Understanding Reading First

What We Know, What We Don’t, and What’s Next

| Corinne Herlihy, James J. Kemple, Howard Bloom, Pei Zhu, Gordon Berlin

In 2008, the Institute of Education Sciences (IES) at the U.S. Department of Education published research findings on Reading First, a centerpiece of the No Child Left Behind (NCLB) Act that provided $1 billion per year to help all children read at or above grade level by the end of third grade. The findings were interpreted by many in the media and the policy community as saying that Reading First did not work. Although the story is more nuanced than that, funding for the program was eliminated in the fiscal 2009 spending bill that was signed by President Obama in March. NCLB is up for reauthorization in 2009. In the meantime, the American Recovery and Reinvestment Act provides tens of billions of dollars to states and localities for spending on education, meaning that federal, state, and local policymakers face critical choices today about how best to use this money to support early reading instruction and achievement.

This policy brief describes what Reading First was, sets the context in which it was implemented and the studies were conducted, summarizes the findings, and discusses the implications both for federal and state policy and for future research in the teaching of early reading.

The bottom line is that Reading First did increase the provision of professional development for teachers and of reading coaches and supports for struggling readers in schools that received funding. The program did influence how teachers taught — in ways that are aligned with scientifically based reading research (as summarized by the National Reading Panel in 2000), a key goal of the legislation. Unfortunately, these improvements did not produce higher reading comprehension scores on average among students in the Reading First schools. Nonetheless, there is some suggestive evidence that Reading First funding may have improved comprehension in schools in which the effects on teacher instruction were larger.

Reading First’s lack of impact on overall reading comprehension test scores may be related to two connected issues. First, the type of reading instruction that was promoted by Reading First was already in wide use when the program came on line in 2002 — in part due to the influence of the National Reading Panel recommendations and an earlier National Research Council report, Preventing Reading Difficulties in Young Children, as well as the availability of funding from the Reading Excellence Act of 1998, the predecessor to Reading First. As a result, the IES studies found that teachers in schools that did not receive Reading First funding — that is, the schools that formed the comparison group in the evaluations — also spent the majority of their class time for reading focusing on the core components of scientifically based reading instruction recommended by the National Reading Panel and supported by Reading First.

Second, the increase that Reading First produced in instructional time devoted to the core elements of scientifically based reading instruction — about 7-10 additional minutes a day on top of the 50 minutes already devoted to teaching in this way — may have been too small, at least on average, to induce improvements in students’ reading comprehension. Intriguingly, however, Reading First did appear to produce improvements in reading comprehension in schools where it produced larger increases in the recommended instructional practice. These schools tended to be ones that (1) served more educationally disadvantaged children, (2) spent less time using scientifically based methods of instruction in the absence of Reading First funding, and (3) received larger grants. But because the impact study was not explicitly designed to answer questions about the magnitude and determinants of variation in program effects, these findings are only exploratory and suggestive.