The effectiveness of industrial policy in developing countries: causal evidence from Ethiopian manufacturing firms

Prioritising the growth of particular sectors or regions is often part of a low-income country's growth strategy. We study a prototypical example of such policies in Ethiopia, exploiting geographic and sectoral variation in the form and scale of the policy for identification. Using product-level data on Ethiopian manufacturing firms, we show that the policy was unsuccessful: in the best case scenario its benefits were around one-tenth of its cost. Subsidised loans did not improve productivity, leading only to an increase in fungible assets not machinery. Tax-breaks improved productivity but reduced firms' capital levels. Further results suggest that these were both the consequence of volatility and the lack of effective bankruptcy protection. There are two key policy implications of our findings. Firstly, we highlight that ineffective industrial policies can be extremely expensive and thus may impede rather than promote countries' pursuit of their development objectives, given that 'blunt' industrial policies like the one we study are common. Second, our results suggest that policies that rely on improving access to capital will be more successful in a stable economic environment with effective bankruptcy protection and that access to capital is not necessarily the key constraint to improving productivity in many of the firms we study.