Wednesday, September 9, 2009

Seniority pay

In the course of arguing here and here for merit pay for teachers Yglesias dismisses the usual seniority based pay scales for teachers as an ineffective version of merit pay (because merit and seniority are not strongly related). But I don't think this is the main argument for seniority pay.

Research has shown people's happiness depends more on whether their condition is improving or deteriorating than on its absolute level. Seniority based pay systems take advantage of this to increase worker satisfaction. Compare a pay scale that rises from $40000 to $80000 over the course of a 30 year career to one than is a flat $60000 independent of seniority. In the seniority based system workers will get annual raises of about 2.3% (in addition to any increases in general pay levels). So their pay will go up a little every year increasing satisfaction. Unions like this as it makes the contracts they negotiate appear better than they actually are.

Of course seniority pay has disadvantages too. It requires workers to trust their employers to stay in business and not to lay off their older workers. So it is especially attractive for jobs like teacher where these aren't major worries. I doubt it will be easy to get rid of.


  1. These comments build on comments recently made regarding previous blogs discussing teachers in the public school system. (See “Teachers”, 8/30/09, and “Rubber Rooms, 8/27/09.) My opinions on all of these subjects is based on discussions I have had with a middle school teacher in a suburban school in the San Francisco Bay Area. The school is not a poorly performing school as judged by the “No Child Left Behind” Act.

    My friend thinks that teachers should receive periodic raises based on seniority and rises in the cost of living. (Pay raises based on continuing education were not discussed.) Although starting salaries need to be higher to attract a competent pool of applicants (see Comments on “Rubber Rooms” 8/27/09), having no raises during one's career would not be satisfactory. She thought that the concept of merit pay had many pitfalls because of the difficulty of measuring performance. It needs to be based on where each individual student started at the beginning of the school year and what level had been reached at the end of the year. Normally a student should advance one grade level in one year, but this still may put them below the school average and the national average. Also the student's past rate of advancement needs to be considered in evaluating their current progress. (For more discussion of student progress, see the comments on the Teachers blog, 8/30/09.)

  2. I agree that merit pay probably doesn't work. Witb merit pay you create an incentive for teachers to increase their students' scores, sometimes when they don't deserve it. With subjective tests such as Reading Comprehension, for example, the teacher who is testing has an opportunity to say, "The student "almost" understood the question so he should get credit for that answer." This is especially true if teachers ar testing and scoring their own students. If you tie increased pay to the test scores, you then will produce higher test scores that really aren't accurate.