Understanding Education's Influence on Support for Democracy in Sub-Saharan Africa

Abstract

Is education consequential for popular endorsement of democracy in developing societies and, if so, what are the mechanisms that account for this influence? We investigate the micro-foundations of the education–democracy nexus using a survey of 18 sub-Saharan African countries. We demonstrate that educational level is the strongest influence on support for democracy and rejection of non-democratic alternatives via its impact on comprehension of, and attention to, politics. This is consistent with a cognitive interpretation of the effects of education on democratic values rather than one which treats education as a marker of economic resource inequalities.

Notes

1. Glaeser et al. (2007) present evidence linking education with socialisation and therefore build a model where education fosters democracy by shaping incentives for joining political participation.

2. Modernisation theories have taken various forms. Here we are primarily concerned with the approach examining ‘individual modernisation’ (for example, Inkeles and Smith, 1974; Inglehart and Welzel, 2005).

3. A recent study of the impact of schooling on democratic attitudes in Malawi provides evidence of a strong relationship between educational level and support for democracy (Evans and Rose, 2007) although this does not examine the mechanisms through which education impacts on democratic attitudes.

4. The 18 countries are: Benin, Botswana, Cape Verde, Ghana, Kenya, Lesotho, Madagascar, Malawi, Mali, Mozambique, Namibia, Nigeria, Senegal, South Africa, Tanzania, Uganda, Zambia, Zimbabwe. See http://www.afrobarometer.org for further information on the surveys and sample design.

5. The dataset also contains a response category referring to informal education. This includes 4.4 per cent of respondents, the majority of whom are Muslims, which suggests that it is probably composed mainly of pupils at madrasahs. We estimated models with ‘informal education’ distinguished from ‘no education’ but found no significance differences.

6. The proportion of those aged 18–25 with post-primary education is lower (8.4%), probably because some of this age group are still in secondary school (half of this age group have achieved this level of education).

7. One of our initial hypotheses was that the impact of education on democracy would be stronger in Anglophone compared with Francophone and Lusophone countries because of the difference in their colonial inheritance. Interaction between effects of education on measures of support for democracy and type of colonial inheritance proved not to be significant.

8. We carried out extensive checks on the robustness of estimates to case selection. We found no evidence that our estimates were substantially affected by outliers.

9. It is possible that there is variation across and within countries with regard to the political content found in mass media. There is no reason to think that this would change the general pattern of substantive findings regarding the strength of education effects reported here, although it could be a useful area for further research.

10. Ordered Probit can also be used for analysing such coarsely scaled dependent variables, but these models bring in further assumptions of their own. The advantage of OLS is its robustness to violations of its assumptions and general interpretability.