October 14, 2009

The comparability fly in the Ouchi/principal-autonomy ointment

Yesterday from a "stakeholders" meeting (I think at the USDOE), Charlie Barone tweets,

Richard Laine of Wallace Foundation: forthcoming Rand study will show [principal] autonomy in hiring a key factor in student achievement.

I've been expecting something like this for a while, not because I'm connected to a RAND insider (I'm not) but because this is the obvious new version of decentralization form that would marry the 1980s-90s site-based management fad with new managerial fads in education.

To some extent I am attracted to Bill Ouchi's argument about principal autonomy leading to lower total student load. Ouchi's claims about total student load is essentially one of Ted Sizer's central arguments from Horace's Compromise, that the number of students a teacher sees is a key factor in the ability to push student achievement. But... and here's a fairly important but... Ouchi's work is tantalizing rather than definitive (because it has not be replicated substantially in terms of total student load), and the temptation to manage large urban districts as "portfolios" with quasi-independent school-level management may push a single form of decentralization at the cost of comparability in expenses and access to great teachers.

What the heck do I mean by that? In a sentence, we may not want principals to have complete autonomy in a task where they have relatively weak skills: knowing which novice teachers are going to be great teachers.


Everyone and her or his grandmother is focusing on the problem of where senior teachers work. This is an intellectual sleight of hand if you simultaneously argue that teachers with seniority are taking advantage of contracts with seniority privileges on transfer to avoid schools who need them and also insinuate that experience means nothing. Let me get this straight: we need to prevent experienced teachers from exploiting labor-market choice to move to schools with more comfortable teaching situations because... they're not inherently any better than teachers with only a few years of experience? This is an inconsistency ripe for Jon Stewart-like treatment.

More important than the intellectual sleight of hand is the way that this argument ignores an opportunity for a simple but politically sensitive intervention we could make that could simultaneously improve the lives of poor children and new teachers: create regional new-teacher clearinghouses and matching services. Here's the thought experiment: Far from decentralizing, I think it would be a healthy system for schools to require new teachers go into a large regional market where vacancies for relatively new teachers (e.g., those with fewer than three years of experience) would be balanced with a matching process akin to matching of med-school graduates to residencies. This would require collective bargaining and regional agreements between districts (or changes to statute), but here's the idea:

Brand-new teacher's perspective: A new teacher registers with the regional teaching market clearinghouse, with all of the stuff you'd want applicants to provide. The clearinghouse is directly tied to vacancies in the region, and that would probably include multiple districts in most parts of the country. The clearinghouse matches teachers to jobs for the first year. The teachers and administrators are told, explicitly, "This is a one-year arrangement. In the second year, the teacher is headed to a new school, and the administrator provides an evaluation knowing that the teacher is not coming back to that school until at least two years down the road." And that's what happens. At the end of the first year, the clearinghouse matches jobs to teachers who want to continue teaching and whom the first-year administrators recommend continue. Same with the end of the second year. And the clearinghouse's job is to make sure that by the end of a new teacher's third year, that teacher has worked in multiple settings, with different characteristics of students (at least within the range of the region), in areas of the teacher's documented expertise (i.e., no out-of-field matches). 

At the end of Year 3? Open market in the spring, in most places, and administrators wanting to hire on the open market must hire teachers with at least three years' experience -- in other words, teachers for whom there is a record of evaluations from different administrators and for whom there is a record of performance for students in different settings (within the range of the region's student population). Schools are allowed to hire teachers who worked in their schools before... if the now-third-year teacher wants to work there again.

Benefit to teachers: first-year teachers stuck with horrible administrators (or generally toxic environments) know that they'll be moving on if they survive. They'll get experience with multiple settings where they'll be able to demonstrate their chops. At the end of their third year, they'll have some variation in experience with administration to be able to judge people better when applying in an open-market situation. Disadvantage to teachers: if you happen to get lucky and get a great job in Year One, you have to move on.... and let another new teacher get the benefit of that experience.

Benefit to administrators: because new teachers are forced to move on after a year, honest evaluations are less likely to result in social backlashes. When you hire on the open market, you'll know you'll have evaluations and (where this is gathered) other performance data that is from school settings with a range of student populations. Disadvantage: you don't get to hire absolutely new teachers; you get whom you get, and if you were great spotters of talent, or you think you're better than the average principal at spotting good talent, you'll be upset.

(Personally, I think I would prefer this as an administrator: if you've read Moneyball, you know the sabremetricians' rule of thumb: you can predict a baseball player's professional performance from college experience, but someone straight out of high school is just a raw bet without college experience. Why would you want the authority to make hires in a situation where you're almost guaranteed to be a worse judge of talent/skill than any other personnel situation? Then again, I'm sure many principals think of themselves like the [very poorly-predicting] old scouts of baseball, making seat-of-the-pants judgments.)

Advantages for systems: See advantages for administrators above. In addition, you have lower risk with variation in administrators' skills in talent judgment, while principals would still have the autonomy to pick more experienced teachers, after they pick up enough of a record for administrators to see who has more talent. You could also get development of evaluation skills in a regional context without diseconomies of scale. If clearinghouses have to track teachers, they could also be tasked with additional evaluation responsibilities across a region. Advantage for relatively poor systems: you know that wealthier districts will not be able to be as much of a magnet for new teachers, because of regional rotation, and you could push administrators to do what is necessary to convince teachers that they want to return to your district after their initial three-year rotation is done. Disadvantages: there would need to be legal agreements to cover this, and there would be some logistical challenges in identifying vacancies (and making sure those vacancies are reported accurately and promptly) as well as the operation of a clearinghouse. School districts would have to delegate hiring authority for some of their jobs to a regional body, and if school systems really thought that they were hot stuff in terms of talent scouting, that might be hard to swallow. (See above and Moneyball on the egos of baseball scouts and possibly school administrators.) Disadvantage for wealthy districts: poof goes your advantage in recruiting brand-new and relatively-new teachers, because they'll spend some time in your districts but also some time in poorer districts.

Now, the payoff in terms of debates about comparability: a regional new-teacher clearinghouse/matching process would instantly equalize a significant part of the teaching staff across a region, because of rotation among jobs and districts. Yes, there would still be an advantage of wealthier districts in attracting teachers with three or more years of experience, but poorer districts would know that they at least have a shot of persuading new teachers that they can make a good career inside a district... if the relatively new teachers have an experience that is supportive. 

Remember that this is a thought experiment: I don't know of any places with regional new-teacher clearinghouses/matching services, and I dreamed it up out of whole cloth (plus some inspiration from what happens with med-school students). But I think it points out a structural problem with giving principals entire autonomy: with complete autonomy, there is no balancing out of regional needs. Equality of opportunity would depend entirely on the skills of individual principals, and while principals are extraordinarily important, that's putting a heck of a lot of eggs in a single basket. If you care about making sure that a broad range of students have access to great teachers, there are serious dangers in the Ouchi principal-autonomy approach. Listen to this article
Posted in Accountability Frankenstein on October 14, 2009 11:19 AM |