THE POTENTIAL AND LIMITS OF FEDERAL POLICY: A RESPONSE TO LADD

Ladd identifies several legitimate shortcomings of No Child Left Behind (NCLB). While I disagree with some of the specific claims she makes, I agree with her broader point that NCLB was not sufficient to accomplish its extremely ambitious goals. However, I would not characterize NCLB as a “deeply flawed” federal policy. I simply believe that NCLB highlights the challenges of using federal policy to address complex and deep-rooted social issues that play out at a local level. More importantly, I strongly believe that its replacement, Every Student Succeeds Act (ESSA), is first and foremost the solution to a political problem that will not help to promote high-quality and equitable education for all children.

ASSESSING THE IMPACT OF NCLB

Ladd acknowledges that National Assessment of Educational Progress (NAEP) scores rose throughout the NCLB era, but argues that because of the positive trend started in the 1990s this improvement should not be attributed to NCLB. There are two problems with this claim. First, as Ladd acknowledges, there is compelling evidence that NCLB itself had a positive impact on math scores (Dee & Jacob, 2011). Ladd claims that the positive effects are due entirely to an implausible jump in achievement in 2003. However, Dee and Jacob (2011) demonstrate that NCLB is associated with a change in achievement growth, particularly for disadvantaged students and at lower points in the achievement distribution. And the Lee and Reeves (2012) analysis, cited by Ladd, is problematic for several reasons, including its failure to include state fixed effects, its reliance on endogenous post-NCLB measures of implementation fidelity, and its discussion of results for subject-grades that do not meet the minimum requirements for a Comparative Interrupted Time Series (CITS) analysis.1 Second, there is considerable evidence that at least part of the positive trend in the 1990s was due to the rise of pre-NCLB accountability policies implemented at the state level (Carnoy & Loeb, 2002; Hanushek & Raymond, 2005). More generally, Ladd ignores a large body of evidence indicating that accountability policies have positive (though modest) impacts on student achievement, and even on some aspects of educational practice.

Ladd describes three specific flaws of NCLB: (i) its narrow focus, (ii) its unrealistic expectations, and (iii) its reliance on pressure without support. With regard to its narrow focus on math and reading, my view is that the moderate shift of time toward these high-stakes subjects in elementary and middle school was not a bad thing. I would not want schools to focus entirely on these skills, but for the most part I think that schools have continued to devote considerable time to a wide variety of curricular areas. And despite the attention devoted to concerns of “educational triage,” the majority of studies actually do not find that accountability led schools to focus on the proficiency bubble (see Ballou & Springer, 2017, for the most recent analysis of this issue). And, indeed, studies that use low-stakes test outcomes find no evidence of such triage.

On the other hand, I agree wholeheartedly with Ladd that accountability regimes (including NCLB) encourage narrow teaching-to-the-test that is counterproductive. But, even here, it is important to acknowledge that such “teaching-to-the-test” can have positive long-term effects. The best evidence for this comes from an analysis of the Texas accountability system (Deming et al., forthcoming) that served as a model for NCLB.

With respect to the expectations embodied in NCLB, I agree that it was unrealistic to expect that 100 percent of children would be proficient by 2014. However, these high expectations—combined with the public reporting of subgroup scores—did spur schools and districts to action, which included positive changes that likely would not have occurred in the absence of looming sanctions. The fact that some states lowered their proficiency standards in response to NCLB is unfortunate, but the fact of the matter is that before NCLB there were no standards for many schools, particularly those serving poor or minority children. And the fact that close to half of all schools in the country were failing by 2011 is completely consistent with the dismal proficiency rates we see on NAEP and the glaring inadequacies many reformers point to in high-poverty schools. While the high failure rates led to a political backlash, they correctly reflect the state of our educational system.

Ladd cites the decline in teacher morale as one of the negative consequences of NCLB. While there have been a number of anecdotal accounts of teacher frustration with high-stakes testing, there is no compelling evidence that NCLB (or any other state accountability policy) systematically decreased teacher morale. For example, using data from the nationally representative Schools and Staffing Survey, Grissom, Nicholson-Crotty, and Harrington (2014) find that teacher reports of various work environment measures improved over the period coinciding with the implementation of NCLB.

Finally, Ladd argues that test-based incentives only make sense if one believes the problem of low-performing schools is attributable to teacher shirking or low expectations. Instead, she argues that low achievement is primarily due to the limited capacity of schools to meet the challenges of children from disadvantaged backgrounds. I agree that the social and economic conditions are a major factor—perhaps the most important factor—holding back disadvantaged schools.

Still, I believe that there are many potentially valuable reforms that districts could have made in response to the incentives provided by NCLB. They could have changed teacher transfer policies that limit the flexibility of administrators. They could have changed school finance systems so that more affluent schools bear the true cost of hiring more experienced (and thus more expensive) teachers. They could have changed school start times to better match the natural sleep patterns of adolescents. They could have lengthened the school day or year. They could have relied more on research than textbook sales reps in selecting curriculum. They might even have radically reorganized schooling to rely more on individualized tutorials than whole group instruction, creatively using a mix of highly paid and certified teachers with less expensive assistants or even volunteers. Of course, most districts did not attempt such changes, for a combination of political, social, and bureaucratic reasons. But was this because the incentives under NCLB were too tough or not tough enough?

THE ROLE OF THE FEDERAL GOVERNMENT IN EDUCATION

Ladd contends that a key flaw of NCLB—and perhaps of federal education policy more generally—is that it fails to confront the social and economic conditions in which schools operate. She argues for a “broader and bolder” approach to education that addresses the problems that disadvantaged children bring to school. I wholeheartedly agree. But this is not simply a matter of providing more pre-school or after-school programs. Over the past decade, the proportion of poor children participating in formal pre-school has jumped dramatically (National Institute for Early Education Research [NIEER], 2015; Klein & Whitehurst, 2015). Yet achievement gaps have stubbornly persisted. As Ladd recognizes, the key is to provide “high-quality” programs. But this brings us back to the same issues raised by NCLB since its main goal was to ensure “high-quality” elementary and secondary schools.

Social policy that increased labor force participation, eliminated poverty, and dramatically reduced income inequality would be excellent education policy. But other than such broad social reforms, what role can and should the federal government play in education policy? Should the federal government's role in education be limited to writing checks?

The federal government has neither the resources nor localized expertise to directly support the operation of schools. In my view, setting up a broad framework of standards and accountability is exactly the type of role the federal government should play, and NCLB attempted just this. Another important role for the federal government is to support high-quality, impartial research that has positive externalities. And NCLB did this with the establishment of the Institute for Education Sciences (IES) whose mission is to “conduct and support scientifically valid research” (Jacob & Ludwig, 2005). NCLB was clearly not perfect, and even a perfect federal school accountability policy would be insufficient to close the racial and economic achievement gaps in our country. We should use the lessons of NCLB to refine federal education policy to support local efforts to improve the academic performance of all children.