In this study of the origins of age-related biases in contemporary social policy, Julia Lynch sets out to answer three interrelated questions that have thus far been inadequately explored. First, how do social policies in rich democracies differentially protect different age groups? Second, how does this ‘elderly/non-elderly’ public policy bias vary across the OECD welfare states? Third, how best to explain these variations across time and place? The first two chapters set out to define, measure and map the book's main dependent variable: the ‘age of welfare’ in 20 OECD democracies between 1985 and 2000. This variable is operationalised as the ENSR: Elderly to Non-elderly Spending Ratio. The numerator includes pensions and services for the elderly (defined to be those either aged 65 and above or those in formal retirement). The denominator includes primarily unemployment benefits, active labour market policies, family allowances and family services. The resulting ENSR data and rankings constitute a noteworthy contribution in and of themselves. Japan, the United States, Italy, Greece and Portugal come out as being the five most elderly oriented welfare states in the sample. Denmark, Sweden, Ireland, Belgium and Finland occupy the bottom five ranks. Not only is there very substantial observed ENSR variation within the OECD, interestingly this variation does not neatly accord with any of the immediately intuitive explanations, such as welfare regime type, levels of GDP per capita or of general social spending, or even the share of aged citizens within the total population.

Chapter 3 then explores a number of competing existing hypotheses about why countries may vary in the age orientation of their social policies: modernisation theory, power resources theory, familialism, neo-corporatism and contemporary grey voters’ influence all pass the review, all are found wanting. Instead, Lynch proposes an innovative historical-institutional explanation, grounded in path-dependent developments presumed to have driven policy trajectories over the early twentieth century. Key here is how the institutional design of welfare programs has had the largely unintended side-effect of shaping the electoral incentives and concomitant social policy emphases of elite politicians in their capacity as policymakers. Two momentous critical junctures on the supply side of democratic policymaking are argued to have shaped the age orientation of contemporary policies. First, in the early twentieth century, the structure of welfare state programs was enacted along either narrow occupational or more universal citizenship-based lines. According to Lynch, the former set of welfare states became progressively more conducive to elderly oriented policies as their protected core of labour market insiders grew older. A second watershed institutional moment came shortly after the Second World War, when the early pathways of the occupational regimes were reinforced by the particular model of political-electoral competition adopted. The critical bifurcation here was between particularistic (or clientelistic, or patronage-based) and programmatic (or general) models of competition. All of the particularistic models stuck to their occupational mode of welfare design, some of the programmatic models switched towards more citizenship-based design.

These mutually reinforcing constellations of electoral competition and welfare state structure, Lynch argues, have interacted over the course of the twentieth century to crystallise into distinct and relatively stable age orientations of social policy. Thus, occupational/particularistic constellations fostered clientelistic and targeted policies disproportionately favouring the elderly. Italy's myriad pension systems catering for narrow groups are a case in point here. By contrast, citizenship-based/programmatic constellations gave rise to much more broadly cast safety nets, precluding tit-for-tat exchanges of votes for targeted benefits. Modes of taxation mattered crucially, as they shaped the feasible formats for social spending. Universalistic taxes, unlike more targeted forms of taxation, facilitated more broadly spread benefits.

How does the theory fare empirically? At first sight, there appears to be an acceptable fit with the evidence. Four of the top five ENSR scorers mentioned earlier (Portugal excepted) happen to combine occupational welfare regimes with particularistic models of political competition. Four of the bottom five scorers (Belgium excepted) can be classified as citizenship-based and programmatic regimes. Surprisingly, however, no sustained explanation is given for the anomalous positions, from the point of Lynch's theory, of Portugal and Belgium. The latter country, for instance, is characterised by precisely the same institutional constellation as the most elderly oriented welfare states (p. 57), despite scoring fourth lowest in the ENSR ranking (p. 5).

Lynch's fact-finding strategy in these chapters is certainly helpful in establishing precise values for the ENSR and in raising additional theoretical puzzles. But it might have been criticised for being overly reliant on aggregate macro-quantitative indicators or, more precisely, ratios between such indicators. Commendably, therefore, the author supplements this early analysis with three rigorous chapter-length studies of family allowances, unemployment benefits and pensions in two countries: Italy and the Netherlands. These case studies, too, are admirably executed, and they provide a wealth of policy detail and historical background to back up the book's general claims in both cases. Not surprisingly perhaps, the highly patronage-ridden and tax-evading case of Italy is where the theory appears to work best. Italy is placed on one extreme of both Lynch's empirical scale (third-highest ENSR value) and her theoretical scale (strongly occupationalist/particularistic). Picking the Netherlands as a second case study is somewhat more idiosyncratic, however. The Netherlands resembled Italy until the Second World War, after which a more programmatic model of electoral competition led to more citizenship-based, and less elderly oriented, welfare policies. Today, the Dutch regime is a middle-of-the-pack hybrid on both Lynch's ENSR scale and on her theoretical continuum. In other words, these two case studies help to illustrate the key role of the second, post-War, juncture in diverging the Dutch from the Italian regime. But they convey little additional information about the origins of the least elderly oriented social policy regimes, as none of these countries – a set including all Scandinavian and all Anglo-Saxon countries except the USA – is covered. For future empirical tests of Lynch's general theory, quantitative and/or qualitative comparative analyses on the entire OECD sample might therefore be more desirable, and they would be equally feasible.

In sum, Julia Lynch has made an unusually creative and insightful contribution to comparative social policy theory. The great virtue of Age in the Welfare State is that it succeeds in answering all three of its major research questions in a robust, systematic and thought-provoking way. The book's main concern, public policy age bias, is of ever-growing relevance to sociologists, political scientists and economists as well, not just because of the accelerated population aging affecting most OECD countries today, but also because such bias is likely to have significant repercussions for variables such as labour market participation, fertility and family planning, voting behaviour, and skills and human capital formation. In addition, this book is highly recommended for anyone interested in the interplay of liberal democracy and public policy, in modes of historical-institutional explanation, and in positive and normative theories of intergenerational justice.