Study Finds Perfomance Pay Does Not Affect Teacher Motivation

A new study released this month in Educational Evaluation and Policy Analysis found that pay-for-performance programs do not affect teacher motivation. The article, “Incentive Pay Programs Do Not Affect Teacher Motivation or Reported Practices: Results From Three Randomized Studies,” looked at three schools that were testing pay-for-performance programs. Metro Nashville Public Schools took part in this study. The study was conducted with five researchers from the RAND Corporation and professors from University of Southern California and Vanderbilt University.

The abstract of the paper lays out the major findings of this project:

“This study drew on teacher survey responses from randomized experiments exploring three different pay-for-performance programs to examine the extent to which these programs motivated teachers to improve student achievement and the impact of such programs on teachers’ instruction, number of hours worked, job stress, and collegiality. Results showed that most teachers did not report their program as motivating. Moreover, the survey responses suggest that none of the three programs changed teachers’ instruction, increased their number of hours worked or job stress, or damaged their collegiality.” (emphasis mine)

When reading the article, the authors do a great job of explaining the three main rationales behind pay-for-performance.

  1. Performance pay will improve student achievement by motivating teachers to improve or innovate their teaching practices.
  2. Performance pay will improve student learning by changing the work environment of teachers.
  3. Changes the supply of teaching candidates and retain high performing teachers.

The research took place at three different schools systems.

  1. Project on Incentives in Teaching (POINT) – Metro Nashville Public Schools
  2. Pilot Project on Tea Incentives (PPTI)- Round Rock Independent School District- Texas
  3. School-Wide Performance Bonus Program (SPBP) New York City Public Schools

The researchers wanted to answer two research questions:

  1. Did teachers find these three incentive pay programs to be motivating?
  2. In response to the implementation of these programs, did teachers report changes in their practices or their working conditions?

 

Results

-The majority of POINT and SPBP teachers agreed that rewarding teachers based on student test scores were problematic because those scores did not “capture important aspects of teaching performance.”

-Half of POINT and PPTI teachers said that they believed teachers were limited on what they could do because family environment played a larger role in student achievement.

-A little over 40% of POINT teachers and 20% of PTTI teachers reported that the chance of a bonus would energize them to improve their teaching.

-“In addition, the majority of incentive eligible teacehrs in all three programs reported that their programs had no effect of teaching, 85% POINT, 78% in PTTI, and 90% SPBP.”

We see that these pay-for-performance programs won’t change how teachers teach or even motivate these teachers to change their teaching styles. The majority of teachers who were participating in this program thought standardized test scores were a bad way to measure the bonus and half of the teachers believed home environment played a bigger role than teachers. If teachers don’t agree with the measurement, they won’t agree with the program.

The authors believe that the way pay-for-performance is designed right now is not the best.

“The lack of program impact on teacher’s practices suggest that more careful thinking about the logic model of incentive pay programs is necessary.”

The authors suggest that based on this study and others with weak effects, that policy makers should be looking at other ideas of reform.

If bonus-based policy is pursued, policymakers need to recognize this lack of evidence and take steps to monitor program implementation and evaluate program impact on targeted outcomes.

We know that some people are trying to bring pay-for-performance to Tennessee. Will this latest research slow them down? Doubtful, but at least we can show these people the research and open their eyes to some programs behind pay-for-performance.

 


 

 

 

Pay for Performance Coming to Tennessee?

Sen. Dolores Gresham, Chair of the Senate Education Committee, and Rep. Glen Casada (somewhat curiously, since he isn’t on the House Education Committee) have filed legislation to create a pay-for-performance system for teacher pay in Tennessee tied closely to the new evaluation system.  The key here?  Compensation would not based directly on value-added scores, but rather on how a teacher does on his/her evaluation (which itself is composed of at least 35%, but up to 50% value-added scores)

Senate Bill 0827/House Bill 0619 establishes a new compensation system based on a “maximum base salary schedule” to be set by a district, plus performance-based “salary adjustments” and “supplements” for various other things (teaching in high-need subjects, teaching in high-needs schools, taking on leadership roles, etc.)  The broad strokes are that this bill would mandate that local Boards create new salary schedules adhering to certain minimum requirements and restrictions (the most important of which is that pay for years of experience or tenure is absolutely barred).

A few important things to get out of the way first:

(1) This applies to licensed teachers.  The new salary schedule would apply to “instructional personnel,” which is defined as “any person with a license to teach in an LEA” under state rules and regulations but excluding substitute teachers. This would include counselors, librarians, etc.

(2) This applies to new hires, rehires, and new teachers.  The new salary schedule would apply to new hires in 2014 and beyond OR personnel “returning to the district after a break in service without an authorized leave of absence” (for example, parents who take a year or two off from teaching to stay home with young kids) OR personnel “appointed for the first time to a position in the district in the capacity of instructional personnel.”

(3) Anyone can opt-in, but there’s no going back if you do.  Any “instructional personnel” can opt-in to the plan, but “any employee who opts into the performance salary schedule may not return to the grandfathered salary schedule.”

With that out of the way, here’s the way the new pay schedule would work:

(1) A Board would establish a base salary schedule with a cap (the “maximum base salary”).  The Board has a certain amount of freedom to do this: (1) for opt-in folks, it will be their salary from the previous year, plus up to a 5% cost of living adjustment and (2) for new folks, it will be whatever the Board wants.

    • It is important to recognize the difference between salary adjustments versus salary supplements.  Salary adjustments ratchet forward — they increase the “base salary” of the employee (see section (c)(2)(B): “The base salary under the performance salary schedule for instructional personnel shall be recalculated each year to include the prior year’s salary plus any salary adjustments earned by the employee.”  Salary supplements are one-time payments that must be earned year-by-year.
    • Here’s the crucial point: Once you reach the “maximum base salary,” you’re no longer eligible for future salary adjustments, only salary supplements.
    • However (and this is a pretty big “however”), a local Board may “recalculate a maximum base salary schedule each school year, as needed.” (section (c)(3)).

(2) The Board then establishes its salary adjustment.  There’s not a lot of specificity as to what these would be, but presumably it’s an across-the-board compensation bump for those who qualify.  The “salary adjustment” comes with requirements designed to get folks to opt-in to the new payment system:

    • Each “salary adjustment” under the performance plan must be greater than the available step-raise under the old plan (section (c)(4)(A)).
    • Each “salary adjustment” can be no less than 10% of the starting salary under the old plan (section (c)(4)(B)).  In Nashville this would be a minimum “salary adjustment” of $4,000 (10% of MNPS’ $40,000 starting salary).
    • Teachers of tested vs. non-tested subjects cannot have different schedules or salary adjustments (section (c)(4)(C).
    • Salary adjustments are only available to teachers who receive a 3, 4, 5 on their evaluation (no teacher who scores “below expectations” or “significantly below expectations”) (section (c)(4)(D)).

(3) Finally, the Board establishes salary supplements.  These also come with requirements designed to entice  teachers into high-need schools, high-need subjects, etc.  (Note: Many systems, including MNPS, already offer some or all of these types of bonuses (sometimes referred to as “combat pay”)).  Salary supplements are to be available for the following reasons, and only to teachers scoring a 3 or above on their evaluation (i.e., “meets” or “exceeds expectations”):

    • Teaching in  a “Title I eligible school” (MNPS has 122 of them)
    • Teaching in a school in “restructuring” or “reconstitution” status (meaning a school hasn’t made “adequate yearly progress” under No Child Left Behind for at least 5 consecutive years)
    • Teaching in a “critical teacher shortage area” as defined by the State Board of Education (usually this is Math, Science, and Special Education, among other areas).
    • “Assignment of additional academic responsibilities,” presumably up to the local district.  MNPS has the ASSET program, under which (again, presumably) participating teachers would be eligible for a supplement under this part for taking on leadership responsibilities.

That’s the basic structure: Base salary + salary adjustments (up to a cap) + salary supplements = total salary.  There are some obvious methods to try to entice current teachers to opt-in (e.g., salary adjustments MUST be great than the existing step-raise under the old plan), as well as some efforts to get teachers allocate themselves where they are needed (e.g., supplements for high-need subjects, Title I schools, etc.).

The restrictions:

    1. Low-evaluated teachers aren’t eligible for raises/supplements.  Any teacher who receives a 1 or 2 (below or significantly below expectations) is not eligible for either a salary adjustment OR a salary supplement.  This means that if you go teach at a high-needs or Title I school, you don’t get the salary supplement just for being there.  You still have to get a 3 or above on your evaluation (which, given the way things worked out last year, doesn’t appear to be that difficult).
    2. Low-scoring teachers get reimbursed for professional development for the following year.  Any teacher who receives a 1 or 2 (below or significantly below expectations) “shall be provided professional development reimbursement for the year following the evaluation,” capped at $1,000.  This is in line with the structure of the new evaluations, which are supposed to provide targeted feedback, coaching, and PD to teachers who aren’t doing well.
    3. Cost of living adjustments are permitted, but capped.  These would adjust the base salary.  This provision seems to be a bit redundant given the freedom of the Board to recalculate the base salary yearly, but I suppose it’s supposed to operate in tandem because cost-of -living adjustments are capped at 5% of annual salary AND 25% of the annual salary adjustment available (which means cost-of-living adjustments are capped at whichever is less).
    4. Advanced degrees, except content/certification degrees, cannot be used in setting salary adjustments or supplements.  This appears to be a compromise provision.  In essence, the provision is an effort to get away from raising salary for any advanced degree, whether it relates to teaching or not (i.e., the mythical “underwater basket-weaving” Master’s Degree, earned online solely for the salary bump).  As the State Board and Commissioner argued recently, however, advanced degrees and years of experience are not correlated with increased student achievement as measured by our value-added and evaluation system. This last point is important.  To believe that years of experience and advanced degrees don’t, by themselves, lead to increased student achievement, you must believe that our current value-added model accurately captures whether students are learning, because that’s the data on which the conclusion is based.  There are studies on both sides of this point, but there is generally a consensus that (1) years of experience do increase student achievement early on, and to a point (see the last 1/3 of this post) and (2) some advanced degrees can help; others don’t.
    5. Any pay based on years of experience or tenure is absolutely barred.  No real other way to say this: “A local board may not use the length of service or tenure of any instructional personnel hired on or after May 1, 2014, for the purposes of setting salary, adjustments, or supplements.”
    6. Budget cuts can’t be directed disproportionately at the new salary schedule.  If a Board has to deal with a tight budget, it is not allowed to put the majority of the cuts on the new compensation system.

Some thoughts, though there will be a bit more analysis/discussion later: There are certainly other compensation systems that are even more closely tied to test scores, with the same base salary + adjustment + supplement regime.  In other states (e.g., Colorado), some supplements are explicitly based on a teacher’s value-added scores, a school’s value-added scores, etc.  This kind of direct pay-for-performance was examined by our own Peabody College, and found, experimentally, to be ineffective at raising student achievement.

Rather, this is a pretty tight fit with (and big investment in) our new evaluation system.  Rather than paying strictly for increased value-added scores (as some reform advocates would like), the new compensation system would weigh heavily on the outcome of a teacher’s evaluation.  Given that the State is adjusting the evaluation system as well, to decrease the prevalence of 3, 4, and 5 scores, the success of this new compensation system will depend largely on the success of the underlying evaluation system, for good or ill.

Note: For further reading, the Comptroller has a pretty extensive recent report on alternative salary schedules.  Disclaimer: I haven’t read the whole thing yet.

Edited to include item number 5 under “Restrictions.”  An earlier version of this post included this information, but it was inadvertently deleted in the final version.