Once a month, BSI members present an academic research article in a simple and accessible way. Our goal? To provide a better understanding of how research can help answer current questions and shed light on the complexity of answering questions that may sometimes seem simple.
Mitch Downey, Did the war on terror deter ungoverned spaces? Not in Africa, Journal of Development Economics, 2021
Abstract:
- Can the threat of invasion lead a government to better control territories where its influence is weak and where armed groups can develop?
- Mitch Downey examines the effect of threats of US intervention after September 11, 2001.
- In Africa, there is no empirical evidence of any increased government control over ungoverned territories after 2001 and the threats of US intervention in the context of the war on terror.
Ungoverned territories (those where the officially responsible state has little or no presence) are now a major issue for both developing and developed countries. For the former (and for international institutions promoting economic development), these territories are unstable regions where essential services (health, education, etc.) are not always provided. For the latter, these territories can serve as operational bases for groups of combatants who can then initiate actions against these countries or harm their interests on the ground. The French intervention in northern Mali and the recent fighting near the city of Palma in Mozambique are evidence of the topicality of this issue.
So what can developed countries do? In the article « Did the war on terror deter ungoverned spaces? », Mitch Downey analyzes the impact of the war policy implemented by the Bush administration after September 11, 2001, on these ungoverned territories. However, rather than focusing on territories directly invaded by the United States, he looks at those that could have been invaded. More specifically, the question examined by the author is: was the threat of an American invasion (which would have been followed by a change of government) enough to convince African governments to make efforts to better control their territories? [1]
The difficult task of measuring the effect of the war on terror
Does threatening states push them to better control their territories? Mitch Downey suggests that they do not, at least not in Africa. But how did he arrive at this conclusion? The author proceeds step by step.
First, he must justify that the policy being evaluated (the war on terrorism, which began in 2001) could have prompted governments to better control their territories. In other words, he must show that the threat of invasion was credible and targeted states that did not (or poorly) control all of their territories. To do this, Downey draws on statements by senior US officials to establish that: (a) the United States threatened a large number of states with invasion, including some in Africa, (b) this threat was credible (after all, some of the targeted states—Afghanistan and Iraq—were actually invaded), (c) these threats were aimed in particular at failed states, where there were ungoverned areas that could serve as refuges for armed groups, (d) these states were targeted in particular because of the existence of these areas.
Figure 1: Government violence and efforts to control territory
Source: Mitch Downey, Did the war on terror deter ungoverned spaces? Not in Africa, Journal of Development Economics, 2021 (Figure 1, page 10)
Once these points have been established, the next step is to identify the ungoverned areas. There are two possible approaches. Downey relies on both US government reports and indirect measures. Indirect measures are based on the fact that these ungoverned territories are often close to borders, without mineral resources, sparsely populated, etc. Therefore, territories with these characteristics can be suspected of being « ungoverned, » even if they are not explicitly designated as such.
With the credibility of the threat established and the « ungoverned » areas identified, the author needs only one more variable to measure the effect of the policy: government presence. How can the presence of a government within its borders (in ungoverned territories and elsewhere) be measured? Here again, two types of measures can be considered: variables that directly reflect the presence or action of the state, such as the number of births in public hospitals, the percentage of individuals vaccinated, voter turnout in elections, etc., and indirect measures related to government violence. For the latter, the general idea is that if a government strengthens its control over a formerly « ungoverned » territory, it will probably have to confront groups (rebels, terrorists, etc.) that had established themselves there. These conflicts will then cause a (temporary) increase in government violence. Figure 1 illustrates this logic with the case of Senegal. In this country, the Senegalese state sought to increase its presence in two regions (Sédhiou and Kolda) between 2008 and 2015. This can be seen through « traditional » measures but also through peaks in violence between 2008 and 2010 (visible in Figure 1).
Once these measurement issues have been overcome, an econometric analysis is possible. This consists of comparing the evolution of the presence of states (potentially targeted by the US) in their ungoverned territories and comparing it with the evolution of the same variable in other territories. A stronger increase in the presence of the state in (previously) ungoverned territories than in others (where the state was already well established) would suggest that the threats of invasion did indeed motivate states to better control these « ungoverned » areas. [2]
This method and Downey’s results are illustrated in Figure 2. The horizontal axis corresponds to time. The area between -4 and 0 shows that there was no significant difference in the evolution of state violence between ungoverned territories and others (the vertical axis, the coefficients measuring these differences in evolution, are close to 0) before 2001 . [3] This highlights the relevance of the empirical analysis, as the territories compared had similar trends before the policy was implemented. After 2001 (between 0 and 8), we observe no increase in violence in these ungoverned areas (the coefficients remain close to zero and are not statistically significant). This result suggests that states did not take action to better control areas where their authority was contested. Consequently, the threats of invasion after September 11, 2001 do not seem to have had an effect in Africa on ungoverned territories.
Figure 2: State violence in ungoverned territories before and after 2001.
Source: Mitch Downey, Did the war on terror deter ungoverned spaces? Not in Africa, Journal of Development Economics, 2021 (Figure 2, page 13)
Conclusion
Downey’s findings suggest that threatening a state is not enough to make it exercise better control over its territories. While several mechanisms may explain this « lack » of results and further research is needed to better understand them, it appears that in order to prevent the proliferation of « ungoverned » territories, other policies—such as strengthening development aid or military cooperation between Western and African states—must be undertaken.
[1] This question boils down to analyzing one of the forms of the military adage « to have peace, prepare for war. » Indeed, a (credible) threat of invasion by the United States would lead the targeted states to better control their territories. These could then no longer serve as a refuge for armed groups, which would render American intervention unnecessary. As Mitch Downey discusses, this doctrine remains influential within the US Department of Defense.
[2] More generally, Mitch Downey draws inspiration from the double difference method. He also conducts a large number of tests to ensure the validity of his empirical strategy.
[3] This does not mean that there is no difference in the level of violence, but rather that there is no difference in the trends in violence. This amounts to testing the hypothesis of parallel evolution before treatment.
