The Credit-Driven Household Demand Channel
Atif Mian∗
Princeton University and NBER
Remarks at Nobel Symposium On “Money and Banking”Stockholm. May 26, 2018
A Striking Empirical Regularity
The great recession of 2008 left behind a striking empirical regularity that you can see in the left
panel of Figure 1. States within the U.S. that had a larger increase in household leverage between
2002 and 2007, ended up experiencing a much more severe recession between 2007 and 2010 as
measured by the increase in unemployment. Remarkably, we find exactly the same relationship
across countries. In particular, countries that had a larger increase in household leverage between
2002 and 2007, experienced a much more severe recession between 2007 and 2010 (right panel of
Figure 1).
One might complain here that I am conditioning on a recession in figure 1 and as such it is
not that surprising that leverage, conditional on a recession, hurts. However, as I will discuss, this
empirical regularity is not an artifact of cherry picking data, but is the result of a systematic force
that I am going to refer to as the credit-driven household demand channel.
The credit driven household demand channel has three stages. It starts with an expansion in
the supply of credit, i.e. for some reason, financial markets are willing to lend more to the same loan
applicant. This increase in credit supply fuels a boom that drives an outward shift in household
aggregate demand. However, the expanding credit boom also sows the seeds of its own destruction
and ultimately results in a macroeconomic slowdown.
I will present a wide range of empirical findings, both from the U.S. and around the world
and covering the last half century, that hopefully will convince the reader of the prominence of
credit-driven household demand channel. I will first present international evidence from business
∗I thank Michael Varley for excellent research assistance and Julis Rabinowitz Center for Public Policy and Financeat Princeton for financial support. Mian: (609) 258 6718, [email protected].
1
cycles worldwide, including a new out-of-sample test of previous findings. I will then present results
from a natural experiment within the U.S. during the 1980s when states started deregulating their
banking systems. And finally i will discuss U.S. evidence from the Great Recession episode. The
last part of my remarks will be devoted to discussing the theoretical and policy implications of
these empirical findings.
International Evidence
I start with stuctural VAR evidence on credit shocks and their impact on business cycles world-
wide from Mian et al. (2017b). We use data from 30 mostly advanced countries over the last half
century and run VAR using log real GDP, and the two components of private credit to GDP, house-
hold credit to GDP and non-financial firm credit to GDP. Two points are worth noting. First, the
response of GDP to a household credit shock is a muted boom followed by a strong and persistent
slowdown that i have emphasized with an arrow (left panel of figure 2). Second, there is a clear
asymmetry in the response of GDP to household credit shock versus a shock to non-financial firm
credit. The latter does not produce the strong cyclical response (right panel of figure 2). This is
the first indication that a household credit-driven demand channel is important for understanding
business cycles.
Table 1 pools together evidence from Mian et al. (2017b,a) and highlights some noteworthy
elements of this credit-induced business cycle. First, household credit growth, but not non-financial
firm credit growth, is contemporaneously associated with “consumption booms, an increase in the
consumption to GDP ratio and a deterioration in the trade balance with the consumption share of
import goods rising significantly (columns 1 through 3).
This is suggestive of a household credit-driven local demand boom. As discussed in Mian et
al. (2017a), we can explicitly test for whether credit pushes local aggregate demand by testing if it
disproportionately expands non-tradable relative to tradable sector - both in terms of the relatize
size of the non-tradable sector and also in terms of relative price of non-tradable to tradable sector.
This hypothesis is confirmed by columns 4 and 5. There is expansion in non-tradable sector to
tradable sector employment (column 4) and increase in non-tradable to tradable sector prices
(column 5) when household credit expands. No such relationship holds for non-financial firm credit
that is more likely to operate on the supply-side of the economy. Column 6 summarizes the key
2
result I already highlighted in the VAR plot: growth in household credit predicts subsequent GDP
slowdown, whereas non-financial firm debt does not. Notice again the asymmetry between the
effects of household credit and non-financial firm credit throughout table 1.
The results in table 1 are based on a paper we wrote in 2015 and published in the QJE in 2017.
Since then, and just this month in fact, IMF has released a new global debt database that covers an
additional 105 countries with a breakdown of private credit into household credit and non-financial
firm credit. So we have done a literal out of sample test for the Nobel symposium in column 7 of
the core finding of Mian et al. (2017b). The results are essentially identical. As an example of a
recent country that experienced a credit-driven household demand channel boom-bust, interested
readers may look at the case of Brazil.
Figure 3 shows the non-parametric relationship between change in household debt to GDP
between year (t-4) and (t-1) and change in real GDP between t and (t+3). The non-parametric
relationship in the original 30-country sample of Mian et al. (2017b) is shown in blue, while the new
out-of-sample relationship is shown in red. The figure once again illustrates how similar the out
of sample result is. Figure 3 is also a more generalized version of the striking empirical regularity
that I highlighted in figure 1 for the Great Recession.
Another noteworthy feature of figure 3 is that the relationship is non-linear: reduction in house-
hold debt to GDP does not predict an increase in GDP growth. This non-linearity is consistent
with macroeconomic theory where frictions such as ZLB or downward wage rigidity only bind in
one direction.
Evidence From A Natural Experiment During the 1980’s
I will next present evidence from a natural experiment that produces a well-known plausibly
exogenous variation in credit supply expansion across U.S. states during the 1980s: the staggered
wave of banking deregulation across states. The left panel of figure 4 shows the first stage from
Mian et al. (2017a). States that deregulated earlier - for reasons unrelated to expected GDP trend
- experience a much larger expansion in bank credit. The increase in bank credit also resulted in a
much larger increase in household credit in early deregulating states. The right panel shows that
the credit supply expansion results in a, now familiar, boom-bust cycle. Early deregulating states
that see a faster expansion in credit, also experience a sharper decline in unemployment, but only
3
to be followed by a stronger subsequent recession.
Figure 5 shows that the credit-induced boom during the 1980s was driven by an expansion in
household demand. States more exposed to deregulation experienced a stronger expansion in non-
tradable employment, but no relative expansion in tradable employment, between 1982 and 1989.
Similarly, the price of non-tradable goods increased in states that were more exposed to banking
deregulation, but no such relationship exists for tradable sector prices (Figure 6). The combined
increase in non-tradable to tradable sector employment and non-tradable to tradable sector prices
suggests an expansion in local household demand driven by the increase in credit supply.
Figure 7 should look very familiar to the audience by now. It has the same empirical regular-
ity that we saw for the Great Recession in figure 1. States with a larger expansion in household
leverage from 1982 to 1989 experience a much more severe subsequent recession. Moreover, this
time the source of credit expansion is plausibly exogenous as it is driven by the staggered wave of
banking deregulation across U.S. states.
U.S. Evidence From The 2000’s
Evidence from the U.S. Great Recession episode is also quite supportive of the credit-driven
household demand channel. A number of authors have argued, some using very detail micro-
empirical evidence, that there was a large expansion in the supply of credit in the U.S.1. At the
macro level the expansion in credit supply can be seen from the fact that credit spreads in the
U.S. declined significantly prior to 2007, even as the quantity of all types of risky credit expanded
aggressively.
The expansion in credit again lead to an increase in household demand. This has been shown
by Di Maggio and Kermani (2017) who report that the non-tradable sector expanded relative to
tradable sector during the credit boom. However, ultimately rising household credit creates a
predictable collapse as we saw before. The severity of the household credit-driven collapse is driven
by Irving Fisher’s famous “debt deflation hypothesis”. There is an initial fall in demand, that leads
to a fall in employment and ignites a large fire-sale of houses. This feeds back to further reduction
in demand, thus amplifying the initial negative shock
1See e.g. Adelino et al. (2014); Demyanyk and Van Hemert (2011); Favara and Imbs (2015); Justiniano et al. (2015,2017); Keys et al. (2010); Levitin and Wachter (2012); Mian and Sufi (2009, 2017)
4
Figure 8 from Mian et al. (2013) shows the fall in spending, or local demand, against the fall in
household net worth at the county level between 2006 and 2009. Counties where households had
levered up more experienced a larger decline in net worth both because of the direct leverage effect
and because house prices fell more in these areas due to foreclosure fire sales. Counties experiencing
a larger reduction in household net worth experience a sharp contraction in local aggregate demand.
Figure 9 from Mian and Sufi (2014b) shows that the fall in local demand directly translates into
a loss in local employment. There is a stronger decline in non-tradable employment between 2007
and 2009 in counties with a stronger decline in household net worth. The focus on nontradable
employment is useful here since by definition nontradable employment must rely on local demand
for revenue. As such this strong positive relationship between change in nontradable employment
and change in household net worth reflects the direct impact and a ‘macro spillover’ of the drop in
local demand on employment.
It is interesting to contrast this result with a similar analysis using tradable employment. Unlike
for nontradable employment, there is no relationship between the local change in tradable employ-
ment and household net worth. This makes sense since tradable employment does not depend
exclusively on local demand for generating sales. The asymmetry in result between nontradable
and tradable employment further strengthens the interpretation that it is indeed local demand
shocks that are causing the change in local employment. Moreover the tradable employment is also
being affected by the demand shocks it is just that since those shocks are aggregate in nature –
they shift tradable employment downwards for all counties proportionately. A number of papers
have confirmed these effects in other contexts and using alternative empirical strategies2.
Macro Theory Implications Of The Credit-Driven Household Demand Channel
I have discussed a broad range of empirical evidence that describes the importance of the credit
driven household demand channel in practice, not only in the most recent 2008 global recession,
but even prior to that. I am next going to discuss the theoretical and policy implications of this
evidence.
A natural theoretical implication of empirical evidence on the importance of household credit
2See e.g. Andersen et al. (2014); Bahadir and Gumus (2016); Bunn and Rostom (2015); Drehmann et al. (2017);Giroud and Mueller (2017); Glick and Lansing (2010); IMF (2012, 2017); Di Maggio and Kermani (2017); Martinand Philippon (2014); Mian and Sufi (2010); Mian et al. (2017a,b); Verner and Gyongyosi (2017)
5
for macro aggregates is that heterogeneity matters, and in particular heterogeneity in the behavior
of borrower versus creditor households. For example, as right panel of figure 10 from Mian et al.
(2013) shows, marginal propensity to spend in response to a dollar decline in housing wealth is
much stronger for levered households than unlevered households. Similarly, the left panel of figure
10 from Mian and Sufi (2011) shows that during the boom phase, marginal propensity to borrow for
a dollar increase in home value was much stronger for low credit score individuals than high credit
score individuals. It is thus no longer sufficient to model aggregate dynamics using a representative
household economy. The empirical evidence suggests that macro aggregates fundamentally depend
on the covariance of shocks with the underlying household heterogeneity. Thus if a negative shock
falls disproportionately on levered households that have the highest marginal propensity to respond
then the net effect on the overall economy would be much stronger.
A number of recent papers in macroeconomic theory have emphasized such heterogeneity and
how it interacts with macro frictions such as zero lower bound constraint or downward nominal
wage rigidity 3. The work has also highlighted the presence of an aggregate demand externality, or a
pecuniary externality such that ex ante individual households will tend to over borrow from a macro
perspective. Individuals fail to fully internalize the negative future macroeconomic consequences of
their collective borrowing decisions leading to over-borrowing that may require macro-prudential
interventions.
The newer class of models rationalize why expansion in credit supply for the household sector
leads to a boom bust pattern. However, there is one potential problem. These models are based
on rational expectation and common belief assumptions. Thus the models suggest that market
participants should also predict the slowdown that follows a household credit boom. However
when we look at the data, there is no evidence that the market or households correctly predict a
subsequent slowdown during the boom. Infact, as figure 11 from Mian et al. (2017b) shows there
is evidence to the contrary. Growth in household credit predicts GDP forecasting errors by the
IMF and other professional forecastors. Models with heterogeneous beliefs or behavioral biases
can help address this issue, suggesting that these forces are also important to fully understand the
relationship between credit expansion and business cycles4.
3See e.g. Eggertsson and Krugman (2012); Farhi and Werning (2015); Guerrieri and Lorenzoni (2017); Huo andRıos-Rull (2016); Korinek and Simsek (2016); Lorenzoni (2008); Schmitt-Grohe and Uribe (2016)
4see e.g. Baron and Xiong (2016); Bordalo et al. (2017); Burnside et al. (2017); Geanakoplos (2010); Gennaioli et al.
6
So far I have focused on the credit driven household demand channel at the business cycle
frequency. However there is also a longer-term “super cycle in the background, driven by a persistent
expansion in credit as shown by the work of Jorda et al. (2016). Moreover, the super cycle is largely
driven by the growth in household credit, and has been accompanied by a strong decline in long-
term real interest rate. The latter fact suggests that the long term trends are driven by an expansion
in credit supply forces.
What might be the drivers of the increase in credit supply? And what are its longer term
economic consequences? The rise in global credit coincides with the rise in global inequality, par-
ticularly the top 1% versus the rest, and the appearance of a global savings glut. Since top incomes
save at a very high rate, channeling these savings into the financial sector is naturally going to
increase credit which can only be sustained at continually declining interest rates5. Indeed we find
in ongoing work, see figure 12, that the rise in household credit is concentrated in the bottom 99%
and not the 1%, while income gains since 1980’s have largely gone to the top 1%. How long can this
process continue without leading to liquidity trap-like situations and lower growth is an important
open question facing us today.
Policy Implications Of The Credit-Driven Household Demand Channel
The credit driven household demand channel has important implications for public policy. I
will particularly focus on implications for crisis response, monetary policy and macro prudential
policy.
Policy response to the 2008 crisis centered on interventions to promote provision of market
liquidity and public support for bank capital injections. However, a recession that is the result of
the credit-driven household demand channel requires that attention should be paid to repair and
restructure household balance sheets as well.
Our main criticism of the administration’s response to the 2008 crisis is that not sufficient
attention was paid to restructure loans for under-water and distressed homeowners. Similarly,
efforts should have been made to reduce the more than four million foreclosed homes that ended
up worsening the downturn considerably Mian et al. (2015).
(2012); Kindleberger (1978); Kindelberger and Aliber (2005); Krishnamurthy and Muir (2016); Lopez-Salido et al.(2017); Mian and Sufi (2018a); Minsky (2008); Nathanson and Zwick (2017)
5see e.g. Favilukis et al. (2012); Jorda et al. (2016); Kumhof et al. (2015)
7
Looking forward, there is a need to design better regulatory and financial framework that pro-
motes risk-sharing between creditors and debtors. We discuss this with relation to state-contingent
contracting in Mian and Sufi (2014a). As we discuss, moving in this direction requires institu-
tional changes, such as eliminating the favorable treatment of debt and re-designing bank capital
regulation.
The credit driven household demand channel is also one of the most powerful channels through
which monetary policy impacts the real economy. In normal times, expansionary monetary pol-
icy lowers rates and increases house prices, enabling high MPC constrained households to boost
spending, thus raising aggregate demand. A tightening cycle can work in reverse, making mone-
tary policy impotent as high MPC households fail to respond to monetary easing due to hightened
risk-aversion and borrowing constraints. For example, Di Maggio et al. (2017) show that lower
interest rates post-2008 were not passed-through to many constrained households who were unable
to refinance, thus putting a real drag on aggregate demand6. The failure of traditional monetary
policy tools to find traction in boosting aggregate demand means that alternative approaches need
to be considered.
Finally, a natural policy implication of credit-driven household demand channel is that ex-ante
macro-prudential policies that constrain household credit growth are useful. A number of countries,
most notably the U.K., have gone in this direction since 2008, putting limits on household credit
growth based on a combination of loan to value and debt to income constraints.
6For broader evidence on household demand channel constraining the efficacy of monetary policy, see Agarwal etal. (2017, 2018); Aladangady (2014); Baker (2018); Cloyne et al. (2017); Ganong and Noel (2017a,b); Jorda et al.(2014); Liu et al. (2018); Mian and Sufi (2018b, Forthcoming)
8
References
Adelino, Manuel, Antoinette Schoar, and Felipe Severino, “Credit supply and house prices:
Evidence from mortgage market segmentation,” Available at SSRN 1787252, 2014.
Agarwal, Sumit, Gene Amromin, Itzhak Ben-David, Souphala Chomsisengphet,
Tomasz Piskorski, and Amit Seru, “Policy Intervention in Debt Renegotiation: Evidence
from the Home Affordable Modification Program,” Journal of Political Economy, 2017, 125 (3).
, Souphala Chomsisengphet, Neale Mahoney, and Johannes Stroebel, “Do Banks Pass
Through Credit Expansions to Consumers Who Want to Borrow? Evidence from Credit Cards,”
Quarterly Journal of Economics, 2018, 133 (1), 129–90.
Aladangady, Aditya, “Homeowner Balance Sheets and Monetary Policy,” FEDS Working Paper,
Federal Reserve Board 2014.
Andersen, Asger L., Charlotte Duus, and Thais L. Jensen, “Household Debt and Consump-
tion During the Financial Crisis: Evidence from Danish Micro Data,” 2014. Working Paper.
Bahadir, Berrak and Inci Gumus, “Credit decomposition and business cycles in emerging
market economies,” Journal of International Economics, 2016, 103, 250–262.
Baker, Scott R, “Debt and the Response to Household Income Shocks: Validation and Application
of Linked Financial Account Data,” Journal of Political Economy, 2018.
Baron, Matthew and Wei Xiong, “Credit expansion and neglected crash risk,” National Bureau
of Economic Research, 2016.
Bordalo, Pedro, Nicola Gennaioli, and Andrei Shleifer, “Diagnostic Expectations and
Credit Cycles,” Journal of Finance, 2017.
Bunn, Philip and May Rostom, “Household Debt and Spending in the United Kingdom,” 2015.
Bank of England Working Paper.
Burnside, Craig, Martin Eichenbaum, and Sergio Rebelo, “Understanding Booms and
Busts in Housing Markets,” Journal of Political Economy, 2017, 124, 1088–1147.
9
Cloyne, James, Clodomiro Ferreira, and Paolo Surico, “Monetary Policy When Households
Have Debt: New Evidence on the Transmission Mechanism,” 2017.
Demyanyk, Yuliya and Otto Van Hemert, “Understanding the subprime mortgage crisis,”
Review of financial Studies, 2011, 24 (6), 1848–1880.
Di Maggio, Marco, Amir Kermani, Benjamin J. Keys, Tomasz Piskorski, Ramcharan
Rodney, Amit Seru, and Vincent Yao, “Interest Rate Pass-Through: Mortgage Rates,
Household Consumption, and Voluntary Deleveraging,” American Economic Review, 2017, 107
(11), 3550–88.
and , “Credit-induced boom and bust,” The Review of Financial Studies, November 2017,
30 (11), 3711–58.
Drehmann, Mathias, Mikael Juselius, and Anton Korinek, “Accounting for debt service:
the painful legacy of credit booms,” BIS Working Papers No 645 June 2017.
Eggertsson, Gauti B. and Paul Krugman, “Debt, Deleveraging, and the Liquidity Trap: A
Fisher-Minsky-Koo Approach,” Quarterly Journal of Economics, 2012, 127 (3), 1469–1513.
Farhi, Emmanuel and Ivan Werning, “A Theory of Macroprudential Policies in the Presence
of Nominal Rigidities,” 2015. Working Paper.
Favara, Giovanni and Jean Imbs, “Credit supply and the price of housing,” The American
Economic Review, 2015, 105 (3), 958–992.
Favilukis, Jack, David Kohn, Sydney C Ludvigson, and Stijn Van Nieuwerburgh, “In-
ternational capital flows and house prices: Theory and evidence,” in “Housing and the Financial
Crisis,” University of Chicago Press, 2012, pp. 235–299.
Ganong, Peter and Pascal Noel, “Consumer Spending During Unemployment: Positive and
Normative Implications,” 2017. Working Paper.
and , “The Effect of Debt on Default and Consumption: Evidence from Housing Policy in
the Great Recession,” 2017. Working Paper.
10
Geanakoplos, John, “The Leverage Cycle,” NBER Macroeconomics Annual 2009, Volume 24,
April 2010, pp. 1–65.
Gennaioli, Nicola, Andrei Shleifer, and Robert Vishny, “Neglected risks, financial innova-
tion, and financial fragility,” Journal of Financial Economics, 2012, 104 (3), 452–468.
Giroud, Xavier and Holger M. Mueller, “Firm Leverage, Consumer Demand, and Unemploy-
ment during the Great Recession,” Quarterly Journal of Economics, forthcoming, 2017.
Glick, Reuven and Kevin J. Lansing, “Global household leverage, house prices, and consump-
tion,” FRBSF Economic Letter, 2010, (11).
Guerrieri, Veronica and Guido Lorenzoni, “Credit crises, precautionary savings, and the
liquidity trap,” The Quarterly Journal of Economics, 2017, 132 (3), 1427–1467.
Huo, Zhen and Jose-Vıctor Rıos-Rull, “Financial Frictions, Asset Prices, and the Great
Recession,” 2016. Working Paper.
IMF, “Dealing with Household Debt,” in “World Economic Outlook, April” April 2012.
, “Household Debt and Financial Stability,” in “Global Financial Stability Report, October”
October 2017.
Jorda, Oscar, Moritz Schularick, and Alan M. Taylor, “Betting the House,” 2014. Federal
Reserve Bank of San Francisco Working Paper.
Jorda, Oscar, Moritz Schularick, and Alan M Taylor, “The great mortgaging: housing
finance, crises and business cycles,” Economic Policy, 2016, 31 (85), 107–152.
Justiniano, Alejandro, Giorgio E. Primiceri, and Andrea Tambalotti, “Credit Supply
and the Housing Boom,” Working Paper 20874, National Bureau of Economic Research January
2015.
, , and , “The Mortgage Rate Conundrum,” 2017. Working Paper.
Keys, Benjamin J, Tanmoy Mukherjee, Amit Seru, and Vikrant Vig, “Did securitization
lead to lax screening? Evidence from subprime loans,” The Quarterly journal of economics, 2010,
125 (1), 307–362.
11
Kindelberger, Charles P and Robert Z Aliber, Manias, panics and crashes: A history of
financial crises, Palgrave Macmillan, 2005.
Kindleberger, Charles, Manias, Panics and Crashes: A history of financial crises, New York,
Basic Books, 1978.
Korinek, Anton and Alp Simsek, “Liquidity trap and excessive leverage,” The American Eco-
nomic Review, 2016, 106 (3), 699–738.
Krishnamurthy, Arvind and Tyler Muir, “How credit cycles across a financial crisis,” Working
paper, 2016.
Kumhof, Michael, Romain Ranciere, and Pablo Winant, “Inequality, Leverage, and Crises,”
American Economic Review, 2015, 105, 1217–1245.
Levitin, Adam J. and Susan M. Wachter, “Explaining the Housing Bubble,” 2012. Working
Paper.
Liu, Ernest, Atif Mian, and Amir Sufi, “Low Interest Rates, Market Power and Productivity
Growth,” 2018.
Lopez-Salido, David, Jeremy C Stein, and Egon Zakrajsek, “Credit-market sentiment and
the business cycle,” The Quarterly Journal of Economics, 2017, p. qjx014.
Lorenzoni, Guido, “Inefficient Credit Booms,” The Review of Economic Studies, 2008, 75 (3),
809–833.
Martin, Philippe and Thomas Philippon, “Inspecting the Mechanism: Leverage and the
Great Recession in the Eurozone,” 2014. Working Paper.
Mian, Atif, Amir Sufi, and Francesco Trebbi, “Foreclosures, house prices, and the real
economy,” The Journal of Finance, 2015, 70 (6), 2587–2634.
and , “The Consequences of Mortgage Credit Expansion: Evidence from the U.S. Mortgage
Default Crisis,” The Quarterly Journal of Economics, 2009, 124 (4), pp. 1449–1496.
and , “Household Leverage and the Recession of 2007-09,” IMF Economic Review, 2010, 58
(1), 74–117.
12
and , “House Prices, Home Equity-Based Borrowing, and the US Household Leverage Crisis,”
American Economic Review, August 2011, 101 (5), 2132–56.
and , House of Debt: How They (and You) Caused the Great Recession and How We Can
Prevent It From Happening Again, University of Chicago Press: Chicago, 2014.
and , “What Explains the 2007–2009 Drop in Employment?,” Econometrica, 2014, 82 (6),
2197–2223.
and , “Fueling a Frenzy: Private Label Securitization and the Housing Cycle of 2000 to 2010,”
Working Paper 2018.
and , “Inequality, Surplus Savings and Credit Creation,” 2018.
and , “Finance and Business Cycles: The Credit-Driven Household Demand Channel,” Jour-
nal of Economic Perspectives, Forthcoming.
, Kamalesh Rao, and Amir Sufi, “Household Balance Sheets, Consumption, and the Eco-
nomic Slump,” The Quarterly Journal of Economics, 2013, 128 (4), 1687–1726.
Mian, Atif R, Amir Sufi, and Emil Verner, “Credit Supply and Business Cycle Ampli cation:
Evidence from Banking Deregulation in the 1980s,” 2017.
, , and , “Household debt and business cycles worldwide,” forthcoming, Quarterly Journal
of Economics, 2017.
and , “Household Debt and Defaults from 2000 to 2010: The Credit Supply View,” Evidence
and Innovation in Housing Law and Policy, 2017.
Minsky, Hyman P, Stabilizing an unstable economy, Vol. 1, McGraw-Hill New York, 2008.
Nathanson, Charles G. and Eric Zwick, “Arrested Development: Theory and Evidence of
Supply-Side Speculation in the Housing Market,” National Bureau of Economic Research June
2017.
Schmitt-Grohe, Stephanie and Martın Uribe, “Downward Nominal Wage Rigidity, Currency
Pegs, and Involuntary Unemployment,” Journal of Political Economy, 2016.
13
Verner, Emil and Gyozo Gyongyosi, “Household Debt Revaluation and the Real Economy:
Evidence from a Foreign Currency Debt Crisis,” 2017.
14
Figure 1: Introduction
AL
AK
AZ
AR
CA
COCT DE
DC
FLGA
HI
ID
IL
IN
IAKS
KYLA
ME
MD
MAMI
MN
MSMO
MT
NE
NV
NH
NJ
NMNY
NC
ND
OH
OK
OR
PA
RISC
SD
TNTX
UT
VT
VA
WAWV
WIWY
0
2
4
6
8
10
∆ U
nem
ploy
men
t, 20
07−
10
0 10 20 30 40 50 60∆ HH debt to GDP, 2002−07
Source: Mian and Sufi (IMF, 2010).
United States
AUSAUT
BELCAN
CHECZE
DEU
DNK
ESP
FINFRAGBR
GRC
HKG
HUN
IDN
IRL
ITA
JPNKOR
MEXNLDNOR
POL
PRT
SGP
SWE
THA
TUR
USA
−5
0
5
10
15
∆ U
nem
ploy
men
t, 20
07−
10
−20 0 20 40 60∆ HH debt to GDP, 2002−07
Source: Mian, Sufi, and Verner (QJE, 2017).
World
15
Figure 2: Effect of credit shock on GDP
−.8
−.6
−.4
−.2
0
.2
.4
.6
0 2 4 6 8 10Years after Shock
Source: Mian, Sufi, and Verner (QJE, 2017)
GDP Response to HH Debt Shock
−.8
−.6
−.4
−.2
0
.2
.4
.6
0 2 4 6 8 10Years after Shock
Source: Mian, Sufi, and Verner (QJE, 2017)
GDP Response to NF Debt Shock
16
Figure 3: Increase in household debt to GDP predicts GDP slowdown
−20
0
20
40
GD
P g
row
th, t
to t+
3
−10 0 10 20 30Household Debt to GDP Expansion, t−4 to t−1
MSV2017 30 CountriesIMF2018 Additional 105 Countries
Source: Mian, Sufi, and Verner (QJE, 2017).
17
Figure 4: Deregulation Experiment in the 1980s
100
150
200
250
1980 1982 1984 1986 1988 1990 1992
Early DeregulationLate Deregulation
Source: Mian, Sufi, and Verner (WP, 2018).
Total Bank Credit
50
60
70
80
90
100
1980 1982 1984 1986 1988 1990 1992Source: Mian, Sufi, and Verner (WP, 2018).
Unemployment Rate
18
Figure 5: Non-tradable vs Tradable Employment in the 1980s
AK
AL
AR
AZ
CA
CO
CT
DC
FL
GA
HI
IAID
IL
IN
KS
KY
LA
MA
MD
ME
MI
MN
MOMS
MT
NC
ND
NE
NH
NJ
NM
NV
NYOH
OK
ORPA
RI
SC
TN
TX
UT
VA
VT
WA
WI
WV
WY−0.10
0.00
0.10
0.20
0.30
0.40
−2 −1 0 1 2Deregulation exposure
Source: Mian, Sufi, and Verner (WP, 2018).
Non−Tradable Employment Growth, 82−89AK
ALAR
AZ
CA
CO
CTDC
FL GA
HI
IA
ID
IL
INKS
KY
LA
MA
MD
ME
MIMN
MO
MS
MT NCND
NENH
NJ
NM
NV
NY
OH
OK
OR
PA
RI
SC
TN
TX
UT
VA
VT
WA
WI
WVWY
−0.30
−0.20
−0.10
0.00
0.10
0.20
−2 −1 0 1 2Deregulation exposure
Source: Mian, Sufi, and Verner (WP, 2018).
Tradable Employment Growth, 82−89
19
Figure 6: Non-Tradable vs. Tradable Price in the 1980s
CA
CO
CT
FL
GAHI
ILIN
KS KY
MAMD
ME
MI
MNMO
NH
NJ
NY
OH
OR
PA
TX
WAWI
0.15
0.20
0.25
0.30
0.35
−2 −1 0 1 2Deregulation exposure
(Alaska excluded)Source: Mian, Sufi, and Verner (WP, 2018).
Non−tradable CPI Inflation, 84−89
CA
CO
CT
FLGA
HI
ILIN
KSKY
MA
MD
MEMI
MN
MONHNJ
NY
OHOR
PATX
WA
WI
0.05
0.10
0.15
0.20
0.25
−2 −1 0 1 2Deregulation exposure
(Alaska excluded)Source: Mian, Sufi, and Verner (WP, 2018).
Tradable CPI Inflation, 84−89
20
Figure 7: Increase in household leverage predicts 1990/91 recession
AK
ALAR
AZ
CA
CO
CTDC
FL
GAHI
IA
ID
ILIN
KSKYLA
MA
MDME
MI
MNMO
MSMT
NC
NDNE
NHNJ
NM
NV
NY
OH
OK
OR
PA
RI
SC
TN
TX
UT
VA
VT
WAWI
WV
WY
−2.00
0.00
2.00
4.00
6.00
∆ U
nem
ploy
men
t, 19
89−
92
−2 −1 0 1 2 3∆ HH Leverage, 1982−89
Source: Mian, Sufi, and Verner (WP, 2018).
21
Figure 8: Fall in demand during the great recession
−.4
−.2
0
.2
∆ E
xpen
ditu
re, 0
6−09
−.3 −.25 −.2 −.15 −.1 −.05 0 .05∆ HH net worth, 06−09
Source: Mian, Rao, and Sufi (QJE, 2013).
22
Figure 9: Local employment falls due to lack of demand
−.15
−.1
−.05
0
.05
.1
∆ T
rada
ble
Em
ploy
men
t, 07
−09
−.15
−.1
−.05
0
.05
.1
∆ N
on−
Tra
dabl
e E
mpl
oym
ent,
07−
09
−.3 −.25 −.2 −.15 −.1 −.05 0 .05∆ HH net worth, 06−09
Source: Mian and Sufi (ECMA, 2014).
−.15
−.1
−.05
0
.05
.1
∆ T
rada
ble
Em
ploy
men
t, 07
−09
−.15
−.1
−.05
0
.05
.1
∆ N
on−
Tra
dabl
e E
mpl
oym
ent,
07−
09
−.3 −.25 −.2 −.15 −.1 −.05 0 .05∆ HH net worth, 06−09
Source: Mian and Sufi (ECMA, 2014).
23
Figure 10: Heterogeneity in marginal propensity to borrow and marginal propensity to consume
0
.05
.1
.15
.2
.25
< 700 700 − 799 800 − 899 900 − 999
Source: Mian and Sufi (AER, 2011).
MPB out of Housing Wealth by Credit Score, ’02−06
0
.005
.01
.015
.02
.025
.03
<= 30% 30%−50% 50%−70% 70%−90% > 90%
Source: Mian, Rao, and Sufi (QJE, 2013).
MPC out of Housing Wealth by HH Leverage, ’06−09
24
Figure 11: Rising household leverage predicts forecasting errors
THA2002
HUN1993
NOR1994
THA2001FIN1996
HUN1994
FIN1997
CHE2009
SGP2007
NOR1996
DEU2009
SWE1996
NOR1995
HKG2007
NOR1993
DEU2008
SGP2008
FIN1998
FIN1995SWE1997
SWE1992
SWE1993
HKG2008NOR1997
NOR1992
THA2003
HKG2006
FIN1999SWE1995
DEU2007
SWE1994NOR1998
HUN1997
JPN2005HUN1998
HKG2009
CHE2008CZE1999
HUN1996JPN2003JPN2004
JPN2007
HKG2005
JPN2009SWE1998GBR1997
JPN2006
MEX1999
MEX2000
GBR1996
FRA1997
DEU1991
JPN2008
HUN1995
HUN1999
DEU2006
MEX2001
ITA1996
KOR2000
KOR2001ITA1997
CAN1996
DEU1992
DEU2005
SGP2009
GBR1995
MEX2002
CAN2002BEL2003NOR2001DEU1993
BEL2002
FRA1998
TUR2004
DEU1990
JPN2002ITA1998DEU2004
HKG1995
FIN1994
MEX1998TUR1997
THA2009
TUR1991FIN2000
DEU2003
CAN2001
ITA1992
CZE2000
TUR1990FRA1996GBR1998
TUR2003
ESP1993
NOR2002
CZE2001ITA1995HUN2000
BEL1996
TUR2002
FRA1999
TUR1992
THA2000CHE2007TUR1993
IDN2009
ESP1995
ESP1996
THA2008
JPN1999
GBR1994
CHE2003MEX2003
TUR2000
JPN1994
AUT2009
PRT1991
JPN1998FRA2001
BEL2004
ITA1991FRA2000
SGP2006
CAN1997
KOR2006
ESP1997
IDN2008
ESP1994BEL2001
AUS1993
TUR1995
TUR1999
KOR1999
PRT1992
ITA1990
NOR1999
TUR1996
BEL1997
USA1995
JPN1993ITA1994
NOR1991
CAN2003
CZE2002
FRA1995JPN2001ITA1999
AUS1991
TUR2001
USA1994
CAN1995SWE1999AUT1999HUN2001
MEX2004
FRA2002
JPN2000
AUT2004
FIN2001GBR1999USA1999
TUR1998
BEL2005BEL1995ITA1993
USA1998
MEX2005FRA2003
USA2000
HKG2004
JPN1995
FIN1993
ESP1992
CAN1998AUS1992
PRT1993AUS1990
SGP2001
MEX2006
AUT2003USA1993
IDN2007
POL2006
JPN1997
TUR1994
SWE1991NOR2000
MEX2009
THA2007
BEL1998
HKG1996
HKG2003
AUT2001THA2004
POL2002
POL2000MEX2007
POL2001
KOR1996
USA1996
PRT1990
AUT2000SWE2000
GRC1998
TUR2005FRA1993
FRA2004
USA1997
FIN2002CAN2000NOR2009BEL1994USA1992
BEL2000AUT2008
AUS1994
FIN2003
FRA1994
AUT2005ESP1998
USA1991
JPN1996
USA1990DEU1994
NOR2007
GRC1999
BEL1992
GBR1993
FRA1992
FIN1992
DEU2002
SGP2000
MEX2008
BEL1993
CZE2003
AUT2002
ITA2000
SWE2001USA2001BEL1999
KOR1997
GRC2000
SGP2005
POL1999
FRA1991
CZE2004
BEL1991BEL2006
HUN2002
KOR1998
FRA1990
CAN1994
HKG1997ESP1991
HKG2002
CAN1999
FRA2005
POL2005
IDN2006
SGP1999
NLD1994
ITA2003
USA2009
BEL1990ITA2001
TUR2009
IDN2005NLD2009GBR2000
ITA2004
SGP2003
KOR1994
CZE2006
KOR1995
JPN1992
HKG1994
ITA2002
CZE2005
POL2007
FIN1991
THA2006
DNK2002GBR2001
TUR2006
CAN2004
BEL2007
AUS2009
AUT2007
POL2004
AUT2006NOR2008
GRC2001
SWE2002
ITA2005
AUS1995SWE2003
FRA2006
KOR2007DEU2001
ITA2009
NOR2006
NOR2003SGP2002
FIN2009SGP1998
DEU1995
TUR2007
TUR2008GBR1992BEL2009POL2003
DNK2001THA2005BEL2008
SGP1995
PRT1994
FIN2004
CHE2006
KOR2005
USA2002ESP1999
SWE2004
HKG2001
HUN2003
DEU1998
NLD1995
GBR2008
CZE2007
FRA2009
KOR1993
FRA2007ESP1990
ITA2006
SWE2005
DEU1999
SWE1990
DEU1997CAN2005CAN1993THA1999
AUS1996CHE2005
FRA2008
CHE2004
DEU2000CAN2006
KOR2002
DEU1996GBR2009
JPN1991
CAN1992
AUS1998
SWE2008
NLD1996
DNK2000
CAN2007
CAN1991
AUS1997SWE2009
ITA2007
ITA2008
GBR2002
PRT1995
NOR1990
FIN1990
GRC2002
CAN1990
FIN2008
KOR2009
SGP1996
ESP2003
POL2008
HUN2007
SWE2006
AUS1999ESP2002
USA2008
SWE2007
CZE2008
CZE2009
FIN2005
SGP1997GBR1991
DNK2003
ESP2000
PRT2009CAN2008KOR1990
HUN2008
NLD1997
USA2007
USA2003
PRT1997
CAN2009
DNK1999
PRT1996
KOR2008
DNK1998KOR1992
HUN2004
GBR2007
NLD2008
KOR1991
ESP2004
ESP2001
SGP2004
FIN2007
NOR2005
AUS2001
GBR1990
ESP2009
GRC2003
NLD2003
GBR2006
AUS2000
HUN2005
AUS2002
HKG1998
HUN2006
THA1998
PRT1998
GBR2003
FIN2006
DNK2004
PRT2008
USA2006
NLD1998
GRC2004
JPN1990
THA1997
PRT2005
NLD2002
AUS2008
THA1996
HUN2009
PRT2007HKG2000
PRT2004
USA2005
DNK2005AUS2003
PRT2006
GRC2005NOR2004
THA1995
HKG1999
USA2004
POL2009
PRT2003
GBR2004KOR2004
NLD2007
NLD2004
PRT1999
ESP2005
AUS2007
GRC2009
NLD1999
DNK2006
GBR2005
GRC2006NLD2001
NLD2005
DNK2009
NLD2000
ESP2008
KOR2003AUS2004
GRC2007
PRT2002ESP2006PRT2000
NLD2006
GRC2008
AUS2006
DNK2008
DNK2007
ESP2007
PRT2001
AUS2005IRL2009
IRL2008
IRL2006
IRL2007
−30
−20
−10
0
10
20
IMF
WE
O t
to t+
3 G
DP
For
ecas
t Err
or
−10 0 10 20 30 40Household Debt to GDP Expansion, t−4 to t−1
Source: Mian, Sufi, and Verner (QJE, 2017)
25
Figure 12: Rising leverage concentrated in the bottom 99%
0
.5
1
1.5
2
Deb
t to
Inco
me
1960 1970 1980 1990 2000 2010Year
bot 99%top 1%
Source: Mian and Sufi (WP, 2018).
26
Table 1: International Evidence
MSV2017 30 Countries, 1962-2012 IMF2018
(1) (2) (3) (4) (5) (6) (7)
∆3CitYit
∆3NXitYit
∆3sMCit ∆3 ln
(LNTit
LTit
)∆3 ln
(PNTit
PTit
)∆3yi,t+4 ∆3yi,t+4
∆3dHHit 0.058∗ -0.15∗∗ 0.055∗ 0.36∗∗ 0.38∗∗ -0.34∗∗ -0.37∗
(0.024) (0.051) (0.025) (0.056) (0.097) (0.089) (0.17)
∆3dFit 0.038∗∗ -0.00036 -0.012 0.0085 -0.065 -0.032 -0.019∗∗
(0.012) (0.031) (0.021) (0.064) (0.059) (0.038) (0.0072)
Country fixed effects X X X X X X XR2 0.087 0.062 0.012 0.17 0.067 0.11 0.056Observations 816 832 858 639 670 840 964
Standard errors in parentheses
Source: Mian, Sufi, and Verner (QJE, 2017).+ p < 0.1, ∗ p < 0.05, ∗∗ p < 0.0127