Merit Pay

Every few days, another proposal for performance-rated pay for teachers pops up, sometimes in concert with claims that schools should be run more like a business. (Actually, I agree with that although almost certainly not in the way intended by the speaker.) That good workers should earn more than poor workers does seem appealing. However, when re-phrased as "more pay for better performance", alarm bells should ring. Good teachers should be paid more, significantly more, and at the same time, dismissing weaker teachers should be much simpler. I do not believe that merit pay for teachers can ever succeed.

Teacher merit pay comes in many guises. A "master" teacher can earn more. A teacher who meets specific targets can earn more. A superior teacher can move over a bar or salary cap. However, teacher merit pay has three insoluble problems which is why schemes are introduced and soon fail, or introduced and soon dropped.

Firstly, studies have shown that people who are motivated by money tend not to become teachers; instead they sell cars, real estate or work on Wall Street. (More on that below.) In particular, money for teachers comes in something like number six after good leadership, supportive colleagues, supply of resources and so on. Interestingly, money comes in as number one as a dis-satisfier, but that is another subject.

Secondly, as we know from Wall Street, what gets rewarded gets done. Attach pay to selling insurance, and insurance is sold. Attach pay to mis-selling sub-prime mortgages and hey presto, sub-prime mortgages are mis-sold. Recent exposes of teacher cheating in standardised tests, or changing student responses, can be traced directly to the test-score relationship with pay, job security and educational  funding. Teaching to the test is the inevitable response to the same thing, as is focusing purely on the four percent of borderline students at the expense of both the top (who will pass any way) and the bottom (who will fail). The NCLB linkage of money to English and Math led to the decline, in some cases demise, of music, art and foreign languages. The loss of afterschool clubs in UK and US schools can be seen as the result either of paying for the coachign of some (eg, football) and not others, or the removal of coaching from a teacher's job description. What is not rewarded is not done.

The second problem is simply that a teacher's role cannot easily be quantified, and therefore pay cannot easily be attached. Extra pay for test scores means other things are sacrificed. Pay for punctuality means turning up to class takes precedence over helping that student crying in the stairwell. Pay for student ratings means populism and playing to the audience. Pay for "outstanding" lessons, however they are defined means not sharing with colleagues and a return to education's secret garden.

Thirdly, and a development of the second point above, teaching is just too collaborative. An excellent 10th grade teacher inherited those students from someone, so are her excellent results due only to her or because of the inputs? Do those inputs come from the 9th grade teacher, or are they an accumulation of several years prior? Is an excellent AP English Literature teacher responsible for his results, or does the fact that his students also take AP History and AP Spanish Literature? Is that formerly under-performing student suddenly doing so well because of the outstanding Social Studies teacher? Or because his PE teacher has taken her under  his wing and helped her gain confidence and self-belief?

Teacher merit pay can not work. Instead, schools should pay teachers more, make becoming a teacher harder, and removing an ineffective teacher easier.