The Evidence on Race to the Top
Are the naysayers right?
Who would have imagined that a Democratic president with broad support from teachers’ unions and education scholars would devise a policy as distasteful to his supporters as the No Child Left Behind Act? Yet President Barack Obama appears to have done just that with the Race to the Top initiative.
In addition to being widespread and heated, the criticism of Race to the Top has focused on an unusual topic: research. The National Education Association wrote in response to the grant competition’s regulations that “we encourage the administration to base its recommendations on research and on what works.” Scholars have been even more dismissive. The education historian Diane Ravitch writes: “What is extraordinary about these regulations is that they have no credible basis in research. They just happen to be the programs and approaches favored by the people in power.” Others point to the apparent hypocrisy of U.S. Secretary of Education Arne Duncan’s claims to be a data-driven reformer.
What has caused the firestorm? To receive Race to the Top funds, states must make “assurances” that they will take four steps: adopt common standards and high-quality assessments; develop and use state longitudinal-data systems; improve evaluations of teachers and principals (incorporating student-achievement scores) and use these to inform high-stakes decisions; and turn around failing schools using, for example, charter schools.
But the research-based criticism of Race to the Top is misleading in three important ways.
First, the federal government is largely limited to two types of reforms: funding and systemic reform. Race to the Top represents both. Along with the rest of the education portion of the 2009 economic-stimulus package, the program is putting more than $80 billion into K-12 public education. In exchange for the money, Race to the Top is inducing significant systemic reform—that is, changing the broad incentives that affect decisions among all stakeholders.
Systemic reforms like this, unfortunately, are hard to evaluate according to rigorous research standards for a number of reasons: (1) Systemic reforms are not adopted randomly, making it difficult to know how students would have fared in the absence of the policies; (2) systemic policies are adopted at the state and federal levels, so in a research sense there are few examples to analyze; and (3) systemic reforms, by design, change schools gradually over time, making it hard to see immediate and noticeable improvement in outcomes when reforms are introduced. For all these reasons, we still have shaky evidence even on the impact of systemic reforms implemented decades ago. Relying on evidence to justify reforms like the Race to the Top measures before they are adopted is that much harder.
The second problem with the criticism is that basic policy analysis requires comparing policy alternatives. Is policy X better than the status quo or policy Y? Analyses of Race to the Top rarely try to answer this question, and when they do, they point to programs that have at best the same weak base of research support.
Some opponents of the federal initiative may respond by saying that I, like Secretary Duncan, am a hypocrite. I am a researcher and spend most of my time carrying out rigorous evaluations of education policies and imploring my students to keep in mind that “correlation is not causation.” How can I then appear to defend policies that are not supported by research that meets those same rigorous standards?
The answer is simple: Researchers and policymakers have different jobs. Researchers should be very cautious about concluding that policy X is better than policy Y. But policymakers do not have that luxury; they make decisions every day, implicitly or otherwise, about whether to stay the course or make changes. These decisions should be based as much as possible on hard evidence, but given the difficulty of evaluating systemic reforms, the possibilities are somewhat limited.
Consider the evidence around Race to the Top. I agree with critics that there is little rigorous evidence showing that common standards and assessments, improved data systems, better educator evaluation (and using the results for merit pay), or school turnaround efforts will improve student outcomes. Charter schools represent a partial exception, because they’ve been the subject of hundreds of studies, some rigorous.
The problem is that the results on charters are mixed. On the one hand, they seem about the same as regular public schools in generating achievement and do little to increase competition among public schools. On the other hand, they are less costly and, as one recent study concludes, attending a charter school increases the likelihood of high school graduation and college attendance.
Does this lack of rigorous research supporting Race to the Top mean policymakers should stay the course and keep searching for a “proven” alternative? The answer depends on which side has the burden of proof. If we have to prove beyond a reasonable doubt that something new will be better, then the status quo will reign forever—again, we cannot easily prove that any systemic reform will be effective.
One thing we can do, however, is show convincingly that the current system does not meet our needs. And there is growing evidence that if we are going to remain economically competitive and equitable, we need to change our education system. We are not “a nation at risk,” as some would suggest, but we are a nation that could clearly do better.
There is also growing agreement that our current approach to ensuring quality teaching is flawed. What educators have long known—that teacher effectiveness varies widely—has now been confirmed by statistical evidence. Unfortunately, teacher effectiveness is only weakly related to factors like certification and degrees, which have been the bulwark of American teacher policy for more than a century. And current teacher-evaluation systems rarely do much if anything to pick up the slack. This is a huge problem, and one that the federal initiative addresses.
If not the status quo or Race to the Top, then what? The Obama administration could have put more money into the existing system. Or it could have pursued the policy that probably has the strongest research support: early-childhood education. Or it could have tried to improve general child well-being through improved health care, for example.
Do these sound familiar? They should, because the administration is already doing all of them in addition to Race to the Top. It has dramatically increased federal spending on early childhood, and on K-12 education more broadly, while universal health care is President Obama’s signature domestic-policy initiative.
The third and final problem with the criticism is that Race to the Top is more voluntary than critics have suggested. Yes, the financial crisis creates pressure for states to go along, but these funds amount to less than 1 percent of current education spending, and the money only lasts two years. So states do have a choice here, and many so far have said no by not entering the competition. It is more reasonable to demand rigorous research for policies that are required than for voluntary policy experimentation like Race to the Top.
In this respect, the problem with the administration’s policy is not the absence of rigorous research, but that evaluation seems to have been an afterthought. The Institute for Education Sciences is working hard to address that problem, but its options became limited once the Race to the Top regulations were set.
Critics might respond that, by my reasoning, we can justify almost any reform. Nothing could be further from the truth. There are only so many reform packages, like Race to the Top, that are based on significant and well-documented education problems and involve solutions designed with clear and reasonable rationales (and some, albeit limited, evidence) to address them. Indeed, if we held the status quo to anything like the standards critics are holding Race to the Top to, the status quo would be finished.
No doubt, Race to the Top is “favored by the people in power.” But it is much more credible than critics give it credit for.
Vol. 29, Issue 27, Pages 26, 36Published in Print: March 31, 2010, as The Evidence on Race to the Top