Regional variation in 1917 health care expenditures

In 1917, the Bureau of Labor Statistics surveyed 11,946 white households nationwide, recording demographic variables and yearly expenses on physicians, medicines, nurses, and hospitals. There was significant variation in these medical expenses among the nine census bureau regions. Using a multivariate analysis, we demonstrated that some of the variation could be explained by household variables, such as household income and size, some of the variability could be explained by the availability of physicians and hospital beds, and some could be explained by interactions between the different types of health care. However, after accounting for these possible explanatory variables, significant regional variation remained. We conclude that regional variation in health care expenditures is not new, and that studying how regional variation has changed during the twentieth century is likely to help explain why that variation exists today.