The Real Problem with Highly Regulated “School Choice”

The only problem is that the cited commentary says precisely the opposite. Describing Indiana’s voucher program, Schaeffer writes: “Because participating schools will have a significant financial advantage over non-participating schools, lightly regulated [non-participating] schools will face increasing financial pressure to participate.” This captures Schaeffer’s concern as well as my own (which I expressed over a decade ago in the political economy journal Independent Review): We do not fear that private schools will refuse to participate in heavily regulated school choice programs. We know that they ultimately will participate, or be driven out of business by their subsidized counterparts.

We know this because there is extensive evidence to that effect from all over the world and across history. Everywhere that private elementary and secondary schools are eligible for government subsidies, the share of unsubsidized school enrollment falls. The higher the subsidy and the longer it has been in place, the more the unsubsidized sector is generally diminished. The Dutch enacted a heavily regulated nationwide voucher program nearly a century ago. Unsubsidized private schooling remains legal, but has been reduced to a statistical asterisk—now making up less than one percent of enrollment, compared to roughly 70 percent for subsidized private schools.

Our reason for concern over this pattern is also grounded in empirical evidence: it is the least regulated, most market-like private schools that do the best job of serving families. That is the consensus of the worldwide within-country research, which I reviewed and tabulated for a 2009 paper in the Journal of School Choice. The Fordham paper does not discuss this evidence.

Despite imputing to Cato scholars the exact opposite of the view we hold, the paper does include some interesting data. In particular, it offers a new corroboration that voucher programs are more heavily regulated than tax credit programs (a difference whose magnitude and statistical significance was previously established here). This will make it even harder for objective observers to cling to the notion that vouchers and credits are functionally equivalent.