Abstract:
We present a simple model explaining corruption on geography and climate conditions. We test the model's validity in a cross-section of 115 countries. Controlling for all other corruption's determinants we find evidence supporting the model's predictions. Corruption increases with temperature and declines with precipitation and non-cultivatable land. Corruption also declines with per capita GDP, democracy, median age and British colonial heritage; and increases with natural resources, bureaucracy and communist past. Finally, corruption declines with the ratio of internet users to total population. This new finding is interpreted as capturing the beneficial interaction of economic development, human capital/education and independent news.