In my recent blog post about an economic experiment in Washington, DC. My anti-union rant received four wonderful replies in the comments and through email, each excoriating me in their own way. My wife added her vituperation later in the evening. I love the debate, and will here respond to my critics.
I should first say that I did not mean to say that the teacher's union is the only problem in primary and secondary education. I realize some might have thought that when I wrote "I think it is a travesty that the U.S. . . . trails most of the first world when it comes to primary and secondary education. I think the problem is the teacher's union." That was a bit strong! I meant to say that I think that's the biggest problem that might be amenable to change via government intervention in economic incentives.
Adrian mentions a number of things that he would list as bigger problems (with Leanne chiming in agreement): general public apathy, commodification of education, high-stakes testing, budgetary priorities, and curricula and licensure in teacher training. I admit that I am uninformed about the nature and extent of these problems. Perhaps I could induce Adrian to write a guest blog post?
Meanwhile Stacey (in the comments) and Topher (in an email) fretted about whether the pay experiment in DC will work. Stacey worries in particular about what system will be used to identify which teachers are meritorious enough to earn bonuses. I agree this is a sticking point. The obvious solution is to use standardized test scores, but this has inherent problems. Teachers may shift their focus from teaching critical thinking skills to teaching students how to memorize facts that appear on standardized tests. This is why the DC experiment is necessary: changing incentives often has unintended consequences. Topher (who knows lots about this stuff by dint of being an education professor) cited some scary data from Texas, which used two testing systems for a while: a national test and a state test. These tests gave scores that correlated very well with each other for years. When the state test only started to be used to determine teacher bonuses, students started scoring much better on the state test relative to the national test. That's not what we want! We want children to score better on standardized tests only if those scores correlate well with learning. When bonuses are linked to the scores, the teachers seem to find ways to improve scores without necessarily improving learning. Topher also points out that the higher education system that works so well uses NO standardized testing and LOTS of tenure. Education is a massive and sticky system that is truly impossible to recapitulate in a few simple equations.
Topher feels that average salary level (rather than the bonus structure for a given salary average) is more of a problem. He quotes a statistic: within 5 years after getting their teaching certificate, more than half of teachers quit the profession. It stands to reason, of course, that those who quit are those who would be the most successful teachers (and who would be the most successful at a variety of other endeavors, as well). Increasing average pay (which the DC experiment is also doing) should increase the quality of teaching by keeping those teachers from departing the profession.
Topher also note that interschool variation (inner city scools compared to suburban schools, say) is much greater than intraschool variation (due to variability in teaching quality from one classroom to the next). I wholeheartedly agree with this and suggest that this might respond well to economic incentives, too. If the average salary were set with that in mind, school systems could raise teacher salaries high enough to overcome the down side of teaching to economically (or otherwise) disadvantaged students.
Despite all the caveats, I think that giving school systems more flexibility in setting teacher salaries will only help education, even if it only helps a little. I think it is unlikely to lead to schools abusing their flexibility to drive down salaries. An interesting harbinger is that the American Federation of Teachers (the second largest teacher's union in the US) now supports merit-based bonuses. Will the National Education Association follow?
I should first say that I did not mean to say that the teacher's union is the only problem in primary and secondary education. I realize some might have thought that when I wrote "I think it is a travesty that the U.S. . . . trails most of the first world when it comes to primary and secondary education. I think the problem is the teacher's union." That was a bit strong! I meant to say that I think that's the biggest problem that might be amenable to change via government intervention in economic incentives.
Adrian mentions a number of things that he would list as bigger problems (with Leanne chiming in agreement): general public apathy, commodification of education, high-stakes testing, budgetary priorities, and curricula and licensure in teacher training. I admit that I am uninformed about the nature and extent of these problems. Perhaps I could induce Adrian to write a guest blog post?
Meanwhile Stacey (in the comments) and Topher (in an email) fretted about whether the pay experiment in DC will work. Stacey worries in particular about what system will be used to identify which teachers are meritorious enough to earn bonuses. I agree this is a sticking point. The obvious solution is to use standardized test scores, but this has inherent problems. Teachers may shift their focus from teaching critical thinking skills to teaching students how to memorize facts that appear on standardized tests. This is why the DC experiment is necessary: changing incentives often has unintended consequences. Topher (who knows lots about this stuff by dint of being an education professor) cited some scary data from Texas, which used two testing systems for a while: a national test and a state test. These tests gave scores that correlated very well with each other for years. When the state test only started to be used to determine teacher bonuses, students started scoring much better on the state test relative to the national test. That's not what we want! We want children to score better on standardized tests only if those scores correlate well with learning. When bonuses are linked to the scores, the teachers seem to find ways to improve scores without necessarily improving learning. Topher also points out that the higher education system that works so well uses NO standardized testing and LOTS of tenure. Education is a massive and sticky system that is truly impossible to recapitulate in a few simple equations.
Topher feels that average salary level (rather than the bonus structure for a given salary average) is more of a problem. He quotes a statistic: within 5 years after getting their teaching certificate, more than half of teachers quit the profession. It stands to reason, of course, that those who quit are those who would be the most successful teachers (and who would be the most successful at a variety of other endeavors, as well). Increasing average pay (which the DC experiment is also doing) should increase the quality of teaching by keeping those teachers from departing the profession.
Topher also note that interschool variation (inner city scools compared to suburban schools, say) is much greater than intraschool variation (due to variability in teaching quality from one classroom to the next). I wholeheartedly agree with this and suggest that this might respond well to economic incentives, too. If the average salary were set with that in mind, school systems could raise teacher salaries high enough to overcome the down side of teaching to economically (or otherwise) disadvantaged students.
Despite all the caveats, I think that giving school systems more flexibility in setting teacher salaries will only help education, even if it only helps a little. I think it is unlikely to lead to schools abusing their flexibility to drive down salaries. An interesting harbinger is that the American Federation of Teachers (the second largest teacher's union in the US) now supports merit-based bonuses. Will the National Education Association follow?
Sure, we could list all kinds of obstacles to better public education that are bigger than the teachers' union. How about the intelligence level of humans? I bet our kids would do much better if everyone were five times more intelligent. I'd rank that ahead of every problem Adrian listed.
ReplyDeleteBut that's irrelevant, because that's a parameter of the system, not a variable we can tweak. When we decry societal apathy, are we complaining about something that can realistically be changed, or are we wasting time whining about something over which we have little or no control? If someone has concrete suggestions for a set of actions to take to improve children's attitudes or parental support or whatever cultural attribute you like, please do, but in the absence of a plausible-sounding plan, I'd rather focus on things that might realistically be changed.
Topher brings up an excellent point -- our secondary education system is unquestionably the best in the world. I would suggest that it got that way primarily due to enormous quantities of choice and competition, not at the level of teacher vs. teacher, but at the institution vs. institution level. And despite my general love for free markets, I do need to acknowledge the important role of state universities and federal support in the form of student aid and bales of research money.
But I still think that choice/competition is the number one reason. And it seems that the natural approach to providing more choice would be voucher programs and/or charter schools. I'm not clear why you think results have been mixed; that article you linked to doesn't really say that. Regardless, we shouldn't expect any solution we implement, including vouchers, to produce clear results in a short period of time -- like you say, education is "a massive and sticky system" and as such it would be impossible to concretely evaluate any major overhaul on a short time scale.
(Oh, and I should point out that JMichael's original statement about the US trailing the industrialized world in quality of primary/secondary education is actually somewhat overstated. To the extent that it can be measured at all, the US generally falls slightly above the median in such rankings. I agree nonetheless that our primary/secondary education system is a pathetically dysfunctional calamity; the fact that many other countries also have lousy education systems is no excuse.)
Topher also makes a good point about standardized testing. It has some serious deficiencies. But if we're going to start giving families a lot more choice in which school to use, how do we expect them to evaluate the quality of their product? And even if we're not giving that much choice, the government (at whatever level) is going to need some way of telling success from failure. I cannot imagine how we could seriously attack the problem of quality of education without some way of roughly judging that quality.
Do we have any evidence to suggest that any of these standardized tests actually predict success, either in tertiary education or in the workforce? If not, do we know of any way of measuring anything that might correlate with that? If not, then what the heck have all the education people been doing for the last few decades? Don't we have some theories of what the purpose of education is, and how one might operationalize that?
Anyway, I don't know how much difference we can make with merit pay for teachers, but I do find it somewhat absurd that people would be so violently opposed to the idea of paying people more if they're good at their job. Most of the rest of society seems to manage OK under that expectation, including higher education (where the higher pay generally comes with job titles, tenure, bigger labs, more prestigious employers, etc.)
I'm curious as to what we should conclude from the statistic about half of teachers quitting within five years. I don't think it's at all obvious what that actually indicates. I certainly wouldn't leap to the conclusion that those teachers didn't find out what teacher salaries are like until they actually got their first paychecks. There are plenty of other possible reasons why the attrition rate might be so high. Even if salary is the primary issue, it might not be average salaries that are the problem as much as average starting salaries. I believe that teachers who have been in the same job for 30 years are usually making plenty of money; that salaries are driven by seniority seems to be another unfortunate consequence of powerful unions.
Hmm, hopefully Blogger will let me post a comment this long. Let's see....