In most countries, households’ consumption expenditure accounts for more than half of GDP. How much households spend reflects their living standards, and typically households try to smooth consumption over time. As a result, consumption usually does not display large variability and, therefore, has found comparatively little academic and policy attention.
Since the onset of the financial crisis, consumption has dropped markedly in many countries. Figure 1 shows the development of per capita real consumption in the main European crisis countries.
Figure 1. Real per-capita consumption expenditure (2000Q1 = 100)
Note: Authors’ calculations, updated from O’Connell, O’Toole and Znuderl (2013). Consumption expenditure are deflated with the CPI (Source: Eurostat) and corrected for population growth (Source: United Nations).
Why does consumption decline during a crisis?
Several explanations for the decline in household spending in the crisis countries have been offered.
- One explanation is that permanent income has declined, causing a decline in consumption as posited by the permanent income hypothesis (Friedman 1957).
- A second explanation suggests that consumption has fallen because of credit or liquidity constraints (Aron et al. 2012; Jappelli and Pistaferri 2010).
If actual income falls and households have neither accumulated savings, nor access to credit, their consumption has to adjust downwards, even if permanent income stays constant. The dynamics of consumption, therefore, may change during financial crises when households find access to credit more difficult because of heightened bank-risk aversion, tightened credit standards, or reduced collateral values.
• A third explanation for depressed consumption is precautionary or buffer-stock savings (Mody et al. 2012). In the context of the financial crisis, savings may be used more to deleverage (see Bornhorst and Ruiz Arranz 2013).
Once households have rebuilt their balance sheets, future consumption can again be financed by new credit. Moreover, even when credit constraints are not currently binding, risk-averse households may try to avoid a situation of binding credit constraints in the future by building up savings beforehand (e.g. Jappelli 1990, Deaton 1991, Carroll 1992, Carroll and Kimball 2001, Carroll and Toche 2009).
In order to formulate policies to revive consumption, two questions become crucial:
- Whether during financial crises something fundamental changes in households’ consumption patterns; and
- To what extent credit constraints account for the drop in consumption.
The macroeconomic perspective
Gerlach-Kristen, O’Connell and O’Toole (2013) examine macroeconomic data in 23 countries over 32 years (1981:Q1-2011:Q4) to assess the impact of financial crises on aggregate consumption. Consumption is modelled in an error-correction setup, in which there exists a long-run relationship between consumption, income, housing, and other wealth. To account for the fact that consumption functions may differ between countries, the analysis uses a mean-group estimator.
Results suggest that during financial crises, only the short-run dynamics of consumption growth change. Consumption growth is lower particularly during banking crises – as defined by Laeven and Valencia (2012) – rather than during crises preceded by credit or housing booms.
There is also some evidence that lower income growth depresses consumption growth during crises. One interpretation of this finding is that households are unable to smooth consumption during financial crises, which suggests a role of credit constraints. However, the analysis does not find a role of variables capturing credit volumes or costs. A potential explanation for this failure is that aggregate macroeconomic data may mask the driving forces behind the consumption decisions taken by individual households (see also Dynan 2012).
The microeconomic perspective
To take household heterogeneity better into account, Gerlach-Kristen and Merola (2013) analyse consumption decisions and credit constraints on a micro level. They first present a small DSGE model with occasionally-binding credit constraints, and then assess whether the predictions made by the model are compatible with Irish household data, collected during the financial crisis.
Figure 2 shows consumption expenditure and income by age, and clearly illustrates the heterogeneity of household experiences during the crisis, which has affected younger households more than older ones. One distinct feature of younger Irish households is that they bought property at the height of the boom. Today many of them are in negative equity, and may be trying to rebuild their balance sheets.
Figure 2. Weekly real consumption and disposable income by age group
Note: Data from the latest for waves of the Irish Household Budget Survey. Values in 2010 euros, age of the household reference person. Average income and consumption by group, taking into account the grossing factors capturing the representativeness of the individual households interviewed in the HBS. Percentage numbers indicate the size of a group in question relative to the full population.
The theoretical model in Gerlach-Kristen and Merola (2013) illustrates that credit constraints can arise from falling property prices, which reduce the value of collateral, and thus raise the leverage ratio of indebted households. Constrained households cease to smooth consumption and use their savings to deleverage in order to improve their future access to credit.
The predictions of the model are corroborated by the empirical analysis. Mortgage households in, or close to negative equity, consume less than the average household, and the permanent income hypothesis is rejected for this population group. The deviation from the permanent income hypothesis is stronger the higher a household’s leverage ratio. This is compatible with deleveraging efforts of highly leveraged households and the existence of credit constraints. Moreover, both the DSGE model and the empirical analysis show that households stop smoothing consumption even if they only expect a decline of house prices in the future, and currently are not facing a credit constraint yet. This suggests a role of precautionary savings.
Interestingly, the permanent income hypothesis is not rejected for the average Irish household, suggesting that most Irish households have continued to smooth consumption in the crisis. It is, therefore, not surprising that analyses at the aggregate level using macroeconomic data find it difficult to identify the exact channel through which credit constraints impact consumption.
In summary, our research papers provide new evidence for policies that may help avoid large declines in household consumption during financial crises. We show that it is crucial to take household heterogeneity into account in order to identify the extent and effects of credit constraints, and, therefore, micro studies are needed alongside macro analyses.
We find that consumption growth is lower during financial crises, particularly during banking crises, and that a drop in income reduces consumption in the short run. This suggests that consumption smoothing is disrupted, and that credit constraints might play a role.
Using Irish data, we show that in a financial crisis that is accompanied by a property bust, consumption smoothing is most disrupted for mortgage households that are highly leveraged in the housing market. One policy measure to reduce the impact of credit constraints is, thus, to limit the exposure households can take, i.e. to enforce a regulatory maximum loan-to-value ratio.
Aron J, J V Duca, J Muellbauer, K Murata and A Murphy (2012), “Credit, housing collateral, and consumption: Evidence from Japan, the U.K., and the U.S.”, Review of Income and Wealth 58 (3), 397-423.
Bornhorst, F and M Ruiz Arranz (2013),“The perils of private-sector deleveraging in the Eurozone”, VoxEU.org, 20 November.
Carroll, C D (1992), “The buffer-stock theory of saving: some macroeconomic evidence”, Brookings Papers on Economic Activity 2, 61-156.
Carroll, C D and M S Kimball (2001), “Liquidity constraints and precautionary saving”, NBER Working Paper 8496.
Carroll, C D and P Toche (2009), “A tractable model of buffer stock saving”, NBER Working Paper 15265.
Deaton, A (1991), “Saving and liquidity constraints”, Econometrica 59(5), 1221-1248.
Dynan, K (2012), “Is the household debt overhang holding back consumption", Brookings Papers on Economic Activity, Spring 2012.
Friedman, M (1957), “A theory of the consumption function”, National Bureau of Economic Research.
Gerlach-Kristen P, B O’Connell and C O’Toole (2013), “How do banking crises affect consumption? Evidence from international crisis episodes”, Economic and Social Research Institute Working Paper 464.
Gerlach-Kristen, P and R Merola (2013), “Consumption and credit constraints: A model and evidence for Ireland”, Economic and Social Research Institute Working Paper 471.
Jappelli, T (1990), “Who is credit constrained in the U.S. economy?”, The Quarterly Journal of Economics 105(1), 219-34.
Jappelli, T and L Pistaferri (2010), “The consumption response to income changes”, VoxEU.org, 2 April.
Laeven, L and F Valencia (2012), “Systemic banking crises database: An update”, IMF Working Papers, International Monetary Fund.
Mody, A, D Sandri and F Ohnsorge (2012), “Precautionary savings in the Great Recession”, VoxEU.org, 22 February.
O’Connell, B, C. O’Toole and N Znuderl (2013), “Trends in consumption since the crisis”, Research Note, Quarterly Economic Commentary (January), Economic and Social Research Institute.