Trends in the housing sector have been a driving force behind the recent financial crisis and associated recession. According to the Case-Shiller/s&p Indices, housing prices fell over 30 percent from the height of the housing bubble to August 2011 across a twenty-city composite, with prices in some markets down by nearly 60 percent. This plunge in housing prices was accompanied by a wave of household defaults and foreclosures, which has led to millions of property owners losing their homes over the last three years. The foreclosure crisis hit hardest in areas that had the largest bubbles, notably parts of Arizona, California, Florida, and Nevada.
The collapse of housing prices plays an important role in spurring foreclosures since “underwater” borrowers, who owe more on their mortgage than their home is worth, have a substantial incentive to walk away. According to CoreLogic, almost one in four borrowers were underwater as of the first quarter of 2011. Households with negative equity perceive their house to be a negative contributor to net worth, resulting both in a higher future default probability as well as lower current consumer spending. The decline in house values has also badly damaged the balance sheets of even above-water borrowers, as housing-related assets still remain the primary form of savings for many households. This link works in both directions: The wealth impacts of lower home prices weaken consumer and business spending while the weak economy in turn feeds into housing price declines. Prices have continued declining in many areas, with downward pressure exerted by the overhang of millions of properties in foreclosure. The problem of elevated foreclosures is now in its fourth year, with the initial wave of subprime-related problems replaced by more traditional foreclosures that reflect the weak economy. In short, the presence of millions of properties in the foreclosure process represents an overhang of supply and remains a major stumbling block for housing price recovery.
Mortgage-related defaults and associated declines in the market value of residential real estate collateral have been responsible for large losses in the financial sector, which led to the collapse of major financial institutions and widespread economic losses. The role of the housing sector as a source of systemic risk throughout the economy has made federal housing responses one of the central aspects of government economic policymaking during the last several crisis years. Indeed, given the large role that the federal government plays in mortgage finance today, it would be basically impossible for policymakers to simply step aside completely. The government provides massive financial assistance to homeowners through tax deductions on mortgage interest and property taxes paid to state and local governments, and through the ongoing sponsorship of Fannie Mae and Freddie Mac. In many ways, the historical roles of the Federal Housing Administration in providing mortgage guarantees and Fannie Mae and Freddie Mac in securitization are responsible for the spread of the 30-year, fixed-rate mortgage, which is now seen as the default mortgage contract.
Yet the housing crisis saw a dramatic escalation in the role played by the federal government. These new interventions can be broadly placed in two categories: First, policies aimed at boosting the demand for housing generally and thus supporting sales and prices; and second, supply-management policies assisting distressed homeowners in avoiding foreclosure.
As the depth of the housing crisis demonstrates, these policies — pursued by both Republican and Democratic administrations — have failed to stem large-scale housing losses. In large part, this failure reflects the magnitude of the housing problem in which tens of millions of American families got into homes they could not afford. Policymakers were consistently behind the curve in estimating the extent of problems in housing, and were unable to tackle the core issue of excess mortgage debt, which was made worse by a precipitous fall in home prices.
However, the difficulties and failures experienced by policymakers combined both tactical problems in dealing with complex issues in a crisis environment with political concerns surrounding the proper scope for federal housing policy. Policymakers faced challenging philosophical questions regarding the degree to which taxpayer money should legitimately be used to prop up the housing sector or assist families that had made speculative financial decisions. Perhaps most striking is that even with the benefit of hindsight, it’s not clear whether any politically feasible and effective policy could have been implemented midstream. A key political obstacle to large-scale interventions to avoid foreclosures is the intrinsic unfairness involved in helping homeowners who are in houses they cannot afford when others who were potentially more responsible with their personal financial decisions are not receiving assistance. This issue of horizontal inequity makes it much easier to say that “more should be done to help avoid foreclosures” than come up with a policy approach that is politically feasible. Understanding the nature of these policies is important as a case study in economic policymaking in a crisis situation, and it is useful in thinking through how to better calibrate future federal housing policies.
Though fears of a housing bubble and bust have been around since 2002, the level of concern among policymakers picked up dramatically as foreclosure and delinquency rates rose in 2007. The lack of focus earlier in the decade was not, as is often alleged, due to some systemic lack of awareness of the housing sector. Rather, policymakers were reluctant to act earlier because they failed to understand how structural shifts in the housing market had fundamentally changed the economic landscape, which in turn made a lot of the existing housing-related data an unreliable gauge for future projections.
For instance, delinquency statistics — even among subprime mortgage products — were quite low by historical standards. In retrospect, it is clear that homeowners were more distressed than this measure indicated at the time. The prospect of refinancing or selling a home in a booming market gave many homeowners a strong financial incentive to keep making payments. Put another way, the anticipation of future housing gains encouraged the vast majority of borrowers to stay current on their mortgages because the value of their house or housing-related investments were growing rapidly. Once refinancing opportunities ended after the Federal Reserve raised interest rates in the middle of the decade, however, subprime borrowers began to default in larger numbers.
There were also relatively few signs, at least on the standard federal policymaker’s dashboard, suggesting that problems in the housing sector were widespread or would have an impact on the broader economy. For instance, an influential paper from the New York Fed in 2004 suggested that though there was a housing boom, its effects were limited to a handful of markets where demand for housing was high and the supply of new housing was scarce. House price gains in other parts of the country seemingly corresponded to changes in shifts in supply and demand. In Las Vegas and Phoenix, to pick two important and booming markets, house prices were tightly correlated with local population growth.
Another survey by Boston Fed economists has stressed the uncertainty inherent in predicting future housing prices. During the height of the bubble, some economists emphasized the role of fundamentals, such as zoning restrictions or rising housing demand. Other economists — including Paul Krugman — noted speculative trends in housing, but associated them primarily with certain areas of the country where geographical restrictions limited construction. Finally, other economists were pessimistic about housing prices, but many of these economists had been predicting a housing bust for years. Indeed, the famous housing bust prediction in 2003 by the Economist proved to be overstated. In fact, the Case-Shiller House Price Index, as of the first quarter of 2011, was only slightly lower than it was when the article came out (by about four percent).
In our view, the failure of the economics profession to clearly predict a housing bust is mostly indicative of the general difficulty of any such forecasting exercise. Yet even if one thinks that providing accurate price forecasts is feasible, it is important to recognize that policymakers were simply not dealing with unambiguous information at the time.
It’s also worth bearing in mind that a narrow focus on housing prices misses the broader picture in the mortgage and real estate markets. Policymakers did consider the possibility of housing-related losses in the market. Their key error lay elsewhere, however: in the belief that the system held enough checks to prevent such shocks from resulting in broader systemic losses. The growth of structured mortgage products — which pooled together the payments from mortgages into mortgage-backed securities that were sold to sophisticated investors — promised to disperse risk and ensure that mortgage-related risks were held by the entities that were best situated to bear them. Meanwhile, the Federal Reserve had demonstrated in the aftermath of the dot-com boom that a fall in asset prices, in and of itself, could be handled through an accommodative monetary policy. The supply of credit did not need to be shut down for the entire economy in response to possible excesses in one sector — or so the thinking went.
Finally, policymakers failed to appreciate, at least initially, the nature of product changes in the mortgage market — toward subprime and alternative mortgages that were frequently lent to borrowers with poor credit histories or low levels of income documentation. Most subprime mortgages were originated as hybrid mortgages, in which a comparatively low initial interest rate paid on the mortgage “reset” after a two- or three-year period to a higher rate. Many of these products were designed to help homeowners initially qualify for a mortgage with the expectation that they would then refinance into another mortgage once the interest-rate reset was about to occur. The short duration “teaser” period was partly designed to allow the borrower to develop a payment history that would ease qualification into a prime mortgage after a couple of years. This was also partially dependent on the expectation that house prices would rise to generate the equity necessary to refinance into another mortgage.
In retrospect, the nature and structure of residential lending during the boom years carried a number of dangers. Many mortgage products and forms of residential lending were geared around exploiting the bubble in housing prices. These hidden problems were initially contained due to low interest rates, steadily rising housing prices, and the continuing flow of mortgage finance or credit. All of these factors enabled homeowners to continue making low monthly payments while waiting for their next refinancing opportunity (often, extracting equity in the process). Finally, the system encouraged homeowners to continue making mortgage payments in the hopes of reselling their homes later at a higher price. All three of these factors had largely stopped by 2006, at which point subprime borrowers started defaulting in larger numbers.
The crisis hits
A series of factors raised the profile of housing in 2007, even as federal concerns remained limited to a handful of states or regions and to certain mortgage product types. The initial policy response was twofold: work to maintain demand for housing generally while extending assistance to some borrowers on the margin who were “in the right house” but had the “wrong, exotic” mortgage product.
A brief timeline of early responses from Congress and the Departments of Treasury and Housing and Urban Development includes the following:
- August 2007. fha Streamline Refinance: A program in which the Federal Housing Administration insures (but does not originate) mortgages for qualified low- and moderate-income borrowers who have less-than-perfect credit and little savings for a down payment. In general, this initiative presents borrowers with a fixed-rate government insured refinancing opportunity.
- October 2007. hope now: A private-sector alliance of mortgage industry participants was launched to encourage servicers, housing counselors, and investors to work together to help streamline the process of modifying mortgages for borrowers with adjustable-rate mortgages who can afford their current payments but will have trouble when their interest rates rise.
- July 2008. hope for Homeowners Act (h4h): A program that provided more accommodative underwriting standards to incentivize private lenders to refinance up to $300 billion in mortgages into new fha-guaranteed mortgages with lower loan-to-value ratios.
As Phillip Swagel has noted in his account of Bush-era decisions, policymakers noted worsening real estate conditions, but incorrectly thought that the worst of the crisis would pass as the wave of subprime-related foreclosures crested in 2008. This prediction was made on the basis of examining the historical relationship between foreclosures and economic factors such as interest rates and unemployment. However, new subprime mortgages were originated using substantially worse underwriting criteria, fundamentally changing the interaction between housing and the rest of the economy. As economists Christopher Mayer, Karen Pence, and Shane Sherlund have found, the originate-to-distribute model was associated with poorer underwriting quality of mortgages. The median subprime mortgage was originated with a loan-to-value ratio of one, meaning that the homeowner put no money down in order to purchase the home. Because such homeowners had no skin in the game, they were particularly susceptible to home price declines, which could immediately put the borrower underwater. Without the financial pain associated with losing equity in the home, borrowers are more likely to become “strategic borrowers” — walking away from their mortgage in times of distress.
Adding to strategic default concerns, many mortgages are originated without recourse, prohibiting banks from going after borrowers for the remainder of their mortgage balance in the event that housing collateral proves insufficient for full creditor recovery. In certain states, the lender may have full recourse, but the legal regime does not allow deficiency judgments to be secured without large transaction costs. Finally, a rising share of home purchases were made by investors seeking second homes, or for other speculative reasons such as hoping to capitalize on rising home prices. All of these factors made owners more likely to stop making payments on their mortgages in times of stress, particularly if they were underwater. This structural break in historic relationships would make the crisis unprecedented in magnitude.
Late in 2007, as the number of homeowners facing delinquency and foreclosure grew, the federal government also initiated projects aimed at helping homeowners stay in their homes. The Hope Now Alliance, which began as an initiative between hud and Treasury in October 2007, worked to connect private mortgage servicers, borrowers, and counselors. The goal was to encourage private mortgage lenders and servicers to voluntarily extend mortgage modifications to borrowers, which would lower interest rates or mortgage balances and enable more homeowners to afford their mortgage payments. In practice, such modifications frequently worked only to lower the interest rates or extend the term on the mortgage. One major goal was to facilitate modification efforts among privately securitized mortgages. The legal agreements governing the servicing of mortgage pools — pooling and servicing agreements — frequently left servicers unclear about their ability and responsibility to pursue different types of modifications.
Other programs allowed the federal government to directly refinance private- sector, adjustable-rate mortgages into government-guaranteed, fixed-rate mortgages with a lower interest rate. Throughout the crisis, policymakers would attempt to implement various refinancing strategies through a variety of different lending agencies. These programs universally failed to work as advertised. Hurdles for mass refinancing programs include the fact that equity and credit requirements (e.g., tighter underwriting standards) prevented many homeowners from qualifying. In addition, many homeowners had already refinanced or otherwise were currently paying relatively low rates, meaning that a subsequent refinancing would not substantially lower their payments, especially after taking transaction and closing fees into account. The h4h program, though it was intended to help 400,000 homeowners through refinancings, eventually ended up completing only 23 endorsements in 2009, 207 in 2010, and 632 in 2011. After being enacted into law in 2008, the program shut its doors on September 30, 2011, having helped fewer than a thousand people.
These federal responses can be classified as supply-management tools to handle the growing surge of delinquent mortgages. Additionally, policymakers took steps to manage the demand for housing by using the government’s leverage over housing finance.
Two of the housing gses — Fannie Mac and Freddie Mae — dramatically stepped up mortgage originations as the number of private securitizations slowed, while the fha grew to become one of the largest sources of new mortgage originations in the country. gse purchases and fha endorsements totaled around 30 percent of originated mortgages in 2005; a figure which increased to 50 percent by 2007, and rose even more by 2008. As of the first quarter of 2011, the U.S. government is involved with the vast majority of new mortgage originations in the country — a combined share of more than 97 percent between the gses, the fha, and other smaller government agencies. These developments were not without financial costs — many of the mortgages originated or otherwise acquired by the gses contained substantial credit risk that was ultimately paid by taxpayers after these agencies were put in conservatorship. For instance, close to fifteen percent of Fannie-backed mortgage holders from 2007 were seriously delinquent on their mortgages within three years of origination — statistics that are far higher than comparable mortgages before or after this period.
As with several other housing programs, the rapid growth of federal mortgage supply has gone from a temporary intervention to a more structural feature of the mortgage market. In some respects, this development is justified by ongoing difficulties in the housing market. However, it is increasingly difficult to imagine an exit strategy for many of these interventions. As long as the housing sector remains weak, there will be incentives to maintain the government’s outsized role in housing — with taxpayers (particularly renters) left paying the bill.
On the housing supply side, difficulties with most mortgage-support programs became quickly apparent. The focus on particular mortgage products or particular legal frictions proved insufficient to handle the scope of foreclosures. In hindsight, it seems that targeting such market inefficiencies did not prove to be an exhaustive solution — and an excessive focus on subprime characteristics may have prevented the government from generating more comprehensive solutions. The scope of housing problems spread to prime borrowers who faced both rapidly declining home prices (placing many more borrowers underwater) and higher levels of unemployment. As a result, more traditional or prime borrowers started to go delinquent in unprecedented numbers. A destructive cycle soon developed, as more foreclosures drove down housing prices, which drove further foreclosures. Problems in housing resulted in losses for financial firms, which also drove down the supply of credit and hurt overall economic conditions. Though the federal government frantically attempted particular innovations to tackle individual aspects of these problems, it was fundamentally unable to develop a strategy to end this destructive cycle.
In particular, the Bush administration was hampered in changing the nature or scope of mortgage alleviation efforts. At the time, most modification efforts did not directly aim to tackle the issue of underwater borrowers. There is a dilemma between assisting homeowners with payment or interest-rate problems — borrowers who incur at least some sympathy — and assisting borrowers with negative equity, which may actually be a more effective approach but involves possibly extending taxpayer funds to strategic borrowers, who have the ability, but not the desire, to continue making mortgage payments once they are underwater and their home no longer remains a good investment.
When it came to gse-held loans, modification efforts (even those that were not subsidized) still often generated a direct loss to taxpayers. Policymakers also had to wrestle with aligning mortgage assistance programs so they were not exposed to fairness concerns by extending relief only to borrowers with private mortgages as opposed to gse-backed mortgages. Federal Flow of Funds data for the second quarter of 2011 suggests that the U.S. government backs over 58 percent of the current mortgage market through its control of mortgage finance entities like Fannie, Freddie, and the fha. In the end, the Bush administration was unwilling to devote large amounts of public funds to mortgage modification efforts aimed at lowering a borrower’s mortgage principal; it recognized the inequity dilemmas with this type of assistance and that many distressed borrowers were in the “wrong house,” as opposed to the “wrong mortgage.”
While the federal government was willing to devote efforts to make mortgages more affordable in 2007 and 2008, it was generally unwilling to transfer taxpayer funds from renters to homeowners, or from homeowners struggling to make payments toward those who had opted not to pay, possibly for strategic reasons. When borrowers take out an outsized mortgage they can’t really afford with the goal of reselling it in the future, but then run into financial difficulty, it is not clear the extent to which homeowner assistance deserves public funding. A taxpayer-financed policy may help irresponsible borrowers at the expense of the more prudent borrowers who avoided onerous mortgage debt. A similar reluctance — or a respect for this concern — can be inferred from congressional Democrats, who designed the h4h program in a way that destined it to serve only a narrow segment of troubled borrowers. The problem is that allowing for a broader targeting would inevitably result in taxpayer dollars going to help more “unsympathetic” borrowers.
Even private servicers face difficulty in constructing appropriate modification programs. As Boston Fed economist Paul Willen and co-authors discussed in a 2009 paper, modification efforts are hampered by the problem of cure-rates and re-default rates. Delinquent mortgages may resume payments even in the absence of a modification if homeowners are able to resume payments through their own effort. However, even a successfully modified loan may re-default. While lowering mortgage principal may raise the probability of a homeowner making payments, it also reduces the final funds received by the lender and could induce other homeowners to stop paying in the hope of obtaining a similar modification. Finally, the practice of processing and extending modifications proved to be an expensive task that was highly dependent on professionals to sift through necessary information. Lenders must balance all of these factors.
One study on modification efforts by researchers at the Chicago Fed found that lenders failed to renegotiate 93 percent of delinquent privately securitized loans within six months, and failed to modify 90 percent of loans held on bank balance sheets. While the difference between these two figures suggests that securitization may inhibit modification to some degree, the overall rate of modifications — even for loans on privately held mortgages — remains quite low. Even gse-held loans are renegotiated infrequently. Among other factors, the presence of other mortgage liens has inhibited the modification process. The owner of a second lien, for example, would often have little expectation of a recovery on a deeply underwater borrower but retained the legal right to prevent a modification of the first mortgage. This presents a so-called “hold-up” problem under which the second lien holder would need to be paid off, raising the cost (and administrative difficulty) of a modification.
During 2008, while the foreclosure crisis was quickly escalating, a number of outside analysts suggested alternative housing resolution tactics that would have resulted in the mass write-down of debt. One popular proposal was creating a new chapter in the bankruptcy code that would allow bankruptcy judges to write-down the principal on mortgages for bankruptcy filers (“cramdown”). This proposal passed the Democrat-held House, but was never enacted into law. Another mass write-down proposal came from Luigi Zingales, who suggested that homeowners in zip codes which experience home price falls above 20 percent should have the right to write down the value of their mortgage by that amount, and let the lender share in a portion of the equity gains upon selling the property.
These mass write-down proposals had many attractive qualities. By relieving the debt burden on borrowers, they would have encouraged borrowers to continue making payments on the mortgage, which could in turn have helped stem the tide of foreclosures, mortgage losses, and (systemic) banking problems. With a lower aggregate housing debt burden, households would experience a gain in wealth, and also be more likely to resume spending — contributing to overall economic growth.
Yet the Bush administration also had good reasons to oppose these proposals. Instituting a policy of mortgage cramdown during the midst of the crisis would have reduced the remaining flow of private sector mortgage finance. Cramdown would have further represented a retroactive change to the terms of contracts, potentially giving rise to “rule of law” concerns by creditors in other markets and fundamentally disadvantaging mortgage lenders relative to other lenders (e.g., credit cards). By renegotiating existing mortgage contracts, it was very possible that credit markets more broadly may have been pushed into further turmoil. In addition, only a few hundred bankruptcy judges would have been tasked with and relied upon to effectively write down the mortgages of thousands of borrowers, along with unsecured debt that may not be a contributing factor behind borrower delinquency. In short, there were legitimate concerns — separate from fairness — about whether this proposal could be scaled appropriately and not suffer from some of the same resource or bandwidth constraints that have plagued mortgage servicers generally over the past few years.
Wider modification plans might also create moral hazard — encouraging a further wave of defaults. There is some evidence that lowering the cost of default and expanding modification efforts may drive further delinquency — one study found that many borrowers who were previously current started to miss mortgage payments in order to qualify for a modification program announced by Bank of America for loans originated or serviced by the subprime servicer Countrywide. The effects were particularly pronounced among borrowers who were underwater on their mortgage, suggesting that such borrowers may be very responsive to programs that lower the cost of delinquency.
As Luigi Zingales has suggested, there are strong moral and social factors behind household decisions. When most underwater borrowers continue to make their payments, they create a strong social norm in favor of doing so, even though many borrowers could make themselves financially better off by going delinquent on their mortgages. By contrast, homeowners personally acquainted with other strategic defaulters are more likely themselves to stop making payments. Shifting social norms may be responsible for many of the defaults in the crisis, and the effect of more widespread measures of modification or cramdown on these norms would be uncertain.
Even programs of mass principal write-downs face serious difficulties. A mass write-down of mortgage debt would result in enormous losses for taxpayers, given that most mortgages are held by the gses, with taxpayers holding the ultimate financial loss. Private financial institutions would also face serious losses, and for some possible insolvency, since principal reductions would force a revaluation of assets. One way to offset the lender loss on a write-down would be to try to convert the residential mortgage so that the lender would share in any future home price appreciation. While these sorts of offsets were considered (and even tried on a limited basis in h4h), the immediate reduction in asset value or principal was generally considered to be the much larger problem for prudential regulators.
Nevertheless, pursuing more radical modification strategies by the private sector during the depth of the crisis would have had the advantage of pushing the administration’s goals at the moment of greatest leverage. As banks were facing crisis and would soon receive federal funding (e.g., tarp), they would be most likely to entertain thoughts of bold modification plans. For instance, Bank of America agreed to the settlement with the states attorneys general with Countrywide in October 2008. By not pressing harder while the banks were still utilizing their tarp loans, policymakers were left with a weaker bargaining position from which to pressure banks once those loans were repaid.
In the absence of such mortgage modification efforts, Congress passed the Housing and Economic Recovery Act (hera) in 2008; it paved the way for a number of piecemeal programs or attempts to patch the housing market. The bill authorized the previously discussed h4h program, while also introducing two new credit programs. On the supply side, a new Neighborhood Stabilization Program (nsp) — administrated by hud — awarded grants to those states and territories facing a disproportionate share of foreclosures and delinquencies; while on the demand side a tax credit program awarded up to $7,500 for first-time home buyers. Despite three grant rounds totaling almost $7 billion, the nsp proved to be too small — divided up among too many areas — to have a substantial impact, while the tax credit was often captured by homeowners who would have bought anyway or otherwise accelerated their purchase decision. Even if the tax credit program shifted some home purchases forward, this came at the cost of lowering future demand. By successively boosting temporary demand at the cost of future demand, this program served as little more than a short-term palliative while creating future problems like a fiscal headwind if an extension or additional subsidies did not follow.
Ultimately, the limited scope and poor targeting of these programs did little to address fundamental problems in housing. However, the (relatively) limited fiscal outlay demanded by these programs would prove attractive to some future legislators seeking to implement additional solutions.
The final months of the Bush administration were focused on stabilizing overall financial markets rather than focusing on housing. tarp was initially aimed at reviving the market for mortgage-backed securities, particularly those that were considered toxic assets. The underlying motive was that the problems in mortgage finance revolved around a lack of liquidity in this market, which could be revived through the price discovery process induced by a large round of government purchases. Yet this logic was highly questionable, as demonstrated by the numerous criticisms the plan attracted in real time. Had the government gone ahead with the plan as scheduled, taxpayers may well have lost hundreds of billions of dollars rather than recouping the vast majority of their investment. The problems in structured finance, it would turn out, had to do with the solvency of mortgage credit, an issue no government solution has managed to successfully address.
The Obama years
Though the obama administration came in with an optimistic outlook and a strong desire to raise the level and scope of federal assistance for homeowners, it ran into the same tactical and political problems faced by the Bush administration. In terms of foreclosure alleviation efforts, the Obama administration dramatically ramped up prior efforts, using tarp funding to finance the Making Home Affordable (mha) initiative and its signature component, the Homeowner Assistance Modification Program (hamp). The program harnessed previous efforts that linked servicers with borrowers, but used taxpayer funds to incentivize certain modification efforts and further the objective of increasing mortgage affordability.
While the Treasury initially estimated that the hamp could help three to four million borrowers, the program has thus far made only about 800,000 ongoing permanent modifications. With its doors open since March 2009, the program is not expected to see any significant improvement in its take-up rate, as new modifications have started to slow down. In fact, a large portion — around 900,000 — of hamp trial and “permanent” modifications have already been canceled. While Treasury initially budgeted $50 billion in tarp funds and $25 billion from Fannie Mae and Freddie Mac for a total of $75 billion for hamp, it later reduced its projected contribution for hamp to $29.91 billion. Separately, cbo estimated that as a result of lower than expected participation in the program, the total cost or spending under hamp (or for housing-related programs financed by tarp) would be more like $12 billion in total.
Under hamp, fewer than 30 percent of borrowers received a principal reduction — despite the fact that academic research has suggested that overall debt burdens are the primary factor driving default. The vast majority of modifications under the program simply extended the term of the loan or reduced the initial monthly interest rate (potentially increasing the total mortgage amount owed over the life of the loan).
The success rate of hamp modifications has been disappointingly similar to that of other subprime modification efforts: The results are just not that great. For example, Fitch Ratings estimates that 75 percent of borrowers will ultimately default again, while Barclays projects a re-default rate of 60 percent. These figures are broadly comparable to estimates based on non-hamp subprime modifications, which are structurally similar. New York Fed economists Andrew Haughwout and Joseph Tracy have calculated the re-default rate and found that an astonishing 56 percent of modifications were back in default twelve months later. These figures are similar to those reported by Fannie Mae and Freddie Mac, which have been tasked with refinance programs.
Ultimately, it seems that the Obama administration has come to view hamp more as a supply management tool for the housing sector rather than a plan to keep millions of struggling borrowers in their homes. For instance, in a meeting with prominent financial commentators, Treasury officials stressed the importance of hamp in preventing a rash of foreclosures at an inopportune time. Even if many borrowers who may have defaulted and faced foreclosure ultimately do fall behind on payments and lose their homes, the administration could still declare success because the banking system as a whole was able to avoid added stress during a critical period. Yet this implementation of mortgage modification as supply management is not how hamp was designed or marketed, which was as a generously funded tarp initiative.
It is easy to see why taxpayers would balk at the prospect of spending tens of billions of dollars to assist banks and merely drag out the foreclosure crisis. Though there are arguments in favor of Treasury’s adjusted narrative for hamp, it is fundamentally based on the counterfactual assessment that more foreclosures (at that time) would have destroyed the system as opposed to speeding the transition in the real estate market. In the absence of any economic literature studying this question, it is difficult to fully evaluate the consequences of a Treasury policy that amounts, more or less, to “extend-and-pretend.” A more cynical interpretation is that the Obama administration was trapped between its own rhetoric and financial realities. Having criticized the Bush administration on this issue, the new president had to do “more” to avoid foreclosures. But the costly and unfair transfers to unsympathetic cases that would be involved in actually doing a lot more left the administration with a hamp system that gave the appearance of action but had a lower than expected cost and modest outcome.
Other modification efforts also came under the mha program. Several other subprograms involved the assistance of various government and private entities in coordinating modification efforts — in particular, to unemployed borrowers, those hoping to refinance into an fha mortgage, and those with second liens. Ultimately, these efforts also failed to translate into mass modification or refinancing efforts for most borrowers.
As with the Bush administration, the Obama administration soon discovered the difficulty with making mass modifications and refinancing work. Faced with these challenges, the Democratic Congress decided to double down on the other supply and demand tools available to legislatures — neighborhood assistance grants and taxpayer credits for home purchases. With the number of struggling households growing, legislators faced immense pressure to “do something” to assist homeowners, and it was often easier to gather political support around ideas that had previously been embraced by Congress.
The Hardest Hit Fund was established using tarp money in February 2010. As with the previous Neighborhood Stabilization Program, money was allocated to states and localities facing severe economic distress. States were allowed to spend money to assist unemployed borrowers or offer mortgage modifications. Over $4 billion was spent over three rounds of grants. Policymakers seemingly ignored the negative data surrounding prior iterations of these programs, and to this day we still lack a proper accounting of how the money appropriated for these initiatives was actually spent.
Even more detrimental to taxpayers was a new round of homebuyer tax credits, which according to irs projections resulted in $24 billion in lost revenue. As with the first round, the bulk of this money — perhaps as much as $21 billion — simply went to homebuyers who would have made purchases anyway.
Yet after such disappointing results, the real failure of the Obama administration has been a commitment to spend meaningful amounts of taxpayer resources without achieving broad-based results. For much of 2011, media discussions of White House economic policy had largely omitted housing policy. It seems clear that the principal economic policymakers in the administration — including Treasury Secretary Geithner and former nec Chair Larry Summers — did not see cramdown as a viable housing strategy. It remained a talking point for the administration but not an agenda item to advance through Congress (even though both the House and Senate were controlled by Democrats in 2009 and 2010). Though many Democrats, including President Obama, came into the office on the platform of advancing mortgage cramdown, the same problems with this policy that hampered the Republicans prevented the Democrats from moving further. Yet while the Republican White House was able to argue for a limited position on taxpayer assistance from the point of view that such spending represented an illegitimate or poor use of taxpayer funds, the Democratic White House has not been limited by such concerns and has failed to deploy such funds wisely or effectively. While the market continues to track housing-related data, housing policy — or the development of new ideas and approaches — has largely fallen out of the national political conversation. The slight resurgence of housing finance related ideas in the second half of 2011 appears to replicate the discussion around the ineffective plans of prior years, focusing on technical changes that could boost marginal participation.
In hindsight, the greatest missed opportunity for the Obama administration came through its handling of tarp funds. Major financial institutions like Bank of America, Citigroup, and Wells Fargo were allowed to exit from tarp funding without resolving their excess mortgage debt — in particular, second mortgage liens, which were frequently worth very little compared with their valuation on bank balance sheets. Despite the fact that second liens remain an impediment to household solvency and some modification efforts, banks were not (and have not been) required to mark these mortgage assets to market value. Again, a major concern for prudential regulators appears to be whether a revaluation of second liens (as with first liens and principal forgiveness) could raise solvency concerns at some institutions.
Early federal policy efforts were instead focused on more discrete issues like “exploding arms.” As the Fed dramatically lowered interest rates, however, many homeowners with arms actually faced lower rates, which should have obviated a large part of this problem. In other cases, lenders and servicers made independent efforts to extend or maintain homeowners’ initial interest rates to enable them to stay in their homes. In the end, it turned out that the problems in housing could not be explained by one particular mortgage type. At best, policies aimed at converting arms to fixed-rate mortgages helped only on the margin, since it was later revealed that fixed-rate mortgages (both subprime and prime) were hardly immune from the crisis.
The financial crisis can only be understood as a part of a cycle connecting real estate, finance, and other sectors of the economy. Problems in the financial sector stemmed from deteriorating conditions in the housing sector, which themselves drove further economic problems, including a new wave of unemployment and foreclosures. The principle role of housing in the crises over the past several years, combined with the structural problems related to the resolution and management of real estate debt, suggests that housing and mortgage finance policy must remain a major focus of federal policymakers moving forward.
Unfortunately, there are not any simple solutions to handling housing-related problems. Through real estate policy was a major area of focus during the later years of the Bush administration and the early years of the Obama administration, neither group was able to pursue effective housing policies on a cost-efficient basis.
The state of play with housing policy today presents a mess. On the one hand, the scope of the problem is just enormous. Yet literally no effective policy plans have been advanced over the past four or five years. Even with perfect hindsight, though, it is difficult to imagine how policymakers could have adequately managed the various technical and philosophical questions in real time. On the other hand, the solutions that have been offered — whether on the side of managing supply (in the form of modifications or refinancing) or demand (in the form of tax credits or federal mortgage underwriting expansions) — have almost uniformly under-delivered and presented taxpayers with an enormous cost.
Losing the fight against the leviathan
This review of housing policy over the past few years provides an illustration of the difficulties policymakers face trying to manage a complex economy with limited information. It also offers lessons for the future of housing policy.
While several of the aforementioned housing programs have since closed shop or are in the process of being phased out, many elements of the government’s approach to credit markets have become entrenched, if not permanent. Even before the housing bust, the federal government had an extensive presence in the consumer credit market — including the provision of large tax benefits to those who take out mortgage debt and low interest rates maintained by the Federal Reserve. As economist Raghuram Rajan has argued in his book Fault Lines, credit support was a popular way for the federal government to help finance otherwise unsustainable household spending habits at a time when many households were starting to struggle.
Yet the degree of federal involvement in this market has now grown to staggering proportions. As noted above, the federal share of the entire mortgage market is now 58 percent, a hike of thirteen percentage points since 2006. The Treasury and Federal Reserve have also drastically raised their holdings of mortgage-backed securities, now totaling over $1.1 trillion, which is more than commercial banks. Every passing month brings us closer to a future in which the government is, in effect, the entire mortgage market. Meanwhile, ongoing economic difficulties present an omnipresent rationale for further government intervention in this area. After all, what legislator is willing to oppose a credit program that will benefit his or her constituents today, at only the risk of taxpayer losses in the future (when he or she may no longer be in office)?
While many of the housing responses to the crisis have been sold as “temporary” government programs, it is clear that their cumulative impact is a further consolidation of an already dangerous level of dependence on government-supplied credit. Hindsight reveals that the economy was perilously reliant on booming house prices for much of the 2000s. Yet rather than attempting to slowly rebalance the economy towards other productive sectors, federal housing policy has been consistently aimed at altering the supply and demand dynamics in an attempt to re-inflate the bubble. This can perhaps best be understood by tracking the policymaking pattern over the past few years, where nearly every programmatic failure has led to a subsequent discussion and push by policymakers for a new or revised program to tackle the original problem (again).
Housing policy can really only move forward when policymakers recognize the limits of and potential fallout from a credit-dominated economy. Moving back to a capital market system that is firmly under the control of private firms should be a near universal aim. To be sure, private firms have made many mistakes as well — both before and throughout this crisis. However, we believe that the evidence of recent housing policy crisis management — not to mention more than $180 billion in taxpayer funds that have been injected into Fannie Mae and Freddie Mac to keep them solvent — suggests that the government is unlikely to prove a better manager of capital allocation in the long term.
The next thrust on housing policy will be a debate about how to establish the right legislative framework for the long-term sustainability of housing finance. While the housing market is still fragile, fundamental reforms need to happen slowly. Yet, it is important for policymakers to understand that the desired transition speed is distinct from starting the legislative debate and committing to a new future system. After all, it is only with clear rules about how the mortgage finance market will function that consumers, lenders, and other investors will be able to make the necessary decisions that will form the foundation of a broader recovery in housing.
A new mortgage compact
The fallout from the housing crisis, as well as the large and growing federal footprint in this space, almost ensures that policymakers will spend a great deal of time debating long-term reforms in the coming years. As a partial step towards addressing these issues, we propose a path to a new mortgage contract or framework. We begin by observing that the housing crisis was driven by a confluence of factors: low down payments encouraged speculation, the lack of mortgage recourse encouraged strategic behavior, while inflexible mortgage contracts prevented household debt renegotiation. Our proposal reverses all three of these factors.
The first change would involve higher mortgage down payments — at least 10 percent of the value for most mortgages, though 20 percent is also worth considering and could be preferable for many mortgages. Though such a policy might restrain homeownership and negatively impact the value of the existing housing stock to some degree in the near term, a requirement like this is the key to ensuring that more homeowners maintain equity in their homes in the long run. Though a prevalent assumption is that buying remains a better option than renting, new research has emphasized that this choice is not so straightforward from a financial perspective, suggesting that it may make sense to take a second look at policies that encourage home ownership generally. Loosening down payments to allow borrowers to acquire a new home really only makes sense if house prices only go up.
Meanwhile, ensuring widespread mortgage recourse with the possibility of cramdown of mortgage debt could present homeowners with an equitable and flexible mortgage contract. With recourse, homeowners would be liable for any underwater value of their mortgage in the event of a default, but their higher equity position at the outset would make any future underwater scenario less likely. The possibility of cramdown under bankruptcy provides a legal mechanism to renegotiate mortgage debt.
The prospect of ensuring mortgage recourse alarms many commentators concerned with equity, while cramdown is opposed by those who worry about the sanctity of contracts, the rule of law, and the impact on future credit availability — or the cost of credit in the future. We feel that policymakers should consider instituting both prospectively over a phased period, while generally rebalancing government policy away from housing in the form of fewer subsidies and higher down payment requirements. To alleviate many of the concerns with this sort of an approach, it is important that these changes get implemented simultaneously and over a number of years. A useful analog is that the major bankruptcy reforms that were enacted in 2005 took more than ten years to get through Congress.
It is also important to recognize that many mortgages are already recourse. Individual state law currently drives this issue and 23 states are already recourse in both judicial and nonjudicial foreclosure. Almost all second mortgages and home equity lines of credit are recourse as well. Expanding recourse status to the remaining mortgages would help balance the worry about persistent mortgaged debt loads against the prospect of being able to legally discharge those debts through bankruptcy. At the same time, moving to a new national standard or contract like this could help mortgage lenders and servicers apply more consistent processes. Evidence from Canada and Spain — where all mortgages are recourse — suggests that legal recourse can play a large role in lowering mortgage default rates even in the face of large price declines. According to one study by economists Andrea Ghent and Marianna Kudlyak, the same is true for American states with recourse status.
Meanwhile, the higher cost of credit generated by cramdown would be weighed against the lower credit cost resulting from making future mortgages consistently recourse. It’s worth noting that nonowner-occupied homes (e.g., second homes or vacation properties) are already eligible for cramdown. Expanding cramdown gradually to primary residences would grant households a legal mechanism for dissolving excess mortgage debts, thus potentially allowing an effective mechanism for mortgage renegotiation that could ease the process of debt deleveraging.
Though there has been near universal frustration with housing policy over the past few years, it is important for policymakers to remember that residential real estate will continue to be an important area of the economy going forward. Despite the crisis, homeownership also remains a central part of the American dream and a key aspiration for families. It is time for policymakers to shift some of their focus and energy from devising the next near-term plan to advancing a long-term legislative framework that can make the private housing market work again.