Charles Wheelan, Ph.D. The Naked Economist

The good news is that the kids are back in school, football season is underway, and my Chicago Cubs are flirting with first place in the NL Central. The bad news is that I'm still not done explaining why it's so hard to make schools better.

Last month, I wrote that we still fumble the most basic task related to education reform -- telling good schools from bad ones. The next discouraging bit is that even when we know which schools need to get much better, the two most intuitive fixes -- spending more money and offering more choice -- have surprisingly modest results.

Maybe this is a case where one plus one equals three.

More Spending, Better Results?

Anyone who's spent time in a struggling school knows that there are lots of things that could be done with more money: adding teachers, fixing decrepit buildings, offering more sports and activities, upgrading books, and so on.

Presumably, doing those things would improve student performance. This is the reasonable view made by those who lean left. Less reasonably, it's also the emphatic view of the teachers unions and the politicians (mostly Democrats) who depend on their support.

Yet the relationship between spending and student achievement is surprisingly tenuous. True, in one famous randomized experiment (yes, children were treated like laboratory rats), students assigned to smaller classes did better than students assigned to larger classes. Obviously, money is what makes smaller classes possible.

The effects were not huge, however. James Heckman, a University of Chicago economist and recent Nobel Prize winner, has argued that the economic benefits of the achievement gains from smaller classes may not even cover the costs of hiring more teachers to make smaller classes possible.

Money Doesn't Change Everything

Meanwhile, there are plenty of other studies showing little or no connection between spending and outcomes, once the research controls appropriately for the backgrounds of the students involved. Schools that spend more generally have better outcomes, but, for reasons I explained last month, it's not necessarily the higher spending that causes them.

(For example, there's also a connection between a student's SAT scores and the number of cars his or her family owns -- but it's not causal. Buying three more cars when Junior gets to high school won't help him get into Harvard. If you understand this distinction between correlation and causation, you can skip last month's column.)

But the most discouraging evidence on the relationship between money and student achievement comes from the large number of states that have had court-mandated changes to their school funding formulas. These remedies have usually involved pumping lots of money into poor schools, with generally limited results. According to a summary of the literature in the Handbook of the Economics of Education, "While spending may have been equalized, there appears to have been no commensurate improvement in the performance of students from poorer districts."

No Incentive for Improvement

On the one hand, this is surprising. I've been involved with enough public schools to know that there's always something worthwhile that requires just a little more funding.

On the other hand, this isn't surprising at all. I've also been involved with enough public schools to know what happens when good intentions meet a voracious and inflexible bureaucracy. If you pour money into a broken system, it's a bit like one of those children's games where the marble goes in the top and then bounces through ladders, wheels, chutes, holes, and assorted objects before landing a long way from where you thought it was going to fall (if it gets to the bottom at all).

Many school districts, including most large urban ones, don't have much incentive to make sure that money flows to where it'll be most productive. True, there's a whole lot more test-taking going on than there was two decades ago. And lots of talk about "accountability." But consider this simple question: Where would you feel more secure as an employee -- in a public school that's performed poorly for five years in a row, or at an airline (or paper company, or Chinese restaurant) with a similar record?

Memos and "improvement plans" would be flying furiously in the former -- and pink slips in the latter.

Given a Choice

So we should change the incentives! And hold schools accountable! And make parents into consumers, bringing market discipline into our schools!

If a typical Democratic political convention consists of cheering delegates from the teachers unions, the Republicans (and most economists) tend to talk more about the elixir of school choice: Let's use markets to make schools better.

Indeed, school choice is one of the most elegant theories in the realm of public policy. In theory, choice gives a powerful incentive to everyone in the system -- administrators, principals, teachers -- to do the kinds of things that would make their schools more attractive to prospective students. If students flee a school when given the option, the teachers and administrators in that failing institution have to either fix the problem or lose their jobs. That can focus the mind.

There's just one problem with school choice: The data aren't that compelling. Choice does seem to have modest positive effects. In places where school choice has been studied with appropriate controls, the test scores of participating students are a few points higher -- though there's often no difference at all in some subjects or grades.

Choice and Competition

For example, Princeton economist Cecilia Rouse compared the achievement of low-income students in Milwaukee who used state vouchers to attend private schools to those who were eligible but turned away because their private school of choice was oversubscribed. She found that the voucher students had higher test scores in math but not in reading.

Those results are modestly encouraging -- but hardly the miracle cure that market enthusiasts would have you believe.

Again, on the one hand this is surprising. Choice is a fast-track way to short-circuit the bureaucracy and align the incentives of everyone working in the system. It rewards schools that innovate in ways that attract new students and punishes those that don't.

And, again, on the other hand this isn't surprising at all. Schools don't necessarily act like airlines or Chinese restaurants. Think about the Ivy League, the pinnacle of the best higher education system in the world. It's built upon choice and competition. Kind of.

Most Ivy League schools accept less than 1 in 10 applicants. What kind of business turns away 90 percent of its eager customers -- while often accepting those who can't afford to pay over those who can? Why hasn't Harvard doubled or tripled in size? Why hasn't Yale quadrupled tuition? Those are the kinds of things that competitive firms are supposed to do when demand exceeds supply.

Of course, the best schools are often the best because they restrict the supply of difficult students. That's part of competition, too. Competitive businesses make money by improving operations -- but also by shedding loser customers. I expect that the Chicago Public Schools would be excellent if they had to accept only 1 of every 10 eligible students. (Indeed, the magnet schools in the system, which are allowed to select students competitively, are some of the best in the country.)

Still Worth a Try

So what to do? The evidence isn't overwhelming for either spending more money or for creating more competition. Yet if I were a legislator, I'd be comfortable voting for a plan that did both at the same time.

Politically, it's a nice compromise. Substantively, the whole may be much more effective than the sum of its parts. More money gives schools greater resources; school choice makes it more likely that those dollars will be used in ways that make students better off.

Can I guarantee it will work? Nope.

But given that 1) We've been wringing our hands over school reform since the Soviets launched Sputnik 50 years ago, and 2) The two political parties are at loggerheads over strategies that might well be complementary, it seems a reasonable thing to try. But in all honestly, I can't promise you that one plus one will equal three.