Those who oppose school vouchers and school choice in general tend to cite lack of positive results of voucherization...But we can learn something about when choice works by looking at Sweden’s move to vouchers.
A recent paper by Bohlmark and Lindahl uses high quality administrative data for the entire country of Swedend for students who attended compulsary school (grades 1 through 9) from 1988 to 2009. Importantly this includes data for the period prior to the 1992 voucher reform. This allows them to control for pre-reform trends...a fact that Bohlmark and Lindahl argues may have biased the results.
Sweden’s voucher policy allowed easy entry of independently run private schools which any student could attend. Prior to this policy less than 1% of Sweden’s students attended private schools, but by 2009 it had increased to 11%. The authors find that the higher percent of voucher students there are in a district the better students do on a variety of outcomes. They find a a positive effect on test scores, compulsary school grades, choosing an academic high-school track, high-school grades, probability of attending college, and average education by age 24. The study is impressive in it’s scope of data, especially in tracking later outcome variables.
Importantly, they find that the primary way that competition effects outcomes is by improving the performance of the nearby public schools, and not by outperforming the public schools. I’ve written before that focusing on static comparisons of charter/voucher and public schools is missing some of the largest gains. And it is also consistent with previous work from David Card et al on school competition in Canada. The effects of real school choice reforms are systemic and long-run. The Swedish study supports the importance of looking at the long-run: they found that positive effects of choice in Sweden didn’t occur until a decade after the reform was put in place.
The authors discuss some important characteristics of the Swedish system that may contribute to the success. First, the Swedish system does not allowing parents to pay additional fees on top of the voucher.Second, there are strong rules about how schools must accept students. They cannot use ability, socio-economic status, or ethnicity. The authors argue that if competition on selection is prevented, schools are more likely to compete on quality:
“The conditions for school choice that are likely to generate the most positive effects on overall school productivity are discussed in MacLeod and Urquiola (2009). Their framework is a reputation model of learning. They argue that in the Chilean system (where schools can select students based on ability), the schools are more likely to compete by selecting the best students instead of with increasing productivity. In a system like the Swedish, where creamskimming is not allowed, the schools are more likely to compete by improving productivity. In fact, MacLeod and Urquiola (2009) state that if the reputation model holds for a school market, then “if schools cannot select on ability, the introduction of school choice will unambiguously raise school performance and student outcomes.” The positive educational performance effects found in this paper and the absence of effects found in Hsieh and Urquiola (2006) [for Chile] support their story.This is an important lesson school choice reformers should seriously consider.
Another important factor is that for each student that attends an independent school, the school received an amount equal to a large majority of the average per-pupil cost of the students public school system, and this is paid by the student’s municipality. This means that the resources available to the local public school are decreased as more students choose independent schools. This increases the competitive pressure, which the results suggest is an important determinant of improving outcomes. In addition, any kind of organization can start a school, including for profit companies. The authors write:
“Importantly, the full financing of the independent schools comes from the local government in the form of a voucher for each student they attract. Hence, we expect a stronger economic pressure on the local public schools the more students that chooses to opt out and attend independent schools.This all suggests we should not be shielding public schools from the pressure of competition, but designing reforms that ensure that competition can have it’s positive effect.
Overall this study and the case of Sweden’s voucher program have lessons that reformers and reform critics in this country should consider.
- "Lessons on School Choice from Sweden", Adam Ozimek (Forbes). (And to anyone who claims that Sweden is having an "education crisis", read this article.)