January 7th, 2012
in Op Ed
As an Australian parent in 2012, the public versus private school debate is hard to avoid. In a society where private schooling is becoming the norm, yet literacy and numeracy skills are stagnating, how does one objectively analyse the costs and benefits of school choice.
First, let me say that school choice is just one factor determining vocational, personal and emotional skills during adolescence. Genetics, parenting, the home environment, peer groups, sports and other club activities, amongst many factors, all contribute to shaping young minds.
Additionally, the composition of students at the school plays a strong role in determining academic outcomes. Many private schools for example, offer academic scholarships. If those students had instead attended the local public school, any difference in average academic results may be greatly reduced.
How then does one separate the impact of school choice from these other factors?
Without the opportunity to conduct controlled studies, for example, by studying twins who attend different schools while holding all else constant, the best analysis of the measurable benefits of private schooling would be a statistical test of various measures of ‘success’, controlling for external factors such as parental intelligence and education, household income and location, and child’s intelligence prior to arrival at the school.
Unfortunately, in this debate one of the most overlooked considerations is what measure of ‘success’ would potentially make private schools ‘better’ than public schools. Is it simply a matter of final grades and tertiary entrance scores, or do parents (and children) value a broader measure of success? Does a public school with more diverse student cultural backgrounds give a better social experience, or does a private school offer more valuable professional connections?
The results of any statistical study will necessarily be narrowly defined to reflect the impact of school choice on a single measure (such as academic test scores), ignoring social benefits and opportunities for extracurricular achievement.
So what do economists and social scientists have to say?
Serious studies isolating school effects from external effects are rare, quite possibly due to lack of reasonable data and the expense and time associated with collecting data for this purpose alone. One pioneering study showed that after controlling for demographic factors public schools were in fact better at increasing grades in mathematics during primary school years than private schools.
A follow up study tracked student achievement through from kindergarten to eighth grade has this to say:
It is worth noting how little variation school type really accounts for in students’ growth in achievement.
On the fringe of the issue, economists have examined the ‘house price premium’ in catchment areas of better performing schools. A process which clearly reinforces the school performance divide through selection bias even between public schools themselves, furthering clouding the issue for researchers and parents alike.
Social scientists have also examined which factors lead to the choice of private schooling and, not surprisingly, found a bias towards private schooling by parents who attended private schools.
What astounds me is that any slim evidence on the subject shows that academic benefits of private schools are, at best, very marginal. While at the same time the costs of private schools are extremely high and increasing rapidly. Irrationality to one side, one must suspect that parents perceive benefits from private schooling far beyond any academic advantage offered.
Is it that parents believe their children will be less likely to get ‘caught up in the wrong crowd’? Is that really worth the cost?
In the end it appears that choosing private schooling may well be a signal of your socio-economic status, could potentially provide important future professional connections for your child, but is unlikely to yield significant academic benefits and carries a high risk of wasting your money.