Readers know that I’m frustrated by the tendency of reformers to turn reasonable discussions about issues like teacher evaluation and pay into polarizing crusades. Too often, this yields “reform” victories that look a lot like crude mandates, and then cements these into statute. The Obama administration is hugely guilty on this count, as it has--in a move I deem patently illegal--required states to adopt Duncan-blessed evaluation systems in order to get an NCLB waiver. Sec. Duncan’s latest bout of muscle-flexing threats (most recently with Arizona) has only aggravated my concerns.

Lord knows I endorse differentiated pay, rigorous evaluation, and tough-minded personnel management, but I don’t think anyone really knows the “right” way to do this--and I’m pretty sure that the answer is going to vary across school systems. I think the folks who ought to dictate teacher evaluation and pay are the leaders who have to hire, support, and manage teachers; that’s district (or potentially school) leaders, not state legislators or board members. To my mind, the point of knocking down anachronistic state statutes and policies governing tenure and pay is not so that reformers can prescribe their new formulas, but so that educational leaders can craft better strategies for their schools and systems. (If those folks choose not to act, that’s where accountability, markets, and elections come in.)

There is a second tension here, though, that receives even less attention--and that’s the way that even well-designed systems may stifle emerging models of schooling. Most of today’s teacher evaluation systems rely upon value-added calculations and observational protocols that are engineered for school environments where one teacher owns 25 or 30 students for 180 days in a traditional classroom. If these conditions don’t hold, value-added metrics can get really messy, as can efforts to simply apply the familiar versions of Danielson- or Marzano-style observation (this is especially true in virtual or hybrid environments). If a school has one fifth-grade teacher do the bulk of math instruction and another take the lead on English language arts, many of today’s teacher evaluation systems break down. If schools are piping in virtual instruction, or making heavy use of in-house tutors (a la High Tech High School or Boston’s MATCH School), the systems break down. If a school adopts the New Classrooms model, with teachers sharing ownership of middle school math instruction in a slew of ways, the systems break down. In short, many of the teacher evaluation systems calling for “21st century” evaluation and pay work only so long as schools cling ever more tightly to the rhythms of Horace Mann’s 19th century schoolhouse.

I know I’m supposed to be so excited that we’re “fixing” teacher evaluation that I shouldn’t worry about such things. But I do, for two reasons. One, the same people who tell me I worry too much are the same kinds of folks who told me in 2011 not to worry about the waiver process becoming a way for ED to insert itself aggressively into managing state education systems; who told me in 2010 that there was no way the Common Core could get politicized; who told me in the Bush years not to worry about the goofy designs of Reading First or NCLB, and so on. Two, it turns out that policies are really hard to change once they’re written into statute. Reformers can say, “Sure, these policies might be a little constricting for some schools, but we’ll adjust them later if we need to.” Unfortunately, I don’t think these folks understand how tough it will be to get legislators to rewrite statute or state boards to upend newly established routines, much less to convince school and system leaders who want to use new models or tools that it’s worth picking a fight with state leaders and trying to win permission to jump through bureaucratic hoops in order to figure out a workaround.

If you’re interested in how we might rethink teacher evaluation and pay without locking ourselves into today’s 1.0 answers for the next millennium, you may want to check out tomorrow’s big AEI conference “Teacher Quality 2.0: Will Today’s Reforms Hold Back Tomorrow’s Schools?”. If you’re in DC, feel free to join us, or you can catch it livestreamed here or follow along on Twitter at @AEIeducation. I’ll be co-hosting it with my all-star AEI colleague, Mike McShane. We’ve got nine new analyses that probe these questions, and stellar lineups of thinkers asking the key questions about what new models mean for staffing, what might constrain the emergence of more promising models, and what all this means for research and practice. The lineup of authors and speakers includes Dan Goldhaber, Tom Kane, Matt Di Carlo, Kate Walsh, Segun Eubanks, Jal Mehta, Katharine Strunk, and many more.