"The causes and consequences of demographic transition" by D. Canning
- In his paper Canning challenges different existing arguments about the causes and consequences of demographic transition. For example, in relation to urbanization transition, there is a view that excess labor pushes people from rural areas while industrialization and high wages pull people in the cities. But Canning notes that there are cases (e.g. Africa) where we see no evidence of such mechanism. Or how there are various possible causes for the fertility transition, such as decrease in child deaths, economic growth, contraception, shift in social norms. But Canning challenges these mechanisms as well. In this context Canning mentions difficulty to establish causality when drawing a link between some factor and demographic transition, in part because of possible reverse causation bias. I wonder if there actually exist appropriate instruments to infer causality in such big picture processes, such as demographic transitions in different countries? Or maybe all we can do is detect associations and build plausible theories based on that?
- I’m a little confused by the use of the words “exogenous” vs. “endogenous” processes. What do they mean in the context of this paper? Does exogenous mean that demographic transition happens on its own, with no particular factor causing it?
- Very interesting to read about the history of demography embedded in the political, economic, and social context of the USA. It looks like for many years demography was holding dearly to the idea of demographic transition, or fertility transition, and trying to adjust the world to its theoretical model. In the conclusion Szreter encourages a more empirical approach in demography. He encourages applying a wider range of analytical tools to study fertility change in particular communities. This sounds very similar to epidemiology. And I guess this could be the answer to my question about causality in demographic studies (addressed to Canning paper). If we want to study causal questions, we need to move from overarching big theories to empirical studies of specific populations and communities.