Guest Post by Ryan H. Nelson

Ryan H. Nelson is an Adjunct Professor of Law at New York Law School where he teaches employment law and in-house counsel at one of the largest global providers of insurance and employee benefit programs where, inter alia, he provides advice and counsel on affirmative action in the workplace.  He received his J.D., cum laude, from the Benjamin N. Cardozo School of Law, Yeshiva University, and his B.S.B.A. with a major in Economics from the University of Florida.


The early 2010s seems like ages ago.  Then-reality television star Donald Trump was hosting Season 4 and Season 5 of The Celebrity Apprentice rather than occupying the Oval Office.  President Obama and others were rightly mocking the prospect of Trump’s campaign for President.  And I was arguing that the absence of LGBT population data from the U.S. Census Bureau served as an insurmountable hurdle preventing the U.S. Department of Labor’s Office of Federal Contract Compliance Programs (“OFCCP”) from promulgating workable regulations requiring certain employers (those holding certain contracts with federal agencies) to take affirmative action to employ LGBT individuals.

Why?  Lawful affirmative action in the workplace requires employers to compare their workforce’s demographics with the availability of certain populations that could change those demographics.  Then, in segments of the employer’s workforce where a certain demographic is underrepresented, the employer can legally engage in outreach to flood its applicant pool with candidates from the underrepresented demographic.  This way, employers increase the odds of changing their workforce to be better representative of the population without considering a candidate’s race, sex, or any other protected classification in its hiring decisions.

For example, if some Silicon Valley tech giant determines that women make up just 8% of its computer programmers, whereas census data demonstrates that women make up roughly 21% of computer programmers in the Bay Area, the company can perform a number of statistical analyses to confirm that women are underrepresented in its workplace.  So, as a best practice (or if the company is a federal contractor, as a legal requirement), the company should try to attract women to apply for computer programmer jobs, hopefully increasing the percentage of women that comprise such roles as a result.

But, without LGBT population data, companies cannot similarly determine whether gay people or transgender people are underrepresented in their workplace.  Even today, there is no reliable data showing the population of gay or transgender computer programmers in the Bay Area.  The best available LGBT population estimates are from UCLA School of Law’s Williams Institute, but these data sets can only estimate LGBT populations based on imprecise census data and, even then, can only slice the data by state (not by statistical areas like the Bay Area or by occupations like computer programing).  Hence, the tech giant in the example above is left to guess whether its workforce accurately represents the LGBT population.

Regulations promulgated by the OFCCP that went into effect in 2015 did prohibit federal contractors from discriminating based on sexual orientation and gender identity, but the agency stopped short of requiring federal contractors to engage in any sort of affirmative action to employ gay and transgender employees.  And rightly so—no LGBT population data existed then, and none exists now, so such affirmative action would be impossible.  Now, the decennial census—the next opportunity to finally fix this problem and secure LGBT population data—is a mere two years away.  Yet, last year, Census Bureau Director John Thompson announced that the bureau would not ask Americans about their LGBT status in the 2020 census, again thwarting the dream of LGBT affirmative action in the workplace until at least the 2030s.  Commenters accurately decried the bureau’s decision for many, many, many reasons, but this post serves as yet another reason why the bureau’s decision is flawed.

Recently, as if to add salt to the wound, Secretary of Commerce Wilbur Ross announced that the Census Bureau would ask all respondents in the 2020 census to identify their citizenship.  Discourse in the wake of this announcement has focused on how asking this question will chill response rates of Hispanics and other racial and ethnic minorities.  True.  And lower response rates for such already-marginalized communities compound existing problems, including the difficulties that companies have engaging in affirmative action to employ racial and ethnic minorities.  Just as the absence of accurate LGBT population data stymies companies from assessing the extent to which gay and transgender people are underrepresented across various segments of their workforce, the absence of accurate race and ethnicity population data would stymie companies from assessing the extent to which racial and ethnic minorities are underrepresented.  What’s worse is that the missteps of this administration with respect to the census will cement inaccuracies in workplace affirmative action and stall workplace diversity efforts for decades to come.

President Trump and the officials that he appointed probably don’t understand the incredible harm that the proposed questions in the 2020 census will inflict upon workplace diversity.  Even if they do understand, I suspect that it won’t bother them much.  But it should bother us.