Poverty Traps and Climate Change

We use a demo-economic model to examine the question of whether climate change could widen or deepen poverty traps. The model includes two crucial mechanisms. Parents are risk averse when deciding how many children to have; fertility is high when infant survival is low. High fertility spreads scarce household resources thin, resulting in children being poorly educated. At the macro level, technological progress is slow because of decreasing returns to scale in agriculture. With high population growth and slow technological progress, the economy stagnates. If, on the other hand, infant survival is high, then fertility is low, education is high, and the economy grows exponentially. Diarrhea and malaria are among the leading causes of infant mortality; both are sensitive to weather and climate. There may thus be a climate-related poverty trap where climate change increases disease burdens that reinforce poverty. We estimate finite-mixture models of per capita income, fertility, and mortality at the national scale. As predicted by the model, the observations are bi-modal. Temperature has statistically significant effects: hotter countries are more likely to be classified as poor; hotter countries are more likely to be classified as high mortality; and the number of children per woman in high fertility societies increases with temperature. We then use the model to simulate a number of different futures, focusing on the question whether climate change may widen and deepen the health/fertility poverty trap. The results suggest that this is unlikely for reasonable parameter choices. Climate change may have a substantial effect on specific causes of infant mortality, but the effect on total infant mortality is more muted. More importantly, the model is driven by infant survival, and climate change has a much smaller proportional effect on survival than on mortality. Furthermore, climate change will be relatively small over the next few decades. In the medium term, the impact of climate change is therefore dwarfed by other factors (health and education in this model). In the long term, climate change is more important, but the long term is primarily shaped by the medium term.