May 12, 2009, 7:02 amShould the Government Subsidize Mortgage Modifications? By Edward L. Glaeser
Edward L. Glaeser is an economics professor at Harvard.
Should the government be bribing mortgage servicers to modify loans and reduce payments?
Under the current housing plan, the federal government pays for one-half of the cost when lenders reduce mortgage payments to 31 percent of borrower income from 38 percent. When this plan was revealed, I was relieved because it seemed so much more reasonable than some of the extreme solutions being peddled, but a new paper by three Federal Reserve economists and a co-author has led me to reconsider my views.
There are two popular rationales for government action to increase mortgage modifications. One argument for supporting loan modifications is that foreclosures are bad for society, perhaps because they tear neighborhoods apart, and that the government should do all it can to reduce foreclosures. A second argument is that the securitization of mortgages broke the loan renegotiation process, and the government needs to fix the problem. The new Federal Reserve paper doesn’t address the first rationale for intervention, but it does leave the second argument in intellectual tatters.
Economists have long believed that dispersed property rights can lead to a breakdown in efficient bargaining. A developer trying to assemble a land parcel owned by 100 separate owners faces immense bargaining difficulties, because one recalcitrant owner can hold up a deal that would be good for everyone.
In the mortgage context, securitization has widely dispersed the rights to a mortgage’s cash flow across hundreds of bondholders. Dispersed ownership made it possible to believe that a similar breakdown in efficient bargaining might occur. Many observers took the view that the banks holding the mortgage, known here as servicers, were not engaging in sensible mortgage modification because they feared investor lawsuits. If that view is true, then there are win-win situations where loan modification can avoid foreclosure and benefit bondholders.
The new Federal Reserve paper challenges this conventional wisdom and argues that servicers are not missing sensible opportunities to avoid foreclosure by reducing payments or principal.
The first piece of evidence supporting their claim is that in their sample, among comparable mortgages, modifications appear to be equally common among loans held “in portfolio” and loans that are securitized. When banks hold loans in portfolio then property rights reside squarely with the bank and there can be no contracting difficulties. If bondholder suits were really restricting the modification of securitized loans, then there should be fewer modifications of such loans, but the rate of modification is roughly the same.
Their second piece of evidence is that “most contracts allowing modifications essentially instructed the servicer to behave as if it were the owner of a single loan.” If this is so, then it would explain why modifications are equally likely for securitized loans and for loans held in portfolio. The paper discounts the threat of investor lawsuits and asserts that “no servicer has yet been sued for making too many loan modifications.”
Perhaps they are right, and the threat of lawsuits has been made up by servicers themselves, because a phantom threat gives servicers an easy excuse for not modifying loans.
But what about the huge losses that servicers take when a foreclosure occurs? Surely the costs of foreclosure mean that renegotiation can be a win-win situation.
The problem with that logic is that it is hard to determine who is actually going to walk away from a house, and it is equally hard to determine whether a modification will permanently avoid a foreclosure. The paper shows that cutting payments has only a modest effect on the likelihood of default, which suggests that even if modification avoids a foreclosure today, a troubled borrower is likely to default again in the future. Moreover, a generous modification policy ends up cutting payments for plenty of borrowers who would never have defaulted.
If this evidence is right, then loan modifications aren’t a win-win situation made impossible by securitization. Servicers, who have a fiduciary duty to bondholders, actually have an obligation not to engage in loan modifications that lose money.
Does this mean that the government program is wrong? Not necessarily, but it needs to be justified by the social costs of foreclosure. Moreover, if these economists’ evidence is right and reducing payments does little to reduce defaults, then the government program will do little to reduce foreclosures.
If each foreclosure carries social costs grave enough to warrant using taxpayer money to subsidize servicers, then the United States also needs to stop encouraging poorer Americans to borrow to buy homes. Instead, the appropriate government policy would be a tax on risky borrowing that would discourage risky borrowing. I’m not quite in that camp yet, but I do believe that too many people, including myself, accepted the view that a contracting failure prevented the modification of securitized loans.