Bas Hollander on the role of a country’s institutional quality on climate-induced migration
About
Climate change, particularly global warming, poses serious challenges for policymakers. One of the most pressing is its influence on migration, which can in turn exacerbate humanitarian crises and regional instability. The migration response to climate change is shaped by contextual factors that influence a nation’s adaptive capacity. One such factor is institutional quality, which may enhance resilience to climate change by improving the implementation and effectiveness of adaptation policies. Using data from 86 middle-income countries between 2002 and 2019, I examine whether a country’s institutional quality moderates the relationship between rising temperatures and emigration.
Results
Fixed effects and long-difference regressions provide limited evidence that stronger institutions may mitigate emigration responses to rising temperatures in lower-middle-income countries. In particular, Government Effectiveness and Rule of Law appear to play a mitigating role. However, this moderating role appears nonlinear: it weakens under extreme temperatures. The climate-migration nexus appears highly context-dependent.
Unlike lower-middle-income countries, upper-middle-income countries show no significant migration response to climate or institutional factors, possibly due to stronger existing institutions and reduced dependence on climate-sensitive livelihoods. Despite insufficient consistent statistical significance and robustness across model specifications, the findings tentatively suggest that institutional quality may enhance climate resilience
Figure 1
Temperature change is measured using annual averages, depicted in figure 1 on the right. As discussed in the section on future research below, the use of monthly data could better capture seasonal climate variation and potentially yield different estimates of migration responses.
Figure 2
Figure 2 below shows that results likely depend on which specific institutional indicator is used, since differences exist in their effects and developments over time. This approach allows for a more nuanced understanding and enables more targeted and actionable policy recommendations. For future research, it remains important to continue including individual governance indicators to uncover potential mechanisms and examine how institutional capacity shapes climate-related migration patterns.
“The strength of a nation lies not in its wealth or power, but in the quality of its institutions to safeguard the well-being of its people.”
Figure 3
Figure 3 illustrates the non-linear moderating role of institutions by plotting predicted migration responses* across the observed temperature range, disaggregated by institutional quality. The panel shows that countries with higher institutional quality (defined as IQ above the median) exhibit a larger reduction in emigration (that is, a higher predicted net migration) as temperature increases. The relationship follows a downward-opening parabola, suggesting that the moderating effect of institutional quality weakens at higher temperature levels. This attenuation is evident around the inflection point, where the blue curve begins to flatten. As discussed earlier, such non-linear responses across temperature levels are consistent with theoretical expectations (Hsiang, 2010). However, the confidence intervals of the two curves overlap substantially. This implies that the estimated migration responses to temperature variation do not differ statistically between countries with low and high institutional quality.
* : An increase in net migration implies a decrease in emigration.
Future research
Above-mentioned insights underscore the need for more targeted, context-specific studies on climate-induced migration to help shape more effective and tailored policy actions. Ideally, future studies should use monthly temperature and precipitation data to better capture seasonal climate variations that significantly influence migration dynamics (Cattaneo & Peri, 2015; Huynh & Hoang, 2024), as annual data may overlook such effects. Accordingly, a graph depicting surface temperature change based on higher-frequency data could reveal different (more severe) patterns than those shown using annual averages.
Extending the sampling period would also align more closely with the gradual nature of climate change and enhance the validity of long-difference regression results. In addition, future research should strive to obtain reliable emigration-specific data rather than relying on net migration as a proxy, as net migration masks the distinct drivers of outflows and inflows. This is particularly relevant in low- and middle-income countries where data quality is often limited.
Whilst making these improvements, it remains important to continue including individual governance indicators for the listed reasons alongside Figure 2.