This paper uses firm level survey panel data to estimate parameters of export propensity and intensity in Kenyan manufacturing. The effects of unobservable factors that would otherwise bias the estimated parameters are removed using a control function regression procedure. The key finding of the study is that export propensity and intensity are strongly responsive to total factor productivity. In particularly a 10% increase in total factor productivity increases export propensity by 54%, but export intensity rises less steeply by 18%. We also find that ownership structure of the firm and unobserved factors specific to firms strongly influence exports. Taken together, the estimation results provide insights into the policies
needed to promote entry and stay of firms in export markets. The findings suggest that policy measures to improve export performance of Kenyan firms should focus on improving total factor productivity, encouraging foreign direct investment and stimulating modernization of manufacturing capital.
Keywords: Exports, manufacturing firms, total factor productivity, control function approach, Kenya.