Science won't vouch for vouchers

Science can be annoying!

Sometimes it gives you an answer that contradicts your beliefs.

When the Department of Education commissioned a study of the Washington, DC, voucher program, a high priority for the Bush Administration, the feds probably weren't hoping for proof that vouchers don't improve student achievement. But that's what they got.

The results, released last week, show students attending private schools (mostly religious) didn't do better than students who stayed in the public schools.

(Despite those results, Secretary of Education Margaret Spellings called vouchers a “lifeline” and said "no one in a position of responsibility can sever this lifeline right now and leave these kids adrift in schools that are not measuring up." But if D.C. public schools are "not measuring up," neither are the private schools when it comes to educating these students. That's what it means to say there's no significant difference between the two groups.)

Voucher advocates have pointed out that some subgroups of the D.C. voucher students showed small, not statistically reliable improvements, which they say is better than nothing. But this comparison test should have been easy for vouchers to pass.

Here’s why: All of the students in the study had applied for vouchers. They either got or didn't get vouchers (up to $7,500 a year) according to a lottery. So the two groups were very comparable, except for one big difference: One group got an educational experience they believed would boost their achievement, while the other group did not. There is extensive research on what happens when people expect scores to rise—they generally do, even if the expectations are based on bogus information.

So when the voucher students got up to $7,500 to take part in a program they expected would work, that should have helped them do better, even if in fact the private schools were no better than the public schools they left.

But that didn't happen.

And the same results have come out of earlier studies in D.C., in Milwaukee, and other places. Of course, test scores are not the only measure of academic achievement, but so far there's no research showing any other benefit.

So if we’re looking for "scientifically-based" strategies for improving learning, vouchers don't make the grade.

Congress is considering extending the D.C. voucher program beyond its scheduled September expiration. You can send your Representative and Senators a message through the NEA web site's Legislative Action Center, asking him or her not to extend vouchers in D.C., and instead use the money to improve public schools for all children.

--Alain Jehlen