The Jedi Path to Formal Economy: Untangling the Causal Effects of Development on Informality in Africa

Abstract

Informality poses significant hurdles, hindering efforts for development in Africa. While it is recognized that informality is negatively associated with development, it is less clear how development is causally impacting informality. In this regard, this paper aims to isolate the causal effect of development on informality in Africa, while effectively addressing endogeneity and heterogeneity issues of this nexus. Leveraging a dataset on measures of informality as a share of GDP over the period 1995-2018 for African countries, derived from the two-sector dynamic general equilibrium and the multiple indicators-multiple causes models, the empirical strategy for causal identification extended use of the instrumental variable approach to encompass a large range of econometric methodologies including, inter alia, sensitivity analyses, Bayesian inference, and causal machine learning. Controlling for tax burden, corruption, political stability, and population density, the findings suggest: (i) a persistent negative causal impact of development on the size of informality in Africa, confirming that as African countries progress in their development trajectories, levels of informality tend to decline; (ii) existence of heterogeneous effects across the African nations that significantly shrink the magnitude of the average causal impact of development on informality.

Keywords: Development, Informality, Instrumental variable, Double/debiased machine learning, Causal random forests, Bayesian inference

JEL Classification: C01, C11, C23, C21, C26, O10, O11, O55