Policy almost always concerns defined populations, e.g., of a city, a province or a country. The first step is to describe the population and understand its demographic trends. Descriptive epidemiology can then measure the health of this population, identifying trends and patterns in health, and assess the population’s health risks and health needs (a topic that our discipline has thought about remarkably little, with the notable exception of the excellent Epidemiologically-based Needs Assessment Reviews from the United Kingdom). This will help to identify health problems and population groups that are priority targets for policy development. It is particularly important for a civilized country to identify and quantify inequalities in health. Analytical epidemiology can determine the causes of health problems, identifying both individual-level and population-level factors; traditionally, we have done fairly well on the former, and almost ignored the latter, but this is changing. A particularly hot and important topic at present is the effect of income inequalities on health (although this may be of more interest outside Japan than within it, since the excellent health of your population is frequently attributed to the relatively low level of income inequality in your country).