The Long-Run Impact of Foreign Aid in 36 African Countries: Insights from Multivariate Time Series Analysis
Research output: Working paper › Research
Standard
The Long-Run Impact of Foreign Aid in 36 African Countries : Insights from Multivariate Time Series Analysis. / Juselius, Katarina; Møller, Niels Framroze; Tarp, Finn.
Helsinki : UNU-WIDER, 2011.Research output: Working paper › Research
Harvard
APA
Vancouver
Author
Bibtex
}
RIS
TY - UNPB
T1 - The Long-Run Impact of Foreign Aid in 36 African Countries
T2 - Insights from Multivariate Time Series Analysis
AU - Juselius, Katarina
AU - Møller, Niels Framroze
AU - Tarp, Finn
N1 - JEL classification: C32, F35, O11
PY - 2011/9
Y1 - 2011/9
N2 - Studies of aid effectiveness abound in the literature, often with opposing conclusions. Since most time-series studies use data from the exact same publicly available data bases, our claim here is that such differences in results must be due to the use of different econometric models and methods. To investigate this we perform a comprehensive study of the long-run effect of foreign aid (ODA) on a set of key macroeconomic variables in 36 sub-Saharan African countries from mid-1960s to 2007. We use a well-specified (Cointegrated) VAR (CVAR) model as our statistical benchmark. It represents a much-needed general-to-specific approach which can provide broad confidence intervals within which empirically relevant claims should fall. Based on stringent statistical testing, our results provide broad support for a positive long-run impact of ODA flows on the macroeconomy. For example, we find a positive effect of ODA on investment in 33 of the 36 included countries, but hardly any evidence supporting the view that aid has been harmful. From a methodological point of view our study documents the importance of transparency in results reporting in particular when the statistical null does not correspond to a natural economic null hypothesis. Our study identifies three reasons for econometrically unsatisfactory results in the literature: failure to adequately account for unit roots and breaks; imposing seemingly innocuous but invalid data transformations; and imposing aid endogeneity/exogeneity without testing.
AB - Studies of aid effectiveness abound in the literature, often with opposing conclusions. Since most time-series studies use data from the exact same publicly available data bases, our claim here is that such differences in results must be due to the use of different econometric models and methods. To investigate this we perform a comprehensive study of the long-run effect of foreign aid (ODA) on a set of key macroeconomic variables in 36 sub-Saharan African countries from mid-1960s to 2007. We use a well-specified (Cointegrated) VAR (CVAR) model as our statistical benchmark. It represents a much-needed general-to-specific approach which can provide broad confidence intervals within which empirically relevant claims should fall. Based on stringent statistical testing, our results provide broad support for a positive long-run impact of ODA flows on the macroeconomy. For example, we find a positive effect of ODA on investment in 33 of the 36 included countries, but hardly any evidence supporting the view that aid has been harmful. From a methodological point of view our study documents the importance of transparency in results reporting in particular when the statistical null does not correspond to a natural economic null hypothesis. Our study identifies three reasons for econometrically unsatisfactory results in the literature: failure to adequately account for unit roots and breaks; imposing seemingly innocuous but invalid data transformations; and imposing aid endogeneity/exogeneity without testing.
KW - Faculty of Social Sciences
M3 - Working paper
SN - 978-92-9230-418-8
BT - The Long-Run Impact of Foreign Aid in 36 African Countries
PB - UNU-WIDER
CY - Helsinki
ER -
ID: 35380486