Click for next page ( 181


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 180
APPENDIX B Retirement-Income-Related Data Sets This appendix describes federally-sponsored data sources of the following types: panel surveys of cohorts of individuals; other panel surveys of families and households; repeated cross-sectional surveys of households; and administrative and survey data sources on employers and employees. PANEL SURVEYS FOR COHORTS OF INDIVIDUALS Asset and Health Dynamics Among the Oldest Old (AHEAD) Ongoing (Hurd et al., 1994~: Panel survey of people aged 70 and older at the time of the original interview and their spouses. Collects detailed information on demographic back- ground, dissaving and Medicaid eligibility, family structure and in-kind and fi- nancial transfers, housing, income and assets, physical and cognitive health, use of community and nursing home services and out-of-pocket service costs; future questionnaires will be more similar between AHEAD and HRS (e.g., both will include employment status), and the section on medical care costs and sources of health insurance coverage will be expanded. Data will be linked with Medicare files, SSA earnings and benefits files, state Medicaid files (if possible), and the National Death Index. First interview conducted in 1993-1994, second interview in 1996; 8,223 respondents, drawn from both the HRS screener sample with an oversample of blacks, Hispanics, and Florida residents and from the Medicare Master Enroll- ment File for people aged 80 and older. Funding has been approved for 2 more interviews and to introduce a cohort in between HRS and AHEAD of people and 180

OCR for page 180
RETIREMENT-INCOME-RELATED DATA SETS 181 their spouses, beginning in 1999 when they are aged 69-75. Conducted by the University of Michigan for NIA. Health and Retirement Survey (HRS) Ongoing (Juster and Suzman, 1995; see also Rust, 1993; Smith, 1988~: Panel survey of people aged 51-61 at the time of the original interview and their spouses. Collects comprehensive information on demographic background, disability, employment status and job history, family structure and transfers, health and cognitive conditions and status, health insur- ance and pension plans, housing, income and net worth, retirement plans and perspectives, and attitudes, preferences, expectations, and subjective probabili- ties; includes experimental modules. Data will be linked with Medicare files, SSA earnings files, and the National Death Index; health and pension benefit plan descriptions were obtained from employers in the first interview. First interview conducted in 1992, second interview in 1994, third interview in 1996; 12,654 respondents with an oversample of blacks, Hispanics, and Florida residents. Funding has been approved for one more interview; funding has also been approved to collect data on a new cohort of people and their spouses, beginning in 1998 when they are aged 51-56. (The same size for the HRS and AHEAD new cohorts will be about 7,000 people.) Conducted by the University of Michigan for NIA. Longitudinal Study of Aging (LSOA) Ongoing (National Center for Health Statistics, 1992~: Based on two studies, the Supplement on Aging (SOA) to the 1984 National Health Interview Survey (NHIS) and the Second Supplement on Aging (SOA II) to the 1994 NHIS. Everyone 55 years of age and older in a 1984 NHIS household was eligible for the SOA; participants aged 70 and older were subsequently interviewed by telephone in 1986, 1988, and 1990 to obtain infor- mation about changes in their functional status and living arrangements. SOA II involved interviewing 10,000 people in the NHIS who were aged 70 and older, beginning in October 1994 through March 1996. Elderly people in black house- holds were oversampled. Participants will be followed up every 2 years, begin- ning in 1997, for one or more waves. Data linked with Medicare records and the National Death Index. Conducted by the Census Bureau for NCHS and NIA. Medicare Current Beneficiary Survey (CBS) Ongoing (Adler, 1994~: Rotating panel survey of Medicare enrollees. First panel began in fall 1991; the original sample of about 15,000 beneficiaries was continued with replenishment each fall for several years. The rotating design began in 1994; a new panel is introduced each fall. Sample members are retained in a panel for 4 years and interviewed 3 times per year; about 16,000 cases are included in each interview. Collects information on health status, access to care, satisfaction with care, and health insurance in the first fall interview; collects information on health care utilization and cost in subsequent interviews and on income and assets each spring. Survey

OCR for page 180
182 ASSESSING POLICIES FOR RETIREMENT INCOME data are linked to Medicare claims records (except for people enrolled in health maintenance organizations, for which no claims records exist). National Longitudinal Survey of Mature Women (NLS-MW) Ongoing (Center for Human Resource Research, 1995a, 1995b; Fahy, 1995~: Panel survey of women aged 30-44 at the time of the original interview; some data collected also for spouse. Collects detailed information (not always at every interview) on current labor force and employment status; characteristics of current or last job, including some characteristics of employer; work experience prior to initial inter- view in 1967; work experience since most recent interview; demographic back- ground, including parents' characteristics; migration; education and training; health and physical condition; marital and family characteristics, including care- giving responsibilities; net family assets and family income by type; attitudes and perspectives (e.g., how feels about job); retirement expectations and self and spouse retirement eligibility; hypothetical job offer; volunteer work; local area labor market conditions. Descriptions were obtained from employers of sample members' pension plans in 1989; information was obtained on post-retirement health insurance coverage for self and spouse in 1992. First interview conducted in 1967; sample of 5,000 women, including over- sample of blacks, followed at 1- to 3-year intervals; most recent interview in 1995 when the women were aged 58-72. Conducted by the Census Bureau for the Ohio State University Center for Human Resource Research under a contract with BLS. National Longitudinal Survey of Older Men (NLS-OM) Completed (Center for Human Resource Research, 1995a, 1995b; Fahy, 1995~: Panel survey of men aged 45-60 at the time of the original interview, who also reported information for their wives. Collected detailed information (not always at every interview) on current labor force and employment status; characteristics of current or last job; work experience prior to initial interview in 1966; work experience since most recent interview; demographic background, including parents' characteristics; migration; education and training; health and physical condition of self and spouse; marital and family characteristics; net family assets and family income by type; attitudes and perspectives (e.g., how felt about job); retirement expecta- tions and retirement eligibility; hypothetical job offer; volunteer work; local area labor market conditions. 1990 reinterview collected information on social sup- port networks and long-term care use and financing. First interview conducted in 1966; sample of 5,000 men, including over- sample of blacks, followed at 1- to 2-year intervals until 1983 when they were aged 62-77; surviving members of the sample (or next-of-kin for decedents) reinterviewed in 1990 when they were aged 69-84. Conducted by the Census Bureau for the Ohio State University Center for Human Resource Research under a contract with DOL; NIA provided funding for the 1990 reinterview.

OCR for page 180
RETIREMENT-INCOME-RELATED DATA SETS 183 National Longitudinal Survey of Young Men (NLS-YM) Completed (Center for Human Resource Research, 1995a, 1995b; Fahy, 1995~: Panel survey of men aged 14-24 at the time of the original interview; some data collected also for spouse. Collected detailed information on current labor force and employment status; characteristics of current or last job; work experience prior to initial inter- view in 1966; work experience since previous interview; demographic back- ground, including parents' characteristics; migration; education and training; health and physical condition; marital and family characteristics; net family as- sets and family income by type; military service; attitudes and perspectives (e.g., how felt about job); hypothetical job offer; local area labor market conditions. First interview conducted in 1966; sample of 5,000 men, including over- sample of blacks, followed at 1- to 2-year intervals from 1966 to 1981 when they were aged 29-39. Conducted by the Census Bureau for the Ohio State University Center for Human Resource Research under a contract with DOL. National Longitudinal Survey of Young Women (NLS-YW) Ongoing (Center for Human Resource Research, 1995a, 1995b; Fahy, 1995~: Panel survey of women aged 14-24 at the time of the original interview; some data collected also for spouse. Collects detailed information (not always at every interview) on current labor force and employment status; characteristics of current or last job, including some characteristics of employer; work experience prior to initial inter- view in 1968; work experience since most recent interview; demographic back- ground, including parents' characteristics; migration; education and training; health and physical condition; marital and family characteristics, including child care arrangements; net family assets and family income by type; attitudes and perspectives (e.g., how feels about job); hypothetical job offer; volunteer work; local area labor market conditions. First interview conducted in 1968; sample of 5,000 women, including over- sample of blacks, followed at 1- to 3-year intervals; most recent interview in 1995 when they were aged 41-51. Conducted by the Census Bureau for the Ohio State University Center for Human Resource Research under a contract with BLS. National Longitudinal Survey of Youth (Young Men and Women) (NLSY) Ongoing (Center for Human Resource Research, 1995a, 1995b; Fahy, 1995~: Panel survey of people aged 14-21 at the time of the original interview; some data collected also for spouse. Collects detailed information on current labor force and employment status and characteristics of current or most recent job; work experience since January 1, 1978, or previous interview; characteristics of jobs with more than 10 or 20 hours per week and more than 9 weeks in duration since January 1, 1978, or previous interview; demographic background, including par- ents' characteristics; migration; education, vocational training, and government jobs or training programs; health and physical condition; marital and family characteristics; family income, assets, and program participation; military ser

OCR for page 180
184 ASSESSING POLICIES FOR RETIREMENT INCOME vice; work attitudes; educational and occupational aspirations and expectations; other social and psychological variables; discrimination problems in employ- ment; significant others; school discipline; delinquency, drugs, and alcohol use; reported police contacts; time use; child care arrangements; looking for work; local labor market conditions. First interview conducted in 1979; sample of 13,000 followed annually until 1994 and biennially beginning in 1996; most recent interview in 1994 when they were aged 29-36. An oversample of people in the military was discontinued after 1984; an oversample of economically disadvantaged whites was discontinued after 1990; the current sample size is about 10,000. Conducted by the National Opinion Research Center (NORC) for the Ohio State University Center for Hu- man Resource Research under a contract with BLS. National Long-Term Care Survey (NLTCS) Ongoing (Manton, Corder, and Stallard, 19931: Panel survey of Medicare eligible people age 65 and older and chronically disabled at the time of the original interview with refreshment, to- gether with a repeated cross-section survey of disability incidence among Medi- care eligible people age 65 and older. For disabled and institutionalized people, collects detailed information on demographic background, health care services used, housing and neighborhood characteristics, income and assets, medical con- ditions and recent medical problems, out-of-pocket medical care payments and other sources of payment, physical and cognitive functional limitations, and ser- vices (formal and informal) received by impaired individuals and source of pay- ment. Linked to Medicare service records. First interview conducted in 1982; 35,000 people were screened to determine disability status (defined by one ADL or IADL impairment of 90 days or more duration), and the 6,400 people who were identified as chronically disabled, community residents received detailed in-home interviews. These disabled people and also people identified in 1982 as living in institutions were reinterviewed in 1984. Also, in 1984,12,100 community residents who were not disabled in 1982 were rescreened to determine disability status in 1984, and a new sample of 4,900 people who turned age 65 between 1982 and 1984 was drawn from Medicare files and screened. Disabled and institutionalized people identified in these screenings received detailed interviews. In 1989, sample members identified as chronically disabled or institutionalized in 1982 or 1984 were reinterviewed in 1989. Also, 5,000 community residents who were not disabled in 1984 were rescreened, and a new sample of 4,900 people who turned age 65 between 1984 and 1989 was drawn from Medicare files and screened. Disabled and institutionalized people identified in these screenings received detailed interviews. The 1994 survey followed the same pattern as the 1989 survey; in addition, oversamples were drawn of people aged 95 and older and nonwhites. Conducted by the Census Bureau for NIA, Duke University, and HHS/ASPE; HCFA and AHCPR provided funding for earlier rounds.

OCR for page 180
RETIREMENT-INCOME-RELATED DATA SETS 185 New Beneficiary Survey (NBS) and New Beneficiary Followap (NBF) Com- pleted (U.S. Social Security Administration, 1993, 1994~: Two-wave panel sur- vey of people who began receiving Social Security benefits in a 12-month period in 1980-1981 and of people aged 65 or older at that time who were covered by Medicare but not receiving Social Security benefits. First interview (NBS) conducted in 1982; sample of 9,100 retired workers, 5,200 disabled workers, 2,400 wife or widow beneficiaries, and 1,400 Medicare recipients not receiving Social Security. Collected information on demographic characteristics, marital and childbearing history, employment history, current income and assets, and health. Second interview (NBF) of survivors and spouses of decedents conducted in 1991. Collected information on economic circumstances, health, family con- tacts, post-retirement employment, and changes in life circumstances that might explain changes in economic status (e.g., death of a spouse, episode of hospital- ization, change in residence). Disabled worker beneficiaries were asked about efforts to return to work, experiences with rehabilitation services, and knowledge of SSA work incentive provisions. Data were linked with selected information from SSA, SSI, and Medicare records. Conducted by Temple University for the Social Security Administra- tion. NIA, HCFA, ASPE, AHCPR, and the Office of the Assistant Secretary for Health also contributed funding support. Retirement History Survey (RHS) Completed (U.S. Social Security Administra- tion, no date; see also Rust, 1993~: Panel survey of people aged 58-63 at the time of the original interview, including men and women with no husband in the household (considerable information was also collected for wives and widows of male sample members). Obtained information on respondent's work experience, retirement and retirement plans, family composition and support patterns, expen- ditures, social and leisure activities (in most interviews), health status, health care utilization, health care expenditures, migration, assets and debts (except in 1973, which had a shortened interview), income sources and amounts, spouse's work experience, labor force participation and financial experience of recent widows. Data were linked with SSA earnings and benefits records. First interview in 1969; sample of 11,150 people selected from the Current Population Survey and followed every 2 years until 1979 when they were aged 68-73. Conducted by the Census Bureau for the Social Security Administration. OTHER PANEL SURVEYS OF FAMILIES AND HOUSEHOLDS National Survey of Families and Households (NSFH) Completed (Sweet and Bumpass,1996~: Interviewed a sample of 13,000 families in 1987-1988, includ- ing an oversample of minorities and households containing single-parent fami

OCR for page 180
186 ASSESSING POLICIES FOR RETIREMENT INCOME lies, stepfamilies, recently married couples, and cohabiting couples. In each household, a randomly selected adult was interviewed; also, a shorter, self- administered questionnaire was completed by the spouse or cohabiting partner of the primary respondent. Collected information on detailed household composi- tion, family background, adult family transitions, couple interactions, parent- child interactions, education and work, economic and psychological well-being, and family attitudes. The original sample was reinterviewed in 1992-1994. Personal interviews were conducted of primary respondents and their spouse or cohabiting partner. Telephone interviews were conducted with children and a randomly selected parent of the main respondent; proxy interviews were obtained with the spouse or other relative of a deceased or seriously ill original respondent. Collected infor- mation on life history since the original interview, health and well-being, family process, kinship, social support and interhousehold exchanges, and current labor force involvement, income sources, assets, and debts. Conducted by Temple University for the University of Wisconsin under a grant from NIA and NICHD. Panel Study of Income Dynamics (PSID) Ongoing (Duncan and Hill, 1991; Hill, 1992~: Follows a sample of families on an annual basis (beginning in 1997, the survey will be biennial). Collects the most detailed information about family heads and, since the late 1970s, about wives and cohabitors. The core content includes demographic characteristics; current employment and employment his- tory in past year; annual income by type for the head (most detail), and other family members; in-kind public assistance; estimate of federal taxes paid; hous- ing value, mortgage and property taxes, or rent; average weekly expenditures on food; financial assistance to people living elsewhere; housework time; geographic mobility; socioeconomic background; health; religion; military service; county labor market conditions. Supplements asked occasionally have covered a wide range of topics, such as disability and illness, extended family and kinship ties, fringe benefits, inheritances, retirement plans and experiences, retrospective his- tories, and wealth. In 1990, there were some links to Medicare records. The current sample size is 9,200 families, including original sample families who were first interviewed in 1968, new families formed from them (e.g., by adult children), and a sample of Hispanic families added in 1990. A sample of immigrant families will be added in 1997. Conducted by the University of Michigan with funding from NSF and other agencies. Survey of Income and Program Participation (SIPP) Ongoing (Citro and Kalton, 1993; Jabine, King, and Petroni,1990~: Repeated panel survey of samples of families interviewed at 4-month intervals. The core content includes demo- graphic characteristics; monthly information on labor force participation, job characteristics, and earnings for each person age 15 or older; monthly informa

OCR for page 180
RETIREMENT-INCOME-RELATED DATA SETS 187 lion on public and private health insurance coverage; and monthly (or 4-month) information on detailed sources and amounts of income from about 65 sources of cash income and seven in-kind programs for each person age 15 or older. Topical modules asked once or twice in each panel cover annual income and taxes; educational financing and enrollment; eligibility for selected programs (includ- ing expenditures on shelter, out-of-pocket medical care costs, and dependent care); employee benefits (1984 panel only); housing costs and finance; individual retirement accounts; personal history; retirement expectations and pension plan coverage; financial and property assets and liabilities. A topical module with variable content in each panel responds to needs of policy analysis agencies. Topics covered to date include characteristics of job from which retired; child care; child support; functional activities; health status and use of health care; home health care; long-term care; and others. The 1984, 1990, and 1991 panels were exactly matched with Social Security earnings and benefit records for the use of SSA analysts; other panels have been exactly matched with IRS earnings data for evaluation of data quality within the Census Bureau. From 1983-1993, the survey followed samples of adults for 32 months; a new panel was introduced each February with an initial sample size of from 12,500 to 23,600 families. Under a redesign introduced in 1996, the survey is following a sample of adults and children in about 37,000 families for 48 months; a new panel will be introduced every 4 years. The 1992 and 1993 panels will continue to be followed at annual intervals from 1997-2002 in the Survey of Program Dynamics (SPD), which was mandated by the Personal Responsibility and Work Opportunity Act of 1996. The SPD will focus on gathering informa- tion with which to monitor and evaluate the effects of changes in social welfare programs on families with children. Conducted by the Census Bureau. REPEATED CROSS-SECTIONAL SURVEYS OF HOUSEHOLDS Consumer Expenditure Survey (CEX) Ongoing (Jacobs and Shipp, 1990): In- terview Survey component interviews 5,000 consumer units each quarter. House- hold addresses remain in sample for 5 quarters.] Collects demographic character- istics; work experience of consumer unit members aged 14; detailed quarterly information for expenditure categories that comprise an estimated 60-70 percent of total expenditures; global or usual quarterly expenditures for categories that comprise an additional estimated 20-25 percent of total expenditures; previous 12 months' expenditures for some items; inventory of real assets at initial interview and changes in ownership and mortgages each quarter; financial assets and 1Surveys like the CEX and CPS have a panel component in that sample units are interviewed more than once. However, the sample units are addresses, not household members. Whoever is residing at the address is interviewed; people who move are not followed.

OCR for page 180
188 ASSESSING POLICIES FOR RETIREMENT INCOME changes in them over past 12 months at 5th interview; income in prior 12 months at 2nd and 5th interviews; taxes at 2nd and 5th interviews. Diary Survey component obtains 2-week expenditure diaries from 6,000 consumer units spread out over the year. Interview and Diary information are combined to form estimates of total expenditures. Conducted by the Census Bureau for BLS. Current Population Survey (CPS) Ongoing (Bureau of the Census, 1996; Pen- sion and Welfare Benefits Administration, 1994, 1995b): Interviews 60,000 households each month about previous month's labor force participation, job characteristics, and demographic characteristics (the sample was recently cut to 50,000~. Household addresses remain in the sample for 4 months in one year and the same 4 months a second year. The March income supplement also includes military in civilian housing and an additional sample of 2,500 housing units that had contained at least one adult of Hispanic origin as of the preceding November interview. The March income supplement collects information on labor force participa- tion and job history in the prior calendar year for each person aged 15 or older; annual income for the prior calendar year for each person aged 15 or older by source (about 30 types of regular cash income); participation in noncash benefit programs; health insurance coverage. The March 1973 and March 1978 CPS files were exactly matched with Social Security earnings records and made pub- licly available. Later March CPS files have on occasion been matched with IRS earnings records for evaluation of data quality within the Census Bureau. The April (or May) employee benefits supplement (administered in 1972 and every 5 years beginning in 1979) collects information for workers about em- ployer-provided pension and health care plan coverage; reasons for pension noncoverage for workers in companies with plans; years of service with em- ployer; size of employer; annual earnings and family income; whether covered by union contract; type of pension plan; self-reported vested status; percent of pay contributed to 401(k)-type plan; percentage of employer match; whether ever received, amount, and uses of lump-sum payments; whether and amount contrib- uted last year to IRA; whether self and/or family covered by employer health plan, spouse's employer plan, or other plan; for experienced unemployed work- ers aged 25-64, whether covered by former employer's health plan; for covered workers aged 46 and over, expectations about availability of health care coverage after retirement. (This supplement may be discontinued in favor of using the pension and health care plan coverage information in SIPP.) The September 1994 supplement collected information on pension benefits received by retirees and health care insurance coverage of retirees. It asked detailed questions of people aged 40 and over about receipt and amounts of pension annuity and lump-sum benefits based on former employment and about continuation of employment-based health insurance coverage during retirement.

OCR for page 180
RETIREMENT-INCOME-RELATED DATA SETS 189 The supplement updated information collected in December 1989 on pension benefit amounts and in August 1988 on retiree health insurance coverage. Conducted by the Census Bureau for BLS; other agencies provide funding for supplements. National Medicare Expenditure Survey (NMES); renamed Medical Expendi- ture Panel Survey (MEPS) Ongoing (Hunter and Arnett, 1996~: 1987 survey interviewed a sample of 14,000 households, including oversamples of blacks, Hispanics, people aged 65 and older, low-income people, and people with func- tional limitations. Households were interviewed 5 times between February 1987 and July 1988 about their utilization, expenditures, and sources of payment for all major forms of medical care and demographic and socioeconomic characteristics. Surveys were also conducted of physicians and health care facilities providing care to members of the household sample during the period and of employers and insurance companies responsible for their insurance coverage. The 1987 NMES also included an institutional survey of 13,000 residents of nursing and personal care homes. Similar surveys were conducted in 1980 and 1977. Beginning in 1996, the survey will be conducted on a continuing basis as a repeated panel survey. Samples of families will be interviewed over 2-year periods, with a new panel introduced every year. The sample will be drawn from households in the Na- tional Health Interview Survey. Conducted by Westat, Inc., for AHCPR, HCFA, and NCHS. Survey of Consumer Finances (SCF) Ongoing (Avery and Elliehausen, 1986; Avery, Elliehausen, and Kennickell, 1987; Kennickell and Shack-Marquez, 1992; Kennickell and Starr-McCluer, 1994; Kennickell and McManus, 1994; see also Curtin, Juster, and Morgan, 1989~: First conducted under the current design in 1983 and every 3 years thereafter (predecessor surveys were conducted as early as 1963~. Originally included a panel component: the 1986 SCF was a reinterview of the 1983 sample, and the 1989 SCF included a subset of the 1983 sample in addition to a new cross-sectional sample. However, the 1983 survey data were not used because of quality problems, and, for cost reasons, the 1992 and 1995 surveys have not repeated the panel feature of the 1989 survey. Col- lects comprehensive information on household wealth from a household sample together with a sample of high-income households drawn from Internal Revenue Service files who agree to participate; total sample size is about 3,000-4,000 households. In most years, has obtained descriptions of sample members' pen- sion plans from their employers. The 1983, 1986, and 1989 surveys were con- ducted by the University of Michigan for the Federal Reserve Board; the 1992 and 1995 surveys were conducted by the National Opinion Research Center for the Federal Reserve Board.

OCR for page 180
190 ASSESSING POLICIES FOR RETIREMENT INCOME DATA SOURCES ON EMPLOYERS AND EMPLOYEES Employee Benefits Survey (EBS) Ongoing (Bureau of Labor Statistics, 1992a, 1994; see also MacDonald, 1995~: Obtains descriptions of various types of benefits offered to full-time and part-time workers in specified occupations by employers from a sample of about 6,000 nonfarm private sector establishments and state and local governments (mid- and large-size private employers with 100 or more employees are surveyed in one year and small private employers and state and local governments the next year). Tabulates numbers of workers cov- ered by or participating in various benefits by occupation. (For private sector employees, the categories are professional, technical, and related; clerical and sales; production and service. For public employees, the categories are regular employees; teachers; police and firefighters.) No tabulations are provided by number of employees, and tabulations by employer characteristics are very lim- ited (geographic region and 1-digit SIC code). The EBS began in 1979 with surveys of benefits provided to full-time work- ers in medium and large establishments (100 or 250 or more employees, depend- ing on the industry), with limited coverage in the service sector. The 1987 survey examined benefits provided to full-time employees in state and local govern- ments with 50 or more employees; the 1988 and 1989 surveys covered full-time employees in medium and large establishments with 100 or more employees in all industries; the present system began in 1990. Work is under way to integrate the EBS more fully with the Employment Cost Trends Survey and the Occupa- tional Compensation Survey Program, which obtains compensation data for spe- cific areas. Conducted by BLS. Employment Cost Trends Survey (ECT) Ongoing (Bureau of Labor Statistics, 1992b): Serves as the basis for the Employment Cost Index (ECI), which provides quarterly estimates of changes in wage and benefit costs and annual estimates of the level of wage and benefit costs, developed from samples of establishments and occupations. The ECT uses the same sample as the EBS. Conducted by BLS. Form 55/)/)1Jata System: Ongoing (Pension and Welfare Benefits Administra- tion,1995a): Private employers are required to file information annually with the IRS about pension and welfare benefit plans. (Fringe benefit plans are a class of welfare benefit plans that include group legal services, cafeteria plans, and educa- tional assistance. Other types of welfare benefit plans are medical and dental insurance, life insurance, apprenticeship and training, scholarship funds, sever- ance pay, disability, etc.) Employers with pension or welfare plans with 100 or more participants must file Form 5500 and relevant schedules; employers with pension or welfare plans with fewer than 100 participants file a shorter Form 5500-C/R; the self-employed with a pension plan that has only one participant or

OCR for page 180
RETIREMENT-INCOME-RELATED DATA SETS 191 one participant and that participant's spouse file Form 5500EZ. Plans for which forms are not to be filed include unfunded and/or uninsured welfare plans with fewer than 100 participants (i.e., plans that are paid for from general assets); unfunded pension or welfare plans whose benefits go only to a select group of management or highly compensated employees; and model simplified pension plans, which include most simplified employee pension (SEP) plans. Also, one- time or special "window" plans are not documented. Information collected includes type of plan; number of participants (plans with 100 or more participants also report the number of active participants, re- tired and separated participants entitled to future benefits, beneficiaries of de- ceased participants, participants who left the employer with less than 100 percent vesting); whether plan ever amended or terminated; total number of employees and those excludable from plan for various reasons; asset amounts by type; liability amounts by type; net assets; income by type; expenses by type; contribu- tions by employers and employees during the year; actuarial assumptions used as basis for plan funding. DOL/PWBA makes available data files that contain the universe of filers of Forms 5500, 5500-C, and 5500-R, including pension and welfare benefit plan filers. These files have been edited to some extent by the IRS for consistency of financial information and arithmetic errors. DOL/PWBA also makes available a research database of all pension plans filing Form 5500 and a weighted, 5 percent sample of all pension plans filing Form 5500-C. These data have been further edited to identify such problems as truncation of financial entries by multi-billion dollar plans and incorrect entries for such items as type of plan. Corrections are made on the basis of analysts' judgment and are not verified with the filer. The database contains some recoded variables, including designation of pension plans as primary or supplemental in the case of employers with multiple plans. Researchers have sometimes prepared exact-match files of the Form 5500 pension plan research database with financial information abstracted by Compustat from annual reports filed with the SEC. A similar research database does not exist for welfare benefit plan filers. DOL/PWBA indicates that inconsistencies in welfare plan reporting severely limit the usefulness of Form 5500 data on these plans for research purposes. Longitudinal Research Database (LRD) Ongoing (Center for Economic Stud- ies, 1995; Troske, 1995~: Contains annual data beginning in 1972 for nearly all large U.S. manufacturing plants from the Annual Survey of Manufactures and data every 5 years for all manufacturing plants from the Census of Manufactures, linked over time to create a panel data set. Includes such information as total value of output, cost of materials, investment, and employment. LRD data have been exactly matched with data on workers from the 1990 census long form, including demographic characteristics, occupation, and in- come by type, to form the Worker-Establishment Characteristics Database

OCR for page 180
92 ASSESSING POLICIES FOR RETIREMENT INCOME (WECD). The WECD contains records for about 200,000 manufacturing work- ers matched to 16,000 establishments. The match rate varied by worker and plant characteristics; weights have been created to adjust for these differences. Plans are to expand the WECD to all workers, which can be done if the proposed American Community Survey (the equivalent of asking the long-form questions throughout the decade) goes forward. Data for other economic sectors would come from an exact match with the Longitudinal Business Database (LBD) that is under development at the Census Bureau. Maintained by the Center for Economic Studies at the Census Bureau; acces- sible only by special arrangement. National Employer Health Insurance Survey (NEHIS) Ongoing (Hing et al., 1995; Hing and Fuller, 1996; Hunter and Arnett, 1996; Sommers, Chapman, and Moriarity, 1996~: First conducted in April-December 1994 in order to provide needed data for the health care reform policy debate (planning work begun in June 1993; final estimates to be released in early 1997~. Will be conducted on an annual basis, beginning in 1997, and coordinated in terms of questionnaire con- tent with the Health Insurance Plans Survey (HIPS) component of MEPS. The ongoing NEHIS sample size will be about 25,000 establishments. Collects information from employers on whether they provide health insur- ance now and in last 5 years; employer characteristics, including type of business, number of branches or locations, number of full-time and part-time employees, wage rate distribution of employees, total payroll, spending on health care; eligi- bility requirements for employee health care coverage (including for retirees); other benefits offered; characteristics of health insurance plans offered (or a subsample in the case of employers with many plans), including type of plan, number of people enrolled (employees, retirees, dependents), whether self-in- sured, total premiums, benefits paid for claims incurred, administrative costs, employee contributions to premiums, deductibles, coinsurance rates, maximums, waiting period for pre-existing conditions, types of services covered. The 1994 survey included 39,000 establishments (36,000 private sector, 3,000 public sector). Sample drawn from Dun and Bradstreet file of private establishments, 1992 Census of Governments, and national sample of self-em- ployed individuals from the 1993 NHIS. 1994 NEHIS conducted by Westat for NCHS; ongoing NEHIS conducted by the Census Bureau for AHCPR, HCFA, and NCHS.