There was a dirt road north of our old farm in Manitoba. Grass grew on the hump between the wheel tracks. Ruts the width of a truck’s axles would get worn into the road. That was all well enough if you were driving the truck, but not so great if you needed to take a car down the track. The axles didn’t fit the tired old ruts.

This week’s debate around the Initiative’s report on building successful schools reminded me of watching people trying to drive cars on those old dirt tracks.

Martine Udahemuka travelled to the UK, Boston, New York, Washington DC and Houston to see how different places turned around failing schools. In DC, she learned about the IMPACT system. Principals evaluated teacher performance on a range of measures, including in-class observations of practice and measures of how well teachers assisted each other. But they also relied on student performance data.

It would be stupid to base a teacher’s performance assessment solely on how well students performed on end-of-year tests. Every classroom has a different starting point, so you would wind up unduly rewarding or punishing teachers for winning or losing that year’s classroom lottery.

DC’s system instead measured what students knew at the start and end of the year. The students’ improvement was then part of the measure of teacher performance.

But it was even more clever than that. Students’ home situations can affect not only their starting points but also their ability to pick up new knowledge over the course of the year. So the DC system also took a lot of these differences into account.

New Zealand’s debates around teacher performance are stuck in tired old ruts that do not fit more modern and better ways of assessing performance. The groove on the right side of the track insists that performance measurement is vital in every other kind of employment and that teaching is no different. The groove on the left points out that basing teacher pay solely on end-of-year test scores yields perverse outcomes.

Both sides of that debate are entirely correct. Old performance appraisal debates based on old performance measurement systems fit those old ruts. But they do not fit when talking about systems like DC’s.

The Initiative’s coming third report will show how New Zealand could do even better than Washington DC. Stay tuned. We’re going to pave that old dirt track.