Foreign sectoral aid fungibility, growth and poverty reduction

If development assistance targeted at specific sectors is not used as intended, aid is said to be fungible. While fungible aid is in general perceived as being less effective than aid used as specified, this has not been formally tested. This paper attempts at filling this gap and hence, tries to assess to what extent fungibility is something donors should be concerned about. Country‐specific estimates of fungibility are obtained for 57 aid‐recipient countries, suggesting that sectoral aid is indeed fungible on average. These estimates are then incorporated into an empirical model of aid and economic growth. I do not find any evidence of non‐fungible sectoral aid working better than fungible aid. Then, I focus on sectoral aid to ‘pro‐poor’ government expenditure sectors to assess the effect on infant mortality. While the results indicate that non‐fungible aid is welfare improving, this is not robust to small changes in the specification. My results suggest that the concept of fungibility may be too narrow and should possibly not be the most central concern when aid is debated or given.