New Evidence on Fungibility at the Aggregate Level
Lukasz Marek Marc
Abstract
Open-access reader
Lukasz Marek Marc
Abstract
Open-access reader
This study examines the fungibility of foreign aid and makes three contributions to the existing literature. Firstly, fungibility of aid at the aggregate level is reexamined on a richer panel dataset of 91 developing countries for 1980-2009, taking into account endogeneity of aid and autocorrelation in residuals. Results indicate that aid is strongly fungible: around 80 % is substituting rather than increasing government spending in the short run. There is also substantial heterogeneity in the sample, with aid being more fungible for countries with a low share of aid in GDP. Secondly, aid is disaggregated into bilateral and multilateral components. Despite substantial differences between both components, there are only very small indications that multilateral aid is less fungible than bilateral aid and estimates are volatile when aid is instrumented. Thirdly, this study attempts to distinguish between off- and on-budget aid at the aggregate level using the value of technical cooperation as a proxy for off-budget aid. While on-budget aid is strongly fungible, off-budget aid is non-fungible. In the long run, around 50 % of aid increases government expenditures, although results are not stable. High levels of fungibility (except for off-budget aid) suggest that resources spent on earmarking may be wasted and donors should rethink the way aid is distributed.
OpenAlex reports 7 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
This study examines the fungibility of foreign aid and makes three contributions to the existing literature. Firstly, fungibility of aid at the aggregate level is reexamined on a richer panel dataset of 91 developing countries for 1980-2009, taking into account endogeneity of aid and autocorrelation in residuals. Results indicate that aid is strongly fungible: around 80 % is substituting rather than increasing government spending in the short run. There is also substantial heterogeneity in the sample, with aid being more fungible for countries with a low share of aid in GDP. Secondly, aid is disaggregated into bilateral and multilateral components. Despite substantial differences between both components, there are only very small indications that multilateral aid is less fungible than bilateral aid and estimates are volatile when aid is instrumented. Thirdly, this study attempts to distinguish between off- and on-budget aid at the aggregate level using the value of technical cooperation as a proxy for off-budget aid. While on-budget aid is strongly fungible, off-budget aid is non-fungible. In the long run, around 50 % of aid increases government expenditures, although results are not stable. High levels of fungibility (except for off-budget aid) suggest that resources spent on earmarking may be wasted and donors should rethink the way aid is distributed.
Key concepts: Fungibility, Endogeneity, Aid effectiveness, Economics, Proxy (statistics), Sample (material), Developing country, Econometrics