Swedish free schools – segregation and division?

Swedish for-profit free schools have been accused of contributing to the increasing segregation of Swedish society. Critics of the policy approach the question of segregation – or more broadly disparity of educational experience and outcome – in two distinctly different ways. Some, such as Institute of Education researcher Dr Susanne Wiborg, focus on indicators appearing to show that the difference in quality between good schools and bad schools is widening. In a LLAKES Research Paper last year Wiborg argued that school choice widens this gap because popular schools end up developing selective admissions systems (often based on merit) in order to manage demand; absorb more of the available state funding; and thus rob less popular schools of both resources and bright pupils. The degree to which this may be happening is of course a matter of debate. If some really good new schools are established, then the difference between schools will of course increase. In support of her thesis Wiborg cites a number of studies that suggest free schools exacerbate social inequalities. The weight of the evidence base (with the exception of Lindbom and Almgren, 2007, and Nordström and Åslund, 2009) does appear to confirm that segregation in the school system is due in part to the presence of free schools, and cannot be put down entirely to residential segregation. However, we should be careful not to over-state the significance of these findings. Wiborg herself concedes that the theory that these studies showing adverse impacts on educational equality have not been substantiated by the PISA studies (2000/2007/2011). Social segregation is low in international perspective. Green, Preston and Janmaat (2006) found that except for Iceland, Sweden has the lowest variation between schools in all of the OECD countries. But presuming Wiborg’s right, is the answer to reverse the reform? Or is it to address the bit that doesn’t work, encouraging for-profits (acknowledged even by critics as the best incentivised to address the access problem), to expand? There is no question that, at a system-wide level, the presence of free schools improves outcomes (though the degree to which they do so is a matter of debate). Are we to deny to all the possibility of choosing better because of the challenge of bringing that choice to every neighbourhood? Are we really saying that it would be better if all schools stayed mediocre, and that poor communities with no option but to send their children to a school they know will fail them, are better off? The other way in which detractors seek to discredit Sweden’s free school reforms is with reference to the socio-economic background of those that benefit most. Böhlmark and Lindahl’s much discussed study (2007; 2008) found that those that have benefited most have been the children of educationally aspirant parents. However, many overlook their qualification of this finding: although positive effects for low-educated parents and those from an immigrant background did indeed appear insignificant, ‘students from low-income families benefit more than those from high-income families’ and ‘none of these sub-groups are losing out due to a higher private school share’. Gabriel Sahlgren offers further insight into these findings, and there implications for the UK, here. Corroborating these findings, a study by SNS, the Swedish think tank, shows that there has been no change in student results based on family background over the last 20 years: inequality cannot therefore we be said to have increased on this measure. Note Last September a research report by Jonas Vlachos, published by the normally pro-market think tank, SNS, argued that free schools had a segregating effect not because of the basis of their admissions procedures, but because of perverse incentives at work encouraging for-profit free schools to lower their standards in order to fill up their rolls, with the inference drawn that low-educated parents are misled into preferring these schools. Does Vlachos’ thesis stand up? Without disputing that the evidence suggests a measure of grade inflation must be operative in the Swedish system, Gabriel Sahlgren, author of the most robust study of the impact of for-profit management on student test scores in Sweden, responded that the differences in grading practice between compulsory schools with different ownership structures is, however, nothing like as marked as Vlachos’ study suggests. Sahlgren finds that for-profit schools give higher final grades to 2.5-3.3% more pupils compared to municipal schools, a difference so slight as to be just as easily attributable to other factors, and nothing like pronounced enough to affect parent choices. What is true, and widely acknowledged by policymakers in Sweden, is that while for-profit school companies work to overcome segregating influences in the wider economy, not-for-profit organisations run by parents and community groups are less inclined to do so. The reason for this is that parent trust schools are focused on supporting their own communities, whereas for-profits ignore community boundaries and characteristics. The profit motive encourages them to expand wherever there is demand. It also encourages them to invest in quality, as expansion is only financially feasible if their service has been well received elsewhere.