Economic Growth in a Cross Section of Countries

In neoclassical growth models with diminishing returns to capital, a country's per capita growth rate tends to be inversely related to its initial level of income per person.This convergence hypothesis seems to be inconsistent with the cross-country evidence, which indicates that per capita growth rates for about 100 countries in the post-World War II period are uncorrelated with the starting level of per capita product.However, if one holds constant measures of initial human capital-measured by primary and secondary school-enrollment rates-there is evidence that countries with lower per capita product tend to grow faster.Countries with higher human capital also have lower fertility rates and higher ratios of physical investment to GDP.These results on growth, fertility, and investment are consistent with some recent theories of endogenous economic growth.With regard to government, the cross-country data indicate that government consumption is inversely related to growth, whereas public investment has little relation with growth.Average growth rates are positively related to political stability, which may capture the benefits of secure property rights.There is also some indication that distortions of investment-goods prices are adverse for growth.Finally, the analysis leaves unexplained a good deal of the relatively weak growth performances of countries in sub-Saharan Africa and Latin America.