Is there an investment motive behind remittances? : evidence from panel cointegration

Remittance flows have become a vital source of foreign exchange for many developing countries. As a result, the issue of whether they act as complements or substitutes for domestic investment remains an important avenue of research. We know that remittances can act as compensatory transfers, in which case altruistic motives may dominate. We also know that they can act as standard capital flows, where self-interest/investment motivates may dominate. Hence, the motives behind remittance flows can have a direct bearing on how they influence domestic capital formation. In addition, the short-run relationship between domestic investment and remittances may be different from their long-run relationship. In light of these considerations, this paper reexamines whether migrant remittances "crowd in" or "crowd out" investment in developing countries, using a sample of 47 developing and emerging economies. The paper employs recently developed panel cointegration techniques given that these can overcome a number of important issues. First, we explicitly account for cross-sectional dependence, outliers as well as cross-sectional heterogeneity. Second, since our variables of interest may be influenced by various factors emanating from, for example, domestic policy changes or global economic trends, we account for structural breaks and regime shifts. Third, the approaches we employ are robust to endogeneity and many forms of omitted variable bias. Fourth, we examine both the long-run as well as the short-run relationship between remittance flows and domestic investment, employing panel error correction model to uncover the short-run dynamics. Finally, we conduct a panel Granger causality analysis to establish whether these relationships are indeed of a causal nature. The results of the paper show that remittances form a long-run equilibrium relation with domestic investment. The results of the panel vector error correction model reveal the absence of a short-run relationship but the presence of a long-run bidirectional link between remittances and investment. Thus, remittances drive investment while investment itself causes more remittances, suggesting that remittances are not only driven by altruistic motives but also investment motives. This long-run (causal) two-way relationship is robust to a battery of sensitivity analyses. However, when the sample is disaggregated into regions, the results of the Asian sub-sample are statistically insignificant. We suspect that this is due to the low number of observations from that region. An important policy implication emanating from this study is that developing countries should improve the effectiveness of remittance inflows given that these can augment the rate of capital accumulation