Why Pay More for Experience and Advanced Degrees?

Incentives are a powerful motivator of human behavior.  And because they are so powerful, we both use and encounter incentives almost daily.  Our email in-boxes are filled with incentives to shop right now (“save 50% today only”).  Governments have established “sin taxes” to discourage the use of alcohol and tobacco.   And as parents we use all kinds of incentives to encourage the right behavior in our children (not all of which are effective!)

Incentives at Work

In employee compensation systems, incentives are foundational.  Sales people earn commissions for selling.  Employees work hard not only out of pride or as part of the employment ‘contract’, but also to earn a larger annual merit increase.   Companies ask employees for cost-saving ideas, and pay out some of the savings to those who contributed the most profitable ideas.  Annual bonuses are used by more than half of all private sector companies, and two-thirds of federal workers are awarded bonuses.  Even the IRS pays out bonuses.

Why are incentives so prevalent?  Because they work.   If a certain behavior is desired, incentives are put in place to encourage it.  This is why almost all organizations use incentives to encourage employees to contribute toward strategic goals.

Incentives for UCF Teachers

We saw in my previous post that UCF teacher pay increases due to three factors:

1.  Longevity
2.  Graduate education
3. Across-the-board increases to the salary schedules.

On one level, these incentives are working very well.  UCF teachers generally stay in the profession – about 30% have maxed out the longevity incentive.  Anecdotally, I am told that it is rare for teachers to leave UCF for ‘greener pastures’.  And UCF teachers aggressively pursue additional education:  87% hold at least a master’s degree, and more than half have earned a master’s degree plus 60 additional credit hours, maxing out the education incentive.

So the incentive system seems to be working — we are getting ‘more’ of what we are rewarding — longevity and advanced education.

But what about student performance?   After all, isn’t the primary mission of our schools to give a great education to our students?  We do not pursue long-service, highly-educated teachers as an end in itself, but rather because (presumably) well-educated teachers with lots of experience make better teachers.  And better teachers will produce higher levels of student achievement.

So what does the research say?  Is there a link between student achievement (the desired outcome) and years of experience (longevity) and post-secondary education of a student’s teachers?

Teacher Experience and Student Achievement

Before I started to read the research, my intuition was that teacher experience would be positively related to student achievement:   the more years of experience a teacher has, the better teacher she would be, and this would show up in student performance.   After all, I can’t think of too many professions where lack of experience is an asset!  But what does the research show?

A 2002 review of 243 prior studies showed that less than half of the studies found any statistically significant relationship between teacher experience and student performance.  Of the ~45% of studies that did find a relationship, almost all found a positive relationship (more teacher experience leads to higher student performance.)

Teacher Experience

These studies also showed that the benefits of experience have diminishing returns.  The largest benefits from teacher experience were realized in the first year.  And at least half of the total returns on teacher experience seem to occur in the first 5 or so years.  This makes some sense — it is the classic exponential learning curve, or perhaps another example of the so-called 10,000 hours rule  (about 6 years of teaching) which says it takes 10,000 hours of deliberate practice to achieve expertise in a chosen field.

So my intuition about teacher experience was not quite right.  Yes – there might be returns to experience, but the evidence is not that strong (remember- more than half of all research studies show no relationship exists).  If there are returns to experience, they come in the first few years and have quickly diminishing returns.

Here is an excellent one page summary of the current research on teacher experience, from a national non-profit.

Teacher Post-Secondary Education and Student Achievement

When I first learned that teachers earned more as they continued their own education, it made sense.  Additional education should make a teacher more effective in the classroom, Graduate Thesisright?  How could additional eduction be a bad thing?  If that’s what you think, you are in for a surprise!

The evidence on teacher education is unequivocal — additional education makes no difference (and may even be a negative).

A 2002 meta-analysis of 170 prior studies showed that only one-third of the studies found any statistically significant relationship between teacher education and student performance.  And more importantly, in the highest quality studies only 9% found a relationship, and the relationship was negative (masters degree holders are less effective than teachers with only a bachelor’s degree.)

Teacher Education

More recent studies have continued to replicate this result.  And because there is so much evidence piling up, there have been calls from many quarters to eliminate extra teacher pay for masters degrees (see an example here) and North Carolina has recently done it for all teachers in the State (see here).


So our current pay system is not well-aligned with student achievement.   Masters degrees do not make teachers more effective in the classroom.  And if experience does make a difference, it appears that most of the gains occur in the first five years.

As one researcher put it:

If a rational system of teacher compensation, aimed at recruiting and
retaining high-quality teachers, were designed from scratch, it is unlikely
it would bear any resemblance to the system that is currently in place.

-Professor Michael Podgursky, University of Missouri-Columbia

Perhaps we should not be surprised — the step & lane pay system is a product of the early 1900s, and was developed for a different era, and was constructed with different goals in mind.

In my view, the step and lane pay system is outdated.  The incentives encourage teachers to pursue degrees and remain in the profession, neither of which seems to drive student achievement. I don’t blame teachers for responding to the incentives … if I were a teacher, I would do the same thing.  But I think teachers, parents, and tax-payers alike should recognize that the current system is not aligned with the organizational goals of our schools, and that we are devoting resources and energy to things (such as night and weekend courses) that are not helping us achieve our goals.

Teaching is a complex and challenging profession.  In other challenging professions, we recognize and reward excellence through our compensation systems (and through many other means).  If an organization wants ‘more’ of something (be it sales, cost-saving ideas, or excellent customer service) it rewards those who contribute to those objectives, and it does so with differentiated pay, bonuses, and other monetary and psychic rewards.

Is teaching so different as a profession that it is immune to the power of performance incentives?  If we want our schools to help all students grow academically and achieve their potential, shouldn’t we have a pay system that identifies and rewards those who do it well?   Shouldn’t we have incentives to encourage great teaching?

(Note:  the same argument applies to incentives for principals and administrators, almost all of whom have no linkage between their compensation and student achievement.)


  • Incentives are widely used to align employee efforts and behaviors with organizational goals
  • In public schools, we pay teachers for longevity in the profession, and for completing additional post-secondary education.
  • There is strong evidence that neither longevity nor graduate education has much to do with achieving our district’s educational mission:  helping all student’s reach their academic potential.

Agree? Disagree?  Leave a comment!

7 thoughts on “Why Pay More for Experience and Advanced Degrees?

  1. FoodforThought

    I’ve found reading your posts informative, but you only cite sources like TNTP. This group is founded by Michelle Rhee, a person at the forefront of the privatization movement of our public schools. I don’t find any information from sources other than those. If you are truly looking to do a service to your constituents don’t you think presenting the entire picture is justified? This post completely ignores poverty, home environment and family. You’re asking teachers to face all the consequences of student performance without actually addressing the factors outside of the classroom that impacts students. If you take away poverty and just compare our student performance with those around the world in the same socio-economic categories, our kids perform just as well globally. What Americans face is huge amounts of poverty (approx 23% of US kids live in poverty, 1 in 5) which is factored into our student performance profiles, so when you put UCFSD kids against DASD kids, they are very similar, but compare UCFSD kids with kids in William Penn, it’s much different. I haven’t read all your posts, but in the one i commented on a month ago and this one, no reference to these additional factors on student performance. Hope you can what this: https://www.youtube.com/watch?v=hf9UVg-TdH0&list=TLLzLPDV2lebvPx_xR-Wu15ay-oQSTYtyR

    1. Bob Sage Post author

      Thanks for the link to the video, which I did watch. I agree with your point that our constituents need full information on the issues. And I appreciate you providing some feedback which will help me make the blog better. Let me try to address two specific points: (1) Sources I cite; and (2) How non-school factors like poverty and the educational level of parents impact student outcomes.

      If you look under the “Research” heading, you will see links to the primary sources that I have read to produce my posts. These are peer reviewed research studies from credible sources. There are about a dozen that I have read and written up, and about a dozen more that I have read but not yet written up. And then many, many more I have not read and not written up. (I have a full time job!) I have not chosen the research studies to support my opinion. Rather, I had no opinion on most of these subjects until I read the research. That being said, I think the research I have studied tends to be cited by the so-called education reformers. But I also provide links to other opinions in my posts. An example of this is my most recent post “Are Board Certified Teachers Highly Effective”. At the end of the article I provide links to research on both sides of the issue, and I link to a teacher’s opinion which states that NBCT process was totally worth it (which runs mostly contrary to the conclusion of the post). I would love for readers to provide citations to relevant primary research in the comments. (I am not too keen on citations of opinion pieces, from any side of the debate.)

      On to your point about poverty and other non-school factors… It is widely known and accepted that the socio-economic background, parental education, and level of parental involvement are hugely important to student outcomes in the classroom. Most estimates indicate that 70-80% of student achievement can be predicted by these factors. And as you said, that is a large reason why UCF performance is much better than at William Penn. The State of Pennsylvania uses PVAAS scores in their School Performance Profiles for that very reason — the PVAAS methodology (and value-added methods in general) controls for those non-school factors statistically. PVAAS leaves us with an estimate of school-specific results (the 20% of student achievement that is not due to socio-economics and parents). Almost all of the research I cite utilizes this same VAM as a research methodology in order to factor out socio-economics so we can study the things we can control, like teacher training, class size, quality of principals, access to technology, etc. From my position as a School Board Director, I cannot do anything to impact who comes to UCF. But what I can influence is what happens in our schools when the students walk in the door. So I am on a quest to make sure we are doing everything we can (within financial constraints) to give the best education to students who walk in the door at UCF. I am not trying to solve the problems of William Penn or education policy for the USA or respond to PISA differentials between Finland and the US. That is Arne Duncan or somebody else’s concern. I am laser-focused only on UCF. And according to the research I have read, the quality of teaching is the single most important determinant of student achievement, after controlling for socio-economic factors, level of parent education, and all of those critical demographic factors. So that is my motivation for the Rewarding Great Teaching series. We need more Great Teaching. It is not a criticism of our teachers, many of whom are already great teachers. Rather I am trying to discover what can we do to attract, reward, recognize, and retain Great Teachers so that we provide the best education to our kids.

      Keep the comments coming … even if you disagree. It helps me learn and to become more effective in my role on the school board.

  2. former quaker

    You should also look at the diminishing marginal returns to the money… look at the PVAAS scores compared to budget – do the dollars spent mean more incrementally? Where do the incentives stop making a difference? A great example is Lower Merion. They’re well beyond the hill of spending for more performance.

  3. Bob Sage Post author

    I think PVAAS / total budget is a great way to measure performance for a school. It captures the most important output of a school (how much annual learning is occurring) compared to the inputs required to accomplish it (total spend). If I had to pick one metric to measure improvement in a school, that might be it.

    Unfortunately, the published PVAAS metric (PDE website) is expressed as % of students meeting a growth expectation. Unionville High School, for example, is at 100% PVAAS in all three measured subjects. But what is the underlying average growth per student? 1.05 years? 1.1 years? Seems like PVAAS as reported publicly will not reflect any future growth improvements at UHS, since we have maxed out the metric. What data do you have for Lower Merion, and where did you find it?

    1. former quaker

      If you’re at the growth expectation, then you may question the incremental budget increases. Why would you go from total spend of $76M to $80M the following year and what does it achieve? What is the outcome? I’d want specifics to the penny, however old timers on the board will be miffed…
      In LM, I reviewed SAT/ACT scores since they are publicly available – although don’t have the value-added-ness… also, total student body / total budget to compare to a similar district. We were spending ~$21K/student but achieving lower test scores then other districts with lower spend rates. It puts the onus back to you business office to better justify YoY increases despite the “sky is falling” millage rate syndrome or property value losses. PVAAS has much more data available privately – especially since its on a SAP ERP Platform. It takes some expertise to draw the data out though (but the proximity to Newtown gives you an advantage I’d assume).

  4. Bob Sage Post author

    If UCF could achieve higher levels of learning (growth) and if could do so while improving the PVAAS / cost per student ratio, I would certainly advocate for it. In our district, I don’t think the expectations are for meeting a academic goal, but to exceed it (within a budget).
    But this is really a theoretical debate because the right measures are in place to relate student growth to expenditure.
    Regarding LM, I think you have to be careful comparing achievement per expenditure. There are many non-school factors that influence achievement (see Food for Thought’s original comment and my response). You can look at LM’s performance trend to see how it is doing, but I think not easy to compare LM against UCF (for example) on achievement metrics like SAT. Growth is a much better indicator for those comparisons, which I think you agree with. Problem is getting the data.
    Could you explain further how to obtain PVAAS data … sounds like you believe there is a way to request it / obtain it?

    1. former quaker

      SAP’s Enterprise Resource Planning software captures quite a bit of data. PVAAS has private access accounts for mining the data however you should have some experience with SAP’s ERP Platform to leverage this. It looks to me that SAP sells various levels of this service to districts. Or you could hire your own local expert, then create the private account and have the bare bones service… it creates a conundrum, since your building new costs to measure costs, is it really worth it? Hold budget constant, while increasing/maintaining # of students and increasing performance? Impossible! lol BTW, I left LM… but still pay property taxes on a home there.

      [removed comments about another board member’s campaign website]

      In LM, there was this group:
      LM when I was there: http://febp.newamerica.net/k12/PA/4214160
      Compared to you: http://febp.newamerica.net/k12/PA/4224210

      Finally, I’ll add this – teaching certifications also raise salaries and create barriers to entry however the certifications do nothing to increase teacher quality. [“Teacher Certification Raises Salaries but not Quality,” NBER Digest, August 03]

Comments are closed.