If a child has a white mother and a black father, then the child is racially...what? How the next Census, to be held in the year 2000, should handle the "multiracial" child will be decided by October. Moreover, every government agency counts races in roughly the same way, and it is the directions on how to count that are up for reconsideration. This working paper describes the interests at stake, mainly those pressed by advocates of a multi-racial category on the one hand and civil rights advocates on the other, and argues that the public's best interest is in fact different from either group's position. At the heart of the problem lies the treatment of racial intermarriage, and the paper argues that a great deal can be learned by comparing the treatment of racial intermarriage to ethnic intermarriage generally. Still, we cannot simply treat racial intermarriage like ethnic intermarriage because clear counts of racial minorities undergirds enforcement of civil rights law. Section 1 of this paper reviews the realities of ethnic blending in America, mostly focusing on white immigrants and their descendants; how has the Census dealt with this ethnic blending? Section 2 contrasts the Census's treatment of ethnicity with its treatment of race. Section 3 provides information on rates of racial intermarriage today. These first three sections, then, explain the issues. Sections 4 and 5 are the practical core of the paper, where the arguments for certain policies and against others are laid out. Section 4 contains the proposals for counting the multi-racial people. By contrast, section 5 is concerned with an issue rarely discussed in connection with OMB Directive #15, namely Census forecasts of the future racial composition of the United States. This issue too regularly makes its way to the front page but in misleading and confused ways. What links sections 4 and 5 is the argument that evading discussion of racial intermarriage distorts our understanding of race data, whether we are discussing 1997 or 2050. Finally, two sections, 6 and 7, can be thought of as "addenda" because these sections form extensions of the main argument. The first addendum, section 6, reviews the long-term history of racial blending in America and its (minimal) implications for the race data covered in present and proposed OMB Directives. The second addendum, section 7, offers some observations regarding another important change proposed for OMB Directive #15, namely making "Hispanics" one of the race categories
