Median voter theory says that to predict changes in collective decision making, one should focus on the median voter. Between 1973 and 1993, average incomes in the United States increased, whereas the income of the median family remained roughly stagnant. (Since then, the median income has increased slightly, but not enough to erase the increased gap of the previous two decades.) How might an economist focusing on median voter behavior and an economist focusing on average incomes differ in their predictions concerning changes in the level and composition of public expenditures?