This thesis consists in a series of theoretical essays on the macroeconomics of health and offers a reflexion on the recent history of economic thought. Its main contribution is to show that population health is not a simple by-product of economic development. The first chapter investigates contemporary differences in mortality between the United States and European countries. It provides a rationale to the American Puzzle, the fact that life expectancy is lower in the United States despite a greater share of GDP devoted to health care. By considering the deleterious effect of long hours of work on health, it shows how differences in working time can account for both the poorer health of Americans and their higher health expenditure and emphasizes a trade-off between health status and consumption. The second chapter takes a more historical perspective and offers a theory for the episode of deskilling experienced during the British Industrial Revolution. In particular, it considers the political incentives of the capitalist class to undertake public health investments that improve workers' longevity and gives them incentives to acquire education. The direction of technological progress is crucial for such incentives, and its unskill-biased nature during the 19th century may explain why life expectancy did not take off earlier for the working class. The third chapter is a reflection on the history of modern macroeconomics. It documents more than thirty years of controversy around the value of the labor supply elasticity, the aggregation of individual behaviors and the use of a representative agent, and questions the standards narrative of the discipline.