Did the GSEs Make the Difference Between a Not-So-Hot Post-Crisis Recovery and a Truly Dismal Pseudo-Recovery?

by Reihan Salam

I think Matt Yglesias is arguing that because Ireland experienced a housing boom followed by a financial crisis, it is silly to believe that Fannie and Freddie played a significant role. The title of his post is “Why Your Pet Theory About the Financial Crisis is Wrong.” 

A couple of thoughts:

(1) One pet theory is that as Jeffrey Friedman and Wladimir Kraus have argued, the Recourse Rule (in the U.S.) and Basel II (more broadly) played a big and significant role. This pet theory seems safe. But this is a parenthetical point.

(2) Might Fannie and Freddie have made matters worse than they otherwise would’ve been? Consider the following from Guaranteed to Fail by Viral Acharya, Matthew Richardson, Stijn Van Nieuwerburgh, and Lawrence J. White:

In Italy and Korea, mortgage debt is below 15% of GDP, whereas in the Netherlands it is 100% of GDP. Updated numbers show that Denmark’s ratio had grown to 95% by 2008, comparable to the Netherlands’ ratio of 99%. The U.K. and Ireland both had mortgage-debt-to-GDP ratios of 80%. For comparison, the U.S. mortgage-debt-to-GDP ratio at the end of 2008 was 92%, twice as high as the EU average (27 countries) of 47%. Some of this size difference is attributable to the fact that countries such as Denmark, Germany, and Italy maintain a minimum 20% down payment requirement (or a maximum 80% loan-to-value ratio; see Table 7-1). Nevertheless, as numbers for Denmark and Netherlands illustrate, the U.S. is no outlier given its much larger government involvement, to which we turn next.

Clearly, the United States is an outlier in its government involvement in mortgage markets, providing much more support than does any other country. We discuss the various programs to support home ownership in Chapter 9, and of course, there are the GSEs. While many countries have affordable housing programs for low-income households, like the Federal Housing Administration (FHA) in the United States, very few have GSE Godzillas like Freddie Mac and Fannie Mae. Two exceptions are Canada and Japan, but the market share of government-backed institutions is significantly less than that of the U.S., and the securitization market is substantially smaller to begin with. In particular, since 1945, the Canada Mortgage and Housing Corporation has insured the principal and interest on around $135 billion of first residential mortgage loans, a far cry from what Fannie and Freddie have guaranteed. And in Japan, the Government Housing Loan Corporation directly lent to households, but has been replaced by the Japan Housing Finance Agency to focus on securitization instead.

Importantly, unlike the experience in the United States, none of the GSEs of other countries have experienced exceptional losses or required government capital injections. The reason is that none of these institutions takes on significant credit and interest rate risk, as they have limited or no investment portfolios, and none has a formal affordable housing policy mandate (like Fannie and Freddie’s mission goals). Ultimately, the U.S. government’s being hand-in-glove with the GSEs, rather than at arm’s length, has allowed these institutions to grow to a size and complexity where the financial markets perceive them to be guaranteed fully for all practical purposes. This has lowered the GSEs’ cost of borrowing and allowed them to grow even further. Hence, in spite of experiencing similar housing price collapses to that of the U.S., no other country faces the uphill task of overhauling mortgage finance and reconsidering the extent of government involvement.

Because home ownership has been such an important goal in the U.S., one might argue that (at least some of) the trauma might have been worthwhile, because all of the subsidies and policies might have allowed the U.S. to have a far superior home ownership rate, as compared to that in other developed countries. As the data in Table 7-2 indicate, this is simply not the case. Other developed countries (Germany is the lone substantial outlier in the table) have been able to sustain comparable or even higher rates of home ownership, without the elaborate subsidies and policy apparatus that has characterized U.S. housing policy and mortgage finance policy. And to the extent that home ownership rates also are indicative of comparative house affordability, the same conclusion stands. [Emphasis added]

One reason Ireland has had a hard time in recovering from the financial crisis is that it doesn’t have its own currency, thus making adjustment that much more painful. We don’t have that problem.

It could be that the outsized role of the GSEs has contributed to making our post-crisis economic landscape far worse than it might otherwise have been. Had mortgage debt been a somewhat smaller share of GDP, had poor and near-poor households taken on somewhat less onerous debt burdens, it seems at least possible that we wouldn’t have taken as hard a hit to demand. If we’re concerned at all about the fact that the stagnant economy has been particularly hard on less-affluent households with underwater mortgages, this strikes me as a salient concern. A contributing factor, the authors suggest, was the role of the GSEs:

No other country saw a comparable slide in the quality of mortgages that were issued. The main reason is that the subprime market did not develop to the same extent in other countries, due to stricter regulation of the financial sector and its better enforcement, the absence of a formal government mandate to encourage affordable home ownership, and the small presence (if any) of government-sponsored enterprises in mortgage markets. This may have prevented a U.S.-style race-to-the-bottom. 

Naturally, some people are inclined to focus on the lack of strict regulation and better enforcement while others are inclined to focus on the formal government mandate and the role of the GSEs. My impression is that all of these things played a role. 

The Agenda

NRO’s domestic-policy blog, by Reihan Salam.