Business registration from inception and employment indicators using causal machine learning

Employing regression analysis and the honest causal tree (HCT), this paper provides the first empirical evaluation of the average treatment effects and the heterogeneous treatment effects of business registration from inception on firm employment, employment change, and employment growth. The theoretical framework formally demonstrates that formalization allows firms better access to capital, intellectual property protection, and collaboration with other firms and government agencies that may boost their organizational efficiency, leading to expansion and the need to employ more workers. The findings suggest that all things equal, formal firms that registered from the start have, on average, approximately 12 to 18 more employment and employment change than their peers that did not formalize from the beginning. The lower bound is robust to model dependence using the doubly robust nonparametric preprocessing matching. However, the decision to start formally from inception is estimated to reduce employment growth by 54 percentage points. This implies that factors that ensure organizational efficiency should be made available to Nigerian enterprises across industries to ensure they reap the fruits of formalization and stimulate employment growth in Nigeria. These results are robust to omitted variable bias and misspecification bias. The HCT discovered interesting heterogeneity by showing that formal firms that started registered and did not face obstacles with the labor regulations have 37 more employees than their counterparts that delayed formalization and did not encounter the impediments. The policy recommendation is to adopt programs to incentivize firms to start registered, provide support services to registered firms, and reform labor regulations to generate employment for the teeming Nigerian youth population.