Estimating the relationship between informal sector employment and formal sector employment in selected African countries.
Ntlhola, Mpho Anna.
MetadataShow full item record
Very little research evidence exists with respect to the informal sector in African countries. Although (mixed) theoretical evidence does exist that postulates a relationship between formal sector employment and informal sector employment, very little is understood about the exact nature of such a relationship. The research problem to be answered by this study thus constitutes two parts: Firstly, to estimate the relationship between informal sector employment and formal sector employment in selected African countries, and, secondly, to compare and contrast the estimated coefficients for the sample of countries with respect to statistical significance, sign and magnitude of such estimated coefficients. The study makes use of a fixed effects or least squares dummy variable (LSDV) panel data regression model, in double-log form, that comprises observations for informal sector employment, formal sector employment and exports (as a possible proxy for the "trade cycle‟ effect on informal sector employment). The sample of countries includes: South Africa; Kenya; Namibia; Zambia; Botswana and Mauritius, for the study period, 1998 – 2008. Theoretically, the expectation is a negative relationship between informal sector employment and formal sector employment as these are (plausibly) "substitute‟ activities in the labour market. However, there is mixed evidence to support/negate this hypothesis. Further, the expectation is a positive relationship between informal sector employment and exports. Including formal sector employment and exports as explanatory variables in a linear regression framework, poses a possible problem of strong collinearity between the explanatory variables (i.e. multicollinearity) as formal sector employment and exports are, generally, strong positively correlated. This study uses suitable ratio transformation to remedy this problem. The general findings of the study are that South Africa, Namibia and Mauritius had statistically significant levels (or average changes therein) of informal employment as a proportion of population not dependent on changes to formal employment as a proportion of population and exports. In Namibia and Zambia, informal employment as a proportion of population was statistically related to formal employment as a proportion of population, with negative sign, and "elasticity‟ greater than 1. In Namibia and Mauritius, informal employment as a proportion of population was statistically related to exports. Namibia had a positive sign and "elasticity‟ barely in excess of 1. Mauritius, however, had a negative sign and "elasticity‟ greater than 1.