Constructing Legitimacy: The Political and Security Effects of Reconstruction Spending in Iraq

Abstract

Political theorists and policymakers have sought to understand how states establish legitimacy for thousands of years, with the question coming to the fore of the policy agenda in the United States in the aftermath of September 11, 2001 attacks. With the continued threat posed by fragile states to national and international security, combating state weakness remains a policy priority in the United States today, with foreign aid and economic assistance serving as one of the main tools through which this goal is pursued. How does foreign reconstruction and development assistance influence the establishment of legitimacy in fragile states? By operationalizing the theoretical notion of state legitimacy through the two proxy measures of voter turnout and incidents of subnational violence, this thesis seeks to evaluate the efficacy of the United States’ reconstruction assistance in establishing legitimacy in Iraq in the wake of the 2003 invasion. In addition to conducting a descriptive analysis of the United States’ reconstruction effort in Iraq, with a particular focus on the geographic and temporal distribution of reconstruction spending in the country, it asks what the effects of reconstruction spending were on voter turnout in four Iraqi parliamentary elections and on incidents of subnational violence over the course of a decade. It additionally examines whether differences in project type and implementing actors influenced turnout and violence differently. To answer these questions, this thesis reports the results of a series of Ordinary Least Squares regression models; contrary to the hypothesized positive impact of reconstruction assistance on proxies of legitimacy, I uncover a more complicated picture.

First, I find a statistically significant negative correlation between reconstruction spending and voter turnout, in an effect that remains significant upon the addition of fixed effects and controls for security, development, and income levels, and persists in the long-term. A similarly significant negative relationship with turnout exists when spending is disaggregated into economic, security, and military-implemented projects, but dissipates in the case of spending on quality of life projects and projects implemented by civilian agencies. Second, I find a statistically significant positive correlation between reconstruction spending and incidents of subnational violence in Iraq, in an effect that also persists despite the addition of fixed effects and controls. This correlation also persists in most spending categories, with the exception of spending on security. Yet, unlike the case with turnout, it dissipates in the aftermath of reconstruction spending, and eventually morphs into a statistically significant negative correlation between spending and violence, such that Iraqi governorates with increased spending were likely to experience lower levels of violence three to four years after its conclusion. I argue that these findings are the product of a distinction between the political and security dimensions of legitimacy, such that while it is possible to bolster the latter through technical capacity-building measures, the state-citizen relationship central to political legitimacy can only be established if policymakers ensure the recipient state is visibly involved in the reconstruction or development effort.