Understanding productivity dispersion: Evidence from a new survey of manufacturing firms in Uganda

Differences in income per capita across countries are mostly accounted for by differences in aggregate total factor productivity (Hall and Jones, 1999; Caselli, 2005). These differences are also present at the firm level, as firms in less developed countries have, on average, much lower labour productivity. However, there is substantial productivity dispersion within developing countries. Indeed, the low average productivity in developing countries is mostly driven by a thick left tail of small and unproductive firms, while relatively productive firms exist even in the poorest countries (Tybout 2000; Bloom et al 2010; Hsieh and Klenow 2009; Hsieh and Olken 2014).

The presence of some relatively high productivity firms even in the least developed countries is encouraging. While productivity differences across countries are in principle very difficult to address since they are possibly due to different institutions or rule of law, productivity differences within countries might be driven, in principle, by factors more amenable to external intervention. However, to design such interventions, it is necessary to understand what explains within-country productivity dispersion.

This project aims to fill this evidence gap by designing and implementing a representative survey of small, medium and large firms and their employees in urban and semi-urban Uganda. The researchers will collect granular information that allows them to precisely measure productivity differences across firms, and most importantly, to understand how high productivity and low productivity firms differ.

The results from the survey aim to generate new key facts about the determinants of productivity dispersion in Uganda, a large developing country in one of the least studied areas of the planet, Sub-Saharan Africa.