Many in our society suffer disproportionately from systemic racism in ways that dramatically alter their daily lives and future possibilities. Institutions have made initial good faith efforts to correct the longstanding effects of systemic racism. Yet most of these efforts rely on antiquated labels that don’t fully capture the diversity of the United States in the 21st century — and don’t help us to revise our fundamental understanding of race as we try to find better ways to achieve equity.

Let me start with a story.

In 1959, my parents were not allowed to obtain a marriage license in the first two states where they applied. They were different races: a blond white guy and a brown Hawaiian/Pacific Islander woman. After they figured out the marriage license (Colorado finally came through for them), they adopted a son and a daughter, and raised mixed-race kids with parents in a mixed marriage in an overwhelmingly white small Midwestern town. What could possibly go wrong?

In elementary school, I was placed into the remedial classes for the “slow” kids. The teacher responsible for placing me into one of three tracks — remedial, regular, or accelerated — seemed to be unsure as to where I fit in. In middle school, students were selected to take IQ tests, but I was not offered to participate, as only the “smart” kids, not the “slow” ones, supposedly needed this test. In high school, I went to the guidance counselor to ask about taking the SATs. She said that I didn’t need to take this test because it wasn’t clear I would go to college, and, even if I did, the colleges that I would apply to wouldn’t require it. She sent me to meet with the Army recruiter instead.

I was pulled over twice by the police in my overwhelmingly white neighborhood (once with a second police car arriving almost immediately). They let me go without even a warning because I had done nothing wrong, but why did they stop me at all? Did they profile me based on my appearance? I definitely didn’t look like anyone else in my neighborhood. Stories of being pulled over for no good reason are common among Black and brown people. Still, while I was pulled over for questionable reasons, I didn’t think, “so this is how my life ends,” as many Black and brown people in the same situation have.

As a Harvard professor, I have been told that it doesn’t look good for me to lead community disparities/equity projects because I’m “white.” Am I really?

The nearly 34 million Americans who are mixed race — called “beige,” “half-breed,” “hapa,” or worse — often fit nowhere. We may choose a racial or cultural identity, but members of that identity don’t always acknowledge or accept us as their own. In 2010, approximately 15% of marriages in the U.S. were interracial. Why, then, does society still make us choose from a historical menu of ancestries?

Race is a social construct, not a biological one. In the 1790 US Census, you could be “free” or “slave” (read: white or Black). By the late 1800s, you could be white, Chinese or Japanese (reflecting immigration patterns), Black, Mulatto (3/8 - 5/8 Black), Quadroon (1/4 Black), or Octoroon (1/8th Black). These designations for people of Black African ancestry were not designed to lift them up. Soon, the descriptors were replaced with the “one drop rule”: any amount of Black African ancestry made you Black. In 1860, an “Indian” category was added to the US Census, and in 1900, “Indians” who lived on reservations were included as well (American Indians literally didn’t count before then).

When I have to check a box on race, do I report the biological ancestry given to me by the Thai father and Swedish mother I have never met, but are the sources of my physical appearance? Or do I answer according to my cultural heritage and social identity as I was raised (Native Hawaiian/Pacific Islander and Midwestern WASP), even though I don’t look like either of these groups and carry none of their DNA? I can check multiple boxes, but really I identify as a whole person, not disparate individual components. I think of myself as a baked cake, not just the eggs, flour, and sugar that were used to make it. Some forms provide the “other” checkbox option, one that reinforces our otherness as mixed-race people. “Rather not say” is there too sometimes, even if that answer implies that being mixed race is something to be ashamed of. “I’d rather not say” I ate a whole pint of ice cream in one sitting last night; my identity should not be equated to that level of shame.

Since I was old enough to talk and still to this day, the top two questions I hear are: “What are you?” and “Where are you from?” Little microaggressions. Even after more than six decades of being asked, I still don’t have an answer. Sometimes I just turn the question back onto the asker and have them guess.

Other mixed race people around the world have similar experiences. Tiger Woods designates himself as “Cablinasian” (that’s Caucasian, Black, American Indian, and Asian), perhaps because he identifies with all of his ancestries and doesn’t want to be defined by others based on his phenotypic features alone. In Russia’s 2002 census, it was possible to define your ethnicity in any way you wanted, so there were Russian Hobbits and Martians. But even if the U.S. Census allowed me to make something up, I’m not sure what label I’d choose. At the very least, I can say that I don’t want to check ill-fitting boxes, and I don’t want to coin a lonely designation for my ancestral-cultural heritage. Other people’s need to bucket your identity into easy-to-swallow but ultimately inaccurate categories, to the point of forcing you to check boxes in every doctor’s office and application form, takes a psychological toll.

Beyond the dilemma that I and many others face every time a form with a checkbox for race is handed to us, no one seems to have a good idea of why they are asking for this racial information. Is it to understand diversity in the organization’s constituents? If so, is it biological diversity, cultural diversity, socioeconomic diversity, diversity in lived experience such as encounters with racism or discrimination, all of the above, or something else, that is being investigated? Is it to identify and correct gaps in the organization’s diversity goals? Or to inform resource allocation for those who belong to a marginalized group? No matter the answer to these questions, the “race/ethnicity” checkbox alone is unlikely to be the right way to gain the desired information. Consider socioeconomic diversity: Some academic applications now ask if you are the first generation to attend college. This is presumably an attempt to identify socioeconomic position in a way that one’s race/ethnicity does not necessarily reveal. It’s a start, although — like the race/ethnicity checkbox itself — this single question is likely inadequate to achieve the larger goal of socioeconomic diversity.

And yet, we still default to a standard census-derived questionnaire on race to measure the vague concept of diversity. If you are a basic laboratory scientist, you don’t use the same assay for every experiment you do. The methods are determined by the research question, the specific hypotheses to be tested, and the analytical approach best suited to produce a meaningful finding. No good scientist will repeat the same experimental setup over and over, regardless of the question they are trying to answer. But this is our default practice when we measure diversity.

We all fill out forms in the name of diversity, and usually we only see the “please check all that apply” end product. I have never encountered a clearly defined purpose for the collection of my race/ethnicity. Nor is it clear to me who will have access to and use this information, or how this information will achieve some pre-specified goal. I imagine this is because specifying purpose and usage is hard to do, and most don’t know where to start (or even recognize they should).

Many organizations, including The Crimson, have made laudable efforts toward increased diversity and inclusion. But even these efforts don’t look beyond traditional labels — of white, Black, Asian, American Indian/Alaska Native, Native Hawaiian/Pacific Islander, or Hispanic — when it comes to race/ethnicity. If we are really serious about achieving diversity and inclusion, we are obligated to do the difficult yet fundamental work of clearly defining the purpose and goals for the labels we use, identifying appropriate metrics, and planning for the future use of collected information. Before we can develop and implement practices, processes, and policies to eliminate the inequities and disparities that marginalized people face, we must take a more considered approach to something as “simple” as the labels with which we are identified.

Timothy R. Rebbeck is the Vincent L. Gregory Professor of Cancer Prevention at the Harvard T. H. Chan School of Public Health and the Dana-Farber Cancer Institute.