An idea for performance pay in education: Guest Post by Avi Waksberg

Here is a guest post by Avi Waksberg. NG

Should we pay teachers performance bonuses for teachers based on standardised testing of their pupils? The teachers I’ve spoken to about this have invariably argued that it encourages them to ‘teach to the test’ whilst neglecting hard to test skills. In contrast, most economists I’ve spoken with favour some kind of ‘merit pay’, often raising promising examples from Israel (Lavy 2004) or Colorado (see de Grow 2007). However, imposing performance-based compensation upon hostile teachers seems a good way to ensure the approach does not work. (Chait 2007).

I am confident that people respond to incentives. However, teaching is difficult to quantify, complex and multidimensional. These are job characteristics that Dixit (2002) found tend to make performance pay less likely to be used. This leads to a situation where we focus on simplified metrics. But if incentives are not well targeted or the desired outcome accurately measured, then the response will be to overly focus on those aspects that are measured at the expense of other responsibilities. This is what underlies the complaint that teachers would be encouraged to ‘teach to the test’.

Moreover, there is no consensus on what constitutes teaching excellence. Is a teacher supposed to maximise: grades, enthusiasm for learning, clear thinking, university entrance, lifetime income, or life satisfaction? Goldhaber (2009) argued that while we may know little about how to objectively and accurately quantify teacher productivity, this problem is surmountable using merit pay programs with several evaluation components (such as Principal or peer evaluation, school-wide analysis, professional development and incentives for hard-to-fill skills and positions). However, any form of merit pay would still require the support of teachers, schools and administrators. Successful programs often emphasize collaboration and improvement rather than dividing teachers into ‘winners’ and ‘losers’.

I have a suggestion that, while incentive-based, seems to have registered a more positive response from teachers I have interviewed. Instead of rewards based on test results, I suggest offering prizes to teachers for posting excellent lesson plans to password-protected teachers’ forums (such as the Ultranet that is used in Victoria). This would potentially have the added advantage of encouraging a sense of sharing and collegiality among teachers, whilst not making them feel under-appreciated and judged in the way that bonuses based on testing can. If only teachers who are logged into secure forums can download the lesson plans, then it would be easy to simply record unique downloads (i.e. the number of downloads from different teacher logins as opposed to total number of downloads) and offer rewards to teachers for the most downloaded lesson plans in each year and subject. This approach has the advantage of addressing the tricky problem: what quantifies good teaching? In essence, we let the people who are best qualified to answer, the teachers themselves, decide what represents the best in their field.

Alternatively or additionally, we could include some kind of feedback / rating system that determines whom among teachers would receive bonuses, although it might be sensible to keep such a rating system private so that no one feels embarrassed if their lesson plans do not score highly. Another reward could be given to the teacher who had the most total number of downloads across all their lesson plans (i.e. a reward for putting many lesson plans up instead of the single most downloaded plan). One advantage of this proposal is that it is relatively inexpensive. A few rewards each semester for each subject in each year in each state of $1000-4000 would be only on the order of a few million dollars. The awarding of certificates or medals would help confer a sense of recognition and personal reward outside of monetary compensation. An ‘Australian Teacher of the Year’ award could be instituted for the top of the top. We could make a ceremony of it and celebrate our best teachers.

Prizes are by no means a new idea, but they are becoming more popular as a way of incentivising behaviour. In recent times the X-Prize has helped development of low-cost, commercial space flight and the Bill and Melinda Gates Foundation has used prizes to incentivise a range of health and governance projects. Because prizes confer psychological rewards that a weekly wage or even an annual bonus does not, people will pursue prizes with surprising verve. Incredibly, a prize underlies the existence of the Cavalier King Charles Spaniel breed of dogs. The breed had been allowed to die out but in 1926 a 25 pound prize was offered and breeders re-bred the Cavalier King Charles Spaniel back into existence.

Prizes are already used in the education sector. One of the teachers I contacted said that Promethean Planet, an interactive whiteboard company, offers prizes for high-quality content that customers share (http://www.prometheanplanet.com/en-us/). Tertiary education institutes commonly offer teaching awards too, such as the Monash Vice-Chancellor’s Award for Teaching Excellence, which uses a peer or student nomination system.

My expertise is not in education policy (or at least, no more so than anyone who has ever had to pass through an education system) and so I cannot claim to have stumbled on the last, great answer for improvement of teaching standards. However, I think this is an interesting proposal, and furthermore, it would be reasonably inexpensive to trial.

With low cost and potentially high reward, is there any reason why we shouldn’t be testing this kind of proposal today?

Avi Waksberg.

References:

Lavy, Victor (2004) Performance Pay and Teachers’ Effort, Productivity and Grading Ethics, http://www.nber.org/papers/w10622.pdf

de Grow, Benjamin (2007) Denver’s ProComp and Teacher Compensation Reform in Colorado, http://education.i2i.org/wp-content/uploads/2011/04/IP-5-2007_web.pdf

Chait, Robin (2007) Current State Policies that Reform Teacher Pay: An Examination of Pay-for-Performance Programs in Eight States, http://www.americanprogress.org/issues/2007/11/pdf/teacher_pay.pdf

Dixit, Avinashi (2002) Incentives and Organizations in the Public Sector: An Interpretive Review, http://www.wcfia.harvard.edu/sites/default/files/Dixit2000.pdf

Goldhaber, Dan (2009) Teacher Pay Reforms: The Political Implications of Recent Research http://www.americanprogress.org/issues/2006/12/pdf/teacher_pay_report.pdf

This entry was posted in Economics and public policy, Education. Bookmark the permalink.
Subscribe
Notify of
guest
28 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Stephen Bounds
Stephen Bounds
10 years ago

I am similarly skeptical of merit based pay for teachers for all the reasons that Avi outlines.

I’m also inclined to support Avi’s proposal but not on the grounds of establishing merit pay. As described, the system would have only a tenuous link to teaching skill in the classroom. However, it would provide a clear incentive to share teaching knowledge with others which could be expected to improve results across the board.

Having known a few teachers at the time of their first posting, it’s been striking how much work is spent in the first few years building up a portfolio of materials that the teacher is comfortable to use. After that, the teacher can “coast”, but of course the best teachers are continually revising and enhancing this portfolio each year.

Taking the prize idea one step further, I think it would be worth having a standard set of bonuses payable for teachers who contributed new material to a shared resource pool each year. Teachers could also be recognised for exceptional online forum presence and be “elected” to a paid moderator/energiser position for a period (eg a year) with an additional stipend paid to recognise their work. This would encourage ongoing engagement and set cultural norms.

john walker
john walker
10 years ago

As Darwin well understood, Sexual display trait Selection ; the selection of display traits needed to gain a chance, to reproduce your genetic information, is not always the same as selection for overall fitness in life (though it mostly is).

It is an interesting approach to paying for what is effectively a new kind of teaching ‘text’ book( a sort of Wiki). However teachers would be ‘selected’ for skill in lesson plan construction for adult teachers by adult teachers ; not necessarily exactly the same thing as selection for skill in inspiring young people.

I am not a teacher but I do remember that the teachers at Sydney boys (40 years ago) who were most inspiring were the ones who did not (visibly)use much in the way of ‘planed notes’ and the lest inspiring were the ones that had lesson plans the size of telephone books
I could be totally wrong… But,
Is lesson plan construction a good match with Classroom performance?

billie
billie
10 years ago

Many teachers fear that merit based pay will be used to pay some teachers, but the overall monies available for teacher salaries will fall.

The idea of providing small prizes for classroom teachers who upload lesson plans to share with other teachers has appeal. The mechanics of what constitutes a unit of work or lesson plan segment will be hotly contested by those who have carved out a professional niche in the field. The strength of this idea is the material would be developed by classroom teachers for classroom teachers as opposed to the current situation where on line materials are developed by people who left the classroom 30 years ago.

I agree with John Walker that the most popular teachers are popular by dint of personality rather than standards of preparation – often slack!

I agree with Stephen Bounds that beginning teachers struggle to develop a body of work that they are comfortable teaching. I don’t think that this idea will obviate that need to develop the body of work but having access to classroom tested lesson plans makes the job easier.

Access to online lesson plans as well as state approved text books helps provide a standard level of education throughout all schools in the system

billie
billie
10 years ago

Forgot to mention this plan would be dynamic, teachers with new ideas posting their plans online and teachers who like the new plans would be free to teach them, perhaps successful plan posters could spend a sabbatical of 1 or 2 years to administer the plan database.

Prizes create the incentive to keep uploading material to the database, after the intial load.

There are existing databases of lesson plans but they have idiosyncratic coding conventions of what constitutes a unit of work and their own vocabulary

billie
billie
10 years ago

Jacques, I can’t see anyone but compsci buffs using GNU Widjt, its worse than the offering I criticised

Paul Bamford
10 years ago

Maybe it’s time for policy makers to take a look on the extensive research on intrinsic and extrinsic rewards. Especially those studies that show that introducing extrinsic reward schemes (such as bonuses) in occupations where the primary motivator is intrinsic reward (e.g. “I like to teach”) are actually counter-productive.

john walker
john walker
10 years ago

Billie
“I agree with John Walker that the most popular teachers are popular by dint of personality rather than standards of preparation – often slack!”

I am a sort of performer ,the people who inspired me were those that understood that, in Hofstadter’s words, “content <is fancy form”. It is the movement that creates the form on the paper not the other way round. A degree of nonchalance is after all essential in painting the ‘big machines’.

billie
billie
10 years ago

Sorry John, I misunderstood you. I wsa thinking of the popular educators I know who spout crap. In fact I used to audit one such class so that I didn’t contradict my colleague. The students laughed that I visibly winced when the person in front of the clss was wrong. When teaching computer subjects there is a right and wrong – [compiler|computer] says “No”.

john walker
john walker
10 years ago

Yes the are a lot of professional educators around these days!

“When teaching computer subjects there is a right and wrong – [compiler|computer] says “No”.”
True ( my attempts years ago at amateur programing mostly ended in small train-wrecks)>
Mind what about Godel? True but mutually exclusive statements, in any system of representation that is complex enough to make statements about ‘itself’, are always a possibility, no?
Not every thing can collapse to a binary solution.

billie
billie
10 years ago

Ideally the teacher can perform and be an expert in the subject they are teaching. You can’t be a teacher a Victorian school classroom without being a registered teacher which means you have completed a degree and a teaching qualification, unless you are part of Teach For Australia.

Sometimes teachers teach subjects they are not expert in and they would benefit from being able to access prepared lesson plans which hopefully will stop them teaching the wrong thing.

Over the years I have heard stories about teachers learning subjects poorly and then they teach that material even worse than they learnt it!

Paul Frijters
Paul Frijters
10 years ago

agreed. Merit pay is too hard to do on the basis of statistics. We will see how the experiments with giving school principles discretion to reward particular teachers goes.
The proposed plan sounds like small beer – giving prizes to a couple of teacher internet gurus. We probably already do things like this in schools; we certainly do in uni.

Pedro
Pedro
10 years ago

Gee the teachers are good at my son’s private school. My daughters grade 1 teacher at the state school, not so much. It seems nothing can be done though. I wonder what is different?

Avi Waksberg
Avi Waksberg
10 years ago

Stephen, I agree that the link between shared lesson plans and a particular teacher’s classroom performance is indirect. The element I like best about paid incentives for shared lesson plans is that it encourages collaboration and improvement. I think that the collaborative tone of shared lesson plans is likely to appeal more to teachers than the competitive tone of test-based merit pay.

I like your idea about elected moderators for the forum. I recognise that the culture that develops on a forum sets the tone and nature of the discussion and could have a large effect on the contributions that are made. Widespread participation in the forum would be helpful to encourage teacher involvement in the program. I had not thought of incentivising a helpful and collaborative culture by monetary incentive but I like the idea of the sites users electing their own to officially recognised and rewarded positions. This needn’t exclude professional moderators and may be very cost effective.

Dehne Taylor
Dehne Taylor
10 years ago

It’s not very hard to determine who are good teachers and who are bad teachers. Anyone with any experience/knowledge of tested performance appraisal systems (and there are many of them out there) could spend half an hour to an hour in a staff room talking with teachers and get a fair idea who is, and who is not, a good teacher in a school (with the remaining 80% being adequate).

But teachers’ unions, academics and consultants are very uncomfortable with normative judgements (primarily because they have never been subject to formal appraisal mechanisms and see them as threatening), hence they claim that the only way to assess someone is via a formal, normative type involving some sort of scoreboard.

This is, of course, crap but sounds reassuring to those who have never experienced (and understood)a formal appraisal.

PS. The term ‘formal appraisal’ does not include the nonsense of feedback surveys from students and/or conference participants at the end of a semester/lecture/presentation.

Dehne Taylor
Dehne Taylor
10 years ago

Oops Second para should read:

But teachers’ unions, academics and consultants are very uncomfortable with normative judgements (primarily because they have never been subject to formal appraisal mechanisms and see them as threatening), hence they claim that the only way to assess someone is via a formal, normative positive type involving some sort of scoreboard.

john walker
john walker
10 years ago

Billie

The way in which teaching in itself became a academic qualification – I.e it became a theory of ‘teaching’- is a interesting topic .
In my profession it led to an redefining of what is a ‘professional artist’ from performance based title to a title based on theoretical qualification.

Stephen Bounds
Stephen Bounds
10 years ago

@Dehne – The big problem arises when governments try to standardise performance across schools, not just within schools.

How do you compare the top 10% of teachers at 2 different schools? They will have different student populations, parent involvement, financial support arrangements etc etc. This iis what unions object to, and I have to say with some justification.

john walker
john walker
10 years ago

We have limited funds for the supply of education and virtually no limits to demand for education ; really it is a political problem, how do you ration without looking like you are being tight?

john walker
john walker
10 years ago

Really do not want to start a war, Ok?
‘Education’ is also a major source of employment, education is also a compulsory item for the consumer, has a special status , not totally defined in economic terms and the qualification for public employment in education is qualification issued by teachers of teaching

Brawls between the teachers and public governments about whose ‘qualification assessments’ rein supreme: eternal.

Patrick
Patrick
10 years ago

Amen Pedro. There is a very simple reason why we just cannot work out how to work out who are the best teachers: we have shut the door on the obvious technique, which is let those closest to them make the call and make sure they have incentives to get it right.

Sometimes called markets.

You could at least let principals hire and fire their own teachers!!

billie
billie
10 years ago

In Victoria principals hire their own teachers. This a very time consuming process as the school has to put in place processes to demonstrate that the hiring process was correctly carried out and the best candidate ws chosen. It takes about 40 man hours of school time to hire 1 teacher. The teachers have to approach each school individually with a CV tailored to each school. The mandatory online application for any government school position in Victoria takes 2 hours to work through for each application plus generous download allowance from your internet provider. I reckon it takes central staffing about 5 minutes to allocate a teacher to a school. Schools near the NSW border have trouble attracting staff because pay is higher across the border and schools in remote towns can’t attract suitable teachers either.

Imagine teachers in NSW having to travel to schools in Moree, Broken hill and Menindee to drop off their CV on the off chance of getting a job.

In Victoria principals collect a bonus for staff savings that can be substantial especially for the principal of the 3000 pupil P12 school whose bonus took 5 years to buy his Porsche Carerra. Anecdotally some schools’ teachers are all obese, another school’s teachers all booze as well as the expected collection of fit young enthusiastic 30 somethings. There is nothing stopping a principal exclusively hiring members of their local church.

Paul Bamford
10 years ago

How about we adopt a more radical solution, based on a proven model employed in the tertiary sector – let the teachers elect the principal?

If you’re going to justify giving principals the right to hire and fire on the basis that, as highly experienced teachers, they’re best placed to assess aspiring and continuing teachers, then it’s reasonable to acknowledge that the collective knowledge and wisdom of school staff is the best guide to selecting school leaders.

Avi Waksberg
Avi Waksberg
10 years ago

Dehne, I am not sure it is easy to tell who is a good teacher and who is not by talking to them or observing them in the classroom. Linda Darling-Hammond (2010) is critical of teacher assessment by principals who observe classrooms in the US. She states that “[principals] differentiate little among teachers and offer little useful feedback”. It seems to me that helping teachers be better prepared can yield benefits at relatively little cost.

Darling-Hammond, Linda (2010) Evaluating Teacher Effectiveness,
http://www.americanprogress.org/issues/2010/10/pdf/teacher_effectiveness.pdf

Pedro
Pedro
10 years ago

Patrick, I think market techniques only work in markets. Frankly, I don’t think is there any way to be sure State principals can properly control teacher quality, the capture is just too easy, and where are the principal’s promoted from and how will you create the incentives for them to do it right. Same with performance pay, any scheme will largely turn into a rort. As I recall, the ALP policy a few years ago was all about teachers doing summer courses to earn their bonuses. It’s not hard to work out what that policy would lead to.

I agree with billie about the hiring as well. The process billie describes is an inevitable result of trying to put any discretion into the system. My first mother-in-law work for the Australia Post and Telecom Promotions Appeal Board. As dumb as that sounds, what else do you do when bad decisions are not punished in the market?

If your kid gets a dud teacher then there is nothing you can really do about it except move schools or wait till the next year. My girl has had to suffer through numerous temp teachers as her teacher has been “sick” at least one day a week on average. Complaining to the deputy has achieved very little. The school is generally very good, but we got the one dud teacher and that’s fucked year 1. At my son’s private school if there is a problem they are all over it like a rash. Just fantastic.

Patrick
Patrick
10 years ago

Sorry Pedro you are right, my idea only works coupled with a voucher system so that parents can choose the school. Ideally there would also be a more flexible approach to the curriculum.

Dehne Taylor
Dehne Taylor
10 years ago

Avi at 24 and Stephen at 18. In a round about way that is the point I was trying to make.

No system of (normative) assessment can be perfect because we humans aren’t perfect.

For example, it is highly likely that there is some person out there somewhere who could do each of our jobs better than we do so ourselves, but if we are subject to ongoing appraisals designed to improve our skills, it becomes less important over time (hint: it’s called productivity).

In my view, appraisal systems are not (and should not) be designed to compare capabilities across different schools. They should apply across an individual school and if the principal and senior staff are not competent to undertake these assessments in a meaningful way, I would assume good teachers would seek reassignment to a school that did so. Obviously, for very small schools, the appraisal system would need to be altered – even to the point of not allowing access to performance pay unless a previous track record existed.

Finally, I exit this debate (knowing how fraught it is with danger!) by noting two things (from experience).

First, our state education bureaucracies are composed overwhelmingly by teachers, many of whom are on some sort of long term leave from teaching due to various ‘pressures’. Unfortunately many of them have little experience, or desire, about how to provide useful policy advice under budgetary constraints.

Secondly, due to successful collaboration by education bureaucracies and teachers’ unions, first year teaching graduates are paid better than nearly every other professional in their first year of work. Naturally, to pay for these high first year wages, subsequent salary rises over time are muted. As such, after three or four years work, bright kids in other professions can be paid more than a high school principal with twenty years of experience who is responsible for more than a 1,000 students.

Dave
Dave
10 years ago

Here’s an idea .. maybe teachers should receive a pension or superannuation “top-up” in retirement that is calculated based on the earnings of students who were under their charge. Of course you’d only pay for “alpha” – outperformance that could be correlated back to the educational impact made by the teacher on the pupil. This kind of long-term incentive could certainly create alignment between the desired action (teachers “educating” their students to be prepared for the world around them) and outcome (an appropriately educated and prepared populace).

I believe it was Einstein that said (along the lines of) ‘if an idea does not seem preposterous at first, reject it’. This just might be one of those :)

Matt Elliott
Matt Elliott
10 years ago

I agree with Avi (24) about not being able to tell the good teachers from the rest through conversations in the staff room. From experience, I know many great people who are not necessarily great teachers. Also, some good teachers are only occasionally even in the staff room.

One of the problems with letting Principals evaluate teaching quality is that, just like teachers, there are good & not-so-good principals out there. Any evaluation process would need to apply to the school leaders as much as the teachers. In Victoria, the Department (of Education) is currently training groups of Principals to evaluate teacher quality in the class room.

In response to Paul (23), one of the issues with letting staff hire Principals is that schools can become stale through a focus on promoting internally. By allowing a rep from each of the stake holders in a school (staff, parents & Dept) the Principal appointment can cover the needs of all these groups.

One of the main problems with evaluating teacher effectiveness is working out how to measure it. When looking at student improvement or academic achievement, which teacher has the most impact? In a primary setting, there is one main teacher, but others have an impact (the music teacher, PE teacher, etc) In a secondary setting, a student has a different teacher for almost all subjects – which one has the most impact on a students’ learning?

While Avi’s suggestion has merit, as a (primary) teacher in Victoria, I spend most of my planning time developing curriculum for my students (the rest is taken up with admin duties – I won’t go there). Any searching of the internet for resources is usually done out of hours. So spending time to trawl through mountains of lessons uploaded by others, let alone finding the time to write and upload my own, will most likely reduce the teaching and learning in my class. Most effective resources are shared amongst teams of teachers, within a teaching level, across the school and with those in my network of colleagues.

Perhaps a solution isn’t to reward teachers directly based on their effectiveness, but to review the current system of pay. In Victoria (and mirrored in other states I believe) a teacher’s pay is almost exclusively based on the number of years teaching. Becoming a leader in the school puts you on a different (and slightly higher) scale, but the new scale is still based purely on years in the job. Maybe this should be modified so that there is more emphasis on the contributions a teacher makes to student learning, whether within a school or the wider education system.

Although most teachers in the state system would much rather see an increase in the funding allocated to schools to improve resources and increase the availablity of professional development, thus making our jobs more effective. While Dehne (27) is correct about beginning teacher salaries being higher than most other sectors, no teacher I know is in it for the money.