Charter schools have been a core element of Philadelphia public school reform. Charter schools get taxpayer money, but operate outside the traditional system's rules.

This freedom from red tape - and union contracts - is supposed to enable charter schools to innovate, to do better by kids while spending less. Charters can, in theory, zero in on kids with special needs or talents. They can, in theory, soar because they're run by educators who share a vision.

Philly has made a big bet on charters - about 32,000 kids attend 61 charters. If viewed as a separate system, the city's charters would be the third-largest in Pennsylvania.

After all the money, the well-meaning effort, the PR, what's the verdict on charters' effectiveness?

The downside has been much on display of late. Philadelphia Academy Charter School in the Northeast, one of the city's biggest (1,200 students), is melting down amid reports of lavish self-dealing and misuse of funds by the school's leaders. This spring, the School Reform Commission canceled two schools' charters for poor performance and put on hold 15 new applications.

Well, you might say, there are always abuses. The embedded power structure is always slow to endorse innovations. But charter schools have to be better than those urban hellholes that kids fled, right?

Not exactly. Not in Philadelphia. Not nationally.

In March, a Rand Corp. report issued a verdict on academic results from the city's charters: "Students' average gains attending charter schools are statistically indistinguishable from the gains they experience while at traditional public schools."

In 2006, the National Assessment of Educational Progress - part of the charter-loving Bush Department of Education - found that, nationally, charters' math and reading scores lagged behind those in traditional schools.

Charter advocates are tireless in dismissing uncongenial facts. The NAEP study compared "apples to zebras," they said, because charters mostly take troubled students from troubled schools. To compare their scores with the mass of public schoolchildren is misleading.

So the Manhattan Institute, a charter advocate, undertook to compare "apples to apples." The best they could come up with was that charters had a tiny edge in test scores over traditional schools with comparable student bodies. Most of that edge derived from Florida and Texas, two taxophobic states with rotten public school systems.

Well, charter ideologues will sniff that I'm obviously a defender of the failed, public-school "educrats." Nope. I welcome giving that crowd a boot in the pants. I support charters as an important experiment.

But the key word is experiment. An experiment tests whether something works. That means demanding reliable data and fair judgment.

For some school-choice advocates - like those who wrote Pennsylvania's original charter law - charters' superiority is an article of faith requiring no proof through data. The mere fact that charters exist is success enough.

Lacking test results to back up their triumphalism, some tout phony measures of success: Look at the waiting lists! Look at the parental satisfaction surveys!

Waiting lists only show the level of discontent with traditional schools. And here's the implicit question such surveys really ask: "Are you a good parent, or a fool who botched the most important decision you make for your child?"

What about that alibi for middling test results - that charters get the most troubled students? This is dubious logic. By definition, a child in a charter has a parent who cares enough about his education to endure the application rigamarole. So, if involved parents flock to charters, doesn't it stand to reason that the traditional schools get stuck with a higher proportion of kids from chaotic, violent, crack-addled homes?

A sad irony: This worship of charters harms them. More rigor and accountability could have prevented foolish charters from opening, helped struggling ones right themselves, and highlighted the innovations of the excellent ones. As the scandal in the Northeast shows, fiscal controls are not all pointless red tape.

By all means, let us support and nurture the charter schools that work for kids. Let's learn from them and try to spread the magic to all schools.

But let's not ignore warning signs and failures just so people nestled in comfy ideological cocoons can go on dreaming, untroubled by the facts.