Additional evidence on Madagascar's tourism-led growth hypothesis

We evaluate the validity of the tourism-led growth (TLG) hypothesis for Madagascar through an examination of the asymmetric relationships among economic growth, tourism activity, and financial development over the period 2007Q1-2024Q4. Nonlinear Autoregressive Distributed Lag (NARDL) modeling reveals a statistically significant long-run asymmetric effect, wherein negative shocks to tourist arrivals exert a stronger adverse impact on economic growth than the expansionary effect of equivalent increases. With respect to causal dynamics, bivariate symmetric Granger causality testing first uncovers empirical support for the TLG hypothesis. Subsequent short-run asymmetric causality then supports the conservation hypothesis, wherein negative economic growth shocks trigger positive tourist inflows. This dynamic suggests that tourism functions as a counter-cyclical shock absorber during domestic downturns. Consequently, our analysis highlights the need to pivot strategy away from mere tourist volume toward high-yield retention and investment in a resilient tourism sector.