This paper reviews the state of the art in firm-level Total Factor Productivity (TFP) estimation by employing an unbalanced panel of 4,501 Senegalese firms in the Construction and Trade Services industries over the period 2008–2018. The three semi-parametric models considered are Olley and Pakes (1996)’s approach (OP hereafter), Levisohn and Petrin (2003)’s algorithm (LP hereafter) and Ackerberg et al. (2015)’s approach (ACF hereafter). Each of these estimators assumes different underlying properties for some inputs, which potentially affect the TFP measurements. We also analyze how corporate taxation, the legal form and the regional competition affect firm level productivity. Our findings are summarized as: (i) the ACF produces insignificant capital and labor coefficients for both Construction and trade services industries, while LP produces the best alternative within semi-parametric models and it is better than OP; (ii) tax pressure adversely affects TFP and the adverse effect is found to be severe in the groups of trades services firms, suggesting that the distortive nature of corporate tax affects disproportionately firms across the sectors. The results also suggest that the local economic environment plays a role on firm performance in the trade industry. The extent of local competition negatively affects the firm performances for both industries. The results indicate that medium-sized and large-sized firms also demonstrate an advantage over small ones. The effect of age of firm is positive but diminishes as firms become older.
