Professional Documents
Culture Documents
The Russell Sage Foundation and The Stanford Center on Poverty and Inequality
I vay l o D . P e t e v, L S Q - C R E S T &
L u i g i P is ta f e rri , S ta n f o rd
Recession Trends • The Russell Sage Foundation and The Stanford Center on Poverty and Inequality
2 Consumption in the Great Recession
to the bleak and suggest major problems in maintaining con- last quarter of 2009). The two main means-tested programs,
sumption. The growth in government transfers didn’t stem other than Medicaid, that experienced substantial increases
the overall decline in consumption because it didn’t address in expenditures were the Supplemental Nutrition Assistance
the falloff in consumption at the top of the income distribu- Program (i.e., “food stamps”) and the Earned Income Tax
tion. Unlike prior recessions, the falloff in consumption was Credit (i.e., EITC). The Unemployment Insurance program also
unusually broad, affecting not just the consumption of dura- increased substantially because of the rise in unemployment
ble goods but also that of nondurable ones. This decline has and the extensions in the length of those benefits.
also proven to be remarkably long-lasting in comparison to
declines in prior recessions. The associated dropoff in con- These transfers, which benefit primarily households at the
sumer confidence was equally spectacular, and the recovery lower end of the income distribution, assisted in propping
in confidence is far from complete even today. Worse yet, up consumption at that lower end. The overall decline in
individual perceptions of economic deterioration were wide- consumption, as shown in figure 1, was nonetheless quite
spread, with no income, age, or racial group spared, even dramatic because the sharp declines in financial income
though each was affected in different ways. We lay out each at the higher end of the distribution lowered consumption
of these five results in more detail below. among the better off. Although one might expect the wealthy
to smooth their consumption during downturns by drawing
Personal Consumption and Personal Disposable Income on their “buffer wealth,” the extreme wealth destruction in the
We start by asking whether trends in consumption spending early periods of the recession may have motivated them to
during the Great Recession are similar to trends in personal rebuild their buffer stock rather than engage in consumption.
disposable income. In figure 1, we plot trends in per-capita It follows that government transfers alone were not enough to
personal consumption expenditure and personal disposable stem the sharp decline in consumption.
income over the Great Recession period. What is remark-
able about this graph is that, while per-capita consumption A Decline in All Consumption Components
declines monotonically until the middle of 2009, disposable In figure 2, we zoom out of the Great Recession period and
income first rises and then, starting in the third quarter of 2008, look at the macro picture for consumption components over
falls precipitously (a 6 percentage point decline from peak to the last 12 years. In particular, we plot the quarterly growth of
trough). The breakdown of disposable income into its three the three components of personal consumption expenditure
components (transfers, wages, and financial income) reveals (durables, nondurables, and services) over the 2000–2011
that the delayed decline in disposable income is explained period. By definition, durable goods are those expected to
entirely by a strong increase in government transfers to last more than three years, whereas nondurable goods (e.g.,
households (+18.6 percent from the last quarter of 2007 to the food, clothing) have a shorter expected lifetime.
figure 1. Consumption and Disposable Income figure 2. Growth Rate of Consumption Components
6
102 4
Quarterly growth rate
2
100 0
-2
98 -4
-6
96
-8
q3 q4 q1 q2 q3 q4 q1 q2 q3 q4 q1 q2 q3 q4 q1 q2 q3 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1
2007 2008 2009 2010 2011
Personal disposable income Personal consumption expenditure Nondurable goods Durable goods Services
Source: BEA, Nipa tables 2.1, 2.3.4, and 2.3.5. Source: BEA, Nipa tables 2.1, 2.3.4, and 2.3.5.
Recession Trends • The Russell Sage Foundation and The Stanford Center on Poverty and Inequality
3 Consumption in the Great Recession
Two important points come out of figure 2. First, the graph 15 quarters from the onset of the recession (normalized to be
shows the well-known fact that spending on durables is much 100 in the quarter immediately preceding the start of each
more volatile than spending on nondurables or services, with recession). Solid lines turn into dashed lines when the reces-
wide upward and downward swings at the onset of booms sion officially ends.
and recessions, respectively. Second, the last two recessions
differ dramatically in terms of the impact on consumption. A number of facts emerge from looking at the top-left panel
The 2001 recession, induced by the deflating of the dot-com of figure 3. First, unlike the 1980 or 1973–75 busts—when
bubble, was very shallow: only durables declined in real terms consumption fell dramatically at the start of the recession—
during this recession. In contrast, the Great Recession, which during the Great Recession, the fall in consumption has been
began with the burst of the housing bubble and the global initially more muted. Moreover, for the Great Recession, con-
financial crisis that ensued, is characterized by a decline in sumption remains below the pre-recession levels for a longer
real terms in all consumption components. At the beginning period than any other recessions represented in the graph.
of the recession, the fall is precipitous for expenditures on The historical comparison illustrates that an economic bust is
durables, and it is substantial for nondurables. Although con- defined not only by the extent of the fall in the components of
sumption growth recovered in the second half of 2009 and Gross Domestic Product (GDP) but also by the time it takes
through 2010, the growth in expenditures on both durables to fully recover. In the 1980 recession, the U.S. economy
and nondurables slowed down again in 2011. The upshot experienced a more dramatic single-quarter fall in consumer
is that the Great Recession, unlike the 2001 recession, has spending, but the recovery was also quite rapid.
involved an across-the-board decline in many types of con-
sumption, not just that of durables. Trends in total consumption mask considerable heterogene-
ity in the behavior of its three components. In fact, because
A Long-Lasting Decline in Consumption there is so much heterogeneity, we’ve had to change the
The Great Recession is one of the longest on record. To put y-axis scale for durables, allowing it to range from 80 to 140
this in perspective and to appreciate the popular reference to (instead of 95 to 115). As shown in figure 3, spending on dura-
this recession as “Great,” we plot in figure 3 the Great Reces- bles falls substantially, while the fall in nondurable spending is
sion next to all the U.S. recessions that have occurred since more moderate, and spending on services falls monotonically
the early 1970s. In each graph, consumption is plotted over but at a substantially lower rate. It is the rapid and significant
115 140
Total Consumption
110
120
durables
105
100
100
95 80
0 3 6 9 12 15 0 3 6 9 12 15
110 115
1973–75
1990–91
110
105
nondurables
1982–83
services
105 2001
100 1980
100
Great Recession
95 95
0 3 6 9 12 15 0 3 6 9 12 15
Recession Trends • The Russell Sage Foundation and The Stanford Center on Poverty and Inequality
4 Consumption in the Great Recession
figure 4. Consumption Growth, Consumer Confidence, and Heterogeneity recovery in services spending that helps
the recovery of total consumption in pre-
Recession years vious recessions (see the bottom-right
ICS and NIPA Growth panel of figure 3). In the Great Recession,
10
120
8 however, spending on services declines
110
100
6 monotonically and fails to recover alto-
4
NIPA GROWTH
90 2 gether, which stands out as a further
80
ics
70
0 peculiar feature of the current period.
-2
60 -4
50 -6 The Sharp Falloff in Consumer
40 -8
30 -10 Confidence
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 The length and severity of the Great
q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1
Recession are likely to make consumers
ICS NIPA Growth
less confident and more uncertain. We
ICS by Income care about uncertainty because economic
120
theory implies that, when consumers are
110 uncertain, they will delay purchases of
100
durable goods and save for precautionary
90
80 reasons. It follows that uncertainty can
ics
70
60 of current and future business conditions,
50 and of favorability of conditions for dura-
40
30
ble purchases. The index ranges from a
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 minimum of 0 to a maximum of 200.
q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1
ICS Black ICS White ICS Hispanic One fundamental aspect of the consumer
confidence data is that, in the case of
ICS by age the U.S. economy, growth in personal
120 spending is trailed closely by consumer
110
100
confidence, as illustrated by the histori-
90 cal trends in the top panel of figure 4. We
80
have measured personal spending in fig-
ics
70
60 ure 4 via the National Income and Product
50 Accounts (NIPA). During the Great Reces-
40
30 sion, figure 4 shows that the ICS declines
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008 2010 dramatically, right on the heels of the
q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1 q1
decline in personal spending. The depth
ICS 18–34 ICS 35–64 ICS 65+ of the decline is rivaled only by the low
level of consumer confidence reached
Source: BEA NIPA Table 2.3.1 and MIchigan Consumer Sentiment Survey.
Recession Trends • The Russell Sage Foundation and The Stanford Center on Poverty and Inequality
5 Consumption in the Great Recession
during the recession of the early 1980s. In other words, simi- these two groups. The same pattern of convergence and then
lar to its impact on actual consumption, the recession marks divergence also appears for age groups.
a complete turnaround in consumer confidence after several
decades of relative optimism, which peaked in the late 1990s. As a final point, it is worth noting the peculiar trend in con-
sumer confidence among Black respondents during the Great
Aggregate indices are likely to conceal potential differences Recession. Their confidence level increased at a remarkable
in perceptions among sociodemographic groups. In times of rate compared to that of White and Hispanic respondents.
economic growth, consumer confidence is predictably lower The effect is robust to differences in income, age, and educa-
among respondents who are poorer, older, and of Hispanic tion. Its timing, from the second quarter of 2008 onward, may
or African American origin. But recessions tend to narrow reflect a complex mixture of economic concerns and of politi-
between-group differences. cal hopes associated with the last presidential election. This
relative optimism resulting from the apparent “Obama effect”
The gap between the reported confidence level of respon- overshadows, without necessarily improving, the concrete
dents from the bottom and of respondents from the top economic consequences of the recession.
income quartiles narrows due to the higher rate of decline
in consumer confidence of high-income respondents. As the Widespread Financial Difficulties
second panel of figure 4 shows, the decline is most abrupt for Behind the decline in consumer confidence lies a concrete
the Great Recession, in which the confidence of respondents deterioration in the financial situations of individuals. Sixty
from the top income quartile lost 50 points between the first percent of the individuals in the Michigan consumer senti-
quarter of 2007 and the last quarter of 2008. For compari- ment sample reported that their financial situation in 2009
son, the level of consumer confidence for respondents from was worse than in the previous year. For comparison, only
the bottom income quartile dropped by approximately 20 half that number expressed that opinion in 2006. In figure
points in the same period. After the recession formally ends, 5, we stratify the sample by income (top and bottom quar-
it’s notable that consumer confidence recovers sharply for tile), age (less than 30, 31 to 60, and older than 60), and race
the top quartile but not for the bottom quartile, a divergence (White, Hispanic, and African American). We plot the portion
that is consistent with the objective economic experiences of of each group that reports that their situation has worsened
Reason why worse in 2009 by income groups by age groups by race groups
10% 4%
3%
Debt 4% 6%
5% 8%
4%
36% 4% 21%
Worse asset position 15% 7%
8% 30% 5%
14% 21%
10%
High(er) prices 18% 14%
30% 27% 27%
49% 27%
25%
Less work, hence less income 36% 51%
28% 13% 36%
7% 10%
Lower income from 11%
9% 5%
self-employment or property 7% 9% 5%
16% 17%
16%
Other 18% 16%
22% 18% 19%
47% 54%
51%
Proportion worse off 56% 60%
57% 52% 48%
Top 25 percent Thirty or Younger White
Bottom 25 percent Thirty-one to 60 Hispanic
Older than sixty Black
Recession Trends • The Russell Sage Foundation and The Stanford Center on Poverty and Inequality
6 Consumption in the Great Recession
(at the bottom of each panel) and, conditioning on reporting rent downturn. First, the rise in government transfers in the
a worse financial situation, display statistics on the reasons early period of the recession raised the disposable income
respondents provide. of low-income individuals, but such transfers weren’t able to
overcome the effects of other consumption-reducing forces
The differences in the reasons provided by specific sociode- (such as the destruction of wealth). Second, the deteriora-
mographic groups paint a coherent picture of the wide impact tion in consumer expenditures lasted longer than in any of
the recession has had on Americans, or at least on their per- the other recessions since the 1970s, and indeed consumer
ceptions of it. No group is spared, though each is affected expenditures still haven’t fully recovered. Third, consumption
in different ways. The historically privileged part of the pop- has also plunged deeper than in the past, leading Americans
ulation—Whites, middle-aged, elderly, and those with high not only to postpone costly purchases of durables but also to
incomes—perceive the recession distinctively as a threat to change their leisure habits and cut back even on subsistence
their wealth, whether in terms of financial assets or, to a lesser spending. Fourth, there was greatly increased insecurity
extent, of lower income from self-employment and property. about the future, an insecurity that’s still in evidence. Fifth, the
“A worse asset position” is the primary reason for 36 per- recession affected virtually all consumers, albeit unequally
cent of the top income quartile, for 30 percent of respondents and in different ways. Some, like the wealthy, experienced a
above 60 years of age, and for 21 percent of Whites. Almost serious shock, one that was both immediate and—as it turns
half of Hispanics and of young respondents cite the job mar- out—temporary. For the rest of us, the shock is likely to linger
ket as a reason for their worse financial situation in 2009. for as long as the problems in the wider economy continue.
Higher prices take precedence among nearly one-third of Afri-
can Americans and of respondents from the bottom income Our reading of the available evidence is that the Great Reces-
quartile. In addition, though to a far lesser extent, these two sion was not consumption driven. The fall in consumption
groups are concerned by worsening debt. occurred after, not before, the financial crisis and the defla-
tion of the housing bubble. However, once the consumption
Conclusion decline was in full swing, it acted to prolong the recession
At least when it comes to consumption, the Great Recession and the continuing downturn. The effects of this consump-
has certainly earned its capital letters. We have focused on tion decline continue to this day and make a recovery a less
five quite distinctive features of consumption during the cur- certain affair.
Additional Resources
Christelis, Dimitris, Dimitris Georgarakos, Stimulus Payments of 2008, NBER Working Congressional Budget Office. 2007. Housing
and Tullio Jappelli. 2012. Wealth Shocks, Paper No. 16684 . Wealth and Consumer Spending. Washington,
Unemployment Shocks, and Consumption DC: U.S. Government Printing Office.
in the Wake of the Great Recession. Netspar Petev, Ivaylo D., Luigi Pistaferri, and Itay
Discussion Paper No. 03/2012-010. Available Saporta-Eksten. 2011. “Consumption and McCully, Clinton P. 2011. Trends in Consumer
at: http://papers.ssrn.com/sol3/papers. the Great Recession: Analysis of Trends, Spending and Personal Saving, 1959-2009.
cfm?abstract_id=2046503 Perceptions, and Distributional Effects,” The Survey of Current Business, 91(6), 14-
Great Recession, edited by David B. Grusky, 23. Available at: http://www.bea.gov/scb/
Parker, Jonathan A., Nicholas S. Souleles, Bruce Western, and Christopher Wimer. New toc/0611cont.htm
David S. Johnson, and Robert McClelland. York: Russell Sage.
2011. Consumer Spending and the Economic
Suggested Citation
Petev, Ivaylo D., & Pistaferri, Luigi. 2012. Consumption in the Great This publication was supported by Grant Number AE00101 from the U.S. Department of Health
Recession. Stanford, CA: Stanford Center on Poverty and Inequality. and Human Services, Office of the Assistant Secretary for Planning and Evaluation (awarded by
Substance Abuse Mental Health Service Administration).
Its contents are solely the responsibility of the authors and do not necessarily represent the
official views of the U.S. Department of Health and Human Services, Office of the Assistant
Secretary for Planning and Evaluation (awarded by Substance Abuse Mental Health Service
Administration).
To further explore the data presented here and to produce customized graphs on recession trends, go to
www.recessiontrends.org
Recession Trends • The Russell Sage Foundation and The Stanford Center on Poverty and Inequality