The need to shift from fossil fuels to renewable energy systems is now broadly accepted, and our understanding of how different policy mechanisms can support this process is growing.

To date, much of this research has focused on developed countries, but the need for this shift is increasingly being recognised as important in countries at all levels of development. The circumstances of these countries are different, however, suggesting that approaches to policy identification and evaluation may also need to be adapted. These differences are found in three areas: first, many developing country governments face severe budget constraints and competing calls on public resources to address poverty. The finances available to support policy mechanisms are therefore more limited than in many developed countries. Second, given the immaturity of financial systems, the most important obstacles to investment in renewable energy may be unrelated to the specifics of these investments, but reflect more general problems. Third, while issues of political economy are important for power sectors everywhere, they can be particularly pronounced in developing country settings.

Before we can assess the potential effectiveness of different policy mechanisms to support renewable energy investment in developing countries, therefore, we need first to understand what the most important constraints to these investments are. Given the very large number of potential constraints, an approach is needed to narrow this set systematically, and identify those constraints which are most important, or ‘binding’. The research presented in this paper adapts and extends the ‘growth diagnostics’ approach developed by Hausmann, Rodrik and Velasco (2004) for this purpose, and applies this to renewable energy investment in two developing countries: Ghana and Kenya. The resulting ‘green investment diagnostics’ thus complements existing work on policy evaluation in developed countries, offering a diagnostic tool specifically designed for developing country settings that could be applied alongside these mechanisms.