Researchers use colonial origin and malaria/settler conditions from ~300 years ago as a statistical instrument for institutions, because today's income level cannot retroactively cause who colonized you centuries ago, which lets them argue institutions causally affect economic outcomes rather than merely correlating.