AUTHOR: | Sharon Sassler |
TITLE: | FEATHERING THE NEST OR FLYING THE COOP? ETHNIC AND GENDER DIFFERENCES IN YOUNG ADULTS’ CORESIDENCE IN 1910 |
SOURCE: | Journal of Family History v21 p446-66 O ’96 |
The magazine publisher is the copyright holder of this article and it is reproduced with permission. Further reproduction of this article in violation of the copyright is prohibited.
PUBLISHER ABSTRACT AB Children’s contributions were an important component of the family economy at the turn of the century. This article uses data from the 1910 Census Public Use Sample to disentangle gender and ethnic variations in coresidence with parents. Bivariate results indicate greater coresidence of women; this reverses after controlling for gainful employment and ethnicity. Work outside the home bought freedom from parents to a significantly greater degree for women. Young men who were best able to contribute financially to the family were more likely to be coresiding, suggesting that they received stronger incentives to remain in the home. Irish and German families benefitted from the presence of sons, whereas Jewish households stood to gain from the contributions of both sons and daughters. Relative to the “new” immigrant groups, Black families relying on the contributions of coresident unmarried children were at a disadvantage.
In early twentieth-century America, many families relied on the contributions of multiple household members to ensure economic sustenance and security. Working children, both sons and daughters, played important roles in supplementing the often inadequate incomes of their fathers or substituting for a father’s absence,(FN1) particularly among immigrant families.(FN2) Children’s contributions to the family economy were often contingent on continued residence with parents, as remitances of either wages or domestic services from non-resident children were far less regular and consisted of smaller amounts.(FN3) Yet research on the family economy has devoted little attention to factors determining whether working-age children remained in the parental home or lived independent of family.
This is surprising, in light of dramatic changes in household living arrangements throughout the twentieth century, particularly among young adults. Evidence indicates that the age of leaving home declined steadily from the 1920s through the 1970s, while the number of unmarried adults living apart from parents increased.(FN4) Although the literature suggests that even before the concluding half of the twentieth century living apart from family has been part of the transition to adulthood for many young people, there is a dearth of empirical research examining factors affecting remaining in the parental home in the early years of the century.(FN5) The present article examines the relationship between life cycle, ethnicity, and residence patterns for never-married young adults living in the United States in the early years of the twentieth century. Data from the 1910 census are used to examine ethnic and gender differences in the living arrangements of never-married young adults. The findings are discussed in light of their consequences for ethnic group stratification.
MAJOR THEMES IN THE STUDY OF LIVING ARRANGEMENTS
Numerous studies of American life in the early twentieth century discuss how families devised strategies for attaining economic security or achieving mobility. Two themes dominate this research. One focuses on how women’s roles altered in adapting to changes in the social, economic, and demographic constraints facing families.(FN6) The second perspective concentrates on the impact of ethnicity in shaping group adaptation strategies.(FN7) To understand differences in the living arrangements of young adults it is necessary to integrate these two themes.
In the industrial era, families often ensured their physical and economic survival by balancing the degree to which women engaged in paid or unpaid labor, depending on overall economic need, opportunities, and the family life cycle.(FN8) Normative gender roles emphasized the home as the optimal location for women, and caring for others as women’s primary vocation.(FN9) Despite widespread social commentary militating against women’s involvement in pursuits outside of the home, substantial proportions of American families in the early years of the twentieth century relied on the wages earned by daughters employed in the paid labor force.(FN10) Reconciling the tension between family needs and cultural norms regarding appropriate behavior for women was especially challenging for families adapting to life in a new country.
Unmarried women living apart from family posed yet another challenge to traditional gender norms. Although social commentators expressed concern over both “wayward girls” and “birds of passage,” as unattached young women and men were often described, the problem of young women living apart from parents was considered especially compelling. “How many of the girls employed in stores, mills, and manufacturing establishments, and other employments of like grade, are practically without homes in the large cities and dependent for a living upon their own earnings?” began the introduction of a special study of women living apart from family situations conducted by the U.S. Department of Labor.(FN11) The rationale for the study soon followed: “there was a moral as well as a material importance in the question of a home and home influence for the young woman forced to earn her own living.”(FN12) Women living outside of their families violated traditional views of the proper sphere for women, suggesting either a need for protection or a desire for autonomy and independence inconsistent with accepted family roles. Although men living independent of families were also cause for concern, their presence did not evoke the same response, such as the creation of the Female Benevolent Society of New York, the Young Women’s Christian Association, Traveler’s Aid Societies, and various rooming situations designed to replicate parental oversight.(FN13)
Family strategies for mediating between economic need and beliefs regarding children’s and women’s roles varied widely by ethnic group.(FN14) The literature frequently asserts that some groups, notably Jews but also Blacks and Irish, assigned a particularly high value to educating children, whereas other groups, such as Italians, preferred to send children into the labor force or retain them in the home.(FN15) However, parental allocation of children’s activities was largely dependent on the continued presence of working-age children. There is a growing body of empirical research on the non-familial living of young children, with a particular focus on racial differences in coresidence.(FN16) There is less systematic documentation of how living arrangements of working-age young adults varied across White ethnic groups. Existing studies for the most part focus either on one location, such as Chicago or New York City, or limit their examination to one or two ethnic groups. This research indicates that large proportions of never-married White young adults, particularly the native-born of native parentage, Irish women, and Italian men, lived apart from family.(FN17)
DEMOGRAPHIC CONSTRAINTS
Residence with parents in the early years of the century was shaped by several factors: age at marriage, mortality of parents, and migration. Much of the historical research on the experiences of young adults comments on the close link between leaving home and marriage, with residence in the parental home a by-product of the hiatus between school completion and assumption of the mantle of adulthood via marriage.(FN18) Evidence of important ethnic variations in the propensity to marry and the timing of that event suggest differential availability of young adults for coresidence.(FN19)
Whether young adults resided with family was also a function of the availability and proximity of relatives. Mortality of parents accounted for substantial proportions of single young men and women living independent of family.(FN20) Yet another factor affecting the availability of parents was migration. In the early years of the twentieth century there was considerable movement, both from farms to the cities as well as international migration. For native-born Americans, relocation from rural to urban areas separated many young adults from their parents. Young women were more likely than their male counterparts to move from rural areas to cities, as they had greater employment opportunities than did their brothers;(FN21) there is also evidence that Black children, especially daughters, left home earlier than did children in White families,(FN22) leaving the farm for life in the city.
International migration also was a frequent cause of family disruption, separating young adults from their parents. Many foreign-born men and women left their families behind in their country of origin to seek their fortunes in the United States. Immigration patterns, as well as migration chains resulting in family reunification, varied widely by both ethnic group and gender.(FN23) Among some groups, male migration was dominant; the dearth of women meant few families with which to live, whether as sons or husbands, and an increased likelihood of non-family living. For example, approximately 80% of all Italian arrivals between 1880 and 1910 consisted of men;(FN24) Italian women were extremely likely to take in boarders, providing living arrangements for compatriots without family.(FN25) Other groups were more likely to arrive in family units, increasing the odds that unmarried adult children would live with parents until marriage; Jewish migration included substantial shares of both women and children.(FN26) Unlike most other ethnic groups, women accounted for a slight majority of migration from Ireland as of the 1890s; however, data indicate that Irish women were more likely to arrive with sisters or cousins than parents.(FN27) Additionally, although chain migration reunited some families, migration strategies varied widely by ethnic group. Whereas some groups either arrived with or brought over family members, others were characterized by high rates of return migration; among those desiring to return to the home country, family leavetaking was a short-term strategy that enabled greater savings.
ECONOMIC CONSTRAINTS
A child’s decision to leave the parental household depended in part on the cost and availability of alternative living arrangements, conditions in the family home, opportunities in another location, and ability to be financially independent. Although young adults may have moved to the city as part of a family earning strategy, others left the home to better their individual prospects, find greater excitement than was generally available in rural areas, or escape from restrictive living situations. From the child’s perspective, living apart from family afforded greater independence and autonomy over earnings, as well as additional financial constraints posed by the need to pay for rent, board, and perhaps other services formerly provided by family members; for parents, having an adult child live apart from family reduced access to both the wages of the secondary worker and other forms of contributions to the family, such as domestic services.
Work activity and schooling were important determining factors in the residential decision making of families and individuals. Although young adults, both those living with parents and those residing apart from family, were quite likely to be employed in the early years of the twentieth century, work in the paid labor force was no doubt even more necessary for residentially independent young adults. Those living with family were better able to pursue schooling or engage in other activities, such as work in the home. In addition to individual need, family considerations also determined young adults’ pursuits. Newly arrived immigrant groups were more dependent on the contributions of working-age children than were established residents, whereas longer-term residents of the United States were better able to allow children to pursue other activities, such as schooling, or to remain in the home.(FN28)
The structure of available housing also shaped young adults’ residential opportunities. Prior to World War I, the great majority of the housing stock consisted of two-to four-bedroom houses or flats; small apartments were rare.(FN29) Young adults living independent of parents often resided with other family members, or boarded in another family’s home. However, employers were often leery of women who lived apart from family; as wages were not adequate for independent support, employers were concerned with the moral character of residentially independent young women. Given social pressures for women to remain under the protection of family, many women informed employers that they were living with family when they were not.(FN30)
Numerous women also took positions as live-in domestics, readily available unskilled work that resolved the need to pay for housing. Black and Irish women were most likely to be employed as domestic servants. The census of 1900 indicates that 43% of Irish-born women working in the United States were employed in domestic and personal service occupations.(FN31) Women from other groups, such as Italians and Jews, were noticeably less likely to engage in domestic work; although some attribute this to cultural differences in attitudes regarding contact of women with non-family men, others assert that the sex-ratio of immigration was an important determining factor, with male-dominated or family migration precluding the need for such alternative living arrangements.(FN32) In addition to domestic service, other occupations also required residence apart from family. Certain higher-status jobs, such as teaching and nursing, often removed women from their family homes; that they were predominantly the domain of native-born or Yankee daughters suggests that more than the availability of parents shaped women’s residence patterns.
DATA AND MEASUREMENT
THE SAMPLE
Data are drawn from the 1910 Census Public Use Sample (PUS), a nationally representative 1-in-250 sample of individuals recorded in household units.(FN33) Its construction allows us to obtain information about both individuals and families, as well as the households in which these families lived. We focus in this analysis on never-married individuals who were age 15 and older in 1910. Fifteen is selected as the start of the period of young adulthood because school attendance was no longer mandatory and working papers not required.(FN34) The sample is also limited to those living in urban areas (populations greater than 2,500). Employment and schooling opportunities were very different in urban and rural locations, and cities offered greater variety of both jobs and living situations.
Because the number from the total PUS is quite large, a subsample of the population was drawn consisting of a random 25% sample of the native-White of native-parentage population, combined with a 100% sample of Blacks and first- and second-generation ethnic groups. Six groups were then selected from this total: native-Whites of native parentage, Blacks, Irish, Germans, Italians, and Jews. The total sample used for this analysis consists of 19,505 never-married young adults age 15 years and older.
MEASURES
The dependent variable, whether an unmarried individual resides with at least one parent, is dichotomous. Coresidence is calculated by examining the relationship codes used in 1910, which categoriezed individuals by their relationship to the household head. All those classified as “child” are considered coresident; they could reside with one or both parents.(FN35)
The key independent variables in this analysis include sex, age, immigrant status, ethnic group membership, and pursuits such as school attendance or gainful employment. Table 1 lists the terms and definitions that appear in the models, as well as the means and standard deviations for the independent variables. Information on family characteristics is only available for those who live with parents; we cannot, therefore, test to determine whether family attributes–such as the number of siblings in the home, whether they were male or female, birth order of the respondent, or the socioeconomic status of the household head–influenced home-leaving for individuals not residing in their families of origin.
Sex is a dichotomous variable, with males serving as the reference category. Due to both social norms and their greater likelihood of arriving with family members, women should be more likely to reside with parents than their male counterparts. Age is included in the model as a continuous variable, beginning at age 15. Graphing the likelihood of coresidence by age indicates a curvilinear pattern; as a result, a variable squaring age has also been included, with the quadratic term divided by 25 to make the coefficient more evident. The likelihood of living with parents is expected to decline with age, as older unmarried adults have greater resources to leave home.
Current pursuits of young adults are designated by dichotomous variables for work or school. If an individual both attended school and was gainfully employed, work status is given precedence.(FN36) Occupational codes are used to determine if the young adult was gainfully employed. School enrollment is captured by the variable designating whether individuals attended school in the year prior to the census.(FN37) Gainful employment is expected to decrease the likelihood of living with parents, whereas school enrollment should increase the chance of coresidence.(FN38)
Designations on respondents’ and parents’ place of birth and mother tongue are used to determine nativity and ethnic group. The foreign born are considered first-generation Americans. Foreign-born individuals have been further divided into child immigrants, those arriving in the United States prior to turning 15, and adult immigrants. American-born individuals whose parents are foreign born are termed second-generation. White ethnics disappear statistically by the third generation, merging with native-born Whites of native-born parents, the reference group heretofore termed NWNP or the native-stock. The first generation should be least likely to coreside, whereas the literature suggests the highest degree of family living among the second generation. Young adults arriving as children are not expected to differ much from the second generation, whereas those arriving as adults should more closely replicate the foreign born.
Ethnic groups are identified in various ways. The Irish and Germans serve as two representatives of the “old” immigrant groups, those who began arriving in the United States in the 1840s; place of birth alone is used for the Irish, whereas both place of birth and mother tongue have been used to characterize Germans. This study also looks at two of the “new” immigrant groups, Italians and Jews, who began arriving in unprecedented numbers in the 1880s. Mother tongue and place of birth serve to distinguish Italians. Because religion has never been asked on a U.S. census, language is the only way of identifying Jews; those who designate their or their parents’ mother tongue as Yiddish are classified as Jewish.(FN39) The last two groups in this analysis are longer-term residents of the United States. Blacks include those classified by race as either Black or mulatto. Native-Whites of native-parentage (NWNP) are captured via the place of birth and parents’ place of birth variable. Among Whites where mother and father differ in ethnic origin, respondents have been assigned to the father’s ethnic group.(FN40) The literature suggests that Blacks should be less likely than Whites to reside with parents, but there are also differences among White ethnics. Because of their imbalanced sex ratio and recency of immigration, Italian men should exhibit lower levels of coresidence than both NWNP and other Whites; among women, the Irish and German are expected to demonstrate greater likelihood of autonomous living.
FINDINGS
In 1910 almost one-quarter of all adults age 15 or older, 22.4%, lived with their parents. The vast majority of coresident children, 89.3%, were never married. The factors affecting coresidence with their parents of currently and formerly married children are different in important ways from those shaping living patterns of never-married young adults, and are therefore not examined here.
Among those who never married, gender and ethnicity strongly differentiated the living arrangements of urban adults. As expected, women were more likely than men to live with their parents, 60.9% compared with 54.2% of all never-married men (see Table 2). Living with parents declines sharply with age. Whereas more than three-quarters of the youngest single men and women lived with parents, only about one-fifth of unmarried adults age 35 or older remain in the parental home. Between ages 15 and 17, unmarried boys are more likely than single girls to live with parents, perhaps because girls left for service jobs. At all subsequent ages, however, unmarried women are more likely than their male counterparts to be living with parents.
With the exception of Italians, and to a lesser extent Jews, the residence patterns for single men and women appear fairly similar. The majority of unmarried native-stock, German, and Jewish children continue to reside with their parents, whereas less than half of all never-married Black and Irish sons and daughters remain in the parental home. Of all ethnic groups, only Irish women are less likely than their male counterparts to live with parents. Italians present a case of contrasts: Whereas almost three-fourths of Italian single women remained in the parental home, only a third of Italian men lived with parents.
Among those living with parents, there are also interesting ethnic and gender variations in the propensity to work and attend school (data not shown). Although there is relatively little difference in the proportions of coresident sons who are gainfully employed, variation is greater among coresident women, with Jewish daughters being the most likely to be gainfully employed and native-White daughters of native parentage least likely. Similarly, there is greater variance in school attendance between ethnic women than among the men. Among sons, native-Whites and Jews are most likely to be enrolled in school through their early 20s, whereas Italians are least likely. Native-White daughters demonstrate the greatest propensity to be in school through their early 20s, followed by Blacks; Italian daughters are least likely to be enrolled in schools.
LOGISTIC REGRESSION ON RESIDENCE WITH PARENTS
Logistic regressions predicting the likelihood of living with parents are estimated, and results are presented below. Table 3 presents the parameter estimates of logit regression on the probability of living with parents for all never-married men and women age 15 years and older, given their individual and ethnic characteristics. Three sequential models are presented. The first examines the likelihood of living with parents, controlling for age and current activities. The second model adds nativity. The final model controls for ethnicity and nativity.
Results from a pooled model including both males and females (not shown) indicate that single women are significantly more likely than men to live with their parents. Nonetheless, this effect weakens considerably after controlling for young adults’ activities, and is no longer significant (and is in fact negative) if nativity and ethnicity are held constant. The results presented in Table 3 demonstrate that the factors affecting living with parents varied in important ways for women and men on every level: age, activities, and nativity. For example, age has a linear effect on women’s odds of living with parents, whereas the effect of age on coresidence is curvilinear for men and exerts a larger effect. Beyond the late teens, age has little influence on the probability that daughters coreside with parents. This result suggests that other factors, such as entry into the labor force or ability to provide domestic services in the home, played a more important role in determining unmarried women’s living arrangements.
In fact, gainful employment does exert a large and significant negative effect on unmarried women’s odds of living with parents. The same is not true for men once nativity is held constant. This finding indicates the need for further examination of the meaning of work for women. Whereas the vast majority of men worked whether or not they lived with their families, participation in the paid labor force was less necessary for women who lived in the parental home, as they could contribute in other ways (caring for younger children, housework, or assisting with boarders). Among employed women, the odds of living with parents are only 47% that of non-working women (Table 3, Model C). The effect of school enrollment on the likelihood of living with parents also varies for men and women. Even though greater proportions of women are enrolled in school, school attendance does not significantly increase women’s odds of living with parents; among men it does. Young men enrolled in school are 59% more likely to be living with parents than are men not in school. Living at home appears to be a strategy used by men to obtain or prolong their education. For women, on the other hand, the balance appears to be between remaining in the home or being gainfully employed.
Nativity and ethnicity also have different effects on the likelihood of coresidence for men and women. Women immigrating at young ages do not differ from the native-stock in terms of their likelihood of living with parents. Unmarried men who arrived in the United States as children are significantly more likely to be living with their parents than are native-White men of native parentage (Model B). Both male and female second-generation ethnics are more than twice as likely to live with parents as third- or higher-generation Whites.
The odds ratio of living with parents among ethnics and Blacks relative to those who are native-White of native parentage, including controls for age and current activities, are portrayed in Figures 1 through 3; these figures are calculated based on Model C of Table 3. The results indicate important gender differences in living arrangements, varying by age at immigration and ethnic group. Falling above and below the X-axis, set at 1, represents a greater or lesser likelihood of living with parents than native-Whites of native parentage. Following the graphs from adult immigrants (Figure 1) to child immigrants (Figure 2) to the second generation (Figure 3) demonstrates how the propensity to live with parents increases with duration in the United States. First, those immigrating as working-age adults are significantly less likely to be living with parents than are NWNP men and women. The negative effect is somewhat smaller for Jewish and Italian women relative to the men, supporting the assertion that their moves were familial, and slightly larger among Irish and German women compared to their male counterparts, as was expected.
There is greater ethnic variation among those arriving in the United States as children. For the Irish in this category, the odds of coresiding are still well below those of NWNPs. German women also appear less likely to live with parents than the native-stock; German men immigrating at young ages, on the other hand, are slightly more likely to be living with parents. Marked gender differences are exhibited by the Italians, with male child-immigrants exhibiting odds of living with parents that are considerably below native-stock men, whereas Italian women do not appear to differ appreciably from NWNP women. By far the most distinctive pattern is demonstrated by Jewish single adults, with child immigrants exhibiting odds of living with parents that are considerably greater than for the Yankee stock, regardless of gender.
Second-generation men and women are universally more likely to be living with parents than their third-generation counterparts, as Figure 3 demonstrates. Again, second-generation Irish and German women are less likely than the men to be living with their parents. In contrast, Italian and Jewish women are more likely than their male counterparts to live with parents. Second generation Jews are also far more likely than both NWNPs and other immigrant groups to be living with their parents. For never-married Blacks, the likelihood of living with parents is far below that of NWNPs, controlling for age and pursuits. The odds among Black women of living with parents are only 34% that of the native stock population, whereas Black men are 37% as likely to be living with their families as native-White men of native parentage.
Are the differences in male and female residence patterns, so evident in Figures 1 through 3, significant? Table 4 presents results of tests for significant differences across gender, obtained by pooling men and women and interacting gender and each variable. Results indicate that age, work, and nativity all exert significantly different effects for men and women. Coresidence rates decrease by age at a significantly different pace for men and women. Perhaps most important, never-married women who work are significantly less likely to be living at home than working men: Although the odds among working men of living with parents are 52% of those of non-working men, women who are gainfully employed are only 42% as likely to be living with parents as those who do not work.
Gender differences are also apparent upon looking at the impact of immigrant status on coresidence. Women who immigrated as adults are significantly more likely than their male counterparts to be residing with parents (though the difference is small). On the other hand, men immigrating as children and those who are second-generation are significantly more likely than their female counterparts to live with parents, and the difference is noticeably larger. Among child immigrants, the odds of living with parents for men are 1.04, whereas for women the odds are .81. Although both male and female second-generation Americans are significantly more likely than their third-generation counterparts to live with parents, the odds for second-generation men are 2.55, whereas for women they are 2.10.
ETHNIC AND RACIAL DIFFERENCES IN CORESIDENCE
Although the above results indicate important gender differences in the factors affecting young adults’ living arrangements, they do not inform us of ethnic-specific differences in gender patterns of coresidence. To examine in greater detail how gender distinguished the living arrangements of young adults from various ethnic groups, interactions by ethnic group are presented (Table 5). Irish and German women are significantly less likely to be living at home than their male counterparts after controlling for activities and nativity. The odds of living at home for Irish women are only 76% that of Irish men, whereas the odds of living at home among German women are 86% that of German men. Women from the “new” immigrant groups, on the other hand, are more likely to be living at home than are Italian or Jewish men. This is particularly evident among Italians. Black and NWNP women are no more or less likely to live at home than are men from these groups.
Although living arrangements vary widely by gender and ethnicity even upon controlling for work and school pursuits, the continued salience of gainful employment as a factor differentiating men’s and women’s living patterns highlights the importance of labor force participation as a means of freeing some women from the parental home. Work in the paid labor force played a significant role in shaping the residence patterns of Irish, German, NWNP, and Black women. Gender discrepancies are strongest for the Irish and Germans. German women who work are 11% less likely to be living at home than German men, relative to their non-working counterparts. The gender gap in living arrangements is even larger for Irish men and women. Whereas Irish men’s living arrangements are not significantly different whether they work or not, working Irish women are only 69% as likely to live with parents as non-working Irish women, and 29% less likely than working men to live with their parents; this gender difference is highly significant (p [less than or equal] .01). In part because work increases the odds that Black men will live at home but has a negative effect for Black women, Black women are 16% less likely than Black men to be living at home if working, although this is only weakly significant. The odds of living at home if working are greater, but only weakly significant for Italian women, whereas work does not exert significantly different effects on the living arrangements of Jewish and native-White men and women.
Age at immigration also appears to exert different effects by gender and ethnicity, at least for the Italians and Irish. As expected, Italian women who immigrated as adults were significantly more likely than male adult immigrants to be living with their parents; the effect is in the opposite direction for Irish adult immigrants, as well as for second-generation women. The prevalence of domestic service occupations among Irish women as well as the disproportionately female immigration from Ireland no doubt accounts at least in part for this. Whether these factors also explain the significant difference between second-generation males and females seems less probable, as native-born women were less likely to work in service positions. Perhaps it reflects Irish daughters’ movement into white-collar positions such as teaching or nursing, which were usually limited to unmarried women and often required moving away from parents.
These results indicate that significant gender differences in the effect of pursuits and nativity upon living with parents are not uniform across all ethnic groups. Representatives of the “old” immigrant groups present the most evidence of gender differences in factors affecting living arrangements. Such findings raise questions regarding the nature of household allocations of resources, gender differences in children’s participation in the family economy, as well as women’s acceptance of traditional gender roles.
DISCUSSION AND CONCLUSION
Life course plays an obvious role in shaping the residential experience of unmarried young adults at the turn of the century. Those most likely to be living in their parents’ home at the turn of the century were in many ways more “dependent”: younger, still in school, less likely to be working. However, life course alone cannot account for variations across gender and ethnic group. The evidence suggests that, all things being equal, among certain ethnic groups encouragement or tolerance of remaining in the parental home was stronger for men than for women.
GENDER FINDINGS
Although bivariate results indicate that women were more likely than men to reside with parents, this difference disappears after controlling for gainful employment and ethnicity. A closer examination of who remained in the home reveals that men who were better able to contribute financially to the family economy were more likely than their female counterparts to be living at home. Those immigrating as children and born in the United States to foreign-born parents had greater earning power than their foreign-born counterparts. The youngest men residing with parents also benefitted more than did women through extended school attendance. Perhaps families encouraged sons to remain in the home to a greater extent than they did daughters. This would seem to be corroborated by the wealth of evidence that families provided sons with different resources within the home than they did daughters, whether in the from of room and board, cooking and cleaning services, more spending money than their sisters received, or looser supervision.(FN41)
An alternative possibility is that women who worked in the paid labor force tired of their often double contributions in both productive and domestic labor and chose to leave the familial nest. Work outside the home buys freedom from parents to a significantly greater degree for women than for men, highlighting the importance of work for women’s autonomy. Societal norms that determined the home as women’s proper location evidently operated by discouraging daughters from working and becoming financially independent. Another plausible explanation is that the structure of job opportunities required women to leave the parental home more often. Earned income appears to have allowed single women at the turn of the century to depart from less-than-optimal living situations with family, just as employment enables married women to leave undesirable marriages in the late twentieth century.
ETHNIC FINDINGS
Distinctive ethnic variations in residence patterns suggest that the status of women varied widely across groups. The most obvious divide is between the “old” and “new” immigrant groups. For example, although Italian and Jewish women are significantly more likely to live at home than their male counterparts, attending school or working does not increase or diminish their likelihood of coresidence. On one hand, this suggests that greater gender equality marked the “new” immigrants. A more likely explanation, however, is that the “new” immigrants relied on coresidence as a means of adapting to life in a new country, even into the second generation, and that coresidence benefitted children as well as parents. Irish and German families appear to have had less success in monitoring their daughters, particularly if they were gainfully employed, relative to their ability to retain sons in the parental home. This gender difference may reflect a cultural pattern of greater authonomy for women from these groups. Alternatively, that men from the “old” immigrant groups stayed in the parental home well into their 20s and 30s suggests they received inducements to remain, rather than marry and establish independent households.
Gender distinctions for Blacks are more difficult to interpret. Some have suggested that the employment of Black daughters enabled Black sons to remain in school. Our findings indicate that school enrollment does not exert significantly different effects on the living arrangements of never-married Black males and females. Nonetheless, the wide gender discrepancy does indicate that Black males were more likely than Black daughters to attend school, perhaps because of a dearth of employment opportunities. In conjunction with Black daughters’ greater propensity to be gainfully employed, it also suggests that Black daughters’ education received secondary priority to that of Black sons, perhaps because daughters could obtain jobs and contribute more to the family than their brothers. The causal direction may not be from school to living at home, but rather from job opportunities through residing at home to school enrollment.
RAMIFICATIONS FOR GROUP STRATIFICATION
Variations in the living arrangements of young adults have implications not only for the composition of ethnic households, but for group social mobility. Certain groups are more likely to be characterized by the inclusion of working-age young adults, enabling them to raise their standard of living, buy property, consume more leisure goods, or educate children. Black families appear to be disadvantaged in that working-age single children, both sons and daughters, were not likely to be living at home. Irish and German families benefitted from the presence of sons, who married relatively late and could therefore contribute for an extended period of time. Italians were less likely to benefit from the presence of sons, and although daughters remained with parents, they married at young ages, thereby curtailing their wage contributions. Unlike all other groups, Jewish families stood to gain from the coresidence and hence contributions of both sons and daughters; although this may have been due to discrimination in the employment and housing market, the arrangement appears to have benefitted both children, who gained through schooling, and parents, who received coresident children’s wages.
In light of these findings, various portrayals of ethnic group cultural values also need to be reexamined. Third- and higher-generation White women and men appear to be uniformly affected by activities, if not age, suggesting an even-handed allocation of household resources between sons and daughters. Jewish families also appear more egalitarian than the literature credits them, with school attendance and employment exerting similar effects for men and women. Other groups demonstrate much greater gender discrepancies than expected. Blacks and the Irish are characterized by significant gender differences. More attention should be paid to examining household and gender dynamics for Germans, as our findings indicate that sons were advantaged from an early age even among the foreign-born German. The results point to the need to further examine gender dynamics and how they shape the constellation of activities performed by all family members, the allocation of and contributions to household resources, and the process of group stratification.
Added material
Sharon Sassler is an assistant professor in the edepartment of sociology at Wellesley College. She is the author of “Trade-Offs in the Family: Sibling Effects on Daughters’ Activities in 1910,” Demography 32 (November 1995) 557-75, and coauthor (with Michael J. White) of “Ethnicity, Gender, and Social Mobility in 1910,” Social Science History (forthcoming). Professor Sassler’s current research examines both historical and contemporary factors contributing to changes in union formation and neighborhood status.
Table 1 Definitions, Means, and Standard Deviations for Variables Used in Analysis
Variable Description M (SD)(FNa) Female Female = 1, Male = 0 0.46 Age Current age 26.11 (10.92) Age-squared Current age, squared, divided by 25 32.03 (31.79) Current activities: Work Gainfully employed (occupational code GE 10) 0.79 Attend school Enrolled in school at any time since Sept. 1, 1909 0.09 Nativity: First generation Place of birth outside of U.S. or territories 0.22 Second generation Born in the U.S., at least one foreign-born parent 0.44 Third generation Native-born to native-born parents 0.34 Age at immigration: Child immigrant Foreign-born, arrived prior to age 15 0.09 Adult immigrant Foreign-born, arrived at or after age 15 0.13 Ethnicity: Irish Place of birth Ireland 0.25 German Place of birth Germany, mother tongue German 0.26 Italian Place of birth Italy, mother tongue Italian 0.07 Jewish Mother tongue Yiddish 0.08 Black Race Black or mulatto 0.14 NWNP Place of birth/parents' place of birth U.S., race White 0.34
FOOTNOTE
a. Continuous variables.
Table 2 Percentage Coresiding Among Never-married Men and Women, by Age and Ethnicity
Independent Variables Men (% ) Women (% ) Age 15-17 86.8 84.4 18-21 67.7 68.1 22-25 48.4 59.2 25-29 40.1 52.1 30-34 34.8 44.4 35+ 18.3 21.0 Ethnicity: Irish 50.0 47.5 German 61.2 66.7 Italian 33.2 72.0 Jewish 61.4 74.8 Black 35.4 41.1 NWNP 58.3 64.9 Total 54.2 60.9 N 10,621 8,884
Table 3 Logistic Regression of Residence with Parents in 1910
Women Men Variable Model A Model B Model C Model A Model B Model C Constant 3.486(FN*) 3.380(FN*) 3.332(FN*) 4.431(FN*) 4.260(FN*) 4.675(FN*) Age -0.100(FN*) -0.117(FN*) -0.090(FN*) -0.205(FN*) -0.215(FN*) -0.218(FN*) Age-squared 0.002 0.005 -0.003 0.039(FN*) 0.039(FN*) 0.041(FN*) Activities Work -0.917(FN*) -0.819(FN*) -0.750(FN*) -0.304(FN*) -0.144 -0.144 Attend school 0.065 0.110 0.156 0.572(FN*) 0.571(FN*) 0.466(FN*) Nativity Native-born (Ref) Child immigrant 0.076 -- 0.277(FN*) -- Adult immigrant -1.497(FN*) -- -1.891(FN*) -- Second generation 0.971(FN*) -- 1.213(FN*) -- Ethnicity and nativity NWNP (Ref) Irish, child immigrant -1.523(FN*) -0.549(FN*) Irish, adult immigrant -4.719(FN*) -3.030(FN*) Irish, second-generation 0.336(FN*) 0.787(FN*) German, child immigrant -0.298(FN**) 0.084 German, adult immigrant -2.954(FN*) -2.448(FN*) German, second-generation 0.599(FN*) 0.871(FN*) Italian, child immigrant -0.013 -0.631(FN*) Italian, adult immigrant -1.099(FN*) -2.700(FN*) Italian, second-generation 1.036(FN*) 0.732(FN*) Jewish, child immigrant 0.570(FN*) 0.457(FN*) Jewish, adult immigrant -0.958(FN*) -1.401(FN*) Jewish, second-generation 1.706(FN*) 1.453(FN*) Black -1.089(FN*) -0.983(FN*) -2LL 10,018 9,225 8,846 12,227 10,503 10,268
FOOTNOTES
* p < .01.
** p < .10.
Table 4 Examining Significant Gender Differences in Factors Affecting Living Arrangements
Full Model Model Model Model Model Variable Model A B C D E Constant 4.345(FN*) 4.910(FN*) 4.205(FN*) 4.352(FN*) 4.338(FN*) 4.318(FN*) Female -0.033 -1.250(FN*) 0.142(FN**) 0.050(FN***) -0.019 0.025 Age -0.164(FN*) -0.201(FN*) -0.164(FN*) -0.163(FN*) -0.164(FN*) -0.164(FN*) Age-squared 0.022(FN*) 0.035(FN*) 0.022(FN*) 0.022(FN*) 0.022(FN*) 0.022(FN*) Activities Work -0.811(FN*) -0.794(FN*) -0.655(FN*) -0.816(FN*) -0.809(FN*) -0.809(FN*) Attend school 0.044 0.086 0.091 0.043 0.044 0.040 Nativity Native-born (Ref) Child immigrant -0.060 -0.062 -0.058 -0.059 0.039 -0.060 Adult immigrant -1.998(FN*) -1.996(FN*) -1.998(FN*) -2.148(FN*) -1.997(FN*) -1.993(FN*) Second generation 0.833(FN*) 0.830 0.836(FN*) 0.834(FN*) 0.833(FN*) 0.937(FN*) Interaction terms Female × Age 0.077(FN*) -- -- -- -- Age-squared -0.026(FN*) -- -- -- -- Work -- -0.209(FN*) -- -- -- Adult immigrant -- -- -0.419(FN*) -- -- Child immigrant -- -- -- -0.225(FN*) -- Second generation -- -- -- -- -0.220(FN*) -2LL 33,351 33,316 33,345 33,343 33,347 33,339
FOOTNOTES
* p < .01.
** p < .05.
*** p < .10.
Table 5 Logistic Regression on Residence with Parents in 1910, by Ethnicity
Variables Irish German Italian Jewish(FNa) Black NWNP
Constant 2.700(FN*) 3.812(FN*) 3.524(FN*) 2.605(FN*) 4.134(FN*) 4.507(FN*)
Female -0.274(FN*) -0.147(FN**) 1.041(FN*) 0.340(FN**) -0.073 0.030
Age -0.121(FN*) -0.125(FN*) -0.208(FN**) -0.051 -0.239(FN*) -0.161(FN*)
Age-squared 0.005 0.009 0.033 -0.027 0.046(FN*) 0.022(FN*)
Activities
Work -0.531(FN*) -0.783(FN*) 0.257 -0.026 -0.269(FN***) -0.954(FN*)
Attend school -0.238 0.207 0.351 0.744(FN**) 0.596(FN*) -0.071
Nativity
Adult immigrant -2.652(FN*) -2.532(FN*) -1.860(FN*) -1.803(FN*) -- --
Second generation 1.653(FN*) 0.844(FN*) 1.340(FN*) 1.135(FN*) -- --
Interaction terms (odds)
Fernale age & age-squared (FN**) (FN***) (FN*) NS (FN**) (FN*)
Male 0.86 0.86 0.80 0.78 0.83
Female 0.18 0.16 0.00 0.10 0.11
Female work (FN*) (FN**) (FN***) NS (FN***) (FN***)
Male 0.97 0.69 0.79 1.06 0.51
Female 0.69 0.58 2.66 0.91 0.46
Female school NS NS (FN***) NS NS NS
Male 2.43
Female 1.96
Female adult immigrant (FN**) NS (FN*) NS -- --
Male 0.11 0.12
Female 0.02 0.70
Female second generation (FN*) NS (FN***) NS -- --
Male 3.41 4.49
Female 2.77 7.67
N 4,780 5,075 1,434 1,477 2,641 4,099
FOOTNOTES
* p < .01.
** p < .05.
*** p < .10.
a. For Jews, “work” includes only those working, whereas “attend school” incorporates both those whose sole activity is attending school, and those who combine school and work.
Figure 1. Odds of Living With Parents Among Adult Immigrants (Relative to NWNP)
Figure 2. Odds of Living With Parents Among Those Immigrating as Children (Relative to NWNP)
Figure 3. Odds of Living With Parents Among Second Generation Ethnics and Blacks (Relative to NWNP)
FOOTNOTES
1. A congressional study published in 1910 noted that among families containing both young children and either working sons or daughters, more than a third of the family income was contributed by offspring; daughters 16 years of age and older contributed 39.7% compared to 36.5% contributed by sons. U.S. Bureau of Labor, “Wage-Earning Women in Stores and Factories,” in Report on Conditions of Woman and Child Wage-Earners in the United States (Washington, DC: U.S. Bureau of Commerce, 1910), vol. 5, 652-3. See also Tamara Hareven, Family Time and Industrial Time (Cambridge, MA: Cambridge Press, 1982); Patrick M. Horan and Peggy G. Hargis, “Children’s Work and Schooling in the Late Nineteenth Century Family Economy,” American Sociological Review 56 (1991): 583-96; Robert V. Robinson, “Family Economic Strategies in Nineteenth- and Early Twentieth-Century Indianapolis,” Journal of Family History 18 (1993): 191-211.
2. In 1911, the United States Immigration Commission found that only 64% of native-born American husbands living in New York City supported their household without the help of other family members. Among foreign-born fathers, only 45% of Irish and 42% of German husbands were the sole support of their families, as were 38% of southern Italian husbands. Jewish and Black fathers were half as likely to be their family’s sole provider, with only 20% of Jewish families and 19% of Black families relying on income provided entirely by the father. U.S. Immigration Commission, Immigrants in Cities (Washington, DC: U.S. Bureau of Commerce, 1911), 232. See also Claudia Goldin, “Family Strategies and the Family Economy in the Late Nineteenth Century: The Role of Secondary Workers,” in Philadelphia: Work, Space, Family, and Group Experience in the 19th Century, ed. Theodore Hershberg (New York: Oxford University Press, 1981), 277-310; John Modell, Frank F. Furstenberg, Jr., and Theodore Hershberg, “Social Change and Transitions to Adulthood in Historical Perspective.” Ibid., 311-43.
3. Although there is no data on variations in the domestic contributions of coresident and non-coresident daughters, there is evidence that nonresident daughters contributed far less in wages than those living with parents. Among women living away from home in 1910, more than two-thirds retained all of their earnings. For those living with parents, on the other hand, 86% handed over their entire earnings to their families. Claudia Goldin, Understanding the Gender Gap: An Economic History of American Women (New York: Oxford Press, 1990), 53.
4. For a discussion of changes in household structure and family composition, particularly among young adults, see Frances Kobrin Goldscheider and Celine LeBourdais, “The Decline in Age at Leaving Home, 1920-1970,” Sociology and Social Research 70 (1986), 143-5; Frances E. Kobrin, “The Fall in Household Size and the Rise of the Primary Individual in the United States,” Demography 13 (1973), 127-38.
5. For a study of “wayward women,” one of the terms applied by social commentators to unmarried females living apart from family, see Joanne J. Meyerowitz, Women Adrift: Independent Wage Earners in Chicago, 1880-1930 (Chicago: University of Chicago Press, 1988). The introduction to the 1910 Bureau of Labor report also contains a lengthy description of the process used by investigators to determine where young women lived; see U.S. Bureau of Labor, “Wage Earning Women in Stores and Factories,” 9-14. In 1900, more than a third of single women workers in cities lived apart from their parents; see Goldin, Understanding the Gender Gap, 93; see also Lynn Y. Weiner, From Working Girl to Working Mother: The Female Labor Force in the United States, 1820-1920 (Chapel Hill: The University of North Carolina Press, 1985).
6. For studies of shifts in the female labor from reliance on single working women to married working mothers, see Goldin, Understanding the Gender Gap; see also Valerie Kincade Oppenheimer, The Female Labor Force Participation in the United States: Demographic and Economic Factors Governing Its Growth and Changing Composition (Berkeley: Institute of International Studies, 1970); Lynn Y. Weiner, From Working Girl to Working Mother. See also Robert V. Robinson, “Family Economic Strategies in Nineteenth-and Early Twentieth-Century Indianapolis,” Journal of Family History 20 (1995), 1-22.
7. This research generally deals only peripherally with women’s shifting roles, concentrating instead on men. See John Bodner, Roger Simon, and Michael P. Weber, Lives of Their Own: Blacks, Italians, and Poles in Pittsburgh, 1900-1960 (Chicago: University of Illinois Press, 1982); Hareven, 1982; Thomas Kessner, The Golden Door: Italian and Jewish Immigrant Mobility in New York City, 1880-1915 (New York: Oxford University Press, 1977); Stanley Lieberson, A Piece of the Pie: Blacks and White Immigrants Since 1880 (Berkeley: University of California Press, 1980); Stephan Thernstrom, The Other Bostonians: Poverty and Press in an American Metropolis, 1880-1979 (Cambridge, MA: Harvard University Press, 1973); also Stephan Thernstrom, Poverty and Progress: Social Mobility in a Nineteenth Century City (Cambridge, MA: Harvard University Press, 1964).
8. For an example of how women supplemented the family income by taking in boarders or lodgers, see John Modell and Tamara K. Hareven, “Urbanization and the Malleable Household: An Examination of Boarding and Lodging in American Families,” Journal of Marriage and the Family 35 (1973), 457-79; also Elizabeth H. Pleck, “A Mother’s Wages: Income Earning Among Married Italian and Black Women, 1896-1911,” in The American Family in Social-Historical Perspective, ed. Michael Gordon (New York: St. Martin’s, 1978), 490-510. For an examination of the trade-offs between mothers’ and daughters’ income-supplementing activities, see Robinson, “Family Economic Strategies in Nineteenth- and Early Twentieth-Century Indianapolis.” A more general overview of the family economy framework is provided by Phyllis Moen and Elaine Wethington, “The Concept of Family Adaptive Strategies,” Annual Review of Sociology 18 (1992), 233-51.
9. In his classic work on women and the family in America, Carl Degler asserts that there is a basic tension between the notion of equality for women and men and the institution of the family. “The historic family has depended for its existence and character on women’s subordination,” he writes in his preface. “It means simply that the family’s existence assumes that a woman will subordinate her individual interests to those of others–the members of her family.” Carl Degler, At Odds: Women and the Family in America From the Revolution to the Present (New York: Oxford University Press, 1980), vi-vii. Several studies of working daughters argue that even those employed outside the home viewed their involvement in the paid labor force as temporary, and many looked forward to the time when they could leave the factories and offices; the workplace therefore served as a location where young women reinforced the primacy of marriage and family. Leslie Woodcock Tentler, Wage-Earning Women: Industrial Work and Family Life in the United States, 1900-1930 (New York: Oxford University Press, 1979), 28, 56, 60-2, 69, 71-3; also Elizabeth Ewen, Immigrant Women in the Land of Dollars: Life and Culture on the Lower East Side, 1890-1925 (New York: Monthly Review Press, 1985), 231, 233-4, 250.
10. A 1910 congressional study noted family reliance on daughters’ wages in the following paragraphs:
Although the average earnings of the males in this age group [children who were 16 years of age and over] were, for each nativity and race group without exception, greater than those of the females, yet the per cent earnings contributed to the family fund by the females exceeds that contributed by the males in every case but one…. The figures showing per cent of individual earnings contributed to family income have brought out clearly that females in the age group 16 years of age and older are much more generous contributors to the support of the family than are males.
U.S. Bureau of Labor, “Wage-Earning Women,” 368. For other studies of daughters’ contributions, see Claudia Goldin, “Family Strategies and the Family Economy in the Late Nineteenth Century: The Role of Secondary Workers”; Susan A. Glenn, Daughters of the Shtetl: Life and Labor in the Immigrant Generation (Ithaca, NY: Cornell University Press, 1990); Louise Lamphere, From Working Daughters to Working Mothers: Immigrant Women in a New England Industrial Community (Ithaca, NY: Cornell University Press, 1987); Louise C. Odencrantz, Italian Women in Industry: A Study of Conditions in New York City (New York: Russell Sage Foundation, 1919); Robert V. Robinson, “Family Economic Strategies in Nineteenth- and Early Twentieth-Century Indianapolis.”
11. U.S. Bureau of Labor, “Wage Earning Women in Stores and Factories,” 9.
12. Ibid., 10. For a detailed discussion of the paternalistic climate surrounding women’s living arrangements, see also Meyerowitz, Women Adrift, 40-68.
13. Meyerowitz, Women Adrift, 33, 43-68.
14. For a thorough study of Italian family strategies, see Virginia Yans-McLaughlin, Family and Community: Italian Immigrants in Buffalo, 1880-1930 (Ithaca, NY: Cornell University Press, 1973); a comparison of Blacks and Italians is available in Elizabeth H. Pleck, “A Mother’s Wages,” whereas Italians and Jews are contrasted in Thomas Kessner’s The Golden Door; the effect of family and individual strategies on Irish women are examined in Hasia R. Diner, Erin’s Daughters in America: Irish Immigrant Women in the Nineteenth Century (Baltimore, MD: Johns Hopkins University Press, 1983).
15. The argument that Black parents placed a higher premium on educating their children than did Whites is offered by Claudia Goldin, “Family Strategies and the Family Economy in the Late Nineteenth Century,” 298-300; see also Lieberson, A Piece of the Pie, 123-250. Others assert that Jews placed a strong emphasis on educating children, particularly sons; see Kessner, “The Golden Door,” 88; also Olivier Zunz, The Changing Face of Inequality: Urbanization, Industrial Development, and Immigrants in Detroit, 1880-1920 (Chicago: University of Chicago Press, 1982), 234. Italian treatment of schooling for children is discussed in Kessner, The Golden Door, 84-5.
16. See Linda Gordon and Sara McLanahan, “Single Parenthood in 1900,” Journal of Family History 16 (1991), 97-116; Antonio McDaniel, “Historical Differences in Living Arrangements of Children,” Journal of Family History 19 (1994), 57-77; Andrew T. Miller, S. Philip Morgan, and Antonio McDaniel, “Under the Same Roof: Family and Household Structure,” in After Ellis Island: Newcomers and Natives in the 1910 Census, ed. Susan Watkins (New York: Russell Sage, 1994), 75-103; S. Philip Morgan, Antonio McDaniel, Andrew T. Miller, and Samuel H. Preston, “Racial Differences in Household and Family Structure at the Turn of the Century,” American Journal of Sociology 98 (1993), 798-828; Steven Ruggles, “The Origins of African-American Family Structure,” American Sociological Review 59 (1994), 136-151.
17. See Diner, Erin’s Daughters in America; Kessner, The Golden Door; Meyerowitz, Women Adrift; Morgan et al., “Racial Differences in Household and Family Structure at the Turn of the Century”; Weiner, From Working Girl to Working Mother.
18. For example, the typical young man or woman living in Philadelphia in 1880 would enter the work force upon completing his or her schooling and contribute to the family income for about 7 years, barring mortality. Modell et al., “Social Change and Transitions to Adulthood in Historical Perspective,” 313. See also Goldscheider and LeBourdais, “The Decline in Age at Leaving Home,” 143.
19. The Irish and Germans were particularly noted for their patterns of late marriage and non-marriage, even among the second generation. Similarly, native-born Jewish women with foreign-born parents also married significantly later than did the native born of native parentage. See Nancy Landale and Steward Tolnay, “Generation, Ethnicity, and Marriage: Historical Patterns in the Northern United States,” Demography 30 (1993), 103-26; Sharon Sassler, “Marital Timing and Assimilation in Early 20th Century America,” Brown University PSTC Working Paper Series 93-01. Some immigrant literature, as well as letters to advice columns in the ethnic press, suggests that parents often sought to delay children’s marriage because of familial reliance on their economic contributions; for example, the novel Bread Givers depicts a father reluctant to let his daughter marry because of the importance of her wages to the family economy; he ultimately turns away several of her suitors. Anzia Yezerskia, Bread Givers (New York: Persea Press, 1925); see also Isaac Metzker, A Bintel Brief: Sixty Years of Letters from the Lower East Side to the “Jewish Daily Forward” (New York: Schocken Books, 1971).
20. See Michael Anderson, Family Structure in Nineteenth Century Lancashire (Cambridge, UK: Cambridge University Press, 1971); also Gordon and McLanahan, “Single Parenthood in 1990,” 106; Steven Ruggles, “The Transformation of American Family Structure,” American Historical Review 99 (1994), 113-5. The extent to which mortality accounts for differences in the living arrangements of Blacks and Whites is increasingly being called into question; see Morgan et al., “Racial Differences in Household and Family Structure at the Turn of the Century,” 816. Ruggles contends that among young children, those younger than 14, large racial differences in residence patterns did not usually result from parental death; Ruggles, “The Origins of African-American Family Structure,” 142-3. Whether parental mortality accounts for differential living arrangements among older children (those 15 and up) is nowhere examined.
21. Weiner cites estimates that the rural-urban migration of women exceeded that of men by about 22% between 1920 and 1930; Weiner, Working Girl, 19, 22, 147.
22. See Goldin, “Family Strategies,” 304; Miller et al., “Under the Same Roof,” 164.
23. See Arodys Robles and Susan Cotts Watkins, “Immigration and Family Separation in the U.S. at the Turn of the Century,” Journal of Family History 18 (1993), 191-211.
24. Kessner, Golden Door, 30.
25. For discussions of Italian women’s propensity to take in boarders, see Pleck, “A Mother’s Wages”; also Yans-McLaughlin, Family and Community.
26. Kessner, The Golden Door, 31; see also Arthur Goren, “Jews,” in The Harvard Encyclopedia of Ethnic Groups, ed. Stephen Thernstrom (Cambridge, MA: Harvard University Press, 1980), 581.
27. Diner, Erin’s Daughters in America, 33-4, 38; also Paula Jackson, “Women in 19th Century Irish Emigration,” International Migration Review 28 (1984), 1004-20.
28. See Jerry Jacobs and Margaret E. Greene, “Race and Ethnicity, Social Class and Schooling in 1910,” in After Ellis Island: Newcomers and Natives in the 1910 Census, ed. Susan Watkins, (New York: Russell Sage, 1994), 209-56; Sharon Sassler, “Trade-Offs in the Family: Sibling Effects on Daughters’ Activities in 1910,” Demography 32 (1995), 557-75.
29. Meyerowitz, Women Adrift, 24.
30. In a description of the methodology used by investigators to ascertain whether women wage-earners were in fact “adrift,” the issue of false addresses received considerable attention. The report elaborates:
Department store employers openly express a preference for the girls living at home. A girl endeavoring to secure a position and finding herself rejected because she had no home in the city is under strong temptation, when she finds another vacancy, to say that she is living with an aunt or a cousin or even with her parents when she is really dependent on the boarding or lodging houses. She quickly learns, too, to give an address in the home district, rather than in the boarding house sections.
U.S. Bureau of Labor, Woman and Child Wage-Earners, 13.
31. Jackson, “Women in 19th Century Irish Emigration,” 1008.
32. Yans-McLaughlin argues that conservative cultural traditions that regulated women’s familial roles constrained Italian women to a great extent, particularly in regard to women’s living arrangements; Yans-McLaughlin, Family and Community. For an opposing explanation of ethnic living arrangements, see Stephen Steinberg, The Ethnic Myth: Race, Ethnicity, and Class in America (Atheneum Press, 1981), 156-66.
33. The 1910 PUS was prepared at the Population Studies Center of the University of Pennsylvania. See Michael A. Strong, Samuel H. Preston, Ann R. Miller, Mark Hereward, Harold R. Lentzner, Jeffrey R. Seaman, and Henry Williams, User’s Guide: Public Use Sample, 1910 United States Census of Population (Philadelphia: Population Studies Center, 1989).
34. Despite restrictions on the age at which children could legally work, small proportions of 14-year-olds in the sample were gainfully employed or combined work and school. Among 14-year-olds, 9.9% of the boys worked, as did 8.5% of the girls. Another 6.4% of the boys combined work and school, compared to 3.5% of the girls.
35. Among those living with family even though parents are not the household head are siblings of the head, grandchildren, and nieces or nephews, respectively. Very few unmarried adults age 15 years and older were living in a household headed by their grandparent, only .9% of the entire sample of those never married. Multigenerational households were most prevalent for those who were Black and NWNP. Grandchildren are not included with coresident children, under the assumption that their contributions to and roles within the family differed.
36. Only 2.4% of all single adults in the sample combined work and school, with men more likely to be doing both (2.6% compared to 2.0% for women, a difference that is statistically significant at the .05 level). The combination of working and attending school influence living arrangements in a similar fashion to work alone (regression not shown), and is therefore merged with the work variable.
37. The number of years of level of schooling received by an individual in 1910 cannot be determined, nor do we know whether they attended school part- or full-year. The schooling variable includes only those whose primary pursuit was attending school.
38. Young adults living independently had greater need for resources than those living with parents, who may not have had to work because they resided at home; this is also a problem for the school relationship. Although contemporary studies of living arrangements often exclude measures of employment and schooling to avoid issues of endogeneity, our analysis gives precedence to the impact of young adults’ activities on their residence options as the primary direction of the relationship. For a similar approach, see Roger Avery, Frances Goldscheider, and Alden Speare, Jr., “Feathered Nest/Gilded Cage: Parental Income and Leaving Home in the Transition to Adulthood,” Demography 29 (1992), 375-88; Frances Goldscheider and Julie DaVanzo, “Living Arrangements and the Transition to Adulthood,” Demography 22 (1985), 545-63. Running models excluding the work and school variables yield coefficients similar in both size and direction; including the two variables (work and school) adds significantly to the model.
39. A major drawback to this approximation is that it does not include many non-Yiddish-speaking Jews residing in America.
40. Endogamy rates were extremely high for the “new” immigrants in 1910, as well as for the Irish and German representatives of the “old” immigrant waves. Deanna L. Pagnini and S. Philip Morgan, “Intermarriage and Social Distance Among U.S. Immigrants at the Turn of the Century,” American Journal of Sociology 96 (1990), 405-32.
41. Ewen, Immigrant Women in the Land of Dollars, 106-7, 120, 194-5; Kathy Peiss, Cheap Amusements: Working Women and Leisure in Turn-of-the-Century New York (Philadelphia: Temple University Press, 1986), 25, 68.
ACKNOWLEDGMENTS
This research was supported by NICHD Training Grant HD07338-05 from the Population Studies and Training Center at Brown University and by NIA Grant T32-AG00237 from the Department of Population Dynamics at Johns Hopkins University. An earlier version of this article was presented at the 1994 annual meetings of the American Sociological Association, held in Los Angeles. I thank Frances Goldscheider, Robert Schoen, and Michael White for their helpful comments on earlier drafts.
WBN: 9627502369003