I am very pleased to welcome to ELB Book Corner several contributors to The Oxford Handbook of American Election Law (Eugene D. Mazo ed. 2024). The 30% discount code for ELB readers is ALAUTHC4. The eighth and final contribution is from Justin Levitt:

The extraordinary handbook overseen by Gene Mazo strives to offer a comprehensive review of American election law: constitutional foundations, treatment of particular groups, campaigns, voting rights, election administration, redistricting, campaign finance, you name it. It’s a wonderful resource: a guide to the field and its development, a snapshot of where we stand, and a look to where we might be going.

And so I was particularly delighted when Gene asked me to write a chapter about the Census, which is in many ways the underappreciated constitutional foundation stone. I’ve spoken often about the law of democracy as the infrastructure of infrastructure: the rules by which we do everything else that we choose to do (or not do) together. Which, I think, makes the Census the infrastructure of the infrastructure of infrastructure.

The mandate to conduct a Census appears in the fifth and sixth sentences of the U.S. Constitution. It’s the very first enumerated affirmative obligation of the new federal government: before the power to coin money or regulate commerce, before the power to command armies or make treaties, before the power to decide cases and controversies. It determined representation in government and direct taxation, and tied the two to each other — and even if its role in direct taxation is a bit less relevant today, it now governs the disbursement of trillions of dollars in funding and sets the informational architecture for countless public and private decisions. The census helps us understand who We the People are.  And that understanding drives what we choose to do together, and how.

The chapter covers an array of topics related to the Census, including who gets counted where, how the count is conducted (and many of the difficulties and controversies attending the 2020 Census), and the presentation of all that data to the public (including legal safeguards like a protection against adverse use that the courts have repeatedly gotten wrong). The piece owes quite a bit to Nate Persily’s work on the law of the Census, but also to historians and to scholars of big data. And it’s informed by my own service in government; though the Census Bureau made its own decisions with respect to the issues covered in the chapter when I was in federal service, I had the benefit of a front-row seat to some of the decisions it made. (And I’ve not been shy about describing the very different approach taken at other times.)

In writing the chapter, I was perhaps most struck by the degree to which form has repeatedly driven function. (As a scholar of procedure, I’ll readily admit that I should have been less surprised.) The way that the Census presents information affects both the inputs and our understanding of the outputs. That started in the very first Census, in 1790. Congress told enumerators to “distinguish[ ] the sexes and colours of free persons, and the free

males of sixteen years and upwards from those under that age.” But in the next breath, Congress required the use of a table that didn’t actually do what Congress had purported to direct: the form for reporting the data “distinguish[ed]” the age of white men only, distinguished the sex of white people only, and lumped all other unenslaved people together into one gray sludge of imposed featureless fictitious homogeneity.

Form driving function has been a facet of Census decisionmaking — particularly though not exclusively with respect to race and ethnicity — in the intervening 234 years. The chapter follows Census grappling with socially constructed categories (for example, the criteria for determining whether individuals born in Mexico were “white” changed from 1920 to 1930 to 1940), and the ramifications of its decision to move from primary reliance on enumerators to primary reliance on us to say who we are, including how to present the intricate detail with which we define our own identities, when we’re given choices other than “Other.” A change in 2020 effectively expanding character limits for the intake form yielding detailed racial and ethnic responses showed an America with a much more diverse racial heritage, due less to demographic change and more to revealing the diversity that had been present but unpresented all along. New OMB decisions will enrich the portrait further. 

This is also a thread that ties directly to the controversies over differential privacy, also explored in the chapter with particular focus on the impact for controversies in election law. The presentation of the information we get drives the nature of the information we have. And really digging in to those difficulties may help to explain why a commitment to more transparency can mean a bit less transparency along the way.

Writing this chapter was tremendously enlightening, and not a little fun. I hope that comes through in the chapter — we could all use a little more appreciation of Infrastructure of Infrastructure of Infrastructure Week.