Who the 1935 Social Security Act Left Out

The 1935 Act kept agricultural and domestic workers out of old-age insurance, which excluded about 65 percent of Black workers and most Black working women, and scholars still disagree about whether race or tax-collection feasibility drove that choice.

The record

• April 1, 1935. In Grovey v. Townsend, 295 U.S. 45, the Supreme Court upheld the Texas Democratic Party's whites-only primary. It held the party's exclusion was not state action, reasoning that "we are unable to characterize the managers of the primary election as state officers." (Justia)
• January 1935. The Committee on Economic Security (CES), chaired by Labor Secretary Frances Perkins, recommended in its report that compulsory contributory old-age annuities "include, on a compulsory basis, all manual workers and nonmanual workers earning less than $250 per month," "except those of governmental units and those covered by the United States Railroad Retirement Act." The report did not exclude farm or domestic workers. (CES Report, SSA History)
• Before the report. According to DeWitt, the CES's own Old-Age Security staff and its advisory council had recommended excluding agricultural and domestic workers. The Committee overruled them after Perkins and Harry Hopkins objected. (DeWitt 2010)
• February 5, 1935. Treasury Secretary Henry Morgenthau Jr. told the House Ways and Means Committee, speaking for the Bureau of Internal Revenue, that the exclusions were needed because of tax-collection problems. DeWitt calls this testimony a surprise. He writes that "essentially all the members of both committees" backed the exclusion, "not just Southerners," and that "No Southern member of the Ways and Means Committee spoke out either in favor of or against Morgenthau's proposal." (DeWitt 2010)
• August 14, 1935. The Act was approved. Section 210(b) defined covered "employment" so as to exclude "(1) Agricultural labor;" "(2) Domestic service in a private home;" plus casual labor, crew service on vessels, federal and state government service, and service for certain religious, charitable, scientific, literary, or educational nonprofits. (Social Security Act of 1935, SSA History)
• Title I (Old-Age Assistance) as enacted. The administration had proposed requiring state assistance compatible with "decency and health." DeWitt reports that "Some Southern legislators found this language potentially threatening to economic and social arrangements in their region," and he concedes that "Much of this concern may well have been racially motivated." That phrase does not appear in the enacted Title I, which left benefit levels to the states. (DeWitt 2010; 1935 Act text)
• Scale of exclusion. DeWitt reports the Act "applied to about half the jobs in the economy." Using 1930 Census data, "about 65 percent of gainfully employed African Americans worked in the agricultural or domestic sectors," while 27 percent of white workers were also excluded. Of 15,514,449 workers in those two sectors, 3,564,044 (23 percent) were Black and 11,460,906 (74 percent) were white. He also cites a Social Security Board estimate of 20.1 million excluded workers in all categories, of whom "at least 15 million were white." (DeWitt 2010)
• Women and domestic service (1930 Census). A Social Security Board study found 2,326,857 domestic workers in private homes in 1930. Of these, 44.5 percent were Black and 53.9 percent were white, and women made up 90.7 percent. Black women were 3.8 percent of all gainful workers but 41.0 percent of private household domestic servants. (Needleman, Social Security Bulletin v2n3, c. 1939)
• Women in household work (1940 Census). Social Security Board Chairman Arthur Altmeyer wrote that household work employed about 11 percent of white working women and nearly 60 percent of Black working women, and nearly 18 percent of all women in the labor force. In a Chicago survey, fewer than 2 in 100 white household workers were insured by mid-1941. (Altmeyer, Social Security Bulletin v8n1, c. 1944)
• Voting access around 1935 to 1944. The Legal Defense Fund's Thurgood Marshall Institute reports that in 1940 fewer than 1 percent of eligible Black residents were registered in Alabama, Louisiana, Mississippi, and South Carolina, and about 200,000 Black people were registered across the South in 1944. (Thurgood Marshall Institute, LDF)
• December 6, 1937. In Breedlove v. Suttles, 302 U.S. 277, the Supreme Court upheld Georgia's poll tax as a prerequisite to voting. (Justia)
• August 28, 1950. President Truman signed the 1950 Amendments (Public Law 734). They covered "Regularly employed domestic and farm workers." Farm workers qualified with "60 full days of work during a quarter" for one employer, and the amendments extended coverage to "approximately 10 million additional persons" overall. (SSA History, 1950 Amendments; SSA Annual Statistical Supplement 2005, Table 2.A1)
• September 1, 1954. President Eisenhower signed the 1954 Amendments (Public Law 761). They covered about 3.6 million self-employed farm operators. A farm worker was covered if "paid at least $100 in cash wages by that employer in a calendar year." Domestic workers were covered if "paid $50 in cash wages by an employer in a calendar quarter," which dropped the earlier 24-day test. (Cohen, Ball, and Myers, Social Security Bulletin v17n9)

What this case shows

1. The people left out had the least electoral voice. About 65 percent of Black workers were in the excluded sectors (DeWitt 2010). In the same year the bill passed, the Supreme Court upheld a whites-only primary (Grovey v. Townsend), and in 1937 it upheld the poll tax (Breedlove v. Suttles). In four Deep South states, Black registration was still below 1 percent in 1940 (LDF TMI). Whatever the drafters intended, most excluded Black Southerners could not vote against the members who wrote and passed the bill. This point does not depend on resolving the intent debate.
2. The burden fell hardest on Black women. Black women were 3.8 percent of gainful workers but 41.0 percent of private household servants in 1930 (Needleman). By 1940 nearly 60 percent of Black working women were household workers (Altmeyer). These were workers facing barriers of race, sex, and class all at once, and many could vote only where race did not block them.
3. Where Southern influence is documented, it worked to protect local control. DeWitt himself concedes that Southern objections to the "decency and health" standard in Old-Age Assistance "may well have been racially motivated," and the standard was dropped (DeWitt 2010). This shows the political setting in which an unrepresented population's benefits were decided.
4. Feasibility was a choice, not a fixed fact. The CES itself recommended covering these workers (CES Report). Peer-reviewed work concludes that administrative difficulty was "not administrative impossibility" and points to stamp-based collection used abroad (Rodems and Shaefer 2016). Which difficulties counted as decisive was itself a political judgment, made without input from the workers affected.
5. The gap lasted for years. Coverage did not begin for regularly employed farm and domestic workers until 1950, and broad coverage of farm operators and casual farm and domestic workers waited until 1954, nearly two decades after enactment (1950 Amendments; Cohen, Ball, and Myers). Altmeyer's own 1940s data show how few household workers had earned any protection in the meantime (Altmeyer).

The strongest objections, and our answers

1. Objection: The exclusion came from the Treasury, not Southern legislators. DeWitt shows that Morgenthau proposed it for tax-collection reasons, that CES staff had earlier recommended it, and that "At no point did Southern Democrats create the exclusion or push it through Congress" (DeWitt 2010). Mitigation: This is a strong objection, and this case does not claim that Southern members wrote the exclusion. The point here concerns political voice, not authorship. The workers affected lacked representation whoever drafted the clause, and the claim about voice holds even if the motive was administrative.
2. Objection: Most excluded workers were white. At least 15 million of the 20.1 million excluded workers were white, and 74 percent of farm and domestic workers were white (DeWitt 2010). A rule meant to target Black workers would be a poor fit. Mitigation: This is true, and we say so. The disparity is in rates, not totals: 65 percent of Black workers against 27 percent of white workers. Poor white farm laborers and domestic workers were also among those with weak political power, which fits the broader point that class as well as race shaped who was heard.
3. Objection: Other countries excluded these workers too, and wage-tax collection from scattered small employers was genuinely hard in 1935. Rodems and Shaefer note that no compulsory unemployment insurance program before 1935 covered these groups and that Britain excluded them early on (Rodems and Shaefer 2016). Mitigation: Feasibility was a real constraint. Even so, the CES had judged coverage workable, and alternatives such as stamp books existed. The fair claim is that feasibility concerns and an unequal labor market together produced the outcome. A strong claim of intent goes too far.
4. Objection: Silence in the record is not evidence of motive. DeWitt notes that no Southern member of Ways and Means spoke on the proposal and that Pat Harrison's few comments concerned farm proprietors (DeWitt 2010). Racial-intent arguments may rest on inference. Mitigation: This applies in both directions. Silence can neither prove nor disprove motive. Scholars such as Lieberman, Katznelson, and Poole read the broader pattern of New Deal legislation, while DeWitt reads the Title II record closely. This case presents both readings and stakes no claim on intent.
5. Objection: The coverage gap was corrected within two decades. The 1950 and 1954 Amendments brought these groups in (Cohen, Ball, and Myers). Mitigation: That is true, and it is worth noting that the system expanded. Still, workers who retired in the interim lost out, and the 1954 rules kept earnings thresholds ($100 a year from one farm employer, $50 a quarter for domestic work) that the lowest-paid casual workers could miss.

Open questions: what would change our minds

• Would internal Bureau of Internal Revenue or Treasury memos show whether Morgenthau's February 1935 proposal was shaped by consultation with congressional leaders? Evidence of such coordination would strengthen the intent thesis. Evidence of purely internal origin would strengthen DeWitt's account.
• We have not checked Mettler's Dividing Citizens (1998) or Katznelson's book directly for their specific figures and archival claims. A direct reading may sharpen or weaken argument 3.
• Disenfranchisement figures for 1935 itself, by state, would fit better than the 1940 and 1944 figures used here. Primary registration data (such as state reports or U.S. Commission on Civil Rights compilations) would improve precision.
• How many excluded workers still failed the 1954 earnings thresholds? Data on that would show how complete the "completion" really was.
• If evidence showed that excluded workers or their advocates had effective channels of influence (for example, through Northern Black voters or labor allies) and chose not to press for coverage, argument 1 would need revision.

Sources

Found an error or a better source? We keep every correction on the record with its date and reason. Nothing here is quietly rewritten.