After reading Davies & Derthick, comment on the following:
What is “institutional racism” and what is its relationship to current and past social welfare policy?
There have been many different suggestions for how to “fix” Social Security. What exactly is wrong with Social Security and what are some options that you can locate (
yes, you will need to look for these)
and which do you think might work & why?
a) What choices did you make to “fix” social security in Social Security Reform Game?
b) Were they similar or different to the suggestions you made for in Discussion 11.2? In what ways?
c) In the game, were you able to “fix” Social Security?
d) How do you feel about the choices you made?
e) What other choices might you make in addition/instead (e.g., other ideas to “fix” the program)?
1) In your own words, what are the differences between Medicare & Medicaid? (include specifics, eligibility, funding sources, etc!).
Even when paraphrasing, remember to cite all your sources as appropriate.
2) What challenges are each of the programs individually facing currently? Why?
3) What suggestions do you have (or have you heard – cite sources) to address some of these challenges?
4) What impact(s) will (a) the current challenges and (b) the possible solutions suggested above have on:
Your personal life currently? Why?
Your future? Why?
Older adults? Why?
Race and Social Welfare Policy: The Social Security Act of 1935
Author(s): Gareth Davies and Martha Derthick
Source: Political Science Quarterly, Vol. 112, No. 2 (Summer, 1997), pp. 217-235
Published by: The Academy of Political Science
Stable URL: http://www.jstor.org/stable/2657939
Accessed: 02/06/2010 18:04
Your use of the JSTOR archive indicates your acceptance of JSTOR’s Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR’s Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact firstname.lastname@example.org.
The Academy of Political Science is collaborating with JSTOR to digitize, preserve and extend access to
Political Science Quarterly.
Race and SocialWelfarePolicy: The Social
SecurityAct of 1935
GARETH DAVIES AND MARTHA DERTHICK
Recent years have seen an outpouring of scholarship devoted to
the origins of United States social policy, much of it focusing on the alleged
underdevelopment and bifurcated character of the America welfare state. One
group of social scientists has developed what Theda Skocpol calls a “politycentered” approach, according to which the policy decisions and institutional
legacies of one era condition and constrain subsequent policy making.’ A second
group of scholars has highlighted the role that gender distinctions, and challenges
to those distinctions, played in shaping welfare innovations and backlashes during
the formative years of the welfare state.2 And a third cluster of social scientists
has pioneered an approach that centers on the racial politics of welfare.’
It is with this third group thatthe present article is concerned. That the politics
of race and the politics of welfare are intertwined in contemporary America is
incontestable, and it is equally apparent that their interrelationshiptoday has to
be understood in historical context. Scholars such as the sociologist Jill Quadagno
‘ See introductionto Theda Skocpol, Protecting Soldiers and Mothers: The Political Origins of
Social Policy in the UnitedStates (Cambridge,MA: HarvardUniversity Press, 1992); MargaretWeir,
Politics and Jobs: TheBoundariesof EmploymentPolicy in the UnitedStates (Princeton,NJ: Princeton
University Press, 1992); and introductionto Weir, Ann Shola Orloff, and Skocpol, eds., ThePolitics
of Social Policy in the United States (Princeton, NJ: Princeton University Press, 1988).
2 See, for example, Mimi Abramovitz,Regulating the Lives of Women:Social WelfarePolicy from
Colonial Timesto the Present (Boston: South End Publishing, 1988); Linda Gordon, Pitied But Not
Entitled:Single Mothersand the Historyof Welfare,1890-1935 (Ithaca,NY: Cornell UniversityPress,
1994); Gwendolyn Mink, The Wages of Motherhood:Inequality in the Welfare State, 1917-1942
(Ithaca, NY: Cornell University Press, 1995).
3 See next paragraph.
GARETHDAVIES lecturesin Americanpolitics andhistoryat the Universityof Lancasterin England.
MARTHA DERTHICK is the Julia Allen Cooper Professor of Governmentand Foreign Affairs at
the University of Virginia and the author of The Influence of Federal Grants. Public Assistance in
Massachusettsand Policy Makingfor Social Security.
and the political scientist Robert Lieberman have, therefore, opened up a promising and important line of inquiry as they explore the way in which race has
informed the politics of social policy during the period since the federal government first entered this arena in a big way in the 1930s.’ Other contributors to
this debate include social scientists, such as Mimi Abramovitz, Linda Gordon,
and Gwendolyn Mink, whose primary interest is in gender but who also have
been attentive to the racial context of social policy making.’
Quadagno believes that race has played a crucial role in undermining federal
social policy agendas for over sixty years, and that this persistent pattern can
tell us much about the politics of welfare today. Accordingly, while her recent
book The Color of Welfareis primarily concerned with the fate of Lyndon Johnson’s War on Poverty, she is also concerned to place its rise and fall in a broader
historical context: “The motor of American history has been the continual reconfigurationof racial inequalityin the nation’ssocial, political, and economic institutions. It is this characteristicthathas impeded the development of a comprehensive
welfare state. “6
Quadagno believes that the relationship between race and social provision
first became important during the New Deal. Franklin Roosevelt extended new
“social rights”to the industrialwhite working class; but the price for this advance
was paid by African Americans, who were consistently discriminated against in
New Deal programs and agencies. But New Dealers did more than capitulate to
white racism. In addition, they set in place programmaticand institutionallegacies
that widened the existing racial division within the working class, dimming prospects for future social innovation.
Linda Gordon sees the Social Security Act of 1935 as the beginning of a
forty-year process via which the United States developed the stratified welfare
system that we know today:
[I]n1935, SocialSecurityexcludedthemostneedygroupsfromall itsprograms,even
the inferiorones. Theseexclusionsweredeliberateandmainlyraciallymotivated,as
Congresswas thencontrolledby wealthysouthernDemocratswho were determined
to block the possibilityof a welfare system allowingblacks freedomto rejectextremely low-wage and exploitive [sic] jobs as agriculturallaborersand domestic
The same claim undergirds the recent work of Lieberman and Quadagno. They
argue that the Roosevelt administration,though it initially favored a comprehen4 See Jill S. Quadagno, The Color of Welfare:How Racism Underminedthe Waron Poverty (New
York: Oxford University Press, 1994); RobertC. Lieberman,”Race, Institutions,and the Administration of Social Policy,” Social Science History 19 (Winter 1995): 511-542; and Lieberman,”Raceand
the Organizationof Welfare Policy” in Paul E. Peterson, ed., Classifying by Race (Princeton, NJ:
Princeton University Press, 1995), 156-187.
5See sources previously cited.
6 Quadagno, Color of Welfare, 15.
7Gordon, Pitied But Not Entitled, 5.
RACE AND THE SOCIAL SECURITY ACT |
sive, racially-inclusive system of social supports, capitulated to southern racists
upon whom it was politically dependent. The Committee on Economic Security
(CES), which drew up the administration proposal, had favored coverage of
almost all workers under the old-age insurance and unemployment compensation
provisions of the act. This was to be expected, given the president’s universalist
conception of social security. Congress, however, chose to exclude agricultural
and domestic workers, thereby depriving most African Americans of coverage.8
The same “compromise with racism,” to use Abramovitz’s phrase, occurred in
the case of public assistance. Having first envisioned strong federal standards
in the old-age assistance and aid to dependent children (ADC) programs to ensure
adequate and equitable benefits, Roosevelt again bowed to southern pressure,
agreeing to a highly decentralized program that perpetuated discrimination in
Our primary concern in this article is with these claims about the initial
passage of the Social Security Act, but it might also be worthwhile to summarize
the broader historical claims to which they lead, given their contemporary relevance. Gordon argues that although growing numbers of women and African
Americans were incorporated in the federal welfare state between 1935 and the
1970s, they tended to be consigned to increasingly stigmatized public assistance
programs, particularly ADC (later Aid to Families with Dependent Children, or
AFDC). With social security becoming more generous and popular during this
same period, the bifurcation that had been present at the outset grew more acute,
reaching fruition in 1972 with the passage of Supplemental Security Income,
which “left only AFDC as a maligned ‘welfare’ program.”10
Lieberman sees both the initial bifurcation and its subsequentevolution rather
differently. In his view, social security began as a program that excluded blacks,
but the facts that it was centrally administered and that federal administrators
were relatively color blind meant that it had from the outset the potential to
include African Americans, to the extent thatthe act was amendedor occupational
patterns changed. This has in fact happened, and Lieberman believes that social
security today is “perhapsthe closest thing to a race-blind social program that
the United States has ever known.”11With ADC/AFDC, the picture is quite
different; the divisive politics of race have been fanned by discretionary, decentralized administrative arrangements: “potentiallythe most inclusive part of the
Social Security Act, [it] has invited racial discriminationand has developed toward
retrenchmentand endless controversy.”12In the 1960s a powerful welfare rights
8 Manywhitepersonswere also excludedby this decision. Nine percentof the UnitedStatespopulation
in 1930 was AfricanAmerican,while 21.4 percentof employedpersonswere in agriculturaloccupations.
U.S. Bureau of the Census, Historical Statistics of the United States, 1789-1945 (Washington,DC;
U.S. GovernmentPrinting Office (GPO), 1949), 25 (Series B 13-23) and 63 (Series DI-10).
9 Abramovitz, Regulating the Lives of Women,317.
10 Gordon, Pitied But Not Entitled, 5.
” Lieberman, “Race, Institutions,and the Administrationof Social Policy,” 513.
| POLITICAL SCIENCE QUARTERLY
movement challenged this dichotomy between old-age insurance (OAI) and ADC,
demanding a federally guaranteed income for all Americans. However, it was
unable to “overcome the New Deal legacy of institutional fragmentation, and it
ultimately foundered on the turbulentpolitics of racial conflict that continues to
haunt American social policy. “13
Quadagno also sees the 1960s as a decade when the limits of New Deal social
policy were boldly but unsuccessfully challenged. Lyndon Johnson’s War on
Poverty, launched in 1964, was a response to the problem of “how to reorient
the nation’s social policy agenda so that it could eradicate, ratherthan reinforce,
racial inequality.”4 But this bid to create an “equal-opportunitywelfare state”
was foiled by the very institutionallegacies from the New Deal era that it sought
to surmount. Moreover, the political value of exploiting ratherthan ameliorating
racial difference came to the fore after 1968 with the Nixon Southern strategy;
and the demise of Johnson’s Office of Economic Opportunity in 1973 “wiped
the inner cities off the legislative agenda for the next twenty years. “15 Regarding
the racial politics of social policy in the 1990s, Quadagno is pessimistic. Even
social security, pace Lieberman, remains a racially conditioned program if one
examines coverage and benefit levels: “Rather than transferring income from
rich to poor, it transferredincome from African Americans to whites.”16Looking
to the future, and highlighting the stark economic divisions between the races
today, Quadagno warns that “A comprehensive welfare state that enhances the
rights of citizenship cannot be erected on the foundation of racial segregation.”17
Having provided just a brief introduction to the way that a number of able
scholars have explored the link between race and welfare policy, in the remainder
of this article we will take a closer look at their central claims about the origins
of the American welfare system. If race-based accounts of American welfare
state development do not yet represent the conventional wisdom, they clearly
do constitute an important strain of recent scholarship. Our main purpose here,
however, is to suggest the formidable limitations of racially deterministic approaches to United States policy. The following explanation focuses initially
upon the passage of old-age insurance, andthen on the public assistance provisions
of the Social Security Act.
INSURANCE: THE HISTORICAL AND INSTITUTIONAL CONTEXT
According to Lieberman, “The Old Age Insurance provisions of the Social Security Act were founded on racial exclusion.” He continues:
Quadagno, Color of Welfare, 10 – I 1.
6 Ibid., 160. Quadagnoattributesthis patternto the shorterlife expectanciesof African-Americans.
17 Ibid., 186.
RACE AND THE SOCIAL SECURITY ACT |
Inorderto makea nationalprogramof old-agebenefitspalatableto powerfulsouthern
accededto a southernamendment
excludingagriculturalanddomesticemployeesfrom OAI coverage.Thisprovision
alone eliminatedmore than half of the AfricanAmericansin the labor force and
over three-fifthsof black southernworkers. The systematicexclusion of blacks
throughoccupationalclassificationswas crucialto the passageof the act. 8
This version of events is perhaps intuitively reasonable, and considerable
evidence can be mustered in its support. First, it is true that agricultural and
domestic employees were excluded from OAI despite having been included in
the administrationproposal, and that a majority of black workers were thereby
initially barred from coverage. Second, many other countries by 1935 did include
these same groups, implying that the practical obstacles to their incorporation
were not insurmountable. Third, southern legislators played a crucial role in
rewriting the administrationbill in committee. Fourth, the National Association
for the Advancement of Colored People (NAACP) saw these actions as racially
based at the time: its general counsel, Charles H. Houston, pondering both the
social insurance and public assistance provisions of the bill, observed that “from
a Negro’s point of view it looks like a sieve with the holes just big enough for
the majority of Negroes to fall through.”19
Finally, if one considers other New Deal legislation, it becomes clear that
it frequently left loopholes that permitted discrimination against blacks. Walter
White, secretary of the NAACP, was protesting in 1935 against “widespread
color discrimination under the New Deal.”20 He was referring in particular to
the Agricultural Adjustment Act, but Quadagno’s book reveals a broader and
consistent pattern of discrimination: labor legislation, house-financing arrangements, the state employment services, and the public housing programall tolerated
Nevertheless, if one takes a closer look at the rationale for excluding agricultural and domestic workers from social insurance, the waters become muddied
in ways that cast serious doubt on a race-based explanation. Quadagno and Lieberman each refer to the international context, the latter claiming that “More
than half of the other countries with compulsory, contributoryold-age insurance
systems covered farm workers and domestics.”21This may be true, but the point
also needs to be made that these groups had characteristicallybeen excluded from
coverage at the outset. Indeed, the legislators who revised the administration’sbill
in committee were told by AbrahamEpstein, director of the American Association
Lieberman, “Race, Institutions,and the Administrationof Social Policy,” 514-515.
” U.S. Congress, Senate, Economic SecurityAct, Hearingsbefore the Committeeon Finance, 74th
Cong., 1st sess. (1935), 641.
20 “Bias Against Negro Laid to New Deal,” New YorkTimes, 8 January 1935.
21 Liebermanin Peterson, ed., Classifyingby Race, 163. Quadagnorefers to the same evidence in
“WelfareCapitalismand the Social Security Act of 1935,”AmericanSociological Review 49 (October
for Old Age Security, that “No other country on earth . .. dared to try to include
the agricultural workers and the domestic servants at first.”22
In Britain, the Asquith government’s 1909 pension plan had eschewed the
insurance principle altogether and had done so in part on the grounds that its
adoption would require setting up a complex administrative machinery.23And
when the same Liberal administration did decide to take the plunge in 1911,
unemployment insurance coverage was initially restricted to just seven skilled
trades employing little more than two million workers. Some critics asked Home
Secretary Winston Churchill why coverage should not be made universal, given
his extravagantpraise for the insurance principle. In response, he told the House
of Commons that his actuaries were unable to furnish adequate information on
other workers. It would be futile, however, to wait for the results of additional
research: the only way to demonstratethe feasibility of comprehensive coverage
was to make a start, and he had selected “a definite group of trades which will
admit of steady and practical observation.” He told his critics that “Those . . .
who demand . . . an immediate universal system of compulsory unemployment
insurance are in fact as much the enemies of compulsory insurance as those who
are opposed to it in toto.”24Correspondingly, those who counseled caution were
the best friends of universal coverage. Churchill, placing himself in this latter
category, assured an audience at Manchester’sFree Trade Hall that if the modest
beginning he proposed “proves a success the attempt and the system will not
stop there. It will be extended, and in proportion as experience and experiment
justify its extension, in proportionas the people of this country desire its extension,
it must eventually cover, in course of years, the whole of our great industrial
Clearly the administrative and actuarial obstacles to universal insurance applied as much to old-age provision as to unemployment, and the problem exercised
even the most fervent advocates of the contributory principle. In the United
States, Abraham Epstein outlined one particularobjection to universalism in his
book Facing Old Age: “Itis evident that it can only be made to apply to persons
who are in regular employment. It is next to impossible to collect contributions
from persons who are irregularly employed, from agricultural laborers, from
those who are their own employers, from women who work at home not for
wages, from small merchants, and so forth.”26
U .S. Congress, House of Representatives,EconomicSecurityAct, Hearingsbefore the Committee
on Ways and Means, 74 Cong., 1 sess. (1935), 559.
Jose Harris, WilliamBeveridge: A Biography (Oxford: ClarendonPress, 1977), 101.
24 ParliamentaryDebates, 5th Series, House of Commons, vol. 26, 497.
Churchill’senthusiasmfor universal coverage was nowhere more apparentthan in this speech.
He went so far to say that: “If I had to sum up the immediatefuture of democraticpolitics in a single
word I would say ‘Insurance’.”Winston Churchill, “The Budget and National Insurance,”Speech at
the Free Trade Hall, Manchester, 23 May 1909, reprintedin Churchill, Liberalism and the Social
Problem (London: Hodder and Stoughton, 1909), 311.
AbrahamEpstein, Facing Old Age (New York: Knopf, 1922), 221.
RACE AND THE SOCIAL SECURITY ACT |
Related to the question of administrativecapacity was the issue of “actuarial
soundness.” Social insurancewas based on a range of prior private and voluntarist
schemes-trade union funds, friendly societies, private insurance companies.
Although the inadequacy of such schemes had prompted state intervention in
the first place, social insurance enthusiasts were wont to embrace some core
assumptions of the old order, even as they called for risk pooling, flat-level
benefits, and broadenedcoverage. For one thing, these private economic interests
remained politically important: were “social insurance”to appear more akin to
general assistance than to scientific actuarial principles, its political standing
would be greatly diminished. For another, British New Liberals, like American
Progressives, were attractedby the notion that social progress could be achieved
through the cool, rationalistic approach of new knowledge, and social insurance
seemed to exemplify the intellectual advance that underpinnedthe new activism.
Moreover, the British experience with unemployment insurance confirmed the
fiscal dangers of breaking too dramatically with the actuarial practices of the
private sector: the program became bankrupt in the 1920s in the wake of its
politically popular but financially insupportableextension to workers outside of
the industrial workforce. Even these extensions, incidentally, had not gone so
far that agricultural and domestic workers were included. Influenced by this
unhappy experience with blending the social insurance and general assistance
principles, a 1934 Royal Commission insisted upon their firm separation, and
the dichotomy remained evident in the comprehensive welfare state thatthe Attlee
government introduced in Britain after 1945.
But perhaps a more useful internationalcomparison would be with Canada,
which like the United States was a latecomer to the social security scene; its
federal unemployment insurance law was not passed until 1940.27 Here too,
agricultural and domestic workers were excluded from coverage, along with
hospital workers, teachers, and government workers: only 42 percent of the labor
force was included.28In addition to the two factors of administrationand actuarial
soundness, two further issues contributedto this highly cautious approach. Both
had to do with federalism, and each has obvious relevance to the debate over
social policy south of the 49th parallel. First, the Canadian provinces, like the
American states (and notjust the southernstates), were worried aboutthe implications of social innovation for their labor markets. An inevitable product of a
decentralized and economically variegated polity, this concern existed quite independently of the peculiar racial dynamic in the American South. Second, the
degree to which the federal government in each country had the constitutional
authority to introduce nationwide social assistance programs was unclear. In
1935, the Conservative government of R. B. Bennett had claimed thatthe general
welfare clause of the British North America Act, together with the taxationpower,
Canadadid not adopt old-age insuranceuntil 1951.
Leslie Pal, State, Class, and Bureaucracy:CanadianUnemploymentInsuranceand Public Policy
(Kingston and Montreal: McGill-Queen’sUniversity Press, 1988), 39.
| POLITICAL SCIENCE QUARTERLY
permitted the federal government to introduce nationwide social insurance. The
Canadian Supreme Court, however, had declared the Employment and Social
Insurance Act to be unconstitutional, and the Privy Council upheld the ruling in
In the United States, it was the striking down of the National Industrial
Recovery Act and the Agricultural Adjustment Act that called into question the
constitutionality of the entire New Deal. Frances Perkins, the Labor secretary,
suggested in 1946 that it was “difficult now to understand fully the doubts and
confusions in which we were planning this great new enterprise in 1934.” In
1997, it is perhaps still more difficult to do so, but it may be useful to bear in
mind her recollection that “The problems of constitutional law seemed almost
INSURANCE: THE DEBATE OF
These comparisons suggest that the conservatism of the Social Security Act of
1935 when it came to social insurance may have reflected a lot more than racism.
It is highly misleading to say that the “Roosevelt administration acceded to a
southern amendment excluding agricultural and domestic employees from OAI
coverage.”31As Lieberman himself acknowledges elsewhere in his oeuvre, the
initiative came not from Congress but from Secretary of the Treasury Henry
Morgenthau, whose conservatism was fiscally rather than racially based, and
whose Bureau of InternalRevenue had warned that social insurance simply would
not work if these two groups were included within OAI.32 The train of events
was as follows: CES staff members followed precedent in excluding agricultural
and domestic workers on administrativeand actuarialgrounds, but were overruled
by Perkins and Harry Hopkins, members of the CES, both of whom wanted to
get as close to universal coverage under social insurance as was legislatively
possible, as did the president. Once Morgenthauhad been apprisedof the implications of this decision, he urged the House Ways and Means Committee to adopt
an amendment that he suggested, excluding the two groups on traditional
grounds.33The agreement of the southerners in his audience no doubt reflected
racial factors, but the initial suggestion was not based on race.
29 FrankStrainand Derek Hum, “Canadian
Federalismandthe Welfare State: ShiftingResponsibilities and Sharing Costs” in Jacqueline S. Ismael, ed., The Canadian Welfare State: Evolution and
Transition(Edmonton:University of Alberta Press, 1987), 357.
30 Frances Perkins, The Roosevelt I Knew (New York: Viking Press, 1946), 286.
31 Lieberman,”Race, Institutions,and the Administrationof Social Policy,” 514-515. (This is part
of a larger quote that was reproducedabove.)
For the relevant portions of Morgenthau’stestimony, see House Hearings, Economic Security
33 Edwin Witte, The Development of the Social Security Act (Madison: University of Wisconsin
Press, 1963), 152; Perkins, The Roosevelt I Knew, 297.
RACE AND THE SOCIAL SECURITY ACT |
Moreover, committee members seem to have believed that they were following the will of the president as well as the Treasury secretary in amending
the bill, although neither Morgenthau nor Roosevelt acknowledged as much.
The New York Times reported that Morgenthau’s suggested amendments “were
generally thought to bear the stamp of the administration.”Committee members
wanted clarification on this matter and were worried that if they passed the
administrationbill unaltered, the president would accept sweeping amendments
in the Senate in a way that would embarrass both the committee and the House
as a whole. This had happened on a number of recent occasions.34
But even if it had been southernerswhose initiative removed agriculturaland
domestic workers from the administration’sbill, it might be misleading to interpret
this as an importantdefeat for the administration.From the outset, administration
witnesses demonstrateda strikingly relaxed attitudetoward the text of their Economic Security bill. Perkins, Morgenthau, and the executive director of the CES,
Edwin E. Witte, were among those who encouraged legislators to give it the
most careful scrutiny and make whatever changes seemed necessary, so long as
they were congruent with the basic philosophy of the measure. Perkins outlined
the CES’s philosophy in the following terms: “This committee has felt .
it was primarily the duty of those who are elected by the American people to
determine the policy of the American people.”35And on the opening day of House
hearings, Witte emphasized that the CES had “no thought that this program as
presented should necessarily be passed without any changes”:
The Committee[on EconomicSecurity]intendsto presentto you honestlythe entire
problem.It expectsanddesiresthe cooperationof the Congressin workingout this
problem. If, in your judgment,any parts of this programshouldbe changed, of
course, thatis your privilege. Thatis whatwe expect. Ourpurposeprimarilyis to
give you information,to serve you and assist you in every mannerthatwe can.36
This flexibility brought big dividends in the view of Paul Douglas, a strong
supporterof comprehensive social insurance. Writing in 1936, he noted that the
bill had “emerged in a much clearer and more logical fashion after having been
revised by the committee.”37
The administration’sflexibility is nowhere more apparentthan with the thorny
question of agricultural and domestic workers. It was not just racists and not
even just fiscal conservatives who nourished doubts regarding the desirability
of including agricultural and domestic workers in the 1935 bill. Witte told the
Finance Committee that “We recommend that they be included . . . but, frankly,
the administrativedifficulties cannot be disregarded and you may wish to exempt
these groups at the outset.”38The most outspoken critic of the proposed inclusion
“PutSecurity Bill Up to Roosevelt,” New YorkTimes, 6 February 1935.
House Hearings, Economic SecurityAct, 173.
36 Ibid., 8-9.
37 Paul Douglas, Social Security in the United States (New York: McGraw Hill, 1939), 99.
Senate Hearings, Economic SecurityAct, 198.
| POLITICAL SCIENCE QUARTERLY
of these two categories was Epstein, scarcely an exemplar of southern racism.
Addressing the Finance Committee, he told a respectful audience of the dangers
of inclusion. Farmers and domestic workers obviously needed help “as much
and even more so than industrial workers,” but legislators must be aware of the
“terrifically difficult” logistical problem that they posed. Starkly predicting that
“You are not going to collect it,” Epstein feared “anadministrativeproblem which
will become a fizzle and therefore react ultimately against the whole plan.” He
had an alternative suggestion: “Let us wait at least a couple of years and we will
see if the administrationcan really properly take on the job, and we have acquired
some experience, and the farmers themselves see that it is good for them, and
then we will take them in. . . .39
In light of these points, Congress’s exclusion of agricultural and domestic
workers from OAI and unemployment coverage is eminently unsurprising. To
summarize the nonracial factors that contributed to the decision: legislators had
been told that these groups had always been excluded from new social insurance
programs, and that the Treasury was not capable of administering the CES bill;
the most active private advocate of social insurancehadtold them thatthe inclusion
of these groups might endangerthe long-term prospects of the entire social security
program; Roosevelt was suspected of favoring the Morgenthauamendments; and
administrationofficials, at pains to appear deferential, emphasized that their bill
could only benefit from careful congressional scrutiny. If race contributedto the
enthusiasm of southern legislators for revision, then it was joined with these
Moreover, the creators of social security in the United States did not see
these exclusions as permanent, but ratheras temporaryexpedients. Also excluded
from social security in 1935 were the self-employed, seamen, and employees
of churches, colleges, hospitals, and charities, not to mention all government
employees. These decisions were based not on race but on a range of other
considerations that apply equally well to agricultural and domestic workers: the
familiar administrative difficulties, perceived actuarial obstacles, and the belief
that many members of these groups would not desire to be included in this new
scheme of compulsory taxation. Roosevelt viewed the 1935 act as a foot in the door
and publicly anticipatedthe day when social insurance would be universalized and
the dole vanquished. During the next two decades the principle of social insurance
was embedded and popularizedjust as he had hoped, and consequently extended
to new groups, including, in 1950, agricultural and domestic workers.
THE POLITICS OF PUBLIC ASSISTANCE IN 1935
The evidence for the influence of race is stronger for public assistance than
for social insurance, but still considerably less strong than racial interpretations
suggest. We will assume with other analysts that the crucial event was Congress’s
RACE AND THE SOCIAL SECURITY ACT |
refusalto includeas a conditionof grants-in-aidthatthe statespay “areasonable
subsistencecompatiblewithdecencyandhealth.”But it is worthmentioningthat
had seriouslyintendedto confine the discretion
if the Rooseveltadministration
of the statesandcreatea nationalprogram,it wouldhavehadto addresseligibility
as well as benefitlevels. Thisit didnotdo, exceptto limitresidencerequirements
and set a standardof age (at least 65 years) for old-age assistance.
The moststrenuousobjectionsto the “reasonablesubsistence”languagecame
fromsouthernlegislators,notablytwo Virginians-HarryF. Byrdin the Senate,
who invokedthe doctrineof the states’sovereignty,andHowardW. Smithin the
House, whosetestimonybeforetheWaysandMeansCommitteewas surprisingly
forthrightin the matterof race. Liebermanand Abramovitztell us that this
southernresistanceto federalstandardswas drivenby the desireto keep blacks
in their place. Liebermanbelieves that grandrhetoricabout state sovereignty
such as Byrd used was “merelya stand-infor racial supremacy.”Abramovitz
concurs that “beneaththe states’ rights argument, southern states . . . feared that
such a federalstandardwould requireequal treatmentof whites and blacks.”40
of a nationalstandardof needandassistance.Witte,who testifiedbeforecommittees in both houses and attendedtheir executive sessions, wrote thatByrd was
“supportedby nearlyall of the southernmembersof both committees,it being
very evidentthatat least some southernsenatorsfearedthatthis measuremight
serveas an enteringwedgefor federalinterferencewiththehandlingof theNegro
And then there was Smith, saying that”Ofcourse, in
questionin the South.””4
the South we have a great many colored people.
.” Earlier he had remarked
that$30 per monthfor the averagefarmlaborerin a ruralarea”wouldbe affluence.” “Toput him on a pensionat age 65 of $30 a monthis not only going to
take care of him, but a great many of his dependents, relatives, and so on, who
could much betterbe employedworkingon a farm.”42
Our accountwill put the southerners’position in the largercontext of the
interstatepoliticaleconomyandarguethatfor themto acceptthe administration’s
bill would have been irrational,irrespectiveof race. Althoughsouthernersled
oppositionto the decency and health standard,within Congressonly Senator
Robert F. Wagnerof New York explicitly supportedit. With ample reason,
southernersarguedthatthe bill was biasedin favorof the wealthierstates.Most
of themreactedby strivingto protectthe autonomyof theirstates.Others,whom
40 Lieberman,”Classifyingby Race,” 168; and Abramovitz,”Regulatingthe Lives of Women,”317.
41 Witte, TheDevelopmentof the Social SecurityAct, 143-144. See also Thomas H. Eliot, Recollections of the New Deal: Whenthe People Mattered(Boston: NortheasternUniversityPress, 1992), 110111. As counsel to the Committee on Economic Security, Eliot draftedthe administration’sbill. He
recalls southernopposition to detailed grant-in-aidconditions, but does not link it to race.
42 House Hearings, Economic SecurityAct, 976, 974. Much of Smith’stestimony is ambiguous in
regard to race. He could be referring to impoverishedwhite persons as well.
| POLITICAL SCIENCE QUARTERLY
the advocates of a racial interpretationoverlook, reacted by calling for a wholly
national program of aid.
In contrast to the proponents of a racial interpretation, who tend to focus on
Aid to Dependent Children (ADC) because of its putative association with race,
we will focus on old-age assistance (OAA). The association between ADC and
race had yet to be established in 1934-1935. Mothers’aid, the state-level program
on which ADC was built, was concentrated in the urban North, which at that
time was largely white. A study of caseloads in 1931, covering approximately
half of recipients nationwide, showed that 96 percent of the mothers were white,
3 percent Negro, and 1 percent “other.”Mothers’ aid hardly existed for women
who were divorced, deserted, or never married. It was often called “widows’
aid,” because that is what it was. As of 1931, 82 percent of the families receiving
mothers’ aid for which the mother’s marital status had been determined were
families of widows. Nothing in the administration’sbill in 1935 would have
changed standards of eligibility; the bill did not address marital status, nor did
the CES report hint at any such thing. Southerners may have detected a threat
of federal intervention, beginning with pressure, in the form of grants proffered,
to create programs that were not yet well established in the region. Alabama,
Georgia, and South Carolina had not enacted mothers’ aid laws, while Arkansas
and Mississippi did not have them in operation even though enacted.43But the
threat to create a racially inclusive program at this point was more speculative
We choose to focus on OAA, because that is what Congress did in 19341935. In the Depression-era politics of poor relief, ADC was the tail to the
elephant of OAA. The report of the Committee on Economic Security, which
was the charter for social security legislation, recommended the decency and
health standardonly for OAA. It is not mentioned in the section on aid to dependent
children. The administration’sbill applied it as well to ADC, but this was an
afterthought. ADC grant conditions were modeled after those for OAA, except
that the matching formula was less generous -one-third rather than one-half of
the states’ spending. Congress behaved similarly, reaching decisions for OAA
after extended consideration, and then with little furtherthought replicating them
During the Depression the aged mobilized politically on a massive scale,
bringing intense pressure on both state and national officeholders to act on their
behalf. At the extreme, represented by Louisiana’s Senator Huey Long, some
political figures endorsed national pensions for all elderly persons without regard
43 For descriptionsof the programprior to enactmentof the Social Security Act, see Social Security
in America: TheFactual Backgroundof the Social SecurityAct as Summarizedfrom StaffReportsto the
Committeeon EconomicSecurity(Washington,DC: GPO, 1937) 236; WinifredBell, Aid to Dependent
Children(New York: ColumbiaUniversity Press, 1965), chap. 1; and ChristopherHoward, “Sowing
the Seeds of ‘Welfare’: The Transformationof Mothers’ Pensions, 1900-1940,” Journal of Policy
History 4 (1992): 188-227.
RACE AND THE SOCIAL SECURITY ACT |
to need. But the responsible, moderate position, defined by President Roosevelt
and embraced by Democratic committee chairmen in the House and Senate, was
for a shared federal-state program confined to the most needy. It would use
federal grants-in-aid to the states, a method that was expected to contain and
distribute costs by diffusing the pressure from the aged.
Consistent with a common pattern of policy making in the federal system of
the United States, the administration’sbill borrowed from the law and practice
of leading states in order to create a standard for the rest. The “decency and
health”language was contained in the old-age assistance laws of New York and
Massachusetts.44The bill would have required all states to enact it in order to
qualify for aid. Similarly, the bill promised to pay half of the cost of each individual
recipient’s pension, up to a maximum federal payment of $15 per month, a figure
deduced from the fact that the most generous state-level programs were paying
maximums of $30 per month.
The effect of the federal law would have been to relieve the most generous
states, which were also typically the most wealthy, of half of the cost of their
old-age assistance programs. In the twenty states that had no such programs as
of 1935, which included all of the Deep South states, the federal law was designed
to induce their enactment. These states would have had to begin raising revenue
for old-age pensions, and if they wanted to maximize their federal grant, they
would have had to pay pensions equivalent to those of the most wealthy states.45
Southern states protested that they could not afford to meet the standardthat
was being set, which was a standardof the North and West. Per capita income
in 1929 ranged from $269 per year in South Carolina to $1,164 in New York;
per capita tax collections in 1939 ranged from $12.83 in Mississippi to $36.39
in California.46Representative Smith opened his argument before the Ways and
Means Committee by claiming that if half of the Virginia population over 65
qualified for pensions, and if the state’s share for each were $15 per month
(enough to match the federal offer), the state would have to raise an additional
$10 million per year. This approachedVirginia’s annual total of general revenue,
which Smith put at $12 million.47Other southern states made comparable claims.
The Arkansas legislature, formally pleading its inability to raise the matching
funds that federal grants would require, begged Congress not to require them.48
Furthermore-and this was what outraged Byrd-states faced the prospect
that if federal administratorsjudged their programs to fall below a standard of
Senate Hearings, Economic SecurityAct, 67.
Characteristicsof the old-age pension laws of the states as of 1934 are summarizedibid., tables
14 and 15, p. 50.
46 Erik W. Austin, Political Facts of the UnitedStates Since 1789 (New York: ColumbiaUniversity
Press, 1988), 425; Tax Foundation, Facts and Figures on GovernmentFinance (Baltimore: Johns
Hopkins University Press, 1988), 24th ed., 248.
47 House Hearings, Economic SecurityAct, 974.
48 Congressional Record, 79, 15 April 1935, 5707.
“decency and health,” they would be denied federal grants altogether. Federal
“bureaucrats”would be empowered to tell the states how much they must spend.49
This prospect was no figment of the Virginians’ imagination; it was what the
framers of the act frankly intended. This was evident from the testimony both
of Senator Wagner, who sponsored the administration’sbill in Congress and was
the lead witness before the Finance Committee, and of Witte. Wagner implied
that $40 per month was the correct standardof health and decency, and he tangled
with one senator- Daniel Hastings, a Republican of Delaware – whose state was
spending only $9 per recipient. They had the following exchange:
SenatorHastings.You are expectingus to contributeas muchas $15 and I have a
notion that you would expect us to contributeas much as $25 if we are going to
participatein this at all.
SenatorWagner.I thinkDelawarecan affordit, can it not?
SenatorHastings.Thatis what I supposed.
SenatorWagner.In spite of the fact that you can afford it, you have contributed
only $9 per month.50
Wagner and Witte faced deeply skeptical questioning from other members of
the Finance Committee even before Byrd opened his defense of state sovereignty,
charging that the federal administrator of aid would be the “dictator of State
legislation. “51Witte pointed out that if the committee did not like giving federal
administratorsdiscretion to determine when a standardof “decency and health”
had been met, it could choose to set in monetary terms an explicit nationwide
standard of need and assistance, or it could remove from the bill the authority
for federal administratorsto withhold grants as a means of enforcement. Witte
defended federal administrative discretion on the ground that “it would permit
adjustmentsfor all portions of the country,”but with customary tact and deference
added that the decision “is entirely for your determination.”52
Congress defined its own course. Following the lead of the Ways and Means
Committee, within which the administration’sbill was thoroughly recast, Congress dropped the decency and health standardas a grant-in-aid condition, and
instead provided that grants should be given “for the purpose of enabling each
State to furnish financial assistance, as far as practicable under the conditions
in such State. . . . 5 The administration’sbill had attempted an extraordinary
degree of intrusioninto state policy making and finance; the outcome, presumably
in reaction to this attempt, was an emphatic assertion of the states’ independence.
Besides this phrase in the law, the committee reports and the statements of the
committee chairmen on the floor were explicitly reassuring to the advocates of
49 Senate hearings, Economic SecurityAct, 72-78.
U.S. Statutes-at-Large, vol. 49, 620, 627.
RACE AND THE SOCIAL SECURITY ACT |
states’ sovereignty. “[The bill] does not vest dictatorial powers in any Federal
officials,” the Ways and Means Committee report said.54″[Grant-in-aidconditions] do not involve dictation by the Federal Government, but only establish
standardswhich will make it reasonably certain that the States are honestly trying
to meet the problem of the dependent aged. The administration of the pension
grants is left to the States, as is their amount.”55
To suppose that this action is to be explained simply on grounds of white
southerners’defense of racial superiority requires one to ignore or discount very
heavily the economic interest of all of the poorer states in maximizing their gain
from the new offer of federal grants; the interest of all states in maintaining
autonomy in budgeting and finance; the widespread animosity toward central
administrators,as manifested in the remarks even of so loyal and liberal a Democrat as John W. McCormack of Massachusetts;56and the great practical difficulty
of imposing a national standard at a time when social, economic, and cultural
differences among states were far greater than they are today. It should also be
noted that Congress did not adopt the extreme states’ sovereignty position that
Witte hypothetically put to it; it did not eliminate from the bill authority for the
federal administrative agency to withhold grants as a means of enforcing grant
In debate on the House floor, the central argument of southerners and also
representatives of other poor states57was that the bill discriminated against the
aged in the poorest states by requiring of all states, regardless of fiscal capacity,
the same 50 percent level of matching funds. Southerners typically reacted by
asking that the matching requirement be reduced or even eliminated.58 The
southern position received little support from representatives of wealthier states.
A few populist southerners were to be found among those members of Congress who favored a purely national program. Representative John McClellan
of Arkansas, for example, went beyond the position of his state’s legislature to
argue that the responsibility of the federal government was “directto every Amer”The Social Security Bill H. Rept. 615,” 74th Cong., 1st sess. (1935), 16.
“The Social Security Bill S. Rept. 628,” 74th Cong., 1st sess. (1935), 6.
56 “Ido not want professional social workers of the Federal Governmentcoming into Massachusetts
anddictatingto the old people of my State. . . . The law shallbe administeredby local hands, responsible
to local public opinion.” CongressionalRecord, 79, 16 April 1935, 5978.
57 See, for example, the remarksof RepresentativeHarold Knutson of Minnesota, who presumed
to speak for states in addition to his own. Citing North Dakota and West Virginia in particular, he
complained that the old-age assistance program would “work a great injustice in a numberof states
that are bankruptor nearly so.” CongressionalRecord, 79, 10 April 1935, 366. Per capita income in
North Dakota in 1929 was $365; in West Virginia, $460. Austin, Political Facts.
58 For example, see the amendmentsoffered by RepresentativesWilliam M. Colmer and Aaron
Lane Ford of Mississippi, Congressional Record, 79, 18 April 1935, 5969-5970. For a scholarly
argumentthat the federal public assistance program imposed unequal financial burdens on different
regions and should be revised to allocate grants-in-aidaccording to a principle of equalization, see
Clarence M. Heer, “Financingthe Social Security Programin the South,”SouthernEconomicJournal
| POLITICAL SCIENCE QUARTERLY
ican citizen who comes within the class to be benefited, irrespective of State
citizenship. . . . State boundary lines should not be regarded.”59Representative
Aubert C. Dunn of Mississippi argued that “This bill should be amended so as
to definitely assure our people who reach the age of 60 years and are in need
that they will be comfortable and will not be compelled to depend upon local
politics to give them that which is righteously theirs.”60It is quite conceivable
that the South would have accepted a wholly national program of means-tested
public assistance, but there is no way of knowing, because no responsible national
office holder proposed one. Onerous grant-in-aid conditions that applied to the
states’ own programs were more threatening to states’ sovereignty than purely
Writing on the fiftieth anniversary of the Social Security Act, Wilbur Cohen,
one of its architects, judged the omission of the decency and health standard
“the most significant long-range loss.”61 However, even had the language been
included in the act of 1935, it would have been hard to implement in a substantial
way for at least a generation. Ironically, many states proceeded to incorporate
the decency and health standardin their own laws even after Congress dropped
it.62 That is what the authors of the administration’sbill intended them to doand that is probably as much as the federal executive branch would have been
capable of compelling them to do in the short run. Without supportfrom Congress
and the federal courts, it would have been very hard to force poor states to
meet a particular monetary standardof assistance through the use of executive
discretion. Most likely, the language would have begun to bite only in the late
1960s, when the welfare rights movement developed and turned to the courts to
attack the welfare policies of the states. Then, the language would have given
the litigators something to pin their cases on.63
THE INFLUENCE OF RACE: A SUMMARY
Race appeared furtively in the discussions of 1935. Witte seems to have been
surprised by his encounter with it: “The fact is that it had never occurred to any
person connected with the Committee on Economic Security that the Negro
5 CongressionalRecord, 79, 15 April 1935, 5706.
“TheSocial Security Act of 1935: Reflections Fifty Years Later”in TheReportof the Committee
on Economic Securityof 1935 and OtherBasic DocumentsRelating to the Developmentof the Social
SecurityAct (Washington, DC: National Conference on Social Welfare, 1985), 9.
62 Richard Sterner, The Negro’s Share (New York: Harperand Brothers, 1943), 272-274. Sterner
reportedthat in the early years of the old-age assistance program(1937-1940), Negroes in the South
received benefits out of proportionto their shareof the populationbut thatthe degree of theiroverrepresentation was very likely less than what need would have dictated.
63 R. Shep Melnick, BetweentheLines:InterpretingWelfareRights(Washington,DC: The Brookings
Institution, 1994), chaps. 4 and 5; MarthaDavis, BrutalNeed: Lawyersand the WelfareRights Movement, 1960-1973 (New Haven: Yale University Press, 1993).
THE SOCIAL SECURITY
question would come up [in connection with conditions for old-age assistance].”‘
On the floor of the Senate, Huey Long brought the subject up defiantly, with a
demagogue’s brute candor: “The colored people are among the poorest people
we have. . . . How many of them will ever get on the pension rolls? . . . That
may seem like cheap demagoguery, but I am not afraid to say it.”65Before the
Ways and Means Committee, Howard Smith betrayed a casual, artless racism
by asking that the states be allowed “to differentiate between persons.”66
Only two witnesses before the Ways and Means Committee addressed the
subject directly, the NAACP’s Houston, who asked for coverage of agricultural
and domestic workers under social insurance, and one from the department of
race relations of the Federal Council of Churches of Christ in America, who
asked that Congress prohibit racial discrimination in programs shared with the
states. Neither elicited sympathetic questioning, let alone support.67
Unable or unwilling to engage the subject of race directly, policy makers
engaged it, if at all, indirectly through decisions about the locus of government
responsibility. Insofar as policy was nationalized, the potential for racial fairness
increased. The crucial questions are how far nationalization was carried in the
Social Security Act, and, insofar as nationalizing measures failed of adoption,
to what extent the failure was attributableto the white South’sdesire to perpetuate
a racial caste system.
Title II, which created social insurance for retired persons, was of immense
significance-a breakthrough in the nationalization of American social policy.
As such, it met no significant southern opposition. Whether the failure to cover
agriculturaland domestic workers in 1935 was racially motivated is problematic.
We tried to show that the evidence for such an interpretationis less than compelling. In any case, the temporary omission of such workers was less significant
than was the creation of a new, wholly national program of retirementinsurance.
Had the decency and health standard been enacted in the public assistance
titles, that too would have represented a nationalizing breakthrough, though of
uncertain effect in the short term. Southern opposition to it was undoubtedly
made more intense by racial motives, but we have tried to show that a purely
racial interpretationof the Southern position, let alone Congress’s acceptance of
it, is drastically oversimplified. Southerners had ample economic motivation to
oppose the terms of the administration’sbill.
Moreover, to stress what Congress took out of the public assistance titles
overlooks what it left in. The Social Security Act made poor relief a public
responsibility and delocalized it by requiring state financial participation and
mandatory statewide operation of relief programs. This was a major step toward
more generous social provision-“quite at variance with our past record and
Witte, Developmentof the Social SecurityAct, 143-144.
CongressionalRecord, 79, 14 June 1935, 9294.
Hearings, Economic SecurityAct, 974.
67 Ibid., 796, 597.
history,” Senator Pat Harrison of Mississippi, chairman of the Finance Committee, observed in floor debate.68In the past, the needy aged had received help
from private charitableorganizations or from county or city governments, with or
without state participation. This major step, which used the nationalgovernment’s
money and power to lever the state governments’ money and power, was taken
despite numerous pleas from southern congressmen that their states could not
afford to participate.Members of Congress raised no objections to the requirement
of mandatory statewide operation, which administrative planners judged to be
crucial. On the contrary, when Witte suggested that federal administratorsmight
conclude that “reasonable subsistence” could vary as between urban and rural
localities, Senator Thomas P. Gore of Oklahoma objected that this would lead
to unacceptable inequalities.69 If southernershad wanted to discriminate between
counties more and less heavily black, presumably they would have objected to
the requirement of mandatory state-wide operation.
The Social Security Act also limited the states’use of residence requirements,
a significant step toward nationalization, and opened the way for federal supervision of administrationby stipulating that state plans must “provide such methods
of administration . .. as are found by the [Social Security] Board to be necessary
for the efficient operation of the plan.” Proponents of a racial interpretationhave
attached significance to the fact that Congress in 1935 specifically exempted
from federal supervision the “selection, tenure of office, and compensation of
personnel.”70 Strained to begin with, this interpretation overlooks the fact that
only four years later Congress reversed itself and explicitly authorizedthe requirement of merit systems of personnel administration.7”
No more than Congress was the president prepared for a wholesale nationalization of social policy. Roosevelt never advocated any such thing. His administration sought to bring old-age assistance programs into being in the twenty states
where they had not been enacted yet and to make sure that state programs were
effectuated as such. He took the centrality of the state governments in domestic
affairs as a given, and he wanted them to bear their share of the financial burden.
When in 1939 liberals tried to introduce federal minimum grants of $7.50 per
month per recipient, which would have substantially increased benefits in the
poorer states, Roosevelt protested. “Not one nickel more,” he said. “Not one
solitary nickel. Once you get off the . . . matching basis the sky’s the limit, and
before you know it, we’ll be paying the whole bill.”72
CongressionalRecord, 79, 17 June 1935, 9440.
Senate Hearings, Economic SecurityAct, 82.
70 Lieberman, “Race, Institutions,and the Administrationof Social Policy,” 520.
71 For an account of how federal requirementswere applied in one northern state, see Martha
Derthick, The Influence of Federal Grants: Public Assistance in Massachusetts (Cambridge, MA:
HarvardUniversity Press, 1970).
72 James T. Patterson, The New Deal and the States: Federalism in Transition (Newport, CT:
Greenwood Press, 1981; copyright 1969 by the Princeton University Press), 93. Pattersonattributes
the quotationfrom Roosevelt to ArthurJ. Altmeyer, TheFormativeYearsof Social Security(Madison:
University of Wisconsin Press, 1966), 112.
THE SOCIAL SECURITY
To accept the fundamentalsof federalism may have implicitly acknowledged
the importance of race, depending on how much weight one gives to southern
exceptionalism as an underpinning of the institution itself. Accurately to weigh
the significance of race in social policy making would require taking the analysis
to this deeper level, at which the very form of governing institutions is explained.
However, even without going to that deeper level, the case can be made that
race had a latent influence on the politics and policy outcome of 1935. According
to this argument, it would be the very absence of race from the deliberations
that most eloquently attested to the impact of discrimination on New Deal policy
Had there been more than one African-American in Congress, had there been
any from the South, or black voters in the South in numbers sufficient to influence
white Congressmen, or had the NAACP enjoyed greater influence-in short,
had African-Americans been admitted to full political participation-then debate
over social security might have proceeded very differently. Issues of racial fairness might have surfaced explicitly, producing at least a clearer revelation of
motives if not substantively different policy choices. More pressure would have
existed to include agriculturalworkers and domestics, no matterwhat the administrative difficulties. 73
In 1935, the existing form of federalism per se had a powerful influence on
policy. In choosing to have shared programs for public assistance, the president
and Congress were building on precedent. In creating federal grants-in-aid for
public assistance, they were planting a familiar institution of intergovernmental
relations in new policy terrain. With southerners in the lead, Congress limited
the scope of the proposed innovation, but did not for one moment consider
defeating it in defense of state and local autonomy. To what extent Congress acted
proximately out of racial motives is necessarily obscure, because a discriminatory
political system artificially suppressed consideration of race. But it is precisely
this caveat that should make contemporary social scientists cautious in asserting
the primacy of racial motivation in retrospective analysis, the more so since the
range and probable weight of nonracial factors become so obvious when one
takes the trouble to reconstruct the political debate of 1935. *
73 We are indebted to Richard M. Valelly for urging us to develop this point. The lone AfricanAmericanin Congresswas ArthurW. Mitchellof Chicago. His remarkson the floor indicatedsympathy
for the southerners’position on public assistance: “I do not think it means much for us to pass a law
that will help Illinois, my State, because it might have resources to meet the requirementsof the bill;
and Massachusetts,thathas alreadymet them, and a numberof other States, while the Stateof Alabama
and the State of Mississippi and the State of Minnesota, and other poor States could not meet the
requirements.”CongressionalRecord, 79, part 5, 15 April 1935, 5692.
* The authorswould like to thank EdwardBerkowitz, David Hamilton, and Steven M. Teles for
comments, and the White BurkettMiller Center of the University of Virginia for financial supportfor
Purchase answer to see full