Remittances, real exchange rates and Dutch disease: evidence from Albania, 1995–2023
Adela KarapiciPurpose
This paper examines whether migrant remittances have generated Dutch disease in Albania. It tests whether remittance inflows appreciate the real effective exchange rate (REER) (the price channel), whether they shift labour towards non-tradable (the resource-movement channel) and whether these effects persist.
Design/methodology/approach
Using annual data for 1995–2023 from World Bank and IMF sources, the study estimates OLS with Newey–West standard errors, a dynamic autoregressive model, a long-run levels regression and an error-correction model, supported by unit-root, Zivot–Andrews, ARDL bounds and Engle–Granger tests, influence diagnostics, subsample analysis and a wild bootstrap.
Findings
A one-percentage-point rise in remittances/GDP is associated with long-run REER appreciation of roughly 2–4%, depending on the deterministic specification. The estimate survives controls for the pyramid-crisis years, single-observation deletion and alternative end dates, but loses significance once a level shift is allowed at the 2000 change in REER series construction. No resource-movement effect is detected, although the aggregate employment proxy has limited power. Deviations are corrected rapidly (about 86% a year).
Research limitations/implications
With 29 annual observations, strong remittance–trend collinearity and a REER series spliced across data vintages, the price channel remains conditional until a single-vintage REER series is available.
Practical implications
Policy should focus on channelling remittances towards tradable-sector investment and deepening financial intermediation rather than relying on the exchange-rate regime.
Originality/value
The paper extends the Albanian evidence through the post-COVID period to 2023 and, to my knowledge, is the first to test the price and resource-movement channels jointly for Albania.