The Asymmetric Associations of Foreign Aid and Remittances with Real GDP per Capita in Developing Countries: Evidence from a CS-NARDL Approach

This study examines the asymmetric associations of foreign aid and remittances with real GDP per capita in 68 developing countries, comprising 17 low-income countries (LICs) and 51 middle-income countries (MICs), over 1980–2023. It employs a Cross-Sectionally Augmented Nonlinear Autoregressive Distributed Lag (CS-NARDL) approach, controlling for financial development and foreign direct investment. The methodology captures asymmetric short- and long-run responses to positive and negative changes in foreign aid and remittances while accounting for cross-sectional dependence. The findings show positive coefficients on positive changes in both inflows, while the reported negative coefficients on signed decline components imply a positive predicted income response to declines in these inflows. The estimated aid associations vary more substantially across income groups, with LICs showing greater sensitivity to changes in aid flows. In the short-run, the effects are generally weaker and more heterogeneous, with growth responses to positive changes in foreign aid and remittances varying across model specifications. The expansion and contraction specifications yield different associations; their signs do not establish that external inflows cushion downturns. A supplementary Dynamic Common Correlated Effects (DCCE) specification reports broadly similar coefficient patterns. Overall, the findings underscore the importance of accounting for asymmetric responses when examining the relationship between foreign aid, remittances, and economic growth in developing countries. The findings warrant cautious policy interpretation because several estimated associations differ from the proposed economic mechanisms.