- Research blog
In the past couple of decades, the American charter school sector – comprising of autonomous schools similar to academies – has grown considerably. Today, about 5 percent of children in the US attend charter schools nationally, but there is considerable variation across states.
The best evidence on charter schools thus far appears mixed – while urban charter schools generate higher test scores, sub-urban charter schools appear to generate lower test scores. This difference, in turn, appears to be due to the prevalence of the No Excuse paradigm in the former schools – characterised by high behavioural expectations, strict discipline, and longer school days or years – and the more progressive approach that is often prevalent in the latter schools. Charter schools that employ the No Excuse paradigm have been found to improve test scores radically among disadvantaged children, and similar effect sizes have been obtained from injecting these practices into low-performing state schools. The No Excuse charter schools have also been found to raise university enrolment.
However, there is scant evidence on the longer-term effects of charter schools on labour market outcomes. In a recent paper, 'Charter Schools and Labour Market Outcomes', Will Dobbie and Roland Fryer seek to remedy this situation by analysing the effect of charter schools on test scores, longer-term educational outcomes, and labour market outcomes (at the age 24-26) in Texas. While the authors are not able to exploit admissions lotteries to obtain random variation in charter school attendance, they use other forms of non-experimental econometric techniques in order to tease out the causal effects.
The results are intriguing. While the authors find positive effects of No Excuse paradigm schools on test scores and university enrolment – which are especially pronounced among minority pupils – the impact on earnings and employment is small and not statistically significant. Meanwhile, regular charter schools have negative effects on both educational and labour market outcomes. In other words, whereas the negative regular charter school impact is apparent regardless of outcome studied (apart from employment where it is not statistically significant), the positive educational impact of No Excuse paradigm schools does not carry over to the labour market.
There are various potential reasons for these findings. One possibility is that No Excuse paradigm charter schools do not teach skills that are rewarded in the labour market, but only in the educational system. Research finds that there is little evidence of manipulation, such as teaching to the test, in similar charter schools elsewhere. This does not support the crowding-out theory, but the authors of this study cannot use a similar approach to study this issue due to the lack of data. Of course, No Excuse paradigm schools could also ignore certain subjects that may potentially be more valuable in the labour market, such as foreign languages, but the authors are not able to study whether or not this is the case with the available data either.
Nevertheless, it is also important to remember that the authors focus on early labour market success, which may mask a longer-term impact. This is a concern if, for example, pupils attending No Excuse paradigm schools are more likely to continue studying following their undergraduate education (for example to pursue medicine or law), which the authors do not investigate. When analysing labour market outcomes among people aged 28-30, which decreases the sample by over 50 per cent, the impact grows somewhat but is still statistically insignificant (and very imprecisely measured).
However, it may still take longer before the effects appear. For example, whereas smaller classes have been found to have no impact on earnings when pupils are 27 in America (despite having positive effects on university attainment), smaller classes have been found to have considerable positive effects in Sweden if they analyse individuals when they are 27-40 years old. Indeed, when analysing only outcomes at the age of 27, the Swedish researchers find no impact of class size on earnings either. This indicates that it may take longer before the positive educational effects of No Excuse paradigm charter schools spill over in the labour market.
Overall, therefore, the study provides an important first attempt to look at the labour-market effects of charter schools. While the authors do not provide an experimental evaluation, their method appears to capture causal effects reasonably well. Importantly, they provide evidence that regular non-hierarchical charter schools may in fact be harmful for both educational and labour-market outcomes. They also add to the evidence base indicating that hierarchical charter schools generate considerably higher educational outcomes – but also warn that they may not generate better labour-market outcomes. More research is necessary to determine what lays behind this finding and whether or not it can be detected elsewhere.
Gabriel Heller Sahlgren
This comment piece is also the Editor's Pick in the CMRE Monthly Research Digest_06_16. The piece reviews a paper by Will Dobbie and Roland Fryer, 'Charter Schools and Labour Market Outcomes', published as NBER Working Paper No. 22502, a free copy of which may be downloaded here.
You can download free copies of back issues of the CMRE Monthly Research Digest here.