This chapter bolsters the overall ecological theory of court reform by assessing an alternative explanation rooted primarily in top-down, state-centered logic – one assuming that leaders from the central CCP, state, or both have comprehensively determined the nature and extent of institutional change. One mechanism by which legal forms like judicial selection and promotion mechanisms could change in top-down fashion would be the central Party–State’s allocation of judicial personnel headcount (i.e., bianzhi) to local courts through the personnel management system. This chapter, however, presents evidence that this is not the case – despite differences in geography and development, court systems in Chengdu, Shenzhen, and Shanghai all demonstrated robust capabilities during the post-Mao era to expand and diversify their court and judicial structures, suggesting that top-down theories of court politics do not explain the local variation I describe in this book.

In Chapter 12, we explain the course of public policy – specifically, the degree to which governments intervene in citizens’ social and economic affairs. We argue that scale is negatively related to the level of government intervention through decreased social cohesion, representativeness and trust, particularism, and concentration, and through increased economies of scale. We also discuss the role of trade dependence, which we argue likely has ambivalent (countervailing) effects on intervention. In the empirical section, we turn to a variety of empirical terrains including the growth of the American state, the experience of small states (everywhere), and four policy realms – moral, fiscal, personnel, and social. Our results point to a stark contrast between subnational and national-level analyses. While subnational analyses show either a negative relationship between scale and intervention, or no relationship at all, national-level analyses show a strong (negative) relationship between the size of communities and the size of government. This apparent paradox has many possible answers, and so we leave it for future research to tease apart which factor(s) might be responsible.