The Great Recession and Credit Trends across Income Groups
01 May 2015-The American Economic Review (American Economic Association 2014 Broadway, Suite 305, Nashville, TN 37203)-Vol. 105, Iss: 5, pp 147-153
TL;DR: In this article, the authors investigate trends in access to different credit markets, including mortgages, home equity, automobiles, and student loans, in the period surrounding the Great Recession, and disentangle growth rates of new market entrants from the aggregates and analyze overall as well as within-county growth rate differentials across income strata.
Abstract: In this paper, we document trends in credit use across income groups in the period surrounding the Great Recession. We investigate trends in access to different credit markets, including mortgages, home equity, automobiles, and student loans. We disentangle growth rates of new market entrants from the aggregates and analyze overall as well as within-county growth rate differentials across income strata. Our findings may provide insight into the financial well-being of different income groups in the context of the Great Recession.
Summary (1 min read)
Jump to: [Cumulative growth rate differentials in all credit markets] and [Within-county growth rates]
Cumulative growth rate differentials in all credit markets
- The authors augment Figure 3 in the paper with the results for home equity lines of credit.
- For convenience, the authors keep only the differentials for home mortgages and lines of credit.
Within-county growth rates
- In particular, following Mian and Sufi (2009), the authors calculate annual growth rates in total indebtedness for each zip code, regress those growth rates on a series of county-year dummies, and explore the patterns of the residual credit growth, by income quintile.
- Figure A2 shows the result of this exercise for home mortgages, presented as the cumulative growth rate differential.
- The comparison of this series with the solid line in Figure A1 Recession.
- The results for the home equity and auto loans lead to similar conclusions.
- This series exhibits a persistent pattern of outsized relative growth for the bottom quintile zip codes.
Did you find this useful? Give us your feedback
1
Web Appendix for
The Great Recession and Credit Trends across Income Groups
By GENE AMROMIN AND LESLIE MCGRANAHAN
We use data from two different sources—zip-code-level data from the IRS on average
adjusted gross income (AGI) per return and individual data from the Federal Reserve Bank of
New York (FRBNY) Equifax Consumer Credit Panel (CCP). The CCP/Equifax data consist of
detailed Equifax credit-report data for a 5 percent sample of individuals with credit reports and
Social Security numbers. The CCP data are quarterly and available since 1999. We use the
CCP/Equifax data to compute aggregate measures of credit for households in different zip codes
and divide zip codes into income quintiles based on their 2001 average AGI.
Zip-level income statistics are compiled by the Internal Revenue Service and are made
available through the Statistics of Income (SOI) Tax Stats publication in select years.
1
In order to
be able to track transition of zip codes across income quintiles, we focus on a balanced panel of
zip codes that are present in every SOI release between 2001 and 2012. The resulting set of
24,753 zip codes is then partitioned into equal-sized quintiles in each year based on their average
AGI in that year. We choose the 2001 quintile values as the basis for our analysis. As reported in
Table A1 below, the average AGI in the top quintile zip codes is nearly 3 times higher than the
average AGI in the bottom quintile, and the rate of ownership of assets that generate taxable
interest payments is more than twice as high.
Since we rely on a static assignment of zip codes to income quintiles, we are concerned
about zip code transitions across quintiles during our sample timeframe. We first verify that the
1
SOI Tax Stats - Individual Income Tax Statistics - ZIP Code Data (SOI) are available at:
http://www.irs.gov/uac/SOI-Tax-Stats-Individual-Income-Tax-Statistics-ZIP-Code-Data-(SOI)
2
relative average characteristics of quintiles based on the 2001 income did not change appreciably
by 2012. The bottom two rows in Table A1 show that the two extreme quintiles vary just as
much in their 2012 average AGI and propensities of owning interest-generating assets as they did
in 2001. We further drop from the sample those zip codes that moved by more than two income
quintiles between 2001 and 2012. This eliminates 4.8 percent of the 4,951 zip codes in the bottom
quintile because they have transitioned to quintiles 4 or 5 (top) by 2012, as well as 0.4 percent of
the 4,950 zip codes in the top quintiles that fell to quintiles 1 or 2. We also drop zip codes that
experienced a substantial change in their geography between the 2000 and 2010 Censuses.
2
Most
of these transitions occurred because of changes in geographic boundaries of zip codes, which
evolve over time as the U.S. Postal Service re-optimizes mail delivery routes. In some cases,
transitions occur as a result of gentrification of inner-city neighborhoods.
TABLE A1. SELECT STATISTICS FOR ZIP CODES IN THE TOP AND BOTTOM INCOME QUINTILES
Notes: The income quintiles are defined on the basis of average 2001AGI values. Select characteristics of the
resulting top and bottom quintiles are shown for 2001 and 2012 and are expressed in constant 2012 dollars. The
2012 calculations are made before dropping zip codes that moved by more than two quintiles.
Cumulative growth rate differentials in all credit markets
We augment Figure 3 in the paper with the results for home equity lines of credit. For
convenience, we keep only the differentials for home mortgages and lines of credit. The trends in
2
We define a substantial change in geography in one of two ways: growth or shrinkage in land area in the top or
bottom two percent of the distribution of changes in land area or a top two percent move in the location of the zip
code’s centroid.
Tax year Income quintile
Average AGI (2012
dollars)
Fraction with reported
taxable interest income
Fraction moved up/down by
more than 2 categories by 2012
2001 1 (bottom) $31,418 30% 4.8%
2001 5 (top) $91,186 65% 0.4%
2012 1 (bottom) $31,716 16%
2012 5 (top) $101,021 44%
3
the two housing-based credit markets look remarkably similar, although they differ in levels.
With the exception of the housing bubble period (2004-2007), home equity based credit grew
faster in highest income zip codes.
FIGURE A1. CUMULATIVE DIFFERENTIALS IN GROWTH RATES, HOUSING MARKETS TOP AND BOTTOM
QUINTILES
Within-county growth rates
In particular, following Mian and Sufi (2009), we calculate annual growth rates in total
indebtedness for each zip code, regress those growth rates on a series of county-year dummies,
and explore the patterns of the residual credit growth, by income quintile. Figure A2 shows the
result of this exercise for home mortgages, presented as the cumulative growth rate differential.
The comparison of this series with the solid line in Figure A1 suggests that much of the growth
differential has indeed occurred within individual counties. Outstanding mortgage debt in poor
zip codes in a given county expanded faster during the housing price boom relative to high-
income zip codes in the same county, and collapsed much more rapidly during the Great
-0.50
-0.40
-0.30
-0.20
-0.10
0.00
-0.15
-0.10
-0.05
0.00
0.05
0.10
2001q1
2001q3
2002q1
2002q3
2003q1
2003q3
2004q1
2004q3
2005q1
2005q3
2006q1
2006q3
2007q1
2007q3
2008q1
2008q3
2009q1
2009q3
2010q1
2010q3
2011q1
2011q3
2012q1
2012q3
2013q1
2013q3
2014q1
2014q3
Annualized growth rate
mortgage growth in low relative to high income zip codes (left axis), cumulative
home equity growth in low relative to high income zip codes (right axis), cumulative
4
Recession. The results for the home equity (in Figure A2) and auto loans (in Figure A3) lead to
similar conclusions. The dashed line in Figure A3 shows the within-county growth differential
for student loans. This series exhibits a persistent pattern of outsized relative growth for the
bottom quintile zip codes.
F
IGURE A2. WITHIN-COUNTY GROWTH RATES DIFFERENTIALS, HOUSING MARKETS
Notes: The figure plots the cumulates the differenial growth rates between the top and bottom quintile zip codes in
the same county for mortgages (solid line) and home equity lines of credit (dashed line).
FIGURE A3. WITHIN-COUNTY GROWTH RATES DIFFERENTIALS, NON-HOUSING MARKETS
-0.45
-0.4
-0.35
-0.3
-0.25
-0.2
-0.15
-0.1
-0.05
0
-0.12
-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
within-county relative mortgage growth (left axis), cumulative
within-county relative home equity growth (right axis), cumulative
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
-0.16
-0.14
-0.12
-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
within-county relative auto growth (left axis), cumulative
within-county relative student loan growth (right axis), cumulative
5
Notes: The figure plots the cumulates the differenial growth rates between the top and bottom quintile zip codes in
the same county for auto loans (solid line) and student loans (dashed line).
The apparent existence of substantial and systematic income-group heterogeneity in
credit trends within individual counties can be illustrated on the example of student loans in
Cook County, which includes the City of Chicago and some of its neighboring suburbs. The map
in Figure A4 depicts average annual growth rates in zip-level outstanding student loan balances
between 2005 and 2014. The income quintile of each zip code is indicated by numbers 1 through
5. While the relationship between income and credit growth is far from uniform, the map shows
that many of the high-quintile zip codes exhibited relatively subdued rates of growth. In contrast,
many of the poorest inner-city zip codes experienced rapid expansion in student loans, indicated
by deeper shading.
FIGURE A4. GROWTH RATES IN STUDENT LOAN BALANCES IN COOK COUNTY, ILLINOIS, 2005-2014.
Citations
More filters
••
TL;DR: In this article, the authors used two comprehensive micro-data sets to study how the distribution of mortgage debt evolved during the 2000s housing boom and found that the allocation of mortgage loans across the income distribution remained stable, as did the allocation for real estate assets.
Abstract: In this paper, we use two comprehensive micro-data sets to study how the distribution of mortgage debt evolved during the 2000s housing boom. We show that the allocation of mortgage debt across the income distribution remained stable, as did the allocation of real estate assets. Any theory of the boom must replicate these facts, and a general equilibrium model shows that doing so requires two elements: (1) an exogenous shock that increases expected house price growth or, alternatively, reduces interest rates and (2) financial markets that endogenously relax borrowing constraints in response to the shock. Empirically, the endogenous relaxation of constraints was largely accomplished with subprime lending, which allowed the mortgage debt of low-income households to increase at the same rate as that of high-income households.
30 citations
••
TL;DR: In this paper, the authors examined the differences in racial disparities between white and black home seekers in the type of loan products they are offered in the marketplace, and found that white and Asian Latinos performed better than other Latinos.
Abstract: Studies of the mortgage industry’s impact on racial stratification have long focused on racial disparities found between white and black homeowners. Ample research demonstrates that unequal access and treatment between white and black home seekers has created major differences in the type of loan products they are offered in the marketplace. While numerous studies also document disadvantaged Latino homebuyers, studies have yet to examine racial variation within the Latino population. This paper draws on annual data from the Home Mortgage Disclosure Act (HMDA) from 2010 to 2017 to assess variation in racial disparities among Latinos in loan outcomes and compares them to Non-Latino whites, blacks, Asians, and Others. I show that loan rejections and high cost originations are highest among black Latinos and lowest among white and Asian Latinos. Other Latinos perform somewhere in the middle. These trends are particularly true when examining mortgage denials. When comparing Latino racial groups to Non-Latinos, the observed lending patterns provide evidence of a tri-racial hierarchy in the mortgage market.
10 citations
••
TL;DR: In this paper , the authors show that interracial couples involving a black or Latino co-applicant are more likely to experience a high-cost loan or be denied a mortgage than mono-racial white couples.
6 citations
••
TL;DR: In this article, the effect of perceived utilities on attitude differ for existing and potential users, and also for cases of B2C and P2P car-sharing, and the findings of this study have provided managerial, theoretical, and policy implications.
Abstract: The sharing economy, facilitated by technology and social media, has grown rapidly across the world. The sharing economy is still in a fledgling phase, so relatively little research has been carried out on it. By classifying the existing and potential users, the purpose of the study is to investigate effects of perceptions on utilities, attitude, satisfaction, loyalty, and intention to use for both cases of B2C and P2P car-sharing. Study 1 examines the effects of the level of utility related to the levels of satisfaction and loyalty to existing users and intentions to use and expected satisfaction to potential users of B2C car-sharing, while Study 2 investigates the effects of their intention and levels of expected satisfaction to potential users of P2P car-sharing. This study collected the data via online and offline survey and applied statistical analyses including factor analysis, regression analysis, and ANOVA. The results revealed that the effect of perceived utilities on attitude differ for existing and potential users, and also for cases of B2C and P2P car-sharing. The findings of this study have provided managerial, theoretical, and policy implications.
5 citations
Cites background from "The Great Recession and Credit Tren..."
...With this competition, people were prone to expand their credit without hesitation, which was one of the main reasons for the Great Recession (De Nardi, French, & Benson, 2011; Amromin & McGranahan, 2015)....
[...]
•
TL;DR: The authors argues that the economic benefits of low gasoline prices for the U.S. economy have fallen substantially since the reemergence of America as a major oil producer and that consumers are not spending the windfall like they used to because of higher debt levels, limited access to credit, slow wage growth, and an older population.
Abstract: The author argues that the economic benefits of low gasoline prices for the U.S. economy have fallen substantially since the reemergence of America as a major oil producer. The old rule-ofthumb that a 10% fall in the oil price raises inflation-adjusted U.S. GDP by 0.2% is too large—the impact on economic activity should be closer to zero, and may even be negative if consumption grows slowly. The reasons for this change are straightforward, if underappreciated: (i) the value of oil production accounts for a larger share of the U.S. economy; and (ii) consumers are not spending the windfall like they used to because of higher debt levels, limited access to credit, slow wage growth, and an older population. (Submitted as Policy Paper)
3 citations
References
More filters
••
TL;DR: In this paper, the authors conduct a within-county analysis using detailed ZIP code-level data to document new findings regarding the origins of the biggest financial crisis since the Great Depression, finding that the sharp increase in mortgage defaults in 2007 is significantly amplified in subprime ZIP codes, or ZIP codes with a disproportionately large share of subprime borrowers as of 1996.
Abstract: We conduct a within-county analysis using detailed ZIP code—level data to document new findings regarding the origins of the biggest financial crisis since the Great Depression. The sharp increase in mortgage defaults in 2007 is significantly amplified in subprime ZIP codes, or ZIP codes with a disproportionately large share of subprime borrowers as of 1996. Prior to the default crisis, these subprime ZIP codes experience an unprecedented relative growth in mortgage credit. The expansion in mortgage credit from 2002 to 2005 to subprime ZIP codes occurs despite sharply declining relative (and in some cases absolute) income growth in these neighborhoods. In fact, 2002 to 2005 is the only period in the past eighteen years in which income and mortgage credit growth are negatively correlated. We show that the expansion in mortgage credit to subprime ZIP codes and its dissociation from income growth is closely correlated with the increase in securitization of subprime mortgages.
1,217 citations
•
01 May 2010TL;DR: Raghuram Rajan was one of the few economists who warned of the global financial crisis before it hit as discussed by the authors, and he argues that serious flaws in the economy are also to blame, and warns that a potentially more devastating crisis awaits us if they aren't fixed.
Abstract: Raghuram Rajan was one of the few economists who warned of the global financial crisis before it hit. Now, as the world struggles to recover, it's tempting to blame what happened on just a few greedy bankers who took irrational risks and left the rest of us to foot the bill. In Fault Lines, Rajan argues that serious flaws in the economy are also to blame, and warns that a potentially more devastating crisis awaits us if they aren't fixed.Rajan shows how the individual choices that collectively brought about the economic meltdown--made by bankers, government officials, and ordinary homeowners--were rational responses to a flawed global financial order in which the incentives to take on risk are incredibly out of step with the dangers those risks pose. He traces the deepening fault lines in a world overly dependent on the indebted American consumer to power global economic growth and stave off global downturns. He exposes a system where America's growing inequality and thin social safety net create tremendous political pressure to encourage easy credit and keep job creation robust, no matter what the consequences to the economy's long-term health; and where the U.S. financial sector, with its skewed incentives, is the critical but unstable link between an overstimulated America and an underconsuming world.In Fault Lines, Rajan demonstrates how unequal access to education and health care in the United States puts us all in deeper financial peril, even as the economic choices of countries like Germany, Japan, and China place an undue burden on America to get its policies right. He outlines the hard choices we need to make to ensure a more stable world economy and restore lasting prosperity.
946 citations
••
TL;DR: In this paper, the authors move the discussion of student loans away from anecdote by establishing a framework for considering the use of student loan in the optimal financing of collegiate investments and show that enrolling in college is equivalent to signing up for a lottery with large expected gains, but it is also a better investment today than a generation ago.
Abstract: Total student loan debt rose to over $800 billion in June 2010, overtaking total credit card debt outstanding for the first time. By the time this article sees print, the continually updated Student Loan Debt Clock will show an accumulated total of roughly $1 trillion. Borrowing to finance educational expenditures has been increasing - more than quadrupling in real dollars since the early 1990s. The sheer magnitude of these figures has led to increased public commentary on the level of student borrowing. [The authors] move the discussion of student loans away from anecdote by establishing a framework for considering the use of student loans in the optimal financing of collegiate investments. From a financial perspective, enrolling in college is equivalent to signing up for a lottery with large expected gains - indeed, the figures presented here suggest that college is, on average, a better investment today than it was a generation ago - but it is also a lottery with significant probabilities of both larger positive, and smaller or even negative, returns. [The authors] look to available - albeit limited - evidence to assess which types of students are likely to be borrowing too much or too little.
383 citations
Related Papers (5)
Frequently Asked Questions (2)
Q2. what is the gps rate in a given county?
The apparent existence of substantial and systematic income-group heterogeneity incredit trends within individual counties can be illustrated on the example of student loans in Cook County, which includes the City of Chicago and some of its neighboring suburbs.