RALEIGH – To American school reformers of all stripes hoping to engineer a massive improvement in student learning, University of Texas-San Antonio professor John Merrifield has just offered some sobering advice: don’t get your hopes up.
Surveying decades of published research on in-school variables, background characteristics, and test scores, Merrifield concludes in a new Cato Journal paper that costly policy changes such as cutting class size or raising teacher pay produce at best modest gains in student achievement. Close observers of North Carolina education trends shouldn’t be a bit surprised by his findings.
Using the data to established predicted values of test score averages for various subgroups of the student population, Merrifield shows that even the best-case scenario puts the average score on the National Assessment of Educational Progress (NAEP), for example, well below the “proficiency” level.
In other words, if North Carolina and other states do just about everything on the wish list of teacher unions and other interest groups within public education – chop average class sizes by a third, boost average teacher pay by more than that, and jack up per-pupil spending by as much as 100 percent – the available data suggest that the result would be little improvement on the current mediocre level of student performance.
Why so glum an assessment? Well, educational achievement is highly correlated with non-school variables such as parental background. Some analysts on the Left and Right have concluded from such research that only attacking poverty and other conditions outside the walls of the schools has any hope of generating significant improvement.
I think that’s too pessimistic a reading of the data – and Merrifield actually agrees. In his paper, he doesn’t argue that school reform is a doomed enterprise. Instead, he argues that policy initiatives within the current financial and administrative system underlying public education aren’t likely to accomplish much:
…[T]hose school characteristics don’t matter much now, but they could matter in an environment lacking in some of the current system’s widely shared key characteristics. Repeated urgent calls for reform indicate that the United States needs to find a way to make school characteristics matter. In other words, fixed factors are needed that will substantially elevate the average level of test scores around which non-school factors may still explain most of the variability, and perhaps fewer fixed factors so that variable factors become more significant.
Advocates of market-based reform also need to temper their expectations and promises of immediate, radical improvement, too. But at least their focus is on the right set of policy levers to pull. They want to change the institutions, not just shove more money, personnel, and pedagogical fads into the existing institutions.
Local examples of promising success with disadvantaged students aren’t hard to find. What’s hard is to clear away the obstacles preventing these local successes to be replicated. Happy talk and half-measures won’t do it.
Hood is president of the John Locke Foundation