We examine a panel data set for the US states over the period 1950-1990 and use it to assess the effects of growth in personal income and number of students on expenditure on public primary and secondary education. Our analysis suggests that the share of personal income devoted to education is roughly constant, implying that per student education expenditures grow at roughly the same rate as personal income per student. We also find evidence that additional factors accounted for an increase in education expenditures over the period 1950-1970.