May 29, 2013

Breadwinner Moms

Appendix 2: Data & Methodology

Decennial Census and American Community Survey

Analysis of the characteristics of mothers, married couples and newlyweds are based on the most recent American Community Survey (ACS) data (2011). The data set was obtained from the IPUMS-USA database26. Minneapolis: University of Minnesota, 2010.] (http://www.ipums.org/) and constructed by the Pew Research Center.

The analysis of historical trends is based on microdata from the Decennial Censuses of 1960, 1970, 1980, 1990 and 2000 and the American Community Surveys (ACS) of 2010 and 2011. The microdata files were obtained from the IPUMS USA database. Data are a 1% sample of the U.S. population for the five decennial censuses and ACS.

Except for 1960, the data are limited to the head of the household ages 15 and older. The head of household were ages 14 and older in 1960. The spousal information is attached to the household head if that person’s marital status is “married, spouse present.”

The American Community Survey is a household survey developed by the U.S. Census Bureau to replace the long form of decennial census program. It is collected throughout the year using mailed questionnaires, telephone interviews, and visits from Census Bureau field representatives to about 3 million household addresses annually.

Parents with children under age 18: This refers to people who have at least one “own” child under age 18 in the household. “Own” children in the Census Bureau data are biological children, stepchildren or adopted children.

Mothers who are the sole or primary provider (Breadwinner moms): The breadwinner mothers consist of two groups of women with children under age 18: One is married and their income is higher than their husband’s, and the other is single (including women who are never married, divorced, separated, widowed, and married but with spouse absent from the household). Because the information about personal earnings (includes wages or income from own business or farm for the previous year) was not available for samples prior to 1990, the total personal income (INCTOT), instead of total personal earned income (INCEARN), was used to compare the earning power between the husbands and wives. According to the Census Bureau, total personal income (INCTOT) includes each respondent’s total pre-tax personal income or losses from all sources for the previous year.

Married population: The currently married population is selected as individuals whose marital status is “married, spouse present.” The spouse must be present in the data set to ascertain his/her education and income information. In these data sets all married couples consist of a man and a woman. The unit of analysis in this report is the head of the household, the married couples in which neither of the spouses is a household head are not included in this study. The IPUMS database includes linkages of spouse records and supplies “attached variables” that place the value for the spouse’s variable on each record. However, for a married person whose spouse is not in the household (married, spouse absent), the spousal information is not available.

Newly married population: The newly married population is a subset of the currently married population drawn from ACS 2011. Beginning in 2008, the ACS includes questions related to date of marriage. One question asked respondents if they had been married (or divorced or widowed) in the “past 12 months.”27 Those saying they had married are the basis for the analyses of “newlyweds” and “new marriages” in this report.

Race and ethnicity: Unless otherwise noted, the terms “whites,” “blacks,” and “Asians” exclude the non-Hispanic components of their populations.

Weighting: All estimates have been weighted to reflect the actual population.

Public Opinion Survey

The survey findings presented in this report are based on an omnibus survey, conducted April 25 to 28, 2013, with a nationally representative sample of 1,003 adults living in the continental United States. Telephone interviews were conducted by landline (500) and cell phone (503, including 237 without a landline phone). The survey was conducted by Princeton Survey Research Associates International (PSRAI). Interviews were done in English by Princeton Data Source. Statistical results are weighted to correct known demographic discrepancies. The margin of sampling error for the complete set of weighted data is plus or minus 3.5 percentage points at the 95% confidence level.

  1. Steven Ruggles, J. Trent Alexander, Katie Genadek, Ronald Goeken, Matthew B. Schroeder and Matthew Sobek. Integrated Public Use Microdata Series: Version 5.0 (Machine-readable database)
  2. Because each year’s ACS is an aggregation of 12 monthly samples, responses to this question cannot be assigned to a specific year or set of months. Thus the marriages are reported as occurring in the “previous year.”