Commentary: Research Shows Parental Choice Works
According to news reports, a group of public school superintendents from Macomb, Oakland and Wayne counties calling themselves the Tri-County Alliance for Public Education have sent a letter to Gov. Rick Snyder in which they claim: “No quantitative data exists that shows parental choice improves student learning.” This came as a response to proposed legislation that would remove a 1996 cap of 150 on the number of charter public schools that can be authorized by state universities.
School employee unions and other conventional public school special interests have made similar claims, all of which have one thing in common: They are false.
In contrast, abundant high-quality research shows that parental choice programs like charter schools, private school vouchers, tuition tax credits and others positively impact learning. Studies using the “gold standard” of education research — random assignment — create “apples-to-apples” student-level comparisons. Medical researchers use similar methods to test the potential effectiveness of drugs. Among other devices, these school studies rely on admissions lotteries to compare students who enroll in schools of their parents’ choosing to their peers whose parents sought the same choice but had to remain in a state-assigned school.
Several such studies have found evidence that charter schools boost student performance. A study of Boston schools found “large positive effects” in English and math for charter school students compared to their peers in conventional schools. A similar study of charter schools in New York City analyzed nine years of data and found statistically significant positive impacts in both math and English: charter school students’ standardized test scores grew by 5 points in math and 3.6 points in reading each year from fourth through eighth grade.
Another high-quality study (though not random assignment) conducted by the RAND Corporation used longitudinal student-level data from eight states to test long-term impacts of charters. It concluded that students attending charter high schools were more likely to graduate and attend college than their peers. Although middle and high school students in charters did not outperform their peers on standardized tests on average, they did not negatively impact them either — a theory often propounded by charter school opponents.
In its letter to Gov. Snyder, the “Tri-County Alliance” only referenced one student-level investigation of charter schools — a 2009 study by Stanford’s Center for Research on Education Outcomes. This outlier has been cited in talking points by the state’s largest public school employee unions: the Michigan Education Association and the American Federation of Teachers-Michigan. Charter opponents heavily rely on this study because it shows mixed results: 17 percent of charter schools outperformed conventional schools in reading and math, 46 percent did no better and 37 percent did worse.
The CREDO study, however, does not use random assignment, but instead creates “virtual twins” of charter school students for the purpose of comparison. The study’s methodology has been seriously challenged, and has other shortcomings. For instance, the CREDO study only analyzed three years worth of test scores — just one-third the amount of data used in NYC charter school study.
Despite questions of its methodology, the CREDO study actually produced several findings that show positive impacts for charters. For example, as demonstrated in many other charter school studies, CREDO found that performance improves the longer students are enrolled in charter schools. Unfortunately, half of the test scores the study used to compare charter school students to their “virtual twins” came just from students’ first year in charter schools.
Although it is used by charter opponents to make across-the-board declarations about all charter schools, in fact the CREDO study recognized that charter school performance varies significantly by state, due in large part to different state policies. Of the 16 states it studied (Michigan not among them), eight had positive results in reading or math or both. (In fact, similar CREDO studies of Indiana, New Orleans and New York City all show generally positive results for charter schools). Ironically, the original CREDO study found that a state cap on charter schools — the very thing the Michigan Legislature may soon eliminate — correlates negatively with charter school performance.
For private school choice programs, Jay P. Greene of the University of Arkansas points out that there are more “gold standard” studies measuring the impact of vouchers than any other education policy reform idea. And the results are overwhelming positive. According to Greg Forster, a highly experienced school choice researcher, nine of the 10 random assignment assessments of voucher programs found statistically significant positive outcomes for either all or at least one subgroup of participating students. Just one study found neither a positive nor negative impact. Moreover, 19 of 20 empirical studies (not random assignment) of both vouchers and tuition tax credits suggest that these programs also improve the performance of affected public schools.
Finally, there’s evidence that initiatives offering parental choice among conventional public schools, similar to Michigan’s “schools-of-choice” program, also have positive impacts on student learning. A National Bureau of Economic Research paper released this month using random assignment methods shows that in one North Carolina district, students who won an admissions lottery and enrolled in a public school of their choice were more likely to graduate from high school and earn a bachelor’s degree than their peers remaining in state-assigned schools. These improved graduation rates closed the black-white graduation-rate gap by 75 percent.
Public school superintendents and employee unions are hardly disinterested parties on the issue of greater parental choice. Their institutions’ revenue and continued influence, not to mention their own livelihoods in many cases, depend on preserving a system that forces parents to send their children to the unionized public schools they manage. Under Michigan’s system of school funding based on enrollment counts, when parents have other options, conventional public schools lose revenue and their unions lose dues.
The public school establishment’s special interest groups also maintain a double standard when it comes to measuring student performance. They judge charter schools’ academic performance solely by standardized tests, but oppose subjecting their own institutions to similar accountability. For example, the anti-parental choice Michigan Association of School Boards states in its 2010 by-laws that it “opposes the use of testing as the sole criterion for…the determination of school district success.” Likewise, one of the MEA union’s resolutions rejects using a “single indicator such as [standardized] tests from triggering important school decisions.” Except for charter schools, apparently.
Despite this inconsistency, the public school establishment is right: Standardized test score shouldn’t be the only measurement for schools. For example, since every child in a charter school is there because that’s what his or her parents chose, these schools’ success or failure should include measures of how many parents are making this voluntary choice.
Put simply, increased charter school demand and enrollment are the ultimate proof of satisfied parents. Under a market-like system of open enrollment, conventional schools would be judged by the same criterion. If they were, the only conventional public school officials with cause for worry would be ones whose institutions are failing to satisfy parents.
Michigan Capitol Confidential is the news source produced by the Mackinac Center for Public Policy. Michigan Capitol Confidential reports with a free-market news perspective.