As January comes to a close, it’s safe to say that it’s been a rough first year for the Obama administration. On the right, he is hammered for being a big government liberal, and on the left for being too cozy with big business and Wall Street (and don’t forget the two wars). Yet, there is at least one realm where the administration has received rather broad support, and deservedly so: education policy.
With the first round of competitive grant applications due earlier this week, the “Race to the Top” program will give states grants for implementing positive reforms like allowing charter schools to be created at all, or more easily, or be funded at parity with non-charter schools. The National Alliance for Public Charter Schools has endorsed the plan, as have leading policy experts.
Unsurprisingly, it has drawn the ire of an interest group that the Obama team is wont to placate in other circumstances--teachers unions. They resent that charter school teachers and administrators are not subject to the same restrictions and pay as contracted unionized teachers. Unfortunately, any system which pays based on seniority rather than talent discourages the latter. This isn’t just an abstract hypothesis; rather it is the conclusion one reaches when reading the work of one of the most talented economists and education experts in the country: Caroline Hoxby.
Hoxby is famous for demonstrating that competition increases educational performance. In a more recent paper, she finds strong evidence that teachers’ unions have led to a perpetual loss of talent in the teaching profession by compressing wages and driving away many skilled people, who find their ambitions stifled in public schools. This builds off her earlier work, summarized nicely by Robert Barro here, in which she finds that unionization increases the high school dropout rate above what it would otherwise be if unionization was random, and it does so despite increasing school funding.
If Hoxby’s theory is right, then the rise of charter schools should improve education spending efficiency and student achievement. That is exactly what she finds in a national study, with a compelling methodology that involved comparing achievement in charter schools to the nearest non-charted school with a similar racial composition (this was subsequently criticized by the Economic Policy Institute). A recent Brookings reportusing data from Milwaukee reaches similar, though less enthusiastic, conclusions. Crucially, in cases for which the methodology reaches the “gold standard”--randomized trial--Hoxby found significant benefits to children who were randomly assigned to attend charter schools compared to those who were randomly assigned to attend non-charter schools (see results from Chicago and, just recently, New York City).
These randomized trials are the most scientific way to evaluate charter schools, but some voters and policy makers may want a broader question addressed: Are the effects of charter schools large enough to affect scores at the state-level? Prompted by last week’s Washington Post article on efforts by the Center for Education Reform and the National Alliance for Public Charter Schools to document and grade state policies with respect to charter schools, I decided to take a closer statistical look at the state laws on charter schools. The rankings include things like whether or not there is a cap on the number of charter schools, which groups can authorize the creation of a new charter school, and whether or not charters receive the same level of state funding. The CER data also includes the year that the charter school laws went into effect. I matched this up with state level data from the National Center for Education Statistics on charter schools, class size, funding per students, state population, racial demographics, and income and poverty data. There is a strong correlation between year of adoption, the quality of state laws (as measured by the two organizations), and the percentage of charter schools. My goal was to see if the percentage of a state’s schools that are chartered in 2003 predicts changes in test scores by race, adjusting for initial measures of these other state characteristics.
It turns out that the answer is a strong yes. Blacks perform much better and whites perform about the same in states (and Washington DC) with higher percentages of charter schools, which include the nation’s capital, Arizona, Hawaii, Florida, Wisconsin, and Ohio. Overall, the gap between black and white math scores shrunk roughly two points more (or 25 percent of a standard deviation) in states where charter schools were 3.3 percentage points higher as a share of public schools (3.3 percent is the difference in 2003 between states with charter laws and those without them). The effect increases to 4.3 points after adjusting for the fact that states with failing school systems are the most likely to promote charter schools.
The figure above shows that as the percentage of charter schools increases, the gap between whites and blacks decreases (in other words, the y-axis is the gap in 2009 minus the gap in 2003; so a negative number means the gap narrowed). For a more in-depth methodological discussion of these results, click here.
There are many good reasons to cheer on the Obama administration’s efforts to prod states to liberalize laws on charter schools, such as choice, competition, flexibility, and improvements to the pay structure of teachers. The initial work of Hoxby and other education scholars strongly suggests that charter schools are likely to improve student achievement at the level of the individual and school. The less rigorous exercise discussed here points to promising state-level effects for blacks in states with a higher share of charter schools, but no effect for whites. This may be due to the fact that few whites currently attend charter schools or live in districts where they are relevant. More work needs be done to see which aspects of charter school laws are the most important to success, especially accountability, which varies across states and is a major concern of the teacher’s unions.
As to the achievement gap, there are, of course, other factors aside from school choice that contribute to minority underachievement, such as growing up in under-resourced segregated neighborhoods, as recent work has shown. Still, for too long politicians have been bereft of ideas on how to improve failing public schools and reduce racial inequality in education. As a 2008 Brooking report discusses, innovation in education policy is vital for enhancing the prosperity of cities and metropolitan regions, mostly by expanding opportunities for low-income residents. Admirably, the Bush administration tried to address this with No Child Left Behind, but the Obama administration deserves credit for taking an even bolder step towards real education reform with Race to the Top.