The more uncertain people are about their future income, the more they will save. And households with no access to credit markets or that are unable to share risks among their members are much more sensitive to uncertainty, saving even more of their money for precautionary reasons. In the September 1996 issue of the Economic Journal, Philip Merrigan and Michel Normandin provide convincing statistical evidence to confirm these widely accepted beliefs about attitudes towards risk and the ''feel-good'' factor. Drawing on a vast array of household data on consumption from the UK Family Expenditure Survey (FES) since 1968, the authors conclude that from the point of view of society as a whole, more uncertain environments are very costly in terms of consumption. While it is reassuring that individuals are rational enough to protect themselves by substituting consumption with precautionary savings, a less uncertain world would lead to higher consumption and higher overall well-being. Merrigan and Normandin''s finding that greater uncertainty about future income leads systematically to larger current saving is intuitively very plausible. But until now, it has been difficult to find statistically convincing evidence that households do adjust their savings behaviour in the face of uncertainty. Part of the reason lies in the difficulty in obtaining data that permit the investigation of savings behaviour.
Recent theoretical advances in consumption theory and econometrics, as well as the data from the FES, allow the analysis of precautionary savings as distinct from overall household savings. While past studies usually rely on panel data (data with several observations from the same household) to study precautionary savings behaviour, these researchers'' methodology permits identification of this behaviour even if the FES is not a panel. The repeated nature of the survey (approximately 57,000 observations over the period since 1968), the quality of the data, the long time horizon and the sheer size of the data set all explain their new discoveries about savings behaviour.
By splitting their sample into different groups, Merrigan and Normandin find that ''liquidity constrained'' households or households where risks cannot be shared are much more sensitive to uncertainty. Such results provide yet another signal to governments of the importance of providing stable and credible policies. This is particularly vital since uncertainty has more drastic effects on people lacking access to credit markets, which would permit ''smoothing'' of their consumption in difficult financial times. The results also imply that an observed increase in aggregate savings does not necessarily imply good times: part of this increase could reflect a more uncertain future.
''Precautionary Savings Motives: An Assessment from UK Time Series of Cross-sections'' by Philip Merrigan and Michel Normandin is published in the September 1996 issue of the Economic Journal. Merrigan is at the University of Quebec in Montreal and Normandin at the Centre for Research on Economic Fluctuations and Employment. The research draws on data from the Family Expenditure Survey made available by the CSO through the Economic and Social Research Council Data Archive.