Ease of Enactment

Over the past three decades, the greatest support shown for a market education proposal at the ballot box was garnered by a tax credit plan.

No voucher or tax credit program has yet been enacted as the direct result of a popular vote. When market education programs are put before voters at the ballot box, they are generally rejected by a wide margin. The reason is not hard to fathom. Polls that ask citizens what they know about market education reforms reveal an overwhelming lack of understanding. Sixty-three percent of the general public and 66 percent of parents told Public Agenda Foundation pollsters that they knew “little or nothing” about vouchers, and roughly 80 percent of both groups said they needed to know more.[61] So, when it comes time to choose between a devil they know (public schooling) and a reform they don’t (vouchers, UETC programs, and the like) caution dictates that they maintain the status quo.

Both vouchers and tax credits will remain difficult to sell directly to the public until they are more widely understood. Still, it is worth noting that over the past three decades, the greatest support shown for a market education proposal at the ballot box was garnered by a tax credit plan. The bill, which was put before Colorado voters in 1998, received 41 percent support. The greatest support for a voucher bill over that same period was 36 percent (Washington state, 1996). Vouchers do hold the top all-time spot, having received 45 percent of the vote in Michigan, but that was back in 1972, and only 26 percent of the state’s voters supported a subsequent voucher measure in 1978.[62]

Legislative campaigns to pass market-inspired education reforms have of course been more successful. And though there are more voucher programs than tax credit programs currently in existence, the experience of the last decade suggests that tax credits may now be the easier path. Voucher programs were proposed in the late ‘90s in Pennsylvania, Minnesota, and Arizona, but failed to win enough support for passage. After these unsuccessful voucher forays, tax credit legislation was ultimately signed into law in all three states. While these tax credit bills are not optimal, the same can be said of the voucher proposals that preceded them.

Clearly it is possible to enact some kind of voucher or tax credit plan if enough time and energy are expended. But what about enacting an optimal program, one that has a real chance of instantiating a competitive education marketplace? Though neither an optimal voucher nor an optimal tax credit program has yet been put into place, both experience and polling data suggest that it may be easier to pass an optimal tax credit plan.

Looking first at U.S. school choice programs passed in the last few decades, most voucher laws have imposed significantly more intrusive regulation than most tax credit laws. Four of the five modern voucher programs dictate what admissions policies participating schools must follow (either random lotteries or first come first served). The only exception is Florida’s McKay voucher plan for special education students, which leaves schools with more admissions freedom because of the unique complexities involved in serving disabled students. Voucher programs are more likely to forbid (in the Florida A+ and Milwaukee programs) or strictly limit (in Cleveland) the charging of fees by schools to participating students. Voucher programs can forbid schools from making certain subjects compulsory (e.g., religious instruction under the Milwaukee program). They are more likely to impose stricter public school building code standards on participating private schools (in Milwaukee and Colorado). Voucher programs more often require mandatory state testing (in Colorado, Florida A+, and Cleveland), and they more often set graduation requirements (Florida A+) or student performance conditions (Milwaukee). All of these restrictions violate either the parental choice or school autonomy requirements of effective education markets.

As a counter-example to the above evidence, it has been suggested that the federal HOPE higher education tax credits created by the Clinton administration also “raise [the] regulatory burden” on private education.[63] As it happens, though, the requirements imposed by HOPE are strictly concerned with financial filing requirements and do not affect the freedom or financial responsibility of parents, the pedagogical autonomy or competitiveness of schools, or the legality of profit making. In other words, the “burden” imposed by HOPE tax credits is not as burdensome to the operation of market forces as is the case with typical voucher regulations.

Recent Poll Results

While it has already been noted that voter initiatives and referenda may not be the best avenue for passing market education legislation in most states,[64] due to the public’s lack of familiarity with these programs, public opinion data are still relevant. Legislators will not generally support bills that they think will hurt their chances for re-election, so the relative popularity of vouchers and tax credits will likely affect their respective prospects for passage. According to a fall 2003 survey of voters in 10 states,[65] tax credits were consistently preferred over vouchers (see next page).

Table 14

Support for Vouchers and Tax Credits in 10 States

    AR IN FL LA ME MD MI NE NV WI Avg
Individual donation tax credits Support 58 58 50 65 52 57 57 56 53 53 56
Oppose 34 34 44 26 40 36 37 36 38 40 36
Corporate donation tax credits Support 58 60 54 63 57 61 62 56 58 57 59
Oppose 33 32 39 27 35 30 34 34 34 36 33
Universal v. targeted tax credit Universal 65 66 65 67 61 61 73 67 69 59 65
Targeted 26 24 21 22 29 28 20 23 22 29 24
Universal vouchers Support 50 51 47 58 49 49 47 49 50 52 50
Oppose 39 38 45 30 43 45 44 39 43 40 41
Targeted vouchers Support 34 33 30 40 33 40 37 34 29 33 34
Oppose 54 56 63 48 57 52 55 53 62 56 56

Source: Basswood Research, Washington, DC, October 2003. The full text of the questions can be found in the endnotes.[66] For further details, please contact the Mackinac Center for Public Policy at: [www.Mackinac.org].


As Table 14 reveals, majorities in each of the surveyed states approved of donation tax credits,[67] averaging 56 percent support to 36 percent opposition. The greatest margin of approval for tax credits was 39 points, in Louisiana. Vouchers won approval from a plurality of voters in each state, but not always from a majority, and averaged 50 percent support to 41 percent opposition. Once again, the greatest margin of support was to be found in Louisiana, at 28 points. States that support vouchers by a wide margin tend to be strongly behind tax credit programs as well, but the reverse is not always the case. In Maryland and Michigan, for example, tax credits enjoy a wide margin of support but voucher support exceeds voucher opposition by only three or four points. Among Republicans, vouchers and tax credits usually receive comparable levels of support, but the appeal of tax credits is substantially higher than vouchers among independents and especially Democrats.

More striking than the difference in support for vouchers versus tax credits is the gap in support between universal and targeted programs. When asked to choose between a tax credit aimed at helping all students, versus one aimed solely at helping low-income students in underperforming schools, respondents preferred the universal program by more than two-and-a-half-to one (a margin of 41 percent). The margin of preference for universal programs was an astounding 53 percent in Michigan. Universal programs are preferred by all economic and ethnic groups, though the margins are smaller among low-income and African American respondents.

This consistent preference for universal over narrowly targeted programs is good news given that the success of market reforms depends on ensuring substantial vigorous competition. The benefits do not end there, however. Because universal programs serve so many more people, they offer the prospect of much greater lobbying to support passage of school choice legislation. Programs that benefit only a small percentage of families in one or two districts cannot energize the same large base of activists as programs that benefit all families and communities across a state.

The significance of these polling data must not be underestimated, but it must also not be overestimated. A key difference between vouchers and tax credits is that vouchers have born the full brunt of the teachers’ unions’ anti-choice, anti-market advertising and lobbying campaigns. Vouchers, especially among the constituency most receptive to the teachers’ unions’ views (i.e., Democrats), have accumulated a fair bit of negative baggage over the past few decades. Tax credits have also been subjected to criticism from the teachers’ unions, but the attacks have thus far been less intense. Should that change, tax credits’ margin of superiority over vouchers may erode somewhat. What’s more, opponents of market education are aware of the relative popularity of tax credits and are doing everything they can to weaken it. In Florida, several of the major newspapers consistently refer to that state’s donation tax credit program as a “voucher program.” The fact that it is patently not a voucher program is rarely mentioned. The fundamental differences between the programs are seldom emphasized. Florida, whether coincidentally or otherwise, happens to have the lowest support for tax credits of any of the 10 states surveyed. Still, Florida voters support tax credits by a roughly 10 point margin, compared to a 2 point margin for vouchers.

If You Pass It, They Will Sue: Vouchers and Tax Credits in the Courts

Enacting a market education policy is one thing; defending it in the courts is another. So how well do voucher and tax credit programs compare in their ability to survive the legal gantlope?

In Zelman v. Simmons-Harris,[68] the U.S. Supreme Court found that Cleveland’s voucher program does not violate the U.S. Constitution. In the wake of that ruling it is generally accepted that other carefully designed voucher and tax credit programs will also pass federal constitutional muster, even if they include religious schools. The same cannot be said at the state level. Originally, both the Vermont and Maine voucher programs permitted the participation of parochial schools, but they amended their voucher statutes in 1961 and 1981, respectively, to exclude religious schools. During the 1990s, lawsuits were filed in both states to try to restore the ability of religious schools to participate, but lower courts upheld their exclusion. On appeal, the state Supreme Courts of each state confirmed the lower court decisions. The two state Supreme Court rulings were ultimately appealed to the U.S. Supreme Court, and both appeals were denied in 1999.[69] In other words, there are two cases in which a) states ruled against vouchers for religious schools and b) the U.S. Supreme Court permitted the exclusion of religious schools to stand.

An important distinction between these two cases is that while the Maine ruling was based on the federal Constitution, the Vermont ruling rested entirely on state constitutional grounds. The Supreme Court of Maine ruled in 1999 that including religious schools in its voucher program violated the First Amendment’s Establishment Clause. In the wake of Zelman, that ruling is almost certain to be reversed at some point. The one catch is that while the federal Constitution has now been read to permit the inclusion of religious schools in voucher programs, it has not been read to require their inclusion. Since Maine law now excludes religious schools from redeeming vouchers, it would have to be amended by the state legislature in order to once again permit their participation.

Vermont’s ultimate rejection of religious voucher schools stemmed not from the federal Establishment Clause, but rather from the state’s own constitution. A lower court initially invalidated the participation of religious schools on both state and federal grounds, but the Vermont Supreme Court was more cautious and more prescient. Stating that “the construction of the federal constitution ... faces an uncertain future,” and thus anticipating the Zelman ruling that was yet to come, the state’s supreme justices came down against the inclusion of religious schools solely on state constitutional grounds. Vermont’s highest court found that the inclusion of religious schools violated Chapter I, Article 3 of the state constitution which stipulates that “no person ought to, or of right can be compelled to… erect or support any place of worship…, contrary to the dictates of conscience.”[70]

The counter example to the Vermont case is that of Milwaukee, Wisconsin. In Jackson v. Benson (1998),[71] the Wisconsin Supreme Court upheld the Milwaukee voucher program’s inclusion of religious schools despite the fact that Wisconsin, like Vermont, has a constitutional provision prohibiting “compelled support” of religion. Article I, Section 18 of the Wisconsin constitution states that no person shall “be compelled to attend, erect or support any place of worship, or to maintain any ministry, without consent.” Under the Milwaukee voucher program, taxpayers are compelled to pay for schools that are both ministries (in that they inculcate specific religious beliefs) and places of worship (in that they practice organized school prayer). Nevertheless, the Wisconsin Supreme Court concluded that the compelled support clause was not violated. The Court arrived at that conclusion by applying the compelled support clause only to students and not to taxpayers. The Court’s justification for not considering taxpayers, when this was the group to which the plaintiffs had specifically drawn attention, is murky and unconvincing.[72] The case of students was, on the other hand, crystal clear. Since voucher students could voluntarily opt out of devotional religious classes in Milwaukee, the Justices concluded that they were not being compelled in violation of the state’s constitution. Having so concluded, they deemed the entire compelled support clause satisfied and upheld the voucher program’s inclusion of religious schools.

The compelled support clauses that exist in Vermont, Wisconsin, and 17 other states represent only one of two state constitutional barriers to government funding of religious schools. The other is the notorious and ubiquitous Blaine amendment. The original proposal of James G. Blaine, a failed 19th century presidential candidate, was to amend the U.S. Constitution to forbid government funding of religious schools and other institutions. The public schools of the time were pervasively Protestant, and this amendment was aimed at preventing Catholics from gaining their own government-funded schools. Blaine’s amendment was never adopted at the federal level, but 36 states and the Commonwealth of Puerto Rico eventually amended their own constitutions in accordance with his idea, forbidding state funding of religious institutions or practices.

All but three states have a “compelled support” clause and/or a Blaine amendment in their constitutions.[73] Every voucher program enacted in a state with one of these clauses will be litigated. “Compelled support,” for instance, is one of the two legal pegs on which the ongoing Colorado voucher lawsuit has been hung.[74] Readers may judge for themselves whether current and future legal decisions are more likely to follow the Vermont interpretation of “compelled support” (i.e., that it applies to taxpayers) or the Wisconsin interpretation (that it does not). In any event, states that do find that either or both of these provisions preclude the participation of religious schools are not apt to be overruled by the U.S. Supreme Court, given that the Vermont ruling was allowed to stand.

On February 25th, 2004, the United States Supreme Court ruled in Locke v. Davey[75] that Washington state had the right to deny Joshua Davey a government-funded college scholarship. Davey was a theology student studying at a religious college to become minister. Not surprisingly, there is virulent disagreement over the implications of this decision for government-funded K-12 voucher programs. Does it mean that states could exclude religious schools from voucher programs without running afoul of the Free Exercise Clause of the First Amendment? People for the American Way (PFAW), a group opposed to vouchers, suggests that it might, asserting that the ruling has “broad consequences” for all government funding of religious instruction. Richard Komer,[76] of the pro-voucher Institute for Justice, presents a different view:

The Court issued a narrow, historically based decision involving special state-level concerns that deal with funding the training of ministers…. These concerns clearly are not implicated in school choice programs.[77]

On the face of it, Komer would appear to be right. In deciding this case, the majority weighed two competing principles: the Free Exercise Clause (which arguably favored equal treatment of religious and non-religious degrees) and the Establishment Clause (which arguably proscribes state funding for the training of clergymen/women). Free Exercise mitigated for allowing Davey’s scholarship, while Establishment mitigated against it. The Justices, as already noted, concluded that Washington state did have a compelling Establishment Clause interest in not funding clerical training, whereas Davey’s Free Exercise rights were not substantially circumscribed.

The pivotal statement in Chief Justice Rehnquist’s majority opinion was that: “we… cannot conclude that the denial of funding for vocational religious instruction alone is inherently constitutionally suspect.”[78] Rehnquist defended the 7-to-2 ruling, in part, on the grounds that Promise scholarships are not overly burdensome on religion because they are available for students attending pervasively religious schools, even to those taking devotionally religious courses (so long as they are not specifically pursuing a career in the ministry).

Despite the narrowness of this ruling, there are two reasons to suspect that the Court might also uphold a state’s right to exclude religious schools from a K-12 government voucher program. First, the Court already had an opportunity to strike-down Vermont’s exclusion of religious schools from its voucher program, and it elected not to do so. This is not equivalent to the Court explicitly upholding the Vermont law, but it is certainly suggestive.

The second reason is that, in the wake of Zelman, voucher cases will likely turn on a fundamentally different legal argument than did the Locke case. Locke pit Free Exercise against Establishment because training the clergy was argued to represent precisely the kind of government support for religion that the Establishment Clause was intended to prevent. But Zelman has already determined that neutrally available K-12 voucher programs that include religious schools do not constitute an Establishment Clause violation. With that argument gone, future voucher litigation will inevitably pit Free Exercise against Free Exercise. The question will be: does the free exercise right of parents (who want to use vouchers for religious schooling) trump the free exercise right of taxpayers (who object on moral or religious grounds to paying for that schooling). Because the federal Constitution does not provide clear guidance on how to resolve this conflict, the Supreme Court may well defer to state constitutional provisions that forbid compelled support of religion or state funding of devotional instruction.

When it comes to these religion-based legal challenges, tax credit programs have a distinct advantage over vouchers. In Kotterman v. Killian, opponents of Arizona’s donation tax credit program sued the state, alleging that tax credits constituted government spending and hence violated that state’s Blaine amendment.[79] The Arizona Supreme Court rejected the plaintiffs’ premise, ruling that credits were not government spending and hence did not violate the state’s Blaine amendment. In other words, tax credits just let taxpayers keep more of their own money, and if they choose to claim a credit and make a donation to a religious scholarship granting organization, they do so entirely voluntarily. No one is compelled to support educational practices that violate their convictions. The Arizona Supreme Court decision was appealed to the U.S. Supreme Court, but the nation’s highest court refused to hear the appeal, letting stand the lower court ruling and the reasoning on which it was based.

While state supreme court rulings do not constitute legally binding national precedents, there is considerable evidence that other courts share Arizona’s interpretation of tax credits. In reaching their conclusion, the Arizona justices referred to numerous precedents from other states including Maryland, Indiana, Nebraska, South Dakota, California, Kentucky, and New Mexico, all of which agreed that tax credits are not state money (and hence cannot violate Blaine’s compelled support clauses). The issue was also raised in two separate Illinois Circuit Court cases, both challenging that state’s 1999 tax credit law. In each case, the courts rejected the argument that tax credits are legally equivalent to government spending.

Even scholars skeptical of market education as a whole generally acknowledge this difference. Two researchers writing for the predominantly anti-market Center for the Study of Privatization in Education at Columbia University’s Teachers’ College concluded that “ETCs [Education Tax Credits] – because they are not government funds – are less likely to face legal barriers compared to reforms such as educational vouchers.”[80]

Regulatory Encroachment

Once a voucher or tax credit policy has been enacted and the inevitable spate of legal attacks overcome, it faces its most difficult challenge: resisting the gradual build-up of market-suffocating regulations that can destroy its effectiveness. When first introduced, both the Maine and Vermont programs allowed the participation of religious schools, and were subsequently amended to exclude them. The Dutch system suffered comparatively little regulation when first established in the 1920s, but has since succumbed to a dramatic regulatory expansion including government curriculum and testing, extensive personnel regulations, rising barriers to the creation of new schools, forced consolidation of small schools, etc. Maine’s program has also begun to accumulate a curriculum and testing burden. In Chile, significant parts of the education labor force deregulation ushered in by the voucher program during the 1980s were reversed in the 1990s.

It is true that there are exceptions. The Milwaukee program originally forbade the participation of religious schools and was subsequently amended to include them, and Chilean voucher-redeeming schools gained the right to charge co-payments in 1993, but these are exceptions to the overall trend. Historically, every time a government has begun to fund private schools, it has eventually assumed control over what is taught and who can teach in them. The cases of Holland and Chile actually seem rosy when compared to the system in India. India has both government-funded private schools and parent-funded private schools. Though parent-funded schools have preserved most of their autonomy, government-funded private schools are all but indistinguishable from their government-run counterparts. In some Indian states, government-funded private schools cannot even hire their own teachers, being forced to accept whichever individuals are assigned to them by the education authorities. The freewheeling educational market of the early medieval Islamic empire was similarly stifled by encroaching state control that followed the rise of official state funding. An account of this historical pattern spanning the entire 2,500 year history of formal education can be found in my 1999 book Market Education: The Unknown History.

Some school voucher proponents have argued that voucher and tax credits will suffer equally from this threat to their long-term success. John Humphreys, a research economist with Australia’s Centre for International Economics, writes that

Using either a direct expenditure (education voucher) or a tax expenditure (education tax credit), the government will still provide financial assistance to parents who fulfill the necessary requirements. This will necessitate a degree of government regulation as to what institutions or arrangements are appropriate for educational purposes.

It seems unlikely that there would be any stricter standards under a voucher system than a tax credit system…[81]

But, as already demonstrated, non-refundable tax credits are not generally considered to be government expenditures. By eschewing the use of government money, tax credits not only overcome the Blaine amendment hurdle, they remove a chief justification for regulatory encroachment. This is not to say that tax credit programs will be immune from regulatory interference. There have already been calls in Florida, for instance, to add conditions to that state’s corporate donation tax credit program.[82] But credits clearly do not present the same level of danger as government voucher programs.

To understand this difference between tax credits and vouchers, it helps to look at a concrete example. During the summer of 2003, it was discovered that hundreds of students were receiving scholarships from an SGO in Florida and using them to attend an Islamic school co-founded by Sami al-Arian,[83] the alleged North American head of the terrorist group Islamic Jihad. The scholarship organization immediately halted funds to this school pending an investigation. No lawsuit needed to be filed, no legislation needed to be passed. The SGO is a private organization receiving private donations and when it became uncomfortable funding the school it had the right and the freedom to withhold its scholarships immediately. This is true despite the fact that al-Arian is innocent until proven guilty and the school has not been charged with any crime.

Had the al-Arian-linked school been participating in Florida’s A+ voucher program instead of receiving SGO scholarship students, it is unlikely that it could have been so quickly cut off. If the Department of Education had chosen to act so quickly and decisively, it may very well have wound up in court for capriciously violating the voucher statute (because the school would have remained in compliance with the voucher legislation). No matter what the government voucher program did, it would have caused considerable tension. Either it would have continued to allow the school to accept voucher students, in which case many voters would have been furious and would have demanded stricter regulations on voucher schools, or it would have found itself in a difficult court battle. If the state won the court battle, it would have established a legal precedent for the government to arbitrarily exclude schools from the program. If the state lost the court battle, it would have provided public support for voucher opponents who wished to impose stricter regulations on voucher schools.

This Catch-22 is not simply theoretical. It is all but identical to the situation that exists in Holland over that country’s conservative Islamic voucher schools. Recall that some of these schools are asserted by the Dutch intelligence service to be receiving funding from a radical Islamist group, and to have radical Islamists on their governing boards. Despite a considerable and ongoing hue and cry from parts of the Dutch public, all these schools are still receiving voucher funding because a pre-announced inspection by education officials did not find any radical or violent teachings being conveyed on the days the schools were visited. Pressure to find some way of preventing the creation of new Islamic schools of all descriptions has been intense, and some current proposals have a very good chance of both passing constitutional muster and effectively shutting the door to not only all Muslim schools but to any schools catering to particular religious or ethnic minority groups.

Note that in the case of the Florida tax-credit program, there is nothing to stop another SGO that believes in al-Arian’s innocence from offering scholarships to students attending the school he co-founded. Since that new SGO would only be spending private donations that were given to it voluntarily, it would not generate the same kind of animosity as continued Dutch government support is generating against conservative Islamic schools in that country. Certainly the tax credit program could still be legally challenged for allowing scholarship funds to go to the school, but unless al-Arian is convicted, such challenges would be unlikely to succeed. Opponents of the tax credit program will surely use this case to seek more regulations on private scholarship-accepting schools, but their public policy argument is undeniably weaker than it would be under a government-funded voucher program.

Universality and Public Support

Ongoing public support is clearly necessary to protect choice programs from being curtailed or even eliminated by political opponents. The biggest difference in this regard is not between vouchers and tax credits, but between targeted and universal programs. By serving only a limited segment of the electorate, targeted programs not only reduce the effectiveness of the education market, they also limit the program’s political support base. During budget crunches or changes in the political winds, targeted programs that tax everyone to serve only a small fraction of voters are much more vulnerable than universal programs from which all families and taxpayers can benefit. The Milwaukee voucher program, for example, has come very close to being killed by the Wisconsin legislature, arguably for this very reason. In the summer of 2001, the Milwaukee Journal Sentinel reported that that year’s legislative budget process saw

… Milwaukee’s voucher program pummeled and on the ropes as never before. The Democratic Senate sought to cut in half the amount of the voucher payment a needy parent receives to send his or her child to a private school - a move that would have virtually killed [the] program.[84]