Some new research on the empirical determinants of state fragility

 Graziella Bertocchi summarizes some of the recent work she has done with Andrea Guerzoni over at VOX.eu.

The causes and implications of state fragility – also known as state failure – are not yet well understood. This column explores the determinants of state fragility in sub-Saharan Africa and finds that institutions – as measured by civil liberties and the number of revolutions – are the main drivers. It says institutions trump economic, geographic, and historical factors. 

The interesting findings are:

We find that institutional variables are the key determinants of state fragility. The probability of a country having a fragile state decreases with the level of civil liberties and increases with the number of revolutions. Economic determinants such as per capita GDP growth and investment are not significant when we introduce standard controls for omitted variables. Geography is also insignificant while colonial history appears to be only marginally relevant. These conclusion hold after controlling for endogeneity with standard analytical tools.

She ends the article drawing conclusions regarding EU aid and how including an emphasis on social objective as well as economic ons might feed back into economic growth.

 Among the goals declared since the 1992 Maastricht Treaty, the EU emphasises international cooperation, with the objectives to foster and consolidate democracy, the rule of law, and respect for human rights and fundamental freedoms. Our findings suggest that pursuing these objectives may indeed complement economic aid to developing countries, since it implies that the EU should be interested in reducing of the likelihood of state fragility throughout the world, resulting in a positive impact on their aid absorption.

I'm pretty interested in this question of how social objectives facilitate economic growth. From my reading the evidence that they impede on economic growth is not as strong as some might suggest. I might come back to this in another post in a few days.