How Do African Firms Respond to Unreliable Power? Exploring Firm Heterogeneity Using K-Means Clustering - Working Paper 493

In this paper, we apply a novel analytical technique—k-means clustering—to understand the relationship between the growth of firms and the availability of power in sub-Saharan Africa. We develop a classification of firms and show how firm clusters are distributed across industries and countries. Our analysis reveals a surprising degree of within-country heterogeneity in the experience of firms. While previous studies have found a positive relationship between the reliability of power and firm growth, we find that such a clear relationship seems not to prevail. In other words, some firms are able to cope with an unreliable supply of power while many others do not. This may be because firms self-select into industries which promise high returns despite anticipated power problems. Further research on the conditions determining entry would provide insight into why some firms succeed in a poor business environment, while others are unable to thrive.