A major argument in favor of school choice, including vouchers, tax credits, and education savings accounts (like those created by the proposed Oklahoma Empowerment Act), is that choice will result in improved academic outcomes for students who participate in such programs.

Academic outcomes can be measured in different ways. One is to simply look at attendance. Another is to look at grades earned in class. Still another is to look at scores on standardized tests. Long-term outcomes such as high school graduation or college attendance are yet another way to gauge the academic effects of school choice.

Each of these methods of measurement has its advantages and drawbacks. However, the first two, attendance and classroom grades, have the greatest drawbacks compared to advantages. Attendance really says very little about academic success; it’s more about whether schools maximize funding where attendance helps determine that funding. Classroom grades are easily manipulated. If a teacher or a school is evaluated by the grades given to the students, with higher grades being rewarded, it’s easy enough to make sure students receive top marks, whether or not top marks are truly earned.

Thus, standardized test scores and/or educational attainment are most commonly used to measure academic outcomes.

Theoretically, school choice policies should result in better academic outcomes for students for one simple reason: competition. Think about the incentives that competition engenders. On the one hand, each school in a choice system will have to attract students. Presumably, a major attraction for a school would be the positive academic outcomes for students who attend that school, so a major area of competitive pressure will be student academic achievement. On the other hand, parents and students will respond positively to having a choice. Parents able to exercise choice over which schools their children attend would presumably have more reason to monitor their student’s progress and to make concerns known when school administrators have a financial incentive to be responsive.

From the beginning of the earliest voucher experiment, in Milwaukee, research has been conducted to determine if vouchers improve academic outcomes as reflected in standardized test results. These results have been mixed, often because the research conducted has been of low quality, or because test outcomes in a choice environment have been misunderstood. For example, when a child changes schools, it is highly disruptive to the child’s academic achievement, for a variety of social/psychological reasons. It therefore takes time—three years is now standard—for relative academic achievement to be accurately measured on tests. Another problem occurs when test scores of students who exercise school choice are crudely compared to those who do not. This is an apples-to-oranges comparison since school choice is exercised by families for a variety of reasons, including that individual students often thrive in very different environments. A bullied student might do much better in a choice school even as the bullies in the original public school thrive.

The ideal research comparison would be to compare the test results of students attending choice schools to the results that would have occurred if the students had stayed in public schools. Obviously, this is not possible since the data are unobtainable. Researchers must turn to research methods that can only approximate such a controlled experiment. This is done by conducting controlled randomized trials, which sounds as if researchers control more than they do. The nature of voucher programs is such that decisions made in determining those who do and do not participate is outside of researchers’ control. This is actually an advantage for researchers because they can make various randomized comparisons between those who do and do not exercise school choice, statistically achieving something akin to the double-blind studies conducted in medicine, generally considered a gold standard of research methods.

What the Research Says

A comprehensive database of these high-quality studies, conducted over the years, has been maintained by EdChoice, a school choice advocacy organization. Of the seventeen studies EdChoice has identified, only three indicate statistically significant negative standardized test results occurred for students who chose to leave public schools for private schools of choice.* Four indicate no statistically significant change. Almost a dozen studies using the highest-quality methods show statistically significant standardized test score improvement for students who exercised school choice compared to similar students who did not (or could not, due to being unsuccessful in a lottery) exercise such choice.

A comprehensive academic analysis of 19 school choice studies from the United States, India, and Columbia, all of which were of the controlled randomized trial variety, strongly indicates that vouchers result in positive standardized test score outcomes for students who exercise choice, and that the positive differential over those who do not exercise choice grows over time. Thus, whether one likes it or not, the very best research makes it clear that when it comes to improving standardized test scores, vouchers work.

Vouchers do not work spectacularly. There is not an instant sea-change in overall test results for children who exercise school choice; improvements are best seen after three to four years. Students must have a chance to adjust to new environments, new peers, and different teaching methods. In fact, the evidence is that in its first year, school choice might even appear to be a failure when focused just on test scores. However, no child fully develops in a single year, and students who exercise school choice should not be expected to completely change old habits and attitudes in a single year, either. The future is much longer and our collective future is too important not to avail ourselves of the clear advantages of school choice, which can only accumulate over time.

_____________________________________________________________________________________________________________

*Almost as an aside, it should be noted that the authors of one of the high-quality studies that indicated vouchers had negative impacts on test scores believed the negative result might have occurred at least partly due to the low quality of private schools that chose to participate in the program. This study was conducted on the voucher program in New Orleans. Schools that chose to participate often saw significant drops in enrollment prior to participating, indicating problems with these schools. They also had very low tuition rates, an indication of potentially low quality. Voucher programs often put conditions on private schools that participate, not allowing them to apply admission criteria, requiring testing that impacts their curricula, and presenting risks for disciplinary practices like expulsion. This might explain why higher-quality schools did not participate in New Orleans. It is also a warning that if we as a society care about student outcomes on tests, and if those tests are at all indicative of how well-prepared future generations are to safeguard our safety and security, regulation on schools in a choice environment should be modest in the extreme.