One

Domestic Work Between the Wars

The building across the street

has an ordinary facade, a view of the park

and rows of symmetrical spotless windows.

Each morning, the working women come to perform

their duties. They are in starched white,

could pass for vigilant nurses keeping

order and quiet around those about to die.

And each morning, idle women

in pale blue housecoats, frilled and fluffed

at the edges, stare out of double windows,

waiting for something to begin.

With whom would you change places, I ask myself, the maid or the mistress?

—Irena Klepfisz, “Contexts,” in Different Enclosures, 1985

This book describes housework in middle-class homes, and especially relations between housewives and servants, from 1920 until the end of World War II—the final moment when housewives in large numbers could hope to hire another woman to do part of the work designated as “theirs.” In contrast to most accounts of women’s lives in the 1930s, mine stresses that a large group of women continued to employ domestic servants and to define themselves as mistresses, while large numbers of working-class women, especially women of color, saw domestic work as their most likely paid job.

A conservative era so far as women’s roles were concerned,¹ this twenty-five-year span witnessed the perpetuation and heightening of household standards designed for a servant-aided home. Though household technology changed, and though women sought to modernize their homes through better management and equipment, ideals of domestic graciousness and service persisted in popular ideology and were sustained by the chance of hiring at least some servant labor. A majority of households, including those of the women working as domestics, could not afford to buy another woman’s labor; nevertheless, middle-class housewives were considered the norm for all American women. Well-tended domestic space was upheld as the ideal center for social life.

At the end of World War I, feminists, educated middle-class women, and many working-class women had anticipated a liberation from work in the private home. Some had experimented with alternative housekeeping plans during 1917–1918 and expected the war and its aftermath to end access to domestic labor that the private home could afford.² Though Charlotte Perkins Gilman’s demand that the major tasks of the private home be turned over to commercial agencies remained utopian, as did the architectural plans for cooperative apartments, most educated women did not foresee spending most of their adult lives doing housework with house care as their primary concern.

By the mid-1920s, however, alternative possibilities had faded, weakened by the rapid spread of household gadgets and directly countered by political opposition to changing woman’s primary role of non-wage-earning housewife. When the depression added economic policy to popular moralism, most women with education sufficient to obtain good jobs gave up claims to paid work and accepted housework and voluntary activities as their social role, filling days with a rigorous schedule made possible by the expanded number of servants unable to find other jobs.³

World War II freed many working-class women from the necessity of taking domestic work, but it did not liberate middle-class housewives from their primary obligation to care for children, husbands, relatives, and homes—jobs construed as war work such as planning the healthiest use of food, clothing, and morale-raising leisure.⁴ When the second great war ended, these middle-class women found their ideal images and their work firmly attached to housewifery, but with decreased servant labor force. The stage was set for them to follow working-class women into the paid labor force and for increasing tension between full-time housework and more frequent, though often part-time, employment. In the early 1960s, this tension would burst into a renewed critique of private housework, as well as of women’s status in the paid labor force.

The argument thus far summarized is unorthodox. Some historians of housework between the wars have found the central theme in technological innovation and dispersal of goods through commercial advertising and installment buying; they have shown middle-class women accepting household confinement in return for an ease supposedly brought about by newly available home appliances, whether or not these actually lightened housework.⁵ For others, the primary theme is male conspiracy, with manufacturers and physicians, aided by advertisers and abetted by female home economists, giving women an obsessive concern with their children, their furniture, their food, and their attractiveness as sexual and social partners.⁶ Women felt “modern,” according to this view, if they could buy new, streamlined appliances, even if they were doing more of the daily drudgery of cleaning and cooking, as well as more child care.⁷ The common assumption is that women generally were unable to hire domestic service, which became a prerogative of the rich and was, therefore, of only minor interest to women’s history—an evaluation apparently confirmed by the shrinking proportion of women’s employment in private household service. Service remained a significant part of life, this interpretation goes, only in the South, where a labor force was held captive by racial discrimination and a relatively undeveloped economy, and only for the black women who predominated in the occupation.

My story, on the other hand, stresses the transitional nature of the inter-war years. It seeks to understand better how educated middle-class women came to accept a career in the home, rather than a paid job outside, as a sensible use of a college or high school degree and a satisfying adult existence. What happened to the progressive impulses released by higher education and solidified in the suffrage victory? How did an intelligent, well-trained woman come to believe that developing her sexuality (including its maternal side) brought greater emancipation than earning an income? These choices were certainly pushed by many forces, including the cultural power of Freudian psychology’s animosity to lesbianism and praise of heterosexual intercourse; economic recession that reduced job opportunities; political hostility to Bolshevism that was presumed to underlie feminist demands; and feminism’s own confusion about what constituted liberation for women, that precluded escape from, or reevaluation of traditional female roles.⁸ One underestimated factor was the sustained conviction that educated women might become housewives without devoting most of their energy to housework; they would direct servants.

Appliances lured such housewives to accept traditional housework arrangements, but only the ability to turn over much of the work to other women—less powerful in age, class, or race and ethnicity—sealed their acceptance. Instead of cooperatives or commercial services, housewives settled for household equipment and the prospect of some domestic service.

Women able to buy some domestic labor retained thoroughly traditional attitudes about servants. Throughout the period labor reformers and home economics experts advocated businesslike treatment of servants through contracts, limited hours, and higher cash wages, but little improvement occurred in these areas. As women of color came to make up even more of the group of potentially employable servants, patterns of racial dominance and deference formed in slavery, agricultural peonage, and reservation policy may even have reinforced notions of the housewife’s superior intelligence and culture and the domestic’s inferior ability and character. Appliances improved, and relatively scarce workers bargained for shorter working hours and more freedom away from the job. Housework, nevertheless, continued to dominate life for middle-class wives, and they, in turn, sought as much docile service as they could pay for or coerce. Indeed, only the ability to hire some work kept the middle-class woman from looking like a working-class wife whose days were absorbed by housework.

Middle-class matrons not only got work done but also confirmed their self-worth as women through a mistress–servant relationship amplified by class and race distinctions. That such service came from working-class women, the majority of whom were from racially subordinate groups, provided confirmation for the housewife that she was a superior person and would be honored as such so long as she remained a homemaker. Judith Rollins’s Between Women: Domestics and Their Employers describes how social and cultural patterns of maternalism and deference have characterized domestic work throughout its long history.⁹ In the mid-twentieth century, the rewards of the contrast between housewife and servant were as important as the attractions of consumerism to the middle-class woman who decided to spend much of her adult life as a housewife.

The use of a cheap labor source, kept inexpensive by systemic social inequalities, is as important as the spread of electric appliances—perhaps more so as we seek to understand how barriers to interracial and cross-class cooperation were maintained in the past. With middle-class women now in the labor market and facing the need to find caregivers for children and the elderly, the same danger exists in the 1980s—that such women will advance their own economic well-being and social status through the low-paid labor of the low-status, primarily immigrant, older, and dark-skinned females who take these jobs. Documenting the circumstances and means through which middle-class women chose to accept the image of the housewife during the transitional period of the 1920s and 1930s may alert us to contemporary political and cultural dynamics in this era of choice.

A Brief History of American Mistress–Servant Relationships

Housework is women’s work and has been in the United States, as elsewhere, for the past two hundred years. Its antiquity should not lead us, however, to imagine that housework remains essentially unchanged, except for modern innovations of gas and electrical appliances and running water. A domestic revolution accompanied the industrial revolution, and housework changed as much as manufacturing work did. Between 1780 and 1840, much of women’s home production was taken over by factories, leaving the wife to oversee purchase and use of factory-made goods and to turn her attention to child nurturance and household morals. By the turn of the eighteenth century, bourgeois women living on the eastern seaboard in burgeoning cities or small towns along commerical routes became responsible for domestic life more than for family economy. (Women on the frontiers of population movement or in rural economic backwaters continued to be active producers. Civilization’s westward expansion was measured, throughout the nineteenth century, by how much housewives were freed—or displaced—from home production by access to and the means to buy manufactured products.)

Barbara Welter has described ideals of early nineteenth-century, newly urban housewives, norms purveyed through magazines and novels for all proper women, whether married or only aspiring to marriage, as the “Cult of True Womanhood.” Whatever their particular household arrangements or skills, women were uniformly enjoined to be “pious, pure, obedient, and domestic.”¹⁰

By midcentury, growing food, keeping animals, making candles, soap, and cloth—respectable activities for female ancestors—were considered out of place in the urban home. Domesticated women filled their days not with spinning thread but with tending children, shopping for food and furnishings, preparing meals, cleaning up, and supervising the servants who took over as many of these tasks as the housewife felt she could entrust to them.

Directing servants was a new task for many. In the urban environment, as Faye Dudden has shown, a wife’s new or newly enlarged obligations appeared dialectically with new conceptions of servant labor. The wife no longer needed an assistant to help produce goods and process food. She needed a servant to take over much of the physical work of the home so that she could concentrate on expanded social and emotional duties: seeing to her husband’s comfort,¹¹ developing the moral and intellectual capacities of her children, and organizing charities to assist the poor, ill, and untutored. The wife also had to learn to evaluate and purchase food, furniture, clothing, and entertainment for family use. Middle-class wives filled their time with tasks of sociability and consumption that required much of the remaining work of the household to be turned over to a servant (or servants) who could perform manually what the housewife willed.¹²

In northern cities by the mid-nineteenth century, the social rank of households could be ascertained by a domestic as well as an occupational scale. At the top of the scale, homes employed several full-time, live-in servants; at the bottom, homes sent out women to service as “dailies.” Among the wealthiest, the wife did no physical labor. In the middle ranks, wives expected that some portion of their labor could be turned over to other women, even if only for one day a week or for the years when the house was full of young children. In the artisanal working class, wives could not afford to pass on any of the labor, but they might be prosperous enough not to be obliged to work in someone else’s home. Among the poorest, especially newly arrived immigrants and urban Afro-Americans, women were likely to do the work of their own and their employers’ houses.

Southern cities before the Civil War imitated the southern countryside, relying on black slave labor for the work of the house as well as field and factory production. Slave labor made possible an aristocratic ideal of personal service, including a wet nurse to spare the new mother disfigurement from months of breastfeeding. Meals were cooked and served, clothes washed and ironed, children tended, errands run—in the hottest summer heat or the deepest winter cold—for no wages and no restraints but the owners’ good sense or kindness. Even though plantation mistresses remained responsible for production work, much of which they did themselves, service was a job for slaves.¹³ After emancipation the imagery of service remained intact.¹⁴

White Americans were trained in or learned from the southern model after the Civil War, and northern beliefs in the housewife’s duty to supervise labor meshed with southern beliefs in leisured ladyhood. Coincidentally, the rise of corporate wealth created a northern version of the well-staffed household. During the 1890s Mount Holyoke and Wellesley colleges ended a tradition of students’ doing their own housework and hired domestics to tend the students’ dormitories, recognizing that students’ families could afford the extra service and expected their daughters not to need training in domestic work.¹⁵ Racist beliefs in black inferiority combined with racial–ethnic hostilities toward new, non–Anglo-Saxon immigrants. By 1900, white women, as soon as they had enough money, turned their housework over to a variety of less powerful women—Afro-American, Mexican-American, Asian-American, Indian, European immigrant, or rural migrant—whom they perceived as inferior. Just as white men derived status from a racial–class hierarchy, so did white women.

Into the early twentieth century, marriage remained a life goal and domesticity a center of women’s existence, but ideals focused more on man than God. By then, women were encouraged to be beautiful (and youthful), sexually alive (but not immodest), and intelligently charming (but not overly intellectual).¹⁶ Carroll Smith Rosenberg has concluded that the modern woman organized her life much more around men and heterosexual relations, whereas her predecessor lived in a female world designed to complement men’s. By the 1920s, women’s magazines purveyed a picture of sexy, well-educated partners in companionate marriages.¹⁷

The deployment of household labor shifted in accord with a number of economic, social, and demographic changes: substitution of purchased products and services for household-processed goods (e.g., heating canned vegetables instead of cleaning, cutting up, and boiling or steaming fresh ones), using electrical appliances for tasks such as vacuuming, and a generally smaller household because of fewer children and smaller dwellings.¹⁸ The work of housewives became less physically demanding, with or without servants, even though the required time and emotional commitment may have increased, an ironic outcome that Ruth Schwartz Cowan notes in her study of household innovations, More Work for Mother.

As Cowan indicates, the expectation that mothers would spend more time with children was linked to the belief that psychological development depended on mothers’ scientifically informed supervision. Marriages based on the couple’s mutual satisfaction required that wives develop themselves and entertain husbands socially, intellectually, and sexually. In addition to these tasks, middle-class women had to serve meals, keep house at a high level of physical and aesthetic comfort, and manage laundry—all at a higher standard than in preceding decades.

Most housewives sought a subordinate worker to relieve the load, even if only a part-time “daily” or “hourly.” Wives in the 1920s and 1930s still aspired to have domestic help and welcomed having any work removed from their shoulders. Wives reluctantly acquiesced in the transition from servant to wageworker that accompanied their own transition from mistress to housewife.

By 1940, the balance was shifting and housewives were doing more housework, able to hire other women’s labor only for day labor. Postwar films such as Mr. Blandings Builds His Dream House (1948) persisted, though, in depicting typical family life as upper middle class, signaled in this movie by the family’s being served by the live-in, affable black maid Gussie.

A Statistical Profile of Service from 1920 to 1945

“Personal service: Private household” was the category in which the largest number of women worked until the 1950 U. S. census (when it dropped to fourth position). Until the 1980 census, when 51.4 percent of women over age sixteen held paid jobs, the majority of women listed their occupation as “housewife.”¹⁹ The number of domestic servants “increase[d] substantially—roughly by a third—from 1900 to 1940,” with expansion ending during the 1930s. But population expanded even more rapidly so that “the ratio of servants to private families fell 16 percent . . . from 1900 to 1940.”²⁰

Most historians have calculated the proportion of the population able to hire servants by taking the total of servants reported in the decennial census and dividing by the number of families. This method yielded a ratio of about one in fifteen housewives employing servants (about 7 percent) in 1900. Throughout the 1930s, commentators concluded that 5 percent of homes had servants.²¹ By 1950, that possibility had declined to one in forty-two (2.5 percent). All of these figures seem small, and the inference is that most women did most of the work of their households, which certainly explains why power-driven appliances were so eagerly welcomed.

Gross census numbers are deceptive, however, for several reasons. First, all households in the country, including those of domestics, were included in calculating the ratio of servants available, a ludicrous method because it implied that all households, including those of servants, had access to hired domestic labor.²² Planners throughout the period often spoke, indeed, as if every household—or at least those with young, sick, or elderly members or other special problems—should be able to count on domestic aid.

Second, these ratios assumed that each domestic worked as a full-time servant for one family, which was not accurate by the 1920s. After 1920, housewives hired dailies and part-time workers who came for half a day all week or a full day for part of a week. In fact, many domestics worked for two or more families. Though the full-time, regular-duty, live-in or live-out maid remained the image of the domestic, housewives and advice-givers began to admit that a part-time or day worker earning an hourly rate was equally welcome and perhaps more efficient and practical.

Servants were concentrated among particular groups. If the poorest households are excluded from the statistics, the percentage of homes with service increases dramatically, as indicated by 1930–1931 studies of urban, college-educated homemakers, or middle-class families, from 20 to 25 percent of which had a servant.²³ Studies of members of the American Association of University Women (AAUW) and of self-described career women–homemakers published in the late 1920s concluded that two-thirds of the AAUW working wives and 90 percent of the career women who combined marriage, motherhood, and professions had some domestic help (and half the AAUW set had full-time help²⁴). A 1937 survey for Fortune magazine reported that “70 percent of the rich, 42 percent of the upper middle class, 14 percent of the lower middle class, and 6 percent of the poor reported hiring some help.”²⁵

Likewise, a comparison of urban with rural households, which were sharply divided by income during this period, discloses higher concentrations of domestic hiring. Ignoring the 23 percent of farm families in the total population and considering only the 74 percent nonfarm families, that is, people living in small town and urban areas, reveals a distinct urban profile—households with higher incomes and for which servants were the norm. In 1929, 27.8 percent of nonfarm families had yearly incomes of $3,000 and up. Households earning between $2,500 and $3,000 added another 8.9 percent who might compare themselves to families just above them in the income scale and feel that service was something to be bought as soon as added income allowed.²⁶

The young economist George J. Stigler, with the clear eye that would later win him a Nobel Prize, reminded readers that domestic service in 1939 had as many employees as “the railroads, coal mines, and automobile industry combined” and deserved study. Though Stigler found great elasticity in demand for servants, with “savings and expenditures on domestic service . . . the most responsive to increases in income,” he believed that demand was beginning to fall during the 1930s because families were having fewer children and buying more appliances to ease the hardest labor. By 1940, the percentage of paid servants in households out of the total of women in households had fallen from 8 percent in 1900 to 5.9 percent.²⁷

In an effort to reconcile my own impression of continued reliance on domestic servants with the data showing their decreasing availability, I looked for local data and found a source in “Family Disbursements of Wage Earners and Salaried Workers, 1934–1936,” a project conducted by the Bureau of Labor Statistics.²⁸ Drawing a sample consisting of every household in six of the forty two cities surveyed yielded a record of more than sixteen hundred households in six cities from different regions of the country: Portland, Maine; Lansing, Michigan; Jackson, Mississippi; Indianapolis, Indiana; Denver, Colorado; and Los Angeles, California. Three of these, Jackson, Indianapolis, and Los Angeles, had substantial populations of Afro-Americans and Mexican-Americans. The others were racially homogeneous. Interview protocols recorded household composition, employment, earnings, and an exhaustive list of expenditures over a fifty-two-week period.

The sample is especially interesting because it represents the income middle of the population—households with annual incomes between $500 and $3,000, which some commentators considered the minimum needed to hire a servant. In 1929, 22 percent of American households earned more than $3,000 per year and another 19 percent between $2,000 and $3,000.²⁹ Average education levels of wives in these households ranged between tenth and twelfth grade. A scrutiny of women in this lower-middle-class group indicates the depth of middle-class reliance on domestic service and reveals how central domestic labor was to a household’s self-definition as middle class. Moreover, the data point up a truth often hinted at in the 1930s—any woman with enough income would manage to hire some of her work done.

Of these lower-middle-income households, approximately two-thirds in Lansing and Portland could not afford, or did not choose, to hire another woman’s labor. Even in Indianapolis, with a substantial black population, just under two-thirds of households did not hire any help with domestic chores. In Denver and Los Angeles, however, almost half of the households paid for some work, with 45 percent and 42 percent of homes, respectively, having laundry done commercially. In the South, where severe racial discrimination limited black women to domestic work, only 19 percent of Jackson’s white households did not hire some housework labor. Half hired a domestic and half got their laundry done, either by the domestic or by a laundress.

Such a distribution of paid home services is considerably more extensive than the 5 percent figure usually cited on the basis of decennial census statistics. In addition to revealing the persistence of homemakers’ desire for help with their work, it shows the importance of regional economic and racial–ethnic differences in determining how women managed housework.

Many spent regularly for a laundress or commercial laundry service. From a low of 21.8 percent of households in Lansing to a high of 50.7 percent in Jackson’s white homes, women with household incomes insufficient to buy appliances sought relief from those chores by sending the laundry out. Lansing’s low percentage of commercial laundry use may have been supplemented through its domestic service hiring, 16.5 percent of households and the second highest of the six cities (after Jackson), since laundry remained a regular job for many servants.

How much of the laundry chores were done by hired help also varied regionally. In Jackson, almost half of white families paid to have their clothes washed and ironed. In Denver, and among Los Angeles’s Anglo and Mexican families, the most common laundry service purchased was “rough dry,” thus leaving the ironing to the housewife. Poorer families could afford only “wet wash,” which was the most frequent choice among the 16 percent of India-napolis’s black families buying laundry work and was as common a recourse as rough dry and ironed among the 22 percent of Lansing families sending laundry out.

Even these lower-middle-income families purchased some domestic work. Not surprisingly, 50 percent of Jackson’s white households took advantage of the large number of black workers unable to find better jobs to hire some service. Other cities offered considerably less opportunity for housewives to hire help. In Denver, only 6.5 percent of households could afford any domestic help; in Lansing, the percentage reached 16.5, with Portland just over 7 percent, and Indianapolis’s white and Los Angeles’s Anglo households at about 9.0 percent. These households rarely had full-time help.

What factors made a family decide to hire some domestic or laundry help? The most obvious, and the one found to be most significant in my data analysis, was income. Except for black families in Jackson, total household income was a significant determinant of families’ having laundry done; the other significant factor was the number of weeks the homemaker worked at paid employment.

Hiring a domestic servant seems to have been determined by slightly different considerations. Though income remains important, in white and Anglo households the homemaker’s level of education was a significant indicator in every city except Lansing. That a better-educated housewife (one with at least a high school diploma, as my data analysis was organized) had better things to do than spend all her time on housework was part of popular ideology of the 1920s and seems to have been acted on by as many housewives as could afford to do so.

The occupations of households that could hire laundry work out or hire some domestic service vary according to the predominant industries in a particular city; in general, white-collar workers and artisans whose work required literacy and writing or computation hired domestic help.³⁰ In Denver, for instance, most of the husbands able to hire domestic help worked in sales and clerking jobs, including drug clerk, retail clerk, bookkeeper, and credit manager. In Lansing, a more industrial city, husbands supporting households with some paid service worked at clerical jobs but also in skilled artisanal jobs such as machinist, draftsman, toolmaker, crank inspector, and electrician. Portland husbands’ occupations were split between clerical and skilled artisan, including a linesman, chef, and monotype operator. In Jackson, with its ampler opportunities to hire black domestics, the presence of a railway center enlarged the class of regularly employed skilled workers. In addition to clerical, sales, and artisanal jobs, households were headed by men with jobs as motorman, railway switchman, railway engineer, railway hostler, and night yard master assistant. Los Angeles, as the largest city and with the most diverse economy in the sample, in addition to the jobs already mentioned, also had carpenter, electrician, and freezer men’s families with some service.

Many of these households were also in need of domestic help because the homemaker was employed or because the woman designated as homemaker was supporting the home. Homemakers worked primarily in clerical and sales jobs and a few were skilled artisans. Homemakers hiring domestic help were often clerks, retail clerks, stenographers, and salesladies. Jackson, once again, had a diverse list of genteel female occupations: music instructor, commercial telegrapher, elections clerk, and magazine-subscription seller.

Households that could afford only to hire their laundry done were slightly lower on the economic scale. In Portland, homes hiring laundry were supported by carpenters, truck drivers, policemen, laborers, painters, gas station attendants, and mail clerks. Among black households in Indianapolis, those with wives working as maids and domestics and with husbands as janitors, laborers, and servants could hire laundry work done. People who might be expected to get dirty on the job, or who wore uniforms, seem to have hired laundry work done when they could afford it.

Hiring of domestic servants, however, was limited to households supported by workers who did “clean” jobs: supervisory, clerical, technical, and selling. In such homes, not only did working men and women need to have garments appropriate for their work and their work status, but they also did not expect to perform, or to have their spouses limited to performing, dirty physical tasks. Status divisions in the workplace were replicated in private life.

Women who took housework service jobs were usually the least powerful employees in the labor market. Employment in the category “domestic and personal service” declined from 52 percent of women workers in 1870 to 28 percent of the female labor force in 1920.³¹ Separating private household work from personal service work in beauty shops, restaurants, and hotels left an estimated 15.9 percent of women wage earners in service in 1920, 17.8 percent in 1930, and 20.4 percent in 1940. Though exact calculations are difficult because of changes in the census’s instructions to enumerators and its occupational categories, in general the 1920 figure is probably an under-count; private household work, then, probably continued to employ about one-fifth of women throughout the interwar years.³² Its incidence varied greatly among different groups of women. For the majority of both native-born and foreign-born white women, 1920 represented the culmination of a rapid downturn in reliance on domestic and laundry jobs: from 26 percent in 1890 to 8 percent in 1920 for native-born white women and from 52 percent to 22 percent for foreign-born white women.³³

By contrast, domestic work was the primary occupation of women of color—Afro-Americans, American Indians, Mexican-Americans, and Asian-Americans—who were usually untrained to compete for more attractive jobs in clerical work or were excluded from such jobs by racial discrimination. For Afro-American women, the group whose history and geographic concentration were most closely linked with the unremitting servitude of slavery, employment as domestics remained steady from 1890 to 1920, hovering in the high fortieth percentile, and rose in importance in the next twenty years, from 46 percent in 1920, to 53 percent in 1930, to 60 percent in 1940. In the Southwest, Chicanas “worked primarily in domestic service or agriculture until World War II. Forty-five percent of all employed Chicanas were domestics in 1930.”³⁴ Japanese-American women also were heavily represented: 27 percent in the occupation in 1920, 18 percent in 1930, and 10 percent in 1940, though they were dispersing into agriculture, trade, and manufacturing.³⁵ American Indian women, employed as domestics at rates of 11 and 15 percent in 1920 and 1930,³⁶ were not identified separately in occupational statistics, but the content of training programs in Indian schools indicates that Indian women expected to find employment in domestic work.

White women, who remained a significant but not dominant segment of domestic workers, were more likely than before World War I to be native-rather than foreign-born. Women were often forced to take these less attractive jobs because they needed income; they were widowed, divorced, or single, with no male income in the household, or married to men whose earnings were so low that the wife’s income was essential.

My 1934 sample included households of Afro-American domestic workers in Jackson and Indianapolis. Comparing these households with those in which the wife was not employed in domestic work reveals that domestics’ families had lower total income, lower husbands’ income, and lower homemakers’ income than other households. Husbands worked fewer weeks during the year, which was probably the reason for their relatively low earnings. Wives were forced to take domestic work because of family need and lack of alternative occupations.

With the Depression exacerbating family need, breaking up households, and diminishing job options, domestic work that had remained steady through the decade of the 1920s increased during the 1930s. War-related employment promised an end to such work because other opportunities increased, and the occupation became much less significant for women workers in general during World War II, dropping, according to one researcher, from 17.7 percent of all women’s employment in 1940 to 9.5 percent in 1944.³⁷ Even for black women, the occupation was less dominant, employing fewer than half the working population by 1944 instead of 60 percent as in 1940. Ironically, however, domestic work became more completely identified with black women, who increased their labor force participation in all arenas, including an increase of fifty thousand women in domestic service; by 1944, black women made up over 60 percent of domestic workers.³⁸

The years from 1920 to 1945 may be viewed as a transitional period during which middle-class homes changed from being directed by a lady-housewife to being served by the wife. The image of the housewife did not change, but work changed in content, and so did the relations between housewives and servants. Both at the beginning and end of the period the housewife’s existence was to be devoted to her home and family. By the end of the era, however, she was less able to hire another woman to take over many of the physical tasks such devotion entailed. As a consequence, housewives lost some benefits of the angelic image of the middle-class wife, which not only derived from their attachment to the home but was enhanced by the contrast between the housewife and the domestic servant. With declining access to servants, the housewife not only did more work, she also felt herself to be a drudge. Without a servant to emphasize her superiority, this wife found her role considerably less tolerable.

Theoretical Viewpoint

Throughout the interwar and war years, domestic service remained a constant that defined the elevated status of middle-class women and the degraded status of working-class women of color, especially married women of color, who might be expected to have their own homes to care for. Unmarried white working-class women were most likely to seek domestic work, and often in the 1930s they needed a place to live as well as a wage. Few married white working-class women became servants. Unlike the nineteenth century, service did not represent a stage in a woman’s life (doing service as a girl and being a housewife as an adult), but rather sharp divisions in the status of different groups of women. Hiring a maid enabled a middle-class woman to accept confinement to her home and still believe she had escaped traditional domestic servitude. The modern housewife was not so different from the housewife without appliances. Her work status still rested on old patterns of race and class oppression.

This book develops the idea that divisions in status do not represent simply divisions of class or race, systems of domination based on one group’s holding superior resources that enable it to turn over unpleasant or unprofitable work to less well-endowed groups. Rather, inherent in the division of housework with its system of tasks, relationships, and the meanings assigned to them were notions of womanhood, of whiteness and nonwhiteness, and of middle-classness as well as working-classness. Through the social assignment of different jobs to different categories of women and through cultural significances attached to particular jobs performed by different groups of women, women learned their appropriate social identities in relation to those other women. They learned to act and to feel white or nonwhite (a category divided by color, class, and ethnic differences); they learned to fill and to embody their class position; and they learned to exhibit the characteristics of heterosexual females in ways appropriate to their race and class.

Though I think that people of color learn their social parts and internalize them in the same way white people do, how black people do this self-consciously and incorporate resistance to degradation is a topic much investigated since the modern civil rights movement challenged stereotypie notions of “innate” or “inherent” black behavior. The new issues opened to question in the last two decades are how white people learn their social parts and internalize feelings of whiteness. Although this book offers some information about social and cultural structuring of the lives of people of color, my analysis focuses on the question that has been less studied and that is more central to my own identity: how do white women learn to embody whiteness? How are they shaped to fulfill a racially-based (and class-based) ideal of female identity that limits their actions and warps their humanity into narrow shapes?

In this book, therefore, I confront one issue now being studied by numerous feminist scholars: the racialization of white womanhood. Whiteness is now considered as much in need of explanation as blackness or Indianness or Latinness. I assume that no one of the traditional racial-ethnic groupings in America is “natural” or preferable or represents a progressive advance over another grouping. The characteristics of any group are neither static nor essential to that group. Rather, behavior we observe in a group is a social (cultural, economic, legal, and psychological) creation and often one formed in relationship to the behavior and ideals of reciprocal groups. Much of this insight has emerged from gender studies, which have illustrated how notions of male and female are formed as dualisms changing across time and space, with what is female being the opposite, the complement, the Other to what is male. These theories built on earlier work in labor history that pointed out that “class” was a relational concept and not an absolute existence.³⁹ Race and sexuality are the final significant categories of our society to undergo this scrutiny.

The shift to not seeing human life as made up of essential biologically based essences of innate race, homosexuality or heterosexuality, native intelligence, or gender is part of the shift in Western thought sometimes described as postmodernist. This book is grounded in ideas developed in feminist theory that are generally grouped with postmodernist thought—that there is no one single truth or point of view through which historical accounts can be told; that history is an account of how people learn to act as if some parts of themselves naturally require such behavior; that categories of humanity such as race, gender, and sexuality are formed as mutually exclusive dualities in relation to each other; and that historians cannot retreat to biological categories or natural law to explain motivation and behavior.⁴⁰

In following these precepts, Chapter Two seeks to understand what messages and images white middle-class women received about who they were. Chapter Three looks at how, within a society in which race and class divide groups into the more and the less powerful, these women constructed their primary work—housework—and their daily existences so as to confirm their beings through and in relation to other women who had to take jobs as domestic servants. Chapter Four reveals servants’ views of the work and of the women for whom they worked. Chapters Five and Six turn to the public, institutional context within which housewives and servants negotiated their relationships. Chapter Five examines home economics and domestic service training courses funded and overseen by government agencies that taught women their appropriate duties in relation to housework. Chapter Six recounts battles for reform between domestic servants, who organized like other workers to gain regulation of their work conditions, and middle-class women, who responded ambivalently to the call to aid women workers when reform threatened their domestic service arrangements.

Men are present and set the terms of the racial–class negotiations among women by establishing what household-based work men will not do and will not accept as masculine responsibility, and this study points to moments when men’s actions define the limits within which housework can be organized by women. But it focuses on tensions among women and struggles for dominance and equality among women of different race and class groups. In this sense, it examines how white women formed their identities in relation to other women, while leaving their relations with men relatively unexamined. Serving men, however, defines women, and their relationships are built around accomplishing this service. I hope to clarify the postulate that gender is never an identity formed in isolation from other identities that have significance in twentieth-century America, but is an amalgam of race and class with gender. This idea has long been advanced by women of color, whose subordinate position gives them reason to question the “naturalness” of white dominance, just as white women had reason to question the inevitability of male dominance.

The book focuses on the connection in American culture and history between white middle-class women and purity and between working-class, immigrant women or women of color and dirt. These oppositional ideas of womanhood—as angelic or slatternly, frail or strong, virginal or sexual, served or serving—have been connected with images of home and the organization of housework since the mid-nineteenth century. The linked images of good woman and dainty house or bad woman and dinginess show surprising durability and remain vivid even amid the well-furnished consumer paradise of the modern home. The theme of white women’s delicacy runs through more modern representations of them as witty, well-educated, and sexually attractive. They need to be cared for. And they get care from men and from other more robust and vulgar women who seem “made” to serve.

Chapter Seven theorizes about the place of housework in maintaining the psychological differentiation between women of color and white women, between working class and middle class, that is fundamental to the self-esteem of middle-class white housewives. The chapter clarifies why, given this relational identity, middle-class housewives did not protest against doing housework but, nevertheless, denigrated its value. I argue that white women should reconsider the ways in which whiteness has been formed and how it is integrally linked to their subordinate relations with white men. It is pleasant to have privileges, but pedestals are confining. White women have not only given up the masculine parts of their humanity to enjoy their privileges but have thereby accepted narrowly defined notions of womanhood. When class, race, and sexual inhibitions are added to those of gender, women’s freedom to move through the world is greatly limited. When these identities divide women from each other, they reduce their ability to form coalitions for the improvement of all women’s lives.