Here’s an eye-opening school statistic for you: Only half of Oklahoma’s public education employees are teachers. The bureaucracy is now so big, it takes up half the system. It’s the blob that ate the schools.

Teachers’ unions, and the lousy teachers they protect, have become the central villain in the epic drama of education reform. And well they deserve the role—teachers’ unions exist to fatten themselves by destroying children’s lives.

And these days, all the attention is on teacher quality. Again, teacher quality well deserves the attention it’s getting. It sure beats not paying attention to teacher quality, which was the mode we were in until the last few years or so.

But let’s not lose sight of the bigger picture. Protecting lousy teachers from being fired doesn’t even begin to account for all the mischief the unions create. And raising teacher quality won’t save us if the schools themselves are dysfunctional institutions.

Federal data tell the tale: In Oklahoma, only 51 percent of the state’s public education employees are teachers. Another 10 percent are instructional aides and “coordinators”; 7 percent are building-level administration; 7 percent are in “student support” services such as counseling and nursing; 2 percent are in library and media positions; 5 percent are district-level administration; and a gobsmacking 19 percent (more than 15,000 people) are in “other” positions—bus drivers, cafeteria workers, janitors, and so on. (The numbers don’t quite sum to 100 percent due to rounding.)

This is not a local problem. It’s par for the course nationwide, as school systems in every state have been slowly but surely colonized by parasitic featherbedders. Oklahoma ranks 26th out of the 50 states—smack in the middle—for the percentage of public education employees who are teachers.

Actually, the problem is likely to be worse than these numbers indicate. A growing number of people in the system are called “teachers” but don’t have their own classrooms. They float from classroom to classroom (or so we’re told) providing various forms of teaching help. But nobody really knows how many of these phantom teachers there are.

And this problem is not merely in the government system. Unfortunately, private schools are slavishly following where the government leads. We don’t have state-specific data, but nationwide, only 56 percent of private school employees are teachers. The breakdown of the bureaucracy is also extremely similar. That’s one of the worst effects of a government monopoly on schooling—even those private providers who manage to eke out an existence serving niche markets neglected by the monopoly will be radically influenced by what the monopoly does. (For the record, the private schools are still better than the public schools, mainly because they’re not unionized.)

I have to admit that some of my fellow reformers were prophetic on the issue of bureaucratic bloat, while I was among those who ignored it for too long. Back when I got into the education-reform business about a decade ago, we went through a brief period when this was a hot issue. But I pooh-poohed it, preferring a focus on what are called (in the jargon of the business) “governance” reforms like accountability and school choice.

I opposed the focus on bureaucratic bloat out of what I thought was simple even-handedness. Sure, a lot of money gets wasted in the administration. But a lot of money also gets wasted in the classroom! Let’s give the bad teachers their due; they waste our money just as efficiently as any principal.

More broadly, my approach was based on a desire to look at outcomes rather than inputs. We were bashing the unions for caring only about what goes into the system—money, teaching certificates, smaller classes—and not what comes out. Why should we make the same mistake? What we should care about, I said, was the schools’ performance. Suppose schools spend more on administration, and as a result, schools are better administered and their academic outcomes go up?

Implausible as that hypothetical scenario might seem today, I still think the focus on outputs was right. The hot idea at one point was to pass laws requiring at least a certain percentage of school spending (such as 65 percent) to be spent “in the classroom.” That seemed unwise to me then, and it still seems unwise to me now. Pass that law and all you’ll end up doing is facilitating waste in the classroom—partly because the teachers will have legally protected access to a minimum bundle of cash regardless of how they use it, and partly because a bunch of administrative stuff will be cynically relabeled as “classroom” stuff.

However, I now think my approach to outcomes and “governance” reforms was inadequate. While we certainly want to keep the final focus of evaluation and accountability on outcomes, we don’t want to just turn a blind eye to inputs. Outcomes are the only thing that matters ultimately, but that doesn’t mean they’re the only thing that matters at all. If outputs matter and inputs affect outputs, then inputs matter, too.

What we want is to reorient our analysis of inputs toward producing the right outputs. If we don’t like the outputs, we can’t just shake our heads and say to the system: “No, not that. Try something else.” We have to ask tough and uncompromising questions about what we can change on the input side (and in other places as well, of course) that will yield the outputs we need.

At the time, I did say these same things about the importance of inputs. But I didn’t follow through on that in practice, and my failure to pay enough attention to the unchecked growth of bureaucratic bloat is a case in point.

But better late than never—let’s start asking those tough and uncompromising questions.

What do you hear about when you hear about the growth of school bureaucracy? Administration. We have too much administration. We need to cut back on administration.

But it turns out if we sit down with the data long enough to make a nifty pie chart out of them, it becomes clear that administration isn’t where the real bloat is. Add up all the administration at the building and district levels in Oklahoma, and you get 11 percent (after rounding). But guess what? You’re going to need some administration no matter what you do to the schools. Even much of the more recent growth of administration is outside of anyone’s effective control. These days, schools have to do a lot of extra administrative work to preempt litigation (thanks very much, Supreme Court). And the culture war—which drags on from generation to generation in large part because America insists on educating its children through a government monopoly—ensures that schools will remain a persistent battleground for cultural conflict, which in turn ensures the growth of administration.

Still, let’s say (implausibly) you chop total administration at all levels from 11 percent down to 5 percent. Congratulations, Oklahoma, you just freed up about 4,000 positions. If they all became teachers, total teaching personnel would go up by less than a tenth.

You could free up almost as many positions, and do a lot less damage to the schools, by chopping heavily into “student support.” I’ll admit that you’re going to need to keep the nurses. And truant officers—excuse me, “attendance officers”—are one kind of “student support” I’m always willing to get behind. But they’re going to be a pretty small portion of the category.

Why on earth should the government have a huge bureaucracy to provide students with “guidance”? Isn’t that what parents and families—remember them?—as well as pastors, scout troops, and other community leaders are there for? And, come to think of it, teachers?

Don’t you dare reply that we need a government guidance bureaucracy because families and communities don’t do this stuff anymore. The shoe is on the other foot. The family and the community are declining because government is taking over their functions. (If you doubt it, check out Charles Murray’s March 2009 lecture on “The Happiness of the People,” and my response at www.thepublicdiscourse.com.)

But all that is chump change. It’s hardly even worth the trouble of going after. The real oinkathon is that enormous “other” category.

There’s absolutely no reason for any sector of government to directly employ bus drivers, cafeteria workers, janitors, or any of the rest of this category. The whole enchilada needs to be privatized posthaste. You wouldn’t just eliminate unnecessary positions that are there due to featherbedding, although that’s considerable. More important, though, you’d be able to pay the market rate for the positions you kept, instead of hyperinflated civil-service salaries and benefits (think pensions). And you’d be able to fire people if they didn’t deliver good services.

I’ll admit, crusading against the growth of “administration” plays well in the media. But how much better would it play if we went after school cafeteria food? Why let the nutrition narcs have all the fun?

Greg Forster (Ph.D., Yale University) is a senior fellow at the Foundation for Educational Choice.