President Bush’s education policy has been taking some hits recently. A series of articles in the New York Times revealed that Houston schools, where current Education Secretary Rod Paige was once superintendent, grossly understate the percentage of students that drops out of school. And recently the Times featured a long front-page article analyzing test-score increases in Houston, finding that Houston’s gains appear smaller when measured by a national test rather than by Texas’s state test. The suggestion is clear–if something is rotten in the state of Texas, if there was no “Texas Miracle,” then President Bush’s No Child Left Behind (NCLB) Act, which was shaped by his Texas experiences, may be flawed as well.
It’s true that Houston has higher dropout rates and smaller learning gains than official statistics suggest. But neither fact undermines the evidence supporting the effectiveness of NCLB’s accountability and testing reforms. In fact, independent sources document that student test-score gains in Texas after the Bush reforms were among the highest in the country.
Texas says that its dropout rate was 1.3 percent per year in 2001, and Houston claims an annual dropout rate of 1.5 percent. If these figures were correct, more than 90 percent of students who start high school would be graduating. However, research by the Manhattan Institute indicates that only about 67 percent of students in Texas and only 53 percent of students in Houston actually graduate.
But to use this misreporting of dropout rates as an indictment against testing and accountability is just plain wrong. It is an unfortunate fact that the misreporting of dropout statistics is a very widespread and longstanding national problem. It occurred in Texas before President Bush, Secretary Paige, and their accountability reforms arrived on the scene. It occurs today virtually everywhere in the nation, including those places that do not have comprehensive accountability and testing reforms. In Connecticut, for example, the state claims that only 11.1 percent of students drop out of high school, while we estimate that about 30 percent of high-school students fail to graduate.
It doesn’t take testing and accountability programs to get school officials to misreport dropout statistics. They’ve been doing that on their own in many places for quite some time. The misreporting of dropout statistics is largely the product of loopholes in the definition of who counts as a dropout, sloppy tracking of students, and a school bureaucracy that wants to conceal embarrassing information. Arguing that the misreporting of dropouts means we should scrap testing and accountability is like arguing that accounting fraud at Enron means we should do away with audits. We need testing and accountability with accurate statistics just as we need audits for proper accounting.
Critics are also mistaken to think that testing and accountability reforms are undermined by the test-score analysis. It is true that the test-score gains produced by Houston students between 1999 and 2002 on the state’s accountability test were significantly larger than the gains those students made on the Stanford-9, a nationally respected achievement test. But it is worth noting that the Stanford-9 results do confirm that Houston students made real gains. Houston students improved by about 0.2 standard deviations on the Stanford-9 during this period.
That increase is about as large as the one produced by class size reductions in a widely touted Tennessee study. That study has been cited to support class-size-reduction measures in California, Florida, and elsewhere that will cost several billion dollars. If testing can produce gains as great as those produced by class-size reduction but at a very small fraction of the cost, that would tend to favor testing reforms, not undermine them.
Other evidence also points to real progress in Texas. The latest National Assessment of Education Progress (NAEP) data delivers good news for Texas, particularly with respect to black and Hispanic achievement. Between 1992 and 2003, the portion of black fourth graders performing at or above the “Basic” level of achievement on NAEP’s math test rose 42 percentage points, the third-greatest increase in the nation. The portion of Hispanic fourth graders performing at or above the Basic level on the math test went up by 35-percentage points, tied for the nation’s largest increase. These improvements catapulted Texas’s percentage of black fourth graders performing at or above NAEP’s Basic level to the highest in the nation, while the state’s Hispanic students rank fourth in the nation.
Black and Hispanic students in Texas still have a long way to go. Their improvements on the eighth-grade NAEP were not as impressive. And even after the gains achieved on the fourth- and eighth-grade assessments, black and Hispanic students still trailed significantly behind their white classmates. But given how hard it has been to produce achievement gains over the last 30 years, particularly for minority students, the progress made in Texas was impressive even if it was not miraculous.
Despite exaggerations and distortions in official statistics, independent evidence confirms that Houston and Texas students made significant progress under accountability and testing reforms. While it is too early to tell whether No Child Left Behind has had similar effects on students nationwide, the experience of Texas–and of other states that adopted these reforms earlier, including North Carolina, Arkansas, and Florida–suggest that NCLB’s approach works.
–Jay P. Greene is a senior fellow at the Manhattan Institute’s Education Research Office.