Elite Preferences for Allocating Public Good to Former Insurgents

• In Event: Strategies for Improving Data Quality and Micro-qualitative Research

Abstract

Under what conditions would elites distribute public goods to neighborhoods where the residents have collaborated with an insurgent group? How does variation in the attributes of a neighborhood (e.g. class, religion, partisanship or co-ethnicity) and collaboration status affect the prospects for disbursement of public good? Through a survey of 1,150 politicians conducted in Iraq, we assessed local politicians' preferences for allocating public good using a conjoint experiment design (Hainmueller et al., 2013). The experiment randomly varied the identity (partisanship, income class, religion and ethnicity) of a hypothetical neighborhood, in a local politician's province/district/sub-district, and the collaboration status (low-levels v. high-levels of collaboration.

We find that politicians are less likely to distribute public good to neighborhoods with high levels of collaboration. We also find that the underlying mechanism behind this is the perception that allocating public good to such neighborhoods would negatively impact the reputation of the politician.

The results of these experiments fill a critical gap in the existing literature on reintegration and social cohesion in post-conflict settings. Previous research has used survey data to evaluate the conditions under which civilians forgive civilians who have collaborated (Kao and Revkin 2021) but does not explore the determinants of social trust between elites and civilians. Other work has used evidence from lab-in-the-field experiments to measure the effects of civil war-related violence on pro-social behaviors including altruistic giving, public goods contributions, and trust-based transactions (Gilligan et al., 2014), but does not examine the effects of civilian collaboration with the insurgency on these outcomes. Other studies use data on ex-combatants to identify individual-level predictors of successful demobilization and reintegration (Humphreys and Weinstein, 2007; Annan et al., 2011), but do not systematically examine elite attitudes toward former combatants.

Author

• Narmin Butt, University of California, San Diego (UCSD)