Ethnic diversity and development in Kenya: limitations of ethnicity as a category of analysis

Abstract

This paper challenges the relevance of ethnic diversity in explaining the development impasse in Africa, using Kenya as a case study and arguing that the neopatrimonial and ethnic diversity theses are contradictory as explanatory variables of Africa's development crisis. The former speaks to homogeneity while the latter implies heterogeneity in the public policy process. This paper singles out the ethnic diversity thesis as the weaker of the two, arguing that ethnic groups lack a collective ethnic agenda that can be flagged as consequential for the political economy of the polity. Ethnic group intervention in the political arena is not determined by innate ethnic characteristics: it is due to the tendency for certain actors in the body politic to exploit institutional failure. With proper institutions, the assumed burden of ethnicity in African development would naturally evaporate. Focusing on supposedly ethnic problems is the wrong target.

Keywords:

• neopatrimonialism
• ethnic conflict
• institutions
• elite interests
• ethnic identity
• public policy

Acknowledgements

The support of the following institutions is greatly appreciated: Yale University (Council on African Studies), University of Cape Town (EEU), UNU-IAS and JSPS. The author thanks the editor and reviewers of Commonwealth and Comparative Politics for their constructive intervention.

Notes

1. The 1999 population census, for example, collected ethnic data but it was not released supposedly because, to the government, it was not the best thing to do (KTN 1).

2. When African leaders proscribed multipartism, the forwarded rationale was that diversity of political organisations would polarise the polity and thus arrest the development in Africa.

3. Take, for example, the case of health-care reform in the US Congress. Opponents cast Obamacare in the public script as a case of the role of government intervention in citizens' lives; so if it fails, the explanation will be ideological, not race, tribe, religion or other non-ideological cleavages.

4. The other version of this argument is those countries that have had ethnic conflicts even when ethnic diversity is not one of their defining features (Rwanda and Somalia being such cases; for the former, see, for example, Vuningoma, 2010).

5. Analytically, however, a question can be raised why the resultant cleavages take ethnic and not, for example, class lines. As Rawlinson (2003) and Elkins (2008) have pointed out, the ethnic political and economic elites enjoy a lifestyle dissimilar from those of the rank and file (that they claim to represent while using the plural pronoun ‘we’).

6. There are people whose names are Tribe, and are fine with that.

7. Colonialism, however, imposed boundaries not to create tribes, but for purposes of administrative control. 

8. I am grateful to Professor Kimani Njogu for his intervention in this point.

9. Thus, unlike Green's (2011) reservation, we forward that institutions should be conceived both in terms of content and stewardship.

10. Does it become an issue affecting development once induced? No; corruption is a better predictor of the development trajectory than ethnicity. 

11. It is not unusual to hear comments like Moi's vice-presidency having been an exchange of land for power, hence the tranquility in as far as Moi's power was guaranteed, but when it was challenged, and especially by those he allegedly had contracted with, the Kalenjin moved to challenge land ownership by the ‘alien’ communities.

12. KADU's formation has been associated with colonial settlers who wished to dilute KANU's strong nationalist thrust which they feared could dispossess them (as Mugabe recently did in Zimbabwe). Colonial involvement was part of their strategy of preventing any broad-based pan-ethnic anti-colonial formations (see, for example, Berman, 1998, p. 317).

13. Indeed one can argue that the outgoing 2003–2007 electoral cycle witnessed the most undeniably equitable distribution of national resources since independence. That administration introduced a constituency development fund of 2.5% of annual national revenue collections that was shared by all constituencies using a formula designed by a bipartisan committee of parliament.

14. See the quotations from Muite (1991) and Odinga (1991) cited earlier on the single party political system and homogenisation of participation in policy-making.