Featured Articles
Meredith Turner
/ Categories: 562

Changing Face of Diversity: A Discussion of the U.S. Census and Its Implications

Bharati B. Belwalkar, City of New Orleans; and Kisha S. Jones, Pennysylvania State University

At a recruiting event for my organization, I (Bharati) was approached by a job applicant. With the demographic information section of the application in hand, she asked, “I am half White and half Black. Which box should I check? Both boxes for White/Caucasian American and Black/African American, or just the one that says “Other?” This confusion reflects a larger national discussion in the U.S. regarding the definition of the “Other” category: does it mean “multiracial,” “none of the above,” both, or something else? Although pertinent to all people and institutions within this country (and any others with diverse population), this question is particularly relevant to I-O psychologists, as the science and practice of our field makes demographic reporting important. In our opinion, we have not done the best job of keeping up with how demographic changes in the U.S. reflect how we capture the race/ethnicity of people within organizations.

In this article, we attempt to initiate a broader conversation on this topic within our field. We contend that as the makeup of our country continues to change, this conversation will need to be ongoing to ensure accuracy in how we survey demographic characteristics. As a start, we thought that considering how race and ethnicity have been defined and captured in the U.S. Census over time would be useful. The U.S. Census plays a large role in capturing the demographics of this nation, and its decisions around how people are “categorized” can influence the ways other institutions choose to do so. Therefore, we also contemplate what the Census’ categorizations mean for organizations, and for I-O research/practice.

Race Defined in General and by the U.S. Census

The concept of race/ethnicity in the U.S. is complex and difficult to define, resulting in inconsistent meanings and understanding (Chavez & Guido-DiBrito, 1999). Although race is popularly considered a biological construct that is derived from one’s physical characteristics and gene pool (Spickard, 1992), researchers have recognized its social dimension that includes “a sense of group or collective identity based on one’s perception that he or she shares a common heritage with a particular racial group” (Helms, 1993; p. 3). After all, genetic differences within humans are not large enough to support the notion of racial groups.1 Race can be more easily defined as a social construction rooted in historical and anthropological context, and should be considered a social categorization based on phenotypic characteristics to which society attaches importance (Omi & Winant, 2014). Whereas the concept of race has both biological and social components, ethnicity is a purely social construct, regarded as a collection of cultural traditions, behaviors, values, and beliefs of a group one belongs to or identifies with (Waters, 1990). Due to the overlap in their definitions, these distinct yet related constructs (c.f. Helms & Talleyrand, 1997; Phinney, 1996) are sometimes used interchangeably. On the other hand, the U.S. Census Bureau defines race and ethnicity as “a complex mix of one’s family and social environment, historical or socio-political constructs, personal experience, context, and many other immeasurable factors.” Given this definition, a person's racial or ethnic self‑identification can change over time and across contexts (Liebler, Porter, Fernandez, Noon, & Ennis, 2017).

Currently, the U.S. Census Bureau classifies responses to the race question into White, Black or African American, American Indian or Alaska Native, Asian, and Native Hawaiian or Other Pacific Islander. According to the Census Bureau’s 2017 annual estimate of the resident population, among the non-Hispanic/Latinx population, 5.6% of the people in the U.S. identified themselves as Asians, 0.7% as American Indian and Alaska Native, 12.5% as Black or African American, 0.2% as Native Hawaiian or Other Pacific Islander, and 60.7% as White (U.S. Census Bureau, 2017). Across racial groups, 18.1% identified as Hispanic/Latinx2 (U.S. Census Bureau, 2017). Additionally, the U.S. immigrant minority has increased from 9.6 million in 1965 to a record 45 million in 2015, pushing this country’s foreign-born share to nearly 14% of the total population. Going by these statistics, we can safely say that racial/ethnic minorities together are on track to make up the majority of the U.S. population. In fact, according to the U.S. Census population projections (U.S. Census Bureau, 2018a), the non‑Hispanic White population is going to shrink, making these minority groups a larger fraction of the U.S. population by 2060.

In conclusion, it will be interesting to see how the upcoming 2020 Decennial Census turns out and how these figures translate into labor-force participation. These demographic changes are important to monitor due to the implications they have not only for the Census (to provide accurate representation of the U.S. population3) but also for organizations (that are generally mandated to account for their workforce demographics). We walk through a few of the changes and provide suggestions for organizations below.

Some Other Race (SOR) Category: A Solution or a Problem?

A growing percentage of U.S. Americans do not select a race category provided on the form. Instead, they select the SOR category, which first entered the form as simply “Other” in 1910. According to the 2010 Census, as many as 6.2% of Census respondents selected only the SOR category to identify themselves. In fact, in 2000 and 2010, the SOR population was the third largest race group.4 Mathematically, this proportion results in millions of people unaccounted for every 10 years. If nothing is done ahead of the 2020 Census, this category could become the second-largest racial group in the U.S. (Ashok, 2016).

The U.S. Census Bureau recognized that racial categories have changed extensively through the decades and intended to improve the accuracy and reliability of its race and ethnicity data through a research project. This project (also known as the Alternate Questionnaire Experiment) tested different questionnaire-design strategies for the upcoming 2020 Census (U.S. Census Bureau, 2012a). As a result, changes were proposed to update the outdated terminology.

The Office of Management Budget (OMB)’s Standards for Maintaining, Collecting, and Presenting Federal Data on Race and Ethnicity (OMB Statistical Directive 15, 2017) describes the racial and ethnic categories that we have come to see on nearly every government form. Fairly recently, efforts were made to update these categories in advance of the 2020 Decennial Census due to the increased usage of the SOR category in data collected by the Census over the years. Last year, a combined question format for Hispanics/Latinx was proposed as they generally view their ethnicity as race, as well as a new category named Middle Eastern and North African (MENA) for inclusion in the 2020 Census (U.S. Census, 2018b). However, these changes will not be implemented.5 This failure to update question categories to represent how people racially and ethnically identify themselves could result in overuse of the SOR category and impact respondents’ willingness to report their race/ethnicity. It may be that people choose the SOR category because they do not agree with the category options that are provided.6 Moreover, we suspect that this decision is a step back from reaching an accurate count of this country’s demographics. Generally, a thorough and correct calculation of a population’s demographic makeup is useful for developing policies that benefit the community at the local level. 

Other than the unavailability of more appropriate categories, the overuse of SOR category could also be due to legitimate concerns over data privacy/confidentiality. In small geographic areas, one’s response to the Census might identify someone; this perhaps may encourage some individuals to choose the appropriate response instead of SOR. That said, the U.S. Census Bureau will continue its research into strategies for the collection of more accurate race and ethnicity data. To address this, the U.S. government can do a better job of informing everyone that the U.S. Census Bureau keeps personal information confidential (U.S. Code, Title 13) even from other federal agencies or law enforcement entities.

What Should Organizations Do About This Issue?

Agencies, including the Equal Employment Opportunity Commission (EEOC), are required to follow the regulation issued by the OMB for reporting race and ethnicity categories. This means that the OMB’s recommended changes to help reduce use of the SOR category described above would have also been relevant to how organizations collect demographic data from their employees. We spoke to SIOP member Romella El Kharzazi (RE),7 PhD, SHRM-SCP about the issue of the overused SOR category and about potential solutions to the issue of missing/incorrect collection and reporting of employees’ race and ethnicity data by employers. She explains:

Employers with 100 or more employees or federal government contractors and first-tier subcontractors with 50 or more employees and at least $50,000 in contracts must file an EEO-1 Survey annually. The EEO-1 requires a complete accounting of an employer’s workforce. If an employee has decided not to volunteer their personal data, an employer may make a visual assignment [29 C.F.R. § 1602.13]. Obviously, it is preferable for employees to provide their own racial and ethnic information, instead of an employer relying upon potentially inaccurate visual assessments. (personal communication, July 7, 2018)

The EEO-1 filing requirement has implications for the work that organizations do around equity, diversity, and inclusion, and helps to ensure they do not violate Title VII of the Civil Rights Act of 1964 (Amended 1991). Thus, to increase accuracy to the best extent possible, employers should give employees the regular opportunity to update their demographic data, as self-identification can change (Liebler, Rastogi, Fernandez, Noon, & Ennis, 2014). This also helps mitigate the problem of missing demographic information.7 It would also be important to discourage organizations from making unethical discretionary decisions about the racial/ethnic classification of their employees (e.g., incorrectly reporting a multiracial employee as a minority without consent with the intent of reporting a higher proportion of employed minorities).9

To increase the accuracy of workforce demographic records, organizations may want to proactively encourage their employees to report their data and be transparent about the reasons why such data are important. For instance, they can share how accurate demographic records help determine EEOC-related compliance as well as how diversity and inclusion initiatives can be developed based on the gathered demographic information.

Generally, employees are administered a demographic questionnaire before they are hired (i.e., when they are applying to the positions); however, the ideal time to collect this info is during each employee’s onboarding, eliminating missing (i.e., “do not want to disclose” category) responses from them. Keeping in mind that the racial/ethnic identity of an individual may change over a period of time (Liebler et al., 2017), including a brief demographic questionnaire in organization’s employee surveys (with an option to decline self-reporting) could be another way of keeping track of and continuously updating workforce-demographic data. With all this being said, some employees may not be comfortable sharing their demographic information under any circumstances. An employer may only then resort to “guessing” employees’ demographic membership based on observation, as Dr. Kharzazi mentioned.       

Multiracial Identity and the U.S. Census

With a growing non-White population and a long history of immigration (both voluntary and involuntary), U.S. Americans are also entering into interracial relationships and marriages at an increasing rate. Last year, multiracial individuals made up 2.1% of the total population (U.S. Census Bureau, 2017). In fact, the segment of the population reporting multiple races (9 million) grew by 32% from 2000 to 2010, compared with those who reported a single race, which grew by 9.2%. (U.S. Census Bureau, 2012b). According to the 2020 to 2060 projections of the U.S. Census, the multiracial group is projected to be the fastest growing racial or ethnic group over the next several decades (U.S. Census Bureau, 2018a).

Since as early as 1860, efforts to categorize multiracial Americans have taken place. For example, the term “mulatto” (i.e., a mixed White and Black ancestry) and some categories denoting different types of Asian heritage were included as response options in the Census questionnaire (Pew Research Center, 2015). Thereafter, despite repeatedly including multiracial categories, Census officials expressed doubts about the quality of data such categories produced and hence removed them. In the 1960 Decennial Census, enumerators categorized people, who were both White and any other race, in the minority race. People of multiracial non‑White backgrounds were categorized according to their father’s race (with some exceptions). In 1970, respondents were asked to mark the race with which they most closely identified from a selection of single-race categories. If they were uncertain, the race of the person’s father prevailed. We must mention that this was the decade during which the U.S Supreme Court legalized interracial marriage in all states with its landmark ruling on the Loving vs. Virginia (1967) case.10 Later in 1980 and 1990, if a respondent marked more than one race category, the Census Bureau recategorized the person to a single race, usually using the race of the respondent’s mother, if available.

The U.S. Census has allowed people of multiracial identity to choose more than one race category to describe their racial makeup since the 2000 Decennial Census. For the first time, individuals were presented with the option to self-identify with more than one race, and this continued with the 2010 Decennial Census. This change led to a total of 2.4% Americans identifying themselves as multiracial in the 2000 Census (Pew Research Center, 2015). Furthermore, the multiracial population increased by about one-third in size since 2000 (U.S. Census Bureau, 2011). As you can tell, attempts to record the races and ethnicities of multiracial individuals have been rather inconsistent until 2000. The Census Bureau is considering a new approach to asking U.S. residents about multiracial identity in 2020 although at this point very little is known about it.

What Does This Mean for I-O Research and Practice?

The Race Card Project, started by Michael Norris in 2010, has captured peoples’ observations and experiences about race into just six words. Some of these six-word stories on multiracial identity (e.g., “Yeah, but you don’t look White”; “Of both worlds, belong to neither”) highlight the need to explore the extent to which multiracial individuals experience discrimination at work and/or in society. It would be worth exploring the impact of multiracial identities on work processes and attitudes if we want to build a society/workplace that is more inclusive. Certainly, the multiracial identity has some bearing on I-O research and practice.

As a researcher, I (Kisha) often struggle with decisions around racial and ethnic categories; specifically, how to measure and account for bi/multiracial individuals during data collection. Currently, a biracial or multiracial category is a catch-all for multiple races and ethnicities. This could become even more challenging as multiracial groups begin to account for a larger and larger proportion of the total U.S. population. However, there may be a silver lining. Increasing numbers of multiracial individuals may allow for more specific multiracial groupings, providing us with the opportunity to better understand workplace experiences among multiracial individuals. For instance, those who identify as White and Asian and those who identify as Black and Asian can be examined separately as opposed to all of them being considered only as “multiracial.”

As someone who works in the selection and assessment arena, I (Bharati) am curious as to how including multiracial options to categorize racial identity will affect certain statistical analyses (e.g., adverse impact, differential prediction) where the variable of interest is generally test taker’s demographic identity (i.e., race). In these types of analyses, a group with a lower selection ratio is considered a minority group (typically a non‑White demographic group). However, as the number of individuals who identify as biracial or multiracial continues to grow, I wonder what will be more appropriate: examining all biracial or multiracial candidates as one group, or separating them into their specific multiracial combinations? If I choose the latter, I further wonder the implications of sample size on the results of these analyses.    

Closing Thoughts

Increasing diversity in the U.S. means that I-O scholars and practitioners should be cognizant of these changing demographic labels, be aware of the current and future challenges associated with defining race/ethnicity, and be prepared to address the implications. We hope that the information and contemplations provided in this article will encourage some of us in our scholarly and professional pursuits.

In the introduction, we mentioned that this article is meant to start a conversation in the field of I-O psychology about how we racially/ethnically categorize people for scientific and practical purposes. These issues are also directly relevant for SIOP. For instance, during my (Kisha) time as chair of the Committee on Ethnic Minority Affairs (CEMA), I examined SIOP members’ racial/ethnic group data in order to inform some of our committee initiatives. I noticed that many SIOP members have not indicated their race/ethnicity on their SIOP profiles. To encourage action, we would like to address this issue here and urge SIOP members to report their racial/ethnic identities on their profiles. These data are essential for SIOP and SIOP committees to better cater to the needs of our members.

Author Note

We would like to thank Jason Marks (Researcher at Amazon.com, Inc.) for his friendly review of the initial draft of this article.


1 In 1942, Ashley Montagu – an anthropologist - published a book named “Man’s Most Dangerous Myth: The Fallacy of Race” arguing that race is a social concept with no genetic basis. This book is marked as one of the earliest works on debunking claims of "genetic inferiority" of other races; a series of findings have demonstrated supporting evidence.

2 Note that The U.S. Census Bureau does not consider Hispanic/Latinx identity to be a race. Ethnicity is asked as a separate question. Census forms currently contain two questions related to race and Hispanic origin, with one asking Americans whether they are Hispanic, Latino or Spanish, and the other asking “What is this person’s race?” with checkboxes and spaces for response.

3 Concerns about LGBTQ representation is another example of this. The 2020 Census will ask if respondents are in same sex marriages or relationships, but will not directly ask individuals if they are lesbian, gay, bisexual, transgender, or any other sexual or gender identity. This will result in transgender, queer, and/or single LGB citizens going unaccounted for and will reduce specificity among married/partnered LGB citizens.

4 This was primarily due to reporting by Hispanics/Latinx who do not identify with any of the OMB race categories. In addition, other demographic groups, such as Afro-Caribbean and Middle Eastern or North African populations, identified themselves as SOR (U.S. Census Bureau, 2017).

5 In January 2018, the U.S. Census Bureau announced that it would not implement the changes, as the OMB did not officially revise its standards; the Bureau had run out of time to wait for possible revisions to the official OMB policy. Refer this for more information: https://funderscommittee.org/wp-content/uploads/2018/02/FCCP-policy-update-Feb2018.pdf

6 Thanks to the anonymous reviewer who raised this point.

7 The personal opinions expressed in this article by Dr. Romella Janene El Kharzazi are hers alone and are not reflective of any other entity. Furthermore, the opinion and any interpretations herein are not indicative of her opinion regarding any past or pending matter and should not be viewed as an indication of any future opinions that she might offer as an expert witness in EEO litigation.

8 For more information, refer to the official EEOC instruction booklet here: https://www.eeoc.gov/employers/eeo1survey/upload/instructions_form.pdf

9 Thanks again to an anonymous reviewer for bringing up this issue.

10 Although numbers of multiracial individuals in the U.S. appear to have increased since the 1960s, this is likely due to prior underestimates of their population size and decreased societal enforcement of the “one drop rule,” which meant that a person was considered Black if they had any African ancestry, no matter how small it was in comparison to their total makeup. This rule made it such that multiracial individuals did not always have the opportunity to self-identify, as society racially categorized them


Ashok, S. (2016, August 27). The rise of the American “others.” The Atlantic. Retrieved from https://www.theatlantic.com/politics/archive/2016/08/the-rise-of-the-others/497690/

Chavez, A. F. & Guido-DiBrito, F. (1999) Racial and ethnic identity and development. New Directions for Adult and Continuing Education. 84(1), 39-47.

Civil Rights Act of 1964, Pub.L. 88-352, 78 Stat. 241 (1964).

Helms, J. E. (1993). Introduction: Review of racial identity terminology. In J. E. Helms (ed.), Black and White racial identity: Theory, research and practice. Westport, CT: Praeger.

Helms, J. E., & Talleyrand, R. (1997). Race is not ethnicity. American Psychologist, 52, 1246-1247.

Liebler, C. A., Porter, S. R., Fernandez, L. E., Noon, J. M., & Ennis, S. R. (2017). America’s churning races: Race and ethnicity response changes between Census 2000 and the 2010 Census. Demography, 54(1), 259-284.

Liebler, C. A., Rastogi, S., Fernandez, L. E., Noon, J. M., & Ennis, S. R. (2014). America's churning races: race and ethnic response changes between Census 2000 and the 2010 Census. CARRA Working Paper. Washington DC: Center for Administrative Records Research and Applications, US Census Bureau.

Loving v. Virginia, 388 U.S. 1, 12 (1967).

Montagu, A. (2001). Man's most dangerous myth: The fallacy of race. Walnut Creek, CA: AltaMira Press.

Norris, M. (2018, May 31). The Race Card Project. Retrieved from https://theracecardproject.com/

Office of Management and Budget Directive 15 (March 1, 2017). Federal Register, 82(39). Retrieved from https://www.gpo.gov/fdsys/pkg/FR-2017-03-01/pdf/2017-03973.pdf

Omi, M., & Winant, H. (2014). Racial formation in the United States (3rd ed.). New York, NY: Routledge.

Pew Research Center (2015, June). Multiracial in America: Proud, Diverse, and growing in numbers. Washington, D.C. Retrieved from http://www.pewsocialtrends.org/2015/06/11/multiracial-in-america/

Phinney, J. S. (1996). When we talk about American ethnic groups, what do we mean? American Psychologist, 51(9), 918.

Records as to Racial or Ethnic Identity of Employees, 29 C.F.R. § 1602.13 (1966). Washington DC: U.S. Government Publishing Office. Retrieved from https://www.gpo.gov/fdsys/pkg/CFR-2017-title29-vol4/xml/CFR-2017-title29-vol4-part1602.xml#seqnum1602.13

Spickard, P. R. (1992). The illogic of American racial categories. In M.P.P. Root (ed.), Racially mixed people in America (p. 12-23). Thousand Oaks, California: Sage Publications.

United States Census Bureau (2011, March). Population distribution and change: 2000 to 2010. Washington, DC: U.S. Government Printing Office. Retrieved from https://www.census.gov/prod/cen2010/briefs/c2010br-01.pdf

United States Census Bureau (2012a, August). Press Kit: 2010 Census race and Hispanic origin alternative questionnaire experiment. Washington, DC: U.S. Government Printing Office. Retrieved from https://www.census.gov/newsroom/releases/archives/2010_census/press-kits/aqe.html

United States Census Bureau (2012b, September). Two or more races population: 2010. Washington, DC: U.S. Government Printing Office. Retrieved from https://www.census.gov/prod/cen2010/briefs/c2010br-13.pdf

U.S. Census Bureau. (2017). Annual estimates of the resident population by sex, age, race, and Hispanic origin for the United States and States: April 1, 2010 to July 1, 2017 [Data file]. Retrieved from http://factfinder.census.gov

United States Census Bureau (2018a, January 26). Census Bureau statement on 2020 Census race and ethnicity questions. Washington, DC: U.S. Government Printing Office. Washington, DC: U.S. Government Printing Office. Retrieved from https://census.gov/newsroom/press-releases/2018/2020-race-questions.html

United States Census Bureau (2018b, March 13). Demographic turning points for the United States: Population Projections for 2020 to 2060 (Report No. P25-1144). Washington, DC: U.S. Government Printing Office. Retrieved from https://www.census.gov/content/dam/Census/library/publications/2018/demo/P25_1144.pdf

United States Code: Title 13. Census: Protection of confidential information. Washington, DC: U.S. Government Printing Office.

United States Equal Employment Opportunity Commission. (n.d.). Fact sheet for EEO-1 survey flers. Retrieved from https://www.eeoc.gov/employers/eeo1survey/fact_sheet_filers.cfm

United States Equal Employment Opportunity Commission. (n.d.). F EEO-1: Who must file. Retrieved from https://www.eeoc.gov/employers/eeo1survey/whomustfile.cfm

Waters, M. C. (1990). Ethnic options: Choosing identities in America. Berkeley, CA: University of California Press.

5063 Rate this article:
Comments are only visible to subscribers.


Information on this website, including articles, white papers, and other resources, is provided by SIOP staff and members. We do not include third-party content on our website or in our publications, except in rare exceptions such as paid partnerships.