Apparently, not a damn thing:
Social and behavioral research on terrorism has expanded dramatically. However, theoretical work that incorporates terrorism and collection of valid data on it has lagged behind theoretical work on other criminological subjects. Theorizing has been dominated by deterrence perspectives. Threats of severe consequence for terrorist acts in general show little promise, but there is evidence that increasing the certainty of consequences works in some situations. Research on terrorism will be improved if it moves beyond deterrence to include concepts drawn from legitimacy, strain, and situational perspectives. Limitations of traditional criminology data sources for studying terrorism have encouraged the development of open‐source‐event databases. The most comprehensive, created by combining the Global Terrorism Database with RAND‐MIPT data, documents more than 77,000 terrorist incidents from 1970 to 2006. Attacks peaked in the early 1990s and then declined substantially until 9/11. They have since substantially increased. The regional concentration of terrorism has moved from Western Europe in the 1970s, to Latin America in the 1980s, to the Middle East and Persian Gulf in the twenty‐first century. Despite the enormous resources devoted to countering terrorism, surprisingly little empirical information is available on which strategies are most effective.
Source: “Research on Terrorism and Countering Terrorism” from the journal “Crime and Justice”
Does terrorism work?