In his keynote, Andreas Schedler outlined some cornerstones for unified theory building for democracies and autocracies. Stating that there is a qualitative divide between democracies and autocracies concerning basic principles – equality and liberty versus opression and hierarchy – Andreas pointed out that both types of political regimes are inhabited by human beings. This is a reason and an anchor for unified theory building, he argued: There should be one theory to explain how humans act – even under different conditions on the meso- and macro level, micro foundations should be the same.
Even though different on the aggregate level, Andreas identifies real similarities between autocracy and democracy. There are autocratic practices in democracies as well as democratic practices in autocracies. E.g. there are peaceful challenges to rule in autocracies and there are violations of human rights in democracies as well. As the micrologics of behavior are very similar, Andreas argues that both types of political regimes are essentially comparable. Elections look similar, but work differently in terms of vertical accounatbility and responsiveness. A unified theory of electoral accountability thus would ask when elections fail in terms of accountability and responsiveness.
But developing a unified theory faces difficulties: theoretical and methodological ones.
In terms of theoretical foundations, Andreas claims that many studies suffer from very simple hypotheses that are decoupled from deeper theorizing. Thus, we must do better in theory development and inetgrate broader aspects of social theory as an important basis of theories in comparative politics.
Methodologically spoken, testing theories is a tough challenge. We might be incapable to test these theories, because of a lack of good data on autocracies. There are structural obstacles like the shadow of opression and the opagueness of autocracies as such. Also, it is dangerous tot o do field research in autocracies, and „it is not worth dying for social sciences“. Andreas argues that we usually managed these problems by keeping our distance. Nevertheless, we do have at least some information from within those countries from ethnographic work, field research, interviews, some surveys, some event data. But most is desk research that leads to big distance between us and our objects, Andreas argued. And especially proxy variables are problematic in this context.
In this view, comparative autocracy studies face a long list of missing data. The microfoundations have been put to blackboxes and much of the available data (socio-economic, expert judgements, thin event data) leaves out almost everything about things we are interested in: microfoundations of institutions, dynamics, citizen attitudes, elite attitudes and so on.
Citing Milan Svolik, Andreas claims that due to these theoretical and methodological problems, „the days of crossnational regression analysis are over“, as we have thin theory and thin evidence. His proposition then is: we need much more reflection on security, ethics, methodology, inferential uncertainties, field work and collaborative work with people wihin autocracies. We must get hands a bit dirtier, Andreas says, and stop pretending to know things we cannot know!