Part-time employment of married women in the U.S.A.: a cross-sectional analysis

American Journal of Economics and Sociology, The,  Jan, 1996  by Bijou Yang Lester

• E-mail
• Print
• Link

II

Methods for the Empirical Studies

The present data analysis consists of two parts. A factor analysis was run first in order to identify the orthogonal factors which had a significant Pearson correlation with part-time, full-time, and total employment of married women so that representative variables could be chosen for the regression analysis in the second part. The result of the factor analysis is shown in Table 2. The explanatory variables included in the factor analysis are social/economic variables collected for a study of the rates with which married women and men committed suicide and get murdered (Yang and Lester, 1992). The list included the following: the marriage rate, the divorce rate, the birth rate, the interstate migration rate, the percentage of blacks, the percentage of the population which is urban, the suicide and homicide rates, the percentage of Roman Catholics, church attendance, the total, male and female unemployment rates, the male, female and total labor force participation rate, the population and population density, latitude and longitude, an index of southernness, the percentages of elderly (over 65) and children (under 15), the death rate, the infant mortality rate, the percentage of native born and foreign born, alcohol consumption, the median age, the median family income and the personal income. The dependent variables were the percentage of married women working at all, part-time, and full-time.

The explanatory variables were factor-analyzed using Statistical Package for the Social Sciences (SPSSX) with a principal components extraction and a varimax rotation. Seven orthogonal (independent) factors were extracted (see Table 2). The factor scores for these seven factors were then correlated with the percentages of married women working full-time, part-time, and both using Pearson correlation coefficients (see Table 2).

It can be seen from Table 2 that the percentage of married women working part-time was negatively correlated with scores on Factor 3, a factor which appeared to tap "southernness." Married women worked part-time less often (and full-time more often) in the states with more of a southern subculture.

In addition, the percentage of married women working part-time was positively associated with scores on Factor 6, a factor which appeared to tap the participation rate of men and women in the labor force. In states where more people of both genders were in the labor force, married women worked part-time (and full-time) more often. It can be seen that focusing on married women working without separating the work schedule of full-time from that of part-time fails to capture the complexity of the picture.

The second study sought to explore a particular hypothesis, which was proposed by Tilly, that the relative size of the service industry in the gross state [TABULAR DATA FOR TABLE 2 OMITTED] product (GSP) might be associated with the percentage of married women working part-time. In addition, one representative variable was selected from each of Factors 3, 5, and 6 since they are significantly correlated with married women working either part-time or both part-time and full-time. The inclusion of the latter category is because "any work" covers part-time schedules. The list of the independent variables in the first regression thus include: the percent of service sector in the GSP, the percent of blacks in the state population, the female labor force participation rate, and the female unemployment rate. The data for all the variables except the female labor force participation rate (FLFP) are obtained for the year of 1980. The use of 1979's rather than 1980's FLFP is based on the consideration that the impact of the labor force participation rate on job opportunities tends to be a lagged one.