A History of Infrastructure Bank Proposals

August 3, 2016

Investing in national transportation infrastructure became a significant talking point for both Hillary Clinton and Donald Trump as the candidates have proposed improvement programs that would respectively cost $275 billion and $550 billion (Trump promised to double his opponent’s plan).

Although Trump has yet to release details for his plan, Hillary Clinton’s involves enacting a five-year, $275 billion injection of infrastructure spending within her first 100 days in office. According to Clinton’s campaign, business tax reform would fully pay for these investments, $250 billion of which would be direct public investments.

There are no details yet of how that $250 billion would be spent, but the real centerpiece of the plan is the remaining $25 billion, which would endow a national infrastructure bank “dedicated to advancing our competitive advantage for the 21st century economy.” Her campaign estimated that these funds would support up to an additional $225 billion in direct loans, loan guarantees, and other forms of credit enhancement.

Implementing a national transportation infrastructure bank (I-Bank) is no small feat. This article looks at the history of the infrastructure bank idea and part two will examine the specifics of past efforts and potential problems and tradeoffs facing future attempts to create an I-Bank.

A Revolving Fund. Most early proposals for government infrastructure “banks” are not “banks” per se. Rather, they are a separate government fund that is classified as a “revolving fund.” The Government Accountability Office (GAO) describes a revolving fund as a fund that “authorizes an agency to retain receipts and deposit them into the fund to finance the fund’s operations. The concept of a revolving fund is to permit the financing of some entity or activity on what is regarded as a more ‘business-like’ basis.”

In the field of infrastructure, revolving funds first came to prominence in 1987, with the enactment of the Water Quality Act of 1987. Recognizing that sewer and drinking water infrastructure projects have a long lifespan, and that the need for such projects is perennial, Congress moved away from giving grants to localities for such projects in the 1987 law. The grants were replaced by a new program where, in the words of a Congressional Research Service report, “federal grants would be provided as seed money for state-administered loans to build sewage treatment plants and, eventually, other water quality projects. Cities, in turn, would repay loans to the state, enabling a phase out of federal involvement while the state builds up a source of capital for future investments.”

Since the law took effect in 1989, Congress has been appropriating money to the Environmental Protection Agency (EPA) – in turn, the money is distributed to each state for the endowment of state revolving funds (SRFs) that then make safe drinking water and wastewater infrastructure loans. Once the first round of loans are eventually repaid, the proceeds from the repayments are used to make new loans, over and over, forever and ever, amen.

Since the loans and repayments from and to state revolving funds are made at the state level, the federal seed money is classified as a regular grant in the federal budget, not a federal revolving fund.

State Infrastructure Banks. The first occurrence of the phrase “infrastructure bank” in federal legislation was all the way back in February 1983. Sen. Pete Domenici (R-NM) introduced the bill S. 532 of the 98th Congress, the “Public Investment Incentive Act of 1983.”

The bill would have allowed the U.S. Treasury to endow state infrastructure banks (SIBs) and would have created a National Infrastructure Council to allocate the money between states. The SIBs would be allowed to issue bonds to fund “public investment projects” which were defined as “the physical structures and facilities developed or acquired by public agencies to provide water, waste disposal, resource recovery, transportation, and similar services to facilitate the achievement of common social, environmental, and economic objectives, and are capable of producing user income to service debt or loans issued to finance the project.”

Domenici’s bill was referred to the Environment and Public Works Committee, which held hearings on the legislation but took no action. The following year, a similar bill (S. 1619, 98th Congress) was introduced by Sen. Lloyd Bentsen (D-TX) but likewise saw no action. The 1987 action to create water SRFs, that built throughout 1985-1986 and then was enacted into law early in 1987, as noted above, superseded the Domenici proposal.

The phrase “infrastructure bank” then died out until 1995. In February 1995, President Clinton’s budget proposed a wholesale restructuring of federal surface transportation programs, combining the Highway and Airport and Airway Trust Funds into one big Transportation Trust Fund and streamlining most federal transportation programs into a few big block grants. The budget proposed $24.5 billion for the combined programs in fiscal 1996, of which $2 billion was to be used to establish state infrastructure banks to “enable jurisdictions to leverage more easily public and private resources for transportation infrastructure and encourage more business-like strategies for financing the national transportation system.”

The transportation committees in the House and Senate were in no hurry to reinvent the wheel (particularly in the Senate, where the President proposed to combine the highway, mass transit and aviation funding streams, which are under the purview of three different Senate committees), so the highway bill that the Senate passed in June 1995 (S. 440, 104th Congress) made no reference to infrastructure banks. But when the Senate Appropriations Committee was rewriting the House-passed fiscal 1996 transportation appropriations bill (H.R. 2002, 104th Congress) that same summer, the panel added a new section 349 to

(Interestingly, the first bill introduced in Congress in 1995 relating to SIBs came in September 1995 from Rep. Bill McCollum (R-FL) – neither a transportation authorizer nor an appropriator. McCollum said his bill would encourage “innovative financing partnerships between the public and private sectors.” Many 1995 supporters of the SIB concept were Republicans from right-to-work states, because they saw in SIBs a way to use federal money (eventually) without being subject to onerous restrictions attached to federal grants like compliance with Davis-Bacon prevailing wage laws.)

In the end, the Appropriations Committees dropped their provision from the FY 1996 bill and the NHS authorization bill gained a new provision that appeared in the conference report in November as if from nowhere. Section 350 of Public Law 104-59 allowed up to ten states to establish SIBs endowed out of federal-aid highway funds. The 10 states were all selected by June 1996 (Arizona, California, Florida, Missouri, Ohio, Oklahoma, Oregon, South Carolina, Texas, and Virginia).

The following month, the Senate Appropriations Committee included in the FY 1997 transportation appropriations bill a $250 million appropriation from Highway Trust Fund to the Federal Highway Administration to carry out the SIB pilot program to “more than 10 States”. In the House-Senate conference committee, this was reduced to $150 million and drawn from the general fund of the Treasury instead of the Trust Fund, and the $150 million became law in Public Law 104-205. By July 1997, 29 more states had been added to the SIB pilot program, and all 39 states shared in the $150 million one-time SIB general fund appropriation.

Also, in February 1998, Rep. Rosa DeLauro (D-CT) introduced a bill (H.R. 710, 105th Congress) directing DOT to study ways to expand SIBs so that they could fund non-transportation infrastructure. Nothing came of that bill, but remember the name “DeLauro” for later.

Section 1511 of the 1998 TEA21 law allowed a new round of SIB capitalizations, this time limited to just four states named in the law (California, Florida, Missouri and Rhode Island). And then section 1602 of the 2005 SAFETEA-LU law created a third SIB program, this time permanently authorized in 23 U.S.C. §610. But the more recent programs have had diminishing returns of interest by states.

Section 2001(i) of Division A of the FAST Act makes further changes in the SIB program and reauthorizes it through FY 2020.

A presentation from a recent online course run by the Federal Highway Administration explaining the SIB program as of 2016 can be viewed here. Repeated changes in law have made it clear that SIB money now has Davis-Bacon rules attached to it forever, no matter how many times the original federal seed money has been loaned out and then repaid with interest.

All of the federal money used to capitalize state infrastructure banks took the form of federal grants to states, with no need for the states to ever repay the federal government. But just before this happened, something fundamentally transformative took place regarding loans and loan guarantees made by the federal government itself, which would eventually lead to calls for more federal infrastructure loans.

Credit Reform. The 1990 budget agreement produced some groundbreaking changes in the way the federal government spends money. High among them was the Credit Reform Act, which took effect for the fiscal year 1992 budget cycle but took several more years for all of its profound implications to be realized fully by policymakers.

Prior to enactment of credit reform, if the federal government wanted to loan a non-federal entity $10 million, then the face value of that $10 million loan was shown in the budget as an outlay, just like a federal grant of $10 million. And repayments of both interest and loan principal looked just like tax receipts in the budget.

The situation before the Credit Reform Act had two big effects on public policy. First, it tended to discourage the use of federal credit programs in areas where the federal government also made grants, like in transportation – if your legislation has to take the same budget “hit” whether or not the money is ever paid back, why not make a grant instead of a loan? And second, since loans going out were on a completely separate ledger from repayments coming in, it was difficult to distinguish federal programs that made good loans from those that made bad loans.

Under credit reform, the face values of federal direct loans and loan guarantees are no longer part of the budget at all, and neither are interest and principal repayments. Instead, the only part of the loan that shows up in the federal budget is the “subsidy cost,” defined for direct loans as “the estimated loan disbursements, repayments of principal, payments of interest, recoveries or proceeds of asset sales, and other payments by or to the government over the life of the loan. These estimated cash flows include the effects of estimated defaults, prepayments, fees, penalties, and expected actions by the government and the borrower within the terms of the loan contract.”

Subsidy costs for each federal credit program are re-estimated each year by the Office of Management and Budget and published in the annual Federal Credit Supplement.

Before credit reform, a $10 million grant looked just like a $10 million federal loan, and they both cost $10 million against the budget allocation of the Appropriations Committee, or the Transportation Committee, or whatever committee of Congress was providing the money to make the grant or loan. But after credit reform, a $10 million grant still cost Congress $10 million, but a $10 million direct federal loan only cost somewhere between $500 thousand and $2 million (most subsidy costs are between 5 and 20 percent of loan face value).

This budget scoring change explains why calls for a federal infrastructure bank did not really get going until after credit reform was implemented and the new budgetary advantages of credit over grants became fully understood.

Recent Federal Infrastructure Bank Proposals. The public discussion of infrastructure banks during the 1995-2006 period was limited to SIBs. (One interesting proposal in the 109th Congress sponsored by Sen. Hillary Clinton (D-NY) would have established a separate set of SIBs to build and reconstruct schools and libraries – see

The first proposal for a federal I-Bank came on August 1, 2007, when Senate Banking Committee chairman Chris Dodd (D-CT) and Sen. Chuck Hagel (R-NE) introduced the National Infrastructure Bank Act of 2007, which would have established a federal I-Bank to make loans for road and bridge, mass transit, water and housing infrastructure. Dodd’s committee held a hearing on the bill in March 2008, and at that hearing, one of the witnesses was legendary banker Felix Rohatyn, who had once helped create a special off-budget corporation chartered by the New York State government to help make loans to New York City in the 1970s.

Rohatyn co-wrote a widely publicized article entitled “A New Bank to Save Our Infrastructure” in October 2008 emphasizing the need for an I-Bank.

Meanwhile, two months after Dodd and Hagel introduced their bill, Rep. DeLauro and other House Democrats introduced the National Infrastructure Development Act of 2007. That bill would have created something similar to an I-Bank and used an expansive definition of “infrastructure”:

a road, highway, bridge, tunnel, airport, mass transportation vehicle or system, passenger or freight rail vehicle or system, intermodal transportation facility, waterway, commercial port, drinking or waste water treatment facility, solid waste disposal facility, pollution control system, hazardous waste facility, federally designated national information highway facility, school, and any ancillary facility which forms a part of any such facility or is reasonably related to such facility, whether owned, leased or operated by a public entity or a private entity or by a combination of such entities,

During the 2008 presidential campaign, Senator Barack Obama promised to “address the infrastructure challenge by creating a National Infrastructure Reinvestment Bank to expand and enhance, not supplant, existing federal transportation investments. This independent entity will be directed to invest in our nation’s most challenging transportation infrastructure needs. The Bank will receive an infusion of federal money, $60 billion over 10 years, to provide financing to transportation infrastructure projects across the nation.”

After his election and inauguration, President Obama submitted the broad outline of his fiscal 2010 budget and legislative program to Congress in February 2009. This document included a $5 billion per year placeholder request over the five-year 2010-2014 period for a “National Infrastructure Bank designed to deliver financial resources to priority infrastructure projects of significant national or regional economic benefit.”

A $5 billion per year appropriation for a brand-new program is quite a big “ask.” But amazingly, the House Appropriations Committee in mid-2009 under chairman David Obey (D-WI) found the money. See pages 4-5 of the June 16, 2009 Transportation Weekly for details, but the President requested $18.2 billion in discretionary appropriations for USDOT, $45.5 billion for HUD, $5.0 billion for the I-Bank and $296 million for other agencies in the Transportation-HUD bill, totaling $68.9 billion. And the House Appropriations Committee gave them $68.8 billion. The chairman gave the subcommittee enough money to fund the I-Bank request.

However, two weeks later, the White House sent emissaries to Capitol Hill to tell the appropriators that the full $5 billion for an I-Bank would not be necessary. Instead, the request for 2010 was downsized to $2 billion and focused more directly on transportation. (See the article starting at the bottom of page 1 of the July 8, 2009 Transportation Weekly for details.)

The Transportation-HUD Subcommittee had already been given an extra $5 billion for the I-Bank by chairman Obey and was not about to give the money back, so subcommittee chairman John Olver (D-MA) instead increased funding for high-speed rail to $4 billion (quadruple the President’s request) with a caveat that up to $2 billion of that money could be transferred to an I-Bank if an I-Bank was authorized in law prior to September 30, 2010. But as the appropriations bill was negotiated with the Senate, the I-Bank faded in priority, so the high-speed rail money dropped down to $2.5 billion and the rest of the money was spent on TIGER grants, highways, new starts and HUD.

Although the President kept on requesting funding for an I-Bank in future years, and eventually made some more detailed requests, they never came close to getting capitalization money again. A $4 billion I-Bank-ish proposal was made in the fiscal 2011 budget but was requested as authorization legislation, not an appropriation – at the same time the Administration refused to put forth a highway bill proposal during the 2009-2010 period because they wanted it to be part of their agenda in the latter half of the first term. Nothing came of the proposal, and the President did not mention an I-Bank again until a surprise announcement at a campaign rally on Labor Day weekend of 2010. The loss of Democratic control of Congress in the 2010 elections meant the end of that strategy.

And of the problem came from their ever-changing nature of the Administration’s request. The FY 2010 proposal was called a “bank” even though a substantial portion of its money would be spent making grants instead of loans or loan guarantees. The FY 2011 proposal had the honesty to drop the “bank” moniker and call itself a “National Infrastructure Innovation and Finance Fund” instead. The FY 2012 proposal was back to using the word “Bank” but still asked for the ability to make grants as well as loans. By the time of the FY 2013 proposal, the Administration had stopped asking for the ability to use the I-Bank to make grants, but by this point Congress had stopped listening.

(Ed. Note: One unfortunate victim of this period was Michael Likosky, a preeminent expert on public-private partnerships. Likosky wrote a book published in 2010, most of which was a very good discussion of the history and merits of “P3’s.” But because the book was finalized and sent to the publisher at the height of President Obama’s first-term achievements, Likoskiy entitled the book Obama’s Bank: Financing a Durable New Deal and prefaced and ended the book with discussion of the merits of an I-Bank and the implicit assumption that one would be enacted into law quickly and as part of the Obama legacy.)

Search Eno Transportation Weekly

Latest Issues

Happening on the Hill