Citation

Barros CP, Gil-Alana LA. Def. Peace Econ. 2006; 17(2): 95-116.

Copyright

(Copyright © 2006, Informa - Taylor and Francis Group)

DOI

10.1080/10242690500464214

PMID

unavailable

Abstract

Most searchers performing unit root tests on terrorism series reject the null hypothesis of unit roots (I(1)) and conclude that terrorism is stationary (I(0)). In this paper we analyze ETA activity in Spain during the last 30 years by means of examining its degree of dependence across time, using fractional integration or I(d) techniques. The results show that the activity of ETA is persistent to some extent, with an order of integration of about 0.40, implying stationarity, but also long memory behavior. We argue that this strong degree of dependence between the observations might be explained by the historical background underlying the political conflict in the area. In addition, the results indicate that the most significant factors contributing to a reduction of violence are those related to political pacts among the political parties in the Basque region. In order to put an end to ETA's violence, these accords should involve both nationalist and non-nationalist groups.