As part of an exploration of ways to predict what determines the targets of suicide attacks, RAND conducted a proof-of-principle analysis of whether adding sociocultural, political, economic, and demographic factors would enhance the predictive ability of a methodology that focused on geospatial features. This test case focused on terrorist bombing incidents in Israel, but the findings indicate that the methodology merits further exploration.

Predicting Suicide Attacks

Integrating Spatial, Temporal, and Social Features of Terrorist Attack Targets

Walter L. Perry, Claude Berrebi, Ryan Andrew Brown, John S. Hollywood, Amber Jaycocks, Parisa Roshan, Thomas Sullivan, Lisa Kraus

ResearchPublished Feb 18, 2013

Cover: Predicting Suicide Attacks
Download PDF
Additional Downloads

Includes Summary; other ebook formats

The Naval Research Laboratory (NRL) set out to develop ways to predict what determines the targets of suicide attacks. While the ultimate goal is to create a list of areas at risk for the U.S. environment, the first phase of development employed a data set from Israel. Initially, NRL focused on spatial attributes, creating its own risk index, but realized that this focus on the where ignored the broader social context, the why. The lab asked RAND to test, as a proof of principle, the ability of sociocultural, political, economic, and demographic factors to enhance the predictive ability of NRL's methodology. Again using Israel as a sample, RAND created a database that coded for these factors, then conducted both quantitative and qualitative analyses with an eye to determining what puts a given area at greater risk. The quantitative analysis established that these factors are related to the odds of attack within specific neighborhoods and that the relationships held even when controlling for geospatial factors, so they seem to confer risk for reasons beyond their association with geospatial features of neighborhoods. The specifics of the research are limited to the preferences of Palestinian suicide bombers in Israel; however, the methods used to assess target preferences in Israel could be transferred to the United States or other countries. Any results, if proven to be robust, could be used to develop recommendations for heightened public awareness in certain areas.

Key Findings

Examining Sociocultural, Political, Economic, and Demographic Factors Did Enhance the Geospatial Methodology for the Proof-of-Principle Data Set.

  • Socioeconomic, demographic, and political data have statistically significant effects on the odds of attack within specific neighborhoods.
  • These data also explain the variance in in the risk of attack over and above geospatial predictors.
  • Demographics were related to greater risk of attack (in this case, heavily Jewish neighborhoods with immigrants from Asia or Africa).
  • Voting patterns are also good predictors of increased risk of attack. In this study, neighborhoods that voted heavily for right-wing or Orthodox parties in 1999) were at greater risk.
  • The relationship between socioeconomic, demographic, and political variables and attack probability holds even when controlling for geospatial factors.
  • Suicide bombers focused both on accessible crowds (shopping centers, etc.) and well-known or iconic sites (synagogues, mosques, etc.)

Recommendations

  • This study was a proof-of-principle effort: short term and exploratory. A more-comprehensive study including the use of better-performing classification and supervised learning methods should be conducted.
  • All the regression analyses were cross-sectional. Because sociocultural, geospatial, and precipitant event determinants of suicide bomb attack sites change over time, longitudinal or pattern regression analyses should be conducted.
  • The time windows for sociocultural precipitants were specified. Future analytic efforts should allow a more flexible time window and should also consider additional precipitants.
  • The methods used to assess target preferences in Israel should be transferred to the United States and other countries — especially the qualitative data analysis methods.

Topics

Document Details

  • Copyright: RAND Corporation
  • Availability: Web Only
  • Year: 2013
  • Pages: 112
  • Document Number: MG-1246-NRL

Citation

Chicago Manual of Style

Perry, Walter L., Claude Berrebi, Ryan Andrew Brown, John S. Hollywood, Amber Jaycocks, Parisa Roshan, Thomas Sullivan, and Lisa Kraus, Predicting Suicide Attacks: Integrating Spatial, Temporal, and Social Features of Terrorist Attack Targets. Santa Monica, CA: RAND Corporation, 2013. https://www.rand.org/pubs/monographs/MG1246.html.
BibTeX RIS

Research conducted by

This research was sponsored by the Naval Research Laboratory and conducted within the RAND Homeland Security and Defense Center, a joint center of RAND Justice, Infrastructure, and Environment, and the RAND National Defense Research Institute, a federally funded research and development center sponsored by the Office of the Secretary of Defense, the Joint Staff, the Unified Combatant Commands, the Navy, the Marine Corps, the defense agencies, and the defense Intelligence Community.

This publication is part of the RAND monograph series. RAND monographs were products of RAND from 2003 to 2011 that presented major research findings that addressed the challenges facing the public and private sectors. All RAND monographs were subjected to rigorous peer review to ensure high standards for research quality and objectivity.

This document and trademark(s) contained herein are protected by law. This representation of RAND intellectual property is provided for noncommercial use only. Unauthorized posting of this publication online is prohibited; linking directly to this product page is encouraged. Permission is required from RAND to reproduce, or reuse in another form, any of its research documents for commercial purposes. For information on reprint and reuse permissions, please visit www.rand.org/pubs/permissions.

RAND is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.