Morris A. Davis, Stephen Malpezzi and François Ortalo-Magné
MADISON, WI - October 2, 2009

Executive Summary

The foreclosure problem in the United States is getting worse:  From July, 2008 to July, 2009 foreclosure filings increased by 32 percent. 

Research by economists at the Boston Fed has shown that two events typically lead homeowners to default on their mortgage, hence the so-called “double trigger” theory of foreclosure. First, the value of the house must be less than the value of the mortgage (“under water”). Second, homeowners must experience a significant disruption and loss of income.

Our view is that the double-trigger theory of foreclosures applies in the current environment, and the available data suggest there is an increase in foreclosures in the immediate future.  Trigger 1:  Zillow estimates that 22 percent of the 50 million homeowners with mortgages are currently under water.  Trigger 2:  Unemployment rates are high.  There are currently about 14.9 million unemployed people, representing 9.7 percent of the labor force.  The Congressional Budget Office forecasts that the unemployment rate will rise to 10.2 percent in 2010 and will fall to a relatively high 9.1 percent in 2011. 

The rapid decline in the value of housing and surge in the rate of unemployment has created the possibility of a “perfect storm” for foreclosures in the next few months.   Precise estimates are not possible since the current crisis is so different from previous experience.  Our best, albeit rough, estimate is that if unemployment hits 10 percent, we could face something like 1.4 million additional defaults from this cohort of unemployed, of whom 600,000 would be receiving unemployment insurance.

The Obama Administration’s proposal to reduce foreclosures by modifying mortgage payments helps people with jobs whose mortgage payments are now consuming large fractions of their income.  The proposal doesn't directly address the needs of recently unemployed workers with incomes that have taken an immediate and really drastic hit.  Further, mortgage modifications are complex to implement – perhaps impossible to implement in some circumstances – as in many cases, the modifications require agreement from the investors that own the mortgages. Our view is that the Administration’s current plan is incomplete, will take time, is complex to implement, doesn’t provide enough help given the scale of the problem, and misses a large fraction of those truly in need. We need more, in particular we need to target the unemployed much more directly, and we need to do this fast.

We have in mind a simple expansion and combination of two established programs – unemployment insurance and housing vouchers – to prevent a new wave of foreclosures of recently unemployed workers. Our proposal, the “Wisconsin Foreclosure and Unemployment Relief plan” (WI-FUR), developed in the Graaskamp Center for Real Estate at the Wisconsin School of Business, provides a virtuous simplicity at a time when the need for expediency far outweighs the complexity of implementing a more fine-tuned loan plan. Rather than reinventing the wheel this vehicle recognizes the severe variability of the recent downturn and offers a temporary, targeted and timely solution.

If implemented quickly, we believe the WI-FUR plan will prevent at least 600,000 households from entering foreclosure.

The essence of the WI-FUR plan is to temporarily attach a housing “voucher” to unemployment insurance. We propose that this extra payment be provided along with every unemployment check until we're out of the recession. The housing voucher could be directly applied to a mortgage payment. The amount of the voucher would be based on “Fair Market Rent” or FMR, which define baseline housing costs for each county in the United States. The FMR data are already computed by the U.S. Department of Housing and Urban Development. The voucher supplement would be temporary, with a defined sunset date. The tacking-on of a housing voucher to existing unemployment insurance provides a turnkey platform from which to address the immediate and temporary needs of households facing a shock in income.

The reason we believe that a supplement to unemployment insurance is necessary to prevent additional foreclosures is that many recipients of unemployment will have trouble making mortgage payments with their normal unemployment insurance (UI) check.  In Wisconsin, for example, UI is capped at $363 per week ($1,452 per month).  The average mortgage payment (including first and second mortgages and property taxes) in Wisconsin is approximately $1,200.  This means $300 of the weekly UI payment would be devoted to the mortgage, leaving roughly $63 per week available for food, transportation, property taxes, and other necessities.

We suggest using “Housing choice vouchers” as the source of this additional support because these vouchers are an established and important part of the national safety net.  The U.S. Department of Housing and Urban Development (HUD) oversees a system of regular allowances, sort of like food stamps, distributed through local Public Housing Authorities, which help about 2.1 million low-income homeowners pay their housing costs.  But one big difference from food stamps is that housing vouchers are not an entitlement; only about 30 percent of poor households receive any such housing assistance, according to one recent estimate.  Local Public Housing Authorities often have long waiting lists for vouchers; the vast majority of recently unemployed won’t have much of a chance of accessing such assistance.

So what is the cost of the WI-FUR plan?  At the end of the day, the costs and benefits of supplementing UI insurance with a housing voucher are determined by key inputs:  (1) Who receives the voucher; (2) What is the size of the voucher; (3) For how many months will recipients receive vouchers?  Ultimately, the choice of these three parameters determines both the costs and effectiveness of the plan at reducing foreclosures.

Starting with the first point:  Currently, most HUD vouchers are devoted to renters, and only a very few owners receive such vouchers.  Obviously, this must change.  If we wanted to keep the proposal as simple as possible, we could just give a voucher to everybody collecting UI, regardless of whether they owned or rented.  Alternatively, the vouchers could exclusively go to homeowners or even to homeowners with a mortgage, as we discuss below. 

Second, we propose that the amount of the voucher be based on “Fair Market Rent” or FMR.  Fair Market Rent is an estimate of a baseline housing cost that HUD collects for each county in the United States.  This provides a quick way to send households located in more expensive areas higher voucher payments.

Third, we propose that the voucher program should be temporary in nature.  The program should be in place until the unemployment rate returns to more normal levels. 

To illustrate how the WI-FUR plan would work, we consider the case of Wisconsin.  As mentioned, the average mortgage payment in Wisconsin is $1,200 per month and maximum UI benefits are $1,452 per month.  The WI-FUR plan calls for households, on average, to spend 30% of their UI benefits towards their mortgage, equal to $436 in Wisconsin (= 0.30 x $1,452).  This leaves an average shortfall of $764 (= $1,200 - $436).  To make up for this shortfall, on average households would receive a housing voucher for $764.  This voucher would be equal to 81% of the average FMR of $943 in the state ($764 = 0.81 x $943).  Thus, the WI-FUR plan calls for all people in Wisconsin collecting UI benefits to receive a housing voucher equal to 0.81 times the Wisconsin county-level FMR.  Since the FMR varies across counties in Wisconsin, the housing voucher would also vary across counties in Wisconsin.  Thus, the WI-FUR plan calls for people living in higher-cost areas of the state to receive a higher voucher (to reflect larger mortgage payments in those counties).  For example, the FMR in Dane County, WI is $1,135 while the FMR in Manitowoc County, WI is $712.  Assuming all unemployed receive a voucher for 0.81 times the FMR in their county, the unemployed living in Dane county would receive a housing voucher for $919 (= 0.81 x $1,135) and the unemployed living in Manitowoc County would receive a voucher equal to $577 (= 0.81 x $712). 

The “FMR factor,” 0.81 in the case of Wisconsin, would be allowed to vary across states such that on average in each state households spend 30 percent of their UI benefits towards their mortgage.  Our computations suggest that at the median state, the appropriate FMR factor is 0.69.  This masks significant variation across states.  For example, we compute the appropriate FMR factor for West Virginia and Virginia is 0.34 and 0.91, respectively.

Ultimately, the cost of the WI-FUR program depends on the size of the vouchers.  If vouchers are computed such that, on average, households contribute 30% of their UI payments towards their mortgage, and the maximum monthly voucher payment is capped at $1,500, then we estimate the WI-FUR program (had it been in place at the time) would have cost $2.5 billion in August, 2009.  Assuming the number of unemployed remains roughly constant over the next year, this would translate to an annualized cost of $30 billion per year.  The true social costs of the WI-FUR program are much less than simple estimates of cash disbursements, because the proposal prevents foreclosures and the foreclosure process is costly in terms of legal fees, time spent in the court system, possible degradation of housing units during the foreclosure process due to lack of routine maintenance, and social and educational costs to young children living in households forced to move neighborhoods.

Although we believe the main benefit of a voucher program that supplements unemployment benefits is its administrative simplicity, one problem that we can immediately identify is that less than half the measured unemployed currently collect unemployment insurance.  Some unemployed don’t receive UI because their benefits have expired, some because their particular employment is not covered, or some because they haven’t had any recent employment history but are trying to enter the labor force.  If housing vouchers are mailed to these other unemployed, we would prevent more defaults and foreclosures, but we would also have to spend much more, and we would have to set up some system to identify and process checks for unemployed outside the UI system.

Whatever the details of its design, the WI-FUR plan could be paid for in a number of ways: using totally new money; shifting funds from some other, less effective proposal in pending legislation; or, by rescission, i.e. cancelling one or more programs passed in the haste of putting together the recent stimulus package that aren’t likely to address our real and emerging needs. We also note that the WI-FUR vouchers we propose should be considered as separate from the "normal" housing voucher program that HUD administers.

There are some other recent foreclosure relief proposals which are similar in spirit to WI-FUR; namely, they aim to get cash into the hands of the unemployed, quickly. A good example is the plan put forward by a team of Economists at the Boston Fed and the Federal Reserve Board for a program of loans or grants targeted to the unemployed.   We would gladly endorse their plan if it went ahead; either plan would be an improvement over the present situation, since they both get cash into the hands of the unemployed.  The WI-FUR proposal is similar to the Fed proposal in many respects, but the key advantage of WI-FUR is its simplicity. The unemployment insurance system already exists and we already have plenty of precedent for the use of housing vouchers. Administrative costs for voucher programs are low.

We conclude with this:  The WI-FUR voucher program provides ready money in hand that, along with the rest of the standard unemployment insurance benefit, would help more families under stress make their mortgage payment. It would save families from having to choose between their mortgage and their next meal. It is the kind of quick, actionable response that is called for in the face of grave risk.


Footnotes

i There is a long-running debate on the relative merits of “supply-side” housing programs and “demand-side” programs.  Supply-side subsidies (a.k.a. “bricks and mortar programs”) range from old-style public housing to the current Section 42 Low Income Tax Credit program, as well as many elements of recent proposals for a National Housing Trust Fund.  Demand-side subsidies provide cash or allowances for housing supplied in the private market, and are best represented by HUD’s Housing Choice Voucher program.  Most – not all – housing economists broadly favor demand side subsidies, based on their generally higher efficiency.  See Richard Green and Stephen Malpezzi, A Primer on U.S. Housing Markets and Policy, Urban Institute Press for the American Real Estate and Urban Economics Association, 2003.  There are many suggestions for improving the existing "normal" housing voucher program, many made recently by Professor Edgar Olsen of the University of Virginia. See Crews-Cutts, Amy and Edgar O. Olsen.  Are Section 8 Housing Subsidies Too High?  Journal of Housing Economics, 11(3), September 2002, pp. 214-43.  Olsen, Edgar O.  Achieving Fundamental Housing Policy Reform.  In Alan Gerber and Eric Patashnik (eds.), Promoting the General Welfare: American Democracy and the Political Economy of Government Performance.  Brookings Institution, 2006.   Olsen, Edgar O.  Promoting Homeownership among Low-Income Households.  Urban Institute, Opportunity and Ownership Project Report No. 2, 2007. 

ii Chris Foote, Jeff Fuhrer, Eileen Mauskopf, and Paul Willen. A Proposal to Help Distressed Homeowners: A Government Payment-Sharing Plan. Memo, revised draft February 10, 2009. Available here.