Among the less researched and more urgent questions is why and to what extent are various groups of citizens in advanced democratic states better protected against social risks than others. Focusing on unemployment benefits, George Picot finds the explanation for what he terms ‘policy segmentation’ in the competitive behaviour of political parties. These political parties reflect the policy preferences of specific constituencies and seek to win the support of others for the policies they advocate. However, explanations of party policy by such factors alone come up against research that points to varying degrees of party autonomy in the definition of goals and strategies. Furthermore, parties do not act only as agents of interest aggregation or instruments of social change, but also strive to maximise votes and win office. All these are inevitably influenced by electoral rivalry. Consequently, Picot argues that it is party competition itself, largely determined by the number of contenders and their positioning in the political space, which has an independent explaining effect on party policy making. The combination of party competition theory with welfare state theory allows him to generate three hypotheses. First, that the segmentation of social policies will reflect the number of ‘relevant’ (Sartori, Reference Sartori1976) party system components. The smaller the number, the more centripetal is the competition and the more universalistic the social programmes they stand for. The larger the number of parties, the narrower the social interests they represent, and the greater segmentation of unemployment protection programmes they advocate. Second, that ideological polarisation of a party system results in a higher extent of policy segmentation. Parties occupying the poles of the ideological continuum do not expect to win office and are free to attract voters by far reaching proposals. The need of moderates to counter these will result in centrifugal dynamics. Furthermore, since parties occupying the extreme polls tend to concentrate on questions of principle, they may oppose policies advocated by moderates and seek to block them by parliamentary action or extra parliamentary protest. When they do agree to bargain they would settle only for policies carrying particularistic advantages to their electorate, thus resulting in policy segmentation. Finally, Picot argues that in systems characterised by what Sartori (Reference Sartori1976) called ‘moderate pluralism’ rivals seek to attract the electorates of close neighbours in the political space, and hence focus on policies rather than on principles. The resulting segmentation will reflect the specific party configuration, but we can expect it to be less than in polarised pluralism and larger than in centripetal competitions.

Picot uses Italy and Germany to test his hypotheses. Broadly similar patterns of post-World War II socio-economic development, institutional setup and partisan configuration of government allow him to concentrate on the nature and evolution of the two countries’ party systems and on their welfare policy consequences. The development of the Italian and German party systems and policies has been well documented elsewhere. Though interesting, the account contains little that would surprise the political scientist. But the originality of this work does not lie in the description of the respective party or welfare systems. Rather, it is in the explanatory link between the two: the connection between the exercise of equal citizenship and unequal government policy outputs. And from this perspective, the book is original and persuasive. Moving back and forth between the two realms and the two countries, the comparison is facilitated, at the cost of some repetitiveness, by clear summaries that enable the reader to take stock of the argument and chart its development ahead. The analysis stretches across two time periods. During the years of post-World War II welfare expansion, the Italian fragmented and polarised party system brought about a system of unemployment benefits that was highly segmented, with wide differences in benefit levels. In contrast, the centralised competition between the three German parties resulted in a benefit system that was inclusive and fairly equal in its generosity. As of the mid 1970s, austerity forced the two countries to contain welfare costs. At first, the polarisation of the Italian party system led to the blocking of any significant reform. But as of the 1990s, the formation of pre-election party coalitions and the centrifugal competition between them produced a series of gradual reforms that together amounted to significant reduction in generosity differences, even if the persistence of centrifugal tendencies within the party clusters prevented more fundamental shifts. In Germany, a limited rise in the demand for welfare reform moved the CDU and the FDP further to the right. The Social Democrats, who were not held back by any relevant left-wing competitor, were pulled along towards the centre. In the period of welfare expansion, core voters were taken care of first and other voters later; under the conditions of austerity, the latter were the first to suffer retrenchment and centripetal competition resulted in increased segmentation of unemployment benefits.

To examine the wider applicability of the results, Picot utilises data on the party systems and unemployment benefits in fifteen advanced welfare states. With the exception of the US, party systems have had an important impact not only on the emergence of different patterns of segmentation during the era of welfare expansion but also on the changes of benefit segmentation during the period of reform. Such results show the promise of research that would deepen our understanding of the linkage between party competitive behaviour and policy outcomes. Picot points to three questions that require further research. How do the different factors interact or reinforce each other to produce policy? Where potential party policies impact party goals in a dissimilar manner, what would determine decisions? And how do different types of party organisation affect such decisions? Answers may offer, for instance, a better understanding of the combination of two parties and highly segmented unemployment benefits in the USA. The explanation offered here, that the Presidential system and primaries weaken the parties while strengthening the individual members of congress, is tentative at best (after all, even weak parties and strong candidates wish to win elections). In brief, the book constitutes an unusually stimulating contribution towards a more adequate theory of unemployment policy that is ever more pressing in the circumstances of our time.