As my colleague Audrey Amrein-Beardsley pointed out earlier today, there is a battle royale coming up this week in New York, with a hearing on the changes in state teacher evaluation policy mandated by this year’s budget agreement in Albany. Amrein-Beardsley sees the issue focused on value-added measures, in part because that is the obsession of Governor Andrew Cuomo and a number of his policy allies in teacher valuation. I see things a little more holistically, because teacher evaluation becomes much more interesting when you broaden the scope a bit.
In January, New York Governor Andrew Cuomo took contradictory positions on the question of teacher evaluations and due process for weak teachers. On the one hand, he proposed changing the teacher evaluation process to give more weight to student test-score algorithms–wanting to push the proportion of evaluation based on test scores up to 50%, in contrast to many states that are now backing away from such heavy weighting of value-added statistics. At the same time, he pushed for a reversal of the usual due-process presumption that administrators have the burden of proof on dismissal hearings of experienced teachers.
I suppose you could argue that teaching is the target of both proposals (and a number of others, such as lengthening the time that teachers have to spend before being eligible for due process protections). I would argue that the proposals are hopelessly contradictory. One is based on trust of administrators: give them the presumption of correctness in dismissal hearings. The other is based on complete mistrust: administrators are giving too many teachers satisfactory ratings, so we must put greater algorithmic weight on something administrators do not control (test scores).
In school reform, distrust is now the rule. The motivation behind No Child Left Behind’s 100% proficiency requirement was distrust of states by federal policymakers in 2001 (both Democrats like George Miller and Republican advisors to President George W. Bush). If you do not trust others to act, insert a formula and make them act. With both the Republican wave of 2010 in many states and the Obama Administration’s Race to the Top grants, many states then applied value-added measures to teacher evaluations using a formulaic approach, because principals could not be trusted to evaluate teachers. Many states decided they would replace the widget effect of uniform satisfactory evaluations with the widget effect of one-size-fits-all algorithms.
This is technocratic school reform: if we could just create the right mechanism, we could make the watch, sit back, and see schools improve. It is policymaker as Watchmaker, or to those familiar with religious history, a belief in Deist school reform.
Alas, Deist school reform attempts to use technocratic mechanisms to solve a fundamentally human problem: most principals do not like to give teachers poor evaluations. Most (other) businesses have a similar dynamic: managers do not like to give poor evaluations to the people they work with every day. Like other middle managers, many principals would much prefer to counsel weak teachers out of a school rather than give a hard kick in the formal evaluation process. When teachers have due-process protections, the psychological burden becomes harder: a principal has to provide a formal negative evaluation and also ensure that the follow-up is objective and fair and documented properly even while providing day-to-day supervision. To do that unaided, a principal has to be able to compartmentalize to a degree that is tough for most human beings.
The fundamental problem here is that the dynamics of teacher evaluation are typically geared towards a default of “everyone’s pretty good” (aka satisfactory). The attempt to impose formulaic evaluation including test scores is the Deist attempt to change the dynamic, but there are alternatives, and it is far better to focus on alternatives. If value-added scores kick teachers out of a job, you will never give principals an incentive to make hard decisions; it will always be the fault of value-added measures (or whatever replaces value-added measures). The most important reason to avoid a formulaic approach is because it is critical to develop good principals who can make tough decisions and do so fairly.
There is an alternative to Deist school reform: address the human and political side of the evaluation and personnel decision-making process. While the Joel Klein-era NYC Department of Education took that too far with a (largely unsuccessful) team of tenure assassins, also known as the Teacher Performance Unit, the target was the correct one: the human side of dismissing teachers. Similarly, current NYC Chancellor Carmen Fariña is focusing on principal skills in addressing teacher incompetence by her personally urging principals to take command. I think Fariña’s approach is also problematic, because it assumes that her charisma can inspire principals to have a bit more spine. I am sure she is charismatic; I doubt this solves the human problem of evaluating close colleagues.
If I am disenchanted with Big Apple’s attempts to tackle this problem, what are the alternatives? That is the subject of another blog entry, later this week or next week.