Understanding Consumption.

• Author: Cushing, Matthew J.

This book is a lucid and even entertaining introduction to modern research in consumption. The book is an expanded and updated version of lectures delivered by Professor Deaton at Oxford in May 1991 and is accessible to graduate students or advanced (British) undergraduates. The book retains the lecture format. The result is a coherent development of modern consumption theory from first principles to the frontiers of current theoretical and empirical research. Deaton avoids the tedium of a "textbook" or review article treatment that seeks simply to catalogue or summarize existing work. Instead Deaton provides us with his insights into recent empirical and theoretical developments and seeks to provide the intuition behind these developments.

The underlying theme of this book is a research agenda that seeks to explain aggregate consumption behavior as the result of individual optimizing behavior. The theoretical arguments are tied closely to econometric practice and empirical evidence throughout the book. Apart from an admitted reluctance to deviate from the optimizing agent paradigm, the treatment is remarkably undogmatic. Deaton draws insight from a rich class of theoretical structures encompassing, for example, non-time-separable utility (habit formation), liquidity constraints, and uncertainty (precautionary motive). Similarly, Deaton draws his evidence from a diverse set of empirical studies employing panel and aggregate data from across countries and time.

A major focus of this work is the need to pay careful attention to aggregation issues. Deaton has little patience for consumption paradoxes arising from models applying the fiction of a "representative" agent. Although individuals might be thought to optimize, we have very little reason to believe that fictitious representative agents do. Carefully aggregating individual behavior is, to Deaton, not a nuisance, but a rich source of predictions for aggregate consumption behavior.

Chapter 1 is a useful, unified treatment of the pure theory of intertemporal choice. Of particular interest is his discussion of the conditions under which aggregation across individuals, time and commodities affect predictions of aggregate behavior. Many empirical researchers are blithely unaware of the stringency of the assumptions required to justify aggregation.

Chapter 2 focuses on the long run relationship between savings rates and aggregate income growth and also the long run relationship between real...