Policy Documents

Taxpayers Savings Grants: Reply to Fuller

John Merrifield and Joseph Bast –
June 10, 2011

Click the PDF to read the full reply.

On May 18, 2011, a Texas-based education consultant and blogger named Ed Fuller posted an essay titled "Vouchers? A Bad Idea and Here Is [sic] Several Reasons Why" on his blog.1 Fuller offers a variety of objections to the general idea of vouchers and specifically to a Policy Brief written by John Merrifield and Joseph L. Bast about a Texas proposal to use a Taxpayers Savings Grant Program (TSGP) to help close the state’s budget gap.2 The authors of that Brief have been asked to reply.

Peer Review

Dr. Fuller notes that the Merrifield-Bast policy brief is not peer reviewed, which is true in the sense that it was not submitted to an academic journal. Such policy assessments generally are not publishable in academic journals, in part, because they must respond quickly to pending legislation.

In the current case, however, the Merrifield-Bast study was reviewed and specific changes were recommended by several policy analysts and academics and accepted by the authors. Reviewers included Herbert Walberg, George Clowes, Allan Parker, and Barry Chiswick. Of course, any remaining errors are the responsibility of the authors.

Reliance on “One Study”

Fuller claims Merrifield and Bast rely “on ONE STUDY that they fail to adequately describe....” This is wrong on many counts.

Merrifield and Bast cite more than a dozen sources in their paper. The study they cite regarding price elasticity of demand for private schooling (apparently the study Fuller had in mind when writing this, though he erroneously says it was conducted in 1990 rather than 1996), was written by Barry Chiswick and Stella Koutroumanes, two leading authorities in the field. It is the most widely cited study on the topic and is generally regarded as seminal and still authoritative. Merrifield and Bast consulted with Chiswick and Koutroumanes to see if more recent research was available and on how best to update and apply their findings.

Merrifield and Bast checked the Chiswick-Koutroumanes study against real-world results from school choice programs in Milwaukee and the Edgewood district of San Antonio, using publicly available enrollment figures for Milwaukee and scholarly research on the Edgewood program. They then produced three different estimates of the enrollment effects of the Taxpayers Savings Grant program resulting from application of the three different enrollment rates. It is difficult to see how Fuller could have overlooked all this.

Private School Tuition

Fuller claims “the authors underestimate the true cost of tuition in private schools...” and returns repeatedly to this claim over the course of the essay. The claim is false, and Fuller’s focus on private school tuition levels reveals that he doesn’t understand how the analysis of enrollment and tax savings was conducted.

Fuller treats anecdotal evidence of tuition levels of some private schools in Texas as comparable to statewide enrollment-weighted averages, which they are not. Merrifield and Bast acknowledge “we have incomplete information about private school tuition levels in Texas” and so chose instead to use “a national enrollment-weighted average tuition amount as reported by the U.S. Department of Education” (p. 4). That estimate, adjusted for inflation, is $9,030 for 2011-12 (p. 5). This is substantially higher than the reported tuition of all but one of the Catholic schools in San Antonio, and five of the 15 private schools in Houston that Fuller reports.

More importantly, actual Catholic school tuition in Texas is irrelevant to the analysis presented by Merrifield and Bast. The cap on savings grants provided by the program – set at 60 percent of current public school M&O spending, or about $5,143 – is less than the national tuition figures the authors use, and less even than 80 percent of the national figures, which is the author’s best guess as to Texas tuition figures compared with the national average (see footnote 3 on p. 5, and discussion on pp. 18-19). The authors assumed that all students would qualify for the maximum grant. How much actual tuition exceeds that amount is irrelevant to the analysis. The analysis depends on the rate at which private school use changes as out-of-pocket cost to families changes, and not based on current private school tuition levels.

As noted above, three separate data sources (Chiswick-Koutroumanes and school choice programs in Milwaukee and Edgewood) yielded virtually the same result: private school enrollment would increase by an amount equal to between six percent and seven percent of total K-12 school population in the first two years of the program. The main point of discussing incomplete Texas tuition data is to observe that many currently available private schools charge less tuition than the TSGP maximum grant, thus likely yielding larger savings to the state than we estimated on the basis of the maximum grant.

Lack of Accountability

Fuller claims “there will be no accountability for such schools to use our tax dollars well. Students will not have to take any type of assessments, there will be no state oversight of the schools, and there will be no data collection on the students or school. We will be handing over tax dollars and saying, “here – take our money, we trust you.”

Since the political debate about the TSGP started, the bill was revised to require that participating schools be accredited or have filed an application for accreditation. This should address Fuller’s concern, but experience suggests that it will not. It is a standard trope of the public education establishment that school choice cannot achieve the level of “accountability” that the myriad public laws and massive bureaucracies of the public sector have achieved, but this is profoundly anti-parent and is simply wrong.

Perhaps the strongest argument for overhauling K-12 funding and governance is that the supposed accountability of traditional public schools to taxpayers and more generally to the American people is a myth. The supposed “accountability” that Fuller fears losing has created a level of dysfunction and plummeting productivity “that threatens our very future as a Nation and a people.” As more and more is spent, achieving virtually unchanged levels of utterly unacceptable measured performance, mounting anecdotal evidence points to plummeting levelsof performance in the untested subjects.

Public education in America today is a system in crises. It fails to graduate a quarter of the children it is supposed to serve; it fails to recruit, frustrates, or outright rejects talented people who would be great teachers and administrators; it produces test scores that put the U.S. far behind most of the developed countries of the world; and it does all this while burning through taxpayer dollars at an unsustainable pace.

Against this background of public-sector “accountability,” the real accountability created by parental choice in education looks promising indeed. In a competitive education marketplace, schools that fail to satisfy parents go out of business, something so rare in the public sector that it generates headlines whenever it happens. In such a marketplace, teachers who are good get paid more, also something unheard of in public schools. Bad teachers lose their jobs, whereas in many public school systems they simply get transferred from one school to another, the infamous “dance of the lemons.”

The TSGP requires the Comptroller to “adopt rules solely to effectuate reimbursement and prevent fraud in financial transactions under this program.” If Fuller doesn’t believe the Comptroller is competent to perform this task, he should present evidence to this effect rather than indirectly defame a public agency.

“Apples to Oranges Estimates”

Fuller claims the Milwaukee and Edgewood district real-world school choice programs that Merrifield and Bast cite are “isolated programs” whereas “statewide programs would have different effects on enrollment in schools of choice. Yet, the authors never mention this.” This too is wrong.

The enrollment rate predicted by Chiswick and Koutroumanes is plainly applicable to a state-wide program, since it is based on national data. Regarding the Milwaukee and Edgewood district programs, the authors acknowledge that “neither of these programs is exactly the same as the proposed TSGP” (p. 11), but observe that their actual enrollment rates were remarkably similar to one another and to the rate predicted by Chiswick and Koutroumanes. This is circumstantial evidence, as it were, that parents react similarly to the availability of choice whether that choice is offered in rural or urban areas, limited to low-income families (in the case of Milwaukee), or subject to other restrictions on participation.

Importantly, Merrifield and Bast go on to explain that the differences between the Milwaukee and Edgewood experiences and what would occur with the TSGP.  They write:

The analysis uses enrollment rates from other programs that are or were more restrictive or less generous than the TSGP would be. The Milwaukee program offers grants that are worth only 45 percent as much as public school per-pupil spending, does not allow schools to charge more than the amount of the grant, and is open only to low-income families who reside in a single city. The Edgewood program offered grants that ranged from only $2,000 to $4,700 and participation was limited to residents of that largely low-income community. We believe a universal option-demand program such as the TSGP would see a higher enrollment rate, though perhaps not in the first two years of the program.

The shrinking amount of time between possible enactment of TSGP and the next school year  probably will yield smaller first year effects than Merrifield and Bast predicted. Merrifield noted that in testimony to the Senate and House committees processing SB 1 and HB 33.

“Faulty Assumptions about Supply”

Fuller claims that “the schools participating in this program would likely have extreme difficulty in attracting and retaining high-quality math and science teachers,” but he cites no evidence that this would be the case. Fuller fails to understand that public schools have difficulty recruiting and retaining such teachers because of policies that discourage and frustrate talented and motivated educators from entering the classroom or, once there, remaining as teachers.

This is a subject addressed by Bast and two coauthors in a recently released Policy Brief from The Heartland Institute. “Public school teachers often must perform under poor working conditions, are increasingly micro-managed by bureaucrats, and have limited job and career opportunities,” they write. “By some measures, teacher pay has not kept up with the compensation of professionals such as lawyers, architects, and accountants.”

Bast et al. also write:

... greater parental choice in education increases demand for good teachers, which in turn leads to higher compensation. More funds get spent in classrooms rather than on management and overhead. Opportunities are created to allow teachers to specialize and to become education entrepreneurs or change careers. More generally, better relations with parents are likely to come from parents choosing schools, since the act of choosing gives parents a bigger emotional and often financial commitment to the schools their children attend. Students are more likely to be enrolled in schools that can meet their specific needs and interests, rather than in large and impersonal one-size-fits-all institutions.

Whether the supply of private schools would expand rapidly enough to meet demand created by the TSGP is a legitimate concern, but Fuller makes no arguments that are not anticipated and answered by Merrifield and Bast (see “Supply Response,” pp. 11-14). In particular, “the private K-12 schooling sector constitutes a very small part of an education marketplace that includes public pre-kindergarten and K-12 schools, public and private technical and business training, and public and private higher education,” and “the TSGP may not increase the total amount of schooling demanded, but merely reallocate the shares of the public and private sectors” (p. 12). Until he addresses those facts, Fuller’s reservations are not credible.

“Faulty Analysis of the Impact on State Taxpayers”

Fuller wonders why the authors forecast that “only 6 to 7% of families who currently enroll in private schools will send their children to public schools and then return them to private schools” in order to qualify for the savings grants.  He asks “what if rather than 6 or 7% transfer, 20% actually transfer? Or 50%?”

The TSGP restricts eligibility to students entering Kindergarten or first grade or who have been enrolled in a public school for an entire school year. To estimate how many parents might withdraw their children from private schools and put them in public schools for one year, only to move them back to private schools a year later in order to qualify for the grants, requires that we know how many parents with children already in private schools would be influenced by the change in the cost of tuition caused by the TSGP.

The authors’ 6.3 percent to 7.6 percent estimate is explained in some detail on pp. 17-18, pages that Fuller either didn’t read or understand. Here’s the logic: Is it likely that parents paying tuition to enroll their children in private schools are more or less sensitive to the price of tuitionthan parents with children in public schools? Common sense says the answer must be less, since by choosing to pay tuition at their current levels they have already signaled by their choice that the high price of choosing a private schools is not an obstacle to making that choice. Therefore, the enrollment rate found for parents with children in public school must be the highest estimate for the percentage of parents who might move their children from private to public schools in order to take advantage of the TSGP later.

Q.E.D.

Cost of Administration

Fuller says “the authors totally ignored the cost of administering such a program. The administration costs would be rather large due to the initial start-up costs and the need to ensure against fraud.”

The TSGP legislation puts responsibility for administering the program in the hands of the state Comptroller. At the time the Policy Brief was written, the Comptroller’s office had not released an estimate of what it thought its expenses would be. In the past week, the Comptroller reported that it expected its cost to be $4.8 million a year. This is less than 1 percent of the projected savings of the program, so Fuller’s concern is not justified.

“Segregation of Students”

Fuller claims “there is ample research evidence (Ladd, 2002) that finds school choice programs segregate students on the basis of race, ethnicity, socio-economic status, and ability level.” This is a deliberately inflammatory claim that cannot be supported by an objective review of the literature.

The literature on school choice finds its effects on enrollment are complex and highly dependent on enrollment patterns present in public schools prior to the introduction of greater parental choice. Ladd’s claim, cited by Fuller, is presumably based on observations of a school choice program in New Zealand where central government regulation precludes major differences in the school choices. Those findings probably have limited application to the situation in Texas.

Research on the effects of school choice programs operating in the U.S. show the current system’s school choices, often made in the presence of few readily observable differences among schools except student body composition, often lead to segregation on the basis of race and socio-economic status. A system of schools having significant differences in school pedagogyand subject themes, such as what would likely emerge under the TSGP, would produce sorting based on parental insights into what kinds of schools are best for their children, which in turn could have desegregating effects.

Indeed, understanding that school choice can lead to greater racial integration and other types of sorting that benefit children is already embodied in school choice programs created in many places, in part, to achieve integration without busing. For example, in Cambridge, MA, schools are encouraged to differentiate themselves to achieve, through choice, improved racial balance. For other examples, see the several sources cited in Walberg’s 2007 book, School Choice, The Findings.

The bottom line seems to be that the details of a school choice policy determine whether or not it is likely to realize an improvement in racial/socioeconomic balance compared to the current restrained choice among comprehensively uniform traditional public schools.

Having responded to Fuller’s concerns here, we are comfortable saying that we believe the TSGP also would benefit students, parents, and teachers in Texas.

Conclusion

The essay by Ed Fuller raises some reasonable concerns and objections to school choice in general and the TSGP in particular. But they are readily addressed by paying more attention to the original Policy Brief, which explicitly addresses and answers many of them, or by reference to the generally accepted literature on school choice.

We stand by our original analysis of the Taxpayers Savings Grant program: It is likely to save taxpayers approximately $2 billion in the first two years of the program. When asked to produce that analysis, we were not asked to address the effects of the program on student achievement. But having responded to Fuller’s concerns here, we are comfortable saying that we believe the TSGP also would benefit students, parents, and teachers in Texas.