Engineering Economic Analysis Practices for Highway Investment (2012)

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.

12 chapter two EnginEEring Economic AnAlysis Policy And ProcEdurAl guidAncE There are a number of policy statements and guidelines that govern application of EEAs to highway investments nation- wide. There are also research reports that explain methods and procedures available for use. The references are issued primarily by the federal government [e.g., Office of Manage- ment and Budget (OMB) and U.S.DOT/FHWA] and national organizations (e.g., AASHTO and NCHRP). Several of these references will also be cited in the case examples in chap- ter three. Specific state, regional, or local guidance is not included here; however, pertinent references at the state and regional levels will be included as appropriate with each case example in chapter three. • Presidential Executive Order 12893, Jan. 26, 1994: Principles for Federal Infrastructure Investments. This presidential order applies to “federal spending for infrastructure programs,” encompassing “direct spend- ing and grants for transportation,” water resources, energy, and environmental protection. Among the requirements are the following (Executive Order 12893 Jan. 26, 1994): – A systematic analysis of expected benefits and costs of the investment, including quantitative and qualita- tive measures. – Use of discounted CBA “over the full life cycle of each project.” – Recognition of potential uncertainty in estimates of the amounts and timing of costs and benefits, and where these costs and benefits are important, use of appropriate quantitative and qualitative risk manage- ment techniques. – Definition of a comprehensive set of alternatives for evaluation and comparison, including but not lim- ited to “managing demand, repairing facilities, and expanding facilities.” – A series of recommendations addressing the efficient management of this infrastructure. – Consistency with OMB Circular A-94 (discussed next). • OMB Circular A-94: Guidelines and Discount Rates for Benefit–Cost Analysis of Federal Programs. This document was revised in 1992 and includes a recom- mended discount rate of 7% for constant-dollar BCAs. Appendix C of the Circular, which is updated annually, includes rates for cost-effectiveness analyses in which project benefits do not have to be explicitly stated. The implied discount rates using Circular Appendix C data are lower; for example, the 30-year real interest rate on Treasury securities is 2.7% (as of Dec. 2009) (OMB Circular A-94). Several agencies discussed in the case examples have used these Appendix C data, combined with state-specific considerations, in arriving at a dis- count rate for their analyses. • FHWA Life-Cycle Cost Analysis Primer. This guide provides practical information on properly setting up and performing an LCCA between alternative invest- ments. It covers agency costs and road user costs, dis- cusses elements important to setting up the analysis (e.g., length of analysis period, timing of actions, and expenditure stream diagrams), and describes the com- putations involved. The Primer dispels misconceptions and confusions often held about LCCAs by explain- ing the difference between remaining service life and salvage value, between an inflation rate and a discount rate, and between economic and financial analyses. It recommends that the analysis be done in constant dol- lars. The Primer closes with a discussion of issues and reservations regarding LCCA, and suggestions on how an agency may deal with them (Life-Cycle Cost Analy- sis Primer Aug. 2002). • FHWA Economic Analysis Primer. This guide addresses the broader topic of economic analysis, looking at the role of economic analysis in highway decision making and the benefits of its use, explanations of LCCA and BCA, methods to conduct risk analysis, and EIA. Explanations are included for how to handle inflation and the use of differential inflation in a constant-dollar analysis, the concept of the opportunity cost of money as represented in a discount rate, and the importance of including road user costs in an economic analysis (Economic Analysis Primer Aug. 2003). • User Benefit Analysis for Highways. This AASHTO document explains methods to compute the benefits to road users, generally categorized as savings in travel- time costs, vehicle operating costs (VOC), and accident costs. The manual describes how to evaluate different highway improvements that affect user benefits, how to analyze user benefits in each of the three components, how to conduct benefit–cost calculations, and applicable software (User Benefit Analysis . . . Aug. 2003). • These sources represent general guidance relevant to highway investments. The case examples in chapter

13 three will cite additional federal and state laws, guide- lines, and research studies applicable to the particular case at hand. • A compilation of federal laws beyond those cited in the case examples, which call for economic analyses of highway investments, is contained in Appendix D to illustrate the breadth of these requirements for eco- nomic analyses. Other reports of interest include documents by the General Accountability Office (GAO) that have reviewed aspects of highway program management, including use of economic methods; guidance documents by the U.S. Army Corps of Engineers regarding BCA; NCHRP Reports and Synthe- ses of Highway Practice covering specific topics that call for economic analyses; and FHWA case examples on state, regional, and local transportation agency use of economic methods such as the FHWA Highway Economic Require- ments System—State Version (HERS-ST). The federal and national-level agencies and organizations that have been discussed in this section maintain websites with informa- tion framing their perspectives on EEA. Please note that the guidance discussed here relates only to requirements for eco- nomic analyses. Agencies must also comply with other, non- economic requirements on proposed highway investments; for example, provisions governing environmental protection, mitigation of environmental impacts, and environmental jus- tice, to name a few other areas of policy guidance. TExTs on EnginEEring Economic AnAlysis A number of texts are available that readers may consult for information on methods, techniques, and parameters used in EEA. These texts provide comprehensive coverage of a wide range of topics and issues. They guide readers in under- standing the basics, but also address unique or illustrative problems and situations. These texts have been consulted for particular topics addressed in this chapter, including the equivalence in decision results when the various methods of EEA are correctly applied, and the interpretation of the dis- count rate. Cited texts are listed in the References at the end of the report. Other texts, which have provided background information, are listed in the Bibliography. Many of these texts present methods and analyses in large part from a private-sector perspective. Their explanations and interpretations must therefore be adapted to investments in the public realm. Some texts include separate chapters that focus specifically on governmental expenditures for public works; this material would relate more closely to the analy- sis of highway investments. Topics oriented to the public sector might include, for example, discussion of benefits to the public, problem solutions entailing multiple perspec- tives, multi-criteria or multi-objective decision making, and decisions under budget constraints. Textbook material could be reviewed together with guidance from highway-specific documents, such as the FHWA primers described in the pre- vious section, to place subjects in a transportation context. One other observation is that the nomenclature in the texts (as well as in other documents, including some of the guide- lines discussed in the preceding section) may diverge from the definitions in chapter one. A key example is that the term “interest rate” is used many times to refer to “discount rate.” This report will continue to use discount rate as distinct from interest rate to avoid ambiguity between the two, and to maintain a clear distinction between economic and financial analyses. mEThods supporting investment decisions In their textbook on engineering economy, Grant et al. cite several questions that an engineer might raise regarding a potential investment (Grant et al. 1990, p. 4): • Why do this at all? • Why do it now? • Why do it this way? • Will it pay? The first three, credited to General John J. Carty, chief engineer of the New York Telephone Company in the early twentieth century, are essentially more detailed consider- ations leading up to the overarching final question. These questions invoke thought processes that engineers need to go through in determining how to meet an identified need or problem, recognizing that in civil works there are many choices in the type, magnitude, timing, and location of investments. With highway systems as examples, these choices exist within programs (e.g., preservation, mobility, and safety) as well as across them. The structuring of the analytic steps needed to address these questions establishes the foundation of the methods used in EEA. This foundation is reflected in the following concepts by Grant et al. (1990, pp. 5–14), with highway-related examples provided by the synthesis author: • Recognizing and Defining Alternatives. Decisions imply selections among alternatives. For example, highway preservation may benefit from preventive maintenance now or corrective maintenance later. Going further, there are choices among capital versus maintenance actions to preserve existing assets. Capital preservation may entail periodic rehabilitation or less frequent but more expensive reconstruction or replace- ment. Ranges of alternative investments may likewise be defined, say, for mobility (congestion mitigation) and safety improvements. An economic approach to these choices helps to identify courses of action that provide worthwhile benefits for the appropriate costs incurred, and that avoid “gilt-edging”: that is, the ten- dency to seek a perfect or ideal solution that is exces- sively costly.

14 ognized in federal policy. Provisions of Executive Order 12893 acknowledge that nonmonetized (or non- market) and qualitative benefits may need to be included in an analysis (Executive Order 12893 . . . Jan. 26, 1994). All of the case studies in chapter three address cost as a monetized consequence. Although most of them also include monetized benefits, some use other approaches. For example, the Bridge Program- ming and Permitting case applies a utility function to characterize the health of bridge assets as a measure of benefit; the Safety Programming case shows instances where the frequency of fatal and serious-injury acci- dents is used as a surrogate for the monetized economic cost of these collisions. • Only differences in consequences matter when comparing alternatives. This principle has several implications: – Only the future matters; past investments or “sunk costs” are no longer relevant in decisions on future investments. – In situations where discounted benefits streams among all investment options are close in value to one another, a BCA may reduce to a simpler, dis- counted cost-minimization problem. This situation may occur when policy or practice results in essen- tially the same performance trend among all can- didate investments (a situation that may arise, for example, in pavement preservation). – An accurate tally of differences in consequences among candidate investments requires the actual costs (or benefits) associated with each option. Average costs (which often result from simplifying assumptions in management systems) or allocated costs (as computed by cost accounting systems) may lead to incorrect economic conclusions. • Separable decisions could be made separately when- ever practicable. Within the highway transportation field this principle often applies to how project alter- natives are defined. A proposed project may combine several types of highway improvements that otherwise could be performed separately; for example, renewed pavement surface condition, new safety hardware, mobility enhancements, and improved drainage features. It is preferable to evaluate each project component incre- mentally and retain only those that pass economic muster (“each tub on its own bottom”), rather than analyzing the entire project in a single step. An economic result for the project as a whole might mask the weak economic performance of one or more components that would not be viable on their own. • Criteria for decision making are needed. Criteria help structure a decision to meet stated policy goals and targets, and to guide a decision maker in dealing with multiple goals and objectives reflecting competing and sometimes contradictory interests. The primary crite- rion of an economic analysis would relate to making the best use of limited resources. Other criteria can • The Need to Consider Consequences. The conse- quences of highway decisions—costs as well as proj- ect impacts, results, or outcomes—have become more familiar to the highway community with the growing importance of performance management and account- ability reporting. The consequences of a decision neces- sarily follow the decision—therefore, EEA necessarily involves projections or forecasts of the future. Although estimates of the future always involve risk or uncer- tainty, the need to forecast is inherent in many engineer- ing calculations (e.g., pavement design, road capacity design, intersection capacity design, safety projections, and infrastructure deterioration modeling) and tools are available to help with decision making in this context. Later sections of this chapter will deal with risk analysis and identifiable issues in forecasting; several case studies in chapter three illustrate techniques to deal with risk and uncertainty. (Both costs and benefits are treated as con- sequences, because in many transportation analyses ben- efits are construed as reductions in costs or costs avoided. Similarly, “negative benefits” or disbenefits affect the analysis in a manner similar to costs.) • The Need for a Viewpoint—Consequences to Whom? In private-sector analyses, the appropriate viewpoint is often that of business owners seeking profit. Public-sector analyses are more complicated— rarely is the intended benefit limited solely to the gov- ernment agency. (The ability of an agency to perform within its budget is a separate matter—one of many rea- sons for distinguishing between financial and economic analyses in chapter one.) With respect to transportation, the viewpoint is generally that of the public at large. However, it will be clear in subsequent sections that even this broad perspective is not always sufficient— the “public” comprises stakeholders and constituents who themselves reflect different viewpoints. Certain chapter three case studies illustrate particular aspects; for example, the federal versus the regional perspective in the Critical Facilities case example, maritime versus regional highway conveyance of freight in the same example, and different values of time associated with different road users in the Accelerated Project Delivery case example. • The Desirability of Commensurable Measures. Eco- nomic costs and benefits that are all monetized can be compared in a straightforward manner—the challenge is to estimate these as accurately as possible, or at least to be aware of simplifying assumptions. The value of com- mensurable measures is illustrated in almost all cases in chapter three, but particularly in the Economics-based Tradeoffs example. For some consequences, however, the use of noncommensurable measures is unavoidable (see also the next item). • Comparisons of alternatives might consider non- monetized as well as monetized consequences. That important consequences of investment decisions may not be able to be monetized or even quantified is rec-

15 resolve the need for investments to serve additional needs and purposes such as the following: – Various aspects of equity: the need to ensure a fair distribution of investments and impacts thereof across the state. Examples might include geographic equity (e.g., maintaining a fair balance of road investments across rural and urban needs); jurisdic- tional equity [e.g., maintaining a fair balance of road investment distributions among the state DOT or the state highway administration, metropolitan planning organizations (MPOs), rural transportation plan- ning organizations (RTPOs), and tribal nations]; and environmental justice issues that require agencies to avoid actions that would impose disproportionately high adverse impacts on human health or environ- mental conditions affecting minority populations and low-income populations. – The need for network continuity and connectivity, ensuring that a reasonable and consistent level of engineering standard and development is maintained along a link or corridor. – The need to serve nonmarket or qualitative objec- tives, particularly those required by federal or state law and stated agency policy; for example, neigh- borhood cohesion, economic opportunity, and envi- ronmental protection. (Some of these objectives may be reflected in the equity considerations discussed earlier.) These criteria can be organized hierarchically or with appropriate threshold values for invocation to provide the intended guidance to decisions. For example, Washington State’s pavement type selection process (discussed in chap- ter four as part of Implementation) applies solely to the pri- mary (i.e., the economic) criterion to type selection if the cost difference between the pavement alternatives is at least 15%. (Other states also use tolerance thresholds: refer to the Pavement Type Selection case in chapter three.) This method ensures that the economic superiority of the selected alterna- tive is robust and not likely to be affected by uncertainties in the estimates. If the cost difference between alternatives is less than 15%, then other, nonmarket or nonquantitative criteria must be considered in selecting the preferred pave- ment type. Grant et al. (1990) caution, however, that verbal descriptions of advantages and disadvantages be structured properly to avoid a form of “word” power double-counting; that is, where the same basic argument is stated in different ways to make one alternative appear to be vastly superior to another. • Even carefully conducted estimates of consequences may turn out to be incorrect. It was stated earlier that projecting the future is not an exact science, and that tools are available to help a decision maker understand the relative strength or weakness of a prediction of con- sequences. One approach is the application of second- ary criteria relating to uncertainty in estimates of future consequences, as described in the preceding item. Another is the application of risk analysis methods, described in a later section. As the chapter three case examples illustrate, agencies that use EEAs extensively invoke an attitudinal as well as an analytic response to addressing risk and uncertainty in their estimates: – Their clear objective is to have economic analysis results that aid their decisions. They resist the temp- tation to create excuses based on the challenges in performing these analyses perfectly, be they issues of data, assumptions of future conditions and behav- iors, or difficulty in identifying all reasonable alter- natives to meeting a need or solving a problem. – They are willing to devote the time and resources to resolving significant issues of uncertainty or risk. Agency guidelines, analytic tools, review procedures, follow-up studies, re-examination of assumptions, staff training, and website “knowledge resources” are among the steps taken by agencies to ensure the best decisions based on the best information. (Chap- ter four elaborates on agency implementation of eco- nomic methods.) • There is a need for a “system viewpoint” as well as one focused on the problem at hand. This principle relates to possible interactions among a group of deci- sions and their respective side effects. In a highway system context, for example, the congestion caused by a work zone for a project may cause temporary traf- fic diversions to other routes. Diversions owing to a single project may not be significant enough to require analysis of its consequences on other routes. However, if multiple projects occur within the region, their com- bined diversion effects may need to be studied, with implications for the scheduling and perhaps work requirements on affected projects. discounted cash Flow methods Computational methods have been developed to perform EEAs according to the principles described earlier. These methods include net present value (NPV), equivalent uni- form annual cost (EUAC), BCA (or B/C), and internal rate of return (IRR). Only a brief commentary on these meth- ods is provided in this synthesis, consistent with the study scope outlined in chapter one. The focus of this report is rather on the application of these methods to support high- way investment decisions as illustrated in the several case examples in chapter three. The methods are well described in the general engineering–economics literature. They are also described in references providing transportation- and highway-specific guidance. Risk analysis methods are cov- ered in the next section. Although the four methods listed entail somewhat differ- ent data and procedures, they all will yield the same deci- sion when applied correctly (Grant et al. 1990, chapters 4–7).

16 This synthesis adopts this position, which is based on the following precepts: • All of the engineering economic methods are based on a LCCA of investment alternatives. The LCCA is a dis- counted cash flow analysis of monetized cost and ben- efit streams. The analysis period or analysis horizon, in years, is sufficiently long to capture a reasonable rep- resentation of the significant costs and benefits of each alternative through equivalent periods of performance. The four methods identified previously (NPV, EUAC, BCA, IRR) all meet these stipulations. Other methods (such as project payback periods) have their uses in other contexts, but generally do not meet these char- acteristics of LCCA. (Project payback periods may or may not depend on discounted cash flows; they do not analyze alternatives through equivalent performance periods; and they therefore do not capture the full rep- resentation of costs and benefits through a performance period.) • Net benefits and net costs are used to assess the dif- ferences in consequences among alternatives. Both costs and benefits may take on positive or negative values, and bookkeeping conventions could be estab- lished to treat the respective quantities consistently and correctly. For example, the consequences of highway investments on passenger and freight travel are mea- sured in road user costs. When comparing alternative investments, reductions in these costs (or avoidance of these costs) are treated as benefits. Conversely, actions that increase road user costs are said to incur disbenefits. Case examples in chapter three involving use of the California DOT’s (Caltrans) Cal-B/C model will illus- trate how these net-value calculations in tallies of dis- counted agency and road user costs are interpreted for the economic analysis results. • In comparing alternative solutions, BCA and IRR analy- sis are both properly done on an incremental basis. There is considerable literature covering both “simple” B/C or IRR and “incremental” B/C or IRR calculations. If a project is being compared solely with a “No-Build” or “Do Nothing” option, the incremental case reduces to the simple case, and both yield the identical result. The same decision on whether or not the investment is economically justified will also be produced by the NPV and EUAC methods. • When there are a number of investment alternatives addressing a particular need or problem, however, the proper approach is to conduct the B/C or IRR analysis incrementally. The simple B/C result can be used as a screen: a simple B/C of less than 1.0 will not bear out on an incremental basis either. However, in the general case that is not subject to a budget constraint, a solu- tion with the highest simple B/C may not necessarily be the optimal solution. Rather, the theoretically opti- mal result is the investment with an incremental B/C exceeding 1.0. risk Analysis methods “Risk” and “uncertainty” both refer to imperfect informa- tion about future consequences. Risk involves an indefinite result, but one whose characteristics are possible to quantify. For example, the time to failure of a signal lamp or roadway luminaire may not be known precisely, but can be estimated statistically from laboratory failure studies. Uncertainty refers to a result that is not known definitely and is difficult to quantify in a practical sense. Analysts often address uncer- tain consequences through a set of assumptions that define one or more scenarios. For simplicity, both risk and uncer- tainty are addressed in the literature using what are referred to as “risk analysis” techniques. One set of risk analysis methods involves varying selected parameter values and assumptions through repeated applica- tions of the economic analysis for each project alternative. The results indicate to what degree the proposed project solu- tions are affected by this variability. Information technology hardware and software enable hundreds or thousands of anal- ysis cycles to be done economically and efficiently. Different methods that adopt this approach include the following: • Bracketed analysis. If only one or a few parameters are to be tested under different values, analysts may define a minimum, most likely, and maximum value for each, and then obtain three sets of results corresponding to the minimum, most likely, and maximum specifications. The result is a bracketed analysis that indicates the real- istically possible range of consequences, minimum to maximum. Analysts may then investigate the economic implications for the preferred solution if future conse- quences are different from the case that is judged “most likely” today. • Sensitivity analysis. A parameter (or related set of parameters, such as vehicle distribution percentages to express highway traffic composition) is varied across a set of values to assess its effect on the solution. The economic analysis is repeated for each value to deter- mine the effect on project consequences and implica- tions for the preferred project solution. • Scenario analysis. An underlying assumption is varied to represent different possible behaviors or situations; for example, shifts in demographic or market characteristics that affect transportation demand or future changes in the price of critical items such as materials that can affect project costs. These varying assumptions are translated into corresponding sets of parameter values, each of which is tested in the economic analysis to determine the effect on project consequences and the implications for the preferred project solution. • Probabilistic analysis. In lieu of providing a single, deterministic value for each parameter, analysts provide an input probability distribution. The details of this input vary according to the software used, but one approach is to specify the type of probability distribution desired

17 (e.g., uniform, normal, or beta) and a set of parameter values that define the key characteristics of the selected distribution. For example, a beta distribution can be estimated with minimum, maximum, and most likely values to describe its shape. The probabilistic input data are subjected to a Monte Carlo simulation within the framework of the economic analysis, producing a probabilistic output distribution that describes the range of project consequences and their expected likelihood. Analysts can use this information to assess the implica- tions for the preferred project solution. • Option risk; threshold or breakeven analysis. When a preferred option has been tentatively identified through an economic analysis, this method helps deci- sion makers to understand whether it is the best alterna- tive: that is, how robustly it can withstand the effects of risk and uncertainty. Each element of risk or uncer- tainty is considered, with the underlying question of how the preferred option is affected by the occurrence of that possibility. A threshold or breakeven analysis can be per- formed to help decision makers visualize the impact of an uncertain event coming to pass. Each parameter of the problem that might be affected by possible risks or uncertainties is considered in turn. Another way to struc- ture the analysis, particularly if some parameters are not quantifiable, would be to determine what level of project outcomes would be needed to justify the cost of build- ing the project. The threshold is the value of the selected parameter that is needed to establish the tentatively selected option as the best among all alternatives. It is also referred to as the breakeven value because it marks the point where the tentatively selected option becomes more cost-effective than all other alternatives. Although this approach helps to quantify aspects of the problem when risk or uncertainty is present, it still requires judg- ment by the analyst to determine if the threshold value can be realistically achieved in light of the possible sources of risk or uncertainty (adapted from Transport Canada Guide to Benefit Cost Analysis 1994). • Economic Factor of Safety. NCHRP Report 551 suggests ways in which standard economic methods and criteria can be applied to guard against solutions approach- ing a threshold of increased risk too closely (paraphrased from Cambridge Systematics Inc. et al. 2006, p. 86): – Use available models that show how road user costs (e.g., travel time, VOC, and accident costs) vary with different levels of condition or performance. Examining the slopes of these curves can reveal the performance level at which user costs begin to increase rapidly. Select performance criteria to avoid approaching this point of rapid user cost increase. – Make use of the law of diminishing marginal returns to identify the point at which additional invest- ments begin to have a declining degree of impact on improvements in performance—in other words, where the slope of the investment–performance curve begins to level off or decline. Curves relat- ing performance trends to levels of investment can be used for this purpose. Alternately, models can be exercised repeatedly at different levels of avail- able budget to see where diminishing returns begin to appear. Select a level of investment that avoids significantly diminished returns. – A corollary to the law of diminishing returns is that caution could be exercised in setting targets that call for 100% achievement of a particular condition or ser- vice level, because the benefit–cost ratio associated with achieving the last 1% increment will typically be quite low. For example, setting a target of no structur- ally deficient bridges could necessitate costly replace- ment of a bridge that currently is not heavily used and that at most provides a redundant link in the network. All of the risk methods illustrated in the case examples are of one of these types. There is a second group of risk methods that deal with the decision criterion to be employed (e.g., minimax or maximin solutions and their variants). These methods are not included in the case examples and are not discussed further in this report. other methodological considerations Discount Rate Both an interest rate and a discount rate reflect a time value of money. Whereas an interest rate is a charge for the use of money, the discount rate represents a different concept: an opportunity cost. When dollars in the public sector are invested in Project A, that money is not available for Project B or for any other public or private purpose. The discount rate thus ensures that the return on Project A in terms of benefits to the public is at least as great as the minimum return that could be gained by investing in alternative investment opportunities (Project B or other options, whether public or private). Sev- eral reasons to define a nonzero discount rate are as follows (paraphrased from Winfrey 1969, pp. 72–73): • The investor—the public—has alternative uses of the highway tax money for dividend or profit-returning investments or for immediate satisfactions that are for- gone when that money is applied to highway investments. This concept of alternative uses forgone is referred to as the opportunity cost. • The timing of proposed disbursements (cash flows) will vary with the type of disbursement (e.g., construc- tion expenditures and maintenance expenses). A math- ematical discount factor is needed to bring all monetary disbursements throughout the analysis period to an equivalent basis. • A zero discount rate (which is proposed occasionally) is misleading because it does not weigh the relative desir- ability of different cash flows over time. It says that a dollar received today would have the same value now as a dollar that is to be received, say, in 10 years would have now.

18 In Winfrey’s view, average interest rates or bond rates are not appropriate discount rates (i.e., the discount rate is not necessarily equal to these interest or bond rates, although the case examples indicate that rates on selected instruments can be part of the calculation of discount rates). The cost of financing or borrowing is not the economic cost to peo- ple who pay taxes to furnish the debt. They do not reflect the worth of money or the worth of a proposed highway improvement (where the “worth” is interpreted as different from “cost”) (Winfrey 1969, p. 76). A complementary rationale for the use of a discount rate is given by Moavenzadeh and Markow (2007, p. 28): • A discount rate represents a time value of money—people tend to prefer having a dollar now to a dollar in the future. • The discount rate reflects the marginal productivity of capital: that is, a dollar’s worth of resources today can be invested productively to yield more than a dollar’s worth of benefits or outputs in the future. • The discount rate reflects the opportunity cost of capi- tal, as discussed earlier. According to OMB Circular A-94, the current discount rate for BCAs (as of Dec. 2009, the period in which this report was prepared) is 7%. Additional data are provided in Appen- dix C of the Circular and updated annually; these values are intended for cost-effectiveness analyses. These data provide real Treasury borrowing rates on marketable securities; ana- lysts could select the rate for securities with a maturity com- parable to the period of their analysis. For projects with an analysis period of 20 to 30 years or longer, these Appendix C rates are 4.4% to 4.5%, considerably lower than the 7% recommended for BCAs. As the case examples in chapter three will show, several agencies use Appendix C data as part of their basis for determining their respective real dis- count rates, resulting in an overall range across all the case examples of 3% to 4% at the lower bound and 7% at the upper bound. This range is consistent with findings of other sources; for example, 4% to 7% for safety BCAs (Hanley 2004), and 2% to 4% cited by FHWA as an appropriate real discount rate given recent trends (Stephanos Nov. 13, 2008: refer to the Pavement Type Selection example in chapter three). The case examples also illustrate specific methods that agencies use to derive values of their discount rate. It is well understood that the selected discount rate can affect the results of an economic analysis: • Higher discount rates reduce the present value of future costs and benefits more rapidly; • Lower discount rates reduce the present value of future costs and benefits less rapidly; • A zero discount rate does not reduce the present value of future costs and benefits at all (that the future costs and benefits have the same present value as current costs and benefits in this case is the reason that a zero discount rate is opposed by Winfrey in the earlier comment). A sensitivity analysis can reveal to what degree the results of the analysis vary with the choice of a particular rate. Even the relatively modest discount rates within the range used in the case examples can effectively suppress the effects of costs and benefits several decades in the future. For example, with a 5% discount rate, future value declines to 50% of current value in 15 years, to 10% of current value in about 50 years, and to 1% of current value in about 95 years. The equiva- lent time frames for the same percentage declines in future value given a 7% discount rate are approximately 10 years, 35 years, and 65 years. This effect becomes important with highway facilities that have very long lives (e.g., major bridges) or strategies with very long-term consequences (e.g., investments to address climate change, with benefits intended for future generations). The use of discounted cash flow anal- ysis in these cases may present an analytical impediment to evaluating fairly those investments that yield consequences in the distant future. An approach to deal with this problem for very long analysis periods (as encountered in sustainabil- ity analyses with multi-generational benefits) entails treating the proposed investment in two phases: the first, to examine long-term sustainability with the objective of conceiving and shaping project candidates that are specifically designed to satisfy these long-term goals; the second, to conduct tradi- tional economic project-evaluation analyses (discounted cash-flow analyses as well as risk analyses) for those alterna- tives appropriate to long-term perspectives that are identified in phase one (Moavenzadeh and Markow 2007, pp. 29–30). Data, Forecasting, and Assumptions Another area of EEAs considered to be problematic by some potential users is the uncertainty of data, forecasting, and assumptions that describe future performance and con- sequences of alternatives. The key concern is that by get- ting these factors wrong, the resulting decision may also be wrong; the value of the economic analysis among those holding these concerns is thus called into question. Certain comments from the screening survey and the challenges in applying economic methods described by FHWA and GAO, which will be presented later in this chapter, provide examples of these perceptions. There is a degree of objective validity to these concerns. For example, research has demonstrated risk, inaccuracy, and bias in forecasts of transportation demand. Transporta- tion demand is an important determinant of the benefits of transportation projects and of revenue projections for toll facilities. In a statistically significant study of traffic forecasts for transportation infrastructure projects among 14 countries, Flyvbjerg et al. (2005) found that 50% of 183 completed road projects in these 14 countries had actual traffic varying by more than ±20% of forecasted volumes, with 25% of the projects having actual traffic volumes varying by more than ±40% of forecasted volumes. When highway projects were compared with rail projects, the evidence of a greater bias in

19 the rail forecasts was evident, owing to a greater degree of competition for rail funding and “deliberately slanted forecasts” that provided optimistic projections of ridership and revenues, accompanied by underestimation of costs (Flyvbjerg et al. 2005, pp. 138–140). A similar phenomenon of optimistic bias within a varied history of forecast results has been reported in NCHRP Synthesis 364 for estimates of toll road demand and revenue on projects in the United States (Kriger et al. 2006). Flyvbjerg et al. recommend two paths by which the reliability and accuracy of forecasts can be improved, depending on the underlying motivation at work. If the motivation is to develop more accurate forecasts, they recommend a newer methodology, reference class forecast- ing (Flyvbjerg et al. 2005, pp. 140–142). If the underlying motivation is to bias results (typically to favor proceeding with the project), the solution is to impose checks and bal- ances on the forecasting process, accompanied by mecha- nisms that promote greater transparency in assumptions, methods, data, and accountability regarding the analytic results. These mechanisms might include the following examples: to subject the forecast to benchmarking, inde- pendent peer review, public disclosure, and engagement by stakeholders and citizens at hearings, citizen juries, and similar forums; and to present the analysis at professional meetings that subject the methods, data, assumptions, and findings to peer review and comment (Flyvbjerg et al. 2005, pp. 142–143). There is no reason to presume that forecasts for engi- neering economic studies are immune from these problems. However, the following points also could be noted: • Forecasting is involved in many engineering appli- cations. Forecasts of travel demand, resulting traffic volume and composition, and vehicle and driver char- acteristics are central to planning and design of new roads as well as to analyses of projects to improve mobility, safety, and preservation of the existing sys- tem. Although errors in these forecasts can result in incorrect estimates of basic highway elements such as the need for the project itself, the proposed number of travel lanes, pavement thicknesses, traffic operations features, and safety appurtenances, such concerns do not remove the need to perform these kinds of analy- ses. Rather, the solution may lie in the types of recom- mendations posed by Flyvbjerg et al. (2005). • Risk analysis methods that were described earlier enable analysts to assess the implications of data estimating and forecasting errors, as well as potentially erroneous assumptions. These methods can help agency staff to craft more flexible or robust solutions that minimize adverse consequences resulting from risk and uncertainty. • Refer also to the discussion of ex-post analyses in chap- ter four, which discuss forecasting bias in appraisal estimates and how a disciplined program of ex-post analyses can result in more accurate forecasted esti- mates in the future. Compelling illustrations of how several agencies put these ideas into practice across a wide range of highway investment decisions are presented in the case examples in chapter three. Agencies that successfully apply EEAs across their business processes strive to develop and maintain the tools and capa- bilities to perform these analyses effectively, but in a practical, commonsense way. These tools and capabilities include effec- tive guidance and executive support, staff knowledge and skills, financial and administrative resources, methods incorporated in business processes, data collection and analysis, information technology systems, and supporting institutional relationships with other organizations. The perspective is fundamentally one of incorporating economic analyses within the normal, rou- tine business processes that an agency must perform to do its job effectively, rather than one of viewing economic analyses as additional, somewhat isolated, adjuncts to other activities. Beyond marshaling the resources, processes, and tools to per- form these analyses, the agencies highlighted in chapter three also promote an organizational culture that supports EEAs in two ways: a demand for the information that these economic studies can provide, and a supply of talent, time, and dollars needed to produce this information. These aspects of organi- zational culture and the tangible and intangible benefits of per- forming EEAs are revisited in chapter four. Network and Locational Aspects Highway transportation is a network-based activity. It is not surprising, therefore, that several case examples make explicit use of network and locational information. How they do this depends on the type and purpose of the highway need to be addressed and the capabilities of the analytic systems avail- able to the performing agency. For purposes of this discussion, “network” information implies spatial data that enable a man- agement system or model to build a topographically correct virtual network mirroring the physical highway system: that is, descriptions of links and nodes. “Locational” information refers to spatial data that describe where a highway segment or point structure is located; for example, by route-milepost, longitude-latitude, grid system, or similar device—but does not include the link-node relationships that describe a fully formed network (e.g., descriptions of connections among links at a node are not provided). Both locational and net- work information enable use of a geographic information system (GIS) for mapping and locating events on the high- way system, but only network information (as used in this report) enables accurate modeling of the flows of people, goods, and vehicles throughout the highway system as needed, for example, in travel demand modeling. This dif- ference between locational and network information is one of the reasons for distinguishing between “network-level” and “program-level” analyses in the Study Approach in chapter one. Although both categories relate to a highway system, the network level implies network information about the system; the program level allows for locational information, but does not imply network information.

20 With these distinctions, the case examples in chapter three fall into the following categories: • Spatial relationships are not critical for certain cases; therefore, locational or network information is optional; for example, for GIS mapping, or if the analytic sys- tem already includes locational or network descrip- tions for other purposes. Pavement type selection is a case example that does not require spatial information other than to locate and identify the particular highway segment. • Asset preservation also falls into this category. Pave- ment preservation studies may benefit from locational information in presentations involving GIS, for exam- ple, but this requirement is not a general one (apart from GIS use, road segment identification may pres- ent the primary need for locational data). The economic analysis in the Pontis™ bridge management system addresses the road user cost of traffic detours resulting from bridge load or clearance restrictions, a network- level effect. All of the information needed to compute detour costs is included as part of the bridge descrip- tion, and therefore does not require any supporting network-level capability (Cambridge Systematics, Inc. 2005). Other types of highway assets vary in the spatial information that is gathered and available for analysis. Some agencies may locate safety and roadside hard- ware fairly precisely through global positioning system technology as part of field data collection; other agen- cies may not maintain any inventory information at all for these types of assets (Markow 2007). The sched- uling of multiple preservation projects in an area may require a network-level study of work-zone congestion effects to avoid gridlock. Such studies would require a network model, but such models are typically external to the systems analyzing the preservation strategy on each highway segment. • The emergence of systematic, low-cost safety improve- ment programs in several states has highlighted the effective use of locational data to diagnose the under- lying causes of clusters of crash locations, and to pro- pose effective, low-cost treatments. Refer to the Safety Programming case example for an illustration of this capability and economic studies before construction and after construction to validate the proposed solu- tion (median rumble strips). • Some engineering economic studies engage network- level effects directly, and a network model (or an eco- nomic analysis model linked to the outputs of a network model) must be used. The Critical Interstate Facilities case example in chapter three illustrates such an applica- tion. The consultant to the Port Authority of New York and New Jersey applied the Interstate Network Analy- sis model to derive the transportation economic ben- efit of each of the Port Authority’s Interstate crossings between New York and New Jersey. Although other case examples in chapter three involving Caltrans make use of the Cal-B/C BCA, none of these examples entails a network-level analysis. The Cal-B/C software is capa- ble, however, of accepting travel-flow information from an external travel demand model and applying it within what is referred to as an extended corridor analysis. With this capability, Cal-B/C is able to analyze interchange projects, bypass projects, and passing-lane projects, in addition to those that do not require a network capa- bility (Booz-Allen & Hamilton Inc. et al. 1999). • These comments are based largely on the practices reflected in the case examples in chapter three. A similar breakout of different needs for, and uses of, locational and spatial information is given in the documentation of the Cal-B/C model. This documentation notes that BCAs typically apply one of the following approaches to deal with network effects (Booz-Allen & Hamilton Inc. et al. 1999): – Route-based approach: project benefits beyond the immediate project area are ignored. – Extended corridor approach: standard assumptions are used to approximate project impacts beyond the immediate project area. – Network-based approach: project benefits are esti- mated using the output of a regional planning model. This source notes that in certain situations it may be appropriate to employ a route-based approach and ignore any off-route benefits as negligible. These cases arise, for example, with route alignment improvements and resurfacing projects, or in rural areas with a lim ited network such that there are few alternative paths to a given route, making off-route project impacts unlikely. • It could also be pointed out that network-based travel demand models are themselves subject to forecast- ing limitations, adding to the issues discussed earlier; specifically, methodological weaknesses and obsoles- cence in the face of demands for newer types of policy analyses. The four-step travel demand model is not inherently behavioral in nature, and therefore may not accurately capture road users’ responses to the types of policy initiatives now being explored by transportation agencies and political bodies. Furthermore, the models more suitably represent aggregate, corridor-level travel behavior, but begin to break down when more dis- aggregate, individual-level responses must be pre- dicted. The latter requirements would be needed, for example, in estimating the time chosen for travel, individual responses to policies such as congestion pricing and telecommuting initiatives, nonmotorized travel, and freight and commercial vehicle movements. Recommendations include development of new travel demand models, together with organizational and insti- tutional changes among affected governmental bod- ies: MPOs, state DOTs, the federal government, and intergovernmental relationships (Metropolitan Travel Forecasting . . . 2007). This topic is recommended for research in chapter five.

21 chAllEngEs To WidEr u.s. APPlicATion oF Economic mEThods government Accountability office Findings Although the guidance documents listed at the beginning of this chapter identify the advantages of using EEAs and encourage wider application, practical experience raises a number of concerns that challenge this effort. Three studies by GAO examine economic methods used to analyze high- way investments. The reports distill a number of concerns about shortcomings in these methods. One study considered the use of BCA in transportation planning (Surface Transportation . . . June 2004), with con- cerns summarized as follows: • BCA tallies net benefits in the aggregate, without regard to the equity of the distribution of these benefits. • Although impacts such as travel time saved, reductions in emissions, and reductions in accident fatalities and injuries can be monetized, not all researchers may agree on these valuations of impacts. (This topic will be revis- ited in chapter four, following presentation of the case examples.) • The way in which projects are scoped may affect results—examples were provided relating to grouping of independent projects, where not all might survive a BCA individually, and how to account correctly for interactions between complementary projects, where benefits of each project individually might underesti- mate the total benefit of the projects collectively if they were coordinated. • A BCA that considers impacts in several areas creates the need to forecast data in these areas, a task subject to uncertainty. The second GAO study considered a range of possible problems with BCA of highway and transit investments (Highway and Transit . . . Jan. 2005). Following are exam- ples of the issues that were identified: • The inability of models to accurately predict key effects of transportation investments; for example, changes in land use, driver behavior, or diversion to alternate routes or modes. This problem is aggravated by the diverse set of models employed by local agencies. • Lack of complete, accurate data may distort forecasts and lead to erroneous results. Surveys of travel demand are becoming harder to fund and to conduct. • Certain benefits are double-counted; certain expendi- tures are counted as benefits. Sometimes future benefits are cited at their nominal value, not discounted to pres- ent value. The avoided cost of another project may be counted as a benefit of the project that is being analyzed. • The definition of alternatives may miss viable options in the current mode or in other modes (e.g., failure to compare a highway project with a transit option). • Benefits that are difficult to quantify may be overlooked or eliminated from the analysis. This report also included a survey of state DOTs con- ducted in October 2004, which asked their opinions on fac- tors of great or very great importance in the decision to recommend a highway project. The higher-ranked factors in descending order were political support/public opinion, availability of state funds or federal matching funds, cost- effectiveness, and distribution of impacts across social groups. Economic impacts, ratio of benefits to costs, and availability of local funds were among the lower-ranked factors. The result of a BCA fared better in a similar survey of transit agencies. A third GAO study considered the BCAs used to evalu- ate policies encouraging shifts of passenger and freight traf- fic to rail. Two federal grant programs—the TIGER grants, discussed in more detail in the next section, and the High- Speed Intercity Passenger Rail program—included benefit– cost assessments that, in GAO’s findings, “were generally not comprehensive.” The quantification and monetization of benefits “varied widely” among the applications submitted. Selected applications that were examined in more detail by GAO for the most part failed to provide information that is recommended in federal guidelines for economic analyses of public works investments (refer to sections at the begin- ning of this chapter summarizing guidance documents); for example, information concerning risk and uncertainty, data limitations, and assumptions inherent in the methodology. Grant applicants, industry experts, and U.S.DOT officials who were interviewed by GAO pointed out additional chal- lenges: short time periods in which to prepare the economic assessments; lack of access to, or poor quality of, needed data; and lack of standardized values for monetizing certain benefits. Although benefit–cost was one of the stipulated cri- teria for evaluating the TIGER and the high-speed rail grant applications, these problems limited the usefulness of the economic information provided by the applicants. Federal highway Administration Findings In 2009, FHWA had the opportunity to review EEAs sub- mitted by transportation agencies competing for TIGER grants under the American Recovery and Reinvestment Act of 2009 (ARRA, P.L. 111-5, Feb. 17, 2009). The analyses were mandated by law; the TIGER grant application pro- cess demonstrated that state DOTs will perform economic analyses if they are required to do so. Review teams at the federal level included economists from the Office of the Secretary, U.S.DOT and from the modal administrations. There were positive aspects of the review; for example, the demonstration that BCAs are not that difficult, that no modal bias was evident, and that a good economic analy- sis imposes rigor and discipline, thereby reinforcing the project assessment based on policy criteria. However, a

22 number of common errors were evident in the submittals (Timothy 2010): • Incorrect treatment of some economic development and local construction impacts as project benefits. • Improper accounting of project costs. • Unrealistic base cases and project lifetimes. • Incorrect treatment of discounting and inflation. • Incorrectly treating initial-year or design-year numbers as a stream of constant annual amounts. • Incorrectly estimating safety benefits. The resulting conclusions by the TIGER grant reviewers were the following: • A wide disparity in the depth and quality of BCAs. • Insufficient information on expected project outcomes, accompanied by inadequately supported assertions of project benefits. The FHWA noted the following observations and lessons learned: • In future applications of this type, clearer guidance on input values and expectations for the quality and docu- mentation of project benefits and costs will be needed. • The choice of a discount rate did matter in the exercise. • Models can both illuminate and obscure outcomes. • There is a need for research on nonuser benefits and nontraditional impacts of transportation investments. • The distribution of benefits may need to be better under- stood; for example, public versus private; local versus national; and by income levels/population groups. inTErnATionAl ExPEriEncE World road Association The World Road Association (WRA) conducted a comprehen- sive review of economic evaluation methods for road projects in member countries as of 2002 (Economic Evaluation Meth- ods . . . 2004). Countries included Australia, Canada, Czech Republic, Denmark, France, Germany, Hungary, Japan, Mex- ico, the Netherlands, New Zealand, Norway, Portugal, South Africa, Sweden, Switzerland, the United Kingdom, and the United States. Methods of economic analysis and evaluation that were included in the review were as follows: • CBA (equivalent to BCA). • Cost-effectiveness analysis (e.g., used in cases with nonmonetized benefits). • Multi-criteria analysis (equivalent to a multi-objective analysis). • Risk-benefit analysis (similar to CBA, but with explicit allowance for risk). • Environmental impact assessment (EIA) (various forms of this analysis, sometimes incorporating results of a CBA). • Some countries also use LCCA to analyze and select certain components of the project; for example, pave- ment type. Results of the WRA’s survey showed that all 18 countries use BCA, or benefit–cost analysis (CBA), for at least some projects or some components of their analyses. CBA is often combined with multi-objective analyses (multi-criteria anal- ysis), but in different ways and for different types of projects among member countries. BCA may also be combined with other methods, including EIA, computation of socioeconomic benefits, and input–output analyses. In terms of the range of impacts considered: • Most countries include monetized values of travel time, vehicle operation, and accidents. • Environmental impacts are monetized by several nations, particularly for noise and air quality effects. Measures of economic results reported in the WRA study were generally the same as those used in the United States; that is, net present value, benefit–cost ratio, and IRR (although problems can arise with this approach, consistent with discussions of this method in many texts). The one mea- sure that differed from U.S. practice was the First Year Rate of Return, which is used in three countries. First Year Rate of Return is defined as the project benefit in the first full year following completion of construction divided by the proj- ect costs over the evaluation period. Thresholds of project acceptability using this metric varied from “greater than 12 percent” to “greater than 35 percent.” European commission guide The European Commission (EC) has published a guide to CBA that is intended for European national and regional authorities investing in the transportation, other civil and industrial infra- structure, and environmental sectors (Guide to Cost-Benefit Analysis . . . 2008). The EC Guide describes CBA as “applied social science,” which “is not an exact discipline”: It is based largely on approximations, working hypotheses and shortcuts because of lack of data or because of constraints on the resources of evaluators. It needs intuition and not just data crunching and should be based on the right incentives for the evaluators to do their job in the most independent and honest environment. Source: Guide to Cost-Benefit Analysis . . . p. 15. It is important that this wording not be misconstrued as implying a simplistic or informal process lacking in diligent research, data gathering, and quantification. The examples developed in the EC Guide demonstrate knowledge of theo- retical economics principles, engineering know-how, and a rigorous computational approach. Although the mechanics of CBA in the European context are similar to methods in the United States, the scope of the

23 EC Guide and the recommended use of economic analysis within the project life cycle differ from the U.S. examples provided in this synthesis. The EC Guide is more compre- hensive in scope than comparable U.S. publications, encom- passing the following: • Multimodal transportation infrastructure, covering investments in surface transportation networks (road and rail are illustrated in the EC Guide’s examples); high-speed rail in Europe; and ports, airports, and intermodal facilities. • Environmental infrastructure investments in water sup- ply and sanitation, waste treatment, and natural risk prevention (mainly floods and fires). • Productive industrial investments; energy infrastruc- ture investments including energy transport and trans- mission, distribution, and renewable energy facilities; and telecommunications infrastructure. • Investments in other sectors including educational and training infrastructure; hospitals and other health- related infrastructure; forests and parks; and industrial zones and technological parks. The economic analysis called for in the EC Guide diverges from the applications described in this synthesis in the fol- lowing respects: • The CBAs in the EC Guide are envisioned to occur early in the project life cycle; for example, in the pre-feasibility stage. In contrast, the examples in this synthesis occur throughout the project life cycle, as discussed in chapter one. • Investments addressed by the EC Guide are analyzed generally through a 20-year period. Several chapter three examples use analysis periods of longer than 20 years. The process of project appraisal, of which EEA is a part, includes the following steps in the EC Guide: • Identification of project objectives and the socio- economic and European national context in which the project will occur. • Project definition, which may entail civil works, actions, or services, similar to the use of “project” adopted in chapter one of this synthesis. Definition might also specify the perspective to be represented in the appraisal; for example, local, regional, or national (refer to the Critical Interstate Facilities case in chap- ter three for an illustration of this concept). Indirect benefits in secondary markets would not be included in the economic analysis unless covered by appropri- ate shadow prices or wages (this point corresponds to the exclusion of economic impacts from this study in chapter one). • Feasibility and option analysis, which entail identifica- tion of options or alternatives and an analysis of their feasibility to test which options can potentially meet the objectives of the project. Examples of different options to be considered for transportation projects include alternative routes, variations in the timing of construc- tion, and different transportation technologies. • Financial analysis includes items such as investment costs, operating costs and revenues, sources of financ- ing, financial sustainability, and financial returns. • Economic analysis includes rationalization of costs (i.e., corrections of price or cost distortions to their true val- ues), monetization of nonmarket costs and consequences (including externalities such as environmental pollu- tion), careful consideration of indirect or secondary- market effects (e.g., economic development impacts) to avoid double-counting of benefits, selection of a social discount rate, and calculation of economic performance indicators (e.g., economic net present value, economic IRR, and benefit–cost ratio). • Risk assessment techniques, including sensitivity analysis, probabilistic distributions for critical vari- ables, descriptions of analysis techniques, assessment of acceptable levels of risk, and risk prevention. • Other project evaluation approaches and topics are also covered, including cost-effectiveness analysis, multi- criteria analysis, and EIA (more of a macro-economic study, whereas CBA is essentially a micro-economic analysis). EIAs are rare, typically done for mega-projects that are very large in relation to the economy. The EC Guide includes one case study related to road transport: the construction of a new intercity highway to accommodate rapid traffic growth and relieve congestion on an existing link. Two options for the new highway are considered: a toll road and a non-tolled road. The analy- sis is structured to reflect two sources of traffic for the new highway: diverted traffic from the existing road and gen- erated traffic that did not exist before. The percentage of traffic induced to switch from the existing highway to the new road is price-elastic, with the price comprising the per- ceived VOC, including the toll on the toll–highway option. Thus, the benefits provided by the new highway will vary between the tolled and the non-tolled options, because fewer vehicles divert to the toll highway because of its higher per- ceived cost. Because shifts in demand for the new facility will occur as well as changes in the perceived price of trans- port, the benefit measures used in the EC Guide case study include a calculation of consumer’s surplus. (Construction of a new highway is not included in any of the case examples in chapter three, so no comparison of methods is available in this synthesis for this category of project.) The EC Guide includes 10 annexes providing addi- tional information on topics such as selection of a discount rate, project performance indicators, evaluation of projects to be performed by public–private partnerships, and risk assessment.

24 Transport canada guide Overview of Guide Transport Canada has published a guide on BCA (Guide to Benefit–Cost Analysis . . . 1994). Compared with the EC Guide, it is focused specifically on transportation and on BCA as the sole analytic method of interest. The specific economic criterion used is the NPV. The Guide aims to be a practical source of information to transportation analysts and a frame- work within which transportation projects of diverse charac- teristics can be addressed under a common set of principles, methods, and information. Transportation is treated as mul- timodal, encompassing surface, air, and marine examples, and with shifts between modes part of the consideration of alternatives. The value of an economic analysis is that “all of the important effects of investment choices can be made visible and, to the extent possible, quantified; it is a key tool in the quest for value for money” (Guide to Benefit–Cost Analysis . . ., p. 2). The Guide treats the elements of a proper BCA in detail, including illustrations with realistic examples but avoiding a “cookbook” approach. Although its focus is on the practical, it occasionally explains the theoretical ratio- nale for a particular step. It is organized around the benefit– cost methodology: • Part I—The Evaluation Framework: addresses the identification of options for analysis, including the base case, and how to establish a common framework for assessing competing options. • Part II—Measurement of Costs, Benefits, and Other Effects: includes the following: – Costing principles, the identification of project costs for different phases (planning, construction/development, operation, post-forecast), and LCCA. – Benefits and other effects, including measurement principles (basis used is willingness to pay), and transportation benefits in several categories that are discussed in more detail here. – Discounting, including the recommended treatment of nominal and constant dollar projections, the ratio- nale for discounting (including the “dollar more valuable now than later” and the opportunity-cost concepts discussed earlier), the recommended value of discount rate (a 10% real rate is stipulated, with sensitivity analyses to be conducted for 7.5% and 12.5%—more on these values later), and discounting conventions. Part III—Analysis and Presentation: discusses the evaluation of options, dealing with uncertainty, cost-based comparisons to deal with nonquantifiable impacts, presentation of results, and guidelines for structuring a BCA report for Transport Canada. Part IV—Final Summary: provides a bulleted list of items summarizing the entire process of benefit–cost assessment of a project, together with suggested con- siderations at each step. Categories of Transportation Benefits The categories of transportation benefits considered in the Transport Canada Guide are broader than those typically identified in U.S. analyses. Overlap between practices in the two countries regarding highway projects occurs in safety ben- efits and selected transportation efficiency (or mobility) ben- efits; for example, passenger and cargo travel-time savings, congestion mitigation, and VOC savings. The Guide includes a number of suggestions for estimating these quantities; for example, values of passenger time for different system users and for adults versus children, values of travel time for cargo, and methods to estimate avoided costs of collisions for safety benefits. Benefits to non-Canadians are treated as a matter of expected reciprocity and are computed in the same way as benefits to any other passengers. What the Guide labels as transitional effects are essentially disruptive impacts to motorists and others resulting from project construction. The Guide also cautions that economic impacts representing “multiplier effects” of a project would be excluded from a benefit–cost assessment, which is consistent with the guide- line for this synthesis set forth in chapter one. The Guide includes other categories of benefits that are not always found in U.S. highway-related analyses, again recognizing that the Transport Canada Guide applies to several transportation modes. These categories include: • Small travel-time savings: This category is that sub- set of passenger travel-time savings that represents a few minutes per trip. The issue arises because it is not clear whether these small increments of time should be valued proportionally to larger travel-time savings (e.g., so that one minute saved equals 1/60th of hourly savings) or whether small time intervals that fall below some threshold might be valued less than proportion- ally because they are too small to be used productively elsewhere. The recommended approach is to compute the cumulative small time savings (e.g., the sum of all savings of 5 minutes or less) and value them propor- tionally to larger savings, but to isolate them from the NPV analysis. In this way they are reserved for separate consideration by management. • Generated traffic. Benefits to generated traffic are taken as less than those to existing traffic. A factor of one-half is recommended as the multiplier. • Diverted traffic. Treatment of benefits to traffic that diverts to a newly constructed facility or service depends on context; that is, what other transport services are available. The Guide’s recommendation is for analysts to contact Transport Canada’s Economic Evaluation Branch for advice on this point if applicable. • Productivity gains. These are considered for those projects that improve the productivity of government operations. Productivity gains due to a project or other investment (as in training) may result when (1) the same service can now be provided at less cost, (2) for a given

25 cost the level of service is now higher, or (3) a com- bination of these two situations. The range of projects that can yield productivity gains is large, encompassing examples such as new air traffic control systems, new management systems and other information technol- ogy applications, installation of training simulators, the aforementioned improvement in training programs, and improved communications. • Environmental effects. These are discussed in terms of air, water, and noise pollution; degradation of the natural environment and of habitats; loss of amenities such as access to parkland; and disposal of contaminated soil. As the Guide acknowledges, however, identifying envi- ronmental impacts is one thing, but measuring or quan- tifying them is another; and information regarding the values that Canadians place on environmental protection is uncertain and subjective. The Guide offers brief, gen- eral advice on how to address each of the environmental impacts listed earlier. Discount Rate The Guide includes a brief commentary on the 10% real dis- count rate called for by the Canadian Treasury Board. The Guide acknowledges that some have proposed a reduction in the 10% rate when there are future benefits and other impacts that call for significant judgment in assigning a monetary value; for example, the avoided cost of fatal and injury-causing col- lisions, and the avoided cost of environmental damage. The issue is posed as one of intergenerational equity, with some feeling that a 10% discount rate gives too little weight to the interests of future generations. Pending further consid- eration of the issue, the Guide continues to recommend the 10% discount rate for all projects and all categories of costs, benefits, and other impacts. other international sources Other international sources will be cited in the chapter three case examples (e.g., papers on highway safety) and in chapter four regarding implementation issues, particularly the link between EEAs and performance management and accountability. scrEEning survEy motivation and Purpose: screening device Case examples illustrating management applications of EEA are at the core of this synthesis. A screening survey was cre- ated as part of the process to identify potential candidates. Although NCHRP synthesis surveys are typically designed to develop a statistical description of current U.S. highway practice, that objective was not part of this screening survey. The primary objective of the screening survey was to increase the chances of identifying good prospects for the case exam- ples, complementing the other efforts described in chapter one (interviews with topic panel members and other experts, committee presentations, and information gained from the literature reviews). Overall, all of these efforts converged on a select number of agencies with comprehensive, well- developed capabilities to perform and apply EEAs. Selec- tion of the agencies whose practices are described in chapter three was made solely on their merit in meeting the study objectives and on their willingness to provide the informa- tion needed to develop the case descriptions. “Meeting study objectives” referred to a desire to present cases represent- ing a variety of program areas (e.g., preservation, mobility, and safety), stages of the decision process (e.g., planning and programming), and levels of the system analyzed (e.g., link or project, corridor, program, and network), consistent with the scope of work discussed in chapter one. distribution To ensure the widest possible response within the brief time allocated to the survey, copies of the questionnaire were sent to multiple recipients within each state. The screening survey was distributed as follows. For state DOTs, questionnaires were sent to the respective members of the AASHTO SCOH, with copies to each state’s RAC member. Copies were also sent to the FHWA division office in each state. Again, the survey was intended primarily as a screening device to identify agencies with information and experience that warranted additional investigation, and which might serve as additional case examples. The survey did not attempt to develop statistical information on the use of EEA nationwide, because past research had already indicated that this use was limited. The questionnaire was purposely kept brief in an attempt to gain quick turnaround and encourage increased response. The survey questions reflected the require- ments of the synthesis scope of work, as listed here. A copy of the screening questionnaire is reproduced in Appendix A and asked for the following information: • Whether the agency uses engineering economic meth- ods at all in its decisions on highway investment. • For those agencies that do use such methods, to check- off where such methods are used, identifying the following: – The highway investment programs for which the methods are used; for example, pavements, bridges, capacity addition, and safety; – The stage of the program life cycle for which the methods are used; for example, planning, program- ming, project design and development; and – The level within the transportation system or pro- gram for which the methods are used; for example, individual project, corridor, program, network, and cross-program tradeoffs.

26 • Whether this information is used by the agency’s chief executive office in considering decisions on highway investment. • For those agencies that do not employ economic analy- ses in their decision making, the reasons for their reser- vations about economic methods and information. Twenty-three states provided responses; 17 completed questionnaires came from DOTs; eight, from division offices. In two states, both the DOT and the FHWA division office reported results; the two responses in each state were con- solidated and tallied as a single state response. Twenty of the 23 responding states reported that they do use EEAs, whereas 3 states reported not using economic analyses. Written com- ments by respondents indicated that “using economic analy- ses” was interpreted as “routinely applying these techniques.” Representatives of some state DOTs or FHWA division offices reported that although their state DOT did use economic meth- ods in special situations, they nonetheless described the agency as “not using these methods.” The tally of survey results pre- sented here is based directly on the submitted questionnaire responses, with no attempt at further interpretation. Only the 20 states that reported affirmatively their use of economic methods are included in the tally of results. results Key Findings The use of EEAs by the 20 positively responding states is presented in Tables 3 and 4. Table 3 shows the distribu- tion by stage of decision making for each investment pro- gram; Table 4, the distribution by level of highway system at which analyses are performed for each investment program. The last column in each table indicates the number of states that made any use of economic analysis whatsoever for the indicated investment program in that row. These rightmost column entries are identical between the two tables, which would be the expected result. Although entries in each col- umn are summed at the bottom of each table, row entries are not, because these sums would not be meaningful; that is, an agency may use economic analyses at more than one decision stage, and at more than one level of the highway system. Key findings are as follows: • EEAs are most widely used in DOTs’ pavements and structures programs, corresponding to the wide use of pavement and bridge management systems as well as pavement-type selection procedures. Significant use is also reported in safety programs, a result consistent with nationwide practice reported by Hanley (2004), and in urban mobility programs [which include major capacity projects and Intelligent Transportation System (ITS) projects], a result consistent with earlier findings by Neumann (1997) and with the growing interest in ITS analyses (refer to chapter four). • Tables 3 and 4 indicate that the greatest use of EEA occurs at the project level, supporting project design and development. The next highest tier of uses in the program development cycle is for planning and programming, which are done at a program level. These statements are generally true for all individual programs surveyed, as well as cumulatively. Respondents’ Comments on Specific Applications The questionnaire solicited open-ended comments about particular areas where economic analysis is used by report- ing agencies, based on items suggested by the scope of work. Agencies highlighted more specific purposes of their economic analyses as follows: • A BCA for ITS planning and programming. • A LCCA for material selection as part of the state’s pavement type selection. This is done as policy for proj- ects meeting certain scope or budget requirements, as well as other locations when is appropriate. • A benefit–cost approach in the state’s pavement system preservation program for the budget allocation plan, as well as the project location and treatment selection. • LCCA on larger projects when considering alternatives and more routinely in pavement treatment selection. • One state DOT noted that its Value Engineering (VE) Bureau at times gets involved in bid evaluation, when requested in special cases, to determine if bids are rea- sonable. More generally, the VE Bureau and individual units perform the following tasks: – VE is involved in scope development, scoping team meetings, independent VE review of developed alter- natives, and LCCA. – Road User Solutions is involved with road user costs, lane occupancy charges, and incentive/disincentive values that are included in the Special Provisions. – Value Solutions is the review unit for the depart- ment’s Preliminary Design and Final Design submis- sions with emphasis on roadway, bridge, safety, and traffic operations, and with input from the depart- ment’s state maintenance engineer as required. Two agencies focused on specific applications involving the road user-cost component of economic analyses: • On some projects, road user costs are used for A+B bid- ding. (A+B bidding is a cost-plus-time bidding process comprising component A, the cost that is bid to perform work, and component B, the time that is proposed to complete work, with the objective of encouraging proj- ect completion as soon as possible.) On some projects road user costs have also been used to determine the level of early completion incentives on some projects and are involved in comparing construction plans. (On this topic, refer also to the Accelerated Project Delivery case example in chapter three).

27 TABLE 3 USE OF ENGINEERING ECONOMIC ANALYSIS AT DIFFERENT STAGES OF DECISION MAKING Investment Program Planning Programming Resource Allocation Project Design and Development Bid Evaluation (e.g., for Best-Value Procurement) Other Stage of Decision Making No. of States Reporting EEA Use Pavements 9 9 7 18 6 1 20 Bridges, Other Structures 10 11 8 14 4 1 18 Other Asset Preservation 6 4 3 6 2 1 9 Urban Congestion Relief: Capacity Expansion 9 5 3 9 1 1 13 Urban Congestion Relief: ITS Strategies 8 4 3 9 1 13 Urban Congestion Relief: Other Operations Improvements 6 3 2 7 2 10 Rural Mobility 6 2 2 5 1 8 Safety 10 9 6 12 15 Economic Impact 5 3 2 1 1 8 Environmental Impact Mitigation 6 1 3 8 1 9 Other 1 1 2 1 — Totals 76 51 40 91 15 9 — Source: Synthesis screening survey. Notes: No. of States = Number of DOTs or FHWA division offices reporting, excluding duplicates within a state; EEA = engineering economic analysis; — = Total not meaningful. Investment Program Project Level Corridor Level Program Level Network Level Cross-Program Tradeoffs Other System or Program Level No. of States Reporting EEA Use Pavements 20 6 14 11 4 0 20 Bridges, Other Structures 18 5 10 7 5 0 18 Other Asset Preservation 8 3 3 4 2 1 9 Urban Congestion Relief: Capacity Expansion 11 6 6 4 3 0 13 Urban Congestion Relief: ITS Strategies 8 4 8 2 2 1 13 Urban Congestion Relief: Other Operations Improvements 8 2 4 2 1 2 10 Rural Mobility 6 4 3 3 1 2 8 Safety 15 9 11 6 3 0 15 Economic Impact 4 4 2 1 0 2 8 Environmental Impact Mitigation 9 4 2 2 0 0 9 Other 2 1 1 0 0 0 — Totals 109 48 64 42 21 8 — Source: Synthesis screening survey. Notes: No. of States = Number of DOTs or FHWA division offices reporting, excluding duplicates within a state; EEA = engineering economic analysis; — = Total not meaningful. For purposes of this table: Program Level includes all projects collectively within a single investment program. Network Level projects address particular investments (typically in urban congestion relief or rural mobility) that result in spatial or temporal shifts in demand from one link or time period to another. TABLE 4 USE OF ENGINEERING ECONOMIC ANALYSIS AT DIFFERENT TRANSPORTATION SYSTEM LEVELS

28 • The planning unit within another agency uses BCA when alternate routes are being evaluated for selection based on differences in their respective road user costs. Integration Within Business Process New Jersey DOT On an annual basis, the New Jersey DOT prepares asset management plans, which are instrumental in developing the department’s Statewide Capital Investment Strategy (SCIS). The SCIS is a performance-based decision- making process that evaluates alternative investment scenarios and produces desired and constrained investment targets for each asset category. SCIS is used to guide the resource alloca- tion strategy for the development of the 10-Year Capital Plan. A strategic resource allocation process has been con- ducted that applies performance measures to guide the deter- mination of program category investment targets required to achieve agency goals and objectives over the next ten years. It involves classifying all of the capital work done into pro- gram categories and establishing goals, objectives, and per- formance measures for each category. Quantitative performance analyses have been conducted, when possible, for highway and mass transit assets. Qualita- tive performance analyses were used when sound data were not available or could not be technically applied to gauge the per- formance of a particular category. For example, data for state highway infrastructure are inventoried and life-cycle cost per- formance curves developed and analyzed using various man- agement systems data for bridge and roadway assets including pavement and drainage condition information. Performance data are also applied from the congestion and safety manage- ment systems to conduct prioritization evaluations for alterna- tive budget scenario evaluations. The process to select the Recommended Constrained Investment Targets makes every effort to optimize the overall performance of the budget. This approach tries to make certain that scarce financial resources are used as economically as possible to address their most impor- tant needs. Several investment target options are designed to achieve various performance levels for each program. Washington State DOT Washington State DOT (WSDOT) uses economic analyses in the system planning, scoping, prioritization, and investment tradeoff phases. (Refer to several WSDOT case examples in chapter three.) Executive Use of Economic Information Agencies were asked to respond “Yes” or “No” as to whether economic analysis results were reviewed by executive man- agement. Fifteen of the 20 agencies using economic analyses reported executive-level use of this information. Among those that responded “Yes,” several provided additional comments: • Economic data and results for pavement selection are evaluated by a team of senior managers within the SHA when the life-cycle costs analysis results between different alternatives are within 20% over a 40-year period. Other noneconomic factors are included in this discussion when the economic factors and consensus are reached for pavement type selection. With regard to the pavement system preservation program, the chief engi- neer and administrator review the program’s performance targets and budget allocation plan after they are devel- oped and before approval and delivery (Maryland SHA). • Summary results of network/program-level economic analysis for pavements are presented to upper manage- ment (Maine DOT). • When cost analyses are used in a project analysis, the bureau administrators present the results to the Com- missioner’s Office to support their proposed alterna- tive. Program analysis in pavements is nearly ready to implement (New Hampshire DOT). • Information related to bridges is used for the Pennsylvania DOT’s Accelerated Bridge Program (Pennsylvania DOT). • Pavement life-cycle analysis results are used by man- agement to facilitate decisions on pavement type (South Carolina DOT). • Executive-level managers review pavement, bridge, and safety (Utah DOT). • Reporting results also includes presenting some of this information to the public through performance reporting in WSDOT’s Gray Notebook. Recently, WSDOT began presenting a portion of this information to the governor as part of [statewide] performance reporting (WSDOT). Shared Vision Among Agencies California’s response noted that economic analysis plays a role in helping create a shared vision of transportation among agen- cies at different levels of government. It is intended that trans- portation agencies at various governmental levels throughout the state will use engineering and economic analysis from plan- ning through operations, preservation, and maintenance: Through the use of management systems, engineering and economic analysis, and other tools, MPOs/RTPAs [Regional Transportation Planning Authorities] and transit operators can more comprehensively view the big picture and evaluate col- lected data before making decisions as to how specific resources should be deployed. Asset management principles and tech- niques should be applied throughout the planning process, from initial goal setting and long-range planning to development of the TIP [Transportation Improvement Program] and then through operations, preservation, and maintenance. . . . Source: California’s 2010 Regional Transportation Planning Guidelines, as cited in California’s screening survey response. On a related note, the literature review and case exam- ple development conducted for this synthesis indicated that other state DOTs play an important role in assisting their local agencies with methods and tools to conduct EEAs. Other Responses Because this was a screening survey to help identify candidates for further study in chapter three, the results are not intended as

29 statistically representative portrayals of nationwide positions on EEA. In particular, only three states claimed that they do not use EEA in highway investment decision making, whether reported by the state DOT itself or by the FHWA division office. All three provided rationales for this position, which are summarized here. Because of the small sample and that not all responses came directly from the state DOT, these comments are summarized for completeness in reporting the survey and as examples of selected state DOT opinion. They are not neces- sarily indicative of nationwide practice or thought. The states that claimed that they do not use EEA on a routine, consistent basis cited one or more of the following reasons: • Matters of relative priority, lack of a champion, and lack of an effective plan for implementation hinder advances in this area. • Use of economic analyses is limited to occasional, selective applications, with limiting factors including: – political forces driving decisions; – questionable assumptions in modeling, leading to variability of results depending on particular assump- tions and inputs; – lack of available data; – lack of confidence in certain data (with engineering judgment preferred); – infrequent uses of engineering economic methods that focus on special types of projects (e.g., design–build); – while planning to consider the use of economic methods in the future, acknowledging the difficulty of selling the plan now in times of tight budgets (because the need now is to consider initial project costs and to distribute project funding throughout the state, dimin- ishing the importance of the long-term view); – the complexity and time involved in developing an EEA; – the perceived tendency of a BCA to favor a single type of safety treatment rather than a more diverse program of multiple types of treatments; and – litigation concerns with safety programs evaluated through BCA. PrEviEW oF cAsE ExAmPlEs The information in this chapter has provided the founda- tion for the case examples in chapter three, introducing the methods of EEA typically used to inform decisions on trans- portation investments, drawing on U.S. and international experience. It has outlined the broad guidelines issued by the federal government and professional associations at the national level for performing highway-related economic analyses in the United States. It has discussed issues in applying economic analyses, including methods to mitigate the risk and uncertainty that are inherent in these types of assessments; identified shortcomings in the analytic meth- ods and data required; and noted other problems that impede wider use of economic methods. It has established, through a screening survey of state DOTs and FHWA division offices, general characteristics of existing practice that have contrib- uted to the selection of the set of case examples and that pro- vide a context for the cases. The case examples build on this information base. Each case provides more detailed information on governing policies, amplifying the federal requirements discussed in this chap- ter and adding state-level and agency-specific guidance. After developing the context and rationale of a case, the chapter three descriptions explain the particular economic analyses used, the measures of technical and economic performance appropri- ate to the case, and how these results are incorporated within highway investment decision making, including non-economic and non-quantifiable considerations where appropriate. The cases have been selected to illustrate a broad range of examples in terms of highway programs (e.g., mobility, safety, preserva- tion, and program-level tradeoffs); the stage of decision mak- ing (e.g., planning, programming and budgeting, and resource allocation); and the level of the highway system to which the analysis applies (e.g., project, corridor, program, and network). The screening survey results showed that most state DOTs use economic analyses for at least certain types of projects; for example, pavement and bridge preservation, safety improve- ment, major projects, and urban mobility. What differenti- ates the exemplary case examples in chapter three from this broader participation in EEA? One item, mentioned in the previous paragraph, is the wider scope of application across programs, decision-making stages, and levels of the highway system analyzed. The characteristics go deeper, however, into several aspects of an agency’s makeup and its approach to solving transportation problems and needs; for example, the influence of organizational champions and culture; a level of knowledge and comfort regarding use of economic meth- ods; a tighter integration of business and decision processes that make economic analyses a part of routine business rather than a distinct, somewhat isolated task; the willing- ness to be creative in defining alternative solutions, and to innovate in devising the methods and measures of analysis; that upper management asks for and uses the results of eco- nomic analyses in its decision making; the availability of information technology to support not only the economic analyses but also important steps such as diagnosing prob- lems, defining realistic alternatives, and displaying results; the training of staff and provision of resources such as for- mal guidelines and websites on economic concepts and methods; and, perhaps most importantly, the recognition that economic outcomes are an integral part of gauging highway system performance. For purposes of this synthesis, this latter subset of agen- cies will be referred to as those “conversant” with engineer- ing economic methods. This label will distinguish them from other agencies that may apply economic methods, but in a less intensive way to a more limited scope of decisions. The case examples in chapter three will demonstrate characteris- tics of selected conversant agencies; chapter four will revisit

30 Decision-Making Stage Exam ple EEA Applications Selected Illustrations in Case Examples in Chapter Three Set Policy Goals, Objectives, and Perform ance Measures Translate aspirational goals into practical, achievable objectives Define performance m easures that can be incorporated within EEA analyses Set perform ance targets using engineering and econom ic criteria Although EEA is not described at this stage in case examples, the cases illustrate explicitly or im plicitly how goals, objectives, and perform ance measures drive decision making. This guidance is established in a practical, meaningful way for use in subsequent decision and evaluation stages. Planning Assess realistic, economically viable alternatives for future major projects Assess alternatives in corridor-level or network-level transportation im provem ents Assess options for system atic im prove me nts at a progra m level Refer to the Critical Facilities case exam ple Refer to the Critical Facilities case example Refer to the Mobility Planning case example Programming and Budgeting Apply EEA potentially at several steps: 1. Assess project options, select best 2. Rank or prioritize projects in programs 3. Develop realistic candidate programs 4. Conduct program tradeoffs Recommend program budgets based on results of previous analyses Refer to the following case examples: Mobility Programming, Safety Programming, Bridge Programming and Permitting, and Economics-Based Tradeoff Analysis Resource Allocation Following Legislative Budget Approval Update proposed program s and projects based upon legislative budget approval, for use in resource allocation Realign economics-based program or project objectives and performance targets if needed Refer to the Economics-Based Tradeoff Analysis case exam ple — Decisions on Program Delivery: Project Design and Development Assess design options from an econom ic standpoint Assess capital–maintenance tradeoffs, including user costs as a gauge of road perform ance Refer to the Pavement Type Selection and Value Engineering case exam ples — Decisions on Program Delivery: Options in Project Construction Assess the mo st econom ical strategy for project construction (including a design-build option) Assess potential applications of new construction approaches (e.g., perform ance warranties), and use of innovative materials and technology Refer to Accelerated Project Delivery case exam ple — Bid Evaluation; e.g., for Best-Value Procurem ent Incorporation of econom ic analysis as part of a best-value bid evaluation algorithm — System Monitoring, Perform ance Assessment, and Feedback to Policy Stage Validation of project and program decisions in term s of outcomes: engineering perform ance and econom ic results Adjust me nt of assum ptions and para me ters used in econom ic analyses Refer to follow-up studies described in the Safety Programming case exam ple — Notes: — = No case in chapter three corresponds to the example application in the second column. EEA = engineering economic analysis. Decision stages in the first column are illustrative. Actual practices within individual agencies may vary in their order and content, but the intended business processes should be clear. TABLE 5 POTENTIAL APPLICATIONS OF ENGINEERING ECONOMIC ANALYSIS THROUGH THE PROGRAM DECISION CYCLE

31 the themes that differentiate these agencies from the larger population of transportation organizations, to identify more fully the factors underlying successful applications of EEAs. A brief overview of the cases is provided in Table 5, indi- cating for each stage of decision making the types of uses to which EEAs may be applied and the corresponding case example in chapter three. Additional information on the methods used in each case and the level of the highway sys- tem at which they are applied is given in Table 6. All told, there are 12 case examples of engineering economics appli- cations, including a “case within a case” in the Critical Inter- state Facilities example. Furthermore, chapter four draws on lessons gained from the set of case examples plus additional state DOT interviews and literature searches to present sev- eral aspects of successful implementation of engineering economic methods—in effect, an extension of the 12 cases to a “thirteenth case” drawn from current industry practice. Decision-Making Stage Relevant Chapter Three Case Exam ples System Level at Which Analyzed Methods Illustrated Set Policy Goals, Objectives, and Perform ance Measures Although engineering economic analysis is not described at this stage in case exam ples, the cases illustrate explicitly or implicitly how goals, objectives, and perform ance measures drive decision ma king. This guidance is established in a practical, meaningful way for use in subsequent decision and evaluation stages. Planning Critical Interstate Facilities Mobility Planning Corridors and Networks: Regional highways International maritime shipping Program Road user costs avoided Benefit–cost + risk analysis (scenarios) B/C analysis, NPV; adjust for uncertainty Programming and Budgeting Mobility Programming Safety Programming Bridge Programming and Permitting Econom ics-Based Tradeoffs Program Program Project Corridor, Network B/C analysis, NPV Benefit–cost Cost-effectiveness + utility Excess road user costs Resource Allocation Following Legislative Budget Approval Econom ics-Based Tradeoffs Corridor, Network Excess road user costs Decisions on Program Delivery: Project Design and Developm ent Pavement Type Selection Value Engineering Project (two DOTs described) Project (two DOTs described) NPV + risk analysis (probabilistic) B/C analysis, NPV Decisions on Program Delivery: Options in Project Construction Accelerated Project Delivery Project Comparison of discounted costs and discounted benefits + risk analysis* (probabilistic) Bid Evaluation; e.g., for Best-Value Procurement — System Monitoring, Perform ance Assessm ent, and Feedback to Policy Stage Safety Program: Follow-Up Analyses Progra m: Assess perform ance and econom ic results to validate future program guidance Cost-effectiveness Decision stages in the first column are illustrative. Actual practices within individual agencies may vary from these in order and content, but the intended business processes should be clear. For explanation of System Levels in column 3, refer to description of Scope of Work in chapter one. For explanation of Methods Illustrated in column 4, refer to discussion of Engineering Economic Methods in chapter two. B/C = benefit-cost; NPV = net present value. *The objective of this study was to explore the development of a methodology for future application to project delivery options, and not to inform an actual decision. Notes: — = There is no case example at this decision-making stage. TABLE 6 CHARACTERISTICS OF CASE ExAMPLES