Development Blossoming: Industrial Policy and Flower Exports in Ethiopia

As private sector development re-gains centre stage on the development agenda, there is an increasing need to understand the design of optimal industrial policies. This paper presents a preliminary investigation of the effects of the industrial policy that has spurred the development of the Ethiopian horticulture sector. In contrast to existing empirical analysis of entry – which condition on entry – the analysis rests on an originally assembled dataset that contains information on all the approved projects in the industry. We find that less than 20% of all projects approved eventually lead to an operational project. While the entry process has been hindered by the global crisis, we also find large difference in behavior, decisions and outcomes between domestic projects and projects involving foreign investors. Domestic investors are less likely to fill the application forms and, conditional on filling the forms, are less likely to bring the project to the implementation phase and to become operational, even conditional on implementation. Furthermore, domestic investors respond to the industrial policy by increasing the capital-labor ratio of the proposed projects – but not of realized projects. These facts are not driven by cohort effects, activity type or region of investment. Differential access to credit and misallocation of investment projects (are likely to) explain part of these differences. While the industrial policy might have relaxed financing constraints for domestic investors, it is possible that it has failed to maximize its potential due to imperfect targeting. The results show how crucial it is for industrial policies to be designed with “incentive compatibility” as a core objective. Further theoretical analysis is required to shed light on the design of optimal industrial policies in the presence of credit constraints.