This is a scholarly survey of the economic literature on risk and economic behaviour in the context of rural poverty in developing countries. The focus is largely on theoretical approaches, although the issues are placed in a broader context and available empirical studies (including some in anthropology) are cited. The exposition of the theories is clear and well structured, but not for a beginner-some grounding in the techniques of development microeconomics is a prerequisite for fully following the text. Risk, and modeling behaviour under risk, requires the use of technical material; Poisson processes appear very early (p5) and Bellman equations as early as page 24. Although the theoretical exposition is rather terse the intuition is presented clearly; those somewhat daunted by the formalism would still be able to follow the arguments.
The book is organized into five substantive chapters, with a concise introduction that sets the scene and a conclusion that draws out some clear policy conclusions. A brief Chapter 2 defines concepts of risk and explains why the analysis of risk is central to understanding rural poverty. Chapter 3 presents the core models, most based on (expected) utility maximization, of risk coping strategies-measures to reduce exposure, forms of saving, risk sharing behaviour and issues relating to intra-household allocation. Chapter 4 reviews reasons (in the context of the formal models) why these coping strategies may fail given the environment faced by the poor-limits to reducing exposure, difficulties in risk sharing (especially where power is unequally distributed or commitment difficult to enforce), and information asymmetries (e.g. in credit markets). Chapter 5 is the most overtly theoretical, in the sense that the formal models are rather abstract, concentrating on models of the relationship between risk and inequality. Chapter 6 then explores links between risk and aspects of development-nutrition, technical change, commercial versus subsistence farming and constraints on investment and precautionary savings. Although, as with previous chapters, the discussion is framed around formal theoretical models, there is substantive and informative text.
Chapter 7 presents the conclusions, stressing that the 'magnitude and range of shocks that affect rural populations in the Third World is without comparison in developed economies ' (p. 196). Fafchamps argues that we have learned quite a lot from the application of theoretical models: society's ability to share risks is severely constrained, rendering kin and informal networks especially important. Some policies often advocated to reduce poverty may be ineffective because they consolidate or promote inequality: insurance does not reduce inequality, but may reduce social mobility (with an adverse impact on the poor); accumulation of land, manpower and resources promotes inequality (as distinct from accumulation of skills and equipment). Encouraging the poor to adopt new technologies requires that the technologies can be adopted on a small-scale in parallel with traditional methods. The need for the poor to engage in precautionary savings helps to explain both low levels of investment and why the poor hold assets that are relatively unproductive. Patronage societies encourage accumulation of social capital, which is conducive to trade, but discourage widespread accumulation of physical capital, which is required for industrialisation. The theoretical models do help to explain the slow pace of development in rural communities, and especially why many interventions have not been particularly effective. Fafchamps then considers 'what we do not know', in particular 'there is dramatic dearth of work on savings and the use of financial instruments by the rural poor' (p. 203). Although they have become core 'units of analysis' in much of development microeconomic theory, rather little is known about the dynamics of households and networks-how do they form, why do they collapse, and how are resources allocated. Although many risk coping strategies have been identified and studied, there is little