Unemployment and Violence in Northern Ireland: a missing data model for ecological inference.1

Contrary to the body of literature in political violence, and the rhetoric of many parties of the conflict, time-series models of “the troubles” in Northern Ireland by White(1993) and Thompson(1989) have found no evidence that economic conditions effect the intensity, sources or direction of violence. I show that several methodological flaws exist in previous models. They fail to address the discrete, count nature of the data, the contagion present from aggregation over time, pooling issues from different types of violence, and the over dispersal of deaths. However, the key problem, acknowledged even by the authors themselves, is that all measures of unemployment aggregate Protestant and Catholic unemployment rates into one single measure. Clearly, Republican paramilitary violence should be caused chiefly by the Catholic unemployment rate rather than the aggregate unemployment rate. Possibly also Loyalist violence may be a result of the Protestant unemployment rate. However, little historical data exists on disaggregated rates of unemployment, thus previous studies forced to use aggregate measures have found no connection between economic conditions and the incidences of violent death. Using a model that combines methods of Multiple Imputation to recover missing data (King Honaker Joseph Scheve 2001) and the literature of models for Ecological Inference problems (especially King 1997) I estimate the disaggregated unemployment rates by religion from the available data. Unemployment is shown to be a leading cause of the violence by Republican factions in Northern Ireland. I also estimate the effectiveness of the various security forces (RUC, UDR, British Army) in lowering the death rate. I show these forces have differential effectiveness depending on both the sectarian source and target of violence using a model of probabilistically distributed lags appropriate for time-series event count data. 1. ECONOMICS, POLITICS AND VIOLENCE The connection between economic conditions and political violence is central to multiple approaches to the study of conflict. For normative policy makers, economic conditions are often the only policy instruments with the prospect of short term manipulation or improvement. For more positivist observers, measures of aggregate economic conditions are frequently the only variables with consistent explanatory power, other than previous levels of violence. The unfortunate relevance of the topic has brought scrutiny and charged debate to both vantages in this long-standing literature. At one time the connection seemed concrete. Economic conditions have served as a leading and robust predictor of conflict throughout the quantitative literature on political violence, from its earliest roots (Russett 1964, Feierabend et al 1969, Hibbs 1974). Measures of aggregate economic conditions (Fearon and Laitin 1999, Sambanis 2001), economic growth or instability (Olson, or closely related, Hovland and Sears 1940, Tolnay and Beck 1995) or economic inequality (Gurr 1968, Gurr 1971, Muller 1985, Muller and Seligson 1987) are often the only variable with explanatory power, other than previous levels of violence. Most of these studies are cross-national, sometimes additionally across time. However, a number of recent counter arguments have formed, partially as the result of increasingly rich and textured data, and partially as the result of shifting units of analysis. Starting with Russell and Miller (1977, 1983) a number of authors have collected detailed intra-national series of violence, support for violence, or the personal histories of instigators of violence (particularly Kruger and Maleckova 2003, Berrebi 2003) and such intra-national level data tends to add no support to a connection to economics. For example, individual instigators of violence generally tend to come from better economic circumstances than average in their country or region. Kruger and Maleckova, from their own work and a synthesis of other chosen studies rather strongly conclude “Any connection between poverty, education, and terrorism is indirect, complicated, and probably quite weak.” 1 In summary, cross-national aggregate studies find a clear connection between political violence and economic conditions, but in sub-national studies there is a growing literature of null results. 1.1. The Case of Northern Ireland Northern Ireland is probably the most detailed and documented case in the study of political violence, and yet the best empirical analyses yield impossible results. The two most thorough studies, by Thompson (1989) and by White (1993), find little connection between economic conditions and the level of violence, and what little effect they do find is in the wrong direction, with increased unemployment decreasing the level of violence. This exception is important because it stands in opposition to most any theory of political violence (rational actor, modernization, or deprivation theory) and to the preponderance of empirical evidence previously touched on. This has led some outside commentators to speak of Northern Ireland as an innovation in violence, a precursor of instability that will be mirrored in other industrial democracies, or even a “post-materialist insurgency”2. However, these exceptional empirical findings are notable not simply because they oppose In another article, the authors similarly argue that “Instead of viewing terrorism as a direct response to low market opportunities or ignorance, we suggest it is more accurately viewed as a response to political conditions and long-standing feelings of indignity and frustration that have little to do with economics.”