A fairly heavy-duty academic paper by Matt Lowe, January 27, 2014
Abstract
I explore the impact of railway privatization in Sub-Saharan Africa using two spatial
panel datasets – light density measured by satellites at night and geo-coded Afrobarometer
survey responses. The lights data comprises a panel of 0.3x0.3 decimal degrees grid cells
over the period 1992-2010, covering thirty-two countries. I ï¬nd little evidence of an impact
on lights – though the average effect masks considerable heterogeneity across countries. In
contrast, I ï¬nd robust evidence in the survey data that privatization leads to a worsening of
perceived living standards and opinions toward government, with no impact on objective
measures of poverty or employment. Privatization is a controversial, political process –
taken together the results suggest that whilst no objective harm was done, individuals felt
worse off. More generally, the results show the great difficulty of successful rail reform in
Africa.
