No predictable fault lines

Aid organisations such as Oxfam and Welthungerhilfe argue that recent violence in Kenya was not due to ethnic divisions, but rather to a widening gulf between rich and poor. Nonetheless, the perception dominates that the elections were rigged against the Luo candidate, Raila Odinga, by the ruling President Mwai Kibaki, a Kikuyu – even though Odinga had teamed up with Kibaki in the elections five years ago and, at the time, “hauled the Luo votes into the Kibaki camp”, as a German newspaper put it.
But is it true that voter alignment was determined by ethnicity in the latest and earlier Kenyan elections? That is hard to establish, considering a recent paper by Africa scholar Gero Erdmann of the Hamburg-based German Institute of Global and Area Studies (GIGA). Erdmann argues that there is no hard evidence in support of ethnicity being what drives voter behaviour in Africa, though he does admit it is reasonable to think along those lines.

It is reasonable, he says, because – at least in Western societies – the way people vote has so far been satisfactorily explained by social fault lines, such as those between capital and labour, town and country, or religious and secular environments. In the industrialised countries, the past 150 years have seen many political parties emerge, deriving the bulk of their support from such milieus. Only in recent times have old loyalties begun to wane.

In Africa, Erdmann explains, those fault lines never acquired the significance they had in the West. Ethnic background, however, plays an important role for identity formation, though to varying extents. Erdmann stresses that Africans make no secret of this fact, and therefore it seems reasonable to apply the fault-line theory to ethnic groups. That said, the researcher points out that there is currently no clear evidence of voting patterns really being defined by ethnic allegiance.

Erdmann does not find that surprising, as ethnicity is not always clear-cut. In most cases, ethnic groups overlap some way or another. There are splinter groups and sub-groups. What is more, the impact of ethnicity on identity and behaviour is not a fixed quantity. Take Kenya, for example. In 2002, Odinga endorsed Kibaki. This time around, he stressed ethnic differences.

At the end of his study, Erdmann puts down the ethnic lenses and recommends that research focus on a more general question: “What makes African voters tick?” (Also note Mildred Ngesa’s comment on Kenya on page 86.) (ell)