Last week we published a new paper, Can Africa Be A Manufacturing Destination?, that highlights the persistence of high labor costs in many countries in sub-Saharan Africa. This led to a lively debate on Twitter, initiated by Chris Blattman at the University of Chicago.
Several bloggers and academics responded to Blattman’s tweet, sharing their views as to why labor costs are so high. Tom Westland at the University of Cambridge writes:
To which Blattman responds:
Dani Rodrik also weighs in:
Johan Fourie at Stellenbosch University has a different view:
To which economist Grieve Chelwa responds:
Our colleague Paddy Carter chimes in:
Pseudoerasmus, a prolific blogger, reminds everyone that unit labor costs could look very different from wages or total labor costs. He says:
… and concludes in response to data posted by Asif Dowla:
Howard French, an expert on China at Columbia University, comments on the story in Quartz Africa on our paper:
And Neil Gregory from the International Finance Corporation wrote in an email about gender disparities:
Bangladesh, the key comparator, is notable for very high female participation in manufacturing. If there are cultural/economic barriers to such participation in Africa, that would drive up wages as women may not be available for hire to the same extent. You could look at differences in female labor force participation to estimate this effect.
We have tried to look at some of these hypotheses in the context of Purchasing-Power-Parity. Using the 2011 ICP data, Gelb and Diofasi confirm that African countries tend to be more costly than countries with similar incomes in other parts of the world. Aid inflows and a negative current account balance are correlated with higher price levels (the latter less strongly), while FDI and remittances are not. But even though aid dependence is associated with a higher PPP price level, this does not seem to account for the “Africa effect.”
Another possibility is that the Africa effect reflects widespread underestimation of income levels. The recent upwards revisions of GDP in several countries, including Ghana (60%), Nigeria (89%), Kenya (25%), and Zambia (25%) suggest that this could be the case for many countries on the continent. Taking a linear approximation to the relationship between the logs of income and PPP, an illustrative 30 percent increase in income per head would correspond to a 5.3 percent increase in the PPP price variable. This could account for as much as a third of the Africa effect.
Turning to structural features, another possible cause that we considered was low agricultural productivity and weak domestic supply chains. Food is relatively costly in the PPP bundle in poorer countries, and particularly so in many African countries. On average, the food price level is about 50 percent higher in low- and middle income countries than the overall price level in the economy. As shown in the recent ACET report, agricultural imports have been rising with urbanization. Much urban food is imported; this makes it harder to support a low-cost labor platform for manufacturing.
And the point on gender may be very relevant. The Asian countries have been excellent in drawing on the young female labor force. Among other things this requires a degree of operating and transit security, especially for shift work.
We would like to thank all of the researchers who shared their views on this topic! We will be continuing our work in this area and we hope that others will also try to answer the question of why industrial labor costs are so high in many countries in Africa.