Another point…teachers who plan together get the same bonuses to encourage collaboration.

-p

What a clear-eyed and pertinent list of priorities! I’m glad Gotham Schools highlighted your rodent comment.

As a former district public school teacher, I would probably re-order your list in the following way:

1. Stronger leadership (this is what you are really saying in number 3, I believe).

2. improved professional development, especially focused on effectively using data and reaching the most struggling kids

3) more time to collaborate

4) smaller classes

5) Rodent reduction (I agree with the commenter, impossible right now in Manhattan)

On the subject of rodents, there was a really compelling NY Times Magazine lead story devoted to it about five years ago. The lesson: the only way to remove rodents is to remove the food source. The poisoning the city continues to do only spoils our own living environemnt and makes the colony stronger by eliminating the weakest links. Completely removing the food sources will starve the entire colony.

I’ve got a thick skin, but really: no call for the personal digs. Parenthood even? Sheesh. This site hasn’t needed moderation to date for ad hominem attacks, let alone a “comment policy,” and I hope you’d agree it should stay that way.

I would note that I went on to point out in the next sentence — in good faith — better uses of time for any child’s education and development than test prep, and, that I proof-read the above comment to make sure it didn’t read like I was suggesting any REAL teachers would do such a thing.

Last, I note that your comment above eloquently (and with no recognition of the irony) details the real-world pitfalls of a VAMP system. It didn’t even occur to me that teachers would in real life undercut each other. Call it Reason #11.

Sincerely curious:
1) Was your school public, private, or charter; and
2) Did the test scores move that year?

Cheers.

And after the individual bonus was scrapped after that one year, the collaborative spirit returned.

In terms of sandbagging the fall test … that’s a pretty dirty comment. I hope you don’t expect that an even merely decent teacher would do something like that, let alone a good teacher. That’s the kind of attitude that has gotten this country into a hot economic mess. I’m glad you’re not a teacher, Michael M, though disheartened that you’re a parent.

Perhaps – I hope – your comment was somewhat tongue in cheek, and I just don’t get your particular brand of humor.

At my charter school in Boston, we get a merit bonus that has two parts. The first is a school-wide bonus and is given based on how well we help our students meet various achievement benchmarks. The second is teacher-specific and is tailored to different metrics for different grade levels and classes. We have not yet implemented value-added, but are interested in finding a system that works…

Reason #9: How would VAMP handle known achievement gap issues for various demographic subgroups?

Reason #10: How would VAMP handle pulling out of the baseline the nominal fact that ELA scores (at least in District 2) have been going up PRIOR to VAMP? What if the rate of pre-VAMP improvement slows down post-VAMP? Dock everyone’s pay?

Reason #7: If I’m a teacher, I have the kids SANDBAG* the test in the fall. And maybe I don’t tell ‘em, exactly. But during fall test prep, we all go to a museum, or the FDNY training center on Randall’s Island (yes, they allow schoolkids as visitors), or, we go panhandling for our own “hold harmless” enrichment and intervention funds.

P.S. Dan makes an excellent case for asking Tweed employees to take “psychic pay” only. (Kidding.)

* “To downplay or misrepresent one’s ability in a game or activity in order to deceive (someone), especially in gambling: sandbagged the pool player by playing poorly in the first game when stakes were low.” — TheFreeDictionary(dot)com

I take it that you mean that you are agreeing with the points I am trying to make in my first six paragraphs, right?

If that is the case, why do you support merit pay? Do you believe that in spite of all of these problems that merit pay for teachers would have beneficial effects? If so, can you explain how you think that that will work?

To your first point, I’ll do you one better. I agree 100% with the first 6 paragraphs of your post, and the only point where I think we diverge is that I actually think teachers, because we don’t get paid as much as Wall Streeters, might be more sensitive to bonus-pressure. That is, the size of our bonuses are closer to determining whether we can afford essentials. If I get a merit bonus, I’m not hiring a second butler, I’m getting the transmission fixed.

I do agree with you that by nature and by training teachers are less money-focused than businesspeople generally are. But I also think that businesspeople aren’t as driven by money as most of us think, and teachers are more like other people in other industries than most of us think.

NYC Educator, would you take a rodent-free building over a really big bonus, say $50K? You could buy a lot of those buzzing mouse deterrents for $50K.

Yes, I’d take a rodent-free place to work over merit pay.

1) It’s nice to see you write, “Ceolaf is right.” I feel like I’ve accomplished something.

2) Do you believe that there is a sweet spot where bonuses will be large enough to incentivize changes in teachers’ behavior, but not so large that that it would incentivize competition/stifle cooperation? (I don’t mean teachers going into each others’ classrooms to mess with kids’ learning. That’s too obvious. I mean too much noise in the hall, undermining during lunch, filling their heads with garbage to make the next steps harder to teach, etc..) How would we find that sweet spot, if it even does exist?

3) Have you seen value added analyses of scores in your school(s)? If not, how do you know that with the appropriate adjustments that good teachers are responsible for high scores every year?

4) No, we shouldn’t be afraid to use ideas from Wall Street simply because they are from Wall Street. But we should be able to think carefully about whether particular ideas are good ideas for education.

In this case, it seems they there are some big lessons about merit pay bonuses. Wall Street gives bonuses based on short term success (i.e. profits) and does not take them back if the deals that led to those profits generate losses down the road. Many have written that this is one of the primary ingredients of the sub-prime crisis that has made this recession so bad. That is, Wall Streeters were incentivized to book deals with a high likelihood of massive long term-losses, so long as they also had a high likelihood of short-term gains. (In fact, some have even written that they *must* have known how risky these deals were (long-term), and therefore there are elements of fraud.)

A system that incentivizes short term goals is going to shift attentions and efforts away from long term goals. In fact, Prof. Willingham lists that as one of his six reasons in this piece. I think that we can see from Wall Street that this is a real problem, not some the some theoretical concern. That just leaves us with the question of whether there is a reason to believe that teachers will not do what Wall Streeters did.

One of my repeated concerns about merit pay is that it seems to assume that teachers will respond to financial incentives like business types do. I don’t think that that is the case, in which case it is a waste of time and attention. However, if I am wrong — as proponents of merit-based bonuses — then don’t we need to worry about them acting like business types? (Moreover, merit-based bonuses are potentially unlimited in on Wall Street, because they were based upon increased profits for the firm, which made the money available for bonuses. However, better teaching does not directly generate profits or even revenue for schools, meaning that the pool for bonuses is inherently limited, meaning that teachers would be in greater competition for it than Wall Streeters were.)

As to the points Ceolaf says are more compelling to people like me who generally support merit pay, I’m not terribly concerned that teachers would start undermining each other. Maybe if the amount of variable pay were really, really high some of them would, but for a couple thousand dollars I don’t see my colleagues coming into my classroom to mess with my kids’ learning.

To me, and perhaps not to most of my colleagues on the pro-merit-pay side of the debate, my concerns are primarily a) does merit pay work to benefit kids? and b) are we defining “good for kids” right? That is, are test scores sufficient (no) and do bonuses motivate (unclear)?

I’ve seen great teachers get great scores year in and year out, and bad teachers get bad scores. I simply don’t buy that truly great teachers will frequently be “outperformed” by not-so-great teachers. The existence of whole schools that use great teaching to get outstanding results every year seems to me to belie the notion that tests aren’t sensitive to teaching. So Ceolaf is right that I don’t buy that argument.

I find myself arguing a position other than my own. As it happens, I entirely agree with you and love your presentation style. But what I am uncertain of is whether it would be convincing to those who do not already agree with us.

I think that those who are in favor of merit pay don’t expect it to be 100% fair, or need it to be. If they are right about bonuses acting as a motivator for teachers, all they need to hold onto merit pay is for it to be strongly correlated with good teaching. That is, if the people who get the bonuses are almost entirely good teachers and virtually all of the best teachers get bonuses, they would be satisfied.

Value-added models are good enough for them, as they don’t need 100% accuracy. The kinds of issues you point to are — to them — technical issues that we can fix over time, but don’t need to wait on. The fact that some of the better teachers won’t get bonuses doesn’t impress them because they are astronomically more concerned about students test scores than about fairness to or treatment of teachers. No, the wouldn’t be favor of a random system, but value-added is not — in their view — random.

I think that there are other issues that might be more compelling, but I am not certain how well you can explain them in your little presentations — which I *love*, by the way. For example:

* This kind of system puts it in teachers interests to undermine students’ performance before they get to them and after they leave, because kids who do worse the year before and/or after they in a particular teacher’s class would make them look better. It also would put it in teachers’ interest to undermine other teachers, with whom they’d be in competition for a limited/finite bonus pool. Does anyone believe that the path to improved schools is for teachers to undermine each other and each other’s students?

* Tests have been designed to measure student progress, and checked to see how well they do that. We’ve spent decades developing those abilities, in industry, in academia and in our schools. However, we don’t know how to design tests that are “instructionally sensitive.” It might appear that measuring student progress on tests translates to measuring teaching, but it turns out that that might not be the case. When we look by this method, we find wild swings in teachers rankings from year to year, which seems to violate everyone’s expectations about teacher quality. (This suggests that it might be random, or at least have a very large random component to it.) (Yes, this is *strongly* related to the point you make in your comment, but it is not a point you made in your video — though it is related to your third reason, the possible lack of an interval scale for teaching.)

I actually agree with your conclusion, and your reasoning. My concern is whether you are, in effect, preaching to the converted. I think that your last reason (peer effects) might get the most traction, and your 5th reason (short v. long term) is the most foundational, but one I’ve had trouble getting any traction with.

I’ve spent years trying to get people to think a bit more carefully about merit pay for teachers, as you probably have. I don’t offer these ideas to disagree with you, but rather to help make the case.

I didn’t address questions about whether all important aspects of teaching can be captured by standardized tests, or whether teachers respond to financial incentives or the others you mentioned because I think those questions are open to debate–honest people can disagree about them. I don’t think that a cogent argument can be mounted that using standardized test scores is fair. To me, it’s self-evident that if it’s not fair, it shouldn’t be used. We might be able to think of other measures that would motivate teachers and improve test scores–for example, tell a group of teachers “If test scores don’t go up, we’ll randomly select 1/4 of you to be fired.” That might make test scores go up, but I wouldn’t advocate for it. (I’m not suggesting that you would either, ceolaf. I’m just making the point that group increases are not the only criterion we should pay attention to.)

Who was I trying to convince? People who didn’t know much about value-added measures, and to whom the problems might be news. I, for one, have not seen the reduced-reliability-due-to-correlated-scores problem discussed much.