Should the government subsidize mortgages?

Fannie Mae has $1 trillion in assets and owns or guarantees more than one-quarter of all American mortages. While the company is owned by private shareholders, its securities have an implicit guarantee from the federal government. In essence private debtholders, through the company, can fund mortgages at lower risk.

It now turns out that the company has engaged in questionable accounting and financial practices; the Wall Street Journal called them “stunning.”

Put aside the details of the current scandal, should Fannie Mae exist at all? Should we privatize the entity and move away from government guarantees?

There are two core arguments for government involvement in the mortgage market:

1. There are external social benefits to home ownership.

2. The mortgage market will otherwise be stunted by credit rationing.

On the first, while I enjoy suburban sprawl, I see no good reason to subsidize it.

How about credit rationing? Are there Americans who, in economic terms, should get homes but could not in a free market? A key assumption of most (Pareto-inefficient) credit rationing models is that you know your probability of loan repayment better than does your lender. (In that case lenders are afraid to extend a full menu of loans, for fear that the “lemons” will borrow the most.) I doubt if this informational asymmetry is true today. In most cases the lender, through statistical profiling, probably has a better predictive sense of the future than do self-deceiving borrowers. And for some skeptical empirical evidence on credit rationing, see this volume. Or just try reading your spam.

Even if some Pareto-inefficient credit rationing exists, I would rather wait for information technologies to alleviate the problem. The Fannie Mae behemoth involves financial risk for the taxpayer and sets a bad precedent for government involvement in capital markets.