School choice allows families to choose schools that are more suited to their children. These choices may affect non-academic outcomes, including students’ mental health. We empirically examine the relationship between school choice and mental health using two methods. First, we use difference-in-differences to estimate the effects of state voucher and charter school laws on adolescent suicide rates. States adopting charter school laws witness declines in adolescent suicides, whereas private school voucher laws are generally not associated with statistically significant changes in suicides. Second, we use survey data to estimate the effects of private schooling on adult mental health. Controlling for a post-baseline measure of mental health and a variety of individual characteristics, the estimates suggest that private schooling reduces the likelihood that individuals report having mental health issues as adults.
This is the abstract of Corey A. DeAngelis and Angela K. Dills, “The effects of school choice on mental health,” School Effectiveness and School Improvement, December 3, 2020.
Here’s one of the key paragraphs:
Across all specifications, the estimated effect of a charter school law is robust: States adopting charter schools witnessed declines in adolescent suicide rates. The estimated effect of a charter school law translates to about a 10% decrease in the suicide rate among 15- to 19-year-olds. Voucher programs, tax credit scholarships, and ESAs assist 468,199 students (EdChoice, 2019c); charter schools enrolled more than 6 times as many students, 2.9 million, in 2015 (National Alliance for Public Charter Schools, 2015). The larger number of students affected by charter schools suggests more potential for affecting children’s outcomes.
Read the whole thing.
READER COMMENTS
Les Holcomb
Dec 21 2020 at 11:08am
Dr. Paul Farmer, a physician who founded Partners In Health along with Dr. Jim Jong Kim (sp.), now head of The World Bank, have always treated “Economic Poverty” poverty as a medical diagnosis in their international work. For instance early on in Haiti they created “medical food” ( I recall that the first one was called “nourinmanbaha”especially to address cases from children and adults starting out the day (breakfast) with caustic medications in their stomach. I used it decades later with a few dual diagnosis (ETOH-MH)sometimes HIV patients over 50 but under 65, Who were coming out of (deinstitutionalized) nursing homes, prison, and unsupported congregate housing, making the transition to independent living (within 24 hours to 3 weeks). He also used this in families with active TB in-treatment living on dirt floors with barefoot children playing in the mud or dirt by buying simple metal waterched roofing and a concrete pad. There were other levels of economic repression, where farmers were prevented from farming if their nation was receiving funds for “famine food”, which were overcome with it as a “medical food”, and they provided all the components locally. They then formed a little factory to make the stuff into it “medication” without refrigeration on site. Looking back farther, I could see all these principles working within many large and small “deinstitutionalizing” projects around the country. And also sporadinc adolescent suicide. I’m no economist or internationalist, but hope that Partners in Health, come back and apply what they have found on other continents where they focus mainly on recovery and prevention of antibiotic-resistent resistent strains of lethal viruses and bacterial infections TB, HIV, Hepatitis in it’s many forms, etc. using these kinds of approaches that try to “monetize” and code every component of treatment. I hope that I am not an embarrasement as a public health, community development, and re-engineering guy. If you can figure out where I might fit better, let me know. I may be no use to these conversations, although I might be able to get some help from firends at “pricing” some of them in a more global way of treatment. For instance: It seems clear that the financial incentives for treating the general Diabetes with a lifetime medication here is $30 to 40 billion a year (without all the new features advertised on the major media. There is no economic incentive for coming up with a “cure” without jarring the economy.
Alan Goldhammer
Dec 21 2020 at 1:45pm
I did heed David’s advice and read the whole thing. I remain unconvinced by this paper which runs pretty close to being ‘junk science.’ Statistics can be used to prove any hypothesis as this paper demonstrates. the authors seem to have forgotten one of the first principles of statistics, correlation is not causation.
More fundamentally, they do not address some of the issues that confound their findings. Many charter schools have a selective admissions policy and can kick out students for rules violation quite easily (the Success Academy has been written about by other for this). They are not mandated to accept special needs students as public schools are. The attrition rate of charter schools is quite high with 1/4 closing after five years with higher closings in areas where there is more poverty. This was all summarized in this Washington Post story.
MarkW
Dec 21 2020 at 5:44pm
More fundamentally, they do not address some of the issues that confound their findings.
How would those issues (a standard list of objections to charters) confound their findings? Those findings are not ‘Charter schools are wonderful in every way’ or even ‘Charter schools have lower suicide rates’, but rather that where school choice is introduced, there are reductions in teen suicides that follow. Certainly that makes sense, doesn’t it? When teens are being bullied to the point of considering suicide, having an escape hatch ought to make a difference in at least some cases (regardless of considerations of the relative academic quality of the new vs old schools). For some kids, escaping toxic peer groups is surely more important than escaping than lousy instruction.
Charlie
Dec 21 2020 at 8:15pm
Sanity checking this, I came across this paper. From Figure 2 in the paper, it appears that child mortality has increased, not decreased, from 1999 to 2016, right around when charter schools came to prominence. I assume that’s why the authors had to resort to a State-level differences-in-differences approach to find significance, but it definitely seems to undercut the argument presented.
David Henderson
Dec 22 2020 at 9:11am
You misunderstand the reason for the diff-in-diff approach. The point of doing it is to factor out other variables that would affect the variable of interest. So if child mortality increased in a certain time period, one might be tempted to attribute the increase to another factor that also increased, such as charter schools. That’s a temptation that drew you in. It’s precisely to handle this problem that economists, whatever their prior view, use the difference in difference approach. If child mortality is increasing, but it increases less where there are bigger increases in charter schools, that’s telling you something.
Comments are closed.