Overconfidence in the Value of Measurement
If there's one thing that defines today's accountability movement in education, it's the sanctity of measurement. Reformers relentlessly demand hard data that students are learning. Without objective evidence, they claim that schools will never improve.
They point to business as a model. But despite what is widely assumed, numbers are rarely the sole consideration in evaluating employees in the private sector. When they are, they result in collateral damage in the form of depersonalization throughout the ranks and a false sense of assurance in the executive suite.
The latest to address the issue was Samuel Culbert, professor of management at the UCLA Anderson School. His op-ed in The New York Times on Mar. 1 argued that performance reviews "corrupt the system by getting employees to focus on pleasing the boss, rather than on achieving desired results" ("Why Your Boss Is Wrong About You"). In short, "nothing could be less fair."
For teachers facing increasing pressure to boost test scores or face possible dismissal, the implications are especially timely. New York State serves as a case study. On Mar. 1, Gov. Andrew Cuomo introduced a compromise plan that would allow merit to be considered instead of only seniority. His proposal was in response to criticism about using LIFO (last in, first out) as the basis for layoffs. (New York is one of 14 states that makes it illegal to consider anything except a teacher's length of service in making layoff decisions.)
In the absence of length of service, which is totally objective, what would substitute in determining layoffs? If only measurable outcomes were considered, then teachers would overlook non-cognitive results, which are every bit as important in education, and would be given no credit for serendipity, which can yield valuable results.
The Washington Post published a story that illustrates these perils ("Evaluation of D.C. teachers is a delicate conversation," Mar. 17). The District of Columbia uses a system of teacher evaluation known as IMPACT, which contains nine standards. The teacher in question was penalized for not strictly adhering to the objective for the day's lesson. Apparently, IMPACT does not take into account a teacher's sensitivity to cues that arise spontaneously.
Moreover, there is the issue of whether it's better to motivate groups, rather than individuals, in order to get the best overall results. In a working paper issued by the National Bureau of Economic Research, researchers found that group incentive pay and hourly pay play a greater role in motivating workers in manufacturing settings than individual incentive pay ("Social Science Palooza II," New York Times, Mar. 17).
This conclusion has particular relevance to today's debate in education over identifying effective teachers. That's because teaching by its very nature is a collaborative profession. Teachers openly share ideas in department meetings and in faculty lounges. They do so because they are all working toward a common goal. Those teachers who do not do their job well in one grade force other teachers to pick up their load.
In an essay published in Education Week, Andrea Gabor, chair of business journalism at Baruch College, CUNY, put the issue this way: "Individualized pay incentives also run counter to the logic of a systems approach to organizations" ("Why Pay Incentives Are Destined to Fail," Sept. 22, 2010). Teachers are "interlinked actors" who learn from each other and from outside experts.
Numerical goals have their place in schools. But they don't deserve to be elevated to the throne in making decisions. That's the mistake reformers make time and again.