Looking back, much of my professional involvement with the school voucher issue dealt with a single topic: Accountability. I discuss that issue in an upcoming article (not yet published) where I found that voucher school administrators generally have a positive view of the Department of Public Instruction, and a wide variety of takes on the issue of accountability. The concept of accountability fascinates me, mainly because different audiences define it in very different ways.
One of those ways is basic transparency. Here I too have done some work. Last year I published an article in Education and Urban Society entitled “Changes in School Enrollment Patterns After the First-Time Release of School Level Test Scores in Milwaukee’s School Voucher Program: A First Look.” This article is part of a larger agenda of mine in which I explore the organizational characteristics of actors in the Milwaukee Parental Choice Program (MPCP) rather than the issue of whether or not vouchers on aggregate help or hinder education. Why? First, the issue of voucher impacts on test scores is pretty settled…there is little to no aggregate impact …and second I believe the proper unit of analysis is the school, not the program as a whole.
Anyways, in this article I take advantage of the 2010-2011 policy change requiring voucher schools to publicly release the WKCE scores of their voucher students. Under basic voucher theory, you would expect enrollment to increase in higher performing schools, and decrease in low-performing schools due to the transparency intervention. Indeed, I found schools with enrollment growth had significantly higher test scores than those with enrollment declines. I also found that above average voucher schools (as measured by test scores) gained students at a significantly higher rate than their average growth prior to the release of test scores. I also found that schools with lower reading scores were comparatively more likely to close. Interestingly, when I control for school level variables like religious affiliation, these relationships lose their significance.
Hence, my conclusions are a bit preliminary (it is a topic I will revisit). However, there is evidence that enrollment patterns were changed by the release of test scores. And importantly, changed in a good way as parents gravitated toward higher performing schools. Now, it does appear that other factors, many of which go unmeasured in my study, also drive enrollment patterns (see Paul Teske and company’s piece on how parents choose schools for more on this). So what to make of this? My conclusion is straightforward and simple, voucher programs should be required to publicly release test scores in a manner that allows parents and policy makers to make comparisons between schools across sectors. There is no evidence that doing so his harmful, and some evidence that doing so increases enrollment in higher-performing schools. From a practical standpoint, my findings suggest Wisconsin policy makers need to move to a stable universal accountability framework for all publicly funded schools. Most probably don’t need an academic paper to reach that conclusion, and policy makers have indeed articulated the need for such an approach…we just never seem to get there.