The Washington PostDemocracy Dies in Darkness

Neovouchers: The debate continues

.By Kevin Welner

In response to my March 2nd post on the Answer Sheet, Mr. Bedrick of the Cato Institute posted a “rejoinder” on the Hoover Institution’s Education Next blog. Below, I respond directly to some of the claims made by Mr. Bedrick. But I first want to point out why this issue is important.

Generally speaking, Americans know what vouchers are. Cleveland and Milwaukee have had conventional voucher plans for decades. Ballot initiatives to institute conventional vouchers have been voted on (and, I believe, always defeated) in many states. But most of us know little or nothing about neovouchers.

Yet for every conventional voucher used in the US, there are about two neovouchers used.  In a chapter of the 2012 book, “Exploring the School Choice Universe,” Gary Miron and I estimated that in 2011 over 120,000 neovouchers were issued, while fewer than 70,000 conventional vouchers were issued.

So it is important for the public to understand the nature and use of neovoucher policies.

There appear to be three main issues that have emerged from this exchange:

(1) What are the differences between neovouchers and vouchers, and what is the significance of those differences?

(2) What are the effects of neovouchers on the public coffers?

(3) Who receives neovouchers?

As a foundational matter, the first question is probably the most important, since neovoucher policies will likely be found to violate several state constitutions if courts determine that they are merely a back-door way of creating vouchers.

Let me pause here and note that Mr. Bedrick seems to have a Lord Voldemort thing going on. “Neovouchers” is the policy that shall not be namedhe carefully excised the word from every quotation he pulled from my earlier piece. This could have something to do with focus groups and polls that show that Americans don’t support voucher policies. In fact, even voucher policies now use the term “opportunity scholarships.”

Mr. Bedrick also takes issue with my ‘mortgage’ example. I contended that “tax expenditures” are essentially the same whether we’re talking about a state subsidy to help pay a mortgage or a mortgage interest deduction. He suggests instead a comparison between food stamps versus tax deductions for donations to food pantries. But in truth even the mortgage example concedes too much. The similarities between vouchers and neovouchers are greater than that example would imply. But don’t just take my word for it. The legislators in Arizona who came up with the system knew exactly what they were doing. As I describe in the Neovouchers book, John Huppenthal, then-chair of the Arizona Senate’s Education Committee, who had been a longtime voucher supporter, said, “This has turned into something so close to vouchers you almost can’t tell the difference.”

One other point, before I turn to the second and third issues. In my earlier post, I challenged Mr. Bedrick’s statement, “STC programs use private money not public money. Every state supreme court to address the matter has agreed.” His rejoinder pushed back hard on this, but I remain unconvinced. Arizona’s is indeed the only such court to address the matter. The Montana and Illinois cases he points to did not address neovoucher laws (or, if Mr. Bedrick prefers, “STC” laws of the type he’s discussing). The Montana case isn’t even about a tax credit related to schooling. Instead of “Every state supreme court to address the matter has agreed,” I suggest, “Arizona’s supreme court agreed, and some other cases provide some tangential support for the proposition.”

Fiscal Effects

Mr. Bedrick writes, “Even Welner concedes that it is likely that Florida’s STC program does save money, though he is skeptical that Arizona’s does. What I actually wrote was, “I wouldn’t be surprised if Florida’s program does [save the public money].” I am not saying that it’s likely. I’m acknowledging a possibility, while I’m also stressing that the public is ill served by a policy that prevents us from ever knowing the truth.

The available evidence does not in fact suggest that savings are likely. The available evidence is completely inadequate to make any such determination. These policies and laws were designed to be obscure – to block an informed of their effects. It therefore comes with ill grace for their backers to point to extraordinarily weak evidence and suggest a positive finding. If we don’t know the ratio of switchers to stayers, we simply cannot know the financial effects. If I said, “I just paid X people $10, while Y people paid me $4,” and if I don’t tell you the values of X and Y, good luck telling me whether I made or lost money.

In fact, it’s even more complicated than that. In chapter 6 of my Neovouchers book, I attempted to flesh out additional elements that should be considered when trying to figure out the financial impact on the public. The issues that Mr. Bedrick raised are part of this picture: the percent of the tax credit (100% or less) and the per pupil expenditure for public school students, as well as the percent of neovoucher recipients who are switching from public school to private school because of the availability of that subsidy. But there are many other factors that are important. Consider the following four: (a) when the tax credit is for less than 100%, the difference is taken as a deduction; (b) I believe that switching parents in some states (e.g., Louisiana and Indiana) can claim additional tax benefits for private school costs, and a federal law provides a tax benefit as well; (c) switchers have additional effects on taxpayer supported public schools, and the nature and extent of those effects are often dependent on unused enrollment capacity; and (d) the private schools themselves take advantage of additional tax benefits, such as the substantial charitable donations made to Catholic schools that subsidize enrolled students.

So if Mary moves from public school to an Archdiocese school, prompted by the availability of a $2,000 neovoucher, and if the per pupil operating revenue at the old public school was $9,000, this ‘transaction’ may appear to save the state $7,000 (or even more if the tax credit was only for 90%). But assume the new school charges $4,000, meaning that the parent is out of pocket $2,000 and that this is funded through a Coverdell Education Savings Account (allowing for tax-free withdrawals of investment earnings). And assume that the actual cost to the Archdiocese school is $8,000 and that $3,000 of that is funded through charitable donations. The public savings from this transaction, then, are probably about 20% less than originally calculated. And if a substantial majority of the people in Mary’s state who are receiving neovouchers would have attended the private school anyway (something the public is never supposed to find out), then adding up those 20% differences could easily be the difference between black ink and red ink.

Who receives neovouchers?

Valerie Strauss is easily capable of defending herself, but – contrary to Mr. Bedrick’s claim – I didn’t read her statement to be saying that zero low-income families receive neovouchers. Similarly, I want to be clear that in saying that neovoucher programs provide financial assistance to many lower-income families, I am not asserting that the system is of overall benefit to them. Most lower-income families are served by their neighborhood schools, so the benefit of any given system will depend largely on its impact on those schools. This is comparable to saying that the baseball free agency system has provided benefits to small-market teams. The overall system undermines those teams, but it’s the water they all swim in, so each team finds some way to make use of it.

In the absence of a state-authorized evaluation and state-mandated reporting of data, we are left to the tender mercies of advocacy groups. So Mr. Bedrick points us to what he calls an “independent study” from the Center for Arizona Policy. I encourage readers to visit their website and decide for themselves whether this looks like a disinterested group ( Similarly, visit the website of the Commonwealth Foundation ( and the Pacific Research Institute ( – the other folks Mr. Bedrick points to for evidence. We shouldn’t dismiss this evidence just because it comes from advocates, but let’s also not pretend such evidence is in any way a substitute for a true accountability or reporting system. Cato is only one of a large group of advocates promoting these policies, and they have generated a substantial volume of materials. Some are informative; some are misleading – but none can come close to replacing a state-level reporting system.

Finally, I want to acknowledge that Mr. Bedrick is correct that my earlier post inadvertently failed to provide correct context to some statements about Georgia and Arizona. Since the time when I wrote my book (five years ago! damn, time flies…) there has been an explosion in the number of states that adopted neovoucher laws. There’s a new one (Alabama) just this week (though a federal judge has postponed implementation). Much of this change happened in the aftermath of the 2010 midterm elections, when Republicans swept into state offices in very large numbers.

While I was aware of this explosion, I haven’t kept up with the details. So I incorrectly assumed (trusting Mr. Bedrick’s assertion here) that some of those were not means-tested. In looking at a sampling of those new states, I see requirements that the recipients be within 250% (one was 300%) of the federal poverty level. That does sweep in many families generally regarded as middle-income, but it excludes upper-income families. And this is a sensible approach. Moreover, I applaud Mr. Bedrick for suggesting that Arizona and Georgia should perhaps re-think their laws in this regard.