The panel structure of our data allows us to estimate the model thus specified as both a finite lagged model and a dynamic model. The results refute the hypothesis that agricultural investment contributes to structural transformation; however, they confirm the positive role of industrial investment and services lagged one and two periods. They also highlight the positive impact of GDP, the policy regime, and the negative impact of inflation.

The analysis highlighted the high sensitivity of structural transformation to investment in: (i) manufacturing industry, (ii) transport-warehousing and communications and (iii) insurance-real estate and business services. Exploring the implications of these results in terms of economic policy measures through a normative approach, a decentralised intervention in the form of public-private partnership appeared to be the best framework to influence individual investment decisions and to channel investment primarily to these branches