Abstract

Comparative sociologists mostly ignore wide differences in criminality and incarceration rates among modern western societies; with notable exceptions, students of the prison take scant notice of research comparing political economies, welfare regimes, and patterns of inequality. This article outlines an opportunity structures model of imprisonment that bridges this gap by treating incarceration trends as byproducts of the institutional organization of opportunities over the life course. Using a sample of 15 rich democracies observed over four decades, empirical attention focuses on three levels of analysis: the capacities of alternative life course paths, the distribution of political power, and institutional differences in state structures and policy regimes. Hypothesized cross-level interactions call for the specification of a hierarchical model to be estimated within a Bayesian framework. Results conform to the expectations of the opportunity structures model and support many of its specific predictions. Introduction Law making and law enforcement are central functions of the modern nation state, and the wide variation in crime rates and levels of incarceration among industrialized democracies presents an attractive set of puzzles for comparative analysis. But these issues are largely ignored in the literatures on comparative politics, socio-economics, and social policy. The focus of this article is on incarceration, and while the number of cross-national analyses in this area is growing, further development requires scholars to overcome two significant challenges. The first is the theoretically promiscuous quality of the imprisonment literature. Recent accounts have drawn links between punishment practices and social welfare regimes A second challenge to comparative research, particularly involving quantitative data, is the difficulty of comparing punishment practices across nation states. Research on welfare, education, labour markets, and macroeconomics benefits from the rough isomorphism in institutional forms in these domains, and the fact that, thanks to the homogenizing influence of IGOs like UNESCO, the ILO, the OECD, and the World Bank, data are reported in consistent ways across the developed democracies. In contrast, comparability across modern systems of criminal law is problematic because of their disparate and remote historical origins, the tendency of national legal institutions towards self-referentiality and inertia, the lack of internationally valid conventions for recording data, and wide cross-national variation in the structure of institutional fields assigned jurisdiction over criminal behaviour. This is not just a measurement problem. The more fundamental issue is that causal processes determining forms and rates of incarceration may vary in different institutional contexts. Given this possibility, it is reasonable to ask whether any single model can reasonably be applied across a range of societies. This article attempts to meet both challenges. First, in a bid for theoretical synthesis, I outline and test an opportunity structures model that aims on the one hand to weave together strands from the new research on penalty, and on the other to integrate the analysis of imprisonment with recent macrosociological research on mobility regimes, life-course institutions, and social policy. A preliminary version of the opportunity structures model was described in an earlier study of five Common Law democracies Second, this emphasis on context suggests an analytical strategy that not only acknowledges cross-national causal heterogeneity, but also seeks to take advantage of it. I argue that social policies in different countries-including penal regimes-are influenced by historically rooted and relatively stable institutional structures that vary widely among the developed democracies. In making this argument, I borrow freely from the comparative welfare state literature, which has demonstrated that institutional differences in state structures To forecast my argument, democracies with tightly regulated labour markets and bureaucratically powerful national states are more assertive in regulating the distribution of life-course opportunities, with consequences for penal policy directly, and for the political opportunities available to non-state interest groups seeking to influence social policy. An adequate test of this argument requires both a broader range of data and a more sophisticated modelling strategy than was used in the earlier five-country study. This study uses data from 15 rich democracies observed over 40 years to capture wide variation in institutional regimes, and a multilevel modelling approach that can analyse complex interactions between relatively stable institutional structures and more fluid dimensions of the opportunity space within each country, and the consequences of both for incarceration rates. I motivate hypotheses and describe my modelling strategy in more detail in the sections that follow.