I think the most innovative part is a call for merit pay instead of seniority pay, with the full support of the teachers unions. (That is, the plan calls for the unions to work to craft a merit pay plan. I'm not suggesting the unions support the concept today.)
For some background on merit pay, check out Denver. They are probably running the most aggressive merit pay system in the nation. And the teachers unions crafted the plan.
Here's an interview in Education Sector (a neat independent think tank) with Brad Jupp, a labor organizer with the Denver Classroom Teacher Association and a lead negotiator and advocate for the merit pay system.
The whole interview is worth a read, but here are some of the best parts.
ES: What were some of the specific lessons you learned in the pay-for-performance pilot?
BJ: The most important lesson was that you can build pay systems around pragmatic judgments. By pragmatic judgments, I mean decisions that are not necessarily based on researched psychometric standards but reflect common sense and professional judgment to make effective decisions. In fact, almost all pay systems–including the single salary schedule in place in most schools today–are built around pragmatic judgments. We will never create a perfectly objective basis for compensation decisions, but if we rely on the common sense of professionals we can go a long way.
The second thing we learned, which is very important, was that differentiated pay did not destroy workplace morale; it created new challenges, but in our pilot schools, we never saw the plummet in morale predicted by opponents of alternative compensation schemes.
The third thing we learned was that, when teachers set goals and plan to meet them, students perform well whether teachers meet those goals or not. When teachers set high-quality objectives, objectives that have clear, measurable outcomes and well-articulated strategies to meet them, and those objectives are assessed routinely throughout the year–kids learn more. Learning became the cornerstone of the way we built the pay system.
A fourth thing that we learned was that we need to think hard about how to connect the stakes in a pay system to the behavior that we're trying to change. Policymakers often think of pay systems in very simple ways: "If I put a lot of money on the table, it's going to change people's behavior dramatically, so I'll put a lot of money on the table for the behavior I want." But you often don't need to do that, and you may, in fact, be making a big mistake.
We've found, for instance, that a $1,000 incentive to work in a high-poverty school with low-performing kids doesn't motivate teachers in schools with wealthier kids that perform well to move to that low performing school. On the other hand, it does motivate teachers to stay at that high-poverty school after they've been hired there. Maybe what you need to do is to put a small amount of money on the table, stabilize the workforce, and then build the workforce in these schools over time, rather then to assume that what you want the incentive to do is to steal teachers from the suburbs. Another example is that it doesn't take a whole lot of money–only about $330 in the compensation model that we have–to get people to commit to look at their objectives twice a year. But if there's no money, they don't do it. Sometimes smaller stakes make a big difference.
There's a lot more, including why teachers should embrace accountability measures, because then they'll get paid more (as they should).
I also love the call in the plan for longer school days and longer school years.
So, say what you want about the efficacy about licensing the lotto to come up with the money, but this is the best plan that's out there about how to improve education. I think it's a big step forward.