Back to Top

Dollars before scholars? Assessing the impacts of the introduction of a profit-motive to Florida's Charter school sector

Politicians worldwide have in the past decades tried to improve their statefunded schooling systems by introducing more choice and competition. By introducing market forces, the idea is to ensure that bad schools either are forced to improve or go out of business and that good schools expand – creating a virtuous ‘race to the top’.

But introducing a profit motive in this process is still anathema, in England and in many other countries. The fear is that profit-making companies will only focus on reducing costs, thereby decreasing quality as well. Instead of a ‘race to the top’, we would enter a ‘race to the bottom’. Furthermore, people have been arguing that profits will introduce stronger incentives among providers to ‘cream skim’ the best pupils, thereby producing higher school segregation and decrease equity.

In a new paper, 'Putting Dollars before Scholars? Evidence from For-profit Charter Schools in Florida', John D. Singleton analyses the impact of for-profit, state-funded charter schools in Florida, whose enrolment shares have increased by about 80 per cent since 2005. Having access to data from detailed annual independent financial audits and accountability reports enables the author to study differences in spending patterns, pupil composition as well as pupil performance among schools with different ownership structures.

The results display intriguing differences between state schools, non-profit charter schools, and for-profit charter schools. Non-profit and for-profit charters tend to be located in poorer areas than state schools on average. Both non-profit and for-profit charters locate in areas with higher shares of ethnic minorities – and for-profit charters do so even to a larger degree than non-profits. The enrolment patterns support this story: charter schools, especially for-profit ones, take on more pupils from ethnic minorities than state schools. The data also suggest for-profit charter schools especially focus on Hispanic kids, whereas non-profit focus slightly more on African-American children. While the share of pupils on reduced or free school meals is somewhat lower in the charter sector overall, it is slightly higher in the for-profit charters than in the non-profit ones.

A starker difference between non-profit and for-profit charter schools is that whereas only 13 per cent of non-profit charter schools belong to a network, fully 60 per cent of for-profit charters do so. In the published paper, the author only presents differences between the two sectors separately for independent schools and network schools. The argument is that networks may realise economies of scale, which may have an impact by itself apart from profit-making status.

But this ignores the fact that for-profit schools may become network organisations precisely because of their for-profit status; indeed, the ability to scale and grow is an argument often used in favour of such schools. Disaggregating network schools from independent ones when analysing differences across ownership structures may therefore generate selection bias. If successful for-profit schools grow into networks, whereas non-profit schools do not, then it would be inaccurate to only compare network schools or independent schools when studying differences between the two sectors.

The aggregated data presented in the working paper suggest that per-pupil expenditures are somewhat lower in for-profit charter schools than in non-profit charter schools on average. In addition, the former spend less money on instruction and have higher pupil-teacher ratio. At the same time, for-profit schools have the same average test scores and actually higher value added than non-profit schools.

In the analyses adjusting for background characteristics, the author only presents estimates disaggregated by network/independent status. The results reveal that independent for-profit schools perform on par with non-profit network schools and slightly lower than non-profit independent schools in reading but not in maths. On average, for-profit network schools perform better than all other groups, although most differences are not different from zero.

However, when adjusting for differences in expenditures, for-profit network schools do better than all other groups. In other words, for-profit network schools are more efficient than all types of non-profit schools (and independent for-profit charters). This means that for the same amount of money spent, they achieve higher results.

Furthermore, since profits are taxed and thus bring in more money to the state, the total efficiency gains of for-profit network schools are probably undervalued. And the null effects of for-profit independent schools also become important in this perspective: some of the money can be repatriated through taxes, which is not the case among non-profit schools.

The author also finds evidence that network size is only positively associated with performance among for-profit schools; among non-profit schools, network size is associated with lower results. The evidence is mostly consistent with a story in which good for-profit schools join networks or scale up as a result of being good rather than becoming good because they scale up. Among non-profit 3 network schools, however, the findings are both consistent with a story in which poorly performing schools scale up and that they become worse as they do so. These differences may potentially be due to different incentives across the different sectors, which are important for policymakers to consider.

Unfortunately, the data do not permit the author to use a solid methodology and we therefore cannot be entirely confident that the estimated results reflect causal effects. However, they are supported by other research, which, on average, suggests that for-profit schools at least perform on par with non-profit schools and often provide more choice for disadvantaged parents, while at the same time injecting more private funding into the state system. Policymakers should therefore at the very least consider allowing regional randomised trials to investigate the benefits of profit-making, state-funded schools in England as well.

Gabriel Heller Sahlgren

This comment piece is also the Editor's Pick in the CMRE Monthly Research Digest_04_17. The piece reviews a paper by John D. Singleton'Putting Dollars before Scholars? Evidence from For-profit Charter Schools in Florida', published as a working paper by Duke University (a free copy of which may be downloaded here), and subsequently in the Economics of Education Review (June 2017).

You can download free copies of back issues of the CMRE Monthly Research Digest here.


Blog Category: 
About the author