To reduce costs, states (principals) frequently turn to allies, coalition partners, client states and non-state armed groups (agents) to help defend territory, deter threats, compel policy concessions and fight wars. Because interests are rarely in perfect alignment, a principal must incentivize its agents to advance the principal's objective. I develop two arguments about how states approach this agency problem. First, the use of conditional inducements (carrots and sticks) is less common than traditionally believed because typical features of international politics generate severe moral hazard. Second, principals frequently turn to three alternative strategies that shift incentives by manipulating the structure in which power politics play out. My empirical strategy leverages policymaker interviews, process-tracing of prominent historical cases and fixed-effect analyses of panel data on post-World War II interstate arms transfers to build the case that these strategies of manipulation are important solutions to the agency problem in delegation.
Part I of the dissertation develops the theory. Chapter One argues that the concept of delegating power politics describes the structure of a wide range of political relationships, including formal alliances, state-sponsored terrorism and proxy wars. While states have always turned to delegating partners to advance their interests, I show that because of fundamental changes in international relations -- the nuclear revolution and the rise of transnational threats from non-state armed groups -- delegation is now a central component of statecraft.
Chapter Two describes the strategies of conditional inducement and capacity building that principals are traditionally thought to use to incentivize agents. I argue that typical features of international politics generate two sources of moral hazard -- recognition and sanctioning problems -- that inhibit the efficiency of these strategies. I show that recognition problems are particularly common because most of the action of power politics is in implicit threats to use force rather than its actual use. This makes agent behavior hard to observe. Though outcome-based contracts can still be used in this context, I argue that the complex and noisy relationship between agent behavior and political outcome makes it difficult for principals and agents to locate mutually-beneficial agreements. Furthermore, I identify a novel source of moral hazard that arises when principals are unable to sanction shirking because of domestic or bureaucratic actors with a vested interest in maintaining a status quo relationship with the agent. This problem is common in the context of delegation because capacity-building transfers to agents can have spillover effects that cause actors within the bureaucracy or polity to benefit from the continued flow of arms. That delegation remains ubiquitous despite these challenges poses an important puzzle in international relations.
Chapter Three introduces the three strategies of manipulation that I argue states also use to solve agency problems. These strategies work by reducing the payoffs to agents for employing power against their adversaries, thereby creating relative incentives to instead employ power against the principal's adversary. Equilibrating, the strategy I develop in my job market paper, manipulates the balance of power between the agent and its adversaries (e.g. the U.S. policy of maintaining Israel's qualitative military edge ensures that America's Arab partners use military resources to deter Iran and combat terrorism rather than in competition with Israel). Threat manipulation changes the political relationship between the agent and its adversaries (e.g. U.S. policymakers pushed for European integration, in part, to mitigate conflict between France and Germany and focus European militaries on the threat from the Soviet Union). Resource manipulation alters the agent's military capabilities and thus its relative ability to coerce its adversaries (e.g. the U.S. provides counter-insurgency training and light weapons to incentivize Pakistan to fight the Taliban rather than fight India).
Part II provides empirical evidence of these strategies using policymaker interviews, process-tracing of prominent historical cases and panel data on post-World War II interstate arms transfers. Chapter Four shows that principals provide capacity-building arms transfers to agents. In this chapter I introduce the data, including the computations of dyadic and triadic political relationships that serve as key independent variables. Using a fixed-effects model with principal-year, agent-year and dyad fixed effects to control for a range of unobserved confounding factors, I find that when agents are geographically contiguous to an adversary of the principal arms transfers increase by 40-115%. This result holds across different models and operationalizations of the key variables.
The final three chapters offer evidence that equilibrating, threat manipulation and resource manipulation motivate policymakers in important cases of delegation and that these strategies can help explain patterns of interstate arms transfers. Chapter Five focuses on equilibrating, finding that the effect on arms transfers of agent contiguity to the principal's adversary (described in the previous chapter) is conditional on whether the agent has other adversaries (i.e. an agency problem) and whether the principal has incentives to implement an equilibrating strategy to resolve it (proxied as the agent's adversaries being friends of the principal). When principals can credibly and efficiently offer equilibrating defense commitments and arms transfers to the adversaries of the agent, the effect of contiguity on arms transfers is statistically the same as when agents have no other adversaries: transfers increase 96-255%. However, when principals are unable to implement strategic manipulation to resolve the agency problem, contiguity to an adversary of the principal has no effect on arms transfers. A historical case study based on original archival research of U.S. decision-making around the qualitative military edge policy and military aid to Israel is used to validate the quantitative results.
Chapter Six on threat manipulation develops a case study of American strategy in Europe at the onset of the Cold War. I argue that the U.S. confronted an agency problem with its European allies: the U.S. needed strong partners to help deter the Soviet Union but worried (or at least took seriously French fears) that German rearmament would lead to renewed conflict within Europe. I provide evidence that American policymakers pushed for European integration to help solve this agency problem. By mitigating threats between its allies through integration, the U.S. was able to convince them to accept the rearmament of Germany and focus their own military efforts on the Soviet threat.
Chapter Seven on resource manipulation uses the arms transfers data to assess the hypothesis that principals solve agency problems by transferring military capability that is tailored for employment against the principal's adversary. Specifically, I test whether the effect on arms transfers of agent contiguity to an adversary of the principal is conditional on whether the agent's adversaries and principal's adversaries pose distinct military threats (e.g. state vs. non-state, regional power vs. small states, landlocked vs. maritime). When adversaries are qualitatively different, the agency problem should be resolvable through resource manipulation and thus have no mitigating effect on arms transfers; when they are not different, on the other hand, the agency problem should reduce the effect of agent contiguity on arms transfers. Quantitative analysis is supported by comparative case studies of U.S. arms transfers and military cooperation with states around the South China Sea.