Malthus irrelevant?

We estimate and make population forecasts with Foley’s (2000) model in three different ways. The population forecasts for high, middle and low-income countries are quite good and suggest that the omitted variable bias from its simplicity is small. Estimation of the model as a system shows that indeed Malthusian behaviour - defined as increasing population growth through increasing per capita income - cannot be found for any of the income groups of the Worldbank classification nor for Sub-Saharan Africa, and also not for countries with per capita income below $1200 in a panel estimate. For world aggregate data and for the low-income countries we find increasing returns to scale, but for the other groups decreasing returns (outweighed by a positive time trend except for Sub-Saharan Africa and the u1200 group). For the panel of countries with income below $1200, per capita income is stagnant for the period 1970-2002 in spite of the positive growth rates of the period 1991-2002. The time trend is as strong as the population growth in connection with decreasing returns to scale. Together with the absence of Malthusian behaviour this seems to suggest a strong role for the population growth problem as seen by David Ricardo.