Does cellphone coverage make violence more likely in Africa?

Jan Pierskalla and Florian Hollenbach argue that it does in a new article in the American Political Science Review.

Overall, our quantitative models demonstrate a clear positive association between cell phone coverage and the occurrence of violent organized collective action. This effect persists when controlling for a series of standard explanations of violence, as well as unobserved, time-invariant factors at the country and even grid level. Plainly, our results suggest that local cell phone coverage facilitates violent collective action on
the African continent.

This article should set off some interesting debates. I’ll leave it to those more statistically adept to assess their analysis (although I wonder whether the authors will get some pushback for their claim that regulatory efficiency is a good instrumental variable for cellphone coverage and is causally unconnected to levels of violence). Nonetheless, this piece does draw some interesting and potentially important connections between the diffusion of communication technology and ‘real world’ outcomes. As the authors note, we have seen a number of pieces over the last couple of years asserting that new communication technologies have helped e.g. foster the spread of the Arab Spring revolutions. However, we’ve seen precious little work that really tries to demonstrate systematic linkages rather than assert them. Pierskalla and Hollenbach’s piece begins to think about how we might want to investigate these linkages.

4 Responses to Does cellphone coverage make violence more likely in Africa?

  1. Kreator May 21, 2013 at 7:07 am #

    FROM PSR:

    Shae
    Is this a joke? Two basic, basic points:

    1. From a social scientific perspective, ICT is best regarded as politically neutral, as a set of tools that activists can potentially use to promote any kind of political ideology or movement. (Bratton 2013).

    2. According to the same article, cell phone usage ranges from approximately 83% on the low end to 94% on the high end– i.e. across the continent, usage is extremely high.

    Any clarifications?

    4 HOURS AGO # QUOTE 0 YEA 0 NAY

    Marva
    Empirics aside, this article is just incredibly poorly written. Who starts a paper with “This quote exemplifies..”? And the sheer amount of typos is impressive too.

    2 HOURS AGO # QUOTE 0 YEA 0 NAY

    Hale
    What utter crap. What accounts for all the violence of the 1990s?

    2 HOURS AGO # QUOTE 0 YEA 0 NAY

    Alphonzo
    Given the breathtaking spread of cell phone technology worldwide

    2 HOURS AGO # QUOTE 0 YEA 0 NAY

    India
    The sad truth is that this really is the pathetic future of our discipline. Its essentially data mining with no real theory driving the findings. They gesture towards the civil war literature with some trifling talk of the collective action problem in insurgencies, but they cannot (and don’t even try to) square their own findings with the empirical observation that the number of insurgencies and civil wars in Africa has gone down steadily since the cold war.

  2. Jan k May 21, 2013 at 6:13 pm #

    Haven’t read he article, but his almost certainly conditional on a host of negatives within a country. I don’t buy an unconditional argument.

  3. J Leeland May 22, 2013 at 12:12 pm #

    I think there’s a strong correlation between cell phone coverage and violence even without ‘conditioning on a host of negatives’.

    If you want to stage a scene, set a bomb, start a riot or conflict, where would you do it? In a dense urban area or in the middle of nowhere? Probably former. Those are also areas with high cell phone coverage.

    That’s why I bet the two things are correlated, with or without covariates. Yet, this correlation seems purely spurious and shouldn’t be mistaken as causation.

  4. James Conran May 23, 2013 at 3:56 pm #

    Chris Clary points out that violence is likely under-reported in areas lacking phone coverage:
    http://thesmokefilledroomblog.com/2013/05/22/cell-phones-and-conflict/

    (I also like the first comment about placing research in relation to a production possibility frontier “where well-identified is on the Y axis and ‘Importance of question’ is on the X-axis, this study is below the 45 degree line. The question is whether it is on the frontier below the 45 degree line, or somewhere inside.”)