With about an hour to go before I leave for vacation, I can't resist making one last parry in my ongoing debate with Matt Yglesias about standardized tests and teacher merit pay.
First, though Matt is a dear friend, I bristle at his suggestion that my positions uncritically echo a pro-union line. Those who read my work regularly will know that I believe there is little evidence for the claim that teachers' unions are the primary barrier preventing the United States from closing the achievement gap or beating Finland and Canada's butts on international assessments. There are simply too many other deep-seated differeces between those societies' educational systems and our own; what they all share in common is that the teachers within them are unionized. What's more, American states without teacher collective barganing perform worse educationally than those with strong unions. That doesn't mean the unions are responsible for strong academic performance, just that it is doubtful that they are signficantly hindering it.
That said, I carry no water at all for those progressives (and there are not that many of them, by the way) who say there's no use in focusing on school reform until we beat poverty and fix health care, child care, and the like. If I thought that were true, I wouldn't spend so much time reporting on and writing about interesting, innovative educational programs like the Children's Literacy Initiative in Newark or Tech Valley High School in Albany. I believe there are schools, classrooms, and teachers changing children's lives each day, and even fighting inequality.
So I do support efforts to identify and reward successful teachers while simultaenously identifying, remediating, and eventually firing those who are unsuccessful. The question at hand is simply whether standardized test scores are the best means for doing so. I don't think they are, because of Campbell's Law. The education team at the Center for American Progress, where Matt works, has investigated a number of possible ways to evaluate teachers, even publishing a paper by Linda Darling-Hammond that describes a more holistic, comprehensive system, one that relies on evaluators observing and interviewing teachers and looking at examples of their students' work–both tests and other sorts of assignments.
Though such systems are labor-intensive, I'm more hopeful about them than I am about test-score based evaluation and pay. For one thing, from Nashville to New York City , such merit pay schemes have often disappointed researchers, for they seem to have little effect on student achievement.
And we know from the work of sociologists like Dan Lortie and social psychologists like Edward Deci that the kind of people who choose teaching as a career are unlikely to be people whose performance is swayed heavily by relatively small financial incentives.
I have to run or I'll miss my train to Connecticut! But there you have it.