The Electric Commentary

Monday, January 16, 2006

Public Education

Meanwhile, American public education operates a lot more like North Korea than Silicon Valley. I will not wade into the school voucher debate, but I will discuss one striking phenomenon related to incentives in education that I have written about for The Economist. The pay of American teachers is not linked in any way to performance; teachers' unions have consistently opposed any kind of merit pay. Instead, salaries in nearly every public school district in the country are determined by a rigid formula based on experience and years of schooling, factors that researchers have found to be generally unrelated to performance in their classroom. This uniform pay scale creates a set of incentives that economists refer to as adverse selection. Since the most talented teachers are also likely to be good at other professions, they have a strong incentive to leave education for jobs in which pay is more closely linked to productivity. For the least talented, the incentives are just the opposite.

The theory is interesting; the data are amazing. When test scores are used as a proxy for ability, the brightest individuals shun the teaching profession at every juncture. The brightest students are the least likely to choose education as a college major. Among students who do major in education, those with higher test scores are less likely to become teachers. And among individuals who enter teaching, those with the highest test scores are the most likely to leave the profession early. None of this proves that America's teachers are being paid enough. Many of them are not, especially those gifted individuals who stay in the profession because they love it. But the general problem remains: Any system that pays all teachers the same provides a strong incentive for the most talented among them to look for work elsewhere.

That is from Charles Wheelan's Naked Economics, which is proving to be an excellent read.


  • What you say is exacerbated by the fact that most teacher's unions have made it excessively difficult to fire a bad teacher (and arduous even if it is almost certain the firing would succeed). So basically, once you get the job or reach tenure requirements you are guaranteed a set pay structure no matter what your performance.

    BTW, once you finish reading Naked Economics let me know how much it overlaps with the other Pop-Eco books you have read so I can prioritize my other potential reads.

    By Anonymous Scott H, at 2:37 PM  

  • But how do you determine merit?

    Principals? I have had principals who would rate me badly, though I think I'm good, because I have gone toe-to-toe with them on union issues. Some principals have had hot-and-cold relationships with me, so it would even depend on the day of my evaluation.

    Test scores? I have students who do well on tests, and students who do badly on tests, because by the time they reach me, they're either good test-takers or bad ones, and though I do tons of test-prep on top of my regular instruction, some still do badly. If I stopped teaching the "regular" classes and only taught the IB (advanced) students, I'd be the richest guy in school. The exurban students I taught in my first year--by far my worst as a teacher, self-reflectively--likely scored better than my urban kids do, now that I have my shit together. Plus, students I take on next semester will not be tested again until next fall--how can I be judged fairly on those kids? In fact, many students I currently teach aren't even tested, since Wisconsin requires only sophomores to test.

    I have yet to read of a merit system (even John Kerry's) that adequately addresses reality. We're not making widgets here, judged by how many we can make and how uniform they are. We can't be judged like salespeople, or customer service reps, or even CEOs. If you can suggest something, I'll consider it.

    (I agree that, perhaps, in higher-demand fields--like special ed, math, or science--there can be bonus pay. It's also important to note that there are good data suggesting that professional development and experience--what teachers' pay is currently based on--do improve performance, though only to a point; six or seven years' experience, if I remember right, is where it plateaus.)

    By Blogger Jay Bullock, at 8:58 PM  

  • So Jay are you advocating that teacher performance should be entirely unrelated to compensation? Are you saying that job accountability has no effect on performance? Are you saying it is completely impossible to tell good teachers from poor teachers? Are you saying that the idea of being locked into such a pay scale doesn't discourage people who are motivated to succeed and who would like to be rewarded for it from joining the profession?

    Or are you just saying that because we don't have assurances that any changes would immediately result in a perfect system that we shouldn't consider trying to take advantage of these possibilties to see if a merit based system could develop into one that gets better results? Or in other words: why try to fix what is not working?

    Not that I have sufficient expertise to propose a detailed merit system, but one that might work is school choice (even if just among public schools), which would result in schools succeeding based in large part on their ability to attract effective teachers and remove ineffective ones. Sure it would take a while, but when kids perform better at the next level, whether it be high school, college, or simply college entrance exams, the systems or criteria those successful schools use would by copied and modified by other schools seeking to improve.

    By Anonymous Scott H, at 7:58 PM  

Post a Comment

<< Home

Amazon Logo