Document Library

Key Text Is Clientelism at Work in African Elections? A Study of Voting Behaviour in Kenya and Zambia

Author: Daniel Young
Date: 2009
Size: 23 pages (665 KB)

Access full text: available online

Summary

Does clientelism play a major role in the voting decisions of African elections? Which factors are most important to African voters when choosing a candidate to vote for? This Afrobarometer working paper argues that clientelism is not a major factor in voting decisions. Instead, voters focus on issues relating to the provision of local public goods and the frequency of an MP's visits to the constituency.

The concept of clientelism pervades the literature on African politics. It has evolved from ideas surrounding post-colonial Africa's authoritarian era. While there are several relevant definitions, all are variations on a mutually beneficial exchange between patrons and clients. Variations include constituency service, elite hierarchy and personal exchanges.

While historically considered a major factor in African elections, clientelism may not be as important as the literature suggests. First of all, the ability of candidates to connect with constituents on the personal level is small. Given that there are 50,000 to 100,000 people living in electoral constituencies, candidates cannot deliver personal goods to all (or even most) of them. Also, constituents recognise that one-time gifts will not significantly alter their standard of living.

The responses from voters in Kenya and Zambia reflect these arguments. Overwhelmingly, voters are more interested in candidates who offer local public goods than private gifts. Indeed, the responses revealed several facts in support of this argument, such as:

Voters are more likely to choose candidates that visit their constituency. MPs who visited their constituency at least once a year were 37% more likely to be re-elected than those who never visited.
Economic voting was not a factor. Changes in living conditions do not correlate to voting patterns.
Direct contact with candidates does not correlate to more votes. This is significant because clientelism works through exchanges between candidates and the constituency.

These results indicate that clientelism does not significantly impact the choices of voters in Kenya and Zambia. Furthermore, they suggest that:

Voters in Kenya and Zambia are strategic. They are more interested in candidates who can provide public goods rather than private gifts.
Contrary to prior claims, democratisation does not seem to increase clientelism.
While increased contact between candidates and voters did not suggest that clientelism was a major concern, clientelism may be carried out by intermediaries.
In order to determine that clientelism is not a factor in African elections, future research should expand the scope of this study.

Access full text: available online

Source: Young, D., 2009, 'Is Clientelism at Work in African Elections? A Study of Voting Behaviour in Kenya and Zambia', Working Paper no. 106, Afrobarometer.
Author: Daniel Young , djyoung[at]gsu.edu
Organisation: Afrobarometer, http://www.afrobarometer.org