Comments

Caroline Hoxby: Christian A. L. Hilber and Christopher J. Mayer ask an important question: Do school finance laws affect where people live? In other words, if we pass a new school finance law in our state, should we expect a reshuffling of the population so that we will all have new neighbors? Will our state draw in certain people from other states or make certain people depart for other states? The answers to these questions are important because American states regularly revise their school finance laws and sometimes revise them in dramatic ways that might trigger sizable reshuffling. States' school finance formulas allocate $370 billion a year. To see how large an amount this is, compare Medicare at only $245 billion, all federal income support programs combined at only $330 billion, and national defense at a similar $376 billion. The amount of money affected by school finance laws is so large and the laws vary so widely, over time and among states, that school finance could easily be the government policy that affects where people live. This is not to say, of course, that other factors do not affect where people live. It is just that most of those other factors are not in policymakers' control.

Lurking in the background of this research is the implicit assertion that where people live matters. If neighbors do not have a causal effect on a person's outcomes, then it is unclear why we should care who is the neighbor of whom. Here it is important to distinguish between neighbors' correlations and neighbors' causal effects. We have a lot of evidence that there are correlations among neighbors; the evidence on neighbors' causal effects is limited. Indeed, some of the best research, based on the Moving to Opportunity experiment, suggests that neighbors' causal effects are very small.³⁸ The question of neighbor effects is beyond the scope of Hilber and Mayer's study or this comment, but it is important to flag the issue. We know that we care about whether someone is richer or poorer. It is not so obvious that we care about how people reshuffle themselves if there are no causal effects of sorting. After all, John Smith or Jane Doe is always someone's neighbor.

Hilber and Mayer test three hypotheses. First, do middle- and upper- income families prefer to live in states where school districts are relatively independent fiscally? The logic is that fiscally independent districts' resources tend to reflect the resources and tastes of the people who live in them. Therefore, an affluent, well-educated person living in a district with affluent, well-educated neighbors can probably obtain public schools that suit his desires. Conversely, if his district cannot spend money on its schools without giving the state a commensurate amount for other districts' schools, or if his district is forced by the state to spend the same amount as every other district, an affluent, well-educated person's local public schools may fall short of his desires. Indeed, he may abandon the expensive and frustrating task of trying to obtain public schools that suit him and may use private schools or move to another state instead.

The second hypothesis is that, in states that aggressively redistribute funds among their school districts, middle- and upper-income families and the elderly need not avoid living in districts with poor school-aged children as they might in states with financially independent districts. The logic behind this hypothesis is not so clear. If states that equalized school finance always "leveled up," then districts that served poor children would usually (though not always) look more attractive to middle- and upper-income families than they did before equalization.³⁹ However, more than half of school finance equalizations "level down" because their formulas contain parameters that implicitly penalize districts for higher spending.⁴⁰ After a leveling-down equalization such as the Serrano II equalization in California, well-off families have more, not less, reason [End Page 135...