Time Use on Caregiving Activities

Abstract

Working mothers in federal service spend about 20 min per day less on caregiving activities, compared to their counterparts in the private sector. This result holds regardless of the type of job they hold, their educational attainment, marital status, the number and ages of their children, or the employment status of their spouse. This is an important result to federal agency recruitment, which targets a similar labor pool as does the private sector. It is also important to the retention of human capital in federal government, which has sought to establish a reputation as a model employer through the development and implementation of family-friendly workplace programs and a culture that supports overall work–life balance. However, mothers in federal service spend more time at work compared to their counterparts in the private sector, which prompts one to wonder whether less caregiving time and more work time is true balance.

Keywords

benefits family-friendly workplace federal government HRM gender and public personnel administration workplace environment/culture

“The National Government should be a model employer. It should demand the highest quality of service from each of its employees and it should care for all of them properly in return.”

From President Theodore Roosevelt’s 1907 State of the Union Address

The federal government has long aspired to the “model employer” status—for more than a century, in fact. The quote above is from President Theodore Roosevelt’s 1907 State of the Union Address (1907, para. 38), which is arguably the first recognition of a need for policy intervention in light of a changing workforce as women increasingly entered the paid labor force, largely in garment manufacturing. As an employer, Roosevelt implored federal government to set high standards and serve as an exemplar for all employers. Although its efforts to be a model employer have waxed and waned over the years, “caring for all of them properly” has come to mean attending to the whole worker: providing resources and benefits to allow employees to balance time pressures from demands within and beyond the workplace. Every worker faces demands for her time from a variety of sources, but as the demographic characteristics of the workforce have changed government has been the first to respond, with both benefits for its own workers and policies governing all workers.

In this article, I analyze reports of time use on caregiving activities by working parents in federal service and in for-profit firms to see whether the former compares favorably to the latter with respect to its efforts to foster a family-friendly workplace through time-saving programs like providing child care subsidies and facilities or work–life balance efforts such as flexible work schedules and wellness programs. To determine whether such a comparison is justified, I first present a simple analysis of the occupational composition of the federal government workforce and that of the private sector to determine the comparability of the two with respect to the characteristics of their labor forces. The types of workers demanded in federal service more closely resemble the private sector than they resemble state and local governments. Federal service is more similar to the private sector than to the state and the local governments, given what each level of government is tasked with in the economy. Given the appropriateness of comparing the nature of working in federal service to working in the private sector, I then consider whether the amount of time spent on caregiving differs for women in federal service compared to women in the private sector. This is an appropriate question to consider given the efforts by the federal government to establish itself as an attractive place to work—a model employer—through years of personnel management reform of which family-friendly policies and work–life programs have played significant parts.

Using data from a nationwide survey, the U.S. Department of Labor Bureau of Labor Statistics’ American Time Use Survey (ATUS), I compare the amounts of time devoted to caregiving by women and men in the federal service and private sector—not to test the effectiveness of any particular Office of Personnel Management (OPM) program, but rather, to compare the outcomes between groups of parents who face conflicting demands for their time. One group—those in federal service—has access to a range of family-friendly workplace policies that have been developed and promoted heavily for years. Access to such programs varies widely across employers for those in the other group, the private sector workers (Hegtvedt, Clay-Warner, & Ferrigno, 2002; Jang, 2009). This aggregate finding that women in federal service spend less time on care activities, even when accounting for a range of alternate explanations, provides the foundation upon which others can evaluate specific OPM programs for their effectiveness in fostering work–life balance toward their goal of being a model employer. I conclude with a brief discussion of the concept of work–life balance: Does less care time and more work time indicate greater balance, an interpretation rooted in a conflict model (Greenhaus & Powell, 2006)? Or rather, would more time spent on care activities indicate greater work–life balance (Sayer, 2005), an interpretation borne of a theory of work–family enrichment (Greenhaus & Powell, 2006)?

Comparing Workforces

I compare time spent on family caregiving activities reported by workers in federal service to those working in the private sector. Is this a valid comparison? OPM thinks so: Comparisons to the private sector are included in results from biennial Federal Human Capital Surveys (FHCS, recently renamed the Federal Employee Viewpoint Survey, which will be conducted annually starting in 2011). FHCS results compare proportions of positive responses by federal workers and private sector workers to about a dozen questions related to job satisfaction and personal work experiences. This is done, presumably, to show that federal workers are at least as satisfied with their jobs and experience with work as are private sector workers. Continued reforms to federal compensation schemes are meant to make federal government pay competitive to private sector pay (Bowman, 2010; O’Toole & Churchill, 1982; Perry, 2010). Pay system reforms and FHCS reports betray OPM’s awareness that federal agencies compete with private sector firms for much of the same talent; that the federal government workforce is much more similar to that of the private sector than it is to state and local government workforces.

Federal Versus Private Sector: A Reasonable Comparison?

A simple analysis of leading occupations in each sector illustrates this point. Appendix A contains four tables (Tables 1A, 2A, 3A, and 4A): the list of occupations comprising half of total employment in local government, state government, federal government, and the private sector (excluding self-employed) in 2009. Eleven occupations comprised 50.35% of total employment in local governments across the country, mainly in primary and secondary education. Protective services—police and fire—and general administration are also among these leading occupations. In state governments, 17 occupations comprised 50.15% of total employment, primarily in the areas of education, health, and social services, but also in law and general administration. State and local governments share similar functions and their occupational distributions reflect that. Tables 3A and 4A in Appendix A contain the leading occupations in federal government and the private sector, respectively. Twenty-seven occupations comprised 50.23% of total employment in federal service in 2009, and 37 occupations comprised 50.62% of private sector employment. Shared occupations between the two sectors include managerial positions, finance, information technology, and general administrative services. Nearly half of the leading occupations in federal service—in terms of numbers employed—are shared with the private sector. The two sectors overlap and potentially compete for talent in a number of important professional arenas: finance, operations management, information technology, and legal services. The two sectors depart predictably: Leading federal government occupations include jobs with the postal service and Armed Forces, and leading private sector jobs include those in retail sales and goods-producing industries like construction, manufacturing, and agriculture.

In short, it is appropriate to compare the experiences of federal government workers and those in private sector firms. The comparison is meaningful because similar skill sets are demanded by both in many of the leading occupations in federal service and the private sector. What’s more, it is meaningful to compare the work–life balance experiences of workers in these two groups because such policies are important to the federal government’s recruitment and retention efforts. To attract the best and brightest workers who have alternative employment options in the private sector, federal agencies have marketed themselves as among the best places to work, owing significantly to the balanced workplace culture they have developed over time through family-friendly policies and programs. Finally, it is important to compare the amount of time spent on caregiving activities between two groups that might have similar skill sets but who may enjoy very different workplace benefits. Federal workers have access to a range of family-friendly workplace policies, whereas access to such programs varies widely across employers for private sector workers.

Federal Sector Workplace Policies: Access Versus Uptake

Wood (1999, p. 114) concludes that the types of organizations most likely to adopt what he calls “family-friendly management” are those “where top management place a high value on employees’ welfare in respect to the family situation, perceive a bottom line-benefit from providing family-related benefits, keep in tune with their workforce through consultation processes and give a high priority to the achievement of employee commitment.” He finds that the public sector is more likely to embrace the principles of family-friendly management and, indeed, the federal government meets these criteria: Through OPM, federal agencies place a high value on employee welfare, perceive a bottom-line benefit from providing these benefits, and keep in tune with their employees through annual surveys. Evidence of uptake and satisfaction with work–life benefits comes from FHCS results. Agencies have a powerful incentive to encourage participation in work–life programs (Parker, 2010, para. 7):

The surveys have been influential in measuring how successful agencies are at motivating and managing their employees. The Partnership for Public Service uses the data to compile a “Best Places to Work” list every 2 years, which the Office of Management and Budget said in 2009 would be a factor in agency funding requests.

Federal employee job satisfaction is positively related to increases in benefits such as child care subsidies, Employee Assistance Programs (EAPs), and other health and wellness resources (Saltzstein, Ting, & Saltzstein, 2001). Flexible scheduling is particularly associated with higher job satisfaction (Ezra & Deckman, 1996; Facer & Wadsworth, 2008; Jang, 2009; Keene & Quadagno, 2004; Marler & Moen, 2005) and a sense of work–life balance (Shockley & Allen, 2007; Wadsworth & Owens, 2007; Wharton, 1994).

If OPM links job satisfaction as gauged by FHCS results to budget appropriations, then every federal agency faces a strong incentive to increase participation in its benefit programs.

Using the survey response “No Basis to Judge” as an indication that the respondent does not or cannot use a particular benefit asked about in the FHCS, Figure 1 shows that a substantial portion of federal workers make use of at least one of the five main work–life programs available to them at their agencies. The FHCS work–life items are categorized as telework, alternative work schedules, paid leave, child care subsidies, or other work–life support programs, including health and wellness programs, the Employee Assistance Program (EAP), elder care, and various support groups.

Figure 1.

Work–life program uptake by federal employees, 2004-2010

Less than half of federal workers responding to the survey take advantage of child care subsidies available to them, which may be explained by the lack of need for workers whose children are older or who do not have children or if they have alternate child care arrangements. Moreover, no direct questions were asked about Paid Leave in 2010, which is clearly the most popular benefit, perhaps due to its ease of use and its flexibility. Telework is the least exploited benefit and has been since its inception (Newman & Matthews, 1999). The 2010 survey asked specifically about respondent’s views on telework. More than one third responded that they cannot telecommute because they must be physically present on the job (e.g., security personnel); 7% cannot telecommute due to technical limitations (e.g., connectivity or inadequate equipment); however, more important, nearly one quarter responded, “I do not telework because I am not allowed to, even though I have the kind of job where I can telework” (FHCS, 2010, Item 72).

According to Newman and Matthews (1999), these comments reflect a deep lack of trust that telecommuting employees are actually working: “Wariness on the part of supervisors and a distrustful organizational culture are two examples of the forces inhibiting the effective operation of family-friendly workplace policies.” OPM understands that “the missions of many federal agencies require at least a percentage of their employees to be physically present on a daily basis in order to support critical systems and processes” (2011, p. 15) but also that “Presenteeism, the practice of sitting at one’s desk without working, can be just as problematic as absenteeism” (Berry, 2011, para. 8). To change the minds of both workers and supervisors about the merits of telework, Congress passed the Telework Enhancement Act of 2010, which among other things, “requires each agency to designate a senior management official as the Telework Managing Officer to help transform the use of telework” (Berry, 2011, para. 3). Other requirements of the act include a range of administrative reforms to increase accountability and oversight of telecommuting arrangements within federal agencies. These changes should increase uptake, for they are consistent with Newman and Matthews’ conclusions that “the organizational placement of the administration of work/life programs . . . may be a partial predictor of success” (1999, p. 41) and that such programs “are predicated on a foundation of trust, and feature a focus on work outcomes” (p. 41). OPM encourages increased use of telework as a flexibility tool for managers to maintain continuity of operations in the event of adverse weather, public health crises like the influenza pandemic of 2010, or terrorist threats that might restrict access to DC, as well as a strategy to reduce congestion and auto emissions in the area (OPM, 2011).

But organizational norms are difficult to overcome. McCurdy, Newman, and Lovrich find that among public employees that “if you use a work/life policy, you are at risk of losing a competitive edge—of being perceived as being less committed to one’s professional advancement and less willing to make the necessary sacrifices in pursuit of that advancement” (2002, p. 47). This is confirmed by Tower and Alkadry (2008), and Lewis (1997, p. 15) concurs, “If workers are ambivalent about their right to be sick, it is not surprising that they often fear repercussions for using family-oriented provisions which are constructed as a concession for those who cannot conform to ‘normal’ working patterns.” What is more, Barnett (1999) observes how women can be penalized for trying to strike a balance using family-friendly workplace benefits: “Many women felt accurately that these benefits came with strings attached: women who took advantage of them were seen as less committed and less desirable. Their opportunities at the workplace were often curtailed, and their long-term career plans, jeopardized. Thus the informal corporate culture was often more critical in shaping employee behavior than the formal policies were” (p. 147). Trends in program uptake illustrated in Figure 1 reflect these findings.

However, Bruce and Reed (1994, p. 39) note that “the federal government has been the trendsetter in the use of alternative work schedules, part-time career tracks, on-site childcare, and leave-sharing programs . . . [and leads] the rest of the country in percentage of employees on flexible work schedules, and in permanent part-time career opportunities for professional and administrative women.” Program uptake may not be at optimal levels (OPM, 2011), but federal government workers still lead private sector workers in their access to and participation in family-friendly programs.

Data, Model, and Method

Having consistent access to a host of family-friendly workplace programs, and that federal workers take advantage of them more than other workers can or do (Bruce & Reed, 1994; Hoyman & Duer, 2004), and the very strong incentives agencies have to promote participation, I hypothesize that female workers in federal service will be able to spend less time on caregiving activities compared to workers in the private sector who do not have such access. Again, this is not a direct evaluation of any particular OPM program, but rather, it is an assessment of aggregate outcomes using detailed data from time-use diaries completed by respondents across the country. No information currently exists that links time use directly to participation in federal family-friendly workplace programs for a nationwide representative sample of workers. If the hypothesis is confirmed, then these results should prompt further research into potential reasons why. Data from time-use diaries come from the ATUS, which uses a nationally representative subset of households in the Current Population Survey (Abraham, Flood, Sobek, & Thorn, 2008; U.S. Department of Labor, Bureau of Labor Statistics [BLS], 2008). ATUS has been conducted annually since 2003. I use pooled data from 2003 to 2009 with the following model:

LN(Caregiving Time) = f[FEM*FED, FEM, FED, Age, Age Squared, Married, Spouse Employment, Number of Children, Infant, Toddler, Kid, Some Postsecondary, College Degree, Postgraduate, P, A, T, O, Work Time, Year]

The natural log of caregiving time is a function of worker characteristics such as sex, age, marital status, educational attainment, and spouse employment status; family characteristics include number and ages of children, and job characteristics include sector (federal vs. private) and job type (professional, administrative, technical, or other; Mastracci, 2010).

The ATUS definition of caregiving time is the number of minutes spent per day on activities related to children’s personal care, education, and health (BLS, 2008). Appendix B lists the activities included in this definition, which is extensive and covers multitasking, as long as the child is engaged. For example, time spent preparing a meal would be counted if the parent were also asking the child about her day or if the child were helping prepare the meal as well. Separate survey questions gather information on the amount of time spent on housework, shopping, yard work, and other household management tasks.

Worker Characteristics

Female is a dichotomous variable denoting respondent sex (equal to 1 if female, 0 otherwise), Federal is a dichotomous variable denoting whether one works in the federal government (equal to 1 if federal, 0 otherwise), and FEM*FED is a multiplicative interaction of these two. As such, FEM*FED will drop out for all respondents except those who are female and work in federal government. It is the key variable of interest. If the coefficient on this variable is negative and statistically different from zero, the hypothesis that women in federal service spend less time on caregiving activities would be confirmed. The data include both male and female respondents; the former are included as controls to capture overall effects on time use over the 2003-2009 period. It is appropriate to construct the research hypothesis as a statement about women’s time use, however, because women remain the primary caregivers: “married mothers were more likely to provide childcare to household children than were married fathers” (BLS, 2008, para. 7). Women have more potential for work–family conflict than men do (Arrighi & Maume, 2000; Braun, Lewin-Epstein, Stier, & Baumgartner, 2008; Cinamon & Rich, 2002; Coltrane, 2000; Guendouzi, 2006; Keene & Quadagno, 2004; Mannino & Deutsch, 2007; Maume, 2006; Mikelson, 2008; Sayer, 2005; Wood, 1999).

Both age and its square are included in the model because the relationship between caregiving time and age is reasoned to be nonlinear: Workers early and late in their careers spend less time taking care of children compared to those in mid-career. Early-career workers may not have children yet, and late-career workers’ children are likely to be older and require less direct care. Age of parents affects time use on caregiving activities not only because older parents are more likely to have older children who require relatively less time-intensive basic care but also because parents who delayed starting their families also tend to share caregiving activities more equitably (Helms-Erikson, 2001).

Family Characteristics

Number of children is the number of children in the household under age 18; infant, toddler, and kid are dichotomous variables indicating whether there is an infant, child aged 3 to 5, or a child aged 6 to 12, respectively, in the household. The number of children in the household and caretaking time should be positively related, other things held constant, because “women . . . end up doing a larger share of family work as the number of children increases” (Coltrane, 2000, p. 1222). Effects on care time due to children’s ages are less clear (Arrighi & Maume, 2000; Braun et al., 2008; Guendouzi, 2006). Teenagers might assume more of their own care, so, as the reference (i.e., omitted) category, I expect the presence of infants, toddlers, and kids, to demand more caregiving time (Sullivan & Gershuny, 2001), so the coefficients on these variables should be positive. Evidence from Bureau of Labor Statistics analysis supports this expectation: Parents “of children under six spent more than twice as much time providing childcare on an average day as did their counterparts whose youngest child was age six to seventeen” (BLS, 2008, para. 9). Furthermore, I would expect that the magnitude of effect would be greatest for infants, less for toddlers, and even less for kids, as they become progressively self-sufficient (Sullivan & Gershuny, 2001). Some postsecondary, college degree, and postgraduate denote educational attainment: Some postsecondary education, holding a 4-year degree, and having at least some postgraduate education, respectively. High school diploma is the reference category, and it also includes the small portion of respondents with less than a high school education (less than 10%). Prior research suggests that the more years of education attained, the fewer minutes per day spent on caregiving activities: “In general, studies suggest that women with more education do less housework” (Coltrane, 2000, p. 1221); therefore, I expect that the coefficients on these variables to be negative. Married denotes whether the respondent is married or not. “Women tend to feel more obligation to perform household labor . . . when they get married. Single and cohabiting women perform less housework than do married women” (Coltrane, 2000, p. 1222). So the coefficient on married should be positive. Spouse employment status is also a dichotomous variable indicating whether the respondent’s spouse is working or not. The effect of this independent variable on time use should be positive (Arrighi & Maume, 2000).

Job Characteristics

Dichotomous job characteristic variables indicate whether the job is professional, administrative, technical, or other, consistent with the federal government’s PATCO categories. I use a previously constructed cross-walk to translate Census job codes to PATCO categories (Mastracci & Thompson, 2009). Clerical is the comparison category. Wood (1999, p. 114) finds that the availability of family-friendly workplace programs to be positively related to the proportion of workers in this category: “the proportion of the clerical workers in the workforce, many of whom are women, is significantly related to the adoption of family-friendly management.” Wood addresses the supply side, but what about the demand for these policies as indicated by program uptake? Despite the greater availability of such programs due to the proportion of the clerical workforce, Hochschild (1997) finds their uptake to be negatively related to the size of the clerical workforce. Compared to Clerical, P, A, and T should be inversely related to caregiving time, if Hochschild is correct:

[This clerical worker] didn’t want to hire someone to pick up the children, as . . . top women managers did. She wanted to be that person. . . . Employees higher up the . . . ladder happily delegated aspects of their parental role to others [who] had less money and less desire to outsource parts of . . . motherhood. (1996, p. 135, italics in original)

Coltrane (2000, p. 1221) further supports this assumption, “Women’s higher occupational status and income . . . is strongly associated with the purchase of domestic services.” For these reasons, I expect the coefficients on these variables to be negative: Other job types will spend less time on caregiving activities compared to workers in the clerical category. This outcome is anticipated not only based on the findings described above (Coltrane, 2000; Hochschild, 1997; Wood, 1999) but also due to occupational segregation by gender (Kerr, Miller, & Reid, 2002), in other words, the overrepresentation of women in the clerical category and women’s greater likelihood to assume responsibility for child care (BLS, 2008).

Two Important Restrictions

I restrict the analysis in two important ways. First, I limit the data to include households with at least one child less than 18 years of age. Households without children face very different time-use constraints than do those without, and although work–life balance is important to all employees (Hoyman & Duer, 2004), those with children are the primary targets of family-friendly workplace policies. As much as 93% of respondents from households with no children report zero minutes of caregiving time per day.

Second, I limit the analysis to respondents who reported at least 1 min of caregiving time per day. Fully one third of all respondents—even those with children under age 18 at home—reported “zero” minutes of care taking time, resulting in highly skewed data on this variable. The natural log of those data approximates a normal distribution, allowing for the use of ordinary least squares regression analysis (see Appendix C), but it also ends up dropping observations reporting zero minutes of caregiving time because the natural log of zero is undefined. I could have added one to every observation and then conducted the log transformation, but the purpose of the study is to determine the difference in care time spent between two groups, one of which has access to an array of family-friendly workplace policies and the other does not. Respondents reporting zero care time would not benefit from such policies or suffer from their absence, so I determined it appropriate to allow those observations to drop as a result of the log transformation. Working parents with at least one child at home who report zero minutes of caregiving time differ from those reporting at least some caregiving time in expected and unexpected ways. As one might expect, those reporting zero are more likely to be men: 69% of respondents reporting no caregiving time are male compared to only 49% reporting at least some caregiving time; they are also more likely to have a teenager at home (52% compared to 29%) and less likely to have an infant (6% to 11%) or toddler (24% to 42%). Unexpectedly, however, respondents reporting zero caregiving time are about as likely to be single (about 14%) or married (about 80%). Spouse employment status does not factor into whether a respondent reported some or no caregiving time, either.

Descriptive Statistics: Federal and Private Sector

Table 1 contains mean values on a range of indicators for parents working in federal government or in private sector firms who report at least some caregiving time and who hold jobs that are leading occupations in both the federal government and private sector.

Table 1.

Variable Means of Working Parents in Federal and Private Sector Employment.

	Federal government	Private sector
Caregiving time (working mothers)	128 min/day	126 min/day
Caregiving time (working fathers)	111 min/day	103 min/day
Age	39.3 years	36.4 years
Proportion female	47.5%	51.3%
Proportion married	81.4%	78.6%
Proportion with employed spouse	73.8%	77.1%
Number and ages of children
Number of children (age < 18)	About 2 (1.90)	About 2 (1.90)
Infant (age < 3)	8.8%	12.1%
Toddler (age 3-5)	40.5%	42.8%
Kid (age 6-12)	57.4%	53.1%
Teen (age 13-17)	33.2%	29.5%
Educational attainment
High school diploma or less	24.8%	42.7%
Some postsecondary	26.3%	23.9%
Four-year degree	24.0%	20.0%
At least some postgraduate	24.9%	13.5%
Job type
Professional (P)	22.9%	21.1%
Administrative (A)	7.2%	4.0%
Technical (T)	22.1%	16.4%
Clerical (C)	11.3%	11.0%
Other (O)	36.6%	47.4%
Hours worked per day (working mothers)	8.3	7.8
Hours worked per day (working fathers)	8.7	9.1

Note: Mean values for parents reporting at least 1 min of caregiving time per day. Individual categories on educational attainment and job type may not sum to 100% due to rounding, but the age of children categories can sum to greater than 100% because the categories are not mutually exclusive.

Source: Author’s calculations of pooled data from the U.S. Department of Labor Bureau of Labor Statistics, American Time Use Survey, 2003-2009.

On average, working mothers spend more than 2 hr per day on caregiving activities regardless of whether they work in federal service or for private sector firms. They spend about 15% more per day on caregiving activities than men do regardless of sector, as well. Federal workers are a bit older than private sector workers by nearly 3 years, and they tend to be more highly educated than private sector workers as well. This reflects the relative absence of entry-level jobs in federal government, compared to the private sector (Mastracci & Thompson, 2009). The age difference between the two sectors may also explain differences in family characteristics: Compared to parents in private sector firms, parents in federal service are less likely to have infants and toddlers at home and more likely to have children in grade school and high school. Controlling for these characteristics, I run the model using ordinary least squares regression analysis to determine whether mothers in federal service spend less time on caregiving activities than do their counterparts in the private sector. The results are below.

Results and Discussion

Table 2 contains results from regression analysis. Because the dependent variable is in log form, coefficients are interpreted as the percent change in Y (caregiving time) due to a one-unit change in each independent variable X, holding other factors constant.

Table 2.

Effects of Worker Characteristics and Family Type on the Number of Minutes per Day Spent on Caregiving Activities.

	Coefficient (% impact on Y)
Worker characteristics
Female working in federal government (FEM*FED)	−0.18^a
Working in federal government (FED)	0.08
Being female (FEM)	0.21^c
Age	0.01
Age squared	−0.00
Married	−0.02
Having an employed spouse/partner	0.11^c
Hours worked per day	−0.00^c
Year	−0.00
Age of children (compared to teen)
Presence of infant (age < 3)	0.49^c
Presence of toddler (age 3-5)	0.31^c
Presence of kid (age 6-12)	0.06^b
Number of children (age < 18)	0.03^c
Educational attainment (compared to high school or less)
Some postsecondary	0.01
Four-year degree	0.18^c
Some postgraduate	0.20^c
Job type (compared to clerical)
Professional (P)	0.05
Administrative (A)	−0.04
Technical (T)	0.08^a
Other (O)	0.00
Model characteristics
Number of observations	9,787
Model F (p value)	69.70 (0.00)^c
R-squared (Adj. R²)	0.12 (0.12)
Constant	4.03^c

Statistically different from zero at the 90% level of significance (less than 1 in 10 chance of obtaining this result mistakenly).

95% level.

99% level.

The effect of being female in federal service—the coefficient on FEM*FED—is about a 20% reduction (–0.18) in caregiving time, all other things being equal. Based on an average of 2 hr of caregiving time spent per day, a working mother in federal service spends about 20 min each day less on primary caregiving, compared to working mothers in the private sector and men in both sectors. This effect is statistically different from zero at the 90% level, thereby confirming the hypothesis. The effect of working in federal government alone (FED) is not statistically different from zero. Being female (FEM) increases time spent on caregiving activities by 21% and this result is statistically significant, which is consistent with expectations. Marital status alone does not affect time spent on caregiving activities, but the employment status of spouses for those who are married does. Having an employed spouse increases one’s own caregiving burden by 11%. Age of children affects caregiving time in the hypothesized manner: Compared to having teenagers at home, having an infant at home increases time spent by 49%, having a toddler increases time spent by 32%, and having a school-age child increases that amount by 6%. All results are statistically significant. The effects of educational attainment do not conform to expectations: Compared to respondents with no more than a high school education, workers with college degrees and at least some postgraduate education spend 18% and 20% more time, respectively, on caregiving activities. Likewise, job characteristics do not affect caregiving time spent in the ways that were hypothesized: Only one is statistically different from zero and all magnitudes are low. In sum, the hypothesis is confirmed: Women workers in federal service spend less time on caregiving activities compared to workers in the private sector who do not have such access, even when controlling for alternative explanations of time use, including marital status, age and number of children, and educational attainment.

I ran a number of variations of the model presented here, as well. Models including usual hours worked, full-time/part-time status, and earnings did not change the result. Contrary to expectations (Arrighi & Maume, 2000) variables accounting for earnings and hours were not statistically significant and did not contribute to the explanatory power of the model, including variables that capture the ratio of respondent-to-spouse earnings and hours. The simple dichotomous variable indicating spouse employment status had more explanatory power than the more complex ratios of hours and earnings, without problems of multicollinearity that arose when all of these variables were included in one model. The results confirm the hypothesis that women working in federal agencies spend less time on primary caregiving activities compared to their private sector counterparts. This is an important result to federal agency recruitment and employee retention in the federal government, which has sought to establish a reputation as a model employer through the development and implementation of family-friendly workplace programs and a culture that supports overall work–life balance.

Conclusion

Working mothers in federal service spend about 20 min per day less on caregiving activities compared to those in the private sector, regardless of job type, educational attainment, marital status, family composition, or whether or not their spouses are working. Table 1 shows that working mothers in federal service work more hours per day, on average, than do working mothers in private sector jobs (8.3 hr per day compared to 7.8 hr). What is more, the coefficient on work time in Table 2 indicates that the amount of time spent at work has no real impact on caregiving time. The coefficient is statistically different from zero, but practically no different from zero (–0.0007941), indicating a zero-percent decrease in caregiving time due to an increase in time spent at work. Does this indicate that the availability of family-friendly workplace programs offered by OPM allow women in federal service to spend more time at work? And if so, is this true work–life balance?

Further research in this area should attempt to link the reduction in time spent on caregiving activities to actual participation in family-friendly workplace programs. Future research could also use data from multiple sources to avoid problems inherent in self-reports of time use (Beam, 2012; Mikelson, 2008); however, Greenhaus and Powell (2006) “believe that there is value in continuing to assess work–family enrichment through self-reports” (p. 86). I remain confident in these results because the ATUS data on caregiving time spent by women are arguably in the opposite direction one might expect from biased reporting. Gender norms might bias women’s overreporting of time spent on caregiving, not to under report (Guendouzi, 2006; Mikelson, 2008). Finally, further research could consider whether less caregiving time is actually the desired result of participation in family-friendly workplace policies. Perhaps women and men would like to spend more time on caregiving activities and use workplace programs to do so; as Sayer (2005, p. 298) speculates, “It is possible that both women and men would choose to spend less time in paid work and more time in family work.” Perhaps greater balance means more caregiving time, not less (Greenhaus & Powell, 2006). Further research to address these open questions could employ Grzywacz and Carlson’s (2007) framework for work–family balance that is rooted in role theory.

Footnotes

Appendix A

Table A4.

Occupations Comprising Half of Total Employment in the Private Sector.

Census code	Occupation title	Percent employed
10	Chief executives	0.95
*20	General and operations managers	0.79
50	Marketing and sales managers	0.85
*120	Financial managers	0.95
*430	Managers, all other	1.67
*800	Accountants and auditors	1.22
*1020	Computer software engineers	0.78
*3130	Registered nurses	1.58
3600	Nursing, psych, and home health aides	1.44
*3920	Security guards	0.69
4020	Cooks	1.85
4030	Food preparation workers	0.78
4110	Waiters and waitresses	2.14
*4220	Janitors and building cleaners	1.45
4230	Maids and housekeeping cleaners	1.31
4250	Grounds maintenance workers	0.88
4600	Child care workers	0.69
4700	Supervisors and managers of retail sales workers	2.50
4710	Supervisors and managers of nonretail sales workers	0.79
4720	Cashiers	3.29
4760	Retail salespersons	3.18
4850	Sales representatives, wholesale and manufacturing	1.17
*5000	Supervisors and managers of office workers	1.20
5120	Bookkeeping, accounting, and auditing clerks	1.07
*5240	Customer service representatives	1.82
5400	Receptionists and information clerks	1.14
5620	Stock clerks and order fillers	1.43
*5700	Secretaries and administrative assistants	2.06
*5860	Office clerks, general	0.69
6050	Misc. agricultural workers	0.81
6230	Carpenters	0.97
6260	Construction laborers	1.27
6350	Electricians	0.70
7750	Misc. assemblers and fabricators	0.95
8960	Production workers, all other	0.82
9130	Driver/sales workers and truck drivers	2.90
9620	Laborers and freight, stock, and material movers	1.84

Source: Author’s calculations of Current Population Survey merged outgoing rotation group data.

Twelve occupations also among the top half of private sector employment (see also Table A3).

Appendix B

Appendix C

Raw data for the dependent variable—minutes per day of caregiving time—are highly skewed, as Figure 1C illustrates. Ordinary least squares regression fails to generate the best linear unbiased estimator when the dependent variable is not normally distributed. A log transformation corrects for the skewed distribution (see Figure 2C) but because the natural log of zero is nonexistent, all observations reporting zero caregiving time are lost. I decided that losing those observations would not be problematic as the effect of time-saving, family-friendly workplace policies would be small if no time were spent on caregiving in the first place.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

Author Biography

Sharon H. Mastracci is an associate professor at the University of Illinois at Chicago, Department of Public Administration. She conducts research on the public sector workplace and employment policy. An earlier version of this paper was presented before the members of the Project for Equity, Representation, and Governance (PERG) working group at the Texas A&M University Department of Political Science. The author is grateful for their feedback, as well as that of three anonymous reviewers.

References

Abraham

K. G.

Flood

S. M.

Sobek

Thorn

(2008). ATUS data extract system. College Park, MD: Maryland Population Research Center, University of Maryland.

Arrighi

B. A.

Maume

D. J.

(2000). Workplace subordination and men’s avoidance of housework. Journal of Family Issues, 21(4), 464-487.

Barnett

R. C.

(1999). A new work-life model for the twenty-first century. The Annals of the American Academy of Political and Social Sciences, 562, 143-158.

Beam

(2012). The problem with survey research. New Brunswick, NJ: Transaction Publishers.

Berry

(2011, February). Status of telework in the federal government: A message from the director. Retrieved July 21, 2011, from http://www.telework.gov/Reports_and_Studies/Annual_Reports/2010teleworkreport.pdf

Bowman

J. S.

(2010). The success of failure: The paradox of performance pay. Review of Public Personnel Management, 30(1), 70-88.

Braun

Lewin-Epstein

Stier

Baumgartner

M. K.

(2008). Perceived equity in the gendered division of household labor. Journal of Marriage and Family, 70, 1145-1156.

Bruce

Reed

(1994). Preparing supervisors for the future work force: The dual-income couple and the work-family dichotomy. Public Administration Review, 54(1), 36-43.

Cinamon

R. G.

Rich

(2002). Gender differences in the importance of work and family roles: Implications for work/family conflict. Sex Roles, 47(11/12), 531-541.

10.

Coltrane

(2000). Research on household labor: Modeling and measuring the social embeddedness of routine family work. Journal of Marriage and the Family, 62, 1208-1233.

11.

Ezra

Deckman

(1996). Balancing work and family responsibilities: Flextime and child care in the federal government. Public Administration Review, 56(2), 174-179.

12.

Facer

R. L.

Wadsworth

L. L.

(2008). Alternative work schedules and work/family balance. Review of Public Personnel Administration, 28(2), 166-177.

13.

Greenhaus

J. H.

Powell

G. N.

(2006). When work and family are allies: A theory of work-family enrichment. Academy of Management Review, 31(1), 72-92.

14.

Grzywacz

J. G.

Carlson

D. S.

(2007). Conceptualizing work-family balance: Implications for practice and research. Advances in Developing Human Resources, 9(4), 455-471.

15.

Guendouzi

(2006). The guilt thing: Balancing domestic and professional roles. Journal of Marriage and Family, 68, 901-909.

16.

Hegtvedt

K. A.

Clay-Warner

Ferrigno

E. D.

(2002). Reactions to injustice: Factors affecting workers’ resentment toward family-friendly policies. Social Psychology Quarterly, 65(4), 386-400.

17.

Helms-Erikson

(2001). Marital quality ten years after the transition to parenthood: Implications of the timing of parenthood and the division of housework. Journal of Marriage and Family, 63(4), 1099-1110.

18.

Hochschild

A. R.

(1997). The time bind: When work becomes home and home becomes work. New York, NY: Metropolitan Books.

19.

Hoyman

Duer

(2004). A typology of workplace policies: Worker friendly versus family friendly? Review of Public Personnel Administration, 24(2), 113-132.

20.

Jang

S. J.

(2009). The relationships of flexible work schedules, workplace support, supervisory support, work-life balance, and the well-being of working parents. Journal of Social Service Research, 35(2), 93-104.

21.

Keene

J. R.

Quadagno

(2004). Predictors of perceived work-family balance: Gender difference of gender similarity? Sociological Perspectives, 47(1), 1-23.

22.

Kerr

Miller

Reid

(2002). Sex-based occupational segregation in U.S. state bureaucracies. Public Administration Review, 62(4), 412-423.

23.

Lewis

(1997). “Family friendly” employment policies: A route to changing organizational culture or playing about at the margins? Gender, Work, and Organizations, 4(1), 13-23.

24.

Mannino

C. A.

Deutsch

F. M.

(2007). Changing the division of household labor: A negotiated process between partners. Sex Roles, 56(3), 309-324.

25.

Marler

J. H.

Moen

(2005). Alternative employment arrangements: A gender perspective. Sex Roles, 52(5/6), 337-349.

26.

Mastracci

S. H.

(2010). (Un)intended consequences of family-friendly workplace policies. In d’Agostino

Levine

(Eds.), Women in public administration: Theory and practice (pp. 95–110). Sudbury, MA: Jones and Bartlett.

27.

Mastracci

S. H.

Thompson

J. R.

(2009). Who are the contingent workers in federal government? American Review of Public Administration, 39(2), 352-373.

28.

Maume

D. J.

(2006). Gender differences in restricting work efforts because of family responsibilities. Journal of Marriage and Family, 68, 859-869.

29.

McCurdy

A. H.

Newman

M. A.

Lovrich

N. P.

(2002). Family-friendly workplace policy adoption in general and special purpose local governments. Review of Public Personnel Administration, 22(1), 27-51.

30.

Mikelson

K. S.

(2008). He said, she said: Comparing mother and father reports of father involvement. Journal of Marriage and Family, 70, 613-624.

31.

O’Toole

D. E.

Churchill

J. R.

(1982). Implementing pay-for-performance: Initial experiences. Review of Public Personnel Administration, 2(3), 13-28.

32.

Parker

A. M.

(2010, February 25). Uncle Sam kicks off new federal employee survey. Government executive. Retrieved July 21, 2011, from http://govexec.com/story_page_pf.cfm?articleid=44655.htm

33.

Perry

J. L.

(2010). A strategic agenda for public human resource management research. Review of Public Personnel Administration, 30(1), 20-43.

34.

Roosevelt

(1907, December 3). State of the Union Message. Retrieved July 21, 2011, from http://www.theodore-roosevelt.com/images/research/speeches/sotu7.pdf

35.

Saltzstein

A. L.

Ting

Saltzstein

G. H.

(2001). Work-family balance and job satisfaction: The impact of family-friendly policies on attitudes of federal government employees. Public Administration Review, 61(4), 452-467.

36.

Sayer

L. C.

(2005). Gender, time, and inequality: Trends in women’s and men’s paid work, unpaid work, and free time. Social Forces, 84(1), 285-303.

37.

Shockley

K. M.

Allen

T. D.

(2007). When flexibility helps: Another look at the availability of flexible work arrangements and work-family conflict. Journal of Vocational Behavior,71(4), 479-493.

38.

Sullivan

Gershuny

(2001). Cross-national changes in time use: Some sociological (hi)stories re-examined. British Journal of Sociology, 52(2), 331-347.

39.

Telework Enhancement Act of 2010. (2010, January 5). HR 1722, as enacted. Retrieved July 21, 2011, from http://www.gpo.gov/fdsys/pkg/BILLS-111hr1722enr/pdf/BILLS-111hr1722enr.pdf

40.

Tower

L. E.

Alkadry

M. G.

(2008). The social costs of career success for women. Review of Public Personnel Administration, 28(2), 144-165.

41.

U.S. Department of Labor, Bureau of Labor Statistics. (2008a, May 8). Married parents’ use of time, 2003-2006. Retrieved July 21, 2011, from http://www.bls.gov/news.release/pdf/atus2.pdf

42.

U.S. Department of Labor, Bureau of Labor Statistics. (2008b, June). American Time Use Survey questionnaire. Retrieved October 4, 2010, from http://www.bls.gov/tus/

43.

U.S. Office of Personnel Management. (2011, February). Status of telework in the federal government. Retrieved July 21, 2011, from http://www.telework.gov/Reports_and_Studies/Annual_Reports/2010teleworkreport.pdf

44.

Wadsworth

L. L.

Owens

B. P.

(2007). The effects of social support on work-family enhancement and work-family conflict in the public sector. Public Administration Review, 67(1), 75-86.

45.

Wharton

C. S.

(1994). Finding time for the “second shift”: The impact of flexible work schedules on women’s double days. Gender & Society, 8(2), 189-205.

46.

Wood

(1999). Family-friendly management: Testing the various perspectives. National Institute Economic Review,168(2), 99-116.