Why An Infrastructure Bank Isn’t Self-Sustaining

Last week Senators John Kerry (D-Mass.) and Kay Bailey Hutchison (R-Texas) proposed legislation, the Building and Upgrading Infrastructure for Long-Term Development Act (BUILD Act), that would create a national infrastructure bank. The proposal is essentially a new federal program that would issue direct loans and loan guarantees to finance public works projects.

The overarching goal of the proposal is two-fold: it aims to boost infrastructure spending and wrest decisions about the allocation of spending from the congressional appropriations process. These decisions would instead be made by an independent development bank – the American Infrastructure Financing Authority (AIFA) – that would be backed partially by private money. This bank, the theory goes, will make better decisions than pork-barreling congressmen, especially since private investors have their money on the line.

The sponsors of the BUILD Act also argue that the program would not impose any long-term costs on taxpayers because the federal loans and loan guarantees that it issues must be “self-sustaining” arrangements. Projects financed with the loans must pay the federal government a fee that fully covers the subsidy that the loans provide. Put another way, a project that secures a lower interest rate because it obtained a government-backed loan instead of private financing must pay for the full cost of this benefit.

Any rational person would ask: why go through the trouble of setting up a development bank that requires projects to pay the federal government for the “full” subsidy that they get? After all, states and municipalities can borrow in the well-established municipal bond market to finance infrastructure projects, and these bonds are already subsidized by federal taxpayers since the interest on these obligations isn’t subject to federal income tax. That’s a subsidy that bond issuers get to keep, whereas the BUILD Act would require states and municipalities to pay the government for any subsidy they receive. So both financing methods provide lower interest rates for infrastructure projects, but the BUILD Act makes borrowers pay for it so federal taxpayers are presumably kept whole.

BUILD Act supporters will tell you that even after project sponsors pay for the subsidy cost on their federal loans, the interest rates and other terms on the loans will still be better than they could get otherwise. Can that even be possible? No, there really is no free lunch. But there is a loophole in federal budgeting rules that makes it look like there is.

The Federal Credit Reform Act, which details how the federal government must calculate the costs and risks that federal loan programs impose on taxpayers (including loans made under the BUILD Act), systematically excludes a full measure of the riskiness of these loans. That is, by discounting expected loan payments at risk-free U.S. Treasury interest rates, the rules ignore the fact that loan performance is unpredictable over time and that defaults will be more frequent and costly in bad economic climates. Private lenders charge borrowers a premium to take on that kind of risk and uncertainty, called “market risk”, but government budget rules do not.

Nothing in the BUILD Act would reduce the market risk inherent in financing a public works project. What the program does do, however, is shift market risk from one group (would-be municipal bond investors) to another (federal taxpayers). This appears to produce cost savings because the federal government doesn’t charge for the market risk taxpayers take on when they finance projects through loans and guarantees.

Thus, the BUILD Act would charge infrastructure projects a subsidy cost for a federally-backed loan that is less than what the subsidy is truly worth in the private market – and less than what taxpayers would charge to bear the credit risk. The budget rules virtually guarantee that this would happen and it’s how the BUILD Act appears to provide subsidies at no cost to anyone. In short, the program isn’t self-sustaining because taxpayers are bearing a risk for which they have not been compensated.

To make the program truly self-sustaining, lawmakers should require that the federal government purchase full reinsurance in the private market that would cover all default losses on loans it makes and guarantees under the BUILD Act. This shouldn’t change the program’s costs at all if the BUILD Act really does charge project sponsors for the full subsidy cost of the loans they receive.

BUILD Act supporters would probably say that buying reinsurance would defeat the whole purpose of the program because the financing advantage it provides to infrastructure projects would disappear. On the contrary, it would show that the financing advantage was never there.

Jason Delisle is the Director of the Federal Education Budget Project at the New America Foundation.