Transportation Investments in Response to Economic Downturns (2014)

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.

3 Design, Management, and Outcomes of the American Recovery and Reinvestment Act Transportation Grants The experience of the transportation grants provided by the American Recovery and Reinvestment Act (ARRA) is the best available evidence on the functioning of a transportation stimulus program. This chapter describes the provisions and objectives of the grants, the challenges that federal and state government managers encountered in managing them, and ARRA’s impact on transportation spending and on construction employment. Alternative provisions for a transportation stimulus pro- gram that might increase benefits (i.e., stimulus impact and long-term transportation benefits) are described in Chapter 4. ARRA TRANSPORTATION PROVISIONS As Chapter 1 described, of the $831 billion in federal spending and tax relief provided by ARRA, $48.1 billion was for programs to be admin- istered by the U.S. Department of Transportation (U.S. DOT). Most of the U.S. DOT–administered funds were to be disbursed as grants for capital expenditures for state and local governments (Figure 3-1). ARRA also provided federally subsidized credit through Build America Bonds, which state and local governments could use to finance transportation construction; appropriations for the U.S. Army Corps of Engineers civil works program for water projects, including navigation projects; and appropriations for nontransportation infrastructure. The com- mittee’s examination was limited to the U.S. DOT–administered pro- grams. Administrative provisions of the transportation funds provided in ARRA are summarized below (GAO 2011b, 4–6; see also ARRA, Public Law 111-5, February 17, 2009, Title XII). 63 

64 Transportation Investments in Response to Economic Downturns 2.3% 3.1% 0.2% 0.4% Highway grants to states & territories ($26.8 billion) Highways: federal lands, Indian reservations, other ($0.7 billion) Intercity passenger rail grants to states ($8.0 17.5% billion) Amtrak ($1.3 billion) Transit agency grants ($8.4 billion) 2.7% FAA facilities ($0.2 billion) 55.6% Airport grants ($1.1 billion) 16.6% TIGER competitive grants ($1.5 billion) Small shipyard grants ($0.1 billion) 1.6% Total $48.1 billion FIGURE 3-1  ARRA appropriations for U.S. DOT–administered programs. (Note: FAA = Federal Aviation Administration; TIGER = Transportation Investment Generating Economic Recovery. Source: GAO 2011b, 4.) Apportionment of Funds Most ARRA transportation funds were distributed to state and local gov- ernment agencies according to procedures of established federal high- way, transit, and airport grant programs. All highway funds [other than highway grants in the Transportation Investment Generating Economic Recovery (TIGER) program described below] were distributed among the states according to formulas already in use in the regular federal- aid highway program that depend on highway lane miles, vehicle miles, population, and tax payments to the federal Highway Trust Fund attrib- utable to highway users in the state. Of the $8.4 billion of transit funds, $7.65 billion was apportioned among the 152 largest urbanized areas, or among the states for use of small urbanized areas, according to formulas in the existing federal transit program that depend mainly on popu- lation; population density; and transit vehicle miles, passenger miles, and route miles. The Federal Transit Administration (FTA) awarded the remaining transit aid, $750 million designated in ARRA for the existing discretionary transit capital grant program, to projects under way that 

Design, Management, and Outcomes of the ARRA Transportation Grants 65 had already received grants and that FTA judged were capable of spend- ing additional funds quickly (74 FR 89, 21843, May 11, 2009). As a corollary of following the rules of the existing federal-aid pro- grams, ARRA grants could be used only for capital expenditures [with the exception that recipients of certain transit grants were allowed to apply up to 10 percent of the funds to operations (FTA n.d. a)]. Thirty percent of ARRA surface transportation funds apportioned to each state were to be suballocated to local governments for use on proj- ects of the local governments’ choosing. This was larger than the share of funds in the regular federal surface transportation aid program subject to a similar substate allocation requirement [about 9 percent of authoriza- tions in the 2005–2009 Safe, Accountable, Flexible, Efficient Transporta- tion Equity Act: A Legacy for Users (SAFETEA-LU)].1 Two newly created discretionary grant programs, high-speed rail grants and TIGER grants (called Supplemental Discretionary Grants for a National Surface Trans- portation System in Title XII of ARRA), were to distribute $8.0 billion and $1.5 billion, respectively (20 percent of U.S.DOT–administered ARRA funds in total), on the basis of the U.S. DOT evaluation of applicants’ pro- posals. The U.S. DOT defined the criteria for judging TIGER and high- speed rail project proposals to include both long-term outcome (i.e., the transportation-related benefits of the facilities constructed) and immedi- ate economic stimulus (74 FR 115, 28758, June 17, 2009; 74 FR 119, 29902, June 23, 2009). Because these two programs were new (the passenger rail grant administrative framework had been established in the Passenger Rail Investment and Improvement Act of 2008, enacted 4 months before ARRA), the U.S. DOT was obliged to establish procedures for awarding and overseeing grants quickly, and eligible recipients lacked experience in applying for and using them. Distressed Areas Priority States were required to give priority to highway projects located in eco- nomically distressed areas, defined as areas with low per capita income 1 In the regular federal-aid program, the states were required to suballocate about 56 percent of Surface Transportation Program funds, which in turn were 16 percent of all SAFETEA-LU authorizations (FHWA 2007). 

66 Transportation Investments in Response to Economic Downturns or high unemployment relative to national averages or that had expe- rienced recent major job loss from events such as a business closure or disaster. To enforce this provision, the Federal Highway Administration (FHWA) required the states to document how they had searched for distressed area projects and to report which projects were located in dis- tressed areas (FHWA 2009a). Matching Funds Waiver No matching contribution to projects was required from recipients of most ARRA transportation grants; that is, projects could be 100 percent federally funded. This provision was the major exception to procedures of the established grant programs. For example, normally, the federal government reimburses the states for 90 percent of the cost of federal-aid projects on Interstate highways and for 80 percent of the cost of projects on other highways. Obligation and Spending Deadlines ARRA imposed deadlines on obligation of funds by transportation aid recipients. [Obligation occurs when the recipient notifies the U.S. DOT of a specific project to be funded with the federal aid and the U.S. DOT approves the project (GAO 2011b, 7).] Highway funds apportioned to states and transit funds apportioned to states or metropolitan areas that the recipients had not obligated within 1 year were to be withdrawn and redistributed by the U.S. DOT among recipients that had met the dead- line. Half of funds apportioned were required to be obligated within 120 days for highways and 180 days for transit projects. Priority was to be given to highway projects that could be completed within 3 years. These short deadlines were a departure from normal practice; in the federal-aid highway program, funds apportioned to a state normally are available for use by the state for 3 to 4 years. The deadlines for the new competitive grant programs were less rigor- ous, presumably because these programs were seen as serving the rein- vestment objective of ARRA as much as the recovery objective. Passenger rail grants were to be obligated by September 30, 2012. The ARRA appro- priation for TIGER grants expired September 30, 2011 (although a new round of TIGER grants was authorized in November 2011). 

Design, Management, and Outcomes of the ARRA Transportation Grants 67 Maintenance of Effort The governor of a state receiving transportation grants was required to certify that the state was maintaining its expenditure of state funds at previously intended levels for the types of projects funded by the state’s ARRA grants. The intent was to ensure that ARRA increased transporta- tion spending rather than replacing state funding (U.S. DOT 2011, 1). States were required to report in March 2009 their planned expenditures of state funds in each of the nine programs in which ARRA appropriated funds (see Table 3-1) according to plans in effect on February 17, 2009, for the period from February 17, 2009, to September 30, 2010, and then to report actual expenditures at the end of this period. The penalty for states that failed to maintain planned spending was exclusion from the 2011 regular annual redistribution of unused obligation authority in the federal-aid highway program. TABLE 3-1  Appropriations and Expenditures in ARRA Programs Administered by U.S. DOT Appropriation Percent Expended Mode and Use ($ billions) March 31, 2011 March 31, 2012 Highways: Capital grants to states through 27.5 68 88 federal-aid highway program; capital projects on Indian reservations and federal lands Passenger rail    Capital grants to states for intercity 8.0 1 7   passenger rail   Amtrak 1.3 91 100 Transit: Capital grants to transit agencies 8.4 62 86 Aviation   FAA facilities 0.2 65 94    Capital grants to airports 1.1 95 98 Multimodal (TIGER grants): Capital grants 1.5 5 38 to state or local government agencies for highways, transit, rail, or ports; awarded competitively Marine: Grants to small shipyards 0.1 76 92 All programs 48.1 56 73 Source: GAO 2011a, 5; GAO 2011b, 4, 8; Recovery Accountability and Transparency Board 2012a. 

68 Transportation Investments in Response to Economic Downturns Data Collection and Reporting Grant recipients were required to collect data and submit periodic reports on “the number of direct, on-project jobs created or sustained by the Federal funds provided . . . and, to the extent possible, the estimated indirect jobs created or sustained in the associated supplying industries” [ARRA Section 1201(c)]. Jobs were reported quarterly and measured in full-time equivalents, that is, hours worked divided by a standard num- ber of quarterly hours in a full-time job. Recipients also were required to report on funds obligated and expended and numbers of projects begun and completed. ADMINISTRATION OF ARRA TRANSPORTATION GRANTS The following were the apparent purposes of the special rules for trans- portation grants that ARRA imposed: • To ensure timely distribution and expenditure of the federal funds; • To ensure that the benefit of the federal funds was not offset by reductions in recipients’ spending of their own funds, through the maintenance-of-effort requirement; • To target spending to geographic areas experiencing the greatest economic hardship, through the priority assigned to projects in dis- tressed areas; and • To provide transparency through public reporting requirements. These are essential design objectives for a federal stimulus spending program. Timeliness is necessary for effective stimulus. Maintenance of effort is necessary to ensure that the federal grants actually lead to an increase in government purchases (also necessary to achieve the stimu- lus) of the kinds that Congress has determined would be most benefi- cial. Similarly, geographic targeting of funds may in some circumstances increase the stimulus impact [although experience with ARRA described below and with earlier public works spending programs (Vernez and Vaughan 1978, x) indicates that many of the direct hires on a construction project will not be resident in the locale of the project]. Transparency is 

Design, Management, and Outcomes of the ARRA Transportation Grants 69 necessary to motivate federal and state managers to use the funds con­ sistently with Congress’s intent and to allow the ARRA programs to be evaluated so that future stimulus programs can be better designed. As noted in Chapter 1, the spending deadlines and requirements for reporting of spending and employment were recommendations of the 1986 report of the General Accounting Office (renamed the Government Account- ability Office) (GAO) that identified failings in earlier federal public works stimulus spending programs (GAO 1986, 62). GAO (1986, 26) and the Economic Development Administration evaluation of public works stimulus programs (Vernez and Vaughan 1978, viii) had cautioned that the employment impact of the earlier stimulus programs would have been offset by state and local government substitution of federal funds for their own. Federal imposition of these requirements was seen as appropriate in ARRA because Congress was seeking to attain national welfare objec- tives, which were not necessarily consistent with the objectives of the individual jurisdictions receiving the transportation grants. The stimu- lus benefits of transportation construction tend to leak out of the state in which the stimulus occurs. For example, materials may be imported from out of state, and out-of-state contractors may be hired. If state gov- ernments received unrestricted federal grants in place of the transporta- tion grants, their incentives would be to use them in ways that retained the benefits for state residents, perhaps for tax relief or for government purchases with greater local impact but less national stimulus impact than transportation infrastructure. Of course, arguments for the federal requirements are valid only if Congress was correct in its assessment that purchases of transportation infrastructure and equipment were the most effective use of a portion of ARRA funds, with respect to the act’s objectives. The first subsection below describes difficulties that transportation grant recipients reported in attempting to comply with ARRA’s admin- istrative provisions. The second subsection considers how state and local government grant recipients’ responses to the program may have obscured its actual impact on transportation spending, total government spending, project selection, and spending rate. 

70 Transportation Investments in Response to Economic Downturns ARRA Administrative Challenges The experiences of the state and local government recipients in comply- ing with the ARRA transportation grant rules and of the U.S. DOT in administering the rules have been described in reports by GAO (GAO 2011b), the U.S. DOT (U.S. DOT 2011), and the U.S. DOT Inspector General (Scovel 2011) and in the paper on implementation of ARRA commissioned by the committee (Meyer 2012), which was based on interviews with officials in 11 states. All described administrative bur- dens created by certain of the rules, difficulties that some states expe- rienced in complying with the rules, concerns about the consistency of their application, and concerns about unintended consequences. The remainder of this subsection lists reported difficulties associated with each category of ARRA transportation grant administrative provision and finally gives state officials’ suggestions for alleviating the difficulties. Apportionment of Funds Most state officials interviewed for the committee affirmed that dis- tributing ARRA funds according to procedures of established federal transportation grant programs facilitated distribution of the funds and allowed projects to be initiated quickly (Meyer 2012, 10). In distributing funds in this way, the federal government may have forgone opportu- nities to target funds more directly to geographical areas in which the stimulus impact would be greatest. However, the uncertain results of the ARRA distressed areas priority requirement (described below) suggest that the practical potential for geographic area targeting is limited. Some state officials argued that spending could have been faster if greater effort had been made to facilitate or relax temporarily planning and environmental review steps required under the federal-aid program (Meyer 2012, 13). State officials reported that the lapse in the regular federal highway and transit aid program in 2009 affected their ability to maintain spend- ing during the recession (Meyer 2012, 8). In that year, the multiyear federal-aid program authorized in 2005 expired, and spending was con- tinued through a series of temporary extensions until Congress enacted a new program in 2012. The funding uncertainty slowed project design because states cannot afford to develop projects with a low probability of 

Design, Management, and Outcomes of the ARRA Transportation Grants 71 construction. Consequently, the lapse in the federal-aid program prob- ably reduced opportunities to spend ARRA funds on ready-to-construct, high-value projects. Funding uncertainty also complicated compliance with the ARRA maintenance-of-effort requirement. For the period for which each state was required to report its planned expenditures of its own funds (Febru- ary 2009 to September 2010), the states’ overall capital programs were uncertain because they had not known how much (pre-ARRA) federal aid they could expect to receive. Some states reported that the required 30 percent suballocation of state grants to local governments slowed spending because local govern- ments were less prepared to respond and that it made compliance with ARRA reporting and other requirements more difficult (Meyer 2012, 11). GAO’s analysis of spending data (see above) confirmed that the sub- allocated funds were spent more slowly. State officials reported confusion over award requirements for the pas- senger rail grant program administered by the Federal Railroad Adminis- tration (FRA) (Meyer 2012, 11). However, states appreciated the flexibility afforded through the TIGER grant program for seeking aid for a broad variety of types of transportation projects. Distressed Areas Priority GAO found “substantial variation” in how states identified distressed areas in their reporting of ARRA projects to the U.S. DOT and how states gave priority to projects in distressed areas. State officials explained to GAO that they had not received guidance on the objective of the dis- tressed areas requirement, that is, whether the intent was to provide immediate jobs in these areas or to provide infrastructure for long-term development (GAO 2011b, 26–27). In the interviews conducted for the committee, officials in one state reported that changing federal guidance on the definition of distressed areas complicated compliance and that the requirement tended to favor projects in low-population-density areas of the state, reducing the stim- ulus effect of spending (Meyer 2012, 9). State officials reported to the committee that if immediate employment is the objective, conducting highway construction in distressed areas has little effect, because con- tractors operate statewide and their workers travel to job locations. The 

72 Transportation Investments in Response to Economic Downturns evaluation of the Economic Development Administration public works stimulus program of the 1970s found similarly that “the ability of a pub- lic works program to serve local cyclical needs is limited. No more than three out of ten jobs created by a public works project—including direct, indirect, and induced jobs—may be created in the labor market area in which the project is implemented” (Vernez and Vaughan 1978, x). The U.S. DOT reported that in 2009 and the first quarter of 2010, more than 60 percent of all ARRA highway projects were located wholly or partially in distressed areas (U.S. DOT 2010a, 59). Distressed areas were identified according to state reports or according to a list of dis- tressed counties published in 2009 by the Bureau of Labor Statistics. Because the share of population or road miles in the distressed areas is not reported, it is not possible to ascertain from the U.S. DOT report the degree to which project selection favored these areas. GAO reported in 2010 that the U.S. DOT was overstating the fraction of spending that was in distressed areas because some states were not following the FHWA guideline for defining a distressed area (GAO 2010b, 76). Matching Funds Waiver Presumably, the intent of the ARRA waiver of the normal requirement that grant recipients contribute a share of the funding of each project receiving federal transportation aid was to ensure that no state or local government would be obliged to refuse the offer of extra federal aid for lack of matching funds. Neither the 2011 GAO report nor the interviews conducted by the committee examined whether the waiver was needed to ensure acceptance of federal aid or had other consequences. One rationale for the matching requirements that are a common feature of grant programs is that they are a means to add leverage to the grant, that is, to increase funding for the targeted purpose by more than the amount of the grant, or at least to discourage recipients from substituting the grant funds for their own funds. To the extent that the normal federal-aid matching requirements are effective for this purpose, the match waiver may have conflicted with the intent that ARRA funding not substitute for state and local transportation funding. Matching requirements also provide evidence that the grant recipi- ent is committed to a project and has resources available to support it, increasing the likelihood that the project will be carried out successfully. 

Design, Management, and Outcomes of the ARRA Transportation Grants 73 In the highway program, where federal aid amounts to a minority of state spending, the commitment and capabilities of the states may be evident. However, state passenger rail programs generally are less well developed, and therefore the absence of a state match requirement in the ARRA rail program may have increased the risk of grants going to recipients who lacked the resources to provide adequate incidental sup- port or to cope with contingencies like cost overruns. Obligation and Spending Deadlines GAO reported that 95 percent of U.S. DOT–administered ARRA funds had been obligated and 63 percent spent by May 31, 2011, 27 months after enactment. Obligation and spending in the new competitive grant programs, TIGER grants and high-speed rail, had been much slower than in other programs (Table 3-1), and GAO found that obligation and spending of funds that the states suballocated to local governments for locally selected and managed highway projects were slower than for state projects (GAO 2011b, 8–11). GAO also reported that, according to the federal and state officials it interviewed, the deadlines probably affected the types of projects selected for ARRA funding and that a “tension between the purposes of the Recovery Act” (creating jobs and investing in infrastructure for long- term economic benefits) existed; that is, in some instances, projects that could meet the deadlines were selected instead of alternatives that would yield greater long-term benefits but would require more time to initiate (GAO 2011b, 28). The state interviews commissioned by the committee reinforce the GAO observations. State transportation officials reported that the dead- lines were a primary factor in selecting projects for ARRA funding and that the types of projects chosen would have been different if the time constraints had been relaxed (Meyer 2012, 7–8). State officials repeat- edly expressed a preference for at least partial relaxation of the stringent obligation deadlines in ARRA while acknowledging the importance of immediate job creation as one of the goals of the act (Meyer 2012, 9–12). Apparently, these state transportation officials believed that ARRA did not strike the correct balance between the two goals of immediate stimu- lus and long-term economic benefit from transportation improvements. This perspective would be consistent with the responsibility of these 

74 Transportation Investments in Response to Economic Downturns officials, which is to develop and maintain the transportation network rather than to manage macroeconomic cycles. Maintenance of Effort GAO (2011b, 24–26), the U.S. DOT (2010b; 2011) and Meyer (2012, 9–10) documented the difficulties that the U.S. DOT and the states experienced in interpreting, administering, and complying with ARRA’s maintenance- of-effort requirement. Terms in the act such as “state funding” and “types of projects that are funded by the [ARRA] appropriation” required defini- tion. The states and the federal government ordinarily plan and track obli- gations of funds for transportation rather than expenditures, which were the measure of the ARRA maintenance-of-effort requirement. During 2009 and 2010, the U.S. DOT repeatedly revised its guidance on mainte- nance of effort concerning the definitions of planned and actual expen- ditures (GAO 2011b, 25). Some states objected that the requirement was not consistently applied across states and that U.S. DOT interpretations sometimes were unnecessarily rigid. States that failed to comply some- times objected that they were penalized for factors beyond their control, in particular, declines in dedicated transportation revenue during the reces- sion (Meyer 2012, 9–10). The funding uncertainty caused by the 2009 lapse in the regular federal-aid program further complicated compliance with the ARRA maintenance-of-effort requirement. For the time period for which each state was required to report its planned expenditures of its own funds (February 2009 to September 2010), the states’ overall capital programs were uncertain because they had not known how much (pre-ARRA) fed- eral aid they could expect to receive. In a 2010 report to Congress on ARRA, the U.S. DOT summarized its experience with defining and enforcing the maintenance-of-effort require- ment and obstacles that hindered states’ good-faith effort to comply. The U.S. DOT concluded that “it is difficult to craft and apply a quantitative measure that works well across multiple transportation programs involv- ing a large number of grantees with differing organizational, financial, and legal structures. . . . [A] pass-fail [maintenance-of-effort] standard like that found in §1201 [of ARRA] may not be well suited to U.S. DOT modal programs, because of the myriad factors that affect whether and when State funds are expended on a transportation project” (U.S. DOT 

Design, Management, and Outcomes of the ARRA Transportation Grants 75 2010b, 4). This conclusion appears to be based primarily on concern for fairness to the states (i.e., that a state not be penalized for factors beyond its control) rather than on an assessment of the requirement’s effectiveness in ensuring that federal grants translated into increased total spending. In a 2011 report, the U.S. DOT concluded that requirements in any future stimulus legislation should be modified to simplify defining and tracking maintenance of effort (U.S. DOT 2011, 12) but did not propose specific modifications. The U.S. DOT reported that by November 2010, 29 states and the District of Columbia had demonstrated that their spending had complied with the maintenance-of-effort requirement and that spending by 21 states had fallen short (U.S. DOT 2011, 2–3). Buy America Requirement Section 1605 of ARRA provided that “none of the funds appropriated or otherwise made available by this Act may be used for a project for the construction, alteration, maintenance, or repair of a public building or public work unless all of the iron, steel, and manufactured goods used in the project are produced in the United States.” The requirement was to be applied “in a manner consistent with United States obligations under international agreements,” and the act allowed federal department heads to waive the requirement if no U.S. source existed or if it would increase projects costs by more than 25 percent. States were already required to use U.S.-produced iron and steel in federal-aid highway projects, but the ARRA requirement applied more broadly. GAO identified the require- ment as a potential source of delay in the ARRA high-speed rail program (GAO 2010a, 20). Data Collection and Reporting GAO initially observed numerous difficulties of grant recipients in com- plying with ARRA jobs reporting requirements and of federal agencies in providing guidance and concluded that the reports were of unknown validity (GAO 2009, 40). However, by 2011, GAO found that reporting had improved and concluded that it was “generally satisfied with the stability of the data quality” (GAO 2011b, 21). Beyond these administrative difficulties, GAO emphasized that the jobs reports do not indicate the overall employment impact of ARRA. They cover only a minority of ARRA spending, do not include reporting 

76 Transportation Investments in Response to Economic Downturns or estimates of employment by suppliers of the projects or programs receiving the grants or of employment resulting when the workers directly employed spend their wages in the community, and do not take into account how recipients would have behaved if they had not received the grants. Consequently, “the recipient reports are not estimates of the impact of the Recovery Act, although they do provide a real-time win- dow on the results of Recovery Act spending” (GAO 2009, 42). The state officials interviewed for this study endorsed the principle of transparency, but some objected to the burden, which they regarded as excessive, of mandatory reporting to multiple federal agencies with differing reporting requirements (Meyer 2012, 9, 24–25). States were required to report on ARRA-funded activities to the U.S. DOT, to the Recovery Accountability and Transparency Board (created by ARRA to monitor funds), and to the House Transportation and Infrastructure Committee. Also related to transparency and accountability is the GAO finding that the potential long-term transportation benefits of ARRA-funded projects are unknown because the states lack data and methods for mak- ing such assessments. GAO recommended that the U.S. DOT determine the data and measures that would be needed to assess the long-term benefits (GAO 2011b, 15–16). Obviously, such an improvement in evalu- ation capabilities would be of value not only in administration of special programs like ARRA but also in all federal, state, and local transporta- tion activities. State Views on Improving Administration Frequently mentioned preferences of state transportation officials interviewed for the committee concerning the administration of a transportation stimulus spending program included the following (Meyer 2012, 12–13): 1. Select projects on the basis of greatest economic impact; do not arti- ficially constrain the selection with requirements for economically depressed designation and geographic balance. 2. Eliminate unnecessarily burdensome federal reporting requirements, especially duplicative reporting to multiple federal agencies. 

Design, Management, and Outcomes of the ARRA Transportation Grants 77 3. Allow flexibility in reallocating federal grant funds, for example, if a project encounters an unforeseen obstacle or bids come in under the estimate. 4. Allow a longer period of time for projects that will have a strong eco- nomic impact but that cannot be implemented in a short time frame. 5. Relax the process requirements if the goal is to get projects implemented quickly. The Transportation Improvement Program–State Transporta- tion Improvement Program planning requirements, environmental impact, and cultural resource requirements were pointed to in particu- lar as constraints in getting the right projects implemented. 6. Make engineering design eligible for stimulus dollars, which will allow projects that might not be implemented in the short time frame to be expedited as part of the normal investment program. It is to be understood that state officials’ preferences must be weighed against the nationwide objectives that the federal program was intended to achieve. Effect of Grant Rules on ARRA Outcomes The GAO and U.S. DOT reviews did not attempt to evaluate quanti- tatively the effect of the administrative rules on recipients’ behavior or on the overall outcome of the ARRA transportation program. The program’s structure and rules affected its success in achieving the two intended outcomes of immediate economic stimulus and long-term eco- nomic benefit from the transportation assets purchased. The outcomes depend on how ARRA affected governments’ total spending, total trans- portation spending, and transportation spending priorities. Conceiv- ably, altering ARRA rules could have resulted in a mix of projects with greater stimulus benefits (e.g., by accelerating construction) or greater long-term transportation benefits (e.g., by allowing projects that were difficult to fund under ARRA rules), or both. Historically, states have managed their participation in federal trans- portation aid programs with the goal of maximizing the grant funds received and complying with the letter of the federal regulations, while at the same time minimizing the degree to which the federal rules divert them from their own transportation and spending priorities (TRB 1987, 

78 Transportation Investments in Response to Economic Downturns 51–54). The new transportation grants in ARRA created incentives for the states to • Reduce the amount of their own funds devoted to transportation (fiscal substitution), to free funds to satisfy higher state priorities; • Designate as ARRA-funded already programmed or anticipated projects that could be undertaken quickly, to satisfy the timeliness requirement of ARRA; and • Divert administrative, engineering, and other resources to the start-up of ARRA projects and away from projects, including normal federal- aid projects, that were subject to more lenient deadlines, also to satisfy the timeliness requirements. These likely state responses complicate the task of determining the net impact of ARRA on total transportation spending, on total government spending, on the projects selected for funding, or on the acceleration of spending. ARRA’s maintenance-of-effort requirement was intended to prevent fiscal substitution, but as described above, maintenance of effort proved difficult to define and measure. Moreover, the receipt of the ARRA trans- portation grants may have discouraged state actions to maintain trans- portation spending through borrowing or other new sources of funds. Therefore, the effect of the ARRA maintenance-of-effort requirement is uncertain. Diversion of a part of ARRA transportation funding to other purposes would not necessarily indicate bad faith on the part of the states, since recipients had no certain means of determining what their spending priorities would have been in the absence of the grant. It also would not necessarily diminish the stimulus impact, depending on how the freed-up funds were used. (If the ultimate effect were additional state spending for any purpose or lower taxes, there would be a stimulus impact.) There is empirical evidence that, historically, increases in federal high- way aid distributed to the states have not led to equal increases in state highway spending. The structure of federal highway grants does not give the federal government strong leverage over state spending priorities. The state or local matching share is small (10 or 20 percent for most projects), and the total federal aid for which a state is eligible in a time period is capped. As long as a state plans more highway capital spending from its own funds than the minimum needed to match available federal aid, it can qualify for all federal aid for which it is eligible without greatly 

Design, Management, and Outcomes of the ARRA Transportation Grants 79 altering its governmentwide expenditure levels, priorities, or taxes from those that would have prevailed in the absence of the federal aid (GAO 2004; Gramlich 1990; TRB 2006, 174–175). GAO estimated in 2004 that the states had used roughly half of increases in federal highway aid since 1982 to substitute for state and local highway funding (GAO 2004, 3). Because ARRA grants required no state match, federal leverage over state transportation spending through ARRA may have been weaker than in the normal federal-aid program. The GAO estimate applies to changes in the established multiyear federal-aid programs. In the case of a short-duration grant program like ARRA, the structure of state transportation funding presents an obstacle to substitution. Most state transportation spending is funded by revenue from dedicated tax sources. Some states’ laws provide for emergency diversion of these funds, but overall, the trust fund arrangement likely hinders fiscal substitution in response to a short-term extraordinary grant program. ARRA project data show that pavement improvement was the most common ARRA project type, because such projects easily satisfy the ARRA timeliness requirements. However, state officials reported that the states identified as ARRA funded some projects that already had been programmed, or would have been programmed, within the ARRA spending deadline, freeing the federal aid and state funds that had been dedicated to those projects to be used for other purposes. Thus, even if a state reported that it spent all its ARRA funds on resurfacing, the net effect of the grants could have been to increase spending on other kinds of projects as well as on resurfacing. The U.S. DOT recognized that a state could blunt the effect of ARRA’s timeliness requirements by relabeling projects in this way. It instructed the states that it would allow them to convert already programmed, fully state-funded highway projects to ARRA projects only if the state funds freed were then used for other highway projects that met the ARRA timeliness requirements and that already-approved regular federal-aid projects (with shared federal and state funding) could not be converted to ARRA projects (with 100 percent federal funding) (FHWA 2009b). However, resurfacing is programmed on a short cycle, and rather than relabeling projects, a state could attain the same result by reducing the amount of state-funded resurfacing programmed in the period after receipt of ARRA funds. 

80 Transportation Investments in Response to Economic Downturns GAO found in 2010 that obligations and reimbursements in the regu- lar federal-aid highway program had slowed since enactment of ARRA. At the end of the third quarter of fiscal 2010 (June 30), $19.7 billion of regular federal-aid highway funds available to the states remained un­obligated, compared with an average of $12.1 billion at the same point in the 3 previous years. Federal reimbursements to states in the regu- lar highway program in the first 10 months of 2010 were $4.3 billion (18 percent) less than the average of the 3 previous years (GAO 2010b, 69–73). GAO reported that “some state officials told us they had not been obligating regular federal highway formula funds as quickly because they had been focusing on meeting the Recovery Act obligation deadlines and did not have the resources to do both” (GAO 2010b, 72). In summary, the states took action to maximize the federal aid received and at the same time attempted to balance the federal program’s priori- ties with each state’s own priorities in spending decisions. Because of these actions, the data on ARRA project spending levels, project types, and employment do not show the net effect of the grants on the rate and com- position of spending or on construction employment. These effects could be estimated empirically with a model based on historical data on state and local transportation spending by project type and revenue by sources of funds. The committee did not attempt to develop such a model. The next section describes the recent historical trends in spending and funding. TRANSPORTATION SPENDING AND CONSTRUCTION EMPLOYMENT DURING ARRA This section presents trends for indicators related to the ARRA transpor- tation program’s impact through the period of ARRA’s operation: • Obligations and expenditures of ARRA transportation funds and recipients’ reported uses of funds, • Total government highway and transit spending and sources of funds, • Employment on ARRA projects as reported by grant recipients, and • Total highway construction and maintenance employment. The section concludes with summary observations on the responses of state and local governments to the transportation grants and the impact of the grants on spending and transportation construction employment. 

Design, Management, and Outcomes of the ARRA Transportation Grants 81 The trends are suggestive of the impact of ARRA. However, to take into account the likely state and local government responses to ARRA grants described in the preceding section, a model that projected highway and transit spending and employment in the absence of ARRA would be needed. Obligations and Expenditures of ARRA Funds The grant recipient reporting that ARRA mandated indicates how suc- cessfully recipients satisfied ARRA timeliness requirements and the kinds of projects designated to receive ARRA funds. State and local government agencies receiving apportionments of ARRA transportation funds met essentially all the obligation deadlines in the major ARRA transportation programs: 50 percent of highway funds obligated by June 30, 2009, and 100 percent by March 2, 2010; 50 percent of transit funds by September 1, 2009, and 100 percent by March 5, 2010 (Committee on Transportation and Infrastructure 2011). Consequently, there was no redistribution of unobligated ARRA highway funds in 2010 (Mendez 2010). One hundred percent of the funds provided in ARRA for passenger rail grants to states were required to be obligated by September 20, 2012; by August 2012, FRA reported at least $7.4 billion of these funds obligated out of $8 billion appropriated (FRA n.d.). Obligation occurs when the U.S. DOT approves a state or local government agency proposal for use of ARRA funds on a specific project, thereby obligating the federal government to reimburse the agency’s expenditures for the project (GAO 2011b, 7). The expenditure of federal funds occurs when the grant recipient applies to the U.S. DOT for reimbursement for its expenditures on a project. Before this application, contracts are let and contractors start work, begin submitting invoices to the state or local government, and are paid. Expenditures reached 56 percent of U.S. DOT–administered funds provided by ARRA in March 2011 and 73 percent in March 2012. Eighty-nine percent of funds had been spent by March 2012 in U.S. DOT–administered programs excluding the two new competitive grant programs, intercity passenger rail (7 percent spent) and TIGER grants (38 percent spent) (Table 3-1). In comparison, 85 percent of CBO’s estimated budget impact of ARRA in total had occurred by June 30, 2011 (CBO 2011, 1), and 80 percent of all projected outlays to state and local governments under ARRA had been 

82 Transportation Investments in Response to Economic Downturns made by June 3, 2011 (Dodaro 2011). Thus, federal outlays in U.S. DOT– administered programs were somewhat slower than the average rate for all federal spending under ARRA. However, actual transportation con- struction spending occurs earlier than the federal expenditure is recorded in these data, because the federal government reimburses the states for expenditures they have already made; in some ARRA programs (e.g., aid to individuals), the federal outlay preceded actual spending by recipients. Recipients’ Reported Uses of Funds GAO, in its review of ARRA transportation spending, observed that ARRA’s obligation deadlines caused state and local government agencies selecting projects for ARRA funding to favor relatively simple projects that could be initiated quickly, in particular, pavement improvements in the highway program and bus purchases in the transit program. GAO also reported the concerns of officials in some states that projects with greater long-term transportation benefit could have been funded if the deadlines had not forced reordering of priorities (GAO 2011b, 28). Fifty percent of ARRA highway funding was obligated for pavement improvements (Table 3-2 and Figure 3-2). These projects are relatively small and routine, and most jurisdictions maintain a schedule of future pavement projects needed. Therefore, state and local governments could quickly identify, prepare, and contract for these projects and meet ARRA’s obligation deadline and requirement that projects that could be built promptly were to be favored. This share is much larger than pave- ment improvement projects receive in the federal-aid highway program normally: 16 percent of the total cost of all federal-aid highway projects initially authorized in 2006 through 2008 were for pavement resurfacing, restoration, and rehabilitation ($7.6 billion annually on average out of $47.2 billion average total cost of initially authorized federal-aid projects annually). This fraction rose to 19 percent in 2009 and 21 percent in 2010 (FHWA various years, Table FA-10). In the ARRA transit program, 51 percent of obligations were for infrastructure improvements and 23 percent for buses (Table 3-2 and Figure 3-3), a category that transit agencies found to be easy to under- take within the ARRA deadline. For comparison, 16 percent of all transit agency capital expenditures were for buses in 2006 through 2008 ($2.4 bil- lion annually on average out of $15.2 billion average annual total capital expenditures). This fraction rose to 18 percent in 2009 and 20 percent in 

Design, Management, and Outcomes of the ARRA Transportation Grants 83 TABLE 3-2  Highway and Transit Obligations of ARRA Funds, by Project Type Amount Awarded Percent Project Type ($ millions) of Total FHWA Programs (GAO 2011b, 10) Bridge construction, improvement, and replacement 3,100 12 New construction (nonbridge) 1,800 7 Widening 4,700 18 Pavement improvement 13,200 50 Other 3,300 13 Total 26,200 100 FTA Programs (FTA n.d. b) Operating expenses 193 2 Other capital expenses 980 11 Preventive maintenance 765 9 Transit infrastructure construction 4,479 51 Vehicle purchase or rehabilitationa 2,040 23 Rail car purchase or rehabilitation 322 4 Total 8,779 100 Note: All ARRA highway and transit funds were obligated by mid-2010, although some funds may have been reobligated later. Transit obligations exceed ARRA’s $8.4 billion transit appropriation because some states chose to exercise the option provided in the federal-aid program to transfer a share of their highway apportionment to transit programs. a Vehicle purchase or rehabilitation is entirely expenditures for buses (FTA n.d. b, Table 4a). 2010 (APTA various years). In contrast, the share of spending for rail car purchase and rehabilitation was much smaller in the ARRA transit pro- gram (4 percent) than in regular transit capital spending (12 percent in 2006–2008), presumably because rail car purchases are complex, multi­ year acquisitions. As noted above, the ARRA funds obligated for a particular project type do not necessarily indicate the net impact of ARRA on total spending for that type. Grant recipients may have been able to designate already- planned projects as ARRA funded, freeing the funds that would have been spent on those projects for other purposes (e.g., projects that would have been difficult to qualify for ARRA funding). The increase in the pavement improvement share of the cost of initially authorized federal-aid highway projects in 2009 and 2010 compared with the 2006–2008 average suggests 

84 Transportation Investments in Response to Economic Downturns FIGURE 3-2  ARRA obligations for FHWA programs, by project type. (Source: GAO 2011b, 10.) FIGURE 3-3  ARRA obligations for FTA programs, by project type. (Note: rehab = rehabilitation. Source: FTA n.d. b.) 

Design, Management, and Outcomes of the ARRA Transportation Grants 85 that ARRA funding may have accelerated pavement improvement proj- ects. However, state and local governments carry out substantial amounts of pavement projects outside the federal-aid program, and data on total pavement expenditures are not available for 2010. Therefore, the available data are inconclusive with regard to the impact of ARRA on the composi- tion of highway capital spending. Similarly, the increase in the bus expenditures share of total transit capi- tal expenditures in 2009 and 2010 compared with the 2006–2008 average suggests that ARRA funding may have accelerated bus purchases. If the 2006–2008 percentage had prevailed in 2009–2010, bus purchases would have been $1.2 billion less than the actual 2009–2010 expenditure. How- ever, the bus share was increasing throughout 2005–2008 (Figure 3-4); therefore, these data alone do not support a strong conclusion about the impact of ARRA on the composition of transit spending. Timeliness of Spending Assessment of the timeliness of stimulus spending depends on the char- acteristics of the recession that the spending is intended to ameliorate. FIGURE 3-4  U.S. total public transit capital expenditures and bus capital expenditures, 2005 to 2010. (Note: Bus capital expenditures are all expen- ditures for purchase and rehabilitation of vehicles other than rail vehicles. Source: APTA various years.) 

86 Transportation Investments in Response to Economic Downturns FIGURE 3-5  Months to recover to prerecession employment after recessions, 1948 to 2013. (Source: FRB Minneapolis 2013.) The 2007–2009 recession was extraordinary in its depth and in the per- sistence of depressed employment after growth restarted. However, the time required for recovery from recessions appears to be lengthening. In each postwar recession before 1973, the prerecession employment level was regained in 22 months or less from the start of the recession; in each recession from 1973 to the present (with the exception of the first phase of the 1980–1981 double-dip recession), return to prerecession employ- ment required more than 24 months. The three longest employment recoveries are from the most recent three recessions (Figure 3-5).2 ARRA was not enacted until 15 months after the start of the 2007– 2009 recession. Twenty-two percent of ARRA public investment spending [a category of ARRA programs defined by the Council of Economic Advisors and including federal spending and grants for health, education, 2 Possible causes of this pattern have been examined by Coibion et al. (2013), who suggest that “if business cycles have become systematically more protracted affairs, as seems to be the case, then discretionary fiscal policy responses should target longer-lived projects rather than transitory transfer payments” (p. 32). 

Design, Management, and Outcomes of the ARRA Transportation Grants 87 energy, transportation, and other infrastructure (CEA 2013)] and 54 percent of the act’s stimulus (spending and tax cuts) other than public investment occurred while employment was still declining (Figure 3-6). Transportation spending, excluding the two new grant programs (high- speed rail and TIGER), was about as timely as all other categories of ARRA public investment. Virtually all stimulus in both categories will have been disbursed while employment remains depressed. If a program like ARRA had been enacted in response to the much milder and shorter 1990 recession, on the same schedule with respect to the start of spending (i.e., 15 months after the start of the reces- sion) and the rate of spending, 6 percent of public investment spending and 21 percent of other stimulus would have occurred by the month in which actual unemployment reached its minimum (February 1992), and most public investment spending would not have occurred until 10 100 CumulaƟve ARRA outlays and tax reducƟons 9 ARRA sƟmulus other than 90 public investment ARRA public Percent change in employment from 8 (right axis) 80 7 investment 70 (% of total through 2012:Q3) sƟmulus 6 (right axis) 1990 60 5 recession 50 start of recession 4 40 3 30 2 20 2001 1 recession 10 0 0 -1 0 12 24 36 48 60 -2 -3 2007 -4 recession -5 -6 -7 Months from start of recession FIGURE 3-6  Change in U.S. employment from start of recession, 1990, 2001, and 2007 recessions, and cumulative ARRA stimulus spending from 2009 to 2012. (Note: The Council of Economic Advisers defines public invest- ment to include federal spending and grants for health, education, energy, transportation, and other infrastructure. Total ARRA public investment outlays through the third quarter of 2012 were $241 billion. Total ARRA outlays other than public investment (tax cuts, aid to individuals, and state fiscal relief) through the third quarter of 2012 were $527 billion. Q = quarter. Source: FRB Minneapolis 2013; CEA 2013, 6.) 

88 Transportation Investments in Response to Economic Downturns after employment had already passed its prerecession peak. Spending on the ARRA schedule would not have been as badly timed in response to the 2001 recession: 32 percent of public investment spending and 72 percent of other stimulus would have occurred during the months in which employment in actuality was still declining, and 85 percent of public investment and 98 percent of other stimulus would have occurred before the month in which employment actually returned to its prerecession level. (Such comparisons do not take into account any effect that a stimulus program would have had on the course of employ- ment in the earlier recessions.) Figure 3-6 illustrates that the lag between the beginning of the reces- sion and enactment of ARRA was a major impediment to timely spend- ing. If ARRA had been enacted 6 months earlier (9 months after the start of the recession rather than 15), 45 percent of public investment spending and 83 percent of other stimulus would have been disbursed by the date that employment reached its minimum in actuality (Febru- ary 2010). If a stimulus program structured like ARRA had been enacted 9 months after the start of the 2001 recession, most public investment spending would have occurred in the period in which actual employ- ment was declining. Total Public Highway and Transit Spending and Sources of Funds The impact of the ARRA transportation grants depends on how they changed total transportation spending. ARRA required that each state receiving ARRA transportation funds certify that it would maintain the level of spending it had planned before passage of the act, in the period February 2009 through September 2010, for the types of projects eli- gible for ARRA funding. By the act’s standard, if ARRA’s maintenance- of-effort requirement had the full intended effect, total state government highway and transit spending should have increased over the certified planned amounts in the period from February 2009 to September 2010 by an amount equal to the ARRA funds that the states reported spend- ing. However, the meaningful baseline for assessing the impact of ARRA is the level of transportation spending that would have occurred if ARRA 

Design, Management, and Outcomes of the ARRA Transportation Grants 89 had not been enacted. Spending would have been influenced not only by the states’ pre-2009 plans but also by revenue from dedicated taxes and each state government’s overall fiscal condition during 2009 and 2010. This baseline may have been below the pre-ARRA planned spending level or the historical trend. Maintenance of effort is difficult to assess because this baseline cannot be observed (although the fact that at least 29 states satisfied the act’s maintenance-of-effort requirement is evidence that by that standard, ARRA increased total spending above baseline). If total spending increased above the without-ARRA baseline by less than the amount of ARRA aid received—either because state and local governments substituted the new federal aid for other resources or because the administrative and engineering resources that ARRA demanded slowed other projects—the stimulus effect of the ARRA transportation funding would be muted. Depressed economic condi- tions persisted after most ARRA funds had been spent in 2009–2011; therefore, if state and local governments reduced transportation spend- ing in 2012 and 2013 because the ARRA funding eliminated some needs, the overall stimulus impact of ARRA also would be reduced. This subsection presents historical data on U.S. spending and sources of funds for highways and public transit. The committee examined these trends for indications of the effect of the 2009–2011 injection of ARRA funds on total spending and for any indication of substitution of ARRA funds for other sources (i.e., a reduction in transportation funding from other sources in response to receipt of the ARRA funds). The evidence on these questions is not conclusive without modeling to estimate the spending that would have occurred in the absence of ARRA. Highway Spending Total highway spending (for construction, maintenance, and operation) in current dollars grew strongly throughout the past decade, with no pause in the recession years of 2007 to 2009, until growth slowed in 2010, according to data from the U.S. DOT’s regular annual survey of all the states (Figure 3-7). Growth in capital expenditures from 2007 to 2010 was stronger than growth in maintenance and operating expenditures (Figure 3-8). Maintenance and operating spending declined in 2008, the first full year of the recession, before recovering in 2009 and 2010. As 

200,000 180,000 160,000 140,000 120,000 $ Millions 100,000 80,000 60,000 40,000 20,000 - 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Disbursements for highways, excluding debt service, all units of government Highway user revenue, all units of government Disbursements for highways by state governments, excluding debt service and transfers to local governments Federal payments to state and local governments for highways FIGURE 3-7  Highway spending, federal payments to state and local governments, and revenues from users, current dollars, 2002 to 2011. (Source: FHWA various years, Table HF-10.) 100,000 80,000 $ Millions 60,000 40,000 20,000 0 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Capital outlay, all units of government Maintenance and operaƟon, all units of government FIGURE 3-8  Capital expenditures and operating expenditures for highways, current dollars, 2002 to 2011. (Source: FHWA various years, Table HF-10.) 

Design, Management, and Outcomes of the ARRA Transportation Grants 91 with all federal highway aid, ARRA highway funding and other funds added to the federal-aid program in the period were to be used entirely for capital spending.3 Federal payments to state and local governments for highways fluc- tuated in the first part of this period and then grew steadily from 2007 to 2010. Federal payments to state and local governments for highways were $34.1 billion in 2007, $37.1 billion in 2008, $39.5 billion in 2009, and $43.5 billion in 2010. Thus, federal payments in 2009 and 2010 were cumulatively $14.8 billion greater than they would have been if the 2007 level of payments had prevailed in those two years. Presumably, this incre- ment reflects ARRA spending (total ARRA highway aid was $27.5 billion); however, the acceleration of federal aid began in 2008, before ARRA was enacted. (Federal payments lag state spending because, on construction projects eligible for federal aid, the states first pay contractors and then apply to the federal government for reimbursement.) Total highway spending in 2010 was $26.9 billion (24 percent) greater than in 2007 in current dollars, as reported by the U.S. DOT. Federal spending (aid to state and local governments and direct federal spend- ing) in 2010 was $10.7 billion higher than in 2007; therefore, state and local governments were able to increase annual highway spending from their own taxes and borrowing by $16.2 billion in 2010 compared with 2007. The increase was not derived from highway user revenues (motor fuel excise taxes, tolls, vehicle registration fees, and licensing fees, mostly dedicated by law to transportation purposes), the traditional source of most highway funds; user revenues were stagnant after peaking in 2007 (Figure 3-7). Bid prices for highway construction were declining in this period; FHWA’s National Highway Construction Cost Index in 2010 was 18 percent below the 2007 level (Figure 3-9). As federal highway user fee revenue began to decline in 2008, Con- gress acted to sustain the transportation aid program (Figure 3-10) by extraordinary payments to the Highway Trust Fund from the general 3 The Census of Governments trend in state highway spending in this period diverges from spend- ing reported by the U.S. DOT. In Census data, state highway spending in 2010 is 7 percent higher than in 2007 (U.S. Census Bureau 2013); in U.S. DOT data, 24 percent (both including state transfers to local governments for highways). The source of the discrepancy is not evident. 

92 Transportation Investments in Response to Economic Downturns FIGURE 3-9  FHWA National Highway Construction Cost Index, 2003 to 2012. (Source: FHWA 2013a.) FIGURE 3-10  Federal disbursements for highways and federal highway user fee revenue, 2002 to 2010. (Note: Highway user revenue includes revenue dedicated to highways, as well as revenue dedicated to other purposes or deposited in state general funds. Federal disbursements include federal aid to other units of government and direct federal spending. Total disburse- ments by all units of government are net of intergovernmental transfers. Source: FHWA various years, Table HF-10.) 

Design, Management, and Outcomes of the ARRA Transportation Grants 93 fund, amounting on net to $8.0 billion in 2008, $7.0 billion in 2009, and $13.6 billion in 2010 (FHWA 2013b, Table FE-210), in addition to the ARRA funds available for highways in 2009 and 2010. Not all the added funds were spent in these years; the balance in the trust fund increased from $8.1 billion at the end of fiscal year 2007 to $20.7 billion at the end of fiscal year 2010 (FHWA 2013b, Table FE-210). These extra federal contributions, financed by borrowing and replacing the decline in user revenue, were in effect another stimulus, comparable in magnitude with the ARRA highway appropriation. The additions to federal-aid funds were sufficient only to maintain federal aid at the historical level as a percentage of total highway spending and contributed to the decline in reliance in highway user–derived revenue as a source of highway funding after 2007 (Figure 3-11). As noted above, the GAO review of the ARRA transportation pro- grams found that obligations and reimbursements in the regular federal-aid highway program slowed in the period in which the states were obligating their ARRA highway funds. The increases in highway FIGURE 3-11  Ratio of highway user revenue to total highway disbursements and ratio of federal disbursements for highways to highway disbursements by all units of government, 2002 to 2010. (See note to Figure 3-10. Source: FHWA various years, Table HF-10.) 

94 Transportation Investments in Response to Economic Downturns spending in 2009 and 2010 might have been greater if state and local government administrative and engineering resource constraints had not limited the rate at which new federal aid could be absorbed. The reduction in highway user revenues after 2007 was a consequence of the recession. State motor fuel tax receipts declined by 6.4 percent from their 2007 peak to the low point in 2009 (FHWA 2012, Table MF-201). State motor fuel tax revenue had declined year-over-year in the past only in 1974, 1942, and 1943 (FHWA 1997, IV-66). At the same time, highway funding from state general fund appropriations and from special state taxes other than user taxes, which had been rising during the decade, was well below its 2007 peak in 2008–2010 (Figure 3-12). The states were able FIGURE 3-12  State taxes and fees devoted to highways, other than user revenues, and state bond proceeds devoted to highways, 2002 to 2010. (Note: State taxes and fees devoted to highways other than user revenue include (a) general fund appropriations for highways and (b) revenue dedicated by law to highways from taxes other than highway user taxes. Source: FHWA various years, Table HF-10.) 

Design, Management, and Outcomes of the ARRA Transportation Grants 95 to increase highway spending from 2007 to 2010 [$16.6 billion (17 per- cent) higher in 2010 than in 2007] because of increased federal aid to the states ($8.9 billion greater in 2010 than in 2007) and increased bond sale proceeds ($7.9 billion greater in 2010 than in 2007) (FHWA vari- ous years, Table HF-10). Local government bond issues for highways increased in a similar pattern in the period. Transit Spending Transit capital expenditures, in current dollars, grew strongly in the 2005–2008 period, but the trend flattened in 2008–2010, according to transit agency financial data reported by the American Public Transpor- tation Association (APTA) (Figure 3-13). ARRA provided $8.4 billion for transit (see Figure 3-1), all but minor amounts of which was to sup- port capital expenditures. Apparently most of these funds are not yet reflected in the APTA 2009 and 2010 capital funding data. The 2-year total of federal funding for transit capital spending in 2009–2010 was $3.3 billion greater than it would have been if the 2006–2007 level had prevailed, and federal capital assistance had been on an increasing trend before enactment of ARRA. Funding for transit operation in current dollars also grew during 2008–2010, but more slowly than earlier in the decade. Increases in funding from fare collections and from federal assistance offset a small decline in state and local government operating assistance (Figure 3-14). Cyclical Behavior of State and Local Government Transportation Spending The data presented above are insufficient to reveal any cyclical pattern in transportation spending. Analyses of the correlation of government spend- ing with the business cycle over longer time periods have found that state and local government spending, excluding public assistance programs that are designed to increase spending during recessions, is procyclical; that is, spending tends to accelerate in response to expansion of the general economy and to decelerate in response to recessions. State government expenditures during the period from 1960 to 2006 other than safety-net programs (e.g., Medicaid and public assistance) tracked the business cycle with about a 1-year lag, with deviations in state spending from the long- term trend nearly proportional to deviations in personal income (Clemens 

96 Transportation Investments in Response to Economic Downturns FIGURE 3-13  Public transit capital funding sources, 2002 to 2010. (Note: Transit agency–generated capital funding includes funds from fares and other agency earnings, proceeds of bonds issued by the agency, and revenue from taxes imposed by the transit agency. Source: APTA 2012, Table 58.) and Miran 2012, 48–49). Total state and local government capital spending in the 1977–2008 period was procyclical but lagged the business cycle by 1 to 2 years (GAO 2011c, 16). Noncapital highway spending by state and local governments also was procyclical in 1977–2008, with a more imme- diate response to the business cycle (GAO 2011c, 19, 47). GAO (2011c) did not examine highway capital spending or transit spending as separate categories. The long-term trend in these periods was growth in transporta- tion spending and in total state and local government capital spending; the typical response to recessions was slower growth in spending rather than absolute declines. 

Design, Management, and Outcomes of the ARRA Transportation Grants 97 FIGURE 3-14  Public transit operating funding sources, 2002 to 2010. (Source: APTA 2012, Table 63.) The procyclical pattern in state and local government spending may have several undesirable consequences. It may amplify the magnitude of business-cycle fluctuations. Funding cutbacks will delay completion and likely increase the costs of long-term capital projects in progress. As noted above, construction prices generally are higher during expansions. Unless the slowdown in spending during a recession matches a decline in use of transportation facilities, customer service will suffer. Some of the features of transportation programs may have tended to dampen business cycle fluctuations in spending in the past. The federal and state trust funds, reliance on dedicated revenues, and the multiyear predictability (until recent years) of the federal-aid program have been sources of stability. The spending data presented above suggest that, dur- ing and after the 2007–2009 recession, the federal and state actions to maintain funding may have more than offset the historical tendency for spending to slow during a recession. 

98 Transportation Investments in Response to Economic Downturns Employment on ARRA Projects ARRA required grant recipients to submit periodic reports to the federal government on the number of jobs directly supported by the grant funds. For construction projects, such jobs included employees of contractors and subcontractors, but not employees of suppliers to contractors. The recipient jobs reports are compiled by the Recovery Accountability and Transparency Board, a body created by ARRA. The direct jobs supported as reported by recipients of U.S. DOT grants peaked at 67,000 full-time equivalents in the 2010 third quarter (Figure 3-15) (i.e., 67,000 person quarters of full-time employment dur- ing the quarter). Jobs supported declined to 18,000 by the 2012 second quarter, 27 percent of the peak. For comparison, direct jobs supported by all ARRA spending subject to the recipient jobs reporting requirement peaked at 750,000 in the 2010 second quarter and declined to 152,000 (20 percent of the peak) in the 2012 second quarter (Figure 3-16). About a fifth of the total of spend- ing and tax reductions provided by ARRA was subject to the reporting requirement (CBO 2012, 4). As Chapter 1 explained, the grant recipient jobs reports are not a good indicator of the overall employment impact of ARRA because they cover only a minority of ARRA spending and do FIGURE 3-15  Direct jobs supported by U.S. DOT–administered ARRA grants, quarterly: 4th quarter 2009 to 2nd quarter 2012. (Source: Recovery Account- ability and Transparency Board 2012b.) 

Design, Management, and Outcomes of the ARRA Transportation Grants 99 FIGURE 3-16  Direct jobs supported by all ARRA spending subject to the recipient jobs reporting requirement and by U.S. DOT–administered ARRA grants, quarterly: 4th quarter 2009 to 2nd quarter 2012. (Source: Recovery Accountability and Transparency Board 2012c, 2012b.) not reflect secondary effects of the spending. The jobs reports are useful as an indication of the timeliness of spending in the various grant pro- grams funded by ARRA. The federal budget cost per direct job supported was $318,000 per per- son year of employment (for the 2009 fourth quarter through the 2012 first quarter) in U.S. DOT–administered ARRA programs. In all ARRA programs for which recipient reporting of employment is required, the cost was $162,000 (for the 2009 fourth quarter through the 2012 sec- ond quarter) (Recovery Accountability and Transparency Board 2012a, 2012b, 2012c; CEA 2011). The relatively high cost per job reported for the ARRA transportation programs does not show that these programs were less cost-effective than the others; employment by suppliers to the projects is not known, and public benefits of the facilities constructed offset the cost to the government. Supplier employment may be sub- stantial for highway construction projects that require large quantities of materials. 

100 Transportation Investments in Response to Economic Downturns Total Highway Construction and Maintenance Employment In Bureau of Labor Statistics (BLS) Occupational Employment Statistics (OES) survey data, employment in the month of May in the highway, street, and bridge construction industry peaked in 2006 at 361,000 and fell to 297,000 in 2010, a decline of 18 percent. For comparison, BLS-reported employment in the entire construction industry declined 26 percent in the period. Highway maintenance employment (nearly all government employees) rose 4 percent in the same period, from 139,000 to 144,000. Employment in transportation construction other than highway con- struction is not identifiable in the published BLS industry employment statistics. (It would be included within the “other heavy and civil engi- neering construction” industry.) BLS explains that “the OES methodology . . . makes it difficult to use OES data for comparisons across short time periods” and that it “does not encourage the use of OES data for time-series analysis” because the share of total employment reported for an industry or occupation is estimated as a moving average of results of surveys of six semiannual panels of establishments collected over 3 years. Therefore, for example, a sudden and permanent step increase in an industry’s share of total employment would appear in the reported employment estimates as a gradual change over 3 years. In the BLS estimates, employment in the highway, street, and bridge construction industry declined monotonically for 5 years (from May 2006 through May 2011); therefore, the estimates are evidence of an actual decline. It is not evident how the declining trend in the BLS highway construc- tion employment data are to be reconciled with the U.S. DOT data on highway spending, which show current dollar highway spending grow- ing through 2010 (while the FHWA highway construction cost index was falling). One possible source of the difference may be declining employ- ment in privately funded road construction connected with housing construction (which would be reflected in the Census data but not in the U.S. DOT data). Summary Observations The stimulus impact of the ARRA transportation programs depends first of all on how the ARRA funds affected total spending and employment in public transportation construction and operation. The change in total 

Design, Management, and Outcomes of the ARRA Transportation Grants 101 spending was not necessarily equal to the ARRA funds spent. Recipients may have substituted ARRA funds for funds from other sources, and the effort of qualifying for and spending the ARRA funds diverted personnel or other agency resources away from other projects (or, conceivably, the new federal funds may have attracted additional state and local govern- ment resources to transportation and away from other uses). Moreover, the long-term transportation benefit of the ARRA transportation spend- ing does not necessarily depend solely on the benefits of the projects that were designated as ARRA funded, because grant recipients may have designated some already planned projects to receive ARRA funds, freeing funds previously designated for those projects for other uses. These uncertainties with regard to the ultimate effects of the ARRA transportation grants on state and local government spending are not indications that the grants were ineffective, either as stimulus or in their contribution to transportation infrastructure. To the extent that any diverted grants supported other spending or prevented tax increases by financially strapped states and localities, the funds would still have a stimulative effect. Funds retained in transportation programs contrib- uted to infrastructure development, even if the projects ultimately gaining funding did not meet all ARRA special requirements. However, to assess the ultimate impact of the ARRA grants, modeling would be required to isolate the effects of the grants on transportation spending and on total state and local government spending. The trends presented above none- theless suggest some limited conclusions about the functioning and impact of the ARRA transportation programs. Obligations and Expenditures of ARRA Funds • Essentially all state and local government grant recipients met the stat- utory deadlines for obligation of ARRA transportation funds appor- tioned to them. • The rate of expenditure of ARRA transportation funds was somewhat slower than the average rate for all ARRA funds. Part of the difference is attributable to slow spending in the two new transportation competi- tive grant programs, the TIGER grants and intercity passenger rail. The competitive grant programs may be regarded as the component of the ARRA transportation programs that was aimed at reinvestment rather than primarily at stimulus. However, the competitive grant programs 

102 Transportation Investments in Response to Economic Downturns were subject to short deadlines, implying a stimulus objective. The state interviews conducted for the committee (Meyer 2012) suggest that state officials’ preferred means of balancing short-term stimulus and long- term transportation benefits would have been to allow grant recipients greater flexibility with respect to deadlines in all the ARRA transporta- tion programs. • The ARRA deadlines apparently affected the types of capital expen- ditures selected to receive ARRA funds. The share of ARRA highway funds applied to pavement improvement projects was greater than the share of such projects in the normal federal-aid program. Similarly, the share of ARRA transit funds devoted to bus capital expenditures was greater than such expenditures’ average share of all transit capi- tal spending in recent years. However, the data presented above do not clearly show whether total transportation agency expenditures on these categories (pavement and buses) increased by as much as the ARRA funds applied to them, compared with the expenditures that would have occurred in the absence of ARRA. • Requirements for recipients to comply with new administrative rules to obtain stimulus funds were not conducive to timely spending. New rules were applied not only in the new discretionary grant programs but also with regard to record keeping, maintenance of effort, the buy- American requirement, and other ARRA provisions. The need for federal agencies to promulgate new rules and for recipients to com- ply probably retarded the start of some ARRA-funded projects and may have contributed to the slowdown in spending of regular federal transportation aid funds during the ARRA program. The administra- tive requirements were intended to serve legitimate needs for federal oversight and public accountability. However, the ARRA experience suggests ways the requirements could be streamlined in any future stimulus program, as Chapter 4 will propose. Most important, if rules were in place in advance of enactment of a stimulus spending pro- gram, administrative delay and uncertainty could be reduced. Total Transportation Spending • State and local governments were able to maintain growth in highway spending throughout the recession despite flat or declining user tax revenue by means of three sources of funds: increases in state and 

Design, Management, and Outcomes of the ARRA Transportation Grants 103 local government bond issues, ARRA highway funds, and the extra­ ordinary transfers from the federal general fund to the Highway Trust Fund in 2008, 2009, and 2010 that allowed the regular federal-aid pro- gram to continue normal operation. These trend data alone do not indicate the amount by which highway spending was greater in the period compared with the spending that would have occurred in the absence of ARRA. • According to GAO, obligations and reimbursements in the regular federal-aid highway program slowed in the period in which state and local governments were occupied in obligating their ARRA highway funds. This capacity constraint on state and local governments’ ability to process federal aid muted the stimulus effect of ARRA. • The trends for transit construction and operations spending are simi- lar to those for highways: spending growth (in current dollars) was sustained during the recession not only by ARRA funds but also by increased state and local government support and increased funding through the regular federal-aid program. Spending growth slowed in 2010, but capital spending remained at an elevated level compared with earlier in the decade. • Historically, state and local governments’ total capital spending and noncapital highway spending have been procyclical, tending to accel- erate when the economy is expanding and decelerate in response to recessions. The extraordinary transportation funding provided by the federal and state governments in 2008–2010 appears to have offset the spending slowdown that would have been expected on the basis of the historical pattern. ARRA-Supported and Total Highway Construction Employment • The grant recipient–reported jobs supported by ARRA projects indi- cate the timing of spending, but the relationship of these data to the actual direct employment impact of ARRA transportation spend- ing is unclear. To the extent that any of the jobs supported were in regions or in occupations with relatively low unemployment, the recipient-reported jobs data will tend to overstate the direct actual employment impact. However, such figures show only direct jobs, not the jobs at suppliers or in the broader economy that the spend- ing may support. 

104 Transportation Investments in Response to Economic Downturns • The federal expenditure per person year of work directly supported by ARRA spending was higher for projects in U.S. DOT–administered programs ($318,000 per person year) than the average in all ARRA pro- grams for which grant recipient jobs reporting was required ($162,000 per person year). This difference does not indicate that public benefits necessarily would have been increased by shifting ARRA funds from transportation to other programs, because the jobs reports do not include employment by suppliers to the ARRA-supported projects and because the expenditure per job measure does not take into account the long-term benefits of the supported projects. • BLS highway construction and maintenance employment during the 2007–2009 recession appears to be inconsistent with the trend in highway spending in the period reported by the U.S. DOT. The com- mittee did not determine the origin of the discrepancy. Many ARRA transportation construction projects would have been carried out at a later date if the ARRA funding had not been received. In some circumstances, this would be a favorable feature of the use of transportation aid as stimulus—the net budget cost of the stimulus is less than if stimulus spending is for purchases that would never occur in normal times, and spending during the recession gives government the advantage of lower prices. If the accelerated projects would have been carried out during a future period of full employment, there is no loss of stimulus impact. However, if projects are brought forward from a later time in which high unemployment will still prevail, the net stimulus effect is diminished. Transportation spending may have been reduced in 2011 and 2012, when unemployment was still high, because ARRA funding had accelerated projects that would have been completed in those years. REFERENCES Abbreviations APTA American Public Transportation Association CBO Congressional Budget Office CEA Council of Economic Advisers FHWA Federal Highway Administration FRA Federal Railroad Administration 

Design, Management, and Outcomes of the ARRA Transportation Grants 105 FRB Federal Reserve Bank FTA Federal Transit Administration GAO General Accounting Office or Government Accountability Office TRB Transportation Research Board U.S. DOT U.S. Department of Transportation APTA. 2012. Public Transportation Fact Book 2010: Appendix A: Historical Tables. APTA. Various years. Public Transportation Fact Book. (Annual.) CBO. 2011. Estimated Impact of the American Recovery and Reinvestment Act on Employ- ment and Economic Output from April 2011 Through June 2011. Aug. CBO. 2012. Estimated Impact of the American Recovery and Reinvestment Act on Employ- ment and Economic Output from October 2011 Through December 2011. Feb. CEA. 2011. The Economic Impact of the American Recovery and Reinvestment Act of 2009: Eighth Quarterly Report. Dec. 9. CEA. 2013. The Economic Impact of the American Recovery and Reinvestment Act of 2009: Ninth Quarterly Report. Feb. 1. Clemens, J., and S. Miran. 2012. Fiscal Policy Multipliers on Subnational Government Spending. American Economic Journal: Economic Policy, Vol. 4, No. 2, May, pp. 46–68. Coibion, O., Y. Gorodnichenko, and D. Koustas. 2013. Amerisclerosis? The Puzzle of Rising U.S. Unemployment Persistence. Final conference draft to be presented at the fall 2013 Brookings Panel on Economic Activity. Sept. 9. Committee on Transportation and Infrastructure. 2011. The American Recovery and Reinvestment Act of 2009 Transportation and Infrastructure Provisions Implementa- tion Status as of June 30, 2011. U.S. House of Representatives, Aug. 9. Dodaro, G. 2011. Meeting the Fiscal and Performance Challenges Facing Government. GAO, June 15. FHWA. 1997. Highway Statistics Summary to 1995. FHWA. 2007. Fact Sheets for Highway Provisions in the Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users (SAFETEA-LU). Oct. 25. FHWA. 2009a. FHWA Supplemental Guidance on the Determination of Economi- cally Distressed Areas Under the Recovery Act. Aug. 24. http://www.fhwa.dot.gov/ economicrecovery/guidancedistressed.htm. FHWA. 2009b. Questions and Answers on American Recovery and Reinvestment Act of 2009 (ARRA) Issues Raised by State DOTs. Revised Nov. 11. http://www.fhwa.dot. gov/economicrecovery/qandas.htm#a2. FHWA. 2012. Highway Statistics 2010. FHWA. 2013a. Construction Cost Trends for Highways. Jan. http://www.fhwa.dot.gov/ policyinformation/nhcci.cfm. 

106 Transportation Investments in Response to Economic Downturns FHWA. 2013b. Highway Statistics 2011. Jan. FHWA. Various years. Highway Statistics. (Annual.) FRA. n.d. HSIPR Project Funding. Accessed Sept. 10, 2012. http://www.fra.dot.gov/rpd/ HSIPR/ProjectFunding.aspx. FRB Minneapolis. 2013. The Recession and Recovery in Perspective. Updated Aug. 2. http://www.minneapolisfed.org/publications_papers/studies/recession_perspective/. FTA. n.d. a. ARRA Operating Assistance. Accessed Sept. 28, 2012. http://www.fta.dot.gov/ about/12350_10399.html. FTA. n.d. b. ARRA Statistical Summary. Accessed Sept. 13, 2012. http://www.fta.dot.gov/ about/12350_14246.html. GAO. 1986. Emergency Jobs Act of 1983: Funds Spent Slowly, Few Jobs Created. Dec. GAO. 2004. Federal-Aid Highways: Trends, Effect on State Spending, and Options for Future Program Design. Aug. GAO. 2009. Recovery Act: Recipient Reported Jobs Data Provide Some Insight into Use of Recovery Act Funding, but Data Quality and Reporting Issues Need Attention. Nov. GAO. 2010a. High Speed Rail: Learning from Service Start-Ups, Prospects for Increased Industry Investment, and Federal Oversight Plans. June. GAO. 2010b. Recovery Act: Opportunities to Improve Management and Strengthen Accountability over States’ and Localities’ Uses of Funds. Sept. GAO. 2011a. GAO Releases Most Recent Report on the Recovery Act. http://www.gao. gov/recovery/. Accessed Sept. 16, 2011. GAO. 2011b. Recovery Act: Funding Used for Transportation Infrastructure Projects, but Some Requirements Proved Challenging. June. GAO. 2011c. State and Local Governments: Knowledge of Past Recessions Can Inform Future Federal Fiscal Assistance. March. Gramlich, E. 1990. How Should Public Infrastructure Be Financed? In Is There a Shortfall in Public Capital Investment? (A. Munnell, ed.), FRB Boston, Boston, Mass. Mendez, V. 2010. Notice: Subject: Implementation of the 1-Year Redistribution of Highway Infrastructure Investment Funds Pursuant to the American Recovery and Reinvestment Act of 2009. March 17. http://www.fhwa.dot.gov/legsregs/directives/ notices/n4510721.htm. Meyer, M. 2012. Impact of Program Implementation on the Effectiveness of the American Recovery and Reinvestment Act: The Case of Transportation. Paper commissioned by the committee, May 24. Recovery Accountability and Transparency Board. 2012a. Contract, Grant and Loan Pro- grams. Accessed May 29. http://www.recovery.gov/Transparency/fundingoverview/ Pages/contractsgrantsloans-details.aspx. 

Design, Management, and Outcomes of the ARRA Transportation Grants 107 Recovery Accountability and Transparency Board. 2012b. Department of Transporta- tion Estimated Jobs by Quarter. Updated Sept. 19. http://www.recovery.gov/Pages/ textview.aspx?data=AgencyInfoFundedJobs&AGENCY_CODE=69. Recovery Accountability and Transparency Board. 2012c. Jobs Summary—National. Updated Sept. 19. http://www.recovery.gov/Transparency/RecipientReportedData/ Pages/JobSummary.aspx. Scovel, C. L., III. 2011. Statement of the Honorable Calvin L. Scovel III, Inspector Gen- eral, U.S. Department of Transportation, Before the Committee on Transportation and Infrastructure, United States House of Representatives: Ensuring ARRA Funds Are Spent Appropriately to Maximize Program Goals. May 4. TRB. 1987. Special Report 214: Designing Safer Roads: Practices for Resurfacing, Restora- tion, and Rehabilitation. National Research Council, Washington, D.C. TRB. 2006. Special Report 285: The Fuel Tax and Alternatives for Transportation Funding. National Academies, Washington, D.C. U.S. Census Bureau. 2013. Annual Survey of State Government Finances. Revised Aug. 22. http://www.census.gov/govs/state/. U.S. DOT. 2010a. American Recovery and Reinvestment Act of 2009: Agency and Program Plans Update Second Quarter 2010. U.S. DOT. 2010b. Recovery Act §1201 Maintenance of Effort Provision: Preliminary Report on Implementation and Results. Nov. U.S. DOT. 2011. Recovery Act §1201 Maintenance of Effort (MOE) Provision: Preliminary Report on Implementation and Results. Jan. 27. Vernez, G., and R. Vaughan. 1978. Assessment of Countercyclical Public Works and Public Service Employment Programs. Rand Corporation, Sept.