Posts Tagged ‘Biased Research’

School Choice Deniers

April 13, 2017

Critics hype a pair of studies while ignoring other evidence on education vouchers.

Students work on computers while doing their homework at Bridge Ministry in Lafayette, La.

Students work on computers while doing their homework at Bridge Ministry in Lafayette, La.PHOTO: ASSOCIATED PRESS

President Trump has made a cause of public and private school choice, and liberals who oppose evaluating teachers based on student achievement are now hyping a few studies that have found vouchers hurt student performance. A closer look still supports the case for giving parents choice.

More than 400,000 students in 30 states and Washington, D.C., participate in private-school choice programs whose designs and funding sources vary. Over the last two decades dozens of studies have sought to measure these programs’ impact on student growth. Those with the most rigorous methodologies have produced positive findings.

A meta-analysis last year by the Friedman Foundation found that 14 of 18 empirical studies analyzing programs in which students were chosen at random by lottery found positive academic outcomes. Two demonstrated no visible effect, while two recent studies of Louisiana’s voucher program found negative effects. The Louisiana studies are disconcerting since voucher proponents have hailed the program, and the negative effects were large. Math scores declined in one study by 0.4 standard deviations after one year in private schools, representing a 50% increase in likelihood of failing the state test.

But Louisiana’s voucher program is unusual in several respects. Fewer than a third of private schools participated in the first year, and they had already experienced significant enrollment declines. This suggests that voucher students had their pick of the worst private schools. Some higher performing schools may have been deterred by regulations that prohibit them from setting admissions standards and charging families more than the voucher amount—$5,300 on average in 2012.

Liberals also highlight a Fordham Institute study last year of Ohio’s voucher program that found participants performed “worse on state exams compared to their closely matched peers remaining in public schools.” The study wasn’t included in the Friedman meta-analysis because participating students were not chosen at random, and it excluded students attending the lowest-performing schools who might benefit most from vouchers.

The study did find that vouchers “improved the achievement of the public-school students who were eligible for a voucher but did not use it.” These students tended to be more economically disadvantaged and lower-achieving than those who used vouchers. It appears vouchers impelled low-performing public schools to improve to avoid losing students.

This conclusion is bolstered by the Friedman meta-analysis, which demonstrated positive effects in 31 of 33 studies evaluating the impact of vouchers on public schools. An analysis of Louisiana’s program last year found that student performance increased “in the public schools exposed to the threat of competition, with effect sizes growing in magnitude as the competitive threat looms larger.”

These studies rebut the union claim that vouchers harm students left behind in public schools. Notably, one of the outlier studies was of Washington, D.C., which compensates schools for funds they lose from voucher students.

One reason public schools in urban areas are so abysmal is that the predominantly low-income students they serve have no other options, while the affluent can flee to private schools. This reduces the political and economic incentives to shape up. Vouchers level the playing field among income strata—which liberals should support—and create competition for the public-school monopoly.

Progressives who cherry-pick negative data on vouchers are denying the overwhelming social science that shows private-school choice benefits both participants and public school students. These progressives are thwarting educational progress.