Shortchanging Teachers

Shortchanged, a TNTP report released last week, takes a harsh look at current “lockstep” teacher pay systems, which reward teachers for time in the classroom and advanced degrees rather than actual performance. The report argues that these practices pushes out high performers and incentivize poor performers to stay in the classroom—with costly consequences: TNTP estimates that last year alone, districts spent $250 million on automatic pay increases for ineffective teachers.

TNTP proposes new teacher compensation systems that focus less on years of experience and master’s degrees and instead focus on actual teacher performance. Research on teacher quality peaks offers other reasons to support this argument.

Research shows that teachers develop the most in their first few years of teaching. After three to five years, though, most teachers peak. So to a certain extent, schools can predict early on how effective a teacher is going to be for the rest of his or her career. Yet under current policies schools must continue paying ineffective teachers the same automatic raises as highly effective teachers, year after year. This creates an incentive for poor performers to stay and for high performers to leave – which they do. TNTP’s own research shows that 40 percent of teachers with more than seven years’ experience are not as effective as the average brand new teacher.

Compensation structures in most other professions are designed to reward employees for logarithmic growth – they assume that employees will make large gains in the beginning of their career then taper off later on. Doctors and lawyers, for example, quickly ascend to peak earnings in the first ten years of their career, then plateau at that salary level for the next ten to 25 years. With teaching, it’s the opposite. Teachers’ growth is ignored when they’re actually improving, and they’re rewarded after they’ve plateaued. As TNTP points out, that creates a whole host of problems.

–Ashley Libetti Mitchel

6 Replies to “Shortchanging Teachers”

  1. I think it would be difficult to measure many of the variables that influence student performance. I defenitly agree that ineffective teachers should not be receiving that automatic pay increase but the manner in how effectiveness is measured can still become problematic. I mean in the sense that it can become a double edge sword, because the teacher’s effectiveness can be impacted tremendously by the students that he has, If we are simply basing it on performance. If effectiveness is based largely in student growth and not standardized assessment than I think it would be much more effective.

  2. There are two important mischaracterizations of the research which severely undermine this post.

    First, the post quotes TNTP’s report The Irreplaceables, stating, “TNTP’s own research shows that 40 percent of teachers with more than seven years’ experience are not as effective as the average brand new teacher.” TNTP’s research does not show this at all. Rather the report found that, in two districts studied (District A & District C), this was true. The stat appears, stated as this post’s authors does as if generalizable to the entire nationwide population of teachers, on page 13 of the report. But the reference (endnote 25) clearly reference the available data from two very non-randomly selected samples. You cannot generalize from 2 districts to 10,000, or from less than 3% of teachers to the nationwide pool, unless you are engaging in intentional random sampling. This is simply bad statistics. Moreover there is very good reason to believe that these two districts represent outliers and are not near the nationwide mean. For example, endnote 25 states that in District A, 57% of teachers with 7 years experience or more (i.e. more than half) have value-added scores lower than an average first year teacher. How many districts have you been in where the average 7th year teacher is worse than the average first year teacher? I’m not sure I’ve experienced one. But frankly this, like the over-generalization quoted here is pure conjecture. There is simply insufficient data here to make any nationwide claim. The problem here is in TNTP’s claim, but the author of this post quoted it seemingly without fact-checking it by reading the endnote and hence is propagating an important unsupported claim.

    The second mischaracterization of the research is simply a misquotation of data in the TNTP report. The post states, “Yet under current policies schools must continue paying ineffective teachers the same automatic raises as highly effective teachers, year after year. This creates an incentive for poor performers to stay and for high performers to leave – which they do.” There is no evidence in this report (nor anywhere else to my knowledge) that poor performers stay while high-performers leave. According to figure 6 on page 5 of this report, in each of the 4 districts studied, retention rates were HIGHER for high-performing than for low performing teachers. Or in short, the data here suggest the opposite of the author of this post’s claim: high performers stay, while low performers leave. But to be more precise, while high performers stay more, both groups here stayed at high and quite similar rates. So perhaps best put would be, high performers stay and low performers do too. In TNTP’s own words from the same page, “Most schools retain Irreplaceables and low performers at strikingly similar rates.” This mistake, unlike the first, is entirely attributable to the author of this post.

    Pointing out these two mischaracterizations is not mere quibble about a footnote. These two non-facts, one without data to support a generalization, the other with data supporting the opposite generalization, are used as evidence for a policy prescription that is itself without evidence. Specifically, the author claims that current systems of lock step pay incentives reduce the retention of the best teachers, while increasing the retention of the worst teachers. The solution therefore is to do away with such systems of compensation. The problem however is that the evidence is quite the contrary. Generally speaking current systems of financial incentives (and other factors) seem to encourage the encourage the retention of ALL teachers. This also seems more in line with common sense. The policies are the same across teachers and seem to have similar results across teachers.

    For the record I should clarify that I do not support current salary structures in education, and am generally supportive of the “smart retention” policies advocated by TNTP in their Irreplaceables reports. But strong policy analysis, and accurate policy prescriptions must begin with accurate characterization of the evidence.

    In this case, the picture painted by the author of this post, is misleading and frankly insulting. Basically the author implies that those who stick around in teaching are generally the mediocre to poor teachers. We are led to believe that, by means of an anti-filtration system, once length of tenure is inversely related to one’s performance as teacher. This is simply not true. So far as either this report can suggest, or a wide variety of experience suggests, those who stick around in teaching are actually, mostly everybody.

    It is unkind to paint an inaccurately negative picture of those people who have decided to work, over the long haul, to care for and educate other people’s children. We should not deride those who are willing to make a sacrifice we are not. Having taught only 4 years myself, my current work in education policy feels like a vacation compared to the work I did actually dealing with live children with real and pressing needs, day in day out.

    But beyond the derogatory attitude implied by these demonstrable inaccuracies, the broader policy implications are left without supporting evidence. The follow up report to the one cited, focused on Washington DC’s Irreplaceables, suggested that DC’s policies (which involve both performance pay and more aggressive firing policies), did not actually increase the retention of its highest performing teachers. They simply decreased the retention of its lowest performing teachers. With some caveats, I would say that is actually a helpful feat, because the worst performing teachers are often actually harmful to kids and other teachers, so decreasing their retention is generally a good thing.

    What DC’s case did not show is that performance pay meaningfully increased the retention of high-performing teachers. Contrary to the (unsupported) claims of the author of this post, the lockstep salary schedule (while not something I value), does not seem to incentivize high-performers to leave. And there is no evidence that I have yet heard of (nor any offered here) that supports the claim that offering value-added score based performance pay leads to greater retention of high-performers. If such evidence exists the author of this post, and supporters of such policies in general, would do well to produce it. It makes a sort of intuitive sense that it might, but many education ideas that make a sort of intuitive sense don’t produce the intended outcomes (see for example California’s multi-billion dollar effort to reduce class-size which actually seemed to lower student achievement). That’s why a rigorous evidentiary standard from policy analysts is so important and why the mischaracterizations of the evidence we do have by the author of this post are so troubling.

  3. TNTP is run by a fellow who believes Michelle Rhee took her 63 of her 70 students from the 13th percentile on the CTBS to the 90th percentile.
    It is like giving credence to words from an Obama birther.

  4. I wholeheartedly agree with Ray above. I agree that it would be hard figure out who was a “good” teacher or a “bad” one based on just standardized tests. Teachers, in general, are some of the lowest paid professionals in the United States, shouldn’t they get some kind of increase every year just based on Cost of Living? Teachers that get advanced degrees are paying out of their own pocket as a way of making more money to feed their own self and families and make themselves better teachers, we shouldn’t take that away. Maybe we could base an actual raise on the teacher’s impact on the increase of student’s learning or growth over time, that would something I would be behind. Thank you.

  5. I agree with both Ray and Jenny. It would be very difficult to figure out who was a good effective teacher versus an ineffective teacher just by the use of standardized tests. Not every student will come to school everyday feeling or wanting to do his or her best. We can not say that the teacher is a “bad” teacher because the student doesn’t want to learn. Teachers are some of the lowest paid professionals who are always finding ways to advance themselves through that same low paying salary. In my school district there is an incentive pay out at the end of the year based upon attendance, student growth and evaluations which are done by administration. The largest portion of the incentive pay out is the student growth from the beginning of the year until the end of the year. The student growth is measured based upon standardized test given each quarter. I strongly believe maybe an incentive payout program like this would be a great way to distinguish effective versus ineffective teachers without actually effecting anyone’s salary. An incentive payout to motivate the teachers and students to work hard I believe would be a great approach to effective “good” versus ineffective “bad” teachers.

  6. Systems ought to devise various compensation mechanisms to cover the every persuasion of professionals contained in the workforce. In the educational realm, there are indeed a catalogue of expectations to meet and standards to uphold.

    To address the varied conditions of the teaching workforce while ensuring that incentivization processes are left intact, administrators are compelled to institute policies that cover the varying qualifications and set benchmarks for continuous workforce growth.

    Teacher effectiveness should hinge a lot on performance, and not qualifications, but examining the process of “effectiveness” more holistically, leaves a lot for consideration. Impact of “out of school” experiences on students’ performance is often left out of the effectiveness equation. Additionally, there are a list of intervening conditions and situations that might undermine or support student performance not directly open to teachers contribution.

    Howbeit, the rewards framework employed by any institution, must out of necessity, cover every angle of the employment spectrum, so much so that, no teacher feels underserved and that there is room for “climbing the professional ladder.”

Leave a Reply

Your email address will not be published.