Examining the counterinsurgency-counterterrorism tradeoff: the effects of targeted strikes on militant attacks

ABSTRACT

Do targeted strikes against militant groups affect insurgent operations differently from terrorist operations? One of the reasons it is challenging for states to counter militant groups who simultaneously wage an insurgency and a terrorist campaign is that there are offsetting effects between counterinsurgency and counterterrorism. As both an insurgent group and a terrorist group, the Islamic State of Iraq and the Levant (ISIL) poses particular challenges to states. Since 2014, the United States and coalition allies have conducted thousands of targeted strikes against ISIL in Iraq and Syria as part of Operation Inherent Resolve (OIR). In this article, we examine the effects of targeted strikes against ISIL. Using a novel dataset of time-series observations from Iraq and Syria between 2015 and 2018, we examine the relationship between targeted strikes conducted under OIR and the number of insurgent and terrorist attacks conducted by ISIL. Our findings give empirical support to the theoretical argument that the same military strategy can have different effects on insurgency and terrorism.

KEYWORDS:

• Insurgent groups
• terrorist groups
• counterinsurgency
• counterterrorism
• Islamic State
• targeted strikes

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. The following information about the coding of attacks is based on a conversation between one of the authors and the lead researcher of the GTD Perpetrators Team at START.

2. These also include cases where the source materials do not indicate that ISIL is the aggressor, such as sources that report generic clashes between ISIL and military forces.

3. As a robustness check, we also estimated all models using Poisson regression, which is also used for count models for data generated with specific assumptions. All results hold.

4. Some studies will also report the Akaike Information Criterion (AIC) results as well. The AIC for both models returned an appropriate lag length of 16, but the numeric value of this criterion was higher than the reported BIC. Therefore, we utilize the lag length indicated by the BIC.

5. This becomes 1,206 in the statistical analysis once the lags are accounted for.

6. As defined by the U.S. Central Intelligence Agency.

Additional information

Notes on contributors

Brittnee Carter

Dr. Brittnee Carter is an Associate Professor in the Department of Political Science at the University of Kansas.

Thomas Guarrieri

Dr. Thomas Guarrieri is an Associate Research Scientist at the National Consortium for the Study of Terrorism and Responses to Terrorism and the Director of Undergraduate Studies for the Global Terrorism Studies Minor at the University of Maryland.

Daniel S. Smith

Daniel S. Smith is a Ph.D. candidate in the Department of Political Science at The Ohio State University.