Response to Eric Hanushek

Ed Note – Yesterday in a post below Eric Hanushek argued in favor of removing the lowest-performing teachers. Today, Diane Ravitch responds below.  Hanushek will respond tomorrow and Ravitch again on Thursday.

By Diane Ravitch

I had the pleasure of working alongside Eric Hanushek for a decade as a member of the Koret Task Force at the Hoover Institution. During that time, we had many opportunities to discuss these issues.

Rick claims that we are not eliminating enough teachers; that if we fired the “bottom” 5-10% of teachers, our national performance would rise to the top of the world. He assumes that a “low-performing” teacher would be replaced by an average teacher, thus leading to a drastic overall improvement in student outcomes.

Nobody disagrees that there are ineffective teachers and that, if they are unable to improve, they should be removed. But let’s clear a few things up.

When Rick says “research shows” that removing the “bottom 5-10” percent would generate dramatic results, he is referring to his own calculations, not to any actual program that has ever been tried. It’s a hypothetical extrapolation, and it consists of removing teachers based solely on test scores. Rick says he doesn’t advocate for purely test-based dismissals. But the dramatic improvements he foresees are based solely on students’ test scores.

Even in a perfect world where all the assumptions of his calculation hold up, we have very little idea of what the distribution of teacher effectiveness looks like when that effectiveness is measured using an evaluation system comprised of multiple measures. Different measures, such as observations and value-added, measure different things. The “bottom 5-10 percent” on one measure are not necessarily the “bottom 5-10 percent” on another.

At best, Rick’s calculation illustrates that there are more and less effective teachers, about which there is little disagreement. But there’s a difference between trying to show that teachers differ in their abilities and saying that firing people based on a criterion that nobody supports will produce huge results in the real world.

Rick “assumes away” many of the biggest issues. He seems confident that brand new evaluation systems can identify the “worst” teachers in a manner that is accurate enough for us to fire them based on those scores. A teacher ranked “ineffective” one year might be “effective” the next year, and this might happen often enough that we end up firing many good teachers.

Rick assumes that the teachers who are fired will be replaced by “average” teachers. But that is just an assumption. They are even more likely to be replaced by teachers in their first year of teaching, inexperienced teachers who need intensive support and training. Attrition in many high-poverty schools and districts is already at a critical mass. How will these schools manage under a system of ongoing, annual layoffs?

I would like to disagree as to the larger idea that firing teachers is the key to fixing U.S. public education.

The problem, Rick believes, is that we are not firing enough teachers. The problem, as I see it, is that we are not doing enough to recruit those who are well prepared and then supporting them once they are in the classroom.

At present, we have a very inefficient means of recruiting teachers. In many states, people may become teachers with little or no preparation, and they get little support once they are in the classroom. In addition, attrition is huge. Overall, some 50 percent of new teachers are gone within five years. In poor and minority schools, the percentage leaving may be even higher than 50 percent.

Among the 50 percent who leave teaching are surely many who were fired or who realized that they weren’t good at teaching. Rick’s own research finds that the teachers who leave tend to be less effective than those who stay. But many other good teachers leave because they are discouraged by the working conditions (lack of supplies, mentors, poor leadership, physical conditions, etc.)

I think we should take steps to raise entry standards, provide strong on-the-job training and support, and improve the retention of those teachers who make it through the first three-to-five years.

It is hard to see how Rick’s proposal would improve education or the teaching profession. Why would people line up for a job that is poorly paid and under constant attack, where working conditions are often abysmal, where the teacher is often expected to buy essential supplies out of her own pocket, where pensions and healthcare are being whittled away, where class sizes are rising, where librarians and social workers are laid off, and where support is minimal or nonexistent.

Elsewhere, Rick writes that (pdf), “in the long run, it would probably be superior … to develop systems that upgrade the overall effectiveness of teachers.” He worries that efforts to improve the quality of teachers, such as mentoring and professional development, have not thus far been successful, so he prefers what he calls “deselection.” I don’t think we should give up on efforts to improve the profession.

Given the already high turnover of teachers, given the low respect that teachers today command, given the need to develop a large number of well-prepared, skilled, experienced teachers, I maintain that we should improve the recruitment, support, and retention of teachers. We should aim to build the profession, not to make it more insecure and less professional. Other nations have improved the teaching profession by strengthening it, not by annual firings. So should we.

59 Replies to “Response to Eric Hanushek”

  1. To a point, a better question is why the so many teachers are leaving jobs they worked so hard to qualify for. And really, truly understanding why, not simply referencing “data” from an overly-simple exit interview.

  2. Beatrice,

    My answer will only apply to my school system and state. I work in NC, which is a non-union state. In other districts or states, the rules will vary considerably.

    But in my school, I do not have the power to hire or fire, only to make recommendations to human resources (although HR usually agrees with my recommendations). If I have pretty significant documentation that a non-tenured teacher is underperforming, then I can work with HR to have that person’s contract not be renewed at the end of the year. This can be a one-year process, but could also turn into a multi-year process. Documentation typically comprises classroom observations, feedback from parents, and documentation of non-classroom responsibilities. Throughout the process, I also have a responsibility to be doing what I can to provide the staff member with feedback and to put support structures in place to help the person improve.

    For a tenured staff member, the process is much longer, and the level of documentation needed more significant. I would say that the ultimate removal of a tenured teacher—absent something really big, such as hitting a student or having a sexual relationship with a student—is a process that would last at least two years, and likely more than that. It is a process that also consumes a pretty considerable chunk of time, has ramifications for building culture (both positive and negative), and is pretty emotionally difficult for both the administrator and the staff member because they are continuing to work together for a significant period of time.

    I am sure that other administrators have different experiences and rules.


  3. I’m sure that many of the posters on this blog had the same experience when they attended school: Some teachers were great, some were mediocre, and a few were downright bad, yet somehow we all went on to college and to successful careers as adults.

  4. I find it interesting that people can go on and on about teacher quality and teacher effectiveness (which are extremely important on a micro-level, but yes as someone who has taught in two different school buildings on both the primary and secondary level, it is also highly contextual) but no one really wants to address the LEADERSHIP gap. How convenient! What happens in buildings when the vast majority of teachers are not only competent, but “highly effective”, work together outside of school for no pay to collaborate, trust and support each other but have to overcome the ineptitude and cluelessness of a principal who barely spent time in the classroom. The same principal who gets to collect the kudos and the bonafides when we are miraculously able to succeed without support and communication?
    To above commenters on the issue of removing “bad” teachers,I am sure that this can be a challenge. But I will tell you, it’s even more of a challenge when the vast majority of the teaching staff in a school has no respect for the professional opinion of a principal. in addition, the quality of the leadership has a direct impact on building tone and instruction. I have visited many classrooms and you could tell just by walking through which leaders were inspired and which ones were simply collecting a check. But this is seldom addressed.

    On the other hand, I do agree with one poster who said the bigger issue is sometimes “lazy” teaching rather than incompetency. This is what irritates my colleagues and I more than anything else- teachers who know how to “turn on the charm” for observers when they do very little with the students. I will not pretend as though I know all of the answers to that one, but I think increased informal walk-throughs by both admin and other staff could help cut down on that one.

  5. as a south african i was struck by, and sympathetic to, both Hanushek and Ravitch. Our trade unions in south africa (SADTU in the main) ensure ‘cadre deployment’ where union politics trumps professional qualification… so a little bit of hanushek’s hard edged ‘deselection’ is attractive… ravitch at the same time presents a coherent account of all the unintended consequences of getting rid of the bottom 5%, assuming we could identify them… and the need for better attractors and holders for teachers… but as an outside reader i was most struck by the polarized presentation of the debate… extreme positions presented to generate the most response… rather than more nuanced attempts to hold the possibilities of both… that said… these kinds of debates that present cogent arguments that lucidly distill difficult issues into bite sized arguments needs to be encouraged in SA

  6. Should our efforts and money be spent on an evaluations system that WILL CHANGE THE WAY TEACHERS TEACH and not for the better.
    Since the advent of high stakes testing in the 1990s, we have seen a documented decline in creativity among students.

    Despite the rhetorical pairing of ‘creativity and achievement,’ one drawback of emphasizing the latter may be a lack of the former. In other words, creativity may be curtailed by the very data-driven policies that equate achievement with the constant measurement of student performance on standardized tests. The paradoxical outcome of 30 years of test-driven school reform is that it may have made the US less competitive in terms of international education.

    One aspect this is the reported drop in creativity. According to Kyung-Hee Kim, presently at William and Mary and formerly a high school and middle school English teacher in her native Korea, IQ scores in the US continue to rise, but creativity, as measured by the Torrance Tests of Creative Thinking, have been in decline since 1990. The irony is that, while it is fostered by US culture, “creativity is not emphasized by the U.S. educational system.”

    Her overall framework, which began with examining Confucianism in the Korean context, suggests that “culture is more influential than creative strategies and skills for fostering creativity in individuals.”(See her website)

    This does not discount the influence of schools; as I suggest elsewhere, the emphasis on testing indicates a significant change in US culture. Since the 1990s we have had a test-driven, data gathering model. Again there is a long history, but while the most respected theorists on education say ‘Play, play, play’ to promote child development, currently we test, test, test. Tests of English Language (not literature) and Math predominate, something that was consolidated under NCLB.

    Some schools eliminated recess in order to add test preparation sessions. Tests, which are used before we have established they are accurate, determine the fate of schools, principals and teachers, as well as students; this determination not only depends on these evaluation schemes in development, but often precedes diagnosis and the use of the data by teachers.

    So avoid this evaluation scheme — we’ll all be the worse for it because
    teachers will be afraid to teach creatively.

  7. By the way, I am currently writing on this subject and would appreciate hearing from those who want to discuss it.

    There are some very sharp thinkers commenting here.


Leave a Reply

Your email address will not be published.