Causality between terrorism and economic growth

Meierrieks, D., & Gries, T. (2013). Causality between terrorism and economic growth. Journal of Peace Research, 50(1), 91-104.

Keywords: economic growth, institutional capacity, macroeconomic resiliency, panel causality, terrorism (domestic + international)

Method: LSDV (dynamic, fixed effects), Granger causality

Meierrieks and Gries offer another take on the relationship between terrorism and economic growth. They do not attempt to measure the magnitude of the effect of terrorism on growth, or vice versa, but instead propose a number of tests to verify whether such a relationship exists.

The authors’ approach is refreshing in the sense that they account for the possibility of a two-way causality between terrorism and growth which is not expected to be constant neither across counties nor over time. They attribute the causal heterogeneity across space to differences between countries’ development levels, political system and institutional capacity. The variation over time is explained by ideological and geographical shifts. There is some criticism of previous studies which assumed that the relationship between terrorism and growth is constant over time. Overall, Meierrieks and Gries expect that due to the demise of left-wing terrorism, which was socio-economically motivated, the impact of growth on terrorism should be smaller for the post-Cold War period. Conversely, the effect of terrorism on growth should be larger for the post-Cold War times when terrorism shifted away from Western Europe to less economically resilient places.

Their analysis is carried on a large sample of 160 countries over the past four decades. The data on the number of terrorist attacks and victims is taken from GTD, thus it includes both domestic and international terrorist activity. Unfortunately, there is a drawback to this approach. The authors seem to miss the fact that the two types of terrorism are driven by different factors, this means that they may respond differently to economic growth. They also affect different economic agents, for instance international terrorism may affect foreign investors (or aid agencies) directly and thus discourage them from locating their activities in the country in question. The study includes a number of control variables which lead to a separation of several subsamples based on the level of development and regime type, growth performance, intensity of terrorism, political stability, and regional and cultural membership.

The empirical part of the study tests for Granger causality in dynamic models with fixed effects and tweaks it appropriately to examine the heterogeneous and two-directional relationship between terrorism and growth. The applied methodology is carefully described and intuition behind each of the tests is well explained. The authors show that the terrorism-growth relationship changed over time. During the Cold War, growth was Granger-causing terrorism, while in the post-1991 period terrorism was affecting growth. Both relationships are significant at one lag which, according to the authors, means that terrorists quickly adapt to economic changes and macroeconomic effects of terrorism are short lasting. The next exercise tests whether the relationship is homogenous across countries. As expected, countries differ in their economic susceptibility to terrorism and some states remain more resilient than others. Also, growth doesn’t uniformly cause terrorism across countries. Meierrieks and Gries dig deeper and attempt to establish which factors affect the growth-terrorism connection. Their main finding is that growth exerts casual effect on terrorism in poorer countries, regardless of the quality of political system. This is contrasted with the situation in more developed countries where terrorism may be driven by non-economic factors such as foreign policy, separatist conflicts and rapid modernization. When looking at the terrorism’s causality on growth, they conclude that non-democratic countries suffer adverse consequences while sound democracies do not experience negative effects of terrorism on growth. The difference is explained by the level of centralization of government and the way countries’ address population’s needs.

The authors follow on to divide countries by their geographical and cultural affiliation. An interesting finding is that the causation from growth to terrorism is only present in Latin America., while terrorism negatively affects growth only in Islamic and African states (post-Cold War). Most importantly, the causal relationship between terrorism and growth, and vice versa, is only present in countries with above average intensity of terrorism and those that experience episodes of civil war. Other countries seem to be economically resilient to terrorism.

To sum up, the paper’s goal is rather modest as it doesn’t seek to estimate the magnitude of the effect of growth on terrorism, and vice versa, but instead focuses on simply establishing whether there is any causal relationship between the two variables. In spite of an uncomplicated methodology it delivers insightful results and at the same time largely escapes modelling difficulties which plague many studies on the subject. Nonetheless, the authors should recognize the different nature of domestic and international terrorism. Perhaps this is something to look for in follow-up studies.