PART THREE - CONCLUSIONS

Summary

We began this book by contrasting two normative models of governance: decentralism and centripetalism (see Table 1.1). We argued that institutions combining centralized authority and popular inclusion are likely to lead to better governance overall. This is the theory of centripetalism. Three institutions are paramount to this theory, in terms of both their causal impact and their expected causal exogeneity: unitarism (the absence of federalism and bicameralism), parliamentarism (the absence of a directly elected executive with policy-making powers), and closed-list PR (defined in contrast to majoritarian and preferential-vote systems).

In the first part of the book, we attempted to trace the causal pathways that might plausibly connect centripetal constitutions with good governance outcomes across a range of policy areas. We argued that these causal mechanisms could be profitably (though not uniquely) reduced to three intermediate factors: party government, conflict mediation, and policy coordination. Figure 1.2 depicts the expected interrelationships among these concepts.

In the second part of the book, we generated specific research hypotheses and examined empirical results from a series of cross-national regressions employing a wide range of policy outcomes in the areas of political, economic, and human development. These showed strong correlations between centripetal institutions and measures of good governance, patterns that are robust to changes in model specification and measurement techniques. Table 6.12 summarizes the results.