Explaining Conditionality Failure in Principal-Agent Relationships

• In Event: Proxy War

Abstract

Recent literature on counterinsurgency and proxy war models states as principals and agents. The principal, usually a stronger state, arms and equips the agent, usually a weaker state, to confront a shared threat. Because principals and agents’ interests are never fully aligned, principals should structure incentives such that the agent prefers to comply with the principal’s directives. And yet, one of the puzzles emerging from this literature is the empirical observation that principals often fail to use conditional rewards and punishments with their agents. In the years following 11 September, for example, US military assistance to Pakistan flowed despite evidence that Pakistan was actively abetting the Taliban networks that the United States sought to eradicate.

Two competing arguments in the literature attempt to explain this puzzle. One argument is that policymakers have bounded rationality and fail to understand how conditional rewards and punishments are more useful than unconditional support. The second argument is that principals tolerate noncompliance because they have a commitment problem: a principal cannot pay an agent enough to convince it to solve the principal’s problem when doing so creates a new and ongoing problem that the agent will face alone. No amount of US transfers to Pakistan, in this view, could convince it to confront a Taliban threat that it would face alone after US withdrawal from Afghanistan.

We argue that both of these arguments leave important cases unexplained. Although policymakers, like all humans, face inherent bounds to rationality, they are rarely unaware of the problem of divergent interests. Nor do commitment problems seem to explain all cases. The US relationship with Taiwan, for example, does not demonstrate a commitment problem for the principal: solving the United States’ problem—ensuring that China cannot easily conquer Taiwan—also solves Taiwan’s problem. Yet Taiwan invests in military equipment that is not optimal for its defense, and the United States is unwilling to use carrots or sticks to encourage other choices.

In this paper, we explore two potential explanations for the suboptimal use of incentives in principal-agent relationships. These explanations center the undertheorized sources and forms of proxy agency. First, we explore the domestic politics of the proxy. Agents may be unable or unwilling to comply due to domestic political pressures, or may use these pressures as bargaining leverage against the principal, following the logic of a two-level game. Second, we explore limits on how much the principal is worth to the agent. Access to transfers from other sources—as under conditions of competition—can reduce the value of transfers from the principal and impose limits to how much the principal can compensate the agent. Together, these explanations demonstrate that even weak “agents” have levers to push back against the principal and evade imposition of conditions—or undermine the effectiveness of conditions when imposed. We develop this argument using a formal principal-agent model incorporating domestic political constraints on the agent and constraints on the principal’s compensation to the agent. We then illustrate empirical implications from the model using a case study of US security assistance to Taiwan.

Authors

• Renanah M. Joyce, Georgetown University

• Alexandra Chinchilla, Texas A&M, Bush School of Govt and Public Service