A Badger Institute Policy Brief: School Choice and Property Taxes

Executive summary

The estimate by the Department of Public Instruction of $577.3 million in potential added costs to property tax payers from opening Wisconsin’s school choice programs to families of all income levels is deeply flawed.

It rests on an assumption that all children now attending private schools with tuition paid by their parents would switch to using state aid via the school choice program. This ignores the reality that choice schools, which can determine how many seats they offer through the program, usually can’t afford to have no tuition-paying students – because the choice grant is too far below the cost of education. Instead, the ratio of choice to tuition-paying students is likely to change more slowly.

Since the bill does not include an extension of the fixed window to apply for a choice seat next year, families made eligible by the bill could switch their children only if the bill is signed and the DPI revamps its application process in the next six weeks. It is more likely that no additional children will attend next year.

The estimate’s central claim, that such a shift would burden property tax payers, points to a possible reform. The program’s complicated financing shifts the burden in schools outside of Milwaukee to the property tax in districts that choice students live in. It would be simpler and, from a taxing standpoint, fairer to fund school choice statewide in a uniform way that mirrors how most of state government is funded – through general-purpose revenue raised primarily through the state taxes rather than local property taxes.


By Jim Bender and Patrick McIlheran

The estimate by the Department of Public Instruction of potential added costs to property tax payers from opening Wisconsin’s school choice programs to families of all income levels is deeply flawed, producing a figure so unlikely as to be absurd.

However, its flaws point the way toward a potential reform of school choice funding: Simplify the complex funding for choice so it does not shift the burden onto property tax payers, and instead fund Wisconsin’s commitment to broad access to school choice in the same way most other state priorities are funded – through general purpose state revenue from income and sales taxes.

The bill

The DPI on Feb. 16 produced a fiscal estimate for Assembly Bill 970, which opens Wisconsin’s three school choice programs – in Milwaukee, in Racine and in the rest of the state – to families of all income levels.

Right now, families can use state education aid at any participating private school only if their income is under 300% of the federal poverty line if they live in Milwaukee or Racine. Families in the rest of Wisconsin must earn less than 220% of the poverty line to access the choice program. AB970 lets all families access schools through the choice program, just as families of all income levels now can access education through traditional public schools, through charter schools or through inter-district open enrollment.

The bill also lifts the limits on how many children can take part in the statewide school choice program. Right now, choice participation is limited to six percent of the children enrolled in a school district, a figure that phases up until the limit vanishes in 2026-27. The bill, in other words, ends the limit about four years earlier than now.

The estimate

The DPI estimate included a “headline” figure, one reported widely in the press: For the school choice program covering Racine’s metropolitan school district and the program covering the rest of Wisconsin, “the impact on property taxpayers in applicable school districts could be as high as $577.3 million for the 2022-23 school year, assuming 67,869 additional voucher pupils.”

The problem

The figure is at best a gross exaggeration.

The estimate hinges on the complex way in which Wisconsin funds education for children in the school choice programs outside of Milwaukee. The state sends school funding, termed a “voucher,” on a per-child basis to the independent schools that families choose. In turn, it withholds an equal amount of money, per child, from the aid it sends to the school district in which a particular child lives. But the school district is compensated by the state by being allowed for one year to raise its revenue limit – the cap on the total amount of money, both state aid and property taxes, it may take in – by the same amount that its aid is reduced.

In effect, the school district rejected by a family using the choice program loses about $8,500 per child in state aid – but can raise its tax levy by the same amount. As the Legislative Fiscal Bureau puts it, “If a school district chooses to levy to the maximum, its total resources are unaffected by the choice aid reduction, because it replaced the aid reduction with local levy.” The cost of educating children in the school choice program outside of Milwaukee thus is shifted to the property tax.

The DPI attempt to estimate the effect of letting all families access school choice, then, depends on how many additional children will use the choice program. The DPI’s assumption amounted to, “All of them.”

“One approach,” the DPI’s estimate puts it, “to estimating the impacts of the bill is to assume that the total number of additional voucher pupils will be roughly equal to the number of non-voucher pupils currently enrolled in private schools. DPI modeled the impact of the bill using this assumption.”

There are about 56,580 students in kindergarten through eighth grade in private schools in Wisconsin outside of Milwaukee whose parents pay tuition for them, and 11,289 such high schoolers. The DPI multiplied the K-8 enrollment by $8,399, the amount of state aid private elementary schools will get per child in the choice program next year. It multiplied the high school enrollment by $9,045, the high-school per-child funding in choice, and it added the figures to reach $577 million.

The assumptions and methodology are flawed.

A little background first. Schools that register for any of the school programs must allocate a number of seats for each grade for the upcoming year and must tell the state that number. This number is established as part of the requirement of random lottery by which students are admitted to a school with more applications than space. Schools are not required to expand their seat allocations.

In fact, there are numerous reasons why schools would not expand the number of seats available for choice students. First and foremost, the choice grant is well below the per-child cost of operating a school, especially at the high-school level. While per-child aid for students in choice programs ranges from $8,399 to $9,045, the average public school district in comparison spent a total of about $14,259 per child in 2020, the most recent year for which DPI figures are available. In just instructional costs, Milwaukee Public Schools spent $13,317 per child. 

Because schools must accept the per-child choice grant as full payment and cannot charge families any further tuition, many private schools must therefore balance the ratio of choice students and “private pay” students whose higher tuitions reduce how much outside fundraising schools must do to stay afloat.

In other words, simply because the students become eligible for school choice does not mean the program will grow rapidly. Since its inception, demand has always exceeded supply. In the end, the finances, in particular the per-pupil voucher amount, and the physical capacities of schools determine enrollment more than the number of children who are eligible.

The DPI’s estimate, then, of 67,870 new students, whose parents now are paying tuition and relieving taxpayers of any cost of their education, suddenly switching to the choice program is extremely unlikely: Their schools could not afford it, especially at the high school level. The recent announcement by Milwaukee’s Hope Christian Schools that it is closing its high school program because the choice grant was so inadequate compared to costs shows the effect in action already.

The DPI’s estimate also assumed that no part of the expansion of school choice would be students now in public schools, whose parents would choose an independent school if their incomes didn’t disqualify them for choice program. Right now, a family of four with an income of $59,000 is ineligible in parts of Wisconsin outside of Milwaukee or Racine. If school choice were available to all families, it is unreasonable to assume that no families would move their children from public schools to choice schools if there were space available.

If they did, depending on their school district, the move could reduce the burden on local property tax payers.

A district’s ability to levy property taxes is limited by its “revenue limit,” the cap on how much general state school aid and property tax revenue it can take in. That revenue limit is based on enrollment, averaged over three years. A decline in enrollment gradually lowers the revenue a district may take in – since fewer children generally should mean less spending.

For some districts that get comparatively little state aid and, thus, are permitted to levy a great deal of property tax per pupil, the decline in the revenue limit due to falling enrollment could outweigh the upward adjustment in the revenue limit that comes as the state shifts the cost of the $8,300 or $8,946 choice grant onto local taxpayers.

In other words, in some Wisconsin districts, taxpayers already are paying more than $8,946 per child in property taxes for the children in public schools, and so a shift of those children to choice schools would mean less burden on the local levy.

The DPI estimate did not account for this.

Nor did the DPI estimate take into account the timeline on which school choice enrollment takes place.

The fiscal estimate is for the 2022-23 school year, which is only six months away. Students in much of the state who are currently paying for private school and who want to switch into the choice program for the 2022-23 school year would have to do so by April. For those families to switch next year, the bill would have to be signed into law, the DPI would have to extensively revamp its enrollment process, and tens of thousands of families would have to apply – all in six weeks. This is unlikely if not impossible.

The timeline is even more constricted in Milwaukee and Racine, where schools must run lotteries to allocate their available seats among more numerous applicants. Some schools already have closed their applications and are now allocating seats.

In other words, since the bill does not add an extension to the enrollment period this year, there may well be no new choice students enrolled for next year.


It is important to notice that the DPI’s estimate is only for the speculatory effect on property taxes, not taxes as a whole. This matters because the burden of property taxes is distributed differently across the taxpaying public than is that of the state’s other revenue sources, income and sales taxes.

Suppose that, instead, the 67,870 students now in Wisconsin private schools outside Milwaukee whose parents pay their tuition all transferred next year to their local public schools.

The cost of educating these children now does not fall on taxpayers, since they are the kids whose parents pay tuition from their own pockets. Their parents have relieved taxpayers of the cost of their education. But if they all switched to public schools, which in Wisconsin in 2020 spent on average $14,259 per child, simple multiplication in the DPI’s method suggests an added burden on taxpayers of $968 million – or about $391 million more than the DPI’s figure.

That is because, on average, educating a child in Wisconsin’s choice programs costs taxpayers considerably less than educating that child in a Wisconsin public school. The only way that a child’s education would cost Wisconsin taxpayers nothing is for taxpayers to expect that child’s parents to pay for it themselves – an expectation inconsistent with Wisconsin’s long social norm of seeing an educated population as a public good supported by the public.

The question, then, is how public support for education will be funded. One way to prevent any added burden on property tax payers is to continue cutting off access to the school choice program to parents making as little as just over twice the federal poverty level, as we do now. This would be widely viewed as unjust, even unacceptable, in light of polling that shows the expansion of school choice options is broadly popular among Wisconsin families.

Alternately, Wisconsin could simplify its complex mechanism for funding choice, which, outside of Milwaukee, shifts the cost onto property taxes. One path to simplification would be to fund all choice students from general purpose revenue, with the only effect on public school districts’ finances being those associated with change in enrollment, the same as if a family moved to a different city.

This would have the effect of reducing conflicts over shifting burdens, since students would be funded along the same lines as Wisconsin government as a whole.