Chris Blattman

Search
Close this search box.

Can mathematics give us a way out of the Kenyan election crisis?

Kenya has seen a string of inefficient and kleptocratic (self-enriching) rulers, and recent weeks have seen a disputed election and tragic ethnic violence. (See my blog posts on Kenya here.)

Kenya is far from alone. The politics of fear, division and violence are a too common feature of African politics (as well as the West’s not so distant past). How is it that weak, inefficient, and sometimes criminal rulers stay in place in a democratic society? Why can they enrich themselves and stay in power? What path charts a way out?

The answer might lie in a little math and logic from Gerard Padro i Miquel, a Stanford political economist.

Suppose we have a society of two ethnic groups: K’s and L’s, each with a K and L leader. Ethnicity is politicized and divided. Suppose also that the leadership succession process is likely to be unstable and unpredictable, and that it can result in a switch of power between ethnic groups. Finally, suppose that a leader finds it difficult to tax or extort from a single ethnic group (say, because he can only tax industries or trade) but finds it easy to dole out pork and patronage based on ethnic identity.

In an NBER paper, Miquel illustrates that these three assumptions alone are enough to sustain a kleptocratic and inefficient regime.

The logic is simple: if a K is in power, members of K’s ethnic group fear the rise of a L to power more than they fear the thieving or ineptitude of the K leader. Even if his taxation or ineptitude undermines the economy, he can direct some of that back to other K’s at the expense of members of the L group.

In this world, it even makes sense for the K leader to torment and abuse members of the L group. Why? Because that enrages the L’s so much that they are likely to take revenge if an L leader ever gets power. This (possibly well-founded) fear gives the K group even more incentive to support their K leader.

This is the politics of division and fear.

Pretty depressing? Maybe not. There might be a way out of such an equilibrium. Miquel doesn’t discuss a way out in his paper, but to me a few options for the opposition leader or the international community exist.

For instance, if the opposition or international community can enforce a more peaceful and stable transition of power (e.g. through better election monitoring), or guarantee safety after a transfer (e.g. through peacekeepers), it will reduce the incentives for the leader in power to abuse his or her authority.

More importantly, if an opposition leader can credibly commit to not taking revenge and to governing the K’s fairly, then the K’s may see fit to get rid of their leader.

The problem is, how does an opposition leader do so? Anyone can make promises, but can they be believed? I imagine an L leader could signal his honest and nonviolent nature by making great sacrifices. Think Nelson Mandela’s 20 years in prison, or Gandhi’s self-sacrifice. No leader who was truly kleptocratic or selfish would ever opt to endure such hardship. By proving they can be better than the current leader, the Mandelas and Gandhi’s of the world erase the uncertainty of the succession.

Unfortunately, Kenya’s opposition leader Raila Odinga is doing exactly the opposite of their example. At best he is not signalling his honesty and non-violence. At worst, the rumors are true and he is helping mastermind a disenfranchisement (or even a killing) of Kikuyus.

If only Raila could do the math.

3 Responses

  1. One of the issues is that the potential involvement of the “international community” is tied to factors that have nothing to do with pragmatism – peacekeeping missions (when there is a peace to keep…), sanctions, election monitoring are perceived (and rightly so) as interference, and we constantly tread a very, very thin line…. I really do believe that it is up to African leadership (the real deal – the Gandhis and Mandelas) to push for paradigmic shifts.

    I’m very ambivalent about foreign intervention (in the broadest sense of the term) in internal politics – while it seems unimaginable to simply stand by and wait for matters to be resolved by the primary stakeholders, I’m pretty sure anyone who knows anything about relations between developed and developing countries also knows that imposed external solutions pretty much always fail.

    Either because the “outsider” nature of the solution is exploited by politicians, who point to foreigners as the cause of all problems (hence clearing themselves of responsibility), or because the solution fails to meet the socio/political/cultural (etc.) specificities of country X….

    The Economist “correspondent’s diary” offers a great micro perspective on the current events – I highly recommend the interview with the history professor, Alan Ogot: http://www.economist.com/research/articlesBySubject/displaystory.cfm?subjectid=7933598&story_id=10674712

  2. CHRIS BLATTMAN i forgive your academic claptrap about Raila Odinga, you are just one of those naive westerners who roam about Africa thinking they know the continent and so they have the correct panacea. I will not even attempt to respond to you intellectually. Its just a waste of time.
    elly odhiambo
    ellyomondi@gmail.com

  3. Changes in electoral systems can help too.

    Sure the 25% in 5 out of 8 provinces provision is a start when it comes to guaranteeing that no one can limit his support to an ethnic or regional base but it’s not enough.
    2 rounds run-off systems are far better at it. So is proportionnal (single or multi-constituency) for parliament instead of FTTP.

    Or one could get very creative and enginneer ideology-based parties (though, it does happen that ideological lines are close to ethnic lines) or have an indirect system which kills the regional advantage.

    Another very good option is to get rid of directly-elected presidents and operate through the congress (with a PM) and demand large coalitions (only a consensual candidate would be able to gather say 2/3 support).

    But yeah, the paper is pretty much dead-on. I even know of cases where the K leader doesn’t even have to redistribute to the K population. The threat of L’s vengeance over perceived priviledges of K group is more than enough to make the K group support the K leader.

Why We Fight - Book Cover
Subscribe to Blog