Donating the Voucher: An Alternative Tax Treatment of Private School Enrollment

May 6, 2014 • Research Briefs in Economic Policy No. 1
By Andrew A. Samwick

In the United States, parents send about 10 percent of elementary and secondary school‐​age children to private schools, which through their accreditation meet the requirement that students receive an adequate education.1

An important implication is that sending children to private school generates what economists call a “positive externality.” By paying out of pocket for their children’s private education, these families relieve a financial burden on local, state, and federal taxpayers, who would otherwise have to fund the public education of these children. If private schools did not exist, then public schools — and the tax collections to support them — would have to be about 10 percent larger than the current $600 billion spent annually on public education.2

If sending children to private schools generates a positive externality, then parents may be underutilizing private schools because they do not consider the financial benefit they provide others by sending their children to private schools.3

To see the implications of this externality for tax policy, consider that the current arrangement is equivalent to the following: all families with school‐​age children receive vouchers in an amount equal to the per pupil expenditures in their school district. Most families take these vouchers to their local public schools and redeem them for educational services. The others return the vouchers to the school district unclaimed. This voucher donation by parents who use private schools enables public school districts to reduce their expenditures on public education; this is the beneficial externality.

My research examines one way to offset this externality: allow a federal (and possibly state) tax deduction for parents who send their children to private schools, in the amount of the per pupil expenditure in their local public schools. Currently the federal tax code allows a deduction for amounts contributed to charitable organizations — such as governments at all levels — provided the contribution is for public purposes.4 Thus if parents make a cash gift to a public school, they get a tax deduction. The new policy would extend the same tax treatment to the contributions parents make to their public schools by educating their children privately.5

I therefore estimate the changes in federal and state tax revenues that would occur if “donated vouchers” were accorded the same tax treatment as other charitable donations. Individual taxpayers benefit from this policy to the extent that donated vouchers increase their total deductions, thereby reducing their federal and state taxes.

The reduction is approximately the product of the size of the donated vouchers and taxpayers’ marginal tax rates. My analysis combines data from two main sources. The first is the Public Use Microdata Sample of the American Community Survey (ACS), a 5 percent sample of the U.S. population. The ACS contains responses on the presence of school‐​age children and whether they are in public or private school. The ACS has geographic data at the level of a Public Use Microdata Area (PUMA), an area containing approximately 100,000 people with boundaries that conform to counties where possible. The second data source is the Annual Survey of Government Finances, which contains information on enrollments, revenues, and expenditures at the school district level.

Combining these datasets makes it possible to impute a distribution of per pupil expenditures for the schoolage children in the ACS who attend private schools. That is, students in private school can be matched to publicschool expenditure in their geographic area. These combined data are then used as inputs to the National Bureau of Economic Research Internet Taxsim program, which computes federal and state income tax liabilities based on data reported on the individual 1040 form. Comparing tax liabilities with and without per‐​pupil expenditures treated as charitable donations gives the tax effects of recognizing donated public school “vouchers” as tax deductions.

The main results show that the aggregate value of “donated vouchers” was about $48 billion per year between 2006 and 2010 (in constant 2010 dollars). Were these applied as charitable donations for tax purposes, federal revenues would have fallen by $7.75 billion per year, and state revenues would have fallen by another $1.21 billion.

Treating voucher donations in this way has two benefits. Even if no parents change their decision about public versus private schools, the implied redistribution from taxpayers to parents who send their children to private schools will strike many as appropriate, since such parents generate a positive externality (and the tax deductibility only partially offsets this externality). And if deductibility causes additional parents to choose private school, this further relieves taxpayers of the requirement to educate these children. Any resources saved are available for other government purposes or reduced taxes.

Whether this proposed tax treatment of voucher donations is the best way to improve primary and secondary education is a harder question; expanded school choice is a different approach that also has merit. But expanding choice programs requires state‐​by‐​state or locality‐​by‐​locality actions, while the approach examined here needs only one change by the federal government. That change, moreover, simply recognizes that the implicit charitable donations made by parents who send their children to private schools should always have been treated like other charitable donations.

This Research Brief is based on Andrew A. Samwick, “Donating the Voucher: An Alternative Tax Treatment of Private School Enrollment,” in Jeffrey Brown (ed.), Tax Policy and the Economy 27 (Chicago: University of Chicago Press, 2013), pp. 125–160. All works cited are provided there.

1. Snyder and Dillow (2012, Table 2) report that there were 5.49 million students in private school out of elementary and secondary enrollments of 54.89 million in 2009.
2. The statement assumes that the students currently in private schools could be educated for the same average per pupil costs as the students currently in public schools. Dixon (2012, Table 1) reports that total expenditures in 2009-10 were $603 billion, with about 10 percent being capital outlays. Snyder and Dillow (2012, Table 64) present data from the U.S. Department of Education’s “Schools and Staffing” survey showing tuition charges of at least $45 billion at private schools (excluding financial aid and expenditures financed outside of tuition) in 2007-08.
3. This “spillover” is not the only effect of private‐​school use; the children in private school might generate positive “peer effects” if they attend public schools. If so, then sending children to private school generates a negative externality that could, in principle, outweigh the positive fiscal externality. The evidence for such peer effects, however, is weak; see Joshua Angrist (2013), “The Perils of Peer Effects,” National Bureau of Economic Research Working Paper No. 19774.
4. See Section 170(c)(1) of the Internal Revenue Code at http://​www​.law​.cor​nell​.edu/​u​s​c​o​d​e​/​t​e​x​t​/​2​6/170.
5. Note that “public purposes” describes what the government entity does with the donated funds, not the motivations of the taxpayer who makes the donation.

Media Name: economicpolicy-1.jpg

Download the Research Brief in Economic Policy

About the Author

Andrew A. Samwick is a professor of economics at Stanford University.