The House Subcommittee on Highways and Transit this week held a contentious hearing to examine the Federal Transit Administration’s Capital Investment Grant (CIG) program, which provides federal funding for new subway, light rail, and other fixed guideway transit systems (traditionally called “new starts”) as well as projects to increase the carrying capacity of existing systems.

Acting FTA Administrator Jane Williams testified for two hours (see her prepared testimony here) and was then followed by a three-person panel: American Public Transportation Association chief Paul Skoutelas (testimony here), American Road and Transportation Builders Association representative Bob Lane (testimony here), and Kansas City Streetcar Authority chief Tom Gerend (testimony here). Video of the hearing can be watched here.

Public transportation advocates, and Democrats on Capitol Hill, have been sharply critical of the Trump Administration’s handling of the CIG program for over two years now, to the extent that it might seem surprising that new House Transportation and Infrastructure chairman Peter DeFazio (D-OR) waited this long before holding a CIG hearing. But the delay can be explained two ways: DeFazio’s staff were conducting a national survey of transit agencies to get feedback on how the program was being administered (see his staff’s report here), and they also may have wanted to wait until APTA had its formal recommendations for improving the program ready (found in the Appendix to the APTA testimony), and those were not official until APTA’s Legislative Committee met on June 23, 2019.

Instead of relaying the usual back-and-forth, we decided to examine the issues raised by the hearing individually, rather than when they were raised and re-raised by individual members.

Issue: Delays in signing grant agreements. Chairman DeFazio’s staff produced a chart (under Finding 1, here) showing that the timeframe from a proposed project being approved for the Engineering phase of the CIG program until the time it got its multi-year full funding grant agreement (FFGA) signed rose from 172 days in the last three years of the Obama Administration (projects that were “in the pipeline” as of the December 2015 enactment of the FAST Act, which changed some of the CIG program rules) to 391 days under President Trump. The chart also showed the amount of time that a one-off small start project took to get its grant agreement signed rising from 112 days under President Obama to 243 days under President Trump.

Administrator Williams responded that there is a big difference between the speed at which a brand new Administration functions in its first two years in office and the speed at which a second-term incumbent Administration functions in its last few years of office. This certainly seems to be the case. For an apples-to-apples comparison, let’s look at how many CIG projects were actually approved under the first two years of President Obama versus the first two years of President Trump.

• Q: How many multi-year full funding grant agreements did the Obama Administration sign in its first two years of office (2009-2010)?
• A. One (Dulles Silver Line, March 2009). And the answer was almost zero, because the Dulles project was originally on schedule to be signed in late 2008 before a late-term re-rating by FTA in early 2008 forced 11 months of re-negotiation with Virginia and others. Total CIG program cost: $900 million.
• Q: How many multi-year full funding grant agreements did the Trump Administration sign in its first two years of office (2017-2018)?
• A: Four (Caltrain electrification in May 2017, Maryland Purple Line in August 2017, Santa Ana streetcar in November 2018, and Seattle Lynnwood Link in December 2018). And this was very nearly two – the Maryland Purple Line FFGA was supposed to be signed under the Obama Administration in late summer 2016 (a public ceremony was days away) when a federal judge issued an injunction preventing the signing, which was not lifted until the following year, and Caltrain had a last-minute hiccup with one of its local governments postponing a formal meeting to approve its share of the budget, which forced the Obama Administration to postpone signing past Trump’s Inauguration. Total CIG program cost: $2.87 billion.

A better case can be made that the Trump Administration has been slow in approving one-off “small starts” that can be funded using a single year’s appropriation. The Obama Administration approved nine of those in its first two years in office (total CIG program cost: $178 million). But the Trump Administration also approved nine small starts in its first two years in office, albeit many more in 2018 than in 2017 (total CIG program cost: $454 million). However, there were more small starts in the pipeline anticipating funding in 2017-2018 than there were in 2009-2010.

The program delays weren’t just because of President Trump’s first budget (FY 2018), which proposed to kill the CIG program after paying off the currently signed FFGAs. The problem also lay in President Obama’s last budget (FY 2017) and the false expectations it set for the CIG program and projects in the “pipeline.”

In December 2015, President Obama signed the FAST Act into law. That five-year highway and transit authorization bill authorized the Appropriations Committees to appropriate $2.302 billion per year, out of general revenues, for the CIG program. Two months later, President Obama’s FY 2017 budget request for CIG (which was already in the works before the FAST Act was signed) asked Congress to just ignore the FAST Act total and instead appropriate $3.500 billion for the program, a 52 percent increase over the FAST authorized total and a 61 percent increase over the FY 2016 enacted total of $2.177 billion.

$1.4 billion of that request was for proposed new multi-year FFGAs for projects in Los Angeles, San Diego, Seattle, Fort Worth, Chicago, and other cities. It took Congress over a year to finish the FY 2017 appropriations process (the bill was not enacted into law until May 2017, which left less than five months left in the fiscal year), and the FY 2017 appropriations bill only provided $2.531 billion for the program (this was still more than the FAST Act had authorized). Instead of $1.4 billion for new FFGAs, Congress only appropriated $695 million – $714 million less than the request.

That $714 million difference between the expectations raised by the Obama FY17 budget for projects “in the pipeline” versus the actual FY17 appropriation was, of course, stomped on with both feet by President Trump’s FY 2018 budget request, which proposed killing the CIG program after paying off all prior commitments. Congress eventually ignored that, as well, instead appropriating $2.6 billion – but not until March 2018.

It is the combination of these two politically unrealistic budget requests – Obama’s unrealistically optimistic 2017 request (and the false expectations it engendered from transit agencies), and Trump’s completely unrealistic 2018 request to zero out the program – that have resulted in the current delays. The projects that were to get funded by that $714 million in FY17 money proposed by Obama – which Congress refused to provide – got shifted to FY18, which meant that projects farther back in the pipeline that were wanting FY18 money to get started were going to have to wait until FY19, etc. etc.

Since that FY18 appropriations law was enacted in March 2018, FTA has done a much better job in signing new grant agreements (the Seattle Lynnwood Link, Santa Ana Streetcar, and Dallas DART projects that Congress refused to fund in Obama’s FY17 request have now been signed, mostly using FY18 funds, and the only proposed FFGA from Obama’s FY2017 budget that has not been signed by this point is NYC Canarsie Line, which was sent back to the drawing board by Governor Cuomo six months ago when he decided to have college engineering professors instead of transportation consultants design the work).

But, between those two antipodal budget requests and the general slowdown that happens in the first year of every new President’s term when they are trying to get a new team in place and get them up to speed, it’s hard to deny that 2017 was largely a lost year for the CIG program. 2018 and 2019, however, have proceeded much more normally.

To an extent, this problem is caused by the number and size of projects in the CIG “pipeline” being unconstrained, but the size of the annual CIG appropriation being limited each year by the funding available to the Appropriations Committees. The APTA testimony says that “communities are requesting approximately $27 billion of Capital Investment Grants to fund these projects in the pipeline” – and this is in addition to $4.6 billion in future (post-FY19) appropriations needed to pay off the projects with ongoing multi-year FFGAs. Yet the CIG program is only authorized to receive $2.3 billion per year in appropriations, and actual appropriations are only averaging $2.5-$2.6 billion per year (i.e. more than the House Transportation and Infrastructure Committee and the Senate Banking Committee authorized).

It was the authorizing committees in 2005 who decided to throw the CIG program out of the Highway Trust Fund, deny it future access to multi-year guaranteed contract authority, and leave it up to the annual mercies of the appropriations process. And it was the authorizing committees in 2012 who decided to repeal the traditional limit on “contingent commitment” authority that limited the aggregate amount of grant agreements that could be signed at any one time, leading to unlimited expectations for transit agencies of their ability to get a grant on a timeframe they wanted. To the extent these problems were caused by the authorizing committees, they can be addressed by the authorizing committees in future legislation, perhaps by giving the program multi-year funding certainty once again as part of the Highway Trust Fund.

Issue: Delays in formally allocating appropriated funding. The APTA testimony says “For instance, in FY 2019, Congress provided $2.6 billion for CIG investments. To date, more than one-half ($1.3 billion) of these funds remain unallocated (i.e., FTA has not assigned the funds to a specific project). In fact, FTA has not completed allocating its FY 2018 funds—$41 million remains unallocated from last year.” (See the updated FY19 allocations here). Williams did not really address this issue, but the allocation process used to be fairly simple – the Appropriations Committees would usually put a table in the final conference report on the annual DOT appropriations bill with projects and dollar amounts, and that table would then become the formal FTA allocation table in very short order.

Even after Congress gave up on earmarking in 2011, the appropriators usually included a table that denied FTA any discretion in allocating projects for full funding grant agreements (there was sometimes some leeway given for small starts). But under the Trump Administration, while the 2017 appropriations law (based on the Obama request table) did include a project table, the 2018 and 2019 appropriations laws did not, which left the allocation process entirely up to FTA for the first time in years. So, going from having zero discretion in 2017 and prior years to having more or less complete discretion in 2018 and 2019 are two very different things.

Meanwhile, Transportation for America has been waging a Twitter war against FTA every time that FTA announces it has allocated money to a project (example here), alleging that the whole allocation process is completely meaningless. (Ed. Note: If the act of FTA formally allocating funding to a project were really as meaningless as T4A says it is, then why is APTA complaining in its testimony that FTA is too slow in formally allocating FY 2019 funding? If the allocations were truly meaningless, then APTA and its member agencies wouldn’t care.)

Issue: Delays in issuing Letters of No Prejudice. The DeFazio staff memo states that the average amount of time that it takes from the time a transit agency applies for a “Letter of No Prejudice” (LONP) for a pending project has risen from 54 days under late-term Obama to 78 days under Trump (see Finding 4 here). However, this statement (while doubtless true) calls to mind an unanswered question: should LONPs exist at all, so long as the CIG program is funded solely by annual appropriations?

A LONP is basically a document from FTA saying to a transit agency that has a project that is nearing its FFGA, “you guys can feel free to proceed at your own risk and purchase right-of-way or materials for this project using your own money, and if you then get a FFGA, which is not guaranteed, you will be able to be repaid out of the funding provided by the FFGA, if Congress actually appropriates that funding, which is also not guaranteed.” This kind of pre-award authority is routine in grant programs that have guaranteed multi-year contract authority provided by law years in advance.

But CIG is not one of those programs, and legally, FTA can’t promise that a FFGA will be signed before all the legal restrictions of title 49 are met, and legally, a FFGA can’t promise that Congress will ever appropriate any money to provide its share of the funding of the FFGA (that’s why the appropriations are called discretionary, because Congress has complete discretion to zero out the program every year without exposing the federal government to legal liability).

A LONP is an extralegal promise that a hypothetical future grant agreement might allow you to be repaid out of future appropriations that are, legally speaking, even more hypothetical. If the ability to spend pre-award federal funding to gain efficiencies and prevent delay-related cost increases is essential for CIG projects, that seems to be as much of an argument to put the CIG program back into the Trust Fund and give it real guaranteed multi-year funding, not provide extralegal documents promising hypothetical access to appropriations which can never be guaranteed.

Issue: Delays in moving projects from Project Development to Engineering. The DeFazio staff memo also noted that the amount of time it takes to move a project from the Project Development phase to the Engineering phase has gone from 135 days in the last few years of Obama to 289 days under Trump. It would be interesting to know how much of this was due to 2017 being basically a lost year for the program versus delays in 2018-2019. But the PD to E timeline is particularly important because, by law, if a project does not proceed from PD to E within 2 years, it is supposed to be rejected (though there are provisions for extensions).

The DeFazio staff memo (in Finding 2, here) says that transit agencies have reported to the T&I Committee a total of $160 million in cost increases for projects in the pipeline due to delays in moving from PD to E or from delays in moving from Engineering to a FFGA.

One of the APTA recommendations is to amend title 49 and extend the amount of time that a project can be in Project Development before progressing to Engineering from two years to 3 years.

The DeFazio staff memo alleged that transit agencies have seen $195 million additional project costs since 2017 due to delays in the approval process. In response to a question from Rep. Alan Lowenthal (D-CA), Williams said that FTA does not keep specific track of delay-related cost increases.

Issue: Decreased CIG federal cost share. The DeFazio staff memo (in Finding 3, here) noted that the average CIG program cost share of new projects has decreased from 47.9 percent before 2017 to 36.6 percent under President Trump. (The statutory maximum CIG share of project capital cost is 60 percent for a new start FFGA, though the appropriators usually override that with a 50 percent maximum.) The APTA testimony states that “many New Start project sponsors believe that FTA is strongly encouraging significant “local overmatch” of the federal CIG share. Despite current law restrictions, these project sponsors believe that DOT will not move forward with their New Start projects unless the project sponsor requests significantly less than a 50 percent CIG share.” DeFazio and others said during the hearing that transit agencies had reported to them that, while FTA officials never actually said they were enforcing a 40 percent maximum CIG share, they would also note that nothing with a federal share over 40 percent had been approved by the Trump Administration.

Williams was able to respond that FTA had just moved the next BART project into the Engineering phase at a 43 percent CIG share, but after direct badgering from DeFazio, went on record as saying that yes, they are willing to look at projects with a CIG share over 40 percent.

But, generally speaking, this criticism rings true. For starters, if you were to ask OMB Director (Acting White House Chief of Staff) Mulvaney and the Heritage Foundation crowd about the federal CIG share of projects suddenly dropping eleven points, they would probably respond with the old computer programming adage, “that’s not a bug in the system, that’s a feature of the system.”

As to the relative merits of a higher versus lower federal CIG cost share, during the questioning on this topic, Rep. Rob Woodall (R-GA) tossed Williams one of the fattest, slowest-pitched softball questions that you will ever see at a Congressional hearing (he tweeted a link to a video of the exchange here). Woodall asked Williams if the savings from a lower CIG share of a project go back to federal taxpayers, and she responded no, that money stays in the CIG budget account and allows FTA to fund a greater number of projects. He then followed up by saying that he didn’t want to see the big giant mega-projects suck up the entire CIG program and that he appreciated FTA spreading money around for a greater number of projects.

In its testimony, APTA formally recommended that Congress set fixed, statutory shares for CIG programs be established: “i. New Starts: 60 percent or, for New Start projects with significant total project costs, a lesser percentage; ii. Core Capacity: 80 percent or, for Core Capacity projects with significant total project costs, a lesser percentage; and iii. Small Starts: 80 percent.”

Two things about that APTA recommendation: (i) it’s all about the definition of “significant,” which is not defined in their proposal, and (ii) this would make the CIG program more of a “one size fits all” program, which would get rid of a key selling point of CIG, which is that federal grants can be tailored to the unique financial needs of each locality.

Issue: Cost increases from moving back to a 65% probability in risk assessment instead of a 50% probability. The DeFazio staff memo (in Finding 2, here) alleges that transit agencies have seen a total of $650 million in project cost increases due to a change by FTA last year in the risk assessment process. This bears some explanation.

In 2007, FTA started using an internal risk assessment process for proposed CIG projects that, among other things, required the FTA career staff to agree that a project had at least a 65 percent likelihood of coming in at budget and on time (since delays and cost increases are inextricably linked). Local sponsors could either set aside enough contingency in cash in the cost estimate to get to a 65% likelihood, or else identify secondary mitigations they could take if there were cost overruns to achieve “P65.” Secondary mitigations were accepted to cover some or all of the cost estimate from P65 to as low as P50. FTA found however, that recipients were reluctant to implement secondary mitigations.

The risk assessment process worked – according to a DOT spokesperson, Comparing 13 completed projects that did not use the 2007 risk assessment tool with 28 completed projects that did use the tool, the percentage of projects completed within budget increased from 62 to 89 percent and the percent completed within schedule increased from 69 to 79 percent.

In May 2016, under the Obama Administration, FTA decided to go from a 65 percent certainty level down to a 50 percent certainty level (the coin toss level), and drop the secondary mitigation requirement. FTA did not make any public announcement or give any public justification for this change.

Then, in June 2018, the Trump Administration decided to go back to the old P65 threshold – the spokesperson said that many recent bids for projects in Florida, Los Angeles, Sacramento, Seattle, etc. were significantly higher than the engineer’s estimate, so they went back to P65, and a month later made a public announcement of the new policy.

There are two separate issues at play here. The first is which numerical threshold – the two-thirds likelihood of hitting the cost estimate in effect from 2007-2016, or the 50-5o coin flip in effect from 2016-2018 – provides better protections for taxpayers both federal and local? And the second is at which stage of the process – Project Development, or Engineering – should the risk assessment be applied?

With regards to the first issue, the APTA testimony says “APTA is concerned that increasing the Probability Threshold percentage will require project sponsors to identify more contingency funds, adding to the costs for project sponsors.” A DOT spokesperson countered that the 65 percent probability makes it less likely that a project will have its financing fall apart so badly during construction that the project stalls, as has happened in Honolulu (even though Honolulu was originally judged on the P65 standard).

At the hearing, Williams said that P65 gives a better chance of cost estimates being correct, and then raised the specter of the East Side Access Project in NYC as the alternative (that project is now 10 years behind schedule and $5 billion over budget).

Regarding the second issue, APTA recommends that the risk assessment probability test be performed no earlier than 180 days after the project enters the Engineering phase, unless the project sponsor requests it take place earlier. The APTA testimony says that “requiring the Risk Assessment during the Project Development phase provides an additional hurdle to completing Project Development within the two-year time period. Given the significant number of tasks already required to be completed during the two- year period, this change is likely to require CIG project sponsors to conduct and fund even more preliminary work before seeking entry into Project Development. The sponsor’s funding of this preliminary work is not included in calculating the CIG share.”

A DOT spokesperson countered that the earlier that FTA gets risk assessment information, the better for all concerned.

(Ed. Note: Another way to look at this is, to what extent is it FTA’s responsibility to protect the taxpayers in places like Honolulu from their own overly optimistic local politicians and transit officials, who may try to push through projects without adequate safeguards to protect local finances? Yes, FTA’s primary concern should be to protect the federal taxpayer, but how far should they also go to protect the locals from themselves?)

(Further Ed. Note: The U.S. Government Accountability Office is expected to release an analysis as soon as next Monday that examines current FTA procedures and guidelines for estimating project costs and, possibly, recommending improvements in those procedures and guidelines.)

Issue: Defining the non-federal share. Both APTA, and several Democrats on the T&I Committee, took issue with the Trump Administration’s decision to exclude the face value of federal TIFIA loans, if being repaid from non-federal sources, from being considered part of the non-federal share of a CIG project cost. The TIFIA statute does give the Transportation Secretary discretion in this regard – the statute says that the Secretary “may” count TIFIA loan face value as non-federal, not “shall,” which is the word Congress usually uses when writing a law when they want to force someone to do something. But APTA does point out that a separate provision of law directs TIFIA loans to be treated as non-federal for purposes of highway project funding, so the current Trump Administration approach treats highway and transit projects differently.

(Ed. Note: The APTA recommendations did not say that RRIF railroad loans should be treated as part of the non-federal share of a project – and it is RRIF loans, not TIFIA loans, that New York and New Jersey propose to use to fund most of the non-federal share of the $13.7 billion Hudson River Tunnel under the CIG program (see their latest proposed financial plan here). So, unless APTA meant to say “TIFIA and RRIF” and neglected to mention RRIF, their proposed plan won’t help the tunnel sponsors get their current “low” local financial commitment rating increased.)

Issue: transparency. The DeFazio memo and the APTA recommendations do include one major common-sense transparency recommendation that everyone should be able to endorse: the creation of a one-stop “CIG Program Pipeline Dashboard” website that would, under the APTA recommendation, include:

At present, this information is either fragmented and hard to find (spread across multiple DOT and non-federal websites) or else not public at all. The result is confusion, even between project sponsors and FTA on their own projects. See Finding 5 in the DeFazio staff memo (here), which says “A comparison of the data FTA submitted and the data transit agencies submitted revealed large disparities in terms of timelines in the Project Development phase. The dates provided by FTA and transit agencies matched as little as 39 percent of the time.”

Gateway. Rep. Tom Malinowski (D-NJ) asked about the Portal North Bridge, and he and Williams had a bit of a disagreement on how much more the New Jersey local authorities had to do before their funding for the project could officially be considered to be “committed” by FTA. They may have split the difference – the New Jersey Transit Board of Directors met on the day after the hearing to adopt budgets, which may have covered this. This will make it possible for NJT to apply to FTA for a re-rating of the project, which could happen this fall.

Malinowski also asked about the status of the Environmental Impact Statement (EIS) for the Hudson River Tunnel, which he noted was completed by local authorities in record time (14 months) and then submitted to USDOT, where it has been sitting since June 2017. Williams deferred to the Federal Railroad Administration, which is the lead agency on the tunnel’s EIS, not FTA (which begs the question of why project sponsors are trying to get $6.7 billion from FTA, if it’s really a FRA project).

For more information on how the CIG program has been funded since its inception in the 1960s, and why the current system may not be the best way to fund the program, see this August 2017 ETW article.