There’s a Right Way And a Wrong Way to Think About Merit Pay

Paul Bruno has a good post offering a theory for why merit pay often fails to raise test scores. In short, Bruno thinks teachers simply aren’t sure how to increase student achievement, and thus it’s problem of ingenuity, not motivation or effort. This strikes me as the right way to look at things, and I’m not sure why people continue pushing merit pay based on the reasoning that it will lead to immediate test score gains. At this point it’s charitable to call the research on it “mixed,” and even if merit pay does directly lead to increases in student achievement the effects are likely to be small and not very generalizeable.

More importantly, the idea that merit pay improves teacher performance lends an air of legitimacy to the notion that teachers could increase student achievement by simply trying harder. As a result we have pundits making arguments based on the perception that we’re ignoring poverty in our quest to blame lazy teachers for poor schools. The problem is that, as both Alexander Russo and Michael Petrilli have recently pointed out, nobody serious actually believes this. And while I don’t think there’s any excuse for continuing to argue that people are ignoring the role of poverty, there’s also no need to foster such derpitude by continuing to imply that teachers just need to be more motivated.

That said, merit pay is still a good idea because it’s likely to have benefits that have nothing to do with the short-term test score increases that are the focus of most research. Simply put, giving more money to the best employees builds a healthier profession, leads to better employee retention, and attracts better candidates. Instead of looking at test-scores over one or two years, the merit pay study I want to see would examine whether treatment schools attract and retain higher quality teachers (as measured by mixed-methods evaluation systems) over a 5- to 10-year period. My inclination is that schools that pay more effective teachers more money will attract higher quality candidates, and I bet this would remain true even if the total salary pool was held constant so that teachers at the lower end of the distribution were paid less.

The reason for this is that “high-quality” college graduates want a job where they can differentiate themselves from their peers. None of these candidates believes they will be an average or below average teacher, and so the median salary is irrelevant to them. They want a job where the best are paid like they’re the best. Unfortunately, in heavily unionized professions that’s difficult because of “wage compression.” Josh Barro touched on this issue yesterday:

Unionization has two big effects on teacher pay. It drives wages up overall by about 8 percent, but it also leads to a phenomenon known as “wage compression”: It reduces the gap in pay between the highest- and lowest-aptitude teachers. The biggest gains from unionization go to the lowest-aptitude teachers, while the highest-aptitude ones lose out.

Hoxby and Leigh use the average SAT score at a teacher’s undergraduate college as a proxy for aptitude. In the 1960s, there was a big gap in pay between teachers who went to highly selective schools and those who didn’t. But by 2000, “most states had earnings ratios near one for all aptitude groups” — that is, teachers get paid about the same regardless of where they went to college.

Wage compression even happens in a work environment like the NBA. Lebron James makes less money then he would in the absence of collective bargaining while veterans and bench warmers make more than they would. It’s important to note that wage compression isn’t necessarily a bad thing. But while it provides more economic security for some employees — and there plenty of people who merely want a good middle-class job with pay based strictly on experience — it also makes a profession less attractive to high achievers. And the exceptional people that the teaching profession desires want jobs where they can quickly accumulate evidence that they’re truly great at what they do. To outsiders (i.e. college students) it doesn’t currently appear that such evidence exists for teachers. Merit pay is one way to change that.

Merit pay is also likely to be better for teachers and students if it’s based on multi-faceted evaluation systems rather than just test scores, and that’s why getting evaluation systems right is so important. But if we’re able to do that, merit pay can transform the teaching profession by signaling that there is official differentiation between the good and bad employees. When people talk about teacher evaluation the focus is often on the gains from replacing bad teachers with better ones, but I think the real prize is a profession that makes it obvious to the next generation of potential teachers that the best employees are rewarded and the worst are replaced. Nobody wants a job where the quality of their performance is unappreciated.


2 Responses to There’s a Right Way And a Wrong Way to Think About Merit Pay

  1. It may be the case that for merit pay selection effects are more substantial than motivational effects, although I think the evidence on that is pretty thin at this point since most of the research only really examines the latter.

    To be clear, though, what you’re describing – “multi-faceted evaluation systems” – are typically not referred to as “merit pay” or “performance pay” schemes, since those terms are usually reserved for highly standardized compensation systems based on one or two objective and easily-quantifiable criteria. What you’re describing sounds to me more like just shifting control away from the salary schedule and to administrator judgment. I’m somewhat sympathetic to that – and that’s how lots of professions work – but I think we want to distinguish it from “merit pay” per se.

  2. Eric Horowitz says:

    Agreed, although it’s really only in the last paragraph that I’m arguing that any benefits of merit pay would be stronger if we broadened what earns “merit” beyond just test scores. I think the basic argument still holds for conventional test-based merit pay, and that’s what I’m talking about through the majority of the post. As for selection effects, yes, there’s not much research, and I could be wrong as teaching is not exactly a conventional profession. Hopefully we’ll have more research in the future.

    More generally, something worth thinking about is that while there’s a ton of commentary and input from teachers — and that’s great — there’s a lack of consideration for how non-teachers view things. This matters tremendously because all future teachers are currently non-teachers.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s