Electronic copy available at: http://ssrn.com/abstract=1930686

Too Little, Too Late and Too Timid:

The Federal Response to the Foreclosure Crisis at the 5-year Mark

A Working Paper

Dan Immergluck

Professor School of City and Regional Planning

Georgia Institute of Technology Atlanta, GA 30332

dan.immergluck@coa.gatech.edu www.prism.gatech.edu/~di17

September 19, 2011

Acknowledgements

I want to thank Frank Alexander, Kevin Byers, Sarah Greenberg, Alan Mallach, Ira

Rheingold, and Alan M. White for extremely helpful comments on an earlier version of this paper.

Any errors in the paper—as well as all opinions—remain my responsibility.

Electronic copy available at: http://ssrn.com/abstract=1930686

Abstract

The primary federal policy responses to the foreclosure crisis have included programs to

reduce foreclosures as well as efforts to mitigate the impacts of foreclosures on neighborhoods and

communities. This paper reviews policy responses since 2007 in both of these categories. While

there is less information at this point on the outcomes of mitigation polices, the overall federal

response is found to be lacking thus far. At least in comparison to the magnitude of the challenges,

the programs have been modest in scale. They have also suffered from significant problems of

design and implementation. Foreclosure prevention efforts, in particular, are faulted for being too

reliant on marginal incentive payments, for failing to include a key policy (bankruptcy

modification) that would have encouraged lenders to modify loans more aggressively, and for not

sanctioning servicers more aggressively for poor performance or noncompliance. The federal

response is also characterized as moving too slowly in some cases and being too captive to the

structures and policy preferences of the mortgage and financial services industry.

1

As the U.S. approaches the end of the fifth year of the national foreclosure crisis, it is

important to examine the federal policy responses to the crisis. After outlining the major federal

interventions, this paper pulls out some of the key, salient information on the programs and the

evidence thus far on their scale and progress to date. It also describes some of the principal

complaints that have been lodged against the programs and the design and implementation flaws

that appear to have limited their impact.

The paper breaks out federal responses to the crisis into two basic categories: those focused

on preventing or reducing foreclosures and those focused on mitigating impacts of foreclosures on

neighborhoods where they are concentrated. The first set of responses includes major initiatives

under President George W. Bush and President Obama. These initiatives include the creation of the

National Foreclosure Mitigation Counseling program beginning in late 2007; the facilitation of the

Hope Now Alliance in 2007 and the related modification programs spearheaded together with the

American Securitization Forum; and the Making Home Affordable (MHA) programs (including the

Home Affordable Modification program and the Home Affordable Refinance Program, and their

ancillary programs) which began in 2009. The second category of responses, which focuses on

attempts to mitigate the impacts of foreclosures on local communities, includes the three phases of

the Neighborhood Stabilization Program (NSP), NSP 1, NSP 2, and NSP 3.

Overall, whether from liberal or conservative corners, most observers have been generally

critical of the federal responses to the foreclosure crisis both under the Bush and Obama

Administrations, especially some of the programs focused on loan modification. Although some of

this criticism stems partly from unreasonable expectations (which were first established by the

agencies introducing the programs), this paper argues that with substantial programmatic revisions,

one critical complementary change in bankruptcy law, and more forceful execution, such efforts

could have been more effective at reducing foreclosures. Compared to the foreclosure prevention

programs, it will take much longer to gauge the impact of the NSP programs. However, based on

implementation thus far, particularly of NSP 1, some design and implementation issues have

surfaced. Regardless of how effective the NSP programs are or are not, it is certainly fair to say that

$7 billion was a modest sum given the scale of the foreclosed property problem.

2

The focus here is on proactive, direct policies to reduce foreclosures and mitigate their

impacts on households and neighborhoods and not on other, usually longer term, responses to the

foreclosure crisis and broader problems with mortgage and housing markets that were entangled

with the foreclosure crisis. For example, federal involvement in responding to problems of

fraudulent documentation and processing of foreclosures by servicers that gained national attention

in the summer and fall of 2010, while closely related to the foreclosure crisis, are not a principal

focus here. Ongoing efforts to reform and restructure mortgage markets over the long term are also

beyond the scope of this article.

Figure 1 describes the key federal responses to the U.S. foreclosure crisis. It also indicates

some events that may have had particular influence on the crisis and the policy response. It does this

against a plot of the share of mortgages entering the foreclosure process during the period. The scale

of the foreclosure crisis was vast. Because there were on the order of 55 to 60 million mortgages

outstanding at the height of the housing boom, a figure of five percent of outstanding mortgages

entering foreclosure in a year, a rate reached at the beginning of 2009, amounts to almost 3 million

foreclosure starts a year. Later in this paper, I estimate that, from March 2009, when the Obama-era

MHA program was announced, through June 2011, there were more than 5.3 million foreclosure

starts on owner-occupied homes. While not all of these resulted in lost homes, this figure is a good

indicator of the number of households in severe financial distress and at severe risk of losing their

homes.

Foreclosure Prevention

Bush-Era Foreclosure Prevention Policy

While there is no clear, consensus starting date of the U.S. foreclosure crisis, many would

point to late 2006 or early 2007 when foreclosures began to grow rapidly in many regions that had

previously seen low foreclosure rates, especially in the “sand states” of Arizona, California, Florida,

and Nevada. One of the largest subprime lenders, New Century, went bankrupt in the spring of

2007. In late spring and early summer, Federal Reserve Board Chairman Bernanke and U.S.

Department of Housing and Urban Development (HUD) Secretary Alonzo Jackson called for

federal funding for foreclosure prevention counseling (Joint Economic Committee, 2007).

3

Figure 1. Federal Policies Responding to the Foreclosure Crisis

0

1

2

3

4

5

6

Perc

ent o

f Out

stan

ding

Mor

tgag

es En

terin

g For

eclo

sure

(A

nnua

lized

)

2MPbegins

HERA:NSP 1 &

adopted

FHA SECURE

begins

HAFA begins

1stRound HHF

announced

Dodd-Frank:NSP 3 &

adoptedEHLP

H4H

ARRA:NSP 2

adopted

GSEs put into conservatorship

NFMC begins HAMP &

HARPbegins

EESA:TARP

adopted

UP begins

Lehman & AIG collapse

Subprime foreclosures surge in formerly hot markets

New Centurycollapses

robosigning and foreclosure

process scandals

Fed begins MBS

purchases

FHA short refi begins

PRA begins

Note: Lightly shaded boxes indicate foreclosure prevention policies; darker boxes are mitigation policies. Dashed, white boxes are external events; dashed, shaded boxes are policies likely to have significant indirect effects on foreclosures. ARRA= American Recovery and Reinvestment Act; HERA=Housing and Economic Recovery Act; FHA= Federal Housing Administration; NSP=Neighborhood Stabilization Program; 2MP= Second Mortgage Modification Program; HARP=Housing Affordable Refinance Program; HHF= Hardest Hit Fund; EHLP=Emergency Home Loan Program; PRA=Principal Reduction Alternative; HAFA=Housing Affordable Foreclosure Alternatives; GSE=Government sponsored enterprises; UP= Home Affordable Unemployment Program; Dodd-Frank = Dodd-Frank Wall Street Reform and Consumer Protection Act.

4

In August of 2007, the Federal Housing Administration (FHA) announced its FHA Secure

program, in which it would refinance delinquent homeowners into more affordable loans in order to

prevent foreclosures. While overall FHA refinances surged in 2008, due in part to the decline in

credit availability from other lenders, by April of 2008, fewer than 1,800 borrowers with delinquent

mortgages were served through FHA Secure (Swarns, 2008).

In October, as the subprime crisis was continuing to make national headlines, Senator

Richard Durbin (D-IL) introduced the Helping Families Save Their Homes in Bankruptcy Act (U.S.

Senate, 2007). The bill would have allowed bankruptcy judges to modify the balance on owner-

occupied home loans. When borrowers file for bankruptcy under Chapter 13, the court has the

ability to modify certain debts, including “cramming down” or reducing the balances owed to the

fair market value of the property or other collateral securing the loan. However, this authority does

not extend to loans secured by owner-occupied residences. Bankruptcy judges can modify the

balance due on a vacation home or an investment property, but not on a primary residence. The

Durbin bill would have removed the primary residence exclusion on a temporary basis, which

would have provided relief for those filing bankruptcy, and would have given investors in

mortgage-backed securities (MBS) and servicers an incentive to modify loans voluntarily before the

borrower might file bankruptcy. Opponents to the Durbin bill, including much of the lending and

securities industries as well as the Bush Administration, argued that the proposal would raise

mortgage rates substantially, but Levitin and Goodman (2008) estimated that mortgage cram down

would result in an increase of only a 0.05 to 0.15 percentage points in mortgage rates. Industry

lobbyists were successful in blocking the bill.

The Bush administration, in part to offer an alternative to the Durbin bill, announced the

Hope Now Alliance in October of 2007 (Hope Now Alliance, 2007). The Alliance included banks,

lender and investor trade associations, the NeighborWorks network, and other organizations. Hope

Now offered foreclosure prevention counseling to homeowners via a 1-800 number. Again

rejecting calls for stronger interventions, the Bush Administration also announced an effort to

promote “streamlined” but voluntary loan modifications for a subset of subprime mortgages

(American Securitization Forum, 2007). This proposal was developed primarily by the American

Securitization Forum (ASF), a trade group of structured finance investors. The plan was criticized

by many consumer advocates in part because it was entirely voluntary on the part of servicers and

investors and because it carved out a narrow segment of at-risk borrowers as eligible for assistance

5

(Said and Zito, 2007). Estimates of the proportion of subprime borrowers that would fit the

program’s eligibility requirements fell in the range of 3 to 12 percent (Immergluck, 2009).

In December, the National Foreclosure Mitigation Counseling program began. Over five

rounds of funding through 2011, NFMC provided over $500 million to local housing counseling

organizations to counsel homeowners in or at risk of foreclosure. NFMC counselors became key

advocates for borrowers applying for loan modifications during both Bush-era and Obama-era loan

modification initiatives. As of early 2010, NFMC had supported the counseling of over 1 million

homeowners. A congressionally mandated evaluation of the program found that it had significantly

reduced foreclosures among counseled homeowners.1

By the summer of 2008, with the economy faltering and a national election approaching,

there was more political pressure to do something about the foreclosure problem. Due to losses that

they were taking on some of their investments in subprime mortgage-backed securities, it was

becoming increasingly apparent that the government-sponsored enterprises (GSEs), Fannie Mae and

Freddie Mac, would need federal assistance. The stock price of the GSEs had been falling, leaving

the firms even more highly leveraged than they had been.

In July 2008, the Housing and Economic Recovery Act (HERA) was enacted (U.S. House of

Representatives, 2008). HERA was a complex piece of legislation that created a new regulator for

the GSEs that could, if necessary, place the GSEs in “conservatorship,” essentially placing them

under direct government control (Weiss et al., 2008). It also contained tax breaks for residential

builders, a complicated first-time homebuyers tax credit (which was later simplified), and funding

for local government and nonprofit acquisition of foreclosed properties, later called the

“Neighborhood Stabilization Program,” which is discussed below.

The principal foreclosure prevention program in HERA was the $300 billion Hope for

Homeowners (H4H) program, run by the FHA, which was intended to refinance distressed

borrowers out of unaffordable mortgages and into smaller, more affordable, fixed rate loans. The

program also required borrowers to share at least 50 percent of any equity gain with the FHA when

they sold the homes. The program did not address the prevalence of junior mortgages among

distressed borrowers, especially in high-cost housing markets. Junior lenders can hold up the

refinancing process. H4H failed to get any traction. Even after HUD eased the requirements for

lenders and borrowers and tried to make it easier to compensate junior lenders for releasing their

6

liens, only 340 loans were made through the H4H program in fiscal year 2010 (U.S. Department of

Housing and Urban Development, 2011a).

After HERA was passed in July 2008, housing prices continued to fall rapidly in many

regions, and foreclosures continued to increase and began to spread to the near-prime and prime

sectors. By early September, Treasury Secretary Paulson announced that he would provide

financing to the GSEs and direct the Federal Housing Finance Agency (FHFA), the new GSE

regulator, to place both firms in conservatorship, wiping out GSE stockholders and giving the

FHFA substantial operational control of both companies. After further tumult in U.S. and global

financial markets, including the fall of Lehman Brothers and near collapse of the insurance firm

AIG, Treasury Secretary Paulson introduced a $700 billion proposal to establish the Troubled Asset

Relief Program (TARP). The proposal had been prepared earlier for the eventuality that the crisis

might become significantly worse (Nocera, 2008). In early October, after significant Congressional

gymnastics, the Emergency Economic Stabilization Act (EESA), which authorized TARP, became

law. TARP was originally portrayed primarily as an effort to purchase mortgages or mortgage-

backed securities from financial institutions in order to rid them of bad assets. By mid-October,

however, after the United Kingdom initiated its own equity injections into banks, the Treasury

Department directed TARP funds to be used to buy preferred stock in mostly larger financial

institutions. The Bush Administration decided against using any TARP funds to assist struggling

homeowners directly, ignoring provisions in the bill giving it the authority to do so.

In late 2008, even before taking office, the Obama Administration sought to draw down the

second half of TARP funds. Larry Summers, the incoming director of the National Economic

Council, wrote a letter to Congress that the new administration would use $50-100 billion of TARP

funds for foreclosure mitigation (Summers, 2009). The letter also suggested that the new

administration would seek to change bankruptcy laws to permit cram-downs of primary residence

loans.

In January 2009, the Federal Reserve Board began a major program of purchasing

mortgage-backed securities issued by Fannie Mae and Freddie Mac. During the apex of the

financial crisis in the fall of 2008, the market for some mortgage-related securities had essentially

shut down. Moreover, while the conservatorship of the GSEs increased the strength of the GSE

guarantee to investors in the agencies’ MBS, the Board’s decision to purchase what eventually

amounted to $1.25 trillion in MBS from the GSEs lowered mortgage rates (Gagnon et al., 2011;

7

Hancock and Passmore, 2010). While the MBS purchase program was not aimed specifically at

reducing foreclosures, but more broadly at supporting liquidity and low interest rates, and could

hardly be called a foreclosure prevention program, it likely had some dampening effects on

foreclosures. Keeping mortgage rates low also meant that many adjustable rate loans did not reset

to rates as high as they would have otherwise. If rates had risen substantially in late 2008 and 2009,

foreclosures might have risen to even higher levels. The chief beneficiaries of the Fed’s MBS

purchases, however, were almost certainly the millions of middle- and upper-income homebuyers

and homeowners who were able to save hundreds of dollars per month on their mortgage payments,

most of whom were not at risk of foreclosure. The blunt tools of MBS purchases and low interest

rate policies could hardly be called efficient, equitable, or well-targeted approaches to moderating

foreclosures, and they were certainly not able to stem the rise in foreclosures in 2009. However,

they most likely had a nontrivial impact on foreclosure rates.

Obama-Era Foreclosure Prevention Policy

One of the earliest domestic policy priorities of the new Obama Administration was to

create a more forceful foreclosure prevention initiative. The Bush-era Hope Now initiative had been

widely criticized as inadequate and ineffective, and the ASF/Hope Now modification initiative was

responsible for only approximately 9 percent of loan modifications in the first six months of 2008, a

figure which dropped to 2 percent in the latter half of the year (Fitch Ratings, 2009). Moreover, less

than 7 percent of modifications during the second half of 2008 involved principal reductions. In part

because many modifications did not lower mortgage payments, yet alone the outstanding principal,

significantly during this period, the portion that redefaulted within just six months ranged from 34

to 51 percent (Agarwal et al. 2010).2 While loan modifications resulting in reduced mortgage

payments grew in 2008, many interventions still focused on very short-term forbearance, which

reduced payments for as little as one month (often to see them increase afterwards beyond their

original levels), or did not reduce payments significantly.

One important influence on the Obama Administration’s approach to modifications was the

Federal Deposit Insurance Corporation’s initiative with the failed thrift, Indymac, which it had

taken over. FDIC Chairman Sheila Bair had been skeptical of the unstandardized and marginal

approaches to loan modifications under Hope Now. In 2008, the FDIC established an approach that

8

would unilaterally offer loan modifications to borrowers with debt-to-income ratios above 31

percent through a standardized protocol (Federal Deposit Insurance Corporation, 2008).

In March 2009, the Obama Administration introduced the Making Home Affordable (MHA)

initiative with a goal of assisting 7 to 9 million homeowners (U.S. Department of the Treasury,

2009). Initially, MHA consisted of two key components. The Home Affordable Refinancing

Program (HARP) provided for expanded refinancing capabilities for borrowers through the GSEs.

The GSEs have traditionally not made loans above 80 percent of the value of the home unless

private mortgage insurance insured the value of the loan above this level. Because an increasing

number of homeowners were underwater (with home values less than outstanding loan amounts),

they were unable to refinance their loans during a time of historically low interest rates. Such

refinancing could reduce loan payments, thereby reducing the foreclosure risk of such loans. Under

HARP, the maximum loan-to-value limit for refinance loans increased to 105 percent, and a few

months after the program was announced that limit was raised to 125 percent. The Administration

suggested that HARP would result in the refinancing of 4 to 5 million homeowners into lower-cost

loans (U.S. Department of the Treasury, 2009).

The other, and higher-profile, component of MHA was the Home Affordable Modification

Program (HAMP). In HAMP, which was to be funded by TARP funds, borrowers at imminent risk

of foreclosure would apply to have their loan modified to result in long-term reductions in mortgage

payments. HAMP is a complex program in which mortgages are modified in two basic steps. First,

lenders/investors are responsible for lowering loan monthly mortgage payments to the point where

they are equal to 38 percent of monthly gross income. Then, the costs of reducing the loan payments

further to 31 percent of monthly income are shared between the lenders/investors and the federal

government.3 Under the traditional HAMP program, servicers are not required to, but may, reduce

principal owed. Another MHA program described below, the Principal Reduction Alternative

(PRA), was introduced later to encourage such activity.

After meeting other eligibility requirements, HAMP applications are evaluated by using a

standardized “Net Present Value” (NPV) test that compares the net present value to the

lender/investor of modifying the loan to the status quo of no modification. The NPV test uses

specified inputs to compare the two alternative cash flows and discounts them to current dollars. If

the NPV of the cash flow expected from modifying a loan is greater than the NPV of the cash flow

expected from foreclosing then the servicer must modify the loan. The logic of the NPV test is that

9

servicers should be modifying loans at least when it makes sense from the perspective of the

lender/investor. Loan modification programs, while billed in the media as focused on aiding

distressed borrowers, are also a component of a broader strategy of loss mitigation for lenders. A

key argument in favor of loan modifications is that lenders/investors often lose more on a

foreclosure than they would if the loan were appropriately modified (White, 2009). When servicers

foreclose, lender/investors generally recover well under half of the loan amount, with loss severity

ratios typically running over 60 percent for subprime loans and over 50 percent for prime loans.4

It is important to understand here that the NPV test is conducted solely from the perspective

of the lender/investor and not that of the borrower or society. Moreover, since the servicer has some

ability to influence of the implementation of the NPV analysis, if the servicer does not benefit from

the modification process, then it may have a vested interest in skewing the results towards a “do not

modify” result.

All HAMP borrowers are placed in “trial” modifications for a period intended to last three

months, although in practice, almost twenty percent of trial modifications lasted longer than six

months (SIGTARP, 2011). During the trial period, HAMP borrowers must make at least three

payments at the new level. After transitioning to “permanent” modifications, the terms of the loan

remain fixed for at least five years, after which interest rates may increase at 1 percent per year back

up to market rate if the borrower received a below-market interest rate during the modification

period.

A fundamental proposition behind HAMP was that, with small incentive payments to

servicers and a standardized protocol for evaluating and implementing modifications, resistance by

servicers and lender/investors to modifications would be diminished. The design of the program

rested fundamentally on the notion that a more rational set of procedures and incentives could be

layered on the existing loan servicing system, and that the industry would adapt quickly and

respond well to this overlay. However, HAMP did not fundamentally alter the locus of decision-

making or address some fundamental barriers and blockages to modifying massive numbers of

distressed borrowers in a quick and fair way.

Another fundamental feature of HAMP, one that followed from the Bush-era Hope Now

efforts, was that it did not impose strong modification requirements on servicers. Even when some

steps in the process supposedly triggered mandatory modification, there were many ways that the

complex HAMP procedures would allow servicers or lender/investors to limit loan modifications.

10

Moreover, many key fundamental inputs or assessments in the HAMP modification process were

under the control of the servicers. Figure 2 illustrates some of the complexity of the HAMP process.

In introducing HAMP, the Administration suggested that it would be complemented by

adopting the bankruptcy mortgage modification provisions similar to those proposed in the 2007

Durbin bill (U.S. Department of the Treasury, 2009). Such a bill would provide the “stick” needed

to get lenders to the modification table, while the incentives to servicers in HAMP would provide

the carrot. A bankruptcy mortgage modification bill passed in the House, but failed in the Senate.

Rather than pushing hard for the bill, the Obama Administration argued that HAMP could reach its

goals without bankruptcy modification (Lillis, 2009). As a result, Obama-era foreclosure

prevention strategy, while more focused, ambitious and structured than the Bush-era Hope Now

initiative, relied on the carrots of HAMP’s small incentive payments without the stick of bankruptcy

modification.

The Obama Administration projected that the MHA programs would help seven to nine

million homeowners, with HAMP assisting three to four million of those (U.S. Department of the

Treasury, 2009). These ambitious and widely publicized goals would later become a significant

problem for the Administration.

The MHA initiative was soon expanded by adding programs aimed at a variety of barriers to

foreclosure prevention and at creating opportunities for “soft landings,” using short sales, deeds-in-

lieu of foreclosure, and relocation assistance, when keeping a homeowner in a home does not

appear to be feasible. 5 One of these programs was the second lien modification program (referred to

as “2MP”), which focused on eliminating or modifying junior mortgages. When a borrower’s first-

lien mortgage was modified under HAMP, servicers participating in 2MP had to offer to modify the

borrower’s junior mortgage. As in HAMP, servicers followed specified protocols that resulted in

either extinguishing the junior mortgage or reducing it partially and modifying the terms of the

remaining portion. As in HAMP, servicers received incentive payments to execute 2MP

modifications. Although the program was announced in the spring of 2009, it did not begin

operating until the spring of 2010 (SIGTARP, 2011). Moreover, despite the large percentage of

junior mortgages present in the housing market, fewer than 35,000 2MP transactions occurred in the

15 months after its implementation.

11

Figure 2. HAMP Process Flow Chart

Source: U.S. Department of the Treasury, 2010.

YES

NPV eligible according to

investor guidelines

YES

NO

Borrower notification

Trial default

NO

Borrower makes trial payments

Offer permanent modification

GSE imminent default

indicator

Borrower 60 days

delinquent

Borrower requests HAMP

Imminent default

evaluation

Trial not accepted

Borrower does not make first payment

Offer trial modification

Net Present Value (NPV)

Borrower evaluation (waterfall)

HAMP eligibility

evaluation

NPV not eligible

according to investor

guidelines

Notification, explore other

loan mod options

12

One problem that HAMP did not face head on – at least until the latter part of 2010 -- was

that of severely underwater borrowers. Without realigning loan balances with property values,

borrowers in severe negative equity situations would have less motivation to maintain ownership of

the house, especially if it meant defaulting on other debts or severe strains on household finances.

This was, in fact, why allowing bankruptcy judges to modify loan balances would have provided a

strong complement to any large-scale loan modification effort. Without the threat of bankruptcy

modification, servicers and investors had too limited incentive to make substantive and sustainable

loan modifications in large enough numbers to put a large dent in foreclosure volumes. The

Treasury Department’s response to this problem was the Principal Reduction Alternative (PRA)

program, which did not get going until late in 2010 (SIGTARP, 2011). PRA gave servicers

additional incentives beyond those in HAMP to reduce principal; it also gave incentives to loan

investors. However, PRA does not require servicers to reduce principal, even when doing so is

judged to provide greater benefits to investors than other forms of modification.

Another MHA foreclosure prevention program is the Home Affordable Foreclosure

Alternatives (HAFA) program. Rules for HAFA were issued in late 2009 and the program began in

the spring of 2010 (SIGTARP, 2011). HAFA was aimed at encouraging short sales and deeds-in-

lieu of foreclosure as a preferable alternative to foreclosure. In many states, borrowers engaging in

either of these types of transactions may still be personally liable for the unpaid balance on the loans

following the short sale or deed-in-lieu of foreclosure (even though this debt is now unsecured)

through what are called deficiency judgments. As a result, these transactions may be less

advantageous to borrowers than they might appear on their face. One component of HAFA is a

requirement that participating lenders/investors waive any rights to pursue deficiency judgments.

HAFA provides for incentives to servicers and borrowers to execute short sales or deeds-in-lieu of

foreclosure. Servicers are given a $1,500 incentive for executed HAFA transactions and

homeowners are given $3,000 in relocation assistance. However, many servicers and

lender/investors may be reluctant to give up the ability to pursue deficiency judgments after a short

sale or deed-in-lieu of foreclosure. Indeed, the ten largest servicers participating in HAMP reported

that, as of May 2011, they had completed over 112,000 short sales and deeds-in-lieu for HAMP

borrowers whose modifications had failed, more than ten times the volume of all HAFA

transactions through June 2011.

13

One TARP-funded initiative that takes a very different approach from HAMP and its

affiliated MHA programs is the Hardest Hit Fund (HHF), which provides funds to state housing

finance agencies (HFAs), which in turn design and implement detailed foreclosure prevention

programs in their states, subject to Treasury Department approval (Immergluck, 2010; SIGTARP,

2011). HHF was billed as an effort to promote innovative approaches to reducing foreclosures in

ways that could be more customized to regional housing market and economic conditions.

Beginning in February 2010, the HHF program allocated over $7.6 billion to HFAs via four rounds

of funding. The criteria for funding and the eligible uses of funds changed in each round. The first

round provided $1.5 billion to the five states (Arizona, California, Florida, Michigan, and Nevada)

that had seen home values drop by at least 20 percent from their peak. At least initially, the

Treasury Department appeared open to a wide variety of uses for the funds in this round, as long as

they were aimed at reducing foreclosures or mitigating their impact on households. In a second

round a couple of months later, $600 million more in funds were allocated to five more states. This

time states with high percentages of people living in high-unemployment counties were selected,

including North Carolina, Ohio, Oregon, Rhode Island and South Carolina. The eligible uses for

this round of funds were similar to those of the first round. In August 2010, a third round of HHF

funds, amounting to another $2 billion was allocated to states with unemployment rates above the

national rate. This amounted to 17 states plus Washington DC, including all of the previously

funded states except for Arizona. This third round of funds was directed toward programs to assist

the unemployed in paying their mortgages. Finally, just before the Treasury Department’s authority

to allocate TARP funds ended at the end of September, an additional $3.5 billion was allocated to

existing HHF states to be used in existing HHF programs.

By using a wholesale, block-grant-like approach, the HHF programs were intended to

provide for localized innovations in foreclosure prevention and loan modification programs.

However, even though EESA gave the Treasury Department broad discretion in designing

foreclosure prevention programs such as HAMP and HHF, the flexibility of the Treasury

Department in responding to the states' HHF proposals was, in practice, limited (Immergluck,

2010). For example, proposals to utilize HHF funds to support legal assistance to borrowers were

rejected. In addition, servicers and the GSEs resisted some HHF state plans to implement principal

reduction programs, even when such reductions would be subsidized one-for-one with HHF dollars.

14

As in the case of HAMP, the Treasury Department did little to push servicers or the GSEs to

participate in principal reduction programs.

As unemployment increased in 2009 and 2010, it became a more important driver of

foreclosures. In March of 2010, the Treasury Department introduced the Home Affordable

Unemployment Program, known as “UP” (SIGTARP, 2011). Under UP, HAMP borrowers can

receive partial forbearance of their mortgage payments for a period of 3 months. After hearing

complaints that a three-month forbearance period would not be of much use in an economy where

unemployment spells were averaging over six months in many parts of the country, the

Administration increased the period to 12 months. However, the GSEs have refused to participate

in the UP program, making it available only for borrowers with non-GSE loans.

The UP program was complemented both by HHF programs aimed at helping unemployed

pay their mortgages as well as a newer program aimed at unemployed borrowers in states not

funded via HHF. In the Dodd-Frank regulatory reform bill in the summer of 2010, Congress

allocated $1 billion to begin a mortgage-payment assistance program, the Emergency Homeowners

Loan Program (EHLP), aimed at helping unemployed homeowners pay their mortgages. EHLP was

designated for use only in the 32 states where HHF funds were not available. However, it took until

the summer of 2011 for the Administration to get the program launched. Moreover, the GSEs did

not agree to participate in the program until June of 2011 (Prior, 2011).

Examining the Results of Obama-Era Foreclosure Prevention Efforts Thus Far

Table 1 provides activity numbers for the key foreclosure prevention programs initiated

during the Obama Administration through the summer of 2011. Determining which borrowers

received substantive assistance is not a completely straightforward or objective task. For example,

over 1.6 million borrowers entered a HAMP trial, and the Treasury Department has frequently

argued that this is a reasonable measure of the number of households assisted because, even during

the trial period, households receive some assistance in the form of reduced mortgage payments.

Others, notably the Special Inspector General for the Troubled Asset Relief Program (SIGTARP),

have disagreed, arguing that only sustainable, permanent modifications should be considered

meaningful assistance. The SIGTARP has even argued that many borrowers applying to TARP and

either not qualifying for a trial or having their modification cancelled may actually have been

harmed by HAMP (SIGTARP, 2011).

15

Table 1. Borrowers Assisted through Obama-Era Federal Foreclosure Prevention Programs

Program Acronym Date Started Borrowers Assisted as of June 2011 (or date given) Home Affordable Modification Program HAMP Announced March 2009; begins April 2009 772,559 receiving assistance* Home Affordable Refinance Program HARP Announced March 2009; begins April 2009 810,084 completed refinances (as of May 2011) Second Mortgage Program 2MP Announced March 2009; begins in March 2010 33,715 loan extinguishments and modifications Home Affordable Foreclosure Alternatives HAFA Announced March 2009; begins April 2010 10,280 completed short sales or deeds-in-lieu of foreclosure Hardest Hit Fund HHF Announced from February 2010 to September 2010 2,343 receiving assistance (as of March, 2011) Home Affordable Unemployment Program UP Announced March 2010; begins July 2010 6,752 receiving assistance (as of May 2011) Emergency Homeowners Loan Program EHLP Adopted July 2010; begins June 2011 N/A, did not begin until summer 2011 Principal Reducation Alternative PRA Announced June 2010; begins October 2010 26,258 receiving assistance Federal Housing Administration Short Refinance FHA Short Refi Announced March 2010; begins September 2010 257 loans originated

*This includes 115,515 active trials 657,044 active permanent modifications. A total of 1,639,382 modifications had been started from April 2009 through June 2011, including 760,796 trials and106,024 permanent modifications that were cancelled.

Sources: U.S. Department of the Treasury (2009); SIGTARP (2011); U.S. Government Accountability Office (2011); Federal Housing Finance Agency (2011).

16

By the measures used here, which do not include households whose HAMP trial or

permanent modifications had been cancelled, the maximum number of households assisted by

Obama-era MHA programs (including more than 800,000 receiving HARP loans) was just over 1.6

million as of June, 2011. This figure is likely somewhat of an overestimate of assisted households,

because some of the MHA programs (2MP, PRA, etc.) are likely to have been used in conjunction

with HAMP, so that some households have been assisted by more than one program. Regardless of

the details, 1.6 million is not a trivial number of homeowners receiving assistance. Of course, some

of these households will still end up in foreclosure. Others, especially some HARP borrowers, were

never likely to end up in foreclosure in the first place.6

More importantly, from a political perspective, 1.6 million is a far cry from the goal of 7 to 9

million borrowers in the original MHA plan. Production in the HAMP program in particular was

continually judged in the media against the original goal of 3 to 4 million borrowers assisted.

HAMP assistance, as of June 2011, reached no more than 800,000 borrowers with sustainable,

permanent modifications as late as June 2011 (see Table 1). Moreover, a conservative estimate of

the number of borrowers who originally applied for HAMP modifications is well over 3 million,

including more than 1.5 million who were not accepted for HAMP trial modifications (U.S.

Department of the Treasury, 2011a). Of course, many of these borrowers likely should not have

been accepted into the program, as they may not have been in substantial mortgage distress (almost

400,000 of these remain current on their mortgages to date) or had little chance of being converted

to a permanent modification. Some also received non-HAMP modifications from their servicers,

although the benefits and sustainability of such modifications are unclear. Research on non-HAMP

modifications suggests that they have very high redefault rates (Agarwal et al., 2010). Nonetheless,

it remains the fact that from a base of about 3 million applicants, the program has to date produced

800,000 permanent modifications.

Some have suggested that the the widespread portrayal of HAMP as a “lackluster” effort is

partly the product of unwise, overly ambitious goal setting.7 A former White House staffer was

quoted as saying, “I wish the three to four million had never been uttered” (Powell and Martin,

2011). To move away from comparing HAMP activity against a somewhat arbitrary goal, it is

helpful to compare program activity to foreclosure starts since the program’s inception to gain some

sense of the program’s scale relative to the foreclosure problem.8 Figure 3 compares the level of

permanent, active HAMP modifications versus cumulative foreclosure starts on owner-occupied

17

Figure 3. HAMP Active Permanent Modifications Compared to U.S. Foreclosure Starts on Owner-Occupied Homes (Estimated)

0

1

2

3

4

5

6

Millio

ns of

Loan

s/Bor

rowe

rs

Estimated foreclosure starts, cumulative

Active, permanent mods, cumulative

Sources: U.S. Department of the Treasury (2011a): Mortgage Bankers Association National Delinquency Survey (2011); Federal Reserve Flow of Funds (2011).

18

homes since the program’s inception. The HAMP number here includes only active recipients of

permanent HAMP modifications, excluding borrowers falling out of the program in either failed

trial modifications or those becoming 90 days delinquent in permanent modifications. Foreclosure

starts are not a perfect measure of the potential demand for foreclosure prevention efforts, because

ideally such programs are aimed at assisting borrowers before the formal foreclosure process

begins. However, using all delinquent homeowners is an overly generous estimate of demand for

foreclosure prevention services, since some borrowers may go in and out of delinquency repeatedly

without seeking – or even needing -- such assistance. Therefore, foreclosure starts are used here as a

conservative estimate of demand for foreclosure prevention services. Figure 3 shows that

cumulative foreclosure starts on owner-occupied homes from March 2009 through June 2011 are

estimated to have exceeded 5.3 million.9 In this context, a figure of fewer than 800,000 HAMP

permanent modifications hardly seems impressive, yet neither does it seem entirely trivial, as it has

sometimes been portrayed in the media (Hiltzik, 2011; Powell and Martin, 2011).

Beyond the number of borrowers receiving permanent modifications, it is important to

consider what portion of these are likely to be sustainable over the longer term. Figure 4 examines

HAMP redefault rates over time after the permanent modification has begun. It shows that redefault

rates on permanent HAMP modifications declined over time and that, for loans modified in the

second year or later of the program, they are trending towards one-year redefault rates of

approximately 15 percent and 15-month rates of less than 20 percent (U.S. Department of the

Treasury, 2011a). Moreover, as might be expected, the nature of the modification is highly

correlated with redefault rates. One-year redefault rates on HAMP modifications where payments

dropped by 50 percent or more were less than 9 percent, compared to 26 percent for modifications

where payments decreased by less than 20 percent. Redefaults will likely continue to accumulate

well beyond the 15-month period, although the pace of accumulation should decline over time.

Overall, HAMP figures suggest that, of all the borrowers that begin trial modifications, well under

40 percent will likely remain current on their loans over the five-year HAMP period.

During the foreclosure crisis, the discussion of redefault rates in the media has sometimes

ignored the fact that HAMP was, by design, intended to serve a highly distressed set of borrowers,

households that were economically fragile. The level of overall debt that HAMP borrowers carry –

both before and after HAMP modifications – is very high. While HAMP borrowers experience

almost a 15 percent reduction in their back-end monthly debt-to-income ratios (the ratio of total

19

Figure 4. Redefault rates for Permanent HAMP Modifications

Source: U.S. Department of the Treasury, 2011a.

0%

5%

10%

15%

20%

25%

30%

3 6 9 12 15

Perc

ent o

f mod

ified

loan

s tha

t are

90+ d

ays d

elinq

uent

Months After Permanent Modification Begins

Q3 2009Q1 2010Q3 2010

Permanent Modifications Beginning

20

monthly debt payments, including not just mortgages but also credit cards, auto loans, etc., to their

monthly gross income), even after modification the median HAMP back-end ratio is 61.7 percent

(U.S. Department of the Treasury, 2011a). More than half of HAMP borrowers are still carrying

very high levels of overall debt even after modification. To expect such borrowers to have very low

redefault rates is unrealistic.

The back-end debt problem demonstrates all too clearly the importance of the failure to

adopt bankruptcy modification legislation. A key advantage of modifying mortgages in bankruptcy

is that the court could reduce other outstanding credit while it was reducing the mortgage principal,

thus having a much greater impact on back-end ratios and improving the sustainability of the loan

modifications.

Another aspect of redefaults that is infrequently discussed is that borrowers who are able to

remain in HAMP for even two to three years may be substantially better off than if they had entered

foreclosure earlier. During this period they are able to lower their mortgage payments substantially,

remain in their homes, have time to plan for a possible move (including finding a good next home,

schools, etc.). Thus, one benefit of loan modifications is that it can slow down what could otherwise

be a rapid, forced relocation, providing households more time to relocate and adjust to new

circumstances.

HAMP also arguably had a significant impact on reducing or slowing the foreclosure

pipeline. By reducing the flow of foreclosed properties into the housing market during a period of

rapid price depreciation in many places, and keeping homes occupied, some argued that HAMP had

a positive impact on stabilizing housing prices and reducing the flow of vacant properties into

neighborhoods and cities (Hagerty, 2010).

The other key piece of the original MHA initiative was the HARP program, which was

introduced with a goal of refinancing four to five million homeowners with loan-to-value ratios

exceeding 80 percent. Through May 2011, only about 810,000 HARP loans had been made, despite

the fact that the Federal Reserve Board estimates that approximately 4 million borrowers were

eligible (Duke, 2011). Moreover, approximately 93 percent of these loans had loan-to-value ratios

between 80 and 105 percent, the original HARP guidelines (Federal Housing Finance Agency,

2011). Fewer than 60,000 HARP borrowers were significantly underwater (with loan-to-value

ratios between 105 and 125 percent). The Federal Reserve has suggested a number of “frictions”

that may help explain why HARP reached so few borrowers (Duke, 2011). First, the GSEs charged

21

higher upfront fees for refinancing loans with higher risk factors, including higher loan-to-value

ratios. Such fees may have limited the benefits of refinancing through HARP. Second, as the GSEs

became more risk-averse during the mortgage crisis, they began more aggressively requiring

originating lenders to repurchase loans that violated GSE guidelines. This made lenders generally

more risk averse, especially in the case of loans with any higher-risk features, including high loan-

to-value ratios. In addition, some second mortgage holders were refusing to agree to a refinancing

of the senior loans, even though such a refinancing would likely reduce the risk on their junior

loans.

Structural Obstacles in Loan Modification Efforts

The MHA initiative suffered from many flaws in design and implementation. Most of these

concern HAMP and the other HAMP-like programs (PRA, 2MP, UP). Some are also relevant to the

HARP and HHF programs.

Little carrots and no sticks

A key design premise of HAMP and related MHA programs was that servicers could be

induced to modify large numbers of mortgages through two principal vehicles. First, a standardized

protocol sanctioned by the federal government would speed modifications and potentially shield

servicers from fears of legal action from investors in mortgage-backed securities. Second, servicers

would be given incentive payments to perform modifications and would receive additional

payments as modified loans continued to perform. These incentives, however, were generally quite

modest. MHA programs did not provide either a substantial disincentive to proceed with foreclosure

or a system in which servicers were actively held accountable for processing modifications in a

timely and responsive manner. The most promising disincentive to foreclose was originally

proposed by the Obama Administration in the announcement of HAMP, when it recommended

allowing bankruptcy judges to reduce the principal on mortgages on owner-occupied properties. If

servicers and investors faced such a threat of modification via the courts, they might have been

more likely to reduce principal voluntarily and modify loans more aggressively.

MHA rarely employed sanctions against servicers. Despite continual complaints about

servicer actions from the beginning of the program’s operations, the Treasury Department did not

sanction any servicers for noncompliance with HAMP rules for over two years. In earlier statements

22

the Treasury Department claimed that it did not have the power to assess such penalties. However,

in June of 2011, the Department issued a release stating that it was “withholding financial

incentives” for three of the largest servicers in the HAMP program, Bank of America, J.P. Morgan

Chase, and Wells Fargo Bank (U.S. Department of the Treasury, 2011b). This sanction, which itself

is very limited, was the first in the program after two years of numerous complaints about major

servicers as well as disappointing program results.

Substantial variation in HAMP servicer performance may reflect the limited success that

Treasury had in getting all servicers to implement HAMP aggressively. For trial modifications

begun after June 1, 2010, the conversion rate from trial to permanent modifications among the

largest servicers ranged from 62 percent to 88 percent.10 Although some of this may have been due

to differences in the risk of borrowers entering trials, some of the servicers that specialized in

subprime loans, such as Ocwen, had relatively high conversion rates. In addition, some of the

smaller servicers were more aggressive in implementing principal reductions than were some of the

larger servicers. Moreover, in regressions of non-HAMP modifications, Agarwal et al. (2010) find

that, even after accounting for a wide variety of loan and borrower characteristics, modification

behavior varies significantly by servicer. This might be explained by differences in the

organizational cultures or in other characteristics of the servicers.

Reports of servicer problems in handling loan modification requests are numerous (Kiel and

Pierce, 2010). In late 2010, the U.S. Government Accountability Office (2011) conducted a survey

of almost 400 housing counselors involved in the National Foreclosure Mitigation Counseling

program, which is federally funded and managed through the NeighborWorks organization. The

survey asked respondents to list the three most common reasons why borrowers contacted them

about the HAMP program. The most common response (with 59 percent of respondents listing it as

one of their three most common reasons) was that servicers claimed to have lost HAMP application

information. Another top reason (37 percent) was that borrowers had difficulty contacting the

servicer. Overall, 75 percent of counselors characterized borrowers’ experiences with HAMP as

“negative” or “very negative,” with 33 percent falling into the “very negative” category. Moreover,

86 percent of counselor respondents reported that it typically took four months or longer to receive

a decision about a HAMP trial modification from the time the borrower requested it. More than 45

percent reported that it typically took seven months or more.

23

A lack of “sticks” was not just a problem with HAMP. In the case of the HHF program,

several states designed initiatives that included subsidized principal reduction programs, in which

HHF funds would match principal reductions from the servicers on a 1:1 basis (Immergluck, 2010).

However, the large servicers, as well as the GSEs, resisted these programs. Initially the Treasury

Department encouraged states to develop such programs in their HHF plans, but after receiving

significant pushback from servicers and the GSEs, the Department recommended against such

initiatives, although several states persisted in these efforts (Weise, 2010). Moreover, in later rounds

of the HHF, Treasury reportedly told states that they should get servicer “buy in” before proposing a

particular program to Treasury to be funded with HHF funds (Immergluck, 2010).

Fundamentally, the MHA programs were designed around the systems and the preferences

of the servicing industry. Instead of being driven by the needs of borrowers, they programs were

shaped more by the demands and perceived limitations of financial institutions. Certainly, some

consideration to such concerns was merited. However, the federal response was generally tilted

heavily toward the preferences of servicers and investors, often at the expense of borrowers.

A servicing industry built for foreclosure, not modification

The modern single-family mortgage servicing business is a large-scale, “low-touch”

business built around volume and speed (Jacobides, 2005; Levitin and Twomey, 2010). Servicing

industry structures, policies and procedures, and organizational cultures cannot be turned on a dime.

More specifically, the servicing business is designed in some ways to make money on foreclosures,

not to work against them. Overlaying a few incentives and systems on the top of such an entrenched

system did not alter the industry’s response to delinquency and default in a fundamental way – at

least not immediately. Levitin and Twomey (2010) and Thompson (2009) identify a number of

structural reasons why servicers tend to favor foreclosure over loan modification and why servicers

may not always act in the best interests of lender/investors yet alone borrowers. These include

accounting rules, private-label securitization structures, fee structures, and staffing issues.11 The last

two appear to be among the most important obstacles to modifications. Servicing contracts allow

servicers to retain many of the fees charged to delinquent homeowners. Servicers capture these

loan-level fees when the borrower cures the arrearage, or collect them from the proceeds of the

foreclosure sale before the investors are paid. More fundamentally, servicing delinquent or

defaulted loans is substantially more expensive than servicing performing loans, so services often

24

favor foreclosure and liquidation over modification to eliminate more cost-intensive portions of

their portfolios.

It should not be surprising that servicers had difficulty—or resisted—gearing up quickly for

massive numbers of loan modifications. Such mobilization involves heavy fixed costs including

new staff and infrastructure plus outside costs such as property valuation, credit reports and

financing costs (Thompson, 2009). Servicing is a low-margin business that depends on repetition,

speed and few deviations from regular practice. Loan modifications, on the other hand, are more

complex, involve more instances of human intervention and judgment, and require a more complex

regime of compliance with detailed policies and procedures. Minor changes in per-loan costs can

wreak havoc on a servicer’s bottom line, so they may be reluctant to deviate from highly routine

protocols. The crisis required large, performing-loan servicers to act like specialized default

servicers that were more nimble and responsive. However, incrementalism and a failure to

recognize the depth and length of the crisis impeded such a response.

A political climate hostile to helping borrowers

The MHA was introduced in March of 2009, just after the second round of TARP funds was

released and during a time when the Great Recession was peaking. Public backlash against the

TARP program and other financial crisis interventions had been building. Many commentators

portrayed the subprime crisis as one caused principally by the decisions of irresponsible and

spendthrift homebuyers and homeowners (e.g., Percelay, 2007; Tamny, 2008). Anecdotes of

reckless borrowers using home equity loans to buy expensive television sets or go on extravagant

vacations were ubiquitous. The early outlines HAMP spurred CNBC reporter Rick Santelli’s now

famous “losers’ mortgages” rant on the floor of Chicago Mercantile Exchange in February of 2009

(Stolberg, 2009). Santelli’s losers were those who willingly and without distress made greedy,

selfish and stupid decisions that brought them to financial ruin.

In this climate, it became difficult for policymakers to propose bolder approaches to

reducing foreclosures and, especially, to propose large-scale principal reduction programs. With the

change in Congress in 2010, this problem grew more significant, as some Republicans called for

dismantling HAMP as well as other policies responding to the foreclosure crisis, including the

Neighborhood Stabilization Program (Chaddock, 2011).

25

Mixed signals from federal policymakers and the GSEs

As the foreclosure crisis spread from the subprime to the prime mortgage market in 2009

and 2010, more foreclosures were on loans owned by the GSEs. Given that the firms were under

federal conservatorship, this presented federal policymakers with the ability to exercise more

control of over loan modifications and to reduce principal balances where appropriate. However, the

conservator in charge of the GSEs, the independent FHFA, was charged with “conserving” the

short-term assets of the firms and so was resistant to moves, including principal reduction and

unemployment forbearance programs, that might require the firms to take large-scale, short-term

write-downs on their mortgage assets (Weise, 2010; Timiraos, 2011). Even if such modifications

were expected to improve the long-term returns by improving loan performance, the pressures to

conserve short-term assets worked against such a longer-term perspective.

Mitigating Neighborhood Effects: The Neighborhood Stabilization Programs

While the largest set of federal responses to the foreclosure crisis were those focused on

preventing foreclosures, another set of responses focused on addressing some of the negative effects

that foreclosed, vacant properties can have on neighborhoods. The Neighborhood Stabilization

Programs (NSP) are a set of three different programs, each authorized by different pieces of

legislation, that are aimed at the acquisition and reuse of vacant, foreclosed residential properties.

The three programs are referred to as NSP 1 (authorized by HERA), NSP 2 (authorized by the 2009

American Recovery and Revitalization Act (ARRA)) and NSP 3 (authorized by the 2010 Dodd-

Frank Wall Street Reform and Consumer Protection Act). While the precise details of eligible uses

varied among the three programs, funds were generally able to be used for five principal purposes:

1) the acquisition and rehabilitation of abandoned or foreclosed homes; 2) financing mechanisms

such as downpayment assistance programs (to purchase foreclosed homes); 3) the landbanking of

properties; 4) the demolition of blighted properties; and 5) the redevelopment of demolished or

vacant properties.

In NSP 1, Congress gave HUD two months to develop a formula for allocating $3.9 billion

across existing Community Development Block Grant (CDBG) entitlement communities as well as

states. States receiving funds could then could either directly devote resources to purchasing

foreclosed homes, or redistribute the money to localities. HERA required that each state receive a

minimum allocation of 0.5 percent of total NSP 1 funds ($19 million). HUD allocated the bulk of

26

the funds by developing a formula that in turn relied on a number of indirect data sources used to

indicate foreclosure prevalence. If HUD’s formula resulted in a CDBG community receiving an

allocation of less than $2 million, the locality was not awarded a direct grant, but the sum was rolled

into the overall state grant. Among the 1,201 CDGB-eligible governments, over 300 direct grants

were made, including the state grants (U.S. Department of Housing and Urban Development, 2008).

The local government NSP recipients received awards up to $62.2 million (Miami–Dade County)

with a median award of $4.3 million.

As of October 2010, when the vast majority of NSP 1 funds had been obligated (although

not necessarily spent), the allocations of funds obligated by activity broke out as follows: 66 percent

of funds were obligated towards acquisition and rehabilitation; 11 percent was allocated towards the

construction of new housing; 6 percent towards homeownership assistance; 5 percent for

demolition; and 4 percent for land banking and a variety of other activities (U.S. Government

Accountability Office, 2010). This left 9 percent going to administrative costs.

The NSP 1 program has been criticized by community development experts and scholars for

a number of reasons (Newberger, 2010; Mallach, 2009). One complaint was that the 18-month

timeline to obligate NSP funds was too short to develop and execute complex local redevelopment

programs.12 As a result, many NSP 1 recipients obligated a large percentage of their grant funds

within the last three months of the 18-month period. Figure 5 shows that many NSP recipients made

a large percentage of their obligations in the last few months of their 18-month periods. While 99

percent of recipients had obligated at least 80 percent of their funds by October 1 2010, and 90

percent had obligated at 100 percent of their funds, the situation was far different just a few months

earlier.13 At the end of June, only 34 percent had obligated at least 80 percent of their funds, and 38

percent still had obligated less than 60 percent of their funds. Given the complexity and challenges

of implementing NSP1, it is not surprising that a large percentage of recipients had not obligated a

large portion of their funds at the 15-month mark. Rather, it reflects the challenges that many

localities faced in getting their programs off the ground. While the 18-month deadline was

consistent with the urgency intended in HERA, it may not have been the best strategy for making

the most effective use of resources.

One byproduct of the need to obligate substantial amounts of funding in the last few months

of the NSP 1 period was that some localities were forced to change their NSP strategies as they

neared the deadline for obligating funds. An example is that some localities shifted away from their

27

Figure 5. NSP 1 Recipients’ Obligations as of June 24, 2010 and October 1, 2010

Source: U.S. Government Accountability Office (2010).

15%

23%

28%

25%

9%

June 24, 20100.3% 1%

9%

90%

0 to 39.9%40 to 59.9%60 to 79.9%80 to 99.9%100%

October 1, 2010

Percent of Funds Obligated

28

NSP plans that focused on acquiring single-family homes towards purchasing foreclosed

multifamily apartment buildings (Korte, 2010). This may have been one of the rare forays that

some localities have made into the redevelopment of multifamily rental housing.

Another criticism was that Congress required an aggressive recapturing of any earnings

that localities might receive from the program, although these requirements were eliminated in

later phases of the program. The statute also required that local recipients purchase foreclosed

properties from lenders or loan servicers at discounts of 5 to 15 percent below appraised value, a

practice that was difficult to implement during a time of highly volatile property values. This

policy could also make it difficult for NSP 1 recipients to compete with speculative investors or

bulk purchasers who might be able to pay more. (This requirement was later modified to only a 1

percent discount.) It also prohibited recipients from selling properties for prices that exceeded the

sum of their acquisition and redevelopment costs, meaning that a local government could not use

profits from one property to subsidize the redevelopment of another property that might be sold

at a loss. Successful vacant property reclamation initiatives have utilized such cross-

subsidization tactics to increase scale and overall impact (Paul, 2008).

NSP 1 allowed only 10 percent for administrative costs. Given that most local

governments had little infrastructure established for acquiring and redeveloping foreclosed

homes, this modest allocation was unlikely to be sufficient to enable many to develop capacity in

this area. In addition, for the portion of funding passing through the states, a significant portion

of administrative dollars was not passed through to localities. In NSP 2, the shortage of

administrative resources may have been partly rectified in some places by the inclusion of

dedicated technical assistance funding.

At the time of NSP 1 implementation, most local governments did not have established

land banks that could accept properties, erase back taxes and hold properties until redevelopment

plans were implemented or until housing markets recovered (Alexander, 2011). While HERA

encouraged the use of local land banks, it did not establish any form of national land bank for

localities without such entities nor did it provide start-up, operating funding or technical

assistance support for such endeavors.

A key challenge for NSP efforts in some places was the increased appetite in late 2008

and early 2009 that investors had for foreclosed properties (Immergluck, 2011; Coulton et al.,

2008). Some NSP recipients reported that servicers (or their realtors) did not want to work with

29

the more cumbersome and slower NSP process when they could sell quickly to an investor

purchasing a home with cash (Newberger, 2010; U.S. Government Accountability Office, 2010).

Research in Atlanta and Cleveland found that about the time that NSP 1 was being implemented,

sales of foreclosed properties by lenders – especially lower-value properties – were accelerating

rapidly (Coulton et al., 2008; Immergluck, 2011). Thus, many properties that might be targeted

for strategic acquisitions by NSP efforts fell into the hands of investors, before they could be

acquired with NSP funds.

Especially for properties selling for well under $100,000, and even well under $50,000 in

many cases, servicers dealing with thousands of homes may have little patience for buyers who

need to get appraisals and environmental reviews before completing an acquisition. For NSP

efforts focusing on specific blocks or neighborhoods, losing properties to investors can be

particularly problematic. If these properties remained vacant, or were not rehabilitated

sufficiently, they could be impediments to the revitalization of the targeted area.

One response to the investor competition problem was the introduction of “first look”

programs developed by the National Community Stabilization Trust (NCST) in cooperation with

the GSEs, some of the large national loan servicers, and the FHA. In a first look program, a

lender or servicer agrees to offer local governments—and sometimes NSP recipients in particular

– a chance to purchase foreclosed properties before offering these homes for sale on the open

market. In 2010, a “National First Look Program” was sponsored by NCST and HUD. NSP

grantees were notified when an REO property owned by the GSEs, the FHA or some large banks

became available in an NSP targeted area. NSP recipients then had 24 to 48 hours to express

interest in a specific property (U.S. Department of Housing and Urban Development, 2010a).

The first-look period lasted five to twelve business days during which the NSP recipient

conducted inspections and estimated repair costs. The NSP recipient must commit to purchasing

the home by the end of the first-look period or the market would be put up for sale on the open

market.

In early 2009, as ARRA was being drafted, a second phase of NSP (later, “NSP 2”) was

designed as a competitive rather than formula-funded program, in part to address the issue of

grantee capacity (U.S. House of Representatives, 2009). Besides the competitive funding

approach NSP2, funded at $2 billion – about half the level of NSP 1 – also provided $50 million

in technical assistance funding that could be used for help both NSP 1 and NSP 2 recipients. The

30

technical assistance funding came at an opportune time because some NSP 1 recipients were

having difficulties implementing their programs fast enough to obligate all of their funds by the

18-month deadline. NSP 2 also encouraged nonprofits and consortia of nonprofit and/or local

governments to apply for funding.

While NSP 2 addressed some of the problems identified with NSP 1, it also introduced

some additional challenges. One change was that any redevelopment had to be focused only on

housing uses; NSP 1 had allowed redevelopment for uses other than housing. Given the

oversupply of housing and the dilapidated nature of many homes in some communities, some

argued that such a constraint would preclude some sound redevelopment plans. Another

challenge was that NSP 2, like NSP 1, relied upon an existing infrastructure of federal-local

intergovernmental funding regulations and requirements.

NSP 3, which was included in the Dodd-Frank Wall Street Reform and Consumer

Protection Act of 2010, was funded at $1 billion, with $20 million set aside for technical

assistance (U.S. House of Representatives, 2010). NSP 3 returned to a formula-based funding

approach as in NSP 1, although with some focus on areas with high unemployment rates.

Because NSP 2 and NSP 3 are in their early stages, little systematic information is available on

their outcomes at this point.

Key Lessons for Neighborhood Stabilization Policy Design

Despite the many challenges and obstacles in all of the NSP programs, they do hold

significant promise for spurring significant innovation in local efforts to reclaim and redevelop

vacant properties. However, in considering any future policy initiatives in this area, there are a

few major lessons that can be gleaned from the NSP experience thus far.

NSP funding was small in relation to the flow of foreclosed properties in the U.S.

Given the scale of the problem in many metropolitan areas, NSP funds, by themselves,

are unlikely to make a substantial dent in the vacant property problem. The volume of foreclosed

properties held by lenders, investors, and government agencies at any one time has ebbed and

flowed in recent years, but conservative estimates tend to fluctuate between 550,000 and 720,000

properties since 2007.14 Given that properties are continuously entering and exiting lender

31

ownership, the number of foreclosed properties flowing into communities during the life of the

NSP programs will certainly be on the order of magnitude of several million properties.

By contrast, HUD’s projection of how many homes that will be directly affected by all

three NSP programs is on the order of 100,000 properties. (U.S. Department of Housing and

Urban Development, 2011b). As of March 2011, the agency said that more than 36,000 homes

had been or were being purchased or rehabilitated. Thus, the NSP programs, even if highly

effective, will be very limited in scale compared to the aggregate flow of foreclosed properties.

The narrow NSP legislation, and the resulting rules that followed, constrained localities from

being opportunistic and responsive to local market conditions.

The narrow focus of the NSP programs on purchasing lender-owned properties for

redevelopment and reoccupancy hamstrung local governments and NSP recipients from making

the most effective use of NSP dollars over time. In some cities at least, as NSP 1 was rolling out,

many bank-owned properties – especially lower-value properties in the types of neighborhoods

targeted by NSP -- were being sold off rapidly (Immergluck, 2011). By the time NSP recipients

were beginning to acquire properties, many key strategic properties were in the hands of

investors, making it difficult to use NSP funds to acquire them.15 A more robust and adaptive

approach would have involved funding to local intermediaries and redevelopment organizations

to purchase and redevelop not just foreclosed properties, but vacant properties more generally, or

to engage in proactive activities to control properties and keep them occupied before the

foreclosure process was completed. While some administrative rule changes nudged the

programs in this direction, such incremental movements often came well after substantial NSP

dollars had already been committed. Moreover, it remains unclear how effective such changes

were in allowing for more creative, responsive, and efficient uses of program dollars.

Most localities had little experience addressing foreclosed or vacant properties.

Despite the mismatch between the flow of foreclosed properties and the funding levels of

NSP programs, funding levels were large compared to the existing capacity of many local

governments to acquire and redevelop foreclosed properties. The formula-funding approach of

NSP 1 meant that many localities with little experience in dealing with vacant or foreclosed

properties were thrust into addressing a very challenging problem with limited resources and

32

little established capacity. Moreover, many of the localities hit hardest by the foreclosure crisis –

especially in the sand-state suburbs and exurbs of California, Florida, Arizona and Nevada –

were especially inexperienced in community development and in dealing with troubled

properties. In some ways, the older industrial cities of the Midwest and Northeast had at least

been familiar with these sorts of problems, having struggled in the past with issues of

depopulation, foreclosures and vacancies. The technical assistance funding in NSP 2 and 3 was

designed in attempting to address some of these problems, but came at a late stage in the

implementation of NSP 1.

NSP programs have been constrained by cumbersome legacy systems of funding and oversight.

NSP programs utilized Community Development Block Grant and HOME funding

systems to distribute funding and monitor performance. Many federal rules from these programs

were imposed on NSP funds. This made the programs more cumbersome than they might have

been and made it difficult to be opportunistic and entrepreneurial. At the beginning of NSP 1,

there were anecdotal reports of recipients being told that they should be very careful in

implementing the program because HUD’s investigators would be keeping a close eye on

compliance.

Longer-Term Responses to the Crisis

It is beyond the scope of this paper to discuss all of the responses to the foreclosure crisis

that are aimed at reforming lending, servicing, and investment aspects of the mortgage market.

These efforts include those to strengthen the regulation of the origination of mortgages to reduce

excessively risky lending and encourage investors to be more careful in funding high-risk loan

products. Some of this involves implementing provisions of the Dodd-Frank Act to discourage

loans with very high debt-to-income ratios or to require lenders to retain more “skin in the game”

when they originate higher-risk mortgages.

In the servicing arena, there are efforts underway to increase regulatory oversight of the

servicing industry and efforts to change servicer compensation to encourage meaningful loan

modifications when they make sense. While federal agencies took some initiative on reforming

servicer compensation structures, particularly through the FHFA, federal policymakers were less

prominent in investigating problems in the servicing industry -- including faulty foreclosure

33

practices and fraudulent and missing documents (including “robosigning”) – that came to

national prominence in the summer and fall of 2010. The attorneys general from a large number

of states took the lead on servicing and foreclosure process problems. For a time, the new

Consumer Financial Protection Bureau (CFPB), created by the Dodd-Frank Act, appeared to be

involved in cooperating with the state attorneys general in attempting to arrive at a major

settlement with servicers over these and other servicing industry problems. However, the new

agency and, especially its high profile informal head, Elizabeth Warren, were severely criticized

in the spring and summer of 2011 for their cooperation with the state attorneys general (Panchuk,

2011). After being chastised by Congressional foes, the CFPB went out of its way to point out

that the Department of Justice and other federal agencies were taking the federal lead. A

spokesperson for the CFPB told the press that these other agencies “are currently conducting

negotiations of the terms of a potential settlement with mortgage servicers. The CFPB is not.”

(Randall, 2011).

While not a direct result of the foreclosure crisis, the long-term restructuring of mortgage

secondary markets, which will unfold over several years, will certainly be affected by the crisis.

The GSEs solvency was laid low by plunging property values and losses on investments in

subprime mortgage-backed securities, both of which were caused in large part by skyrocketing

foreclosures.

Broader Lessons Learned from Crisis-Based Policymaking

Notwithstanding the important, detailed lessons learned around HAMP, NSP and other

policy responses to the foreclosure crisis described above, there are at least two broader lessons

that can be drawn from the federal policy response thus far. First, many responses—especially

loan modification efforts—were often tentative, incremental and marginal. Moreover, they often

did not accumulate to a sizeable response, particularly in the first two to three years of the crisis

when intervention would have been most promising. Due to the nature of housing price and

foreclosure cycles, it would have been easier to have a material impact on foreclosures if a more

forceful response came much sooner, and especially before job losses and housing depreciation

accelerated, making it more difficult to reduce foreclosures. Moreover, even when policy

initiatives were announced, it sometimes took as long as a year to get them launched. The

political hostility to helping “undeserving” homeowners, the complexity of the mortgage market,

34

and difficulty of working with Congress, especially after the fall of 2010, all contributed to this

problem.

Second, some programs were so excessively engineered to anticipate or prevent certain

types of problems to the extent that the fundamental efficacy of the programs was negatively

affected. In the case of foreclosure prevention programs the concerns often centered on issues of

moral hazard among homeowners. In the case of NSP, the concerns centered on inappropriate

use of funds or even malfeasance. The heavy focus on these issues affected the fundamental

design of the policy responses in ways that sometimes limited program effectiveness and

efficiency and made it difficult to reach scale.

In the loan modification arena, for example, lenders and economists argued that by

offering advantageous modifications or loans to defaulting borrowers, some programs might

encourage others to default. Yet no compelling evidence has been presented that foreclosures

rose substantially due to HAMP or other MHA programs. Excessive concerns about moral

hazard contributed to a cumbersome design of the FHA Hope for Homeowners program and to

continual resistance to principal reduction proposals.

It is too early to render a definitive assessment of the impact of any of the various federal

responses to the foreclosure crisis. It is reasonable, however, to suggest that the foreclosure

prevention efforts – and especially the loan modification programs – deployed in response to the

crisis were no match for the complexity, scale and depth of the foreclosure problem. Without the

stick of bankruptcy cramdown, servicers and lender/investors were in the driver’s seat and were

able to drag their feet. A more forceful and effective strategy would also have employed servicer

sanctions more aggressively in the implementation of federal modification programs. Periodic

servicer report cards and occasional “jawboning” by cabinet secretaries were clearly not

sufficient to get servicers to respond aggressively enough to the crisis (Braucher, 2010).

As suggested above, the policy responses to the crisis suffered from many problems of

design and implementation that could have been avoided. More forceful approaches that did not

yield so readily to the desires of the financial services industry would likely have proven more

effective. In the foreclosure prevention arena, at the point early in the crisis when aggressive

intervention had the best chances of slowing foreclosures more substantially, weak, voluntary

efforts delivered little substantive relief to most distressed homeowners. When proposals to

35

create more promising responses met substantial resistance from the financial services industry,

they were typically not supported by congressional policymakers, federal bank regulators, or the

Bush or Obama Administrations.

Notwithstanding the issues of political will, many policy responses were confounded by a

host of dysfunctional complexities in the mortgage servicing and securitization industries,

complexities that policymakers had a hard time addressing. If the nation is to be more prepared

and more resilient to any future surge in foreclosures, some very basic and fundamental

problems in the mortgage market must be addressed. The market must be restructured in ways

that will allow it to respond more quickly to economic and housing market shocks.

Whether it is in the restructuring of mortgage markets and regulation going forward, or in

designing policy responses to foreclosure or other crises, there is a need to move away from

relying on public policies that are limited by the political economy of financial capital. Relying

on financial market expertise during the foreclosure crisis was essential. However, allowing the

financial sector, after receiving unprecedented federal support, to dictate the boundaries of

federal policy was a recipe for a timid and insufficient response.

36

References Agarwal, S., Amromin, G., Ben-David, I., Chomsisengphet, S., Evanoff, D. 2010. Market-based loss mitigation practices for troubled mortgages following the financial crisis. Working Paper WP 2011-3. Chicago: Federal Reserve Bank of Chicago. October. Alexander, F. 2011. Land Banks and Landbanking. Flint, MI: Center for Community Progress. June. Retrieved September 10, 2011 from http://www.communityprogress.net/filebin/pdf/ new_resrcs/LB_Book_2011_F.pdf. American Securitization Forum. 2007. Streamlined foreclosure and loss avoidance framework for securitized subprime adjustable rate mortgage loans. December 6. Retrieved on January 15, 2009 at http://americansecuritization.com/uploadedFiles/FinalASFStatementonStreamlined ServicingProcedures.pdf. Bocian, D., Li, W., Ernst, K. 2010. Foreclosures by race and ethnicity: The demographics of a crisis. Retrieved August 23, 2011 from http://www.responsiblelending.org/mortgage-lending/ research-analysis/foreclosures-by-race-and-ethnicity.pdf. Braucher, J. 2010. Humpty Dumpty and the foreclosure crisis: Lessons from the lackluster first year of the Home Affordable Modification Program. Arizona Legal Studies Discussion Paper No 09-37. Tuscon: University of Arizona. Calculated Risk. 2011. REO Inventory: Fannie, Freddie, FHA, PLS and FDIC through Q2 2011. Retrieved September 2, 2011 from http://cr4re.com/charts/charts.html?Delinquency#category= Delinquency&chart=REOInventoryQ22011.jpg. Chaddock, G. R. 2011. House looks to cut $62 billion for distressed homeowners, properties. Christian Science Monitor. Retrieved August 29 from http://www.csmonitor.com/USA/Politics/ 2011/0311/House-looks-to-cut-62-billion-for-distressed-homeowners-properties. Coulton, C., Schramm, M. and Hirsch, A. 2008. Beyond REO: Property transfers at extremely distressed prices in Cuyahoga County, 2005-2008. Center on Urban Poverty and Community Development. Case Western University. December. Retrieved on January 2, 2010 at http://blog.case.edu/msass/2008/12/09/20081209_beyond_reo_final.pdf.

Duke, E. 2011. Speech at the Federal Reserve Board Policy Forum: The housing market going forward: Lessons learned from the recent crisis. Washington, D.C. Retrieved September 1, 2011 from http://www.federalreserve.gov/newsevents/speech/duke20110901a.htm. Fannie Mae. 2011. Fannie Mae second quarter credit supplement. August 5. Retrieved September 2, 2011 from http://www.fanniemae.com/ir/pdf/sec/2011/q2credit_summary.pdf. Federal Deposit Insurance Corporation. 2008. FDIC loss sharing proposal to promote affordable loan modifications. November 20. Retrieved August 18, 2011 from http://www.fdic.gov/ consumers/loans/loanmod/.

37

Federal Housing Finance Agency. 2011. Foreclosure prevention & refinance report. May. Retrieved August 22, 2011 from http://www.fhfa.gov/webfiles/21844/MayFPM72711.pdf. Federal Reserve Bank of New York. 2010. FAQ: MBS Purchase Program. June 28. Retrieved August 20, 2011 from http://www.newyorkfed.org/markets/mbs_faq.html. Fitch Ratings. 2009. U.S. RMBS servicers’ loss mitigation and modification efforts. New York: Fitch Ratings. May 26. Gagnon, J., Raskin, M., Remache, J. and Sack, B. 2011. The financial market effects of the Federal Reserve’s large scale asset purchases. The International Journal of Central Banking 7: 3-44. Hagerty, J. 2010. Loan modifications help the housing market. Wall Street Journal. January 19. Retrieved August 29, 2011 from http://online.wsj.com/article/SB100014240527487036266 04575010711922339400.html. Hancock, D. and Passmore, W. 2010. Did the Federal Reserve’s MBS purchase program lower mortgage rates? Washington, D.C.: Federal Reserve Board of Governors. Retrieved August 19 from http://www.federalreserve.gov/pubs/feds/2011/201101/201101pap.pdf. Hiltzik, M. 2011. Homeowners deserve more than halfhearted mortgage relief. Los Angeles Times. June 28. Retrieved August 23, 2011 from http://articles.latimes.com/2011/jun/28/ business/la-fi-hiltzik-20110628. Hope Now Alliance. 2007. Hope Now Alliance created to help distressed homeowners. Press Release. http://www.fsround.org/media/pdfs/AllianceRelease.pdf. Immergluck, D. 2009. Foreclosed: High-risk lending, deregulation and the undermining of America’s mortgage market. Ithaca: Cornell University Press. Immergluck, D. 2010. State foreclosure mitigation strategies: A comparison of round 1 and Hardest Hit Fund plans in states with nonjudicial foreclosure processes. October 20. Retrieved August 31, 2011 from http://ssrn.com/abstract=1695217. Immergluck, D. 2011. Distressed and dumped: The market dynamics of low-value, foreclosed properties during the advent of the federal Neighborhood Stabilization Program. Journal of Planning Education and Research, in press. Jacobides, M. 2005. Industry change through vertical disintegration: How and why markets emerged in mortgage banking. Academy of Management Journal 48: 465-498.

38

Joint Economic Committee. 2007. Momentum builds for Schumer’s call for additional federal funds to avert subprime foreclosure crisis. Press release. Washington, DC: Joint Economic Committee of the U.S. Congress June 5. Retrieved September 3, 2011 from http://jec.senate.gov/public/index.cfm?p=PressReleases&ContentRecord_id=eb2d3916-7e9c-9af9-742d-195da80f550e&ContentType_id=66d767ed-750b-43e8-b8cf-89524ad8a29e&Group_id=1a3081df-5769-4cc9-99e8-a0387a830c5f. Kiel, P. and Pierce, O. 2010. Homeowner questionnaire shows banks violating government program rules. Propublica. Retrieved August 22, 2011 from http://www.propublica.org/ article/homeowner-questionnaire-shows-banks-violating-govt-program-rules. Korte, G. 2010. Clock ticking for foreclosure aid: Local governments must use $1B in federal funds. USA Today, July 28, Page 1A. Levitin, A., and Goodman, J. 2008. The effect of bankruptcy strip-down on mortgage markets. Georgetown University Law Center. Research Paper No. 1087816. February 6. Levitin, A., and Twomey, T. 2010. Mortgage servicing. Yale Journal on Regulation; Winter 2011; 28.1: 1-90. Lillis, M. 2009. Obama Administration abandons cramdown. Washington Independent. July 16. Retrieved August 15, 2011 from http://washingtonindependent.com/51486/obama-administration-abandons-cramdown. Mallach, A. 2009. Saving America’s struggling communities: Defining the federal role in addressing the secondary impacts of the foreclosure crisis. Washington, DC: Brookings Institution. February. Retrieved August 29, 2011 from http://www.brookings.edu/topics/ financial-markets.aspx. Mayer, N., Tatian, P., Temkin, K., and Calhoun, C. 2010. National Foreclosure Mitigation Counseling program evaluation: Preliminary analysis of program effects. September 2010 update. Washington, DC: Urban Institute. December. Mortgage Bankers Association. 2011. National delinquency survey: Facts. Retrieved August 23, 2011 from http://www.mbaa.org/files/Research/Flyers/NDSFAQ.pdf. Newberger, H. 2010. Acquiring privately held REO properties with public funds: The case of the Neighborhood Stabilization Program. Washington, DC: Federal Reserve Board of Governors. Retrieved August 31, 2011 from http://www.bos.frb.org/commdev/REO-and-vacant-properties/101-Newburger.pdf. Nocera, J. 2008. As crisis spread, alarm led to action. New York Times. October 2. Retrieved October 2, 2008 from http://www.nytimes.com/2008/10/02/business/02crisis.htm.

39

Panchuk, K. 2011. House Republicans push back at plan to settle with mortgage servicers. Retrieved March 9, 2011 from http://www.housingwire.com/2011/03/09/house-republicans-push-back-at-plan-tosettle-with-mortgage-servicers. Paul, E. 2008. Brownfields redevelopment toolbox for disadvantaged communities. Northeast-Midwest Institute. Retrieved July 28, 2010 from http://www.epa.gov/brownfields/pdf/bftoolbox _disadvantage_communities.pdf. Percelay, B. 2007. Predatory borrowers. Boston Globe. September 20. Retrieved February 23, 2011 from http://www.boston.com/news/globe/editorial_opinion/oped/articles/2007/09/20/ predatory_borrowers/. Powell, M. and Martin, A. 2011. Foreclosure aid fell short, and is fading. New York Times. March 30, p. A-1. Prior, J. 2011. Freddie Mac authorizes HUD unemployment program. July 14. Retrieved August 22, 2011 from http://www.housingwire.com/2011/07/14/freddie-mac-authorizes-hud-unemployment-program. Randall, M. 2011. Questions on Warren's role in mortgage talks dominate hearing. Dow Jones Newswires. July 7. Retrieved August 29, 2011 from http://www.foxbusiness.com/industries/ 2011/07/07/questions-on-warrens-role-in-mortgage-talks-dominate-hearing/. Said, C. and Zito, K. 2007. Bush’s plan for an interest rate freeze gets and icy response. San Francisco Chronicle. December 7. Retrieved on August 23, 2008 at http://www.sfgate.com/cgi-bin/article.cgi?f=/c/a/2007/12/07/MNQCTPRVP.DTL. Special Investigator General for the Troubled Asset Relief Program (SIGTARP). 2011. Quarterly report to Congress. July 28. Retrieved August 1, 2011 from http://www.sigtarp.gov/reports/ congress/2011/July2011_Quarterly_Report_to_Congress.pdf. Stolberg, C. 2009. Critique of housing plan draws quick White House offensive. New York Times. February 20. Retrieved August 20, 2011 from http://www.nytimes.com/2009/02/21/ us/politics/21obama.html. Summers, L. 2009. Letter of Larry Summers, Director-designate National Economic Council to Senate and House Leadership. January 15. Retrieved January 18, 2009 from http://financialservices.house.gov/summers011509.pdf. Swarns, R. 2008. Federal mortgage plan falls short. New York Times. April 30. Retrieved August 18, 2011 from http://www.nytimes.com/2008/04/30/business/30fha.html. Tamny, J. 2008. The sanctification of irresponsible borrowers. Real Clear Markets. October 30. Retrieved February 23, 2011 from http://www.realclearmarkets.com/articles/2008/10/ the_sanctification_of_irrespon.html.

40

Thompson, D. 2009. Why servicers foreclose when they should modify and other puzzles of servicer behavior. Washington, DC: National Consumer Law Center. October. Timiraos, N. 2011. An accidental housing chief embraces the power of 'no'. Wall Street Journal. August 31. Retrieved September 2, 2011 from http://online.wsj.com/article/SB100014240531 11904199404576538803284685440.html. U.S. Department of Housing and Urban Development. 2008. Methodology for allocation of $3.92 billion of emergency assistance for the redevelopment of abandoned and foreclosed homes. Retrieved on February 20, 2010 from http://www.hud.gov/offices/cpd/ communitydevelopment/programs/neighborhoodspg/nspfa_methodology.pdf. U.S. Department of Housing and Urban Development. 2010a. Guidance for tracking and reporting the use of NSP funds: Obligations for specific activities. Retrieved August 31, 2011 from http://hudnsphelp.info/media/resources/NSPPolicyAlert_Obligations_4-23-10.pdf. U.S. Department of Housing and Urban Development. 2010b. HUD Secretary announces national First Look program to help communities stabilize neighborhoods hard-hit by foreclosure. Retrieved September 3, 2011 from http://portal.hud.gov/hudportal/HUD?src=/press/ press_releases_media_advisories/2010/HUDNo.10-187. U.S. Department of Housing and Urban Development. 2010c. Guidance on the impact of new definitions for NSP-eligible properties. April 2, 2011. Retrieved September 13, 2011 from http://hudnsphelp.info/media/resources/ImpactOfNewDefinitions.pdf. U.S. Department of Housing and Urban Development. 2011a. FHA single-family operations. January. Retrieved August 22, 2011 from http://www.hud.gov/offices/hsg/rmra/oe/rpts/ooe/ ol2011.pdf. U.S. Department of Housing and Urban Development. 2011b. The facts about the Neighborhood Stabilization Program. Retrieved September 1, 2011 from http://blog.hud.gov/2011/03/15/facts-neighborhood-stabilization-program/. U.S. Department of the Treasury. 2009. Making Home Affordable: Updated detailed program description. March 4, 2009. Retrieved August 19, 2011 from http://www.treasury.gov/press-center/press-releases/Documents/housing_fact_sheet.pdf. U.S. Department of the Treasury. 2010. Making Home Affordable. September 16. Unpublished presentation. U.S. Department of the Treasury, 2011a. Making Home Affordable program performance report through June 2011. August 5. Retrieved August 29, 2011 from http://www.treasury.gov/ initiatives/financial-stability/results/MHA-Reports/Documents/ June%202011%20MHA%20Report%20FINAL.PDF.

41

U.S. Department of the Treasury, 2011b. Obama Administration releases May housing scorecard featuring new Making Home Affordable servicer assessments, regional spotlight on Phoenix housing data. June 9. Retrieved August 31, 2011 from http://www.treasury.gov/press-center/press-releases/Pages/tg1205.aspx. U.S. Government Accountability Office. 2010. Neighborhood Stabilization Program: HUD and grantees are taking actions to ensure program compliance but data on program outputs could be improved. GAO-11-48. December. U.S. Government Accountability Office. 2011. Survey of housing counselors about HAMP. GAO-11-376-R. May 26. U.S. House of Representatives. 2008. H.R. 3221 Housing and Economic Recovery Act. July 30. Retrieved September 12, 2011 from http://www.govtrack.us/congress/bill.xpd?bill=h110-3221. U.S. House of Representatives. 2009. H.R.1. American Recovery and Reinvestment Act of 2009, Title XII, Community Development Fund. Retrieved March 17, 2009 from http://thomas.loc.gov/cgi-bin/query/z?c111:H.R.1.enr. U.S. House of Representatives. 2010. H.R. 4173. Dodd-Frank Wall Street Reform and Consumer Protection Act. Retrieved September 1, 2011 from http://docs.house.gov/rules/finserv/ 111_hr4173_finsrvcr.pdf. U.S. Senate. 2007. Senate Bill 2136: Helping families save their homes in bankruptcy act of 2007. October 3. United States Senate. Retrieved September 3, 2011 from http://www.opencongress.org/bill/110-s2136/text. Weise, K. 2010. Fannie and Freddie’s regulator opposes reducing mortgages for struggling homeowners. Propublica. December 17. Retrieved August 29, 2011 from http://www.propublica.org/article/fannie-and-freddies-govt-regulator-opposes-reducing-mortgages-for-strugglin. Weiss, N., Getter, D., Jickling, M., Keightley, M., Murphy, E. 2008. Housing and Economic Recovery Act of 2008. Congressional Research Service. Order Code RL34623. August 19. Retrieved on January 2, 2009 at http://assets.opencrs.com/rpts/RL34623_20080819.pdf. White, A. 2009. Deleveraging the American homeowner: The failure of 2008 voluntary mortgage contract modifications. Connecticut Law Review, Vol. 41, p. 1107. Available at SSRN: http://ssrn.com/abstract=1325534. White, A. 2011. Foreclosures and modifications - securitized mortgage data through May 25, 2011. Retrieved September 2, 2011 from http://www.valpo.edu/law/faculty/awhite/ data/index.php.

42

Endnotes.

1 In a study of the NFMC program’s first two years, researchers at the Urban Institute

found that the odds of counseled homeowners who had begun foreclosure curing their foreclosure were 70 percent greater than if they had not received NFMC counseling (Mayer et al., 2010). Loan modifications received by NFMC clients receiving loan modifications resulted in substantially lower loan payments than would have been received without counseling. The Urban Institute researchers estimate that NFMC clients receiving loan modifications in the first two program years reduced their mortgage payments by $267 per month more than they would have without NFMC counseling. The sustainability of modifications was also greater for NFMC clients than for other borrowers.

2 Twelve-month redefault rates would be expected to be substantially higher than these rates.

3 In the first step of the modification, the loan servicer rolls any unpaid interest and fees into the outstanding loan balance. Then, the servicer reduces the interest rate on the loan down to as low as 2 percent (with half of the loss in interest being absorbed by the federal government). If the 31 percent monthly debt-to-income ratio has still not been reached, the term of the loan can be extended to as long as 40 years. If the 31 percent ratio has still not been reached, the servicer, at its option (and sometimes requiring the permission of the lender/investor) can defer the due date of some of the principal on the loan (called “forbearance”).

4 White (2011) collects data from private-label mortgage backed securities that show loss severity rates ranging from 40 to over 80 percent with a median of approximately 64 percent. Fannie Mae’s quarterly reporting shows that REO sale prices as a share of unpaid principal balances averaged approximately 56 percent in 2009 and 2010 (Fannie Mae, 2011). Because unpaid principal balance does not include interest and fees, loss severity ratios will be greater than 44 percent on such loans, and most likely approach something on the order of 50 percent.

5 A short sale occurs when a borrower sells her house for less than the amount owned on the outstanding mortgage(s) and the lender(s) agrees to release the mortgage on the property for this amount. Deeds-in-lieu of foreclosure occur when borrowers sign over their property to the lender in exchange for the mortgage being released.

6 HARP assisted over 810,000 borrowers as of May, 2011 in refinancing loans with loan-to-value ratios between 81 and 125 percent. However, approximately 93 percent of these loans had loan-to-value ratios between 80 and 105 percent, the original HARP guidelines. This means that fewer than 60,000 HARP borrowers were significantly underwater.

7 For examples of media coverage portraying MHA foreclosure prevention efforts as lackluster or a failure, see Powell and Martin (2011) and Hiltzik (2011). For an academic perspective see Braucher (2010).

8 HARP activity is not included in Figure 3 because most HARP borrowers are unlikely to have been serious risks for foreclosure since the GSEs did not relax other lending guidelines (including credit history) significantly as a part of the program. The focus here is on the non-HARP activity.

9 The estimated number of foreclosure starts was derived as follows. First, the number of foreclosure starts was taken from the Mortgage Bankers Association National Delinquency Survey. These totals were grossed up by 1.156 because the Mortgage Bankers Association estimates that its survey of captures only 85 to 88 percent of the mortgage market (Mortgage Bankers Association, 2011). Then, because mortgages on non-owner occupied properties are

43

generally not eligible for MHA programs, the estimate must be adjusted for owner occupancy. Bocian, Li and Ernst (2010) estimate that 82 percent of foreclosures from 2007 to 2009 were on owner-occupied homes. This factor is used to adjust the estimate of foreclosure starts downward to provide an estimate of foreclosure starts on owner-occupied homes.

10 Requirements for starting trials changed significantly in June 2010 because full verification of income was not required uniformly before this date. As a result, conversions to permanent modifications increased significantly after this date (U.S. Department of Treasury, 2011a).

11 There is considerable debate over how pooling and servicing agreements and the structures of private label securitization structures impede substantial loan modifications. Thompson (2009) is among those who argue that securitization is not a significant impediment to modification, while Levitin and Twomey (2010) argue that securitization structures can effectively impede modifications. Moreover, Agarwal et al. (2010) finds that securitized loans are significantly and substantially less likely to be modified than loans held in portfolio.

12 The 18-month deadline to obligate NSP1 funds did not mean that all funds had to be expended within that timeframe. According to HUD regulations, “obligation” means “the amounts of orders placed, contracts awarded, goods and services received, and similar transactions during a given period that will require payment by the grantee (or subrecipient) during the same or a future period” (U.S. Department of Housing and Urban Development, 2010b). Obligations must be for specific NSP activities that can be linked to a specific address or household. Thus, a subcontract to a third party that would later purchase specific properties would not be considered an obligation.

13 Grantees’ 18-month periods did not all end on October 1 2010. The deadline depended on when their grant agreements with HUD were executed. However, most periods ended around October 2011. 14 One source of national estimates of foreclosed properties is the financial blog Calculated Risk (2011), which relies on estimates from the housing economist Tom Lawler and have been generally consistent with some other estimates. Calculated Risk’s estimates include properties owned by the GSEs, HUD/FHA, private label securities, and FDIC insured banks. Properties owned by credit unions, non-FDIC-insured lenders, and some smaller federal agencies are not included. Thus Calculated Risk estimates that the data represent “at least 90 percent” of all foreclosed properties. To account for the incomplete coverage, I multiply the Calculated Risk totals by 1.11 to come up with the range.

15 As NSP 1 was being implemented, the definitions of “foreclosed” and “abandoned” properties were broadened in the Spring of 2010 to make the funds more flexible and responsive. For example, the definition of foreclosed properties was broadened to include properties for which a mortgage is more than 60 days delinquent as well as tax delinquent properties (U.S. Department of Housing and Urban Development, 2010c).