 to 1968 and served as an economist in the Office of the Secretary at the U.S. Department of Transportation from 1972 to 1973. His fields of research include the impact of transportation regulation on economic efficiency and industrial organization, and public policy. Dr. Allen is past Chairman of TRB's Freight Transportation Section and Freight Transpor- tation Planning and Marketing Committee and was Chairman of the Com- mittee for the Study of the Effects of Regulatory Reform on Technological Innovation in Marine Container Shipping. He also serves on TRB's Surface Freight Transportation Committee. John Arcudi is a Commissioner of the Connecticut Workers' Compensa- tion Commission. He received his bachelor's degree and doctorate from Yale University. He has practiced law in Bridgeport, representing unions and as a plaintiffs' attorney, and was a part-time unemployment commis- sioner in the Fourth District. Mr. Arcudi served as Chairman of the Con- necticut Workers' Compensation Commission from 1976 to 1991. He has been the principal writer of the Connecticut Compensation Review Divi- sion Opinions. 184

Study Committee Biographical Information • 185 RichardJ. Butler is Professor in the Department of Economics at Brigham Young University. He received his bachelor's degree from the University of Utah and his doctorate from the University of Chicago. He has taught at the New York State School of Industrial and Labor Relations at Cornell University, and has been a Senior Research Economist for the National Council on Compensation Insurance. James R. Chelius is a Professor and Chairman of the Department of Industrial Relations and Human Resources for the Institute of Manage- ment and Labor Relations of Rutgers University. He received his bachelor's and master's degrees from the University of Illinois and his doctorate from the University of Chicago. He has taught at the George Washington Uni- versity; the Krannert Graduate School of Management, Purdue Univer- sity; and the New York State School of Industrial and Labor Relations, Cornell University. He has also been an economist for the National Com- mission on State Workmen's Compensation Laws and a Research Director for the Bureau of Economic Research at Rutgers University. Frank J. Dooley is Assistant Professor in the Department of Agricultural Economics of North Dakota State University. He received his bachelor's degree from St. John's University, his J.D. from the University of North Dakota, and his doctorate from Washington State University. He has taught at Moorhead State University, has been a transportation economist for the Upper Great Plains Transportation Institute of North Dakota State Uni- versity, and has served on the Executive Committee for the Mountain Plains Consortium, one of 10 University Transportation Centers. John T. Dunlop is an economics educator, a former U.S. Secretary of Labor, and Lamont University Professor Emeritus of Harvard University. He received his A.B. and doctorate from the University of California. He began his teaching career at Harvard in 1938. He served as Chairman of the Department of Economics (1961-1966) and as Dean of the Faculty of Arts and Sciences (1970-1973) and was appointed Lamont University Professor in 1970 and Professor Emeritus in 1985. He has served as Presi- dent of the Industrial Relations Research Association and the International Industrial Relations Association. He was a public member of the Wage Stabilization Board during the Korean War. Other service included Presi- dential Emergency Boards No. 109, No. 130, and No. 167 and the Presi- dential Railroad Commission, the Missile Sites Labor Commission, the President's Committee on Equal Employment Opportunity, Director of the Cost of Living Council, and Chairman of the Pay Advisory Committee. He is Chair of the Commission on the Future of Worker-Management Rela- tions. He is a member of the American Academy of Arts and Sciences, the

186 • Compensating Injured Railroad Workers Under FELA American Philosophical Society, a senior member of the Institute of Medi- cine, and a life member of the National Academy of Arbitration. Evelyn S. Ferris is an attorney and currently serves as Executive Assistant to the Director of the Oregon Department of Insurance and Finance. She received her doctor ofjurisprudence degree from Willamette University. Her previous experience includes deputy district attorney, municipal court judge, and partner in the law firm of Brand, Lee, Ferris & Embick. Ms. Ferris was appointed a member of the Oregon Legislative Interim Commit- tee on Business Climate and the Governor's Task Force on Liability. She was Chairman of the Oregon Workers' Compensation Board from 1982 to 1989 and President of the International Association of Industrial Accident Boards and Commissions in 1992-1993. Deborah R. Hensler is Senior Social Scientist and Director of the Institute for Civil Justice at RAND and Professor of Law and Social Science at the University of Southern California Law Center. She has been a member of the Institute for CivilJustice's research staff since its establishment in 1979 and served as its Research Director from 1985 to 1990. She has taught at the RAND Graduate School, University of Chicago Law School, and Uni- versity of California at Los Angeles School of Public Health. She received her bachelor's degree from Hunter College of the City University of New York and her doctorate from Massachusetts Institute of Technology. Her research focuses on civil litigation, injury compensation, and alternative dispute resolution. John L. Melvin is Vice President of Medical Affairs at the MossRehab Hospital and Chairman of the Department of Physical Medicine and Reha- bilitation for the Albert Einstein Medical Center, in addition to being Professor and Deputy Chairman for the Department of Physical Medicine and Rehabilitation at the Temple University College of Medicine. He received his bachelor's, M.D., and master's degrees at Ohio State Univer- sity. Dr. Melvin taught at Ohio State University from 1966 to 1973 and was Professor and Department Chairman at the Medical College of Wisconsin form 1973 to 1991. He has been a consultant to numerous U.S. govern- ment agencies and health care institutions. Prior work has included re- search with the Ohio State Research Foundation, Associate Coordinator of the Ohio State Regional Medical Program, and Medical Director of the Curative Rehabilitation Center, Milwaukee. He has served on the Board of Directors of the Milwaukee Visiting Nurses Association, as a member of the Milwaukee Mental Health Planning Council, the Wisconsin Council on Developmental Disabilities, and the planning and evaluation committee of the Elizabethtown Hospital for Children and Youth, as an advisor to the National Multiple Sclerosis Society and as trustee, Vice Chairman, and Chairman of the Easter Seal Research Foundation.

Study Committee Biographical Information • 187 Herbert R. Northrup is Professor Emeritus of Management, the Wharton School, University of Pennsylvania, and author of numerous studies of railway labor relations and public policy. He has been a consultant on personnel and labor relations problems to many companies and an arbitra- tor in labor relations disputes. He received his bachelor's degree from Duke University and his master's degree and doctorate from Harvard Univer- sity. Dr. Northrup taught at Wharton for 27 years and previously taught at Cornell and Columbia Universities. He also spent 12 years in industry, including service as Vice President of Industrial Relations for Penn-Texas Corporation and an employee relations manager for General Electric Company. He is a member of the American Economic Association, the Industrial Relations Research Association, and the American Arbitration Association. Gary T. Schwartz is Professor of Law at the School of Law, University of California at Los Angeles. He was a Special Advisor to the American Law Institute Project on Enterprise Responsibility for Personal Injury; he is now an Advisor to the Restatements of Products Liability and Apportionment. He received his bachelor's degree from Oberlin College and hisJ.D. from Harvard University. Dr. Schwartz was a law clerk on the U.S. Court of Appeals and a legal assistant to the Under Secretary of the U.S. Depart- ment of Transportation. He has taught at the University of Michigan School of Law, the University of Southern California Law Center, the Stanford University Law School, and the University of Virginia Law School. He also has been a Visiting Fellow at Hebrew University and at the Center for Socio-Legal Studies at Oxford University. Douglas F. Stevenson is Executive Director of the National Council of Self- Insurers. He received his bachelor's degree from DePauw University and his doctorate from Harvard Law School. For some 30 years he was with the law firm of Rooks, Pitts & Poust in Chicago. He left in 1984 to start the firm of Stevenson, Rusin and Friedman, specializing in employniiit problems and workers' compensation defense. He served for 20 years as Chairman of the Joint Employers Committee on Workers' Compensation Legislation, testifying in state legislatures and before congressional committees. He is a member of the National Fraud Commission, the Workers' Compensation Committee of the National Conference of State Legislatures, and the Board of Editors of several worker compensation publications. Wayne K. Talley is a Professor of Economics, holding the designation of Eminent Scholar, at Old Dominion University. He has been a faculty fellow at the U.S. Department of Transportation's Transportation Systems Cen- ter, an industry economist at the Interstate Commerce Commission, a faculty fellow at the National Aeronautics and Space Administration, a

188 • Compensating Injured Railroad Workers Under FELA visiting fellow of the Transport Studies Unit at Oxford University, and a visiting research fellow of the Centre for Transport Policy Analysis at the University of Wollongong, Australia. Dr. Talley has written numerous articles in the areas of urban transportation, maritime transportation, and transportation costing and is the author (or coauthor) of the books Intro- duction to Transportation, Transport Carrier Costing, Ocean Container Transpor- tation: An Operational Perspective, and Causes and Deterrents of Transportation Accidents: An Analysis by Mode (forthcoming). Dr. Talley is active in TRB and the Transportation Research Forum, serving as an associate editor of the Journal of the Transportation Research Forum. He holds degrees from the University of Richmond and the University of Kentucky. Edward M. Welch is a member of the faculty of the School of Labor and Industrial Relations of Michigan State University. He was the Director of Michigan's Bureau of Workers' Disability Compensation from 1985 through 1990 and before that a claimants' attorney. He received his bachelor's and master's degrees and his law degree from the University of Michigan. He was elected a charter member for workers' compensation of the National Academy of Social Insurance and has served as Vice President of the International Association of Industrial Accident Boards and Commissions. He publishes the magazine On Workers' Compensation and is the author of numerous books and articles.

The Transportation Research Board is a unit of the National Research Council, which serves the National Academy of Sciences and the National Academy of Engineering. The Board's purpose is to stimulate research concerning the nature and performance of transportation systems, to disseminate the information produced by the research, and to encourage the application of appropriate research findings. The Board's program is carried Out by more than 330 committees, task forces, and panels composed of more than 3,900 administrators, engineers, social scientists, attorneys, educators, and others con- cerned with transportation; they serve without compensation. The program is supported by state transportation and highway departments, the modal administrations of the U.S. Department of Transportation, and other organizations and individuals interested in the development of transportation. The National Academy of Sciences is a private, nonprofit, self-perpetuating society of distinguished scholars engaged in scientific and engineering research, dedicated to the furtherance of science and technology and to their use for the general welfare. Upon the authority of the charter granted to it by the Congress in 1863, the Academy has a mandate that requires it to advise the federal government on scientific and technical matters. Dr. Bruce M. Alberts is president of the National Academy of Sciences. The National Academy of Engineering was established in 1964, under the charter of the National Academy of Sciences, as a parallel organization of outstanding engineers. It is autonomous in its administration and in the selection of its members, sharing with the National Academy of Sciences the responsibility for advising the federal government. The National Academy of Engineering also sponsors engineering programs aimed at meeting national needs, encourages education and research, and recognizes the superior achievements of engineers. Dr. Robert M. White is president of the National Academy of Engineering. The Institute of Medicine was established in 1970 by the National Academy of Sciences to secure the services of eminent members of appropriate professions in the examination of policy matters pertaining to the health of the public. The Institute acts under the responsibility given to the National Academy of Sciences by its congressional charter to be an adviser to the federal government and, upon its own initiative, to identify issues of medical care, research, and education. Dr. Kenneth I. Shine is president of the Institute of Medicine. The National Research Council was organized by the National Academy of Sciences in 1916 to associate the broad community of science and technology with the Academy's purpose of furthering knowledge and advising the federal government. Functioning in accordance with general policies determined by the Academy, the Council has be- come the principal operating agency of both the National Academy of Sciences and the National Academy of Engineering in providing services to the government, the public, and the scientific and engineering communities. The Council is administered jointly by both the Academies and the Institute of Medicine. Dr. Bruce M. Alberts and Dr. Robert M. White are chairman and vice chairman, respectively, of the National Research Council.

TRANSPORTATION RESEARCH BOARD National Research Council 2101 Constitution Avenue, N. W. Washington, D.C. 20418 ADDRESS CORRECTION REQUESTED .- OOQO1Oj * ( Robert N Smith s.st-.-Nat±s--Engr NON-PROFIT ORG. U.S. POSTAGE PAID WASHINGTON, D.C. PERMIT NO.8970