These variations result in a kind of natural experiment, allowing us to study the effects of regionally marginal medicine, that is, the medicine consumed in high-spending regions but not consumed in low-spending regions. And the research is fairly consistent in showing that the extra medicine doesn’t help.
