During the neo-liberal reform wave of the 1990s, scores of developing countries privatized infrastructure services such as highways, utilities and ports. This dissertation provides a new theoretical explanation of the varied outcomes of infrastructure privatization in weak institutional environments and an empirical assessment of the argument in the water and sanitation sector.

While predominant theoretical explanations of economic growth and investment in regulated industries focus on the ability of institutional configurations and policy design to constrain political and economic actors, the explanation advanced in this dissertation centers on the ability of different types of firms to maintain "relational contracts"—contracts that involve direct negotiation in the face of unforeseen circumstances—with host governments. The theory contends that domestic investors with diverse holdings in the jurisdiction of their contract are better able to maintain bargaining relationships with host governments during periods of economic or political turbulence and when state leverage increases. Such investors are better able to keep bargaining relationships alive and continue investing during such periods than international investors or domestic firms with more scattered portfolios, because they have greater patience and leverage and can fall back on a broader set of possible exchanges.

This dissertation evaluates this argument by examining evidence at two levels of analysis. The first empirical section focuses on the divergent trajectories of fourteen water and sanitation privatizations in Argentina. In these cases, the correlation between diverse local holdings and the long-run viability of private sector management is striking, and process-tracing finds that the mechanisms highlighted by the theory are indeed at work. The second empirical section examines patterns of market entry and exit from the water and sanitation sector in low and middle income countries. Cox proportional hazard regressions find strong associations between domestic investor control over a project, the best available proxy for diversified holdings, and lower rates of premature contract cancellation. The project also finds that international investors have seldom entered new contracts after the Argentine and Asian crises. Moreover, even when not canceling their contracts, multinationals are often attempting to sell shares in existing projects.