This book builds from a conference held in Tokyo in June 2006 that was organised by the editors. It has three parts: ‘Population aging and labour markets’; ‘Saving and wealth’; and ‘Policy’. A first observation is that the title is somewhat misleading – it suggests that the only link between population ageing and the macro-economy that is covered by this book consists of intergenerational transfers, which is not the case. Only three of the 10 chapters are specifically on intergenerational transfers while the others focus on different aspects of population ageing (which are certainly as relevant and interesting as objects of economic research). Thus, if a reader is particularly interested in intergenerational transfers, this book is not the first choice, even though the three chapters on the subject are certainly worthwhile. More importantly, the book does provide a timely and broad overview of the literature on how population ageing affects the macro-economy and of the methods used in contemporary economics to analyse these effects. This brings me to a second general comment. Several chapters are highly technical and firmly embedded in the economics literature on ageing and demographic change. The book is therefore particularly suited for economists who work in this area and for advanced graduate students seeking an overview of this important area of research. Readers from other backgrounds may feel less inclined to dig deep into the more technical chapters. In the remainder of this review, I focus on three chapters that are close to my own interests.

The chapter by Prskawetz, Mahlberg and Skirbeek deals with an important but under-researched aspect of population ageing. Very little is known about how productivity changes over a worker's life. Since there are several potential causal mechanisms, there is no clear theoretical prediction. The relationship between age and productivity is thus mostly an empirical question. Unfortunately, empirical studies are plagued by measurement problems since a worker's individual productivity is not easily observed. Moreover, selection problems arise since only active workers can be sampled for a study of individual labour productivity. For instance, it is likely that older workers who are still employed are more productive than others of the same age who have already left the labour force and who therefore are not in the sample. These and other methodological problems imply that it is difficult, if not impossible, to estimate the causal effect of age on productivity using micro-data. Prskawetz and colleagues are fully aware of these problems and stress that their results do not warrant a causal interpretation. With this caveat in mind, they analyse matched firm-worker data from Austria. For small firms, they find a negative relationship between the share of old workers and productivity. For large firms, the results are less clear cut. In some sectors, larger shares of younger workers are associated with lower productivity. Firm-specific human capital may be the causal mechanism behind this observation. Clearly, more research is needed here, but the chapter shows how matched firm-worker data can be fruitfully employed in this endeavour.

The chapter by Bloom, Canning, Moore and Song investigates the effect of subjective survival on retirement and wealth. This analysis builds on earlier theoretical work by the same authors. The starting point for this line of research is the theoretical insight that retirement and savings decisions are jointly determined by such factors as health status, expected longevity, and features of the social security system. In particular, whether or not the retirement age is mandatory influences the degree to which individuals react to increases in longevity by raising their savings. The present chapter contains an empirical analysis of this mechanism. The data were drawn from the United States Health and Retirement Study, a panel study of individuals aged 50 or more years that collected detailed information on households' financial situation as well as individuals' occupational histories, health status, and survival expectations. The main finding is that higher life expectancy (at the individual level) is associated with higher household wealth but does not greatly affect the timing of retirement. This finding is in line with predictions based on the authors' earlier theoretical research. They conclude that the institutional restrictions implied by the social security system shut down an important mechanism by which individuals may adjust to increased longevity (namely working longer). This is certainly an important finding, but in my view the interpretation leaves open the issue of why retirement ages are implicitly or explicitly fixed in so many countries. One is aware that countries such as France, Italy and Germany are currently experiencing strong, if not violent, public resistance against increasing the statutory retirement age. This observation is in stark contrast to a theoretical model that predicts that the public should be lobbying for increases in the retirement age. More research is certainly needed on this important public policy issue, and the analysis presented by Bloom and colleagues in this chapter and elsewhere is a good start.

The chapter by Hock and Weil presents a state-of-the-art analysis of the effects of population ageing on aggregate consumption. This is a broad research question, and the authors naturally concentrate on a few aspects. The central question asked is: What is the level of fertility that maximises the consumption possibilities of the currently living and of all future generations? To answer this question, they construct a ‘continuous-time overlapping generations’ model, and calibrate it using demographic data from the United States and Japan. Their main finding is that current fertility levels are too low – in the long-run consumption would be higher if fertility rates were higher. The authors also show that while governments should implement policies that raise fertility, it will be difficult to get public approval since the benefits of such policies come to fruition only a few decades ahead. In the short run, having more children creates a greater fiscal burden. This chapter highlights a well-known problem: demographic processes evolve very slowly, and optimal policies are hard to implement when they benefit only future generations, not current voters.