[setgree]

I saw this discussed on Marginal Revolution today as well.

My main question is whether the gains are fundamentally zero-sum, at least in the short term. Some districts implement flexible pay and some don’t, and then the best teachers move to get paid more. And the places that get left behind…?

In the long-run, increased pay ought to lead to more high-quality teachers entering the profession. But in the short term, this scheme seems likely to redistribute talent in ways that reinforce existing patterns of inequality.

[masklinn]

An other more direct issue is Goodhart’s law: if teacher compensation is linked to a specific metric, then that metric is what you’ll get. If the metric is test scores, then you’ll see:

- teaching to the test rather than educating

- trying to get rid of left behind, slow, or difficult students (already an issue for generations in test-oriented private institutions — as opposed to the more remedial “last chance” ones)

- ignoring the groups which have the lowest odds of contributing to the metric (which groups it is depends on the weighting / averaging between pupils)

[daniel-thompson]

and cheating

[snarf21]

I agree that this sounds like a different kind of voucher/private school resource transfer. Private schools are able to self-select their students and aren't required by law to take ALL students that want to come. This makes it trivial to be both cheaper and show better test results while at the same time removing resources from the students that actually need help.

The problem is that people in power thinking that this is mostly a teacher driven problem. This would be the same approach as saying that a the amount of money a doctor gets paid should be based on how much weight their patient loses (e.g.) Anyone who thinks teaching is so easy should go try it for a month. This kind of approach also misaligns incentives. The fact is we have good teachers and bad teachers, just like every profession. We need to rethink two things for our education system. Is it to just push kids to college or is it to maximize the number of students who can become gainfully employed and self-supporting adults. Then we need to restructure our resources towards that goal, specifically in terms of class size and ability level and extra help. The main change we need to make is to move investment of resources much more towards elementary age where it can have the longest compounding benefit. (Source: I'm from a family of a dozen current and retired public school teachers)

I do agree that if we increase the pay and benefits for ALL teachers, then we'll create incentive to get better quality teachers in the profession and that competition will get better results in the long term.

[slibhb]

> But in the short term, this scheme might judt redistribute talent in ways that reinforce existing patterns of inequality.

Aren't you assuming that rich districts with higher grades implement flexible pay? I think it would be the poor districts with worse grades that would be most likely to want to shake things up.

[masklinn]

The main characteristic of poor districts is that they’re poor, which means they don’t have the means to implement high pay schemes. Rich districts do.

[dwater]

The school district I worked in had impoverished students but high per pupil funding, which is pretty common. What reinforced the inequality in that district was that teachers who were able to, self-selected to have the best behaved students, who were most likely to see improvements in test scores. The teachers who ended up with 90% of their students starting the year multiple grade levels behind and with severe behavioral problems would rarely stay in the classroom more than 2 years.

[kspacewalk2]

You think of it in terms of inequality, which is one perspective. However, if seen from the perspective of the best behaved students, the ones who want a positive learning environment that will maximize their talents and success, this situation seems greatly preferable to the alternative, which is a sprinkling of problematic students disrupting the learning equally in all classrooms and making overall outcomes more mediocre. This is Ontario's public school system lately, in a nutshell.

[llm_trw]

Regardless of district the US spends more per student than most other countries.

The issue is that this gets eaten up by bureaucracy and teachers are left living in poverty, hence them clinging to unions like a drowning woman to a straw.

[fnordpiglet]

Additionally it’s a side channel way to simply pay most teachers less and some teachers more, saving money in cash strapped districts. Accountants and MBAs love that sort of stuff.

[jackcosgrove]

> this scheme seems likely to redistribute talent in ways that reinforce existing patterns of inequality

Only if school districts that implement flexible pay are also those with higher SES indicators. Which could very well be the case, although the study did try to control for these factors.

I do agree that unless you increase teacher pay vis-a-vis other professions, the effect will mostly be to rob Peter to pay Paul.

[dmix]

My sister is a very talented leader/organizer who went to teachers college and would love to teach but the pay isn't good enough for what she could get elsewhere.

Yet she still obsessively talks about how bad the teaching systems is here in Ontario, Canada and follows all the studies.

I could 100% see a large increase in talent/quality jumping into the system and taking over the culture. Or at least heavily influencing it. The gaming for tests is also probably a short term effect as well, the old guard doing what it knows best, just more of it.

But Canada will be the last place this sort of reform would happen. The unions completely dominate discourse.

[SiempreViernes]

This does definitely seem like a setting with a large risk for zero-sum outcomes, as the current state of the art seems to only be able to identify the bottom percentiles and then have them discarded. So current understanding hasn't identified why anyone became a good teacher, so we don't know how to make the bad ones better.

This is a problem because even a bad teacher is very likely better than no teacher, so we can't actually discard very many of the bottom teachers before making things worse.

[kspacewalk2]

>And the places that get left behind…?

Think of it as government-performance based teaching quality variations.

>But in the short term, this scheme seems likely to redistribute talent in ways that reinforce existing patterns of inequality.

Only if disadvantaged districts are less likely to switch to performance-based pay. What is the basis for thinking this will happen?

[setgree]

see sister comment: https://news.ycombinator.com/item?id=40012447

[iLoveOncall]

Well the whole point of such an experiment would be to generalize it.

[erehweb]

Sure, but if the gains in an experiment come from cannibalizing other districts, then it's not a good guide to what will happen when everyone's in the experiment. Do you attract more people to the profession? Or do people get discouraged by seeing variable outcomes they can't predict ahead of time?

[iLoveOncall]

Given that we have the whole private sector as benchmark for your questions, I'm pretty sure the outcome would be positive.

[throwaway894345]

How would it reinforce existing patterns of inequality if low-income school districts move to a flexible pay model, thereby attracting better teachers?

[dacryn]

good performance and putting in extra effort should be rewarded over those who just do their job at the minimum effort levels.

That's a fundamental problem with teacher pay, there are no bonusses, there are hardly incentives or room for raises. It's all just tenure based. Once you're in, you're in.

[fnordpiglet]

Does standardized test score outcomes measure this behavior or does it favor teachers who teach test taking and have less challenging demographics in their classroom?

Performance based pay in professions whose performance is difficult to measure directly leads to bizarre outcomes. I’m not disagreeing in general but I’ve seen this again and again in my career, and I spent a long time on Wall Street where bonuses -really- matter. That extreme brought out the extremes in how incentive pay distorts behaviors in unexpected and undesirable ways, which gets worse the further you get from a directly measurable outcome like PNL.

As a parallel example, hospitals who specialize in extremely difficult diseases with high fatality rates generally have abysmal patient outcome metrics over hospitals that punt anything complex to a specialty hospital. This plays out in policy spaces punishing the speciality hospitals despite the fact they are well known to be top of the industry in terms of performance and “extra effort.” Nominally though they should be shut down by all metrics.

This absolutely plays out in education. Not every classroom or school or district is equivalent in terms of the challenges they deal with. A teacher with a very challenged class who is a high performer and puts in extra efforts will be punished simply because their baseline was much lower than a teacher who punches the clock in a class of affluent students who have private tutors.

[thaumasiotes]

> there are hardly incentives or room for raises.

There are lots of incentives. If you get a master's degree, your pay automatically goes up!

Does that help anyone but you? Well, it helps the school of education that you paid for your degree. Does it help anyone else? Of course not.

But that doesn't mean there are no incentives. We stated what we wanted, and we got it. We have more incentives than we need.