The Politics of Local Control in Electoral Autocracies

Abstract

In this dissertation, I ask whether local state capacity strengthens electoral autocrats. The prevailing consensus is that local state capacity, and state capacity more broadly, strengthens autocrats. That means that we generally explain declines in regime durability or regime collapse by pointing to shocks or structural changes which affect the strength of the state.

These accounts hinge on the stylized fact that electoral autocrats are hegemonic. If they are hegemonic, incumbents command state capacity to stay in power without any real constraints. I introduce the concept of `local control' --- which party wins local elections so can mobilize local state capacity associated with these offices --- to demonstrate that electoral autocrats face meaningful subnational constraints on their ability to mobilize local state capacity.

I argue that local state capacity is a double-edged sword for regimes. Under incumbent control, local state capacity strengthens the regime's hold on power. However, I show that if a sufficiently sophisticated opposition party wins local control, a strong local state may be a threat to regime durability. Local control empowers opposition parties to win support and survive regime attempts to suppress them. This makes it easier for a credible challenger to the regime to emerge and harder for the regime to contain such a threat to its rule. I therefore demonstrate that opposition parties are more than just co-opted tools of electoral autocrats and in fact pose a challenge to them.

My dissertation relies on qualitative and quantitative data collected during 14 months of fieldwork in Tanzania. I draw on over 200 interviews, administrative data collected from ministries and subnational governments and an original survey. I exploit variation in the phasing of decentralization and geo-coded data on local public goods to demonstrate the importance of local state capacity to autocrats’ electoral toolkit. I use interviews to trace how politicians use local capacity to win support. I exploit administrative data and interviews to show that local control determines parties’ electoral strategies. Finally, I use surveys, election data and interviews to show how local control and regime responses to it influence political behavior and subnational regime type.