Preventing mass killings : optimal strategies for protecting public targets against terrorist attacks

Abstract

Practitioners with limited security resources lack appropriate guidelines when protecting targets against mass-casualty attacks. Existing guidelines about prioritization between targets and protective security measures are either very abstract or consist of roughly collected advice. Combining game theory with practically oriented literature, such as situational crime prevention, crime scripts and crime prevention through environmental design, this dissertation establishes a systematic framework for prioritizing between targets and measures and provides concrete policy recommendations (given certain assumptions about motivation). I argue that:<br><br>

1. If terrorists cannot be deterred from attacking, strategic authorities will ensure that the terrorists attack well-protected targets. Protection is desirable not only when it deters the terrorists from attacking, but also when it causes the terrorists to target sites that are less rather than more damaging for the authorities.<br> 2. When protecting against mass-casualty attacks, the authorities should give priority to potential targets with a high expected number of casualties, many foreigners, low employee density, many hiding places, many access points, high anonymity, high share of earlier attacks, and high system fragility.<br> 3. When protecting against explosive attacks on railway networks, the best protective security measures focus on limiting the damage caused by an explosive attack, rather than on reducing the probability of an attack’s being successful.<br> 4. By thinking counter-terrorism when designing railway carriages, we may significantly reduce the expected damage caused by explosive attacks on railway.<br><br>

Many of this dissertation’s models generate interesting empirically testable implications. Unfortunately, lack of appropriate data prevents proper testing of these empirical implications as well as testing of assumptions underlying the models; available datasets do not distinguish between attacks where the terrorists seek mass-killings and attacks where they do not. My policy recommendations are, furthermore, less concrete because of the very generic depiction of the terrorists in my models. To refine these recommendations, more knowledge is needed about what resources and capabilities terrorists possess.