Women's Health Issues 25-3 (2015) 283–288
www.whijournal.com
Original article
Sick of Inequality: Gender and Support for Paid Sick Days Danielle J. Lindemann, PhD a,b,*, Linda Houser, PhD c, Karen White d a
Center for Women and Work at Rutgers University, New Brunswick, New Jersey Department of Sociology, Lehigh University, Bethlehem, Pennsylvania c Widener University’s Center for Social Work Education, Chester, Pennsylvania d Center for Women and Work at Rutgers, The State University of New Jersey, New Brunswick, New Jersey b
Article history: Received 17 October 2014; Received in revised form 11 March 2015; Accepted 18 March 2015
a b s t r a c t Background: The availability of paid sick days (PSD) is on the forefront of policy issues relating to women’s health and well-being. Previous research regarding PSD and other forms of family–work balance legislation has linked access to paid time off from work for addressing one’s own or another’s health concerns to a range of health benefits for working women and their families. In general, public support for such policies is high, but little work has tested the extent to which support extends to PSD. Researchers have yet to engage in a rigorous statistical analysis of public opinion on PSD, including whether opinion varies by gender. Methods: Using data from a 2013 poll of adults in New Jersey (n ¼ 925), we bridged this research gap by conducting the first multivariate analysis of public attitudes toward PSD. Results: As expected, we found markedly high levels of support for PSD across all respondents, with a preponderance of most sociodemographic categories supporting proposed PSD legislation in New Jersey. We also found that gender was a strong predictor of support for PSD, with women significantly (odds ratio, 1.916; p .01) more likely than men to be in favor of such legislation. Conclusions: We discuss the implications of our findings for future work on PSD as well as for research concerning women, wellness, and work-life legislation more broadly. Published by Elsevier Inc.
As the number of women in the workforce and the proportion of dual-earner families continue to increase, researchers and policy makers have begun to focus on workplace policies that will help women as well as men to be productive workers and responsible family members. On the forefront of policy issues relating to women’s health and well-being is the availability of paid sick days (PSD; also called “earned sick days”)da term used to refer to time off that employees earn and can use to recover from illnesses or care for sick family members. Such policies are of particular importance for female workers. Prior research suggests that women are more likely than men to benefit from sick leave. Women function as society’s main caregiversdnot only for children, but also for the elderly and the disabled (Heymann, 2000; National Alliance for Caregiving & AARP, 2009;
Smith & Schaefer, 2012). Accordingly, prior work has identified clear differences in leave-taking behavior by gender (Henrekson & Persson, 2004), and has indicated that women miss work more often than men to care for sick children (Smith & Schaefer, 2012). The presence of young children increases women’s, but not men’s, likelihood of missing work, and this likelihood is further increased for women with young children. Each child under the age of 6 adds about 5% to the probability that a mother will be absent from work during a single year (Vistnes, 1997). Moreover, poor women are among the sociodemographic groups with the least access to leave: two-thirds of low-income women and three-fourths of very poor women do not get paid when they miss work to care for sick children (Wyn, Ojeda, Ranji, & Salganicoff, 2003).1 In fact, one study (Oxfam America, 2013)
* Correspondence to: Danielle J. Lindemann, Assistant Research Professor and Research Director, Center for Women and Work, Rutgers University–New Brunswick, 94 Rockefeller Road, Piscataway, NJ 08854. Phone: þ1 848 932 1275; fax: 732-445-1011. E-mail address: [email protected] (D.J. Lindemann).
1 Both men and women in lower income brackets are at a disadvantage when it comes to paid sick days; for instance, workers with annual personal earnings at or below $14,999 are less likely than workers with higher earnings to have access (O’Connor, Hayes, & Gault, 2014).
1049-3867/$ - see front matter Published by Elsevier Inc. http://dx.doi.org/10.1016/j.whi.2015.03.009
284
D.J. Lindemann et al. / Women's Health Issues 25-3 (2015) 283–288
found that 19% of low-wage working mothers had lost a job from being sick or caring for a sick child. Previous research regarding other forms of health-related, family–work balance legislation has indicated that public support for these kinds of policies is notably high across sociodemographic groups (Milkman & Appelbaum, 2013), but little work has tested whether this applies to PSD. Although existing scholarship has brought attention to the benefits of providing PSD for workers (Lovell, 2004; Hill, 2013; Miller & Williams, 2012) and has examined patterns in leave-taking behavior (Heymann, 2000; Lovell, 2004; Smith & Schaefer, 2012), no published work of which we are aware has engaged in a rigorous, statistical analysis of public opinion regarding this issue. Some researchers and policy experts have conducted analyses of public opinion regarding PSD, but there has been limited distribution of these results in scholarly arenas and reporting has been primarily confined to brief reports (Institute for Women’s Policy Research, 2010; YouGov, 2013).2 Aside from a publication evaluating employers’ attitudes toward paid leave in San Francisco (Boots, Martinson, & Danziger, 2009), only three publicly available documents, of which we are aware, have investigated public opinions regarding PSD by sociodemographic group. These documents include a 2010 report from the National Opinion Research Center (Smith & Kim, 2010) and two 2011 reports from the National Partnership for Women and Families assessing the opinions of registered voters in Connecticut (2011a) and in Philadelphia, Connecticut, and Denver (2011b). However, this previous workdalthough methodologically soundddoes not probe whether differences between sociodemographic categories are significant and, for the most part, does not explore relationships between explanatory variables. By testing for statistical significance, we contribute additional rigor. Significance testing suggests both that our findings are not simply the artifact of one particular poll and that differences by group would be replicated upon additional polling in New Jersey. Further, these prior reports are largely policy oriented and are not geared for academic audiences; our work creates a critical bridge between academia and the policy sphere. PSDs are a crucial political issue at the crux of gender, work, family, and health that should concern both policy makers and academics. In a recent report comparing sick day laws in 22 countries, the Center for Economic Policy Research found that only three countriesdthe United States, Canada, and Japandhave no national policy compelling employers to provide sick days for employees who needed to miss 5 days of work to recover from the flu (Heymann, Rho, Schmitt, & Earle, 2009). More than one-third of the private industry workers in the United States (39%) are not provided paid time off from work when they are ill (U.S. Bureau of Labor Statistics, 2014). An even greater share of the workforced54% dis unable to take time off from work to care for sick children without losing pay or using vacation time (Galinsky, Bond, & Hill, 2004). Eighty-three percent of workers go to work when they are ill, and 21% do this because they are saving their sick leave to stay home with ill children (ComPsych Corporation, 2008). Public support is one crucial catalyst for policy change. In fact, previous research has indicated that voters are inclined to use a candidate’s position on the issue of PSD legislation to inform their voting decisions (National Partnership for Women and Families, 2011b, 2014). Although prior work has demonstrated the importance of PSD policies, particularly to the health and well-being of 2 For a summary of available PSD polling data, see Anzalone Liszt Research, 2011.
female workers, it is critical from both a policy perspective and a research standpoint to understand where dissent and agreement arise. Our access to unique data from a Rutgers–Eagleton poll of adults in New Jersey (N ¼ 925) has allowed us to provide, for the first time, a systematic, multivariate analysis of public attitudes surrounding this issue of critical importance to working women. Our findings have implications for work on PSD and other work– family balance policies, as well as for broader scholarship at the intersection of gender, work, and health. Material and Methods Study Sample Data were drawn from a sample of 925 adults in New Jersey who responded to a Rutgers–Eagleton Poll conducted from September 3 to 9, 2013. Using random digit dialing (both landline and cell phone), the survey selected a random sample of New Jersey residents. This telephone poll included 782 landline and 143 cell phone adults. Any New Jersey adult 18 or over was eligible for the poll, which had a response rate of 22%. The sampling error was 3.2 percentage points, at a 95% CI. Data were weighted, based on the most recent U.S. Census data, to be representative of adults in New Jersey. Weighting reflected known parameters in the New Jersey population: gender, age, race, and Hispanic ethnicity. An assent statement was read to each respondent and the response recorded. The respondent’s participation was terminated if she did not give assent. Informed consent was not written or signed. The study was approved by the Rutgers University–New Brunswick Institutional Review Board. Variables and Hypotheses We examined the likelihood of favoring (“strongly favoring” or “somewhat favoring,” versus “somewhat opposing” or “strongly opposing”) a proposed PSD bill in New Jersey Respondents were asked a two questions focusing on the topic of PSD. The first asked whether the respondent favored requiring all employers to provide PSD in general: Now I’d like to ask you about paid sick days, also called “earned sick days” or “sick pay.” This is time off employees earn and are able to use when they need to stay home and care for their own illness or a sick family member. They can use this time off without losing their job or a days’ wages. Not all workplaces offer paid sick days. But some cities and states have passed or are considering laws requiring all employers to provide access to paid sick days for employees. In general, do you strongly favor, somewhat favor, somewhat oppose, or strongly oppose requiring all employers to provide employees with a minimum number of paid sick days to care for themselves or immediate family members? The second question asked whether the respondent would favor the specific New Jersey legislation: The state legislature is considering a bill giving all New Jersey employees access to paid sick time. Employees would earn 1 hour of paid sick time for every 30 hours worked. Sick time could be used for the employee or a family member in the event of illness, injury, or domestic violence. This would set a minimum state standard, and employers could provide more sick time if desired. Based on this description, what is your opinion NOW about paid sick days?
D.J. Lindemann et al. / Women's Health Issues 25-3 (2015) 283–288
Patterns of response to the two items were highly similar, so, in the interests of space, we drew our dependent variable from the second question, which references a specific policy immediately relevant to respondents as New Jersey residents. We examined variability by gender as well as other sociodemographic categories. Gender Gender was collected as a binary variable and coded based on the interviewers’ observations. Based on the wealth of literature regarding the link between gender and leave-taking behavior (Henrekson & Persson, 2004; Mastekaasa, 2000; Vistnes, 1997), we hypothesized that women would express significantly higher levels of support for PSD than men, and that this relationship would remain significant even when we controlled for other factors such as race and income. Income Respondents were asked, “Last year, that is in 2012, what was your total family income from all sources, before taxes? Just stop me when I get to the right category.” Income was coded as a sixpart categorical variable. We collapsed it into three categories, the lowest being “less than $50,000.” Collapsing was necessary owing to the small numbers of individuals in the lowest income groups. Based on prior work illustrating the strength of the relationship between income and leave-taking needs and behavior (Wyn et al., 2003), we expected to find that support for the New Jersey bill would be highest among those in the lowest earnings category. Political affiliation Respondents were asked, “In politics today, do you consider yourself a Democrat, Republican, Independent, or something else?” Based on prior research suggesting an inverse relationship between Republicanism and support f or social welfare policies (AuClaire, 1984), we hypothesized that Democrats would be significantly more likely than Republicans to favor the New Jersey legislation. Race/Ethnicity We operationalized “race/ethnicity” based on responses to two questions: “Are you of Latino or Hispanic origin, such as Mexican, Puerto Rican, Cuban, or some other Spanish background?” and “Are you White, Black, or of Asian origin, or are you some other race, or multiracial?” Each respondent fell into one of the following five categories: White, Black/African-American, Asian, Hispanic/Latino/Spanish, or multi-racial. Based on prior work (Smith & Kim, 2010), we hypothesized that African Americans would be the top backers of PSD. However, based on previous research concerning the strong relationship between PSD and income (Heymann, 2000; Wyn et al., 2003), we hypothesized that controlling for income would mediate this relationship. Union membership Respondents were asked, “Do you or any member of your household belong to a labor union?” Prior work has found striking attitudinal differences between union members and nonunion members, with union members more likely to support collectivist principles, such as worker participation in management (Fenwick & Olson, 1986) and union representation itself (Hills, 1985). Based on this previous research, we expected to find that adults in New Jersey who were associated with unions would be significantly more likely than nonunion members to favor the state legislation.
285
Family factors We coded responses to the question, “Are you married, in a committed relationship, widowed, divorced, separated, or have you never been married?” in three categories: married, in a committed relationship, or not married (a category collapsing “widowed,” “divorced,” “separated,” and “never married”). We used “parent” to refer to a respondent who answered “yes” to the question, “Are you the parent or guardian of any children under 18 now living in your home?” We hypothesized that individuals with minor children would be particularly likely to need sick time and thus would be more likely than nonparents to express support for PSD. Age Respondents were asked, “To ensure we are reaching people of all ages, would you please tell me your age?” Each respondent fell into one of five cohort-based brackets, including ages 18 to 29, 30 to 39, 40 to 49, 50 to 59, and 60 and over. We hypothesized that younger cohort members would be more likely to support PSD than older cohort members based on the inverse relationship between age and support for government involvement in solving problems found in previous research (Parker, 2012). Respondents who answered “don’t know” or who refused to respond to the PSD question were dropped from the analysis, bringing the number of respondents to 880. We calculated the percentages of respondents in each sociodemographic category who indicated that they “strongly favored” or “somewhat favored” the proposed New Jersey bill. We then examined the relationship between each demographic covariate and the likelihood of favoring PSD, by running a series of separate binary logistic regressions (not shown). After this, we ran a multivariate model including all demographic covariates, again assessing the likelihood of favoring the New Jersey bill, compared with the reference group in each category. Respondents who had missing data for any of the variables in the analysis were dropped, resulting in a regression analysis based on 650 respondents. Unless otherwise specified, all of the findings we report are from the adjusted, multivariate, binary logistic regression analysis. We used SPSS for all analyses. Model fit was assessed by utilizing a receiver operating characteristics (ROC) curve as a diagnostic. We ran tests of multicollinearity using all variables in our model, finding that no two variables were prohibitively collinear. Results and Discussion As Table 1 illustrates most respondentsd85%dfavored the proposed PSD legislation in New Jersey. This finding is in line with previous research on other forms of work–family legislation. In fact, a preponderance of respondents in all sociodemographic categories indicated support for the legislation. Even within the groups with the least amount of supportdRepublicans (74.8%) and New Jerseyans in their 50s (78.5%)dabout three-fourths of respondents indicated favorability. Although our analyses revealed high levels of support for PSD from most New Jerseyans, they also indicated some interesting differences between sociodemographic groups. These are illustrated by the results of the adjusted binary logistic regression (including all variables) that is presented in Table 2. Gender As expected, we found that womendwho, as society’s caretakers, arguably have the most to gain from such policiesdwere significantly more likely than men to register high levels of
286
D.J. Lindemann et al. / Women's Health Issues 25-3 (2015) 283–288
Table 1 Percentages Favoring New Jersey Bill, by Demographic Group (Weighted Sample; n ¼ 880)
access to PSDdwould be most likely to support relevant policies. However, as Table 2 demonstrates, when we controlled for the other variables, none of these associations were significant.
Demographic Group
Favor NJ Bill (%)
All (880) Gender Male (429) Female (451) Income
www.whijournal.com
Original article
Sick of Inequality: Gender and Support for Paid Sick Days Danielle J. Lindemann, PhD a,b,*, Linda Houser, PhD c, Karen White d a
Center for Women and Work at Rutgers University, New Brunswick, New Jersey Department of Sociology, Lehigh University, Bethlehem, Pennsylvania c Widener University’s Center for Social Work Education, Chester, Pennsylvania d Center for Women and Work at Rutgers, The State University of New Jersey, New Brunswick, New Jersey b
Article history: Received 17 October 2014; Received in revised form 11 March 2015; Accepted 18 March 2015
a b s t r a c t Background: The availability of paid sick days (PSD) is on the forefront of policy issues relating to women’s health and well-being. Previous research regarding PSD and other forms of family–work balance legislation has linked access to paid time off from work for addressing one’s own or another’s health concerns to a range of health benefits for working women and their families. In general, public support for such policies is high, but little work has tested the extent to which support extends to PSD. Researchers have yet to engage in a rigorous statistical analysis of public opinion on PSD, including whether opinion varies by gender. Methods: Using data from a 2013 poll of adults in New Jersey (n ¼ 925), we bridged this research gap by conducting the first multivariate analysis of public attitudes toward PSD. Results: As expected, we found markedly high levels of support for PSD across all respondents, with a preponderance of most sociodemographic categories supporting proposed PSD legislation in New Jersey. We also found that gender was a strong predictor of support for PSD, with women significantly (odds ratio, 1.916; p .01) more likely than men to be in favor of such legislation. Conclusions: We discuss the implications of our findings for future work on PSD as well as for research concerning women, wellness, and work-life legislation more broadly. Published by Elsevier Inc.
As the number of women in the workforce and the proportion of dual-earner families continue to increase, researchers and policy makers have begun to focus on workplace policies that will help women as well as men to be productive workers and responsible family members. On the forefront of policy issues relating to women’s health and well-being is the availability of paid sick days (PSD; also called “earned sick days”)da term used to refer to time off that employees earn and can use to recover from illnesses or care for sick family members. Such policies are of particular importance for female workers. Prior research suggests that women are more likely than men to benefit from sick leave. Women function as society’s main caregiversdnot only for children, but also for the elderly and the disabled (Heymann, 2000; National Alliance for Caregiving & AARP, 2009;
Smith & Schaefer, 2012). Accordingly, prior work has identified clear differences in leave-taking behavior by gender (Henrekson & Persson, 2004), and has indicated that women miss work more often than men to care for sick children (Smith & Schaefer, 2012). The presence of young children increases women’s, but not men’s, likelihood of missing work, and this likelihood is further increased for women with young children. Each child under the age of 6 adds about 5% to the probability that a mother will be absent from work during a single year (Vistnes, 1997). Moreover, poor women are among the sociodemographic groups with the least access to leave: two-thirds of low-income women and three-fourths of very poor women do not get paid when they miss work to care for sick children (Wyn, Ojeda, Ranji, & Salganicoff, 2003).1 In fact, one study (Oxfam America, 2013)
* Correspondence to: Danielle J. Lindemann, Assistant Research Professor and Research Director, Center for Women and Work, Rutgers University–New Brunswick, 94 Rockefeller Road, Piscataway, NJ 08854. Phone: þ1 848 932 1275; fax: 732-445-1011. E-mail address: [email protected] (D.J. Lindemann).
1 Both men and women in lower income brackets are at a disadvantage when it comes to paid sick days; for instance, workers with annual personal earnings at or below $14,999 are less likely than workers with higher earnings to have access (O’Connor, Hayes, & Gault, 2014).
1049-3867/$ - see front matter Published by Elsevier Inc. http://dx.doi.org/10.1016/j.whi.2015.03.009
284
D.J. Lindemann et al. / Women's Health Issues 25-3 (2015) 283–288
found that 19% of low-wage working mothers had lost a job from being sick or caring for a sick child. Previous research regarding other forms of health-related, family–work balance legislation has indicated that public support for these kinds of policies is notably high across sociodemographic groups (Milkman & Appelbaum, 2013), but little work has tested whether this applies to PSD. Although existing scholarship has brought attention to the benefits of providing PSD for workers (Lovell, 2004; Hill, 2013; Miller & Williams, 2012) and has examined patterns in leave-taking behavior (Heymann, 2000; Lovell, 2004; Smith & Schaefer, 2012), no published work of which we are aware has engaged in a rigorous, statistical analysis of public opinion regarding this issue. Some researchers and policy experts have conducted analyses of public opinion regarding PSD, but there has been limited distribution of these results in scholarly arenas and reporting has been primarily confined to brief reports (Institute for Women’s Policy Research, 2010; YouGov, 2013).2 Aside from a publication evaluating employers’ attitudes toward paid leave in San Francisco (Boots, Martinson, & Danziger, 2009), only three publicly available documents, of which we are aware, have investigated public opinions regarding PSD by sociodemographic group. These documents include a 2010 report from the National Opinion Research Center (Smith & Kim, 2010) and two 2011 reports from the National Partnership for Women and Families assessing the opinions of registered voters in Connecticut (2011a) and in Philadelphia, Connecticut, and Denver (2011b). However, this previous workdalthough methodologically soundddoes not probe whether differences between sociodemographic categories are significant and, for the most part, does not explore relationships between explanatory variables. By testing for statistical significance, we contribute additional rigor. Significance testing suggests both that our findings are not simply the artifact of one particular poll and that differences by group would be replicated upon additional polling in New Jersey. Further, these prior reports are largely policy oriented and are not geared for academic audiences; our work creates a critical bridge between academia and the policy sphere. PSDs are a crucial political issue at the crux of gender, work, family, and health that should concern both policy makers and academics. In a recent report comparing sick day laws in 22 countries, the Center for Economic Policy Research found that only three countriesdthe United States, Canada, and Japandhave no national policy compelling employers to provide sick days for employees who needed to miss 5 days of work to recover from the flu (Heymann, Rho, Schmitt, & Earle, 2009). More than one-third of the private industry workers in the United States (39%) are not provided paid time off from work when they are ill (U.S. Bureau of Labor Statistics, 2014). An even greater share of the workforced54% dis unable to take time off from work to care for sick children without losing pay or using vacation time (Galinsky, Bond, & Hill, 2004). Eighty-three percent of workers go to work when they are ill, and 21% do this because they are saving their sick leave to stay home with ill children (ComPsych Corporation, 2008). Public support is one crucial catalyst for policy change. In fact, previous research has indicated that voters are inclined to use a candidate’s position on the issue of PSD legislation to inform their voting decisions (National Partnership for Women and Families, 2011b, 2014). Although prior work has demonstrated the importance of PSD policies, particularly to the health and well-being of 2 For a summary of available PSD polling data, see Anzalone Liszt Research, 2011.
female workers, it is critical from both a policy perspective and a research standpoint to understand where dissent and agreement arise. Our access to unique data from a Rutgers–Eagleton poll of adults in New Jersey (N ¼ 925) has allowed us to provide, for the first time, a systematic, multivariate analysis of public attitudes surrounding this issue of critical importance to working women. Our findings have implications for work on PSD and other work– family balance policies, as well as for broader scholarship at the intersection of gender, work, and health. Material and Methods Study Sample Data were drawn from a sample of 925 adults in New Jersey who responded to a Rutgers–Eagleton Poll conducted from September 3 to 9, 2013. Using random digit dialing (both landline and cell phone), the survey selected a random sample of New Jersey residents. This telephone poll included 782 landline and 143 cell phone adults. Any New Jersey adult 18 or over was eligible for the poll, which had a response rate of 22%. The sampling error was 3.2 percentage points, at a 95% CI. Data were weighted, based on the most recent U.S. Census data, to be representative of adults in New Jersey. Weighting reflected known parameters in the New Jersey population: gender, age, race, and Hispanic ethnicity. An assent statement was read to each respondent and the response recorded. The respondent’s participation was terminated if she did not give assent. Informed consent was not written or signed. The study was approved by the Rutgers University–New Brunswick Institutional Review Board. Variables and Hypotheses We examined the likelihood of favoring (“strongly favoring” or “somewhat favoring,” versus “somewhat opposing” or “strongly opposing”) a proposed PSD bill in New Jersey Respondents were asked a two questions focusing on the topic of PSD. The first asked whether the respondent favored requiring all employers to provide PSD in general: Now I’d like to ask you about paid sick days, also called “earned sick days” or “sick pay.” This is time off employees earn and are able to use when they need to stay home and care for their own illness or a sick family member. They can use this time off without losing their job or a days’ wages. Not all workplaces offer paid sick days. But some cities and states have passed or are considering laws requiring all employers to provide access to paid sick days for employees. In general, do you strongly favor, somewhat favor, somewhat oppose, or strongly oppose requiring all employers to provide employees with a minimum number of paid sick days to care for themselves or immediate family members? The second question asked whether the respondent would favor the specific New Jersey legislation: The state legislature is considering a bill giving all New Jersey employees access to paid sick time. Employees would earn 1 hour of paid sick time for every 30 hours worked. Sick time could be used for the employee or a family member in the event of illness, injury, or domestic violence. This would set a minimum state standard, and employers could provide more sick time if desired. Based on this description, what is your opinion NOW about paid sick days?
D.J. Lindemann et al. / Women's Health Issues 25-3 (2015) 283–288
Patterns of response to the two items were highly similar, so, in the interests of space, we drew our dependent variable from the second question, which references a specific policy immediately relevant to respondents as New Jersey residents. We examined variability by gender as well as other sociodemographic categories. Gender Gender was collected as a binary variable and coded based on the interviewers’ observations. Based on the wealth of literature regarding the link between gender and leave-taking behavior (Henrekson & Persson, 2004; Mastekaasa, 2000; Vistnes, 1997), we hypothesized that women would express significantly higher levels of support for PSD than men, and that this relationship would remain significant even when we controlled for other factors such as race and income. Income Respondents were asked, “Last year, that is in 2012, what was your total family income from all sources, before taxes? Just stop me when I get to the right category.” Income was coded as a sixpart categorical variable. We collapsed it into three categories, the lowest being “less than $50,000.” Collapsing was necessary owing to the small numbers of individuals in the lowest income groups. Based on prior work illustrating the strength of the relationship between income and leave-taking needs and behavior (Wyn et al., 2003), we expected to find that support for the New Jersey bill would be highest among those in the lowest earnings category. Political affiliation Respondents were asked, “In politics today, do you consider yourself a Democrat, Republican, Independent, or something else?” Based on prior research suggesting an inverse relationship between Republicanism and support f or social welfare policies (AuClaire, 1984), we hypothesized that Democrats would be significantly more likely than Republicans to favor the New Jersey legislation. Race/Ethnicity We operationalized “race/ethnicity” based on responses to two questions: “Are you of Latino or Hispanic origin, such as Mexican, Puerto Rican, Cuban, or some other Spanish background?” and “Are you White, Black, or of Asian origin, or are you some other race, or multiracial?” Each respondent fell into one of the following five categories: White, Black/African-American, Asian, Hispanic/Latino/Spanish, or multi-racial. Based on prior work (Smith & Kim, 2010), we hypothesized that African Americans would be the top backers of PSD. However, based on previous research concerning the strong relationship between PSD and income (Heymann, 2000; Wyn et al., 2003), we hypothesized that controlling for income would mediate this relationship. Union membership Respondents were asked, “Do you or any member of your household belong to a labor union?” Prior work has found striking attitudinal differences between union members and nonunion members, with union members more likely to support collectivist principles, such as worker participation in management (Fenwick & Olson, 1986) and union representation itself (Hills, 1985). Based on this previous research, we expected to find that adults in New Jersey who were associated with unions would be significantly more likely than nonunion members to favor the state legislation.
285
Family factors We coded responses to the question, “Are you married, in a committed relationship, widowed, divorced, separated, or have you never been married?” in three categories: married, in a committed relationship, or not married (a category collapsing “widowed,” “divorced,” “separated,” and “never married”). We used “parent” to refer to a respondent who answered “yes” to the question, “Are you the parent or guardian of any children under 18 now living in your home?” We hypothesized that individuals with minor children would be particularly likely to need sick time and thus would be more likely than nonparents to express support for PSD. Age Respondents were asked, “To ensure we are reaching people of all ages, would you please tell me your age?” Each respondent fell into one of five cohort-based brackets, including ages 18 to 29, 30 to 39, 40 to 49, 50 to 59, and 60 and over. We hypothesized that younger cohort members would be more likely to support PSD than older cohort members based on the inverse relationship between age and support for government involvement in solving problems found in previous research (Parker, 2012). Respondents who answered “don’t know” or who refused to respond to the PSD question were dropped from the analysis, bringing the number of respondents to 880. We calculated the percentages of respondents in each sociodemographic category who indicated that they “strongly favored” or “somewhat favored” the proposed New Jersey bill. We then examined the relationship between each demographic covariate and the likelihood of favoring PSD, by running a series of separate binary logistic regressions (not shown). After this, we ran a multivariate model including all demographic covariates, again assessing the likelihood of favoring the New Jersey bill, compared with the reference group in each category. Respondents who had missing data for any of the variables in the analysis were dropped, resulting in a regression analysis based on 650 respondents. Unless otherwise specified, all of the findings we report are from the adjusted, multivariate, binary logistic regression analysis. We used SPSS for all analyses. Model fit was assessed by utilizing a receiver operating characteristics (ROC) curve as a diagnostic. We ran tests of multicollinearity using all variables in our model, finding that no two variables were prohibitively collinear. Results and Discussion As Table 1 illustrates most respondentsd85%dfavored the proposed PSD legislation in New Jersey. This finding is in line with previous research on other forms of work–family legislation. In fact, a preponderance of respondents in all sociodemographic categories indicated support for the legislation. Even within the groups with the least amount of supportdRepublicans (74.8%) and New Jerseyans in their 50s (78.5%)dabout three-fourths of respondents indicated favorability. Although our analyses revealed high levels of support for PSD from most New Jerseyans, they also indicated some interesting differences between sociodemographic groups. These are illustrated by the results of the adjusted binary logistic regression (including all variables) that is presented in Table 2. Gender As expected, we found that womendwho, as society’s caretakers, arguably have the most to gain from such policiesdwere significantly more likely than men to register high levels of
286
D.J. Lindemann et al. / Women's Health Issues 25-3 (2015) 283–288
Table 1 Percentages Favoring New Jersey Bill, by Demographic Group (Weighted Sample; n ¼ 880)
access to PSDdwould be most likely to support relevant policies. However, as Table 2 demonstrates, when we controlled for the other variables, none of these associations were significant.
Demographic Group
Favor NJ Bill (%)
All (880) Gender Male (429) Female (451) Income
