Answers are grounded in this paper and cite their sources.
The share of elderly households is expected to grow strongly in the United States during the next decades. Conventional wisdom, supported by recent research, suggests that spending on local public schools may decline in real terms as a growing percentage of elderly voters becomes more influential.1 However, recent research has found that elderly voters are willing to support education spending at the local level but are often opposed to additional state spending on schools.2 Hilber and Mayer show that house prices may serve as a mechanism to encourage the elderly to support local spending.3 They show that the percentage of elderly residents is positively associated with additional school spending in certain places where higher school spending raises local house prices. However, the elderly are negatively associated with school spending in places where land is freely available and changes in school spending are likely to have little effect on house prices. These findings imply that the elderly may have a negative effect on public school expenditure levels in locations where state policies (as opposed to local choices) are crucial in determining [End Page 107] local spending on schools or in places where school spending is not strongly tied to house prices. A second mechanism through which an increasing elderly population might not lead to drastic cuts in school spending is Tiebout sorting. For example, well-to-do households with children might choose to live in states that encourage local control over school spending. With local control, elderly households can live in communities that spend little on public schools, while households with children can live in places that have higher levels of school spending. In states with little local control of school spending, such sorting might be less effective in allowing households with children to choose communities with high levels of school spending. Thus states with less local control may be less attractive for middle- and high-income households with children. Alternatively, low-income households might choose to live in states with state funding of schools if such funding leads to additional redistribution. Similarly, elderly households would likely prefer to live in places that spend little on public schools. Of course, much of the recent evidence on Tiebout sorting suggests that Tiebout sorting may not be as powerful a force with the strong rise in two-career couples and the dispersion of employment in rings at the periphery of metro areas. In this context, it is crucial to develop a better understanding of what factors affect the location decisions of households with children and the elderly. We use recently enhanced U.S. Census microdata on states and census-designated places from the four censuses between 1970 and 2000 to examine changes in the concentration of households with children and the elderly. In the aggregate, the concentration of both groups of households across states has been little changed over the 1970 to 2000 time period. However, these aggregate statistics mask the fact that low-income households have become increasingly concentrated in a small number of states. A very different pattern emerges when one looks at household concentrations within each state. After weighting by the number of households, the data show a strong deconcentration of most household groups across places within each state. The extent of deconcentration is strongest for the elderly and nonwhite populations but is also strong for households with children and poor households with children (that is, households with children that have an income below 100 percent of the [End Page 108] poverty level). Such facts are consistent with the diminished importance of Tiebout sorting at the local level. Next, we consider how "reforms" in state educational spending policies have affected household location choices. State educational spending policies determine the constraints that local governments...
No takes yet. Share an insight, caveat, or question.
Hilber et al. (2004) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: