Shaky Plan for Student Loans

AMERICA at its best." That is how President Clinton has characterized his national-service initiative. Who could argue with his vision of Americans helping others, rebuilding their communities and being rewarded? A bold, beneficent aspect of the president's national-service initiative is his proposal to provide students with improved financial access to an education beyond high school.

But some aspects of this sweeping change do not add up: in particular, his proposal, which passed in the House of Representatives as part of the budget package, to shift the management and implementation from private-sector lending and guarantor institutions to the government. The administration estimates that this would save billions of dollars.

This is optimism beyond the bounds of credibility: The revenue estimates rely on assumptions about the efficiency of government that do not stand up to scrutiny.

The Congressional Budget Office, which backed away from earlier projections of large savings from direct government lending (from $6 billion to $4.27 billion), has lowered its estimate again, saying that earlier estimates failed to factor in a significant portion of the costs of administering student loans: $2.19 billion. The CBO also did not take into account transition costs of moving to a direct-lending program and relied on overly optimistic projections for servicing costs and likely defaults.

The impartial Congressional Research Service (CRS) has come out with three reports, each noting that the cost-savings argument is unrealistic and warning of the negative consequences if the government does not run the program efficiently.

This is not the first time that rosy assumptions about direct government lending have wilted under close scrutiny. Last fall, the General Accounting Office (GAO) concluded that a complete government takeover of student loans would save $4.8 billion during the first five years of the program. But the GAO staff recognized the fragility of its assumptions: "The values we assume for certain key variables strongly influenced our estimate of the savings achievable from direct lending."

Unfortunately, the staff's cautions were ignored, and the GAO's optimism helped whip up support for a radical shift to direct government lending, leading to its inclusion in the president's national-service plan. Pulling apart the GAO's assumptions layer by layer clearly reveals the study's flaws, some of which were repeated by the CBO. These flaws include:

* Out-of-date economic assumptions. Updated assumptions about interest rates and inflation and corrected miscalculations of the savings that accrue from new provisions in Higher Education Act amendments combine to reduce purported savings by $1.1 billion.

* Understated estimates of the costs of servicing loans. The GAO based these on what the Student Loan Marketing Association charges. But high volume, preferential costs of borrowing, and selective accounts keep those charges artificially low. Nonsubsidized, private guarantors that cover all eligible students from all institutions bear higher costs. If the government made loans directly, it would either have to inflate government personnel rolls or rely on third parties to process, document, and collect t hese loans. When more-realistic service costs are used in the GAO model, another $1.9 billion in projected savings vanishes.

* A failure to factor the costs of auxiliary services that guarantors provide to hold down defaults and make the program effective. These include millions of phone calls and letters to help borrowers prevent defaults, training for school personnel, and quality-control activities. These costs erase another $1.2 billion in savings.

* An assumption that government bureaucrats or their low-bid contractors would have the same incentive to prevent defaults as private-sector employees. Such optimism ignores the record: For a time, the federal government sent money directly to schools under a program called the Federal Insured Student Loan Program. The federal bureaucracy simply could not run the program. Eventually, the Department of Education asked Congress to turn the program over to state agencies and nonprofit organizations. Thus, t he present system was born. Neither the GAO nor the administration has given us reason to believe that a new direct government loan system would not fizzle like its predecessor.

* Ignoring the impact of a direct-loan program on the government's borrowing cost. Investors will demand higher returns for Treasury securities, forcing up the government's cost of borrowing. Just a five basis-point increase in the 10-year Treasury securities funding this program would cost the government another $100 million.

HAVING peeled away the layers, what is left of the GAO onion? A projected loss of $13 million for the first five years. The inescapable conclusion: Not only would a direct government lending program fail to save taxpayers a dime, but it could cost them money. Worse, the GAO's estimates do not factor in the true transition costs: The change would scare many lenders into pulling out of the current system quickly, leaving students scrambling in search of loans and the government on the hook for defaults.

Real savings can be achieved while still meeting the president's goals for education access and national service. The CRS has pointed out that adjustments to the current system would make this possible. In addition, a recent proposal by the Coalition for Student Loan Reform, a group of more than 50 participants in the student loan program, would streamline the processes and cut service fees to offer actual savings. Many of the elements of that plan have made their way into legislation that the Senate wil l be considering.

This alternate vision should be carefully considered before we blindly shift to a government bureaucracy-based approach that would disrupt so many students and educational institutions to achieve uncertain benefits.