Women in the Workforce (1910-1960)

American women are largely considered underprivileged in terms of both the occupations available to them and the salary they make. However this discussion excludes the effects that race and immigrant status may have on women receiving employment. The movement toward intersectionality, according to Mignon Duffy, is imperative to understanding inequality (2007, 314). I am interested in the effects of race and nativity on employment for women in the labor force. In order to understand women’s subordination in the workplace I will analyze the female employment rate for various immigrant and race subgroups on the national and regional level.

Data:
The data used for this analysis are from the census data available on the Integrated Public-Use Microdata Series (IPUMS). My data set is composed of one percent samples from the 1910 to 1960 United States censuses. The IPUMS organized variables SEX, AGE, RACE, BPL, STATEFIP, REGION and EMPSTAT are used for my analysis. I exclude the 1920 census because the EMPSTAT variable is not available for that census year. The EMPSTAT variable measures if a respondent was part of the labor force and whether or not they were employed or unemployed. For years prior to 1940 the EMPSTAT variable measures employment status on a specific day. Starting in 1940 the EMPSTAT variable measures employment status in a specific week. For all census years prior to 1960 the census data were collected by an enumerator.

For my analysis I exclude data collected in Hawaii and Alaska for all years prior to 1960. I will be using the one percent sample from each of the census years.  IPUMS has randomly selected the one percent samples used for my analysis. All analyses are weighted by PERWT, the sample weight of each individual.

Method:

I begin my analysis with the assumption that all women’s employment status were enumerated correctly. This may not have been true however due to biases of the enumerators. In the process of coding the census data, census officials were instructed to discard any defective cards that may report contradictory information. More specifically, if a woman was reported to have a “male” occupation the card would be discarded or recoded into a more feminine occupation. Therefore some women may not have been accurately reported and others not included in the final census report at all (Conk 2001, 67-69). Folbre confirms this phenomenon and extends it by blaming patriarchal norms for significantly undercounting female labor participation (Folbre 1989, 546-547). However, I proceed by narrowing the data to women over the age of 18, who are in the labor force (either unemployed and looking for work or employed). Therefore my analysis only focuses on women in the workforce and how female employment changes over time. I then categorize each woman based on whether she is native-born or foreign-born (BPL<=99 is native born and BPL>99 is foreign born) and if she is white or nonwhite (RACE==1 is white and RACE!=1 is nonwhite). This method of categorization results in four groups.

By categorizing this way I will analyze how race and immigration status influence female employment rates both nationally and regionally. In order to calculate the national employment rate, for each year I graph the proportion of the work force in each race-immigrant category that was employed. In order to calculate the regional employment rate for each year, I graph the proportion of the work force in each race-immigrant category that was employed for each region of the US. The northeast, midwest, south and west regions are defined by IPUMS as follows: The northeast – New England division, middle Atlantic division, mixed Northeast divisions; the midwest – east north central division, west north central division, mixed midwestern divisions; the south – south atlantic division, east south central division, west south central division, mixed southern divisions; the west – mountain division, pacific division and mixed western divisions. I graph the regions’ employment rates over time using a line graph. Code for analysis and visualization is available here.
Results:

Figure 1:

large nat EMP

Figure one graphs the national employment rate for each race-immigrant category from 1910 to 1960. Disparity between all race-immigrant categories is relatively small across all years. Employment rate for women is always above 85%.

Figure 2:

Magnified Nat Emp

Figure two shows a magnified version of the national employment rate for each race-immigrant category from 1910 to 1960. The employment rate is lowest in 1940 for not white foreign-born women. The employment rate is highest in 1950 for white native born women. All race-immigrant groups, except for white foreign-born women, follow the same temporal trends. White foreign-born women unlike all other race-immigrant categories have a higher employment rate in 1940 than in 1930.

Figure 3:

Regional Emp

Figure 3 graphs the regional employment rate for each race-immigrant group from 1910 to 1960. The northeast demonstrates the same trends that were visible with the national employment rates. In the south, employment rates remain fairly consistent over time. However in the south, not white foreign-born women demonstrate the most dramatic changes in employment rate. In the midwest, female employment rate appears to be divided by race. The employment rate for white foreign-born women and white native-born women in the midwest are similar to each other; similarly not white foreign-born women and not white native born women demonstrate similar trends and rates of employment over time. In the west, the employment rate changes very little for all race-immigrant groups. Unlike national or any other region’s trends, not white foreign born women have the highest employment rate for all years except 1960.

Female employment rates over the twentieth century have remained fairly static for all race immigrant categories because the social conditions that influence labor participation decision also influence the demand for labor. For women, the decision for labor force participation is generally influenced by the opportunity cost that women face to perform unpaid household labor. Over the twentieth century as more women obtained higher education, the opportunity cost to being a housewife also rose. Therefore as opportunities for higher education increased an increasing amount of women chose to join the labor force. On the other hand, the increasing opportunities for female higher education is a reflection on changing social conditions that influenced the demand for female labor participation. We see both of these changes in culture conflated in the female employment rates from 1910 to 1960; employment rate conflates labor supply and labor demand together (Cotter 2001, 430-432).

As demand for female labor grew in the twentieth century, the market for female labor was generally limited to reproductive labor. Reproductive labor consisted of work that was considered “necessary to maintain existing life and to reproduce the next generation.” Reproductive labor was not only limited to being a housewive, but could also be paid labor such as domestic service, cleaning, cooking, and child care. This form of labor could be divided into nurturant and non-nurturant occupations. Nurturant work is labor that involves interacting with another person; these occupations were mainly performed by white women and were considered spiritual work. Non-nurturant work on the other hand is labor that maintains daily life; these occupations were mainly performed by racially-ethnic women and were considered dirty or menial work (Duffy 2007, 316-317). Duffy explicates how a woman’s identity impacts the location for the demand of her labor.

The regional trends observed in my analysis are likely a reflection of racial biases in each region. In decades such as 1930 and 1940, following the great depression, nonwhite foreign born women in the midwest and the north suffered sharper drops in employment rate than both the south and the west. Furthermore the racial biases of the north and the south (specifically for african american women) may have an impact on the observed employment rates. The availability of jobs in the developing markets of each region certainly influences employment rates for all groups of women.

Conclusion:

Although the results of my analysis do not demonstrate dramatic disparities in employment rate between race-immigrant groups, it is true that the occupations that were held by women were largely dependent on their identity. Therefore a woman’s race and nativity had more influence on what occupation she had and less impact on her likelihood of receiving employment.

Bibliography:

Conk, Margo A. “Accuracy, Efficiency, and Bias.” Historical Methods 14.2 (1981): 65-72. ProQuest. Web. 28 Feb. 2016.

Cotter, David A., Joan M. Hermsen, and Reeve Vanneman. “Women’s Work and Working Women: The Demand for Female Labor.” Gender and Society 15.3 (2001): 429-52. Web.

Duffy, Mignon. “Doing the Dirty Work: Gender, Race, and Reproductive Labor in Historical Perspective.” Gender and Society 21.3 (2007): 313-36. Web.

Folbre, Nancy, and Marjorie, Abel. “Women’s Work and Women’s Households: Gender Bias in the U.S. Census.” Social Research 56.3 (1989): 545-69. Web.