“Does No Child Left Behind Place a Fiscal Burden on States? Evidence From Texas”

October 2004

Calculated Additional Annual Cost of NCLB in Texas :

$1.7 billion (in 2004 dollars) to meet the 55% passing rate, which is 5.9% of 2004 public school revenue

$4.7 billion (in 2004 dollars) to meet the 70% passing rate, which is 16.7% of 2004 public school revenue

Outcome Measures The authors prefer a two-year view of test scores, rather than the one-year view required by NCLB. Just prior to the study, Texas changed its standardized tests from TAAS (the Texas Annual Assessment of Skills) to the more rigorous TAKS (the Texas Assessment of Knowledge and Skills). Because TAKS is new, the researchers did not have two years of test results on which to base their study. Therefore, they converted the last two years of TAAS scores into TAKS scores, using a schedule developed by the Texas Education Agency. The authors also examined the tests taken by those students with disabilities who do not take TAKS; ACT and SAT scores; and the dropout/retention rate. Moreover, the authors noted that inaccuracies in reported dropout rates in Texas have been discovered.

Efficiency The authors also used an “efficiency” factor in estimating the cost of educating students, ostensibly to isolate those schools whose higher spending is a result of an inefficient use of resources. The researchers defined “inefficient spending” as any spending not directly related to state standards. Therefore, they eliminated all spending on art, music, vocational education and other subjects and activities from their study. This study assumes that schools facing more local competition, e.g. from private schools, spend more efficiently.

Statistical “Cost Function” Model:

The goal of this method is to find the minimum amount of spending needed in each district to meet the required standards, given characteristics of the district. It attempts to do so by quantifying the relationship between per-pupil spending and student standardized test scores, taking into consideration student characteristics (such as poverty, limited English proficiency, disabilities) and characteristics of school districts (such as size, rurality, etc.), as well as the prices school districts must pay for teacher salaries. This methodology attempts to account for characteristics out of a district's control, such as cost of living, as well as those within a district's control, such as certain requirements set by the districts in hiring teachers. In addition, the authors use the “efficiency” factor described above.

The authors conclude that NCLB is under-funded in Texas. The authors note that for 2005-06, the federal government provided Texas with a $519 million increase in federal funding over the preceding year. However, as noted above, this amount is $1.2 billion short of what is necessary to educate children to meet the 2005-06 AYP target . Several factors make this 519 even smaller, according to the authors:

some federal money goes to schools that already meet the study's 55 percent passing rate and may not need additional funds.

in many schools, the federal Title I allocation has gone down from 2002 to 2005.

the state has been investing less in schools, so schools are more under-funded.

The authors explain that their analysis underestimates the studied costs, for three reasons:

Unlike NCLB's requirement that all subgroups meet passing rate targets, the study only considered the passing rate for the school as a whole. The researchers note that many schools have subgroups that did not reach the 55 percent or 70 percent passing rates, and that the additional expenditures needed to raise their scores are not captured in this study.

The study also averages the passing rates in math and reading. Because NCLB mandates that students pass both tests, the researchers acknowledge that additional expenditures would be required for those students who did not meet reach the target in one of the subjects, even though they are counted as “passing” for the purposes of this study.

The researchers noted that there would be additional teacher salary costs associated with the expansion of testing in 2007 to include science and social studies and to expand testing to grades 3, 5, 8 and 11 in 2005-06.

The authors conclude that, given fiscal constraints in Texas and in most other states, and efforts around the country to place limits on property taxes, it is unlikely that states would have, on their own initiative, developed the testing, reporting and other performance requirements to the degree required by NCLB. They also conclude that the cost of implementing these requirements exceed the federal funds allocated. Therefore, they suggest that a national discussion be undertaken regarding how the cost of these measures should be funded by the federal government and the states.

None

Jennifer Imazeki, Department of Economics, San Diego State University ; and Andrew Reschovsky, Robert M.LaFollette School of Public Affairs, University of Wisconsin