Abstract
The hostage-taking behavior by terrorist groups is special in that the lives of people taken as hostages can be saved by the government in a short run, whereas in other forms of attacks, there is little scope to negotiate and potential to save people from death or injury while an incidence is in progress. Four types of hostage-taking missions are identified: kidnapping, taking over ground or water transportation, barricade missions, and skyjacking. The chapter explains the logic of a no-concession policy toward hostage takers by using a game tree. But its application in reality poses a painful trade-off between death or harm “hanging” before those already taken as hostages and future incidence of hostage-taking. Results from empirical studies on the determinants of various aspects of hostage-taking like the success or failure of concession policy, logistic success of capturing hostages, negotiation success, and safety of hostages are reviewed.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
GTD divides hostage-taking into three categories: hijacking , barricade incidents , and kidnapping . Hijacking includes skyjacking and takeover of other means of transport.
- 2.
The birth name of Carlos the Jackal was Ilich Ramirez Sanchez.
- 3.
The leader of the two hijackers, Mohammed Ali Hamadei, a Lebanese citizen, was arrested in 1987 at the Frankfurt airport for carrying explosives in his luggage. He was convicted in Germany for murder and handed a life-in-prison sentence. He was however released in 2005.
- 4.
Our model here is a slight deviation from the theoretical setup of Islam and Shahin and yields this outcome in a stronger form.
- 5.
- 6.
This encouraged the terrorists to capture other academics and journalists in Beirut.
- 7.
Otherwise, the payoff vector at node A would have been (0, −K). However, since K does not change, it is immaterial for comparisons.
- 8.
However, it is typically hard to assess the cost or the value of damage \(\bar g_n\) associated with no concession. Although eventually the decision may depend on a single individual (like the president, prime minister, or whosoever is in charge), expert opinions may vary widely on the value of \(\bar g_n\).
- 9.
Technically, the right-hand side of Eq. (12.4) decreases with π. Hence the chances of this inequality being satisfied fall.
- 10.
In Eq. (12.6) if either t f or t n or both are positive so that the left-hand side as a whole is positive, the terrorists would prefer attack to no attack.
- 11.
In the 1970s and 1980s, Libya reportedly financed terror groups that were anti-American and anti-British. It fired at a US aircraft in 1981 over the Gulf of Sidra that was claimed by Qaddafi to a part of Libyan waters. Five Americans were killed simultaneously at Rome and Vienna airport in 1985. These attacks were believed to be aided by Libya.
- 12.
In another study, by also using intervention analysis , Enders et al. (1990b) estimated the impact of metal-detector installation, US strikes against Libya as well as the legislation of the Reagan’s “get-tough” laws on terrorism in 1984 and enhanced security for US embassies and personnel. Along with skyjacking, terror attacks against US diplomats and protected persons, US interests and UK interests were analyzed. The same conclusions were reached with regard to metal detectors and US attack on Libya; in addition, Reagan’s get-tough laws were ineffective and expenditures to secure US embassies had the desired effect, but it also entailed the unintended effect of putting non-US diplomats at a greater risk.
- 13.
Enders and Sandler (1993) analyzed the evidence on cross effects—substitutes and complements—among the modes of attack in response to security initiatives. Using quarterly data from 1968 to 1988 and combining intervention analysis with vector autoregression , the authors found strong evidence of both types of cross effects. For example, the installation of metal detectors in airports reduced skyjackings and diplomatic incidents but increased other types of hostage attacks and assassinations. Embassy fortifications decreased barricade missions but increased assassinations. These are substitution effects. Metal detectors not only damped skyjacking but also threats and hoaxes (because they became less credible). These are complementary effects. This study included fifteen types of terror attacks to estimate the cross effects.
- 14.
During 1978–2000, the USA gave concessions to kidnappers 23.1% of the time, whereas in 2001–2013, its concessions to kidnappers rate declined 10.7% of the time.
- 15.
They are Austria, Belgium, Canada, France, Germany, Italy, the Netherlands, Spain, Sweden, and Switzerland.
- 16.
- 17.
They enlist five stages of hostage events that involve bargaining: initiation of the hostage-taking act, presentation of demands, bargaining, completion of bargaining (e.g., ransom paid or prisoners released, etc.) and cessation of the incident at which time terrorists depart and hostages are freed.
- 18.
The economic rationale for these determinants lies in how they affect the bargaining cost of the two parties involved. For instance, more injuries and fatalities among hostage takers increase their bargaining cost and hence imply shorter duration. The larger the number of countries whose citizens are hostages, the greater is the pressure on the negotiating country, which means a higher bargaining costs; this would imply more concessions granted to the terrorists. Chapter Appendix 12.A describes the data and the estimation method.
- 19.
Logistic success and negotiation success can be seen in the game tree depicted in Fig. 12.2 along the respective arms “Succeeds” and “Capitulates.”
- 20.
Sandler and Scott (1987) had argued the opposite: the more the number of nationalities, the greater will be the funds available to the terrorists and hence the greater is the likelihood of logistic success.
- 21.
There are two reasons. First, the probability of securing some hostages is higher while others may escape. Second, when the number of hostages taken is large, the authorities are likely to be more reluctant to use lethal force, thereby making such targets more vulnerable at the abduction stage.
- 22.
In addition, “Protected persons” like diplomats, military officials, and government personnel are likely to be accompanied by bodyguards and using armor-plated cars and other protections and thus their vulnerability to be captured as hostages are less compared to other people or professionals like journalists, social workers, etc.
- 23.
Logit regression (see General Appendix B, Sect. B.4.2) is used since the outcome variables are binary: either logistic success or failure and either negotiation success or failure.
- 24.
The authors used 764 cases of terrorist hostage-taking and kidnapping data originally collected by the Institute for the Study of Violent Groups (ISVG) at Sam Huston State University, of which 234 cases were finally selected for the analysis. There is no mention of the sample period however. ISVG aims to build a comprehensive terrorism database of terrorist events by collecting and analyzing various open-source material like books, magazines, court documents, etc.
- 25.
But one may argue that the length of captivity is not an exogenous factor; instead, it proxies bargaining situation schemata.
References
Atikinson, S. E., Sandler, T., & Tschirhart, J. (1987). Terrorism in a bargaining framework. Journal of Law and Economics, 30(1), 1–21.
Brandt, P. T., & Sandler, T. (2009). Hostake taking: Understanding terrorism event dynamics. Journal of Policy Modelling, 31(5), 758–778.
Brandt, P. T., Justin, G., & Sandler, T. (2016). Why concessions should not be made to terrorist kidnappers. European Journal of Political Economy, 44, 41–52.
Enders, W., & Sandler, T. (1993). The effectiveness of anti-terrorism policies: A vector-autoregression-intervention analysis. American Political Science Review, 87(4), 829–844.
Enders, W., & Sandler, T. (2012). The political economy of terrorism (2nd edn.). Cambridge University Press.
Enders, W., Sandler, T., & Cauly, J. (1990a). UN conventions, technology, and retaliation in the fight against terrorism: An econometric evaluation. Terrorism and Political Violence, 2(1), 83–105.
Enders, W., Sandler, T., & Cauly, J. (1990b). Assessing the impact of terrorist-thwarting policies: An intervention time series approach. Defence and Peace Economics, 2(1), 1–18.
Gaibulloev, K., & Sandler, T. (2009a). Hostage taking: Determinants of terrorist logistical and negotiation success. Journal of Peace Research, 46(6), 739–756.
Islam, M. Q., & Shahin, W. N. (1989). Economic methodology applied to political hostage-taking in light of the Iran-contra affair. Southern Economic Journal, 55(4), 1019–1024.
Landes, W. M. (1978). An economic study of US aircraft hijackings, 1961–1976. Journal of Law and Economics, 21(1), 1–31.
Lapan, H. E., & Sandler, T. (1988). To bargain or not to bargain: That is the question. American Economic Review: Papers and Proceedings, 78(2), 16–20.
Lee, C.-Y. (2013). Democracy, civil liberties, and hostage-taking terrorism. Journal of Peace Research, 50(2), 235–248.
Poe, S. (1988). Nations’ responses to transnational hostage events: An empirical evaluation. International Interactions, 14(1), 27–40.
Sandler, T., & Scott, J. L. (1987). Terrorist success in hostage-taking incidents: An empirical study. Journal of Conflict Resolution, 31(1), 35–53.
Shahin, W. N., & Islam, M. Q. (1992). Combating political hostage-taking: An alternative approach. Defence Economics, 3(4), 321–327.
Yourish, K. (2015). The Fates of 23 ISIS Hostages in Syria. New York Times, February 10.
Yun, M., & Roth, M. (2008). Terrorist hostage-taking and kidnapping: Using script theory to predict the fate of a hostage. Studies in Conflict & Terrorism, 31(8), 736–748.
Author information
Authors and Affiliations
Appendices
Appendix to Chapter 12
12.A Details of Atikinson et al. (1987)
The authors used ITERATE dataset from 1968 to 1977. There were 387 incidents that involved negotiation. But not all incidents were followed (in the data description) till the end of the event. Finally, 122 incidents were selected for which data on both duration and ransom (if ransom was demanded, and if yes, whether any ransom was paid, and if yes, how much) were available. In forty-two of the 122 incidents, ransom was demanded. Among these, no ransom was paid in twenty cases. The ransom paid was less than or equal to the terrorists’ initial demand with a cluster at 0. This motivated the use of Tobit regression (see General Appendix B, Sect. B.4.2) by using data on the forty-two observations just described.
Time-to-event regression models (see General Appendix B, Sect. B.21) were used to estimate the duration equation.
12.B Details of Poe (1988)
The author used ITERATE dataset, from which he generated a target-country “response strength” variable taking values from 1 to 5. Capitulation by the target country is accorded a value 1. Stalling with eventual compromise on demand was assigned 2. Bankok solution , meaning safe passage for terrorists in return for hostages and withdrawal of demands, was given value 3. No compromise and yet no shoot out was assigned 4, and finally no compromise and shoot out was accorded a value of 5. Multiple regression was used with the number of hostage-taking events as the dependent variable. The independent variables include response strength as well as dummy for being a developed or a developing country and regional dummy as control variables.
12.C Details of Brandt and Sandler (2009)
It uses the ITERATE dataset on international hostage incidents during 1968–2005 was used, to which a Poisson regression model (General Appendix B, Sect. B.19) was applied. For each of the three time-series on kidnapping, skyjacking, and the rest, the authors first identified “change points” in time—introduction of major counter-terror measures or initiative—from the data rather than exogenously imposed them as did earlier studies like Landes (1978), Enders et al. (1990a,b), and Enders and Sandler (1993). Each hostage series exhibited a different change point caused by a variety of circumstances. For instance, the kidnapping data contained ten change points. These are in February 1970, October 1977, July 1983, July 1988, May 1993, January 1998, December 1998, October 2000, April 2004, and November 2004. Each change point was linked to some exogenous changes. For example, November 2004 was associated with a decline in kidnappings in Iraq.
Three explanatory variables or covariates were considered: negotiation success, response of the target and death, each being a binary variable. Negotiation success took value 0 if some terrorist demands were met or 1 otherwise; response of the target was 0 if there is no shoot out or 1 otherwise; and death was equal to 0 if there are no deaths or 1 otherwise. While negotiation success was the primary variable of interest, the other two variables were included to estimate whether violent endings or deaths can predict immediate and future hostage-taking occurrences. The model included a lag structure for the number of hostage-taking incidents as well as other covariates, which, as we saw in Chap. H, enables the estimation of short-run and long-run effects.
12.D Details of Lee (2013)
ITERATE was the data source on hostage terrorism. The data on civil rights, political rights, and press freedom were obtained from Freedom House, which provides scores on political rights and civil liberties. The author used the data on civil rights, not political rights on the ground that the former measures the degree of democratic values, which are more important than political rights enjoyed by the public. The index ranges from 1 to 7 with 1 as the highest level of civil liberties. This is reversed. Press freedom is indicated by numbers 1, 2, and 3 representing “not free,” “partly free,” and “free.” The data on executive constraints was taken from Policy IV dataset. This index is from 1 to 7 with 7 being the most limitations on the executive leaders.
A long list of control variables were included, e.g., the level of real GDP, the growth of real GDP, land area, population, and number of involvements in foreign crises and variables are lagged by one year except time to election.. Poisson regressions for panel data were used. That is, it was assumed that y it ∼Poisson(λ it), where y it is the number of hostage-taking events in country i during year t.
Questions
-
12.1.
Consider the following deterministic sequential game. There is no chance mechanism like nature deciding if a mission fails or succeeds, and there is no probability attached to any action taken by the terrorists or the government. The left (right)-hand entry at the terminal nodes represents the payoff of the terrorists (government).
(a) What is the solution of the game, that is, given the payoffs, will the firm attempt kidnapping?
(b) Suppose, the government announces a no-concession policy and it is credible to the terrorists. Will they still attempt to kidnap?
(c) Suppose the payoff vector at node R was (0.5, −1). How would the answers for parts (a) and (b) change?
-
12.2.
From reading Yun and Roth (2008), determine by how much hostage safety is impacted if a group is Islamic vis-à-vis not being Islamic.
-
12.3.
“A democratic government as opposed to an autocratic regime is more likely to invite hostage-taking by terrorists.” Defend or refute.
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this chapter
Cite this chapter
Das, S.P. (2022). Hostage-Taking, Ransom, and Negotiations. In: Economics of Terrorism and Counter-Terrorism Measures. Springer, Cham. https://doi.org/10.1007/978-3-030-96577-8_12
Download citation
DOI: https://doi.org/10.1007/978-3-030-96577-8_12
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-96576-1
Online ISBN: 978-3-030-96577-8
eBook Packages: Economics and FinanceEconomics and Finance (R0)