President Obama's Labor Day infrastructure initiative combined some appealing business tax proposals with a national infrastructure bank. Some NIB proponents estimate the total need for major investments in infrastructure to be as high as $2.2 trillion and the timing is intended to "create jobs" in construction. But why another government-sponsored enterprise?

The proposed NIB is the same as a national economic development bank, which is generally used as a credit allocation mechanism to implement a state industrial policy. Such institutions are first cousins to national housing banks, but with different interests.

We should expect similar results. Development banks direct credit in politically correct ways, crowding out traditional banks and bond issuers by using their central government backing to remove accountability and distort markets. The NIB is also a good candidate to make politically correct loans to borrowers who can't repay, causing the next systemic financial crisis.

Infrastructure is primarily a state and local government responsibility and arguably should be the primary one. The federal government's role is mostly in constructing interstate projects and providing fiscal transfers. State and local government borrowing is authorized and appropriate for such capital projects. But what specific funding problem of insufficient infrastructure investment would a new GSE address?

Not the ability of state and local governments to borrow! State and local public debt has almost doubled since 1980 from 12 percent of GDP to 22 percent currently.

Not the ability of the federal government to borrow! Federal public debt outstanding as a percent of GDP more than doubled since 1970 to over 60 percent, financing a fivefold increase in federal grants to state and local governments from 1977 to 2007.

Not a shrinkage of government generally! Government spending at all three levels has grown from a sixth to a third of GDP since 1950, with taxes generally keeping pace.

The failure of governments to meet their most basic responsibility to provide public infrastructure doesn't reflect their lack of funding or access to finance, but rather a shift in priorities directed primarily to public-sector employees. Political pressure rather than relative productivity improvements have approximately doubled public-sector per-worker wage and benefit costs since 1950 relative to the private sector, while simultaneously approximately doubling the percentage of the labor force in public-sector employment.

Rather than control labor costs and face the wrath of public-sector employees and unions, many state and local governments have found numerous ways around legal borrowing limits. But borrowing governments exhaust taxing capacity as a result of raising white flags to labor, municipal bond insurers, credit rating agencies and investors raise red flags. That is already happening at the state and local level. It hasn't yet happened at the federal level as the Chinese haven't found an acceptable alternative to U.S. Treasuries. But they are beginning to seriously question the size of the federal deficits while looking for alternatives.

This is where the NIB comes in, essentially providing federal fiscal stimulus transfers but with off-budget federal debt. Proponents argue that the NIB will include charter limitations to safeguard that funds get directed to infrastructure projects with the highest (public) returns, but that's never worked.

Federal government policy entrenches rather than bypasses the labor cost problem. The deficit-financed $867 billion stimulus was pitched at the time as targeted to "shovel-ready" infrastructure projects, but it was not designed in the spirit of FDR's public works projects to get the "most jobs for the buck." Davis-Bacon Act hiring requirements and project labor agreements, maintaining the artificially high union wage rates for private-sector employees to spend the "most bucks for the job."

Optimistically assuming the "public risk for private profit" motivation can be permanently avoided, the proposed NIB may be a closer first cousin to the International Bank for Reconstruction and Development, which most people know as the World Bank. Its charter is similarly limited to lending to governments to finance infrastructure projects that are supposed to produce economic benefits that will be indirectly reflected in the ability of government borrowers to repay. In practice however, there is typically little or no link between a project's forecast economic benefits and loan repayment, and the loans have become increasingly politically correct rather than economically viable.

Creating another GSE much like the GSEs that, with political sponsorship caused the worst systemic financial system crash since the Depression, isn't the best long-run policy either.

Subscribe Now

Access to authoritative analysis and perspective and our data-driven report series.

14-Day Free Trial

No credit card required. Complete access to articles, breaking news and industry data.