Unmade by the State: How Early Labor Department Records Constructed an Invisible American Workforce
The history of American labor is, in large part, a history of paperwork. Payroll ledgers, factory inspection reports, union membership rolls, and wage surveys form the documentary backbone upon which generations of historians have built their understanding of how this country worked—and who did the working. Yet the archive is not a neutral repository. It reflects the assumptions, priorities, and prejudices of those who created it. Nowhere is that reflection more distorted, or more consequential, than in the records produced—and deliberately not produced—by the early United States Department of Labor.
Founded in 1913, the Department of Labor arrived with a mandate to track, protect, and improve the conditions of American workers. What it actually documented, however, was a carefully bounded category of worker: one who was male, urban, industrial, and, in most cases, white. Everyone outside that category—the hundreds of thousands of Mexican nationals harvesting sugar beets in Colorado, the Black domestic workers cooking and cleaning in Southern households, the Puerto Rican laborers cutting cane in fields that had been American territory since 1898—was largely absent from the official record. Their labor was essential. Their existence, as far as the federal archive was concerned, was optional.
The Architecture of Absence
Understanding how these gaps were produced requires looking not only at what the Department of Labor recorded, but at the frameworks it used to decide what was worth recording in the first place. Early twentieth-century labor statistics were organized around the concept of the "gainful worker"—a category that, in practice, excluded most agricultural labor, virtually all domestic service, and any work performed outside a formal wage relationship. The Bureau of Labor Statistics, operating under the Department's umbrella, focused its surveys almost exclusively on manufacturing and mining industries. The result was a statistical portrait of American labor that captured perhaps half of the actual workforce.
This was not purely a matter of bureaucratic convenience. Congressional debates from the period reveal sustained resistance to extending labor protections—and, by extension, labor documentation—to agricultural and domestic workers. Southern legislators, in particular, fought to keep these categories outside the reach of federal oversight, understanding that the workers concentrated in those sectors were disproportionately Black and, increasingly, Mexican. To document them would have been to acknowledge them as subjects of federal concern. To exclude them was to affirm that they were not.
Fragments Against the Silence
For researchers attempting to reconstruct these vanished workforces, the challenge is methodological as much as historical. When the primary source simply does not exist, the historian must become an archaeologist of absence—reading the gaps in surviving records as evidence in their own right, and triangulating between documents that were never intended to speak to one another.
Census records offer one avenue of partial recovery. The manuscript schedules of the 1910, 1920, and 1930 censuses do enumerate agricultural laborers and domestic servants, though with significant undercounts that historians have extensively documented. By comparing these population figures with the Department of Labor's own estimates of the industrial workforce, researchers can begin to quantify the scale of the exclusion: the difference between who existed and who was counted as a worker.
Newspaper archives provide another fragmentary but valuable source. Spanish-language newspapers such as La Prensa of San Antonio and El Heraldo de México published accounts of labor conditions, wage disputes, and employer abuses that never appeared in any federal report. Black newspapers, including the Chicago Defender and the Pittsburgh Courier, similarly documented the working lives of domestic servants and agricultural migrants with a granularity that the official record entirely lacked. These publications were not created as labor statistics, but they function, in retrospect, as a shadow archive—one maintained by communities that understood they could not rely on the government to remember them.
Company records and grower association documents, where they survive, add yet another layer. The records of the California Farm Bureau Federation and similar agricultural industry organizations occasionally contain employment tallies, wage schedules, and contractor agreements that name and enumerate workers the Department of Labor never saw fit to count. That these records were created to serve the interests of employers rather than workers does not diminish their evidentiary value; it simply requires that researchers read them against the grain.
The Persistence of Statistical Invisibility
The consequences of these early archival choices have not remained safely in the past. The exclusion of agricultural and domestic workers from federal labor documentation in the early twentieth century established a template that shaped subsequent policy in durable ways. The Social Security Act of 1935, for instance, explicitly excluded agricultural laborers and domestic servants from its coverage—a provision widely understood at the time as a mechanism for keeping Black workers outside the New Deal's protections. That exclusion was itself made easier by the fact that no robust federal documentation of these workers existed; they had already been rendered statistically marginal.
The Immigration and Nationality Act of 1965 and the debates surrounding it were similarly shaped by the thin evidentiary record of earlier labor migration. Because the federal government had never systematically documented the scale, patterns, or economic contributions of Mexican agricultural migration in the 1910s and 1920s, policymakers operated with a distorted picture of what "undocumented" labor actually looked like historically. The phantom workforce of the early twentieth century cast a long shadow over the legislative imagination.
Contemporary historians of immigration and labor have increasingly turned their attention to what scholars call "negative evidence"—the study of what records were never created, and why. Projects housed at institutions including the Bancroft Library at the University of California, Berkeley, and the Benson Latin American Collection at the University of Texas at Austin have worked to compile alternative documentary bases, drawing on consular records from the Mexican government, church baptismal registers, and employer contracts filed in state courts. These efforts are painstaking and necessarily incomplete, but they are beginning to produce a more accurate accounting of who was actually present in the American labor force during the decades when the federal government chose not to look.
What the Record Owes
Archival research is, among other things, an act of reckoning. When historians excavate the deliberate gaps in early Department of Labor documentation, they are not merely correcting a statistical error. They are tracing the outlines of a governmental choice—a decision, made repeatedly and at multiple levels of policy, to treat certain workers as administratively beneath notice. The records that were never created did not simply fail to capture reality; they actively participated in constructing it, lending bureaucratic authority to the fiction that these populations were peripheral to American economic life.
Recovering that history requires working at the edges of the archive: in the columns of immigrant newspapers, in the ledgers of agricultural corporations, in the manuscript schedules of a census that was itself imperfect. It requires treating absence as evidence and silence as testimony. And it requires acknowledging that the project of historical documentation is never finished—that every archive reflects a set of choices, and that understanding those choices is inseparable from understanding the history itself.
The phantom workforce was never truly invisible. It was made so, deliberately and consequentially, by the institutions charged with making it seen.
