Teacher Unionism And Student Achievement: The Grounds For Skepticism

One of the more negative qualities of contemporary debates over educational policy is the vulgar politicization of research. For some, it is sufficient that a study reaches the “right” conclusions about an issue to embrace it as scholarly, rigorous social science. Conversely, when a study reaches the “wrong” conclusions about an issue, it becomes suspect scholarship. Jay Greene’s response to a post of mine last week, in which I pointed that the weight of scholarly literature on the relationship between teacher unions and student achievement was quite the opposite of what he suggested with his citation of a single study by Caroline Hoxby, fits this pattern to a tee. Greene simply dismisses all of the studies which reach a conclusion other than the one he wants, using ad hominem argument and caricature which ventures into the absurd. Little surprise, then, that the only study left standing is the one that reaches the political conclusions Greene favors.

It is a simple enough matter to point out the ideological biases in that sort of argument. But it would be a mistake, I want to argue, to join in Greene’s vulgar politicization of education research, and embrace the substantial body of scholarly work in this case just because it reaches conclusions teachers and unionists find much more amiable. There is very good reason to be skeptical about all of the literature on this subject, regardless of its conclusions.

There are at least three grounds for this skepticism. First and most importantly, the factors that shape student achievement are so many and the relationships among them are so complex and interdependent, that it is hard to see how one could isolate the effects of teacher unionism in a meaningful and accurate way. Just as Massachusetts, with its strong teacher unions and high teacher union density, leads the nation in student achievement, Mississippi, a “right to work” state that prohibits collective bargaining for educators, consistently ranks near the bottom in student achievement. But teacher unionism is far from the only variable in these two states: among many other things, Massachusetts spends more on pre-K-12 education than Mississippi. In significant measure, Massachusetts spends more on education because strong teacher unions engage in political action and lobbying on behalf of education funding. But it is also the case that the political culture of Massachusetts is one that more highly values education and Massachusetts is a wealthier state than Mississippi, so there are other factors that contribute to the education funding differential. In that confluence of factors, how can one isolate and quantify the teacher union effect?

Secondly, teacher unions are not independent actors on the educational stage. By their very nature, what they can accomplish is circumscribed by the management of their school district. There has never been a teacher union local which enacted a significant education reform agenda without having a progressive partner on the management side. By the same token, many a teacher union local that wanted to pursue such an agenda has found itself stymied by a management that refused to collaborate. In that context of mutual determination, how can one distinguish in quantifiable ways what is caused by union and what is caused by management?

Thirdly, the aggregation of different teacher unions into a single “teacher union” effect assumes that there is not significant variation, one teacher union local to the next. Whether the local continues in an industrial union tradition of focusing solely on wages, working conditions and due process, or has taken up educational issues as a central component of their work, the presumption is that their effect is so similar as to justify a sweeping generalization on the effect of all teacher unions.

So if we are intellectually honest, what we are left with is a number of correlations between strong teacher unionism and various measures of student achievement. On the whole, they point to a positive connection, but they are far from establishing a causal relationship. In responding to the vulgar politicization and research cherry picking of a Jay Greene, we should resist the temptation to become his mirror image.

Tagged: featured

 Print