Full text of Current Issues in Economics and Finance : The Homeownership Gap, Volume 16, Issue 5 : The Homeownership Gap

View original document

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

www.newyorkfed.org/research/current_issues
✦

May 2010
✦

Volume 16, Number 5

IN ECONOMICS AND FINANCE

current issues

FEDERAL RESERVE BANK OF NEW YORK

The Homeownership Gap
Andrew Haughwout, Richard Peach, and Joseph Tracy
Recent years have seen a sharp rise in the number of negative equity
homeowners—those who owe more on their mortgages than their
houses are worth. These homeowners are included in the official
homeownership rate computed by the Census Bureau, but the savings
they must amass to retain their home or purchase a new home are
daunting. Recognizing that these homeowners are likely to convert
to renters over time, the authors of this analysis calculate an
“effective” rate of homeownership that excludes negative equity
households. They argue that the effective rate—5.6 percentage
points below the official rate—may be a useful guide to the future
path of the official rate.

omeownership is often seen as an integral part of the American dream, and
encouraging homeownership has historically been an important feature of U.S.
public policy. In 1995, the rate of homeownership in the United States began a
steep rise and between 2004 and 2006, peaked at 69 percent (Chart 1).1 The last three
years, however, have seen a marked reversal of this trend. As the housing boom collapsed and the recession fueled a sharp rise in unemployment, the homeownership
rate fell to 67.2 percent in the fourth quarter of 2009—its most recent reading and
a reversion to its second-quarter 2000 level. Strikingly, the ongoing decline in the
homeownership rate is approaching in magnitude the 2.3 percentage point slide
observed in the early 1980s.
A question of broad interest is how large the decline in the homeownership rate
will ultimately prove to be. In this edition of Current Issues, we assess the downward
pressure on this rate and introduce the notion of a “homeownership gap” as a useful
gauge of the possible extent of the rate’s decline over the next several years.
Our concept of a homeownership gap reflects the dramatic growth in the number
of negative equity homeowners—those who owe more on their mortgages than their
houses are worth—in the current housing market. While the official homeownership
rate tabulated by the Census Bureau includes negative equity homeowners in its count
of owner-occupied houses, our calculations suggest that these homeowners would
need to ramp up their savings by formidable amounts in order to retain their homes
or purchase a new home. Thus, we calculate an “effective” homeownership rate that
excludes negative equity homeowners from the sum of owner-occupied houses and
counts them instead as the renters they are likely to become over time. We find that
the difference between the official and the effective rates—the homeownership gap—

1 The U.S. Bureau of

the Census tabulates quarterly homeownership rates for the nation and for individual
states and metropolitan statistical areas. The measured home ownership rate is the ratio of the number of
owner-occupied housing units divided by the total number of occupied housing units. Second homes and
vacation homes are excluded from the calculation. In addition, properties that are currently vacant—even
if previously owned or rented—are also excluded. See http://www.census.gov/hhes/www/housing/hvs/
annual08/ann08ind.html.

CURRENT ISSUES IN ECONOMICS AND FINANCE ❖ Volume 16, Number 5

period, the government chartered a new financial institution
devoted to providing mortgage credit—the thrift—and created
the Federal Home Loan Bank System as a funding source that
would help thrift institutions manage the problems associated
with making fixed-rate loans scheduled to last for decades. After
World War II, the GI Bill established the Veterans Administration
(VA) mortgage program to provide veterans with high loan-tovalue mortgage loans insured by the federal government.

Chart 1

Aggregate Official Homeownership Rate
Percent
70
68
66
64
62
60
1965

Source: U.S. Bureau of the Census, Housing and Economic Statistics Division.

is significant, measuring 5.6 percentage points for the nation as a
whole and rising as high as 39 percentage points for the metropolitan areas that have been hit hardest by the housing crisis.
While such gaps have most likely existed before at the regional
level, the current national gap has no apparent precedent in the
postwar period.2
Taking our argument one step further, we contend that the
current effective homeownership rate is a good guide to the
future path of the official rate. That is, unless house prices increase substantially, many negative equity homeowners will
in fact convert to renters in the years ahead, and the measured
rate of homeownership will decline toward the effective rate.
We begin our analysis with a look at government initiatives to
encourage homeownership, followed by a discussion of the rationale for this support. If homeownership rates do indeed decline
in the coming years as we suggest, then the larger social benefits
that arise when individuals have an equity stake in their homes
and communities may be reduced.

Homeownership and Public Policy
Since at least as far back as President Roosevelt’s New Deal,
governments at the federal, state, and local levels have enacted
policies to encourage people to become and remain homeowners. In response to the surge in mortgage foreclosures during
the Great Depression, the government created the Federal Housing
Administration (FHA) and the Federal National Mortgage Association (FNMA, or Fannie Mae) to establish a standard mortgage
product—the thirty-year fixed-rate, fully amortizing mortgage—
that would allow borrowers to make modest fixed payments over
an extended period. Moreover, the FHA insured these mortgages,
thus limiting expected losses for investors. During the same
2 The gap is analogous in some ways to the developing country “debt overhang”
problem, which received extensive analysis in the late 1980s. See Sachs (1990) for
a discussion.

In the late 1960s and early 1970s, as thrift institutions came
under stress from rising inflation, the government played a
central role in the creation of the market for mortgage-backed
securities. The Government National Mortgage Association began
issuing federally guaranteed mortgage pass-through securities
backed by FHA and VA loans in 1970. Soon after, the Federal
Home Loan Mortgage Corporation (Freddie Mac) started issuing
mortgage participation certificates backed by conventional mortgages. Ultimately, the securitization of the bulk of new mortgage
loans fell to the government-sponsored enterprises Fannie Mae
and Freddie Mac, largely because of the implicit federal guarantee on the mortgage-backed securities and debt issued by these
institutions.3
The tax code is another channel through which homeownership is encouraged. For homeowners, the gross imputed income
from their home is not subject to taxation while the two major
expenses of owning a home—mortgage interest and property
taxes—are allowable itemized deductions. Moreover, most homeowners are now effectively exempt from taxes on capital gains
realized on the sale of their home(s). Another feature of the tax
code intended to spur homeownership is the ability of state and
local governments to issue tax-exempt mortgage revenue bonds.

The Benefits of Homeownership
The case for government support for homeownership rests in
large part on the view that ownership promotes “economically
efficient” actions—actions that produce the greatest return for
the resources invested. Because owners have a financial interest
in their property, they have incentives to take measures that will
maintain or increase the value of that property. Some of these
measures—such as fixing a leaky roof—are closely related to the
house itself. Others, such as investing resources in the betterment
of the neighborhood and the community, have broader beneficial
effects on the local area, creating what economists call “positive
externalities.” All of these measures will be reflected, or “capitalized,” in stable or rising home prices.
The notion that these capitalization effects prompt homeowners to act in the best interest of the property and the community underlies the “homevoter hypothesis” advanced by William
Fischel (2001). Asserting a close connection between homeownership and civic engagement (hence the term “homevoter”), Fischel
argues that homeowners take an active interest in the policy
decisions of the local government because these decisions affect
the long-term value of their property. Homeowners will support
3 See McCarthy and Peach (2002).

efficient public policies and projects—say, those that do the
most to enhance the quality of the services and schools in their
communities and thus to maximize the value of their homes—in
much the same way that a corporation’s shareholders will support
private projects that have a positive net present value for the firm.
However, the incentives that, in this view, motivate most homeowners will not operate for one subset of homeowners—negative
equity homeowners, or those whose mortgage balance exceeds
the value of their home. For these homeowners, any increase
in the value of their house will accrue not to them, but to the
mortgage lender (up to the value of the mortgage). Thus, with
little to gain, negative equity homeowners will be much less likely
to pursue improvements in their homes or communities. Their
situation is essentially analogous to that of renters, who have little
incentive to make improvements to the homes they occupy since
it is the landlord who reaps the economic benefits.
The homevoter hypothesis is compelling, but is there evidence
for the view that house price capitalization induces homeowners
to act in the best interests of the property and the community?
Researchers have documented that homeowners typically spend
several thousand dollars a year in maintenance and repairs to
offset the depreciation of their house over time (Gyourko and
Tracy 2006; Harding, Sirmans, and Rosenthal 2007). Conversely,
negative equity homeowners have been found to under-maintain
their property relative to other homeowners during regional
house price declines (Gyourko and Saiz 2004). Also consistent
with the homevoter hypothesis are studies showing that elderly
home-owners who have no school-age children still support local
education bond issues. While altruism may be a factor, the homeowners appear to be motivated mostly by a belief that backing
local schools will increase the value of their house (Bergstrom,
Rubinfeld, and Shapiro 1982; Hilber and Mayer 2009). Other
research has demonstrated that children of homeowners are more
likely to finish school than the children of renters and less likely
to become teenaged parents (Green and White 1997). Finally,
home-owners have been found to vote at higher rates in local
elections and to be more aware of local issues and the identities
of state and local civic leaders (DiPasquale and Glaeser 1999).4
To be sure, not all researchers are persuaded that homeownership leads to increased civic engagement or improved maintenance of homes and neighborhoods. Engelhardt et al. (2010)
maintain that the measured benefits from homeownership stem
from the fact that people who choose to buy homes are simply
more likely than others to value investing in social capital. Contending that the homevoter hypothesis and similar arguments
“overstate the impact of homeownership on political involvement,” the authors find that for the small sample of low-income
households in their study, the effect is “zero or negative.”

Still, although dissenting views exist, the preponderance of
research evidence at this point upholds the social benefits of
homeownership. And continuing public support for homeownership makes clear that policymakers regard the advantages for
neighborhoods and communities as substantial.

Equity and the Homeownership Gap
The role of house price capitalization in encouraging homeowners to support economically efficient actions depends on
the homeowner having positive equity in the house. For a homeowner in a negative equity position, this capitalization effect is
likely small or nonexistent. If we assume that the homeowner
will seek to move within five years,5 then unless that homeowner
either expects to be back in positive equity by the time of the
move or intends to use other assets to pay off the loan in full
upon sale of the property, changes in the value of the house will
only affect returns to the lender (or the investor, if the mortgage
has been securitized).6
The idea that having a positive equity stake in one’s house is
critical to the positive externalities from homeownership leads
us to propose an alternative way to measure the homeownership
rate. Specifically, we seek to calculate an effective homeownership rate, defined as the number of owner-occupied housing
units in which the household has a positive equity stake divided
by the total number of occupied housing units.7 This measure
of homeownership assumes that negative equity owners are, in
effect, renters—hence the notion of an effective homeownership rate. Owners with negative equity create a split between
the official homeownership rate compiled by the Census Bureau
and the effective homeownership rate—a split that we term the
homeownership gap.
Since the homeownership gap reflects the extent of negative
equity in the housing market, it is also a gauge of the potential
downward pressure on the official homeownership rate. Assuming that house prices do not appreciate over the next several
years, negative equity households will very likely convert to
renters when they move out of their current homes because
they will be unable to save enough to cover the negative equity,
the transaction costs of selling their existing home, and a down
payment on another home.8 As these transitions from owning to
renting take place, the homeownership gap will narrow, with the

5 According to the most recent U.S. census, nearly half

(47 percent) of all

homeowners moved in the last five years.
6 The extent to which the capitalization effect is shut off

may be a function of the

magnitude of the negative equity position.
7 Thus, we remove negative equity homeowners from the numerator of

the official

homeownership ratio but retain them in the denominator.
4 Recognition that property ownership carries with it particular interests is as old
as the republic itself. In Federalist 10, James Madison writes, “Those who hold and
those who are without property have ever formed distinct interests in society. Those
who are creditors, and those who are debtors, fall under a like discrimination.”

8 If

the homeowner either defaults on the mortgage or negotiates a short-sale with
the lender, then the damage to the homeowner’s credit will likely prevent him or
her from buying a house for several years, even if sufficient funds are available
for a down payment.

www.newyorkfed.org/research/current_issues

CURRENT ISSUES IN ECONOMICS AND FINANCE ❖ Volume 16, Number 5

Table 1

Chart 2

Number of Months Required for Debt Amortization
to Lower the Nonprime LTV Ratio to 94

Aggregate Official and Effective Homeownership Rates
Percent
70

Percentiles
Current LTV

10th

Official

25th

50th

75th

90th

> 100

145

198

242

> 105

120

161

208

249

> 110

114

137

174

216

256

68
66
64

Effective

Note: Calculations assume constant house prices.
62

official homeownership rate dropping toward the effective rate.9
In this sense, the effective homeownership rate is a useful guide
to the future course of the measured homeownership rate. Of
course, negative equity homes that come onto the market may be
purchased by individuals who are currently renters—an outcome
that would mitigate the effect on the official homeownership rate.
However, the number of foreclosed houses purchased by former
renters is likely to be limited.

Measuring the Extent of Negative Equity
To construct the effective homeownership rate, we need to estimate the extent of negative equity across local housing markets.
We start with loan-level data on nonprime mortgages from First
America LoanPerformance (LP) and on prime mortgages from
Lender Processing Services (LPS) Applied Analytics (formerly
McDash). These data indicate the loan-to-value (LTV) ratio for
each mortgage at origination.10 We update the loan-to-value ratio
by adjusting the loan amount(s) to account for debt amortization—the reduction in mortgage balances that accompanies
scheduled payments—and to reflect changes in the value of the
house as indicated by a repeat-sale price index for the metropolitan statistical area (MSA) or, if the property is located outside
an MSA, for the state.11 The house price data are updated quarterly, allowing us to construct a quarterly estimate of the current
LTV ratio for every mortgage in our data. We restrict our equity
calculations to owner-occupied primary residences since these
are the homes captured in the numerator of the Census Bureau’s
homeownership rate.
Having constructed the estimates of LTV ratios, we need to
specify the level of the current LTV ratio that is associated with
9 Public policy initiatives such as mortgage modification and the tax credit for
first-time home buyers can affect the speed of the decline in the official rate.
10 The LTV ratio is measured as the cumulative value of

the mortgage balance
across the first lien and any subordinate lien mortgages divided by the value of
the house. We capture subordinate liens for nonprime mortgages exclusively and
only if the lien was present at the origination of the first lien. If the value of the
mortgage(s) equals the value of the house, we set the LTV to equal 100 (rather
than a value of 1).
11 We use the Office of

Federal Housing Enterprise Oversight/Federal Housing
Finance Agency (OFHEO/FHFA) repeat-sale price indexes. A widely cited
alternative set of repeat-sale price indexes, discussed later in the article, are
the S&P/Case-Shiller indexes, which are available for only twenty MSAs.

60
2005

Sources: U.S. Bureau of the Census; LPS Applied Analytics and LP data;
authors’ calculations.

an owner behaving more like a renter. By convention, a mortgage
is judged to be in negative equity if the current LTV exceeds 100,
but a key consideration is the value of the current LTV that would
allow a household to break even when it eventually sells its home.
Therefore, some additional factors need to be taken into account.
First, we need to consider the transaction costs involved in
selling a house.12 If we assume that these costs amount to 6 percent of the sale price, then the LTV at the date of the sale would
need to be no higher than 94 for the household to break even on
the sale. Second, for mortgages whose current LTV is above 94, we
need to assess how long it would take to bring the ratio down to
94 through scheduled debt amortization, assuming no further net
changes in the price of the home.13 Our findings for nonprime
borrowers are presented in Table 1, which divides the negative
equity mortgages of this group into percentiles on the basis of the
number of months that would be required to bring the LTV down
to 94. The distribution reflects both the differing magnitudes of
negative equity and the remaining payment periods for mortgages in our data. Of the nonprime mortgages whose current LTV
is greater than 100, 90 percent would take longer than five years
to reach an LTV of 94 through the scheduled debt pay-down
process. The median mortgage in this group would take more
than twelve years to reach an LTV of 94. If we look at mortgages
with even higher current LTVs, the length of time required to
reach the break-even point would increase quite significantly.
For the purpose of constructing our alternative homeownership rate, we conclude that the incentives to behave like an owner
are very weak if the benefits from this behavior require living in
the house for more than five years. Thus, we identify a current
LTV of 100—that is, the standard definition of negative equity—
as our marker for households that are likely to behave more as
12 These include the fees to brokers as well as taxes and transfer fees.
13 This is consistent with house prices continuing to decline over the next year
but then recovering by the sale date.

Chart 3

Official and Effective Homeownership Rates: Metropolitan Statistical Areas Hit Hard by Volatile House Prices
Percent
80
Las Vegas
70

Percent
80
Miami
70

Census Bureau

FHFA

Case-Shiller

10
80

Census Bureau

10
80

Los Angeles

Phoenix

Census Bureau

FHFA

60
50

FHFA
40

Case-Shiller

30
20

10
2005

2005

Sources: U.S. Bureau of the Census, Current Population Survey; LPS Applied Analytics and LP data; authors’ calculations.
Note: The FHFA (Federal Housing Finance Agency) and Case-Shiller rates are effective rates.

renters than owners. Using this definition, we calculate quarterly
estimates of the number of owner-occupied prime residences in
negative equity over the period from the first quarter of 2005 to
the first quarter of 2009. We then subtract these negative equity
households from the quarterly counts of owner-occupied housing
units in the official homeownership rate to arrive at the aggregate
effective homeownership rate over the same four-year period.
The effective rate that we compute follows a very different
path than the official homeownership rate (Chart 2). The effective
rate begins to diverge from the official rate in 2006. This homeownership gap widens in 2007 as the pace of the house price
decline accelerates, pulling more households into negative equity.
By the end of fourth-quarter 2009, the effective homeownership
rate has fallen to 61.6 percent, creating a homeownership gap of
5.6 percentage points.
Significantly, the homeownership gap in Chart 2 may understate the true gap for two reasons. First, the price indexes that we
use to calculate the updated LTVs—repeat-sale indexes put out by
the Office of Federal Housing Enterprise Oversight (OFHEO) and
the Federal Housing Finance Agency (FHFA), hereafter termed the
FHFA indexes—have declined considerably less from their recent
peaks than have competing home price indexes. The methodology

used by FHFA to construct these price indexes involves measuring price changes for houses financed with prime, conforming
mortgages purchased by Fannie Mae and Freddie Mac at two or
more points in time. However, in many metropolitan areas in the
weakest housing markets, nonprime mortgages became much
more prevalent in the first half of this decade, while more recently
foreclosures have emerged as an important component of overall
housing transactions. In contrast to the FHFA indexes, the S&P/
Case-Shiller (hereafter Case-Shiller) repeat-sale price indexes
cover homes financed with nonprime as well as prime loans and
cover most foreclosure sales.14 The second reason that our estimate may understate the homeownership gap is that the coverage
of subordinate liens in our database is most likely incomplete,
since it excludes all subordinate liens on prime mortgages and
some subordinate liens on nonprime mortgages.15
The gap between the official and effective homeownership
rates is even more striking when we turn our attention from the
nation to metropolitan areas that experienced a severe collapse in
housing prices (Chart 3). Measured from the FHFA indexes, the
14 Specifically, the Case-Shiller methodology includes all “arms-length” housing
transactions.
15 See footnote 10.

www.newyorkfed.org/research/current_issues

CURRENT ISSUES IN ECONOMICS AND FINANCE ❖ Volume 16, Number 5

Table 2

Official and Effective Homeownership Rates for Large Metropolitan Statistical Areas
Official Homeownership Rate
(Percent)
Metropolitan Statistical Area

Peak

Current

Homeownership Gap
(Percentage Points)

Effective Homeownership Rate
(Percent)
FHFA

Case-Shiller

FHFA

Case-Shiller

Atlanta

70.8

66.4

61.3

57.4

5.1

9.0

Boston

67.7

66.5

64.2

64.3

2.3

Charlotte

69.1

68.0

66.9

63.2

1.1

4.8

Chicago

71.3

69.5

64.2

61.6

5.3

7.9

Cleveland

78.6

74.2

71.4

69.8

2.8

4.4

Dallas

64.5

62.8

62.4

61.6

0.4

1.2

Denver

72.0

63.3

60.8

59.8

2.4

3.5

Detroit

78.4

75.2

59.1

48.9

16.1

26.3

Las Vegas

65.0

58.6

19.3

14.7

39.3

43.9

Los Angeles

55.2

50.5

40.1

35.8

10.4

14.7

Miami

71.0

67.1

48.3

44.6

18.8

22.5
14.6

Minneapolis

74.8

71.2

64.1

56.5

7.0

New York

55.9

51.2

48.7

47.5

2.5

3.7

Phoenix

74.7

68.8

49.1

40.6

19.6

28.2

Portland

72.7

67.6

63.3

61.9

4.4

5.7

San Diego

63.3

55.3

39.3

35.0

16.0

20.3
14.9

San Francisco

61.7

58.3

49.1

43.4

9.2

Seattle

65.7

60.8

55.8

53.4

5.0

7.4

Tampa

74.1

67.6

56.1

51.2

11.4

16.3

Washington, D.C.

70.9

66.5

58.8

52.3

7.7

14.2

Sources: U.S. Bureau of the Census, Current Population Survey; LPS Applied Analytics and LP data; authors’ calculations as of 2009:Q3.

effective homeownership rates for Los Angeles, Miami, Phoenix,
and Las Vegas ranged from 10 to 39 percentage points below
the corresponding official rates in the third quarter of 2009.
The smallest homeownership gap—that for Los Angeles—was
almost double the size of the homeownership gap for the country
as a whole. Moreover, like the estimates of the national homeownership gap, these metro area estimates might understate
the difference between the official and effective rates. As Chart 3
makes clear, the MSA homeownership gaps calculated from the
Case-Shiller house price indexes are much larger than those
produced using the FHFA house price indexes.
Significantly, very large homeownership gaps are not confined
to just a few metro areas. The effective homeownership rates for
half of the metro areas covered in the Case-Shiller indexes are
at least 10 percentage points below the corresponding Census
Bureau homeownership rates (Table 2).

Implications of the Homeownership Gap
Earlier in this article, we suggested that homeownership gives
individuals a financial stake in the long-run outlook for their
homes and communities. If this is the case, then the homeowner-

ship gaps that we have documented for the nation and some
metro areas may have significant implications for civic welfare.
Consider, for example, that the Case-Shiller-based effective
homeownership rates for the four metro areas shown in Chart 3
and for Detroit, New York City, San Diego, and San Francisco
(Table 2) are all under 50 percent. That is, the median household
in these areas is in a negative equity position and no longer has
strong financial incentives to behave as an owner. While the
effects will vary with the distribution of negative equity households across the municipalities within these metro areas, a high
share of these households could result in reduced maintenance
of the housing stock, an increased risk of housing vacancies,
and less stable neighborhoods over time—developments that
could have repercussions for local law enforcement.16 Moreover,
the predominance of “non-homeowners” in these metropolitan
areas could lead to a decline in citizen participation in local
affairs, with a concomitant loss of vigilance over the quality
and efficiency of public services and institutions.

16 See, for example, Millman (2009).

By contrast, households whose regular debt amortization will not
reduce the mortgage balance sufficiently will need to save enough
to pay off the current mortgage before buying again.

Table 3

Savings Required to Remain an Owner
If Moving in Five Years

All borrowers with LTV > 100
Borrowers with 100 < LTV < 111

Monthly Savings per
Household (Dollars)

Total Annual Savings
(Billions of Dollars)

1,222

92.3

602

22.7

Sources: LPS Applied Analytics and LP data; authors’ calculations as of 2009:Q4.
Note: The total annual savings are for the full sample of negative equity households.

The large homeownership gaps that have emerged during this
housing market crisis will likely have significant effects on the
macroeconomy as well. One possible consequence is an increase
in the national saving rate. Homeowners seeking to escape a
negative equity mortgage and purchase a new residence will need
to make a substantial commitment to save. They must remain
current on their mortgage payments and pay off any remaining
negative equity balances upon the sale of their current home.
In addition, they will need to provide cash to cover the down
payment on a new home as well as the transaction costs of the
purchase. Given the large number of households currently in
negative equity, a broad-based movement among these households to increase saving would have the potential to boost the
nation’s savings significantly.
To shed light on the magnitude of this increase, we estimate
how much the negative equity households in our sample would
need to save in order to close out their existing mortgage and
buy a new home. The amount will depend, of course, on the value
of each household’s current and prospective homes, the lending
standards in effect at the time the household moves, and the cost
of the transaction. For our analysis, we assume that the household’s “desired” down payment equals 20 percent of the current
value of its existing house,17 and that transaction costs total
6 percent of that value.
Note that even absent any house price appreciation, homeowners who remain current on their mortgage payments build
their equity position through debt amortization. For each
negative equity homeowner in our sample, we can project the
reductions in debt balances that result from making the scheduled payments for a given period of time, and then incorporate
these reductions in our analysis. Of the households that continue
to make payments, more than a third (36 percent) will assume a
positive equity position within three years, and more than half
(51 percent) within five years.18 For these borrowers, housing
equity could serve as part of a down payment on a new home.
17 Our assumption would allow the household to purchase a residence of

equivalent value under the current tight lending standards. Since a new home
could be more expensive, this is a conservative assumption.

Table 3 reports the net savings required for the average negative equity household in our sample to buy a new home in five
years. Again, these figures represent the sum of the amounts
required to make a new down payment, pay all transaction
costs, and pay off (or receive) the difference between the current
house price and the mortgage balance at the time of sale. Even
accounting for the benefits of debt amortization on the borrower’s equity position, we find that the typical household must save
more than $1,200 more per month if it wishes to buy again in
five years. (For a detailed example of the calculations underlying
Table 3, see the appendix.) Because we estimate that more than
6 million households are in negative equity, these figures imply
an annual savings increase for the nation of $92 billion for five
years. Personal saving as defined in the National Income and
Product Accounts averaged roughly $465 billion during 2009,
yielding an average personal saving rate of 4.3 percent. All else
equal, we calculate that for these borrowers to remain homeowners under our assumptions, personal saving must rise about
20 percent a year for five years. The personal saving rate would
have to rise about 0.8 percentage points, to 5.1 percent.
Since the savings required are so large at both the household
and aggregate level, it seems unlikely that all of today’s negative
equity households will be able to remain owners unless they
defer moving for several years. The second row of Table 3 reports
similar figures for the “better” half of the negative equity distribution—that is, homeowners with LTVs below 111, whose chances
of remaining owners seem more realistic. Even here, however, the
average monthly saving requirement, at $602, is quite large.
Another implication of the homeownership gap for the larger
economy is that household mobility is likely to be significantly
reduced. Negative equity households that are saving for a new
down payment need to delay a move during the period they are
rebuilding their savings. Studies of past regional housing cycles
suggest that household mobility may fall by as much as a third
for households in a negative equity position.19 Recent Census
Bureau data confirm the downward trend in mobility, putting the
number of households moving at its lowest level since the 1960s.
While many factors are likely weighing on household mobility
now, the prevalence of negative housing equity is surely high
on the list.
It is hard to predict with much precision how the homeownership gap will ultimately affect measured homeownership, savings,
and mobility. Our analysis suggests that either savings must rise
and mobility must fall or, more likely, the official homeownership
rate will decline toward the effective rate, narrowing the homeownership gap.

18 Of

course, house price appreciation would hasten this process of equity gains,
while continued price declines would slow it.

19 See, for example, Ferreira, Gyourko, and Tracy (2008).

www.newyorkfed.org/research/current_issues

CURRENT ISSUES IN ECONOMICS AND FINANCE ❖ Volume 16, Number 5

Conclusion
The severe decline in house prices in the last few years, combined
with the large number of borrowers who had little or no equity at
the origination of their mortgages, has led to a dramatic rise in
homeowners with negative equity. This rise in turn has opened
a large gap between the Census Bureau’s official homeownership
rate and a measure that we term the effective rate. The effective
rate recognizes that negative equity homeowners are likely to convert to renters over time and thus excludes them from the count
of owner-occupied housing. The effective homeownership rate
for the nation is currently 5.6 percentage points below the Census
Bureau rate, and in some of the metropolitan areas hurt most by
the housing crisis, the effective homeownership rate falls short
of the official rate by a striking 20 to 39 percentage points.
Public policy has long promoted homeownership, and subsidies for owner-occupants are a key feature of the tax code. But
these recent developments present many challenges to policymakers. Absent any action, the high saving requirements for
remaining an owner make it likely that the current effective
homeownership rate will foreshadow the future official rate.
A drop in the homeownership rate may create a large set of
residents who are less invested in the long-run outlook for their
homes and communities—an outcome that could lead to lower

levels of home maintenance and civic participation, as well as
more short-sighted decisions in local affairs. While the national
saving rate may well rise as negative equity households who prefer
to own their own home try to save up a down payment on a new
house, the task of setting aside sufficient funds will be daunting
for these households.
Public policy initiatives such as mortgage modification can
help to support the homeownership rate by reducing foreclosures
and easing conditions for negative equity borrowers to save for
a future down payment. However, the efficacy of these modification programs depends in part on their structure. Programs that
encourage principal write-down will do more to support the
homeownership rate than those that focus solely on the monthly
mortgage burden to the borrower, and will allow maintenance
of homeownership without producing steep declines in consumption.20 Addressing the problems of negative equity and
low effective homeownership rates is most important for those
metropolitan areas that suffered the worst house price declines.
The current large homeownership gaps in these housing markets
will make it especially difficult to maintain the broader social
benefits that stem from a high homeownership rate.
20 See the comparison of

mortgage modification programs in the appendix.

Appendix: When Negative Equity Mortgage Holders Save for a New Home
For negative equity mortgage holders, remaining a homeowner
requires a substantial saving commitment, but mortgage
modifications—particularly those that reduce the principal
balance—can help.
Negative equity borrowers who want to remain owners but
already have difficulty meeting their mortgage payments may
find that saving for a down payment on a new home is not
feasible. Mortgage modification programs can assist these
households, to a degree, by reducing the required monthly
mortgage payment, thus freeing up funds that can be saved for
a new down payment. But the structure of the modification
program is important. Modifications that reduce interest rates
alone will lower the monthly payment, while those that also
reduce principal balances lower the monthly costs and provide
for additional saving through debt reduction.
For example, consider a household whose home is currently worth $181,818 (see the first column in the top panel of
the appendix table). The household has a nonprime thirtyyear fixed-rate mortgage at a 7 percent interest rate that was
originated two and a half years ago, and has a current balance
of $200,000.1 The household’s monthly income is $4,474. The
required monthly mortgage payment is $1,367, and the monthly
taxes and insurance are $333. This gives the household a fairly
high ratio of debt service to income (DTI) of 38 percent, so this
household is financially stretched in its current mortgage.
Now assume that the household would like to buy a new
home in five years and that the value of its current house will
not change over this period. To be able to make a 20 percent
down payment on a new house of equivalent value, the household needs to accumulate $36,364. The household also anticipates that the sale of its existing home will entail a 6 percent
transaction cost, or $10,909. The household is currently in a
negative equity position of $18,182; five years of payments on
the original mortgage will reduce its negative equity to $3,823.
To be able to sell the house, pay off its mortgage, and make a
down payment on a new house, the household must accumulate $51,096 in savings.
Assuming that the household tries to save this amount
over a five-year period and that it earns 1.6 percent on its savings, it would have to set aside an additional $819 per month.
This additional claim on the household’s income would raise

its DTI to 56 percent—a level that would necessitate a significant reduction in consumption and is likely to be unsustainable. Even if the household is not straining to meet its current
mortgage payments (if, say, it has a DTI of 28 percent rather
than 38 percent), saving to remain a buyer would push its
DTI to a high level (46 percent).2
Now, consider the benefit to the household if it qualifies
for a loan modification program. Suppose that there are two
programs that target a DTI of 31 percent and so reduce the
monthly payment from $1,367 to $1,049. The first program
accomplishes this by reducing the interest rate to 4.8 percent
and extending the mortgage term an additional thirty months,
to thirty years. The household remains in a negative equity
position, but the lower interest rate allows the household to
build equity slightly more quickly, so that after five years the
remaining mortgage balance will exceed the house value by
$1,312. If the household wants to save for a new down payment
over this five-year period, it must accumulate $48,585, for an
effective DTI of 48 percent—lower than the 56 percent without
the loan modification, but still quite high.
The second modification program, like the first, lowers the
interest rate on the existing mortgage and extends the term of
the loan; in addition, however, it reduces the principal balance
to the current value of the house. Under this program, the principal declines by $18,182 and the new interest rate is 5.6 percent. The new monthly payment is the same as under the first
modification program. To save for a new down payment over
a five-year period, the household must accumulate $33,885—
markedly less than under the first program. Moreover, this
amount of required saving would raise the household’s effective DTI to 43 percent—again, a level lower than the 48 percent
under the interest-rate-only modification program.
Clearly, a loan modification program that lowers the principal balance on a mortgage will do more to support homeownership than a program that simply eases the terms of the
loan. And the demand it places upon a household to cut consumption is appreciably less than that imposed by the interestrate-only program. Still, even a reduced DTI of 43 percent will
leave households financially stretched, and it is unlikely that
many negative equity mortgage holders will be able to sustain
the high rate of saving needed to remain a homeowner.

1 In this example, then, the current LTV is 110, very close to the median LTV

2 This example assumes that the household has no other financial assets that

(111) among negative equity mortgages in the fourth quarter of 2009.

it can use to help fund its next purchase.

www.newyorkfed.org/research/current_issues

CURRENT ISSUES IN ECONOMICS AND FINANCE ❖ Volume 16, Number 5

Appendix Table

Modifying Negative Equity Mortgages for Affordability
Dollars Except as Noted
Household and Mortgage Characteristics

Original

Modification Program 1

House value

181,818

Mortgage balance

200,000

181,818

Interest rate (percent)

Modification Program 2

7.0

4.8

5.6

Mortgage principal, interest, taxes, and insurance (PITI)

1,700

1,382

Monthly income

4,474

Debt service–to–income (DTI) ratio (percent)
Saving for a New Down Payment

Original

Modification Program 1

Modification Program 2

Borrower equity after five yearsa

(3,822.83)

(1,312.06)

13,387.86

Down payment to buy a house of this price

36,363.64

Transaction cost at 6 percent

10,909.09

Savings required to buy again in five years

51,095.55

48,584.78

33,884.86

818.55

778.33

542.84

56.3

48.3

43.0

Savings per month (over five years, assuming 1.6 percent interest rate)
“Full” housing cost–to–income ratio (percent)b
a Values presented assume no house price growth.

b Full housing cost includes both the mortgage PITI and the savings required to purchase a new home of

References
Bergstrom, Theodore C., Daniel L. Rubinfeld, and Perry Shapiro. 1982.
“Micro-Based Estimates of Demand Functions for Local School Expenditures.”
Econometrica 50, no. 5 (September): 1183-1205.
DiPasquale, Denise, and Edward L. Glaeser. 1999. “Incentives and Social Capital:
Are Homeowners Better Citizens?” Journal of Urban Economics 45, no. 2 (May):
354-84.
Engelhardt, Gary, Michael Eriksen, William Gale, and Gregory Mills. 2010. “What
Are the Social Benefits from Homeownership? Experimental Evidence for LowIncome Households.” Journal of Urban Economics 67, no. 3 (May): 249-58.
Ferreira, Fernando, Joseph Gyourko, and Joseph Tracy. 2008. “Housing Busts
and Household Mobility.” Federal Reserve Bank of New York Staff Reports,
no. 350, October.
Fischel, William A. 2001. The Homevoter Hypothesis: How Home Values Influence
Local Government Taxation, School Finance, and Land-Use Policies. Cambridge:
Harvard University Press.
Green, Richard, and Michelle White. 1997. “Measuring the Benefits of Homeowning: Effects on Children.” Journal of Urban Economics 41, no. 3 (May): 441-61.

equal value in five years.

Gyourko, Joseph, and Albert Saiz. 2004. “Reinvestment in the Housing Stock:
The Role of Construction Costs and the Supply Side.” Journal of Urban
Economics 55, no. 2 (March): 238-56.
Gyourko, Joseph, and Joseph Tracy. 2006. “Using Home Maintenance and Repairs
to Smooth Variable Earnings.” Review of Economics and Statistics 88, no. 4
(November): 736-47.
Harding, John C., C. F. Sirmans, and Stuart S. Rosenthal. 2007. “Depreciation
of Housing Capital, Maintenance, and House Price Inflation.” Journal of Urban
Economics 61, no. 2 (March): 193-217.
Hilber, Christian A. L., and Christopher Mayer. 2009. “Why Do Households
without Children Support Local Public Schools? Linking House Price Capitalization to School Spending.” Journal of Urban Economics 65, no. 1 (January): 74-90.
McCarthy, Jonathan, and Richard W. Peach. 2002. “Monetary Policy Transmission
to Residential Investment.” Economic Policy Review 8, no. 1 (May): 139-58.
Millman, Joel. 2009. “Immigrants Become Hostages as Gangs Prey on Mexicans.”
Wall Street Journal, June 10.
Sachs, Jeffrey. 1990. “A Strategy for Efficient Debt Reduction.” Journal of Economic
Perspectives 4, no. 1 (winter): 19-29.

ABOUT THE AUTHORS
Andrew Haughwout is a vice president in the Microeconomic and Regional Studies Function and Richard Peach a senior vice
president in the Macroeconomic and Monetary Studies Function of the Federal Reserve Bank of New York; Joseph Tracy is
an executive vice president and senior advisor to the Bank’s president.
Current Issues in Economics and Finance is published by the Research and Statistics Group of the Federal Reserve Bank of New York.
Linda Goldberg and Charles Steindel are the editors.

The views expressed in this article are those of the authors and do not necessarily reflect the position
of the Federal Reserve Bank of New York or the Federal Reserve System.
10

RECENT FEDERAL RESERVE BANK OF NEW YORK RELEASES ON THE HOUSING MARKET

A longer, more technical version of this Current Issues article
appeared in the Research and Statistics Group’s working paper
series. See Andrew Haughwout, Richard Peach, and Joseph Tracy,
“The Homeownership Gap,” Federal Reserve Bank of New York
Staff Reports, no. 418, December 2009.

Bypassing the Bust: The Stability of Upstate New York’s
Housing Markets during the Recession

Other recent New York Fed publications and papers consider additional dimensions of the housing crisis: the rating of
mortgage-backed securities, the regional experience of house
price volatility, and the effects of mortgage modification on
re-default rates:

Over the past decade, the United States has seen real estate
activity swing from boom to bust. But upstate New York has been
largely insulated from this volatility, with metropolitan areas such
as Buffalo, Rochester, and Syracuse even registering home price
increases during the recession. An analysis of upstate housing
markets over the most recent residential real estate cycle indicates
that the region’s relatively low incidence of nonprime mortgages
and the better-than-average performance of these loans contributed to this stability.

MBS Ratings and the Mortgage Credit Boom
Adam Ashcraft, Paul Goldsmith-Pinkham, and James Vickery
Federal Reserve Bank of New York Staff Reports, no. 449,
May 2010
The authors study credit ratings on subprime and Alt-A mortgagebacked securities (MBS) deals issued between 2001 and 2007,
the period leading up to the subprime crisis. They find that the
amount of credit enhancement increases with the amount of
mortgage credit risk (measured either ex ante or ex post),
suggesting that ratings contain useful information for investors.
However, the authors also find evidence of significant time
variation in risk-adjusted credit ratings, including a progressive
decline in standards around the MBS market peak between the
start of 2005 and mid-2007. They observe, conditional on initial
ratings, underperformance (high mortgage defaults and losses
and large rating downgrades) among deals with observably
higher-risk mortgages based on a simple ex ante model and
deals with a high fraction of opaque low-documentation loans.
These findings hold over the entire sample period, not just for
deal cohorts most affected by the crisis.

Jaison Abel and Richard Deitz
Federal Reserve Bank of New York Current Issues in Economics
and Finance 16, no. 3, March 2010

Second Chances: Subprime Mortgage Modification
and Re-Default
Andrew Haughwout, Ebiere Okah, and Joseph Tracy
Federal Reserve Bank of New York Staff Reports, no. 417,
December 2009
Mortgage modifications have become an important component
of public interventions designed to reduce foreclosures. This
paper examines how the structure of a mortgage modification
affects the likelihood of the modified mortgage re-defaulting over
the next year. Using data on subprime modifications that precede
the government’s Home Affordable Modification Program, the
authors focus their attention on those modifications in which
the borrower was seriously delinquent and the monthly payment
was reduced as part of the modification. The data indicate that
the re-default rate declines with the magnitude of the reduction
in the monthly payment, but also that the re-default rate declines
relatively more when the payment reduction is achieved through
principal forgiveness as opposed to lower interest rates.

INFORMATION ON THE FORECLOSURE CRISIS AVAILABLE ON NEW YORK FED WEBSITE

The Federal Reserve Bank of New York’s U.S. Credit Conditions
website (data.newyorkfed.org/creditconditions) offers detailed,
timely data on the incidence of mortgage foreclosures and delinquencies in the nation and in individual states and counties.
The information, presented through charts, interactive maps,
and spreadsheets, is designed to help government agencies, community groups, commercial institutions, and other practitioners
better understand and respond to local conditions associated
with failed and troubled mortgages.
The site offers a range of informative features. Visitors can compare
delinquency rates across geographical areas and across types of
mortgages—for example, prime, subprime, or Fannie Mae and

Freddie Mac loans. Red and green “heat maps” illustrate whether
conditions have worsened or improved over the past year. In addition, a sequence of charts shows the likelihood that subprime and
Alt-A mortgages will roll from their current status to thirty days
late, from sixty to ninety days late, or from ninety days late to
foreclosure. The roll rates are presented in terms of the number
of mortgages likely to roll from one status to the next and in terms
of dollar volumes.
The goal of the U.S. Credit Conditions website is to provide information that will help public and private sector decision makers
identify the best strategies to resolve the delinquency and foreclosure problem and to mitigate its impact on communities.