Can Development Programs Effectively Counter Insurgencies?

The research examines whether development programs can counter insurgencies, using a randomized controlled trial of Afghanistan’s National Solidarity Program (NSP) from 2007-2011. The study analyzes security incidents from a U.S. military database (2,890 events) and three rounds of household surveys (over 13,000 respondents). Findings show NSP reduced violence in central, northern, and western Afghanistan but increased attacks near the Pakistani border, where insurgents relied less on local support. Economic benefits improved government perceptions in most areas. Effect sizes indicate a significant reduction in violence (4-13 km from villages) in non-border regions, but a rise in attacks near insurgent strongholds.

Full Citation and Link to Article

Beath, Andrew, Fotini Christia, and Ruben Enikolopov. Can Development Programs Counter Insurgencies? Evidence from a Field Experiment in Afghanistan. American Economic Journal: Applied Economics (Forthcoming). https://www.aeaweb.org/articles?id=10.1257/app.20230150&from=f

Extended Summary

Central Research Question

The study investigates whether development programs can effectively counter insurgencies by reducing violence and increasing support for the government. Specifically, it examines Afghanistan’s National Solidarity Program (NSP), the country’s largest development initiative, to assess its impact on insurgent activity, economic outcomes, and government legitimacy. The key question is whether providing public goods and infrastructure in conflict-affected areas can reduce violence by increasing the opportunity costs of insurgent participation and fostering greater public support for the state. The research also explores whether these effects vary based on local insurgency dynamics, particularly in regions near the Pakistan border.

Previous Literature

The role of development programs in countering insurgencies is widely debated. Some scholars argue that improving economic conditions can reduce violence by increasing opportunity costs for insurgents and fostering goodwill toward the government. This perspective is supported by studies in Iraq (Berman, Shapiro, and Felter, 2011) and the Philippines (Crost, Felter, and Johnston, 2016), which found that aid programs reduced insurgent attacks. However, other research suggests that development assistance may provoke violence by making government-affiliated projects and individuals targets for insurgents. For example, Crost, Felter, and Johnston (2014) found that a community development program in the Philippines increased attacks, as insurgents sought to disrupt government outreach. Studies in India (Khanna and Zimmerman, 2017), Colombia (Weintraub, 2016), and Africa (Nunn and Qian, 2014) similarly found that aid could exacerbate conflict, particularly in areas where insurgents operate independently of local populations. Given these mixed results, the study aims to provide clearer evidence by leveraging a randomized field experiment in Afghanistan.

Data

The study relies on two primary sources of data: security incidents recorded by the U.S. military and survey responses from Afghan villagers.

1. Security Incidents: The authors use the Combined Information Data Network Exchange (CIDNE), a declassified U.S. military database, which records the date, location, and type of security incidents, including firefights, IED explosions, assassinations, and sabotage. The dataset includes 2,890 incidents occurring within a 15-km radius of 500 Afghan villages from 2006 to 2014.

2. Household Surveys: The researchers conducted three rounds of household surveys (baseline in 2007, midline in 2009, and endline in 2011) across 500 villages. These surveys collected data from over 13,000 respondents on economic conditions, perceptions of security, and attitudes toward the government, NGOs, and international military forces.

The combination of military event data and civilian surveys allows for a nuanced analysis of both the objective and perceived effects of development programs on insurgencies.

Methods

The study exploits a randomized controlled trial (RCT) in which 250 of 500 villages were randomly assigned to receive NSP starting in 2007, while the remaining 250 villages served as a control group until 2011. The program provided each treatment village with a block grant of up to $60,000 to finance projects such as road construction, irrigation, electricity, and drinking water supply. Projects were managed by newly formed Community Development Councils (CDCs)elected through secret ballot, introducing an element of participatory governance.

The researchers employ ordinary least squares (OLS) regression to estimate the impact of NSP on violence, economic outcomes, and political attitudes. The primary specification includes:

• Treatment Assignment: A dummy variable indicating whether a village received NSP.

• Time Periods: Separate estimates for midline (2009), endline (2011), and post-program (2014).

• District Fixed Effects: Controls for regional differences in conflict intensity.

• Clustering Adjustments: Accounts for geographic spillovers between villages.

To explore heterogeneous effects, the authors examine whether NSP’s impact varies by proximity to the Pakistani border, where insurgents rely less on local support and more on cross-border sanctuaries.

Findings/Size Effects

The study finds that the impact of NSP on violence and public perceptions varies by region.

1. Overall Effects on Violence

   • Across all 500 villages, NSP had no statistically significant effect on the total number of security incidents.

   • However, in central, northern, and western Afghanistan, NSP significantly reduced insurgent attacks by 4-13 km from villages at both midline and endline.

   • In contrast, in districts near the Pakistani border, violence increased, suggesting that insurgents targeted NSP-affiliated villages to deter collaboration with the government.

2. Economic Outcomes

   • NSP improved access to public goods, particularly electricity (+19%), irrigation (+28%), and road infrastructure (+22%).

   • Household income increased, but employment effects were mixed, with some shifts away from agriculture.

   • Perceptions of economic well-being improved, especially among women.

3. Attitudes Toward Government

   • In non-border regions, NSP increased trust in the Afghan government and allied NGOs at both midline and endline.

   • In border regions, there was no significant improvement in political attitudes, suggesting that insurgents’ intimidation tactics may have negated NSP’s positive effects.

4. Security Perceptions

   • In non-border regions, villagers felt safer and were more likely to believe the security situation had improved over the previous two years.

   • In border regions, NSP had no effect on security perceptions, aligning with the increase in insurgent attacks.

5. Effect Heterogeneity and Mechanisms

   • The reduction in violence in non-border areas appears driven by declines in active insurgent participation rather than increases in intelligence-sharing with security forces.

   • In border districts, NSP-funded projects became insurgent targets, leading to more attacks within 2 km of villages.

   • The presence of foreign fighters and cross-border sanctuaries in border areas likely explains the differing effects, as these insurgents were less reliant on local civilian support and more willing to use coercion.

     
     

Conclusion

The findings suggest that development programs can reduce insurgent violence, but only under certain conditions. In areas where insurgents depend on local communities for support, NSP successfully reduced violence, improved economic conditions, and increased government legitimacy. However, in border regions where insurgents had external funding and sanctuaries, NSP backfired, leading to more attacks against development projects and civilians.

These results highlight the importance of local insurgency structures in shaping the effectiveness of aid programs. Policymakers designing counter-insurgency strategies should consider:

• Targeting development aid in areas where insurgents rely on local support rather than in regions where they operate independently.

• Providing security for aid projects in high-risk regions, particularly near international borders.

• Combining development initiatives with broader governance reforms to maximize public trust in state institutions.

Overall, the study provides strong empirical evidence that development programs can counter insurgencies—but only in contexts where local insurgents are responsive to shifts in civilian sentiment. In areas where insurgents are externally supported and coercive, development programs alone may be ineffective or even counterproductive.