Government debt and excess sensitivity of private consumption: estimates from OECD countries

Economic Inquiry, Oct, 2004 by Lorenzo Pozzi, Freddy Heylen, Maarten Dossche

This article investigates the role of government debt in the degree of excess sensitivity of private consumption to current disposable income. Because this type of excess sensitivity is generally attributed to liquidity constraints, this amounts to a test of the impact of government debt on the amount of credit extended to individuals. Controlling for financial liberalization we find that, for a panel of OECD countries in the 1990s, a high government debt leads to more excess sensitivity. This result supports the idea that a high government debt induces lenders to tighten credit conditions. Our findings survive several robustness checks. (JEL E21, E62, C33)

I. INTRODUCTION

Assuming rational forward-looking consumers and perfect capital markets, Hall (1978) demonstrated that under the permanent income hypothesis, consumption should follow a random walk. Most studies during the past two decades have, however, rejected this prediction. In particular, they have concluded that private consumption is more sensitive to current disposable income than is consistent with the permanent income hypothesis. In the literature several explanations for this "excess sensitivity" have been put forward.

Typically these explanations come down to dropping one or more of Hall's assumptions. For example, Flavin (1985) and Romer (2001, 540) have referred to myopic behavior from a significant part of the consumers, that is, a deviation from the basic postulate of rational forward-looking agents. Other authors, like Flavin (1981, 1985) and Campbell and Mankiw (1990), have attributed excess sensitivity to credit market imperfections and liquidity constraints, preventing rational consumers from realizing their desired consumption. Other potential explanations for observed excess sensitivity to income relate to precautionary savings as in Barsky et al. (1986) and Carroll (1992), to imperfect information as in Goodfriend (1992) and Pischke (1995) and to, as noted by Campbell and Mankiw (1990), the misspecification of the estimated consumption function. The focus of this article is on liquidity constraints and myopia.

Building on the idea of liquidity constraints, several authors have more recently endogenized the degree of excess sensitivity. Cross-sectionally, Jappelli and Pagano (1989), Campbell and Mankiw (1991), and Sarantis and Stewart (2003) find that countries with better developed capital markets and easier access to credit have lower excess sensitivity of private consumption. Haliassos and Christou (2000) cannot reject that countries with high concentration and low efficiency in the banking sector have higher excess sensitivity. Evans and Karras (1998) show for 66 countries that the excess sensitivity of consumption to disposable income is lower in countries with high savings rates. A high savings rate implies that consumers accumulate more wealth, which makes them less vulnerable to liquidity constraints. A number of studies have investigated the hypothesis that the deregulation of credit markets in many countries during the last decades has over time lowered the fraction of credit constrained consumers and the excess sensitivity of private consumption. Bayoumi and Koujianou (1990), Blundell-Wignall et al. (1995), McKiernan (1996), and Girardin et al. (2000) can confirm this hypothesis for several Organisation for Economic Co-operation and Development (OECD) countries. However, Campbell and Mankiw (1991) cannot. Finally, Bacchetta and Gerlach (1997) have demonstrated the role of (endogenous) liquidity constraints for private consumption from a different perspective. They show excess sensitivity of consumption to credit aggregates in the United States, Canada, the United Kingdom, France, and Japan. As to the evolution of excess sensitivity over time, they only observe a clear tendency of decline in the United States. Despite financial liberalization, they do not observe this tendency in the other countries.

This article focuses on the role of government debt for the excess sensitivity of private consumption. To the best of our knowledge, the empirical literature on excess sensitivity has until now disregarded government debt. Theoretically, however, a number of channels exist. On the one hand, a (very) high or (rapidly) increasing government debt ratio may reduce the fraction of myopic consumers. Excess sensitivity of private consumption to current disposable income should then fall. On the other hand, high government debt may undermine the ability or willingness of banks to extend credit to individuals. The incidence of liquidity constraints and excess sensitivity of private consumption may then rise. In section II we develop these theoretical channels somewhat further and put them into a workable econometric framework. In particular, we derive an equation for the growth rate of private consumption, with the degree of excess sensitivity being a function of the government debt ratio (and the degree of financial liberalization). In section III we estimate this consumption equation for 15 to 19 OECD countries. We make use of panel estimation methods that allow us to correct for simultaneity and heterogeneity. Our results support the idea that a high government debt ratio implies tighter credit conditions and an increase of excess sensitivity. This unfavorable effect of rising government debt on credit conditions seems to exist especially when debt is already at a high level. Section IV summarizes our main results and their policy implications.