What lessons can America learn from Sweden's experiment with vouchers?
By S.H. | STOCKHOLM
SCHOOL vouchers are a divisive subject in America. Proponents claim that vouchers not only grant parents the opportunity to send their children to a private school, but also raise the quality of all education by creating more competition between schools. Critics complain that these subsidies divert necessary resources from public schools, and rarely cover the full cost of a private education. To settle this debate, many have looked to Sweden, where vouchers were introduced in 1992. The results there have been cited as both acase for and against vouchers. So, what has been the actual effect of this Swedish experiment?
Swedish students used to lead international rankings, but the country’s education standards have been declining for years. Indeed 15-year-olds in Sweden perform well below average in mathematics, reading and science when compared with students from other OECD countries, according to the most recent global ranking. Critics of vouchers blame school choice for these dismal results. Raymond Fisman of Columbia Business School recently called the Swedish voucher scheme a disastrous experiment and warned Americans not to go down the same path.
But there are good reasons to believe the problem is not school choice. This is because Sweden's voucher scheme coincided with a host of other reforms, most significantly a change in the national curriculum in 1994, which emphasised individualised learning over teacher instruction. A comprehensive study (in Swedish) published in 2010 found that this was among the most plausible explanations for the drop in student performance. (Sweden duly changed its national curriculum again in 2011.) Norwegian schools implemented similar curriculum changes in the 1990s and saw similar unfortunate results, whereas Finland concentrated on teacher-led pedagogy and saw improvements in student performance.
Earlier this year, the OECD published an assessment of Swedish schools. The report came up with several reasons why Sweden has seen the steepest decline in student performance of all the countries ranked. First, the “disciplinary climate” in classrooms is poor; teachers seem to have little control over unruly students. Second, Sweden has the highest proportion of students who are late for school among all OECD member states. Third, students study less and report lower levels of perseverance than peers from other countries. Fourth, a typical 15-year-old in Sweden receives 741 hours of instruction time in school per year whereas the average OECD student receives 942 hours. The 1994 curriculum change cannot be the sole cause of these problems. But it should not come as a surprise that relieving teachers of some of their responsibilities and authority took a toll on discipline in classrooms and vastly reduced the instruction time that students receive.
In sum, it is unlikely that school choice is the culprit. Indeed, new research suggests that the Swedish vouchers have had a positive, albeit small, impact on student outcomes.
To measure the impact of school choice on student outcomes, one has to isolate the effect of the vouchers net of other reforms. This is what Karin Edmark of the Research Institute of Industrial Economics and her collaborators do in a recent study. They look at both students who were in school when the vouchers were introduced and those who had already graduated, comparing the two groups based on outcomes such as grades in elementary school, criminal convictions and college enrolment. Overall, the authors find that school choice has had a small, but positive impact, particularly for minority and low-income students.
This study does not explain why disadvantaged students appear to benefit more from school choice than their peers. But a plausible reason is that many poor Swedish neighbourhoods have been plagued with bad schools, and vouchers meant students were no longer forced to attend them. Indeed, the authors find that after school choice was introduced, disadvantaged students were more likely than other students to attend schools that were private and far from home.
These findings are supported by previous research. In another study analysing student results in different Swedish municipalities, researchers found a correlation between test scores and the number of independent schools available: the greater the number, the higher the scores. Significantly, these gains were not concentrated among the students in independent schools, which suggests that there are positive effects from competition. A high number of independent schools in a municipality seems to put pressure on all schools to improve their standards.
So what can America learn from Sweden’s experimentation with school choice? First, vouchers appear to have had small but positive effects on educational outcomes. Swedish students thus perform terribly in spite of school choice—not because of it. Second, since the effects are small (and hindered by problematic changes to curriculum), vouchers may be useful but they are no silver bullet.
