Data Proposals Threaten Education and Civil Rights Accountability
Using 2005 data collected by the National Assessment of Educational Progress (NAEP), which has data on students both under the new rule as proposed by the August 7 guidelines as well as the current method of only one race or ethnicity per child, we have been able to examine the impact of the proposed guidelines on the collection of racial and ethnic data compared to the existing method. Our findings show that the proposed changes would suddenly produce vast changes in the apparent racial composition of our educational system, create a large new category that is a grab bag of many forms of multiracial backgrounds, and would very seriously undermine both research and policy analysis work that is essential to understanding racial change and racial inequality as well as to monitor civil rights enforcement. We believe that these changes are not supported by good research and that the goal of including multiracial students in our counts can much better be accomplished with procedures that parallel those devised for the Census. We also find that the policy changes that would rely on fourth graders to respond to complex questions about their race and ethnicity are mistaken and will produce data that is questionable and often meaningless, undermining the valuable data on achievement by racial subgroups that is a valuable result of the No Child Left Behind law. Suddenly schools and their communities would find subgroups growing and shrinking and changing size or even suddenly emerging or disappearing in ways that would be confusing and disruptive to the accountability system
We recommend that these proposed changes be rescinded and that better procedures for accurate counts be devised in collaboration with researchers and community leaders. In addition, we also recommend that the processes for collecting accurate data from children, families and teachers be carefully researched and tested before implementation. We must avoid procedures that cause a serious loss of vital information and accountability for educational institutions and a fundamental undermining of civil rights research and enforcement.
Three relatively simple parts of the new policy create the problems that will be described in the statistics presented in this report:
The decision that institutions should report Hispanics in a separate category and not to report them as members of racial groups or multiracial. This breaks with the Census policy that Hispanic is not a racial category but that Hispanics can be of any race or of mixed racial background. The U.S. Census reports millions of Hispanics who see themselves as white, or black or "other." Ending this reporting mechanism will cut down the number of students in those races. Since the Hispanic question is asked first, the proposed system tends to produce large Hispanic counts.
The decision that students reporting more than one racial background should only be reported as mixed race without the breakdown of their racial backgrounds. This is in contrast to the U.S. Census policy of reporting the racial combinations of multiracial individuals so that researchers can look at a category, for example, of blacks and black mixed race students or Asians plus those with mixed Asian and Pacific Islander backgrounds. Since "mixed race" does not define any kind of ethnic community, it will be impossible to interpret statistics that will combine unknown groups of students from extremely different backgrounds.
The decision that fourth graders should be asked to respond to the question about whether they are Hispanic or not and then to choose from one or more of five racial categories, without any research to show that these questions are intelligible to young children in an extremely wide range of community contexts.
If the proposed guidelines are adopted, it would become the only available source of educational data on race and ethnicity and would make it impossible to compare future patterns and trends with past ones or to know whether various institutions were making progress in educational outcomes by racial and ethnic group. Moreover, the proposed system of accounting is incompatible both with the data that has been collected consistently for the past 40 years for the Office for Civil Rights and with the much more thoughtful way the Census has handled the collection and reporting of multiracial statistics after very extensive work and research with experts across the country. Our findings show that because the proposed system treats Hispanic as a preferential category and precludes counting multiracial students within the various racial categories, the resulting data will exclude from the various racial counts many students who may consider themselves and be seen as "white", "black", "American Indian" or any other race or ethnicity.
The data collected under the proposed guidelines would very seriously undermine both research and policy analysis work that is essential to understanding and successfully dealing with racial change and racial inequality. It would undermine civil rights enforcement. It would make the nation's white enrollment appear to have suddenly dropped substantially and would have a similar impact on the black population and on American Indian and Asian enrollments in some states. The mixed race category would be an essentially meaningless category for civil rights and research purposes since it would include large and unknown numbers of people who were from two historically underrepresented groups, others from two groups that are not underrepresented, and various other combinations. It would provide, of course, no information on how many of these students were overwhelmingly of one race with some distant connection with another versus those who now have two parents from different races and ethnicities.
The problems with these categories are not only apparent in terms of the huge changes in the number of students in different groups in many states but also that even within the same state there are often very large changes between the 4 th and 8th grade, suggesting the categories may not be intelligible for small children. The changes also make it impossible to track achievement and graduation trends within a racial or ethnic category over time within a school, a district or a state. Data tracing trends over time is, of course, a central requirement of the No Child Left Behind Act, essential for judging compliance with various civil rights court orders, and required by the special education law, IDEA. In some states, the change will make it appear that individual racial groups suddenly are performing substantially better or worse on some achievement tests even when nothing has changed about actual test results.
We request a delay in implementing this radical change in policy and believe that full congressional hearings are needed before embarking on changes that will produce a fundamental change in the description of American school and college populations. We see no serious risk in continuing the present method of counting until a full review can take place. We believe that the most appropriate method of implementing an accounting solution that would reflect the growth of multiracial families would be to use the procedures for data collection and reporting adopted by the U.S. Census, subject to serious research and investigation on the issue of the intelligibility of the categories to young children. That research is an obvious prerequisite before any change should be made in counting methods.