Deterring Terrorism: Lessons Learned in Social Science Research

Laura Dugan University of Maryland Department of Criminology and Criminal Justice & The National Consortium for the Study of Terrorism and Response to Terrorism (START)

Some of this research was supported by the U.S. Department of Homeland Security (DHS) Science and Technology Directorate’s Office of University Programs through START. Any opinions, findings, conclusions or recommendations presented here are solely the authors’ and are not representative of DHS or the United States Government. Conclusions

Terrorists differ from other lawbreakers

 They have a larger mission that goes beyond immediate need.  They innovate.  They rely heavily on their constituency, which may be an important intervention point. Deterrence: Rooted in Rational Choice Theory

E(uterror) = p U(y-F) + (1-p) U(y) where p = perceived probability of punishment y = anticipated benefits of perpetration; and F = perceived penalty of the act

Lesson: Raise the costs of perpetration through increased certainty and severity so that the utility of perpetration falls below the benefit of the act. In other words: DETERRENCE Testing a Rational Choice Model of Airline Hijacking

Dugan, Laura, Gary LaFree, and Alex Piquero, Criminology 2005 Policies Suggesting Certainty and Severity Certainty Severity  January 1972:  October 1970: FAA orders tighter Hijacking is a screening crime in Cuba (affects US cases) (affects Cuba events)  February 1973:  February 1973: Metal detectors Cuba-US and law agreement enforcement (affects Cuba events) (affects US cases) Figure 2. US and Non-US Successful Hijackings, 1946 to 1985

40 Non-US US 35

30

25

20

15

10

5

0

1947 1949 1951 1953 1955 1957 1959 1961 1963 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 Series Hazard Models (estimating hazard of hijacking attempt)

For: Time until Event next event 1. Total Characteristic Hijackings 2. US Origin 3. Non-US Origin 4. Cuba Diverted 5. Terrorism Policy 6. Non-terrorism Hazard Ratio of Hijacking Attempt for each Policy by Type

Cuba Policy Tighter Screening 1973 Policies

1.05

0.85

0.65

0.45

0.25

0.05 Total US Origin Non-US Cuban Terrorist Non- Origin Diverted Terrorist Conclusions on Deterrence

 A new hijacking attempt is less likely when the certainty of apprehension is increased.  Compared to those who hijack for other reasons, hijacking attempts by terrorists will be less affected by counter terrorism measures that raise the severity or certainty of punishment.

The Impact of British Counterterrorism Strategies on Political Violence in Northern :

Comparing Deterrence and Backlash Models

LaFree, Gary, Laura Dugan and Raven Korte, Criminology 2009

British Government Actions in

Terrorist Attacks by Republicans

250

200

150

100 Frequency

50

0

1969 1970 1971 1972 1973Operation1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 Criminalization Falls Motorman Loughall Incident Gibraltar Internment Incident Results of Series Hazard Model

2 1.872

1.8

1.598 1.6

1.417 1.4

1.2

1.000 1 Internment Criminalization Falls Curfew Operation Gibralter

0.8 Motorman

0.6

0.543 0.4 Conclusion on Deterrence

 Backlash more common than deterrence.  (massive military deployment) seemed to have a deterrent effect.  Governments should be cognizant of efforts that could sabotage perceived legitimacy. Moving Beyond Deterrence: The Effectiveness of Raising the Expected Utility of Abstaining from Terrorism in Israel (ASR, 2012)

Dugan, Laura and Erica Chenoweth, American Sociological Review 2012 An Underutilized Component of Rational Choice: Raising the Benefits of Abstaining from Terrorism

E(unonterror) = q U(x+G) + (1-q) U(x) where q = perceived probability of rewards from abstention x = value of current situation; and G = anticipated rewards of abstention

Dimensions of Countering Terrorism TARGET

Discriminate

Repressive Conciliatory

ACTION ACTOR TYPE Justice Politician

Police Military Indiscriminate Material M/NM Nonmaterial Targets of Punishment and Rewards in Israel

Punishment Rewards Repressive Actions Conciliatory Actions Discriminate repression Discriminate conciliation directed toward the guilty directed toward the guilty Specific (direct deterrence) (direct benefits of abstention)

Indiscriminate repression Indiscriminate conciliation directed toward the directed toward the General Palestinians in general Palestinians in general (indirect deterrence) (indirect benefits of abstention) Tactical Regimes of the Israeli- Palestinian Conflict The First Intifada (1987-1993)  Started as nonviolent  Dominated by secular nationalists  Hamas became active near the end The Oslo Lull (1993-2000)  Negotiators established Palestinian Authority  Palestinians recognized 1967 borders  Neither side held to agreement The Second Intifada (2000-2004)  Violent from the beginning  Dominated by religious groups  Known for deadly suicide attacks Figure 1. Quarterly Repressive and Conciliatory Actions by Israel and Palestinian Terrorist Attacks 200 80 Conciliatory Actions Second Intifada 180 Repressive Actions 70

160 Palestinian Terrorist Attacks 60

140 First Intifada Oslo Lull

50 120

100 40

80 30

60 Attacks Palestian of Number Number of Israeli Actions Israeli of Number 20 40

10 20

0 0

Quarter, Beginning in Third Quarter 1987 and Ending in the Fourth Quarter 2004 Examining the Relationship Between Actions and Terrorism

Attacks Against Israelis (Current Month)

We first test the relationship parametrically (Negative Binomial) and then examine it non-parametrically (GAM). Together and separately for each of the tactical regimes Results for All Actions for Entire Period

GAM 3 df smooth for allla

1.61374

NS

-.360915 1 121 Lagged all actions Conciliatory and Repressive Actions—Entire Period

a. Conciliatory Actions b. Repressive Actions

GAM 3 df smooth for concla GAM 3 df smooth for reprla

.142908 1.57031

NS 0/- quadratic

-1.04907 -.171465 0 27 1 80 Lagged Conciliatory acts Lagged Repressive Acts a. Conciliatory-Discriminate b. Conciliatory-Indiscriminate

GAM 3 df smooth for cdisla GAM 3 df smooth for cindla

.376635 .165583

0/- NS quadratic

-.691987 -1.48244 0 8 0 26 Lagged Conciliatory Discriminate Lagged Conciliatory Indiscrimina

GAM 3 df smooth for rdisla GAM 3 df smooth for rindla

1.21061 .6765 NS NS

-.356292 -.212605 0 32 0 48 Lagged Repressive Discriminate Lagged Repressive Indiscriminate c. Repressive-Discriminate d. Repressive-Indiscriminate Results by Tactical Regime Conciliatory Actions a. All Months b. First Intifada

GAM 3 df smooth for concla GAM 3 df smooth for concla

.142908 .752482

0/- quadratic +/- quadratic

-1.04907 -1.00882 0 27 0 13 Lagged Conciliatory acts Lagged Conciliatory acts

GAM 3 df smooth for concla GAM 3 df smooth for concla

.378618 .59527

+/- - quadratic linear

-1.89564 -2.39703 0 27 1 26 Lagged Conciliatory acts Lagged Conciliatory acts c. Oslo Lull d. Second Intifada Repressive Actions a. All Months b. First Intifada

GAM 3 df smooth for reprla GAM 3 df smooth for reprla

1.57031 .591733

NS NS

-.171465 -.378483 1 80 1 31 Lagged Repressive Acts Lagged Repressive Acts

GAM 3 df smooth for reprla GAM 3 df smooth for reprla

1.07254 1.59288 + NS linear

-.768185 -.766279 1 35 7 80 Lagged Repressive Acts Lagged Repressive Acts c. Oslo Lull d. Second Intifada Conciliatory-Discriminate a. All Months b. First Intifada

GAM 3 df smooth for cdisla GAM 3 df smooth for cdisla

.376635 .369726

+/- NS quadratic

-.691987 -.696348 0 8 0 5 Lagged Conciliatory Discriminate Lagged Conciliatory Discriminate

GAM 3 df smooth for cdisla GAM 3 df smooth for cdisla

.631996 .23344 NS

NS

-.680878 -.91429 0 8 0 5 Lagged Conciliatory Discriminate Lagged Conciliatory Discriminate c. Oslo Lull d. Second Intifada Conciliatory-Indiscriminate a. All Months b. First Intifada

GAM 3 df smooth for cindla GAM 3 df smooth for cindla

.165583 .586609

0/- quadratic NS

-1.48244 -.629806 0 26 0 10 Lagged Conciliatory Indiscrimina Lagged Conciliatory Indiscrimina GAM 3 df smooth for cindla GAM 3 df smooth for cindla

.375812 .638318

+/- - quadratic linear

-2.00882 -2.4676 0 24 0 26 Lagged Conciliatory Indiscrimina Lagged Conciliatory Indiscrimina c. Oslo Lull d. Second Intifada Repressive-Discriminate a. All Months b. First Intifada

GAM 3 df smooth for rdisla GAM 3 df smooth for rdisla

1.21061 1.07024

NS NS

-.356292 -.527211 0 32 0 13 Lagged Repressive Discriminate Lagged Repressive Discriminate

GAM 3 df smooth for rdisla GAM 3 df smooth for rdisla

.318519 1.3249 NS

+/- quadratic

-2.45229 -.504824 0 10 1 32 Lagged Repressive Discriminate Lagged Repressive Discriminate c. Oslo Lull d. Second Intifada Repressive-Indiscriminate a. All Months b. First Intifada

GAM 3 df smooth for rindla GAM 3 df smooth for rindla

.6765 .840002 NS NS

-.212605 -.801232 0 48 1 25 Lagged Repressive Indiscriminate Lagged Repressive Indiscriminate

GAM 3 df smooth for rindla GAM 3 df smooth for rindla

1.66116 .82789 + NS linear

-.995514 -.541126 0 30 4 48 Lagged Repressive Indiscriminate Lagged Repressive Indiscriminate c. Oslo Lull d. Second Intifada Conclusions on Deterrence  Repression can be harmful  Tactical regime matters  Overall findings are driven by the Second Intifada  Repression only seems to matter during the Oslo Lull (i.e., time of peace)  Discriminate-Conciliation during the First Intifada seems to lead to more attacks  Indiscriminate actions matter more  Especially during the Second Intifada  Conciliation should be a serious policy alternative  Especially when directed toward terrorists’ constituency  Conciliation should be sustained (0/- quadratics) Relevance of These Findings to the Efforts of ADSA

Terrorists differ from other lawbreakers

 They have a larger mission that goes beyond immediate need.  They innovate.  They rely heavily on their constituency, which may be an important intervention point.