STRATEGY ADDRESSING THE WORK-LIFE INTERFACE: STRATEGIC IMPLICATIONS FOR SMALL BUSINESS Sarah W. Jacobson Harriette S. McCaul North Dakota State University ABSTRACT Although some research has assessed the involvement of corporate America in work-life policies and programs, little is known about the involvement ofsmall businesses in this area. The purpose of this study was to learn more about how small businesses are responding to work-lifeissues and to examine the relationship ofvarious organi aiional characteristics to the implementation of work-life policies and programs. The results indicate that small businesses have tespondedto work life issues and concerns in a piecemeal manner and have been relatively non-responsive overall, even though participating would carry linle cost. Traditional economic benefits (e.g., health insurance) were most likely to be offered, even li mired care for dependents (eg., resource and referral services) least likely. Further, fewer than 10% of the firms have conducted any kind offormal assessment of worklife needs and issues. (Ye suggest that a lack of strategic attention by small businesses to the work-life interface is short sightedin view of quality andproductivity issues, changing demographics and the recruitment and retention of well-qualified employees. INTRODUCTION Organizational concern with work-life issues is a recent phenomenon, based largely on dramatic changes in the age and gender composition of the workforce and evidence that historical sex-role definitions are undergoing transformation (Lobel, 1991; Spence, Deaux & Helmreich, 1985). Recognition of these shifls has led to the suggestion that, in order to remain competitive,organizationsmay need to become more accommodatingof the needs of workers who are attempting to juggle the demands of both work and increasingly complicated private lives. However, although there has been considerable popular rhetoric concerning the importance of providing work environments that address private life concerns, relatively few companies have actually moved in this direction (Galinsky et al., 1991; Morgan & Milliken, 1992). Those firms that have decided to address issues faced by employees at the interface between work and private life, have implemented a variety of policies and programs ranging from comprehensive (e.g., flexible scheduling in terms of time and/or work place, child care centers, job sharing and parenting programs) to more limited offerings (e.g., tuition reimbursement, flexible spending accounts and resource and referral services). In recent years, some research has been conducted to assess the extent of work-life responsiveness in corporate America. However, little is known about the involvement of small businesses in this area. 29 Our primary goal in the research we describe here was to consider work-life support in organizations generally overlooked in previous research - small businesses. Since 80% of working Americans are employed by companies with fewer than one thousand employees(U.S. Bureau of the Census, unt Business Patterns 19 I) work-life issues experienced by managers and employees in this group represents a substantial concern. Moreover, the Small Business Administrationhas predicted that by the year 2000 more than hal fof workers in small business will be women (Bureau of National Affairs, 1990), which may necessitate an even greater strategic concern with the work-life interface for small business. LITERATURE REVIEW A recent survey by the Society for Human Resource Management (1992) reported that most companies cite "expense" as the major obstacle to adopting work-life initiatives. Also, companies may have been slow to take action because of the limited evidence on the economic benefit of doing so (Vanderkolk & Young, 1991). Recent research, however, demonstrates that work-life programsand policies build employee loyalty and commitment, improve retention, and may reduce absenteeism and turnover (c.f., Galinsky et al., 1991; Ho(Terth, Bayfield, Diech, & Holcomb, 1991; Vanderkolk & Young, 1991). Work-family support can also enhance productivity (Friedman, 1991; Katz & Piotrkowski, 1983). Small businesses, in particular, have been encouraged to otTer more work-life programs in order to compete with larger companies for well-qualified employees (Bureau of National A1Tairs, 1990). Finally, even if it is difficult to definitivelyshow tangible benefits of work-life policies and programs, organizations have been encouraged to consider the strategic cost of not doing so in terms of potentially increased turnover, absenteeism and lowered productivity (Grover & Crooker, 1994; Lambert, 1990). An additional research goal in the work-life domain has been to describe differences in the degree to which selected organizations are addressing concerns in this area. Based on 0 188 u 31000000 0, 0 18 30 dff 0 manufacturingcategories,the Families and Work Institute was the first to benchmark "family friendliness"'nd to catalogue the experience of major U.S. corporations with numerous policies falling under the work-family umbrella (Galinsky et al., 1991). Although the study provided a better understanding of practices in larger organizations, no elTort was made to describe the relationship between various organizational characteristics and work-family support. Moreover, the study neglected smaller companies altogether. Morgan and Milliken (1992) sought to address the first deficiency by first scoring companies on three categories of work and family policies and benefits —family leave policies, flexible work options, and dependent care benefits. All policies and benefits were assigned equal importance, although additional points were allocated to more generous versions of the ' lere we will use the term "work-life interface" to describe the juncture between work and private life experienced over the complctc life cycle by all employees. This term is broader and more inclusive than "family friendliness" as it has been generally described. This is not to say that "family friendly" programs and policies are not important but, rather, that they constitute just one area of interest in a broader domain of concern. 30 policies(e.g., a flexible work option available to afl employees versus one available only to a few). Scores on the three categories were summed to produce an overall "work-family responsiveness" score. The researchers then examined the influence of various external and internal organizational factors on work-family responsiveness. The most important factors included company size, industry, geographic region of the country, and degree of managerial attention given to work-family issues. For example,companies with more than 500 employees, companies in the health care, finance, insurance and real estate businesses, and companies located in the Northwest, were found to be most family responsive. The study also concluded that companies that actively assess the work-family needs of their employees are more generous in terms of work-family programs and policies than those that do not. These results, however, were based on a relatively low response (175 usable surveys from a sample of one thousand, or 18'/0). Further, company size was framed as a dichotomous variable(fewer than 500 employees versus more than 500 employees), thus possibly masking importantdifferences within each group. Finally, the study included only a limited number of policies and benelits (16 in all) specifically relating to issues having to do with families, omitting more traditional beneflts such as health and life insurance, programs and policies benefitting employeeswithout dependents, and aspects of organizational climate and culture. In this study, we expanded on previous research in two ways: (I) we considered work-life support in small businesses, and (2) we included a wider variety of work-life policies and programs as well as organizational characteristics in our research design. RESEARCH MODEL We conducted this research project in two phases. In Phase I, we analyzed results of telephone interviews with representatives of 403 randomly selected small and medium sized firms. In Phase 2, we analyzed the results of in-depth interviews with key informants in 17 firms. In the initial phase of the project, we first examined the influence of a series of macro- organizational factors: size, industry type, rural/urban location, perceived competitive strength, and independent/subsidiarystatus on the provision of various work-life policies and programs. Previous research has established a connection between organizational size, industry type and work-life support. Largercompanies have tended to be the leaders in the provision of family related support (Morgan & M i I liken, 1992), and certain types of industries (particularly banks and insurance companies) have implemented more generous work-life policies and programs (Morgan & Tucker, 1991). We were unable to locate studies that examined the influence of rural/urban location, perceived competitive strength, and independent/subsidiarystatus on the presence of work-life programs and policies, but these were determined to be particularly salient issues in small 31 business'. We hypothesized that companies in urban locations would be more likely to have adopted work-life policies and programs enabling them to compete more successfully for employees in a tighter labor market. We also hypothesized that organizations that perceive themselves to be doing as well as or better than their competitors would be more likely to expend organizational resources for work-life initiatives than those organizations that see themselves as struggling for survival. Finally, we hypothesized that subsidiaries of larger organizations would be more supportive of work life issues than would independent companies because subsidiaries would likely imitate their larger parent organizations in terms of benefits, including work-life initiatives. We also examined the impact of micro-organizational factors, including the percentage of(a) female employees,(b) professional, managerial,and technical employees,(c) employees under age 40, and (d) part-time employees. Auerbach (1988) suggested that companies with a higher percentage of female employees will be more "family friendly," because work-life issues may be more salient to female employees. Surprisingly, Morgan and Milliken (1992) found no evidence to support this suggestion when other factors, such as organizational size and industry type, were controlled. Given the higher level of competition for upper-level employees, particularly in smaller companies and in rural communities, we predicted that companies employing a higher percentage of professional, managerial and technical employees would be more likely to offer work-life support for recruitment and retention purposes. We also hypothesized that companies with a higher percentage of employees in childbearing and child-rearingyears, no matter their gender composition, would be more likely to be aware of the potential influence of work-life policies and practices on recruitment, retention, and productivity and would, therefore, be more supportive of private life issues generally. Finally, we hypothesizedthat companiesemploying a higher percentage of part time employees do so in part to avoid paying for traditional fringe benefits, and therefore, would be less likely to provide work-life support. METHODOLOGY The Surve In trument A revised version of the Family Friendly Index (Galinsky et al., 1991)was developed that included additional dimensions of work-life support identified by a panel of human resource managers. The instrument was pretested with 35 human resource practitioners. Questions assessed organizational characteristics and work-life support in five categories: Ilexibility, leave, dependent care, organizational climate, and economic benefits. We measured the degree 'Our researchwasconductedat the land grant institutionsin four Midwestern states -Nonh and South Dakota, lowe and Nebraska. Family Science Extension stafl'at each University were a part of the research team. Based on extension contact with businesses in the 1'our states and requests by managers for support, we learned that location, perccivcd competitive strength and independent/subsidiarystatus were key factors in determining the degree to which a company is concerned with, and involved in, work-life issues. Therefore there was solid, though anecdotal, reason to believe that these factors should be included in the study. While other issues might also be of interest (e.g., age of the company, growth rate, gender of the owner etc.) we do not consider them here. 32 of work-lifesupport based upon two factors used in the studies by Galinsky et al. (1991)and Morgan and Milliken (1992): (I) whether or not a given benefit, program or policy was present in an organization, and (2) whether the benefit, program or policy was available to some, most, or all employees. Points were assigned for each of the policies and practices currently being offered in each organizationon the following basis: No=0 points, Yes=I point, and if available to all=2 points. Based on input from the human resource managers, these work-life support policies and practices were then grouped in five categories: I) Flexibility (e.g., job sharing, flextime, flexplace). 2) Leave policies (e.g., parental leave with or without pay, bereavement leave). 3) Dependent care (e.g., on-site or off-site child care centers, child care vouchers, child care referrals, elder care). 4) Organizationalclimate(eg.,EmployeeAssistanceProgram,supervisorytraining community agency referrals, parenting and lifestyle workshops, tuition reimbursement, company-sponsored family events). 5) Traditional economic benefits(e g., health insurance, life insurance, sick leave). Our scoring system, as did the systems used by Galinsky et al. (1991), and Morgan and Mifliken (1992), summed the five submeasures to yield an overall work-life support score. ~Th S In Phase I of the project, a random sample of 525 businesses, from four Midwestern states, stratified by size and SIC code was obtained from a national mailing list (American Business Lists). Letters soliciting participation were mailed to the companies three weeks in advance. A team of three telephone interviewers, trained by the research team to insure consistency, collected the data. Ultimately, either the owner/managerof the firm or a designee in 403 (72%) of the firms contacted were interviewed. (A detailed description of the Phase I firms and their key organizationalcharacteristicsappears in Appendix 1.) Reflecting business patterns in the Midwest, the Phase I sample included a large number of very small businesses (72% employ fewer than 50). The largest category in terms of industry type was wholesale/retail trade (42%). In Phase 2, an in-depth interview was conducted with a convenience sample of 17 respondents, selected from within the various size and SIC code categories, to provide a better understanding of alternative approaches to work/private life issues. Three of these firms employed 20 - 49 people, five employed 50 - 99, six employed 100 - 249 and three employed 250 - 499. Nine of the companies were in manufacturing, two in health services, two in transportation related businesses and four in business and professional services. Control Variables The measure for company size used in Phase I sample construction was number of employees, as indicated above. The companies in the sample were grouped into four categories based on SIC code to reflect similarity of industry and to insure adequate cell sizes. In a preliminary analysis, however, it was determined that the only significant industry 33 differences were between the Finance, Insurance, Real Estate, Business and Professional Services category and all others (p & .001). Thus, for this analysis. these data were collapsed into a dichotomous variable ( I = Business and Professional Services, 0 = other). To measure competitive strength, respondents indicated if they perceived their company as doing better, about the same, or worse than their competitors. Organizational type became a categorical variable based on answers to a question as to whether the firm was independently owned (=0) or the subsidiary or satellite of another company that determines or influences policy(=l). The rural/urbanclassificationwas determined by the researchers based on a U S. census definition (I = business located in a community with a population greater than 2500, 0 = other). Questions concerning micro-organizational factors were also asked in the interviews. Participants reported the percentage of their employees in the following categories: female, less than 40 years of age, part-time, and professional, technical, or managerial. lldl h I Informants in Phase 2 were either the owner/manager in the smallest firms, a general manager or administrator, or (in the three largest firms) the human resource manager. Interviews lasted approximately one hour. Initially, participants were asked the series of standard questions that had comprised the telephone interview of the larger sample. Then, several additional open-ended questions were posed regarding the respondent's overall satisfaction with the manner in which existing work-life programs were working out, the perceived effect of these programs on turnover and/or absenteeism, the manner in which decisions had been made concerning establishmentof work-life programs, means the firm used to evaluate them once the programs had been implemented,and special issues faced by the firm in dealing with work-life issues. RESULTS AND DISCUSSION The average score for overall work-life support in the Phase I firms was 37.57 (SD= I 2 46) out of a possible 102 points. Scores ranged from I I to 83. This relatively low average score for work-life support is consistent with prior research (Morgan & Milliken, 1992). Multiple regression analysis was used to examine the isolated effect of single variables on total work-life support while controlling for the elfect of all other variables. We used a hierarchial analysis by category of variable which allowed us to test the contribution of the two sets of variables (macro-and micro organizational factors) to the explained variance in overal! work-life support aRer controlling for organizational size and type. Table I shows the correlationsamong independent variables in Phase I of the study, and Table 2 shows the results of the regression analysis. These results indicate that company size and industry type explain about 15h/dh of the variance in work-life responsiveness (p&.001; p&.001). Our results are consistent with previous studies (Gal insky et al., 1991; Morgan and Milliken, 1992) in that larger companies were more responsive. Firms operating in the 34 business and financial services category and firms that are satellites or subsidiaries of larger companies were also more supportive of private life concerns in general (p&.001). However, there were no significant dilTerences in overall work-life support between those companies doing business in urban versus rural locations(only 6% of the sample was designated as rural). The only micro-organizational factor that proved significant was the percentage of professional, technical, and managerial employees. This factor was positively related to work- life support (p&.01). None of the other micro factors (i.e., the percentages of women, employees under age 40, and part-time workers) was significant. In general, the respondents were most likely to offer traditional economic benefits (e.g., 97% offered bereavement leave and 86% offered health insurance coverage for full-time employees) and programs without a direct cost to the organization (e.g., 96% offer unpaid leave for the care of sick children but only 33% provide paid leave for such care). Most respondents also reported that they provide flexibility in a variety of ways. For example, 95% allowed employees to take time off to attend school functions,97% allowed for flexible vacation scheduling, and 85% provided cross training and flexiblejob design. Further, there was a positive relationship between flexibility and the percentage of female employees. Policies relating to flexibility were usually implemented case-by-case, rather than as formal policy. Support for dependent care was rarely offered, even if little organizational cost was involved. For example, only 15% of the companies offered resource and referral services concerningchild care options and just 29% offered flexible spending accounts for child care. With regard to organizational climate, the only policy offered by more than half of the firms was one allowing employees to make and accept personal phone calls at work. Finally, fewer than 10% of the firms reported that they have conducted any kind of formal assessment about work-life needs and issues. In the in-depth interviews conducted in Phase 2, several dominant themes emerged. First, most of the informants in the expanded interviews expressed general satisfaction with the eflicacy of their present work-life programs and policies. This was true whether the programs were generous or more limited. Informants also expressed the importance of customizing the program to the particular needs of the business and its employees. For example, the establishment of cross-functional teams has enabled one flexible manufacturing company to turn over scheduling of work time and paid leave to the employees themselves. Management believed that this action has encouraged the kind of empowerment and self-management required for the rapid turnaround of product perceived to be critical to the success of the company. In another example, a small nursing home located in an isolated rural community had experienced extremely high turnover among certified nursing assistants (most of whom were women) who were literally irreplaceable given the size of the existing labor pool. 35 Table I Correlation nal is: Ke Inde endent Variables and Total Score TOTAL PCTVYOAIEN PCTPROF PCTLT40 PCTPART ORGSIZE TOTAL 1.00000 PCTVYOialgiV 0.11693'.00000 PCTPROF 0.19640"' 01827 1.00000 PCTLT40 -0.10817'.12474'0.12775'" 1.00000 PCTPART -0.13917" 0 32092"'0.13376" 0 31815"' 00000 ORGSI/E 0.31951""'.04780 0.02288 -0.16500" -0.11561'.00000 KEY: Total= extent of organization's provision of programs and policies addressing work-life interface, pctwomen=percent women, pctprof=percent professional, pctlt40 = percent employees under age 40, pctpart=percent part-time employees, orgsize=number of employees in organization. Correlations with dummy variables for business type have not been included. «p&.05 ««p&.01 «««p&.001 36 Table 2 Addressin the Work-life Interface'ierarchical Re ression Results VARIABLE beta F R'or step Group I Control Variables A. Organizational Size 3.13 28.84ea* B. Industry Type 8.42 27.9*a* .149 Group 2 Organizational Variables A. Organizational health -2.13 3.23 B. Independent/Subsidiary 6.31 23.48~ev C. Rural/urban location -3.95 2.81 .066 Group 3 Company Demographics A. Percent women .040 3.27 B. Percent professional .82 7.61v~ C. Percent less than 40 - .011 .20 D. Percent part-time - .04 4.04* .034 NOTE: R'.25, F = 3.37" ep&.05, va p&.01, *vs p&.001 tgt In an earlier ANOVA, it was determined that the only significant distinction in industry type was between the business/financecategory and others(F=10 99, p& 0001).Therefore, a categorical variable was created for the regression analysis (bus/fin, other). Organizational Health = company says it is doing better, about the same, or worse than competitors, Organizational Type = independent/subsidiary or satellite. 37 The nursing home was able to significantly reduce turnover, however, after instituting a flextime program. The administrator of the nursing home indicated that management was able to offer this program at relatively low cost, with an excellent payback. Results of the in depth interviews indicate that it is in the strategic selection and implementation of work-life programs that dramatic returns can be realized. OBSERVATIONS AND RECOMMENDATIONS I licati ns for Practice Most small companies can do more to assist their employees in managing their work and private lif'e responsibilities. They are not taking advantage of the many opportunities alTorded by their size to be more flexible and responsive. This might be because, as other authors have noted, they view work-life issues as "competing for agenda space with an array of other issues from which companies must select the important, not the trivial, and then formulate responses to those deemed most relevant" (Morgan & Milliken, 1992, p. 239). In this sample with a preponderance of firms employing fewer than 250 people, issues other than work-life support may be viewed as more critical. In such settings, programs and policies relating to work-life issues have been adopted piecemeal as a given need is identified. It is our contention, reinforcedby the Phase 2 interviews, that such an approach bears substantial hidden costs to the organization. First, if small businesses do not conduct a need's assessment or a cost/benefit analysis, it is easy to miss the potentially significant impact that the implementation of a work-life support perspective may have on organizational success. Second, in responding piecemeal to individually identified needs rather than developing a coordinated effort, the overall eITect of work-life programs may be muted. Finally, based on a well-articulated needs assessment, money spent on some traditional programs offered by a substantial number of firms (e.g., life insurance) might, in some instances, be reallocated to programs and policies more consonant with the needs of the firm's employees. Therefore, a formal process of gathering information about employee work and private life needs should serve as the firm's starting point in determining overall policy in this area. In addition, evaluation against such measures as absenteeism,turnoverand/orproductivityshould be adopted after implementation. This may be even more important for smaller businesses where the margin for error in adopting such change is narrower than in larger firms with slack resources Many work-life programs and policies are relatively simple to establish and carry a low cost, while at the same time encourage higher productivity. Yet, based on results of this and other research projects, it is clear that a majority of small to medium sized firms are not engaged in these activities. It would be a foolish generalization, however, to suggest that all small businesses should be engaged in providing comprehensive work-life programs and policies. Resultsofourin-depthinterviewssuggestthatcustomizingwork-lifeprogramsyields the greatest return. In Table 3 we identify four business types and outline specific strategic justification for choosing to address the work-life interface in each case. 38 Table 3 When Does it ake ense for mall Businesses to Address the Work-life Interfacev ~MA UFACTURIN PROFESSIONAL SERVICES To create a culture consistant with R To compete for talented employees, high quality and empowerment particularly professionals, in this particularly in flexible manufacturing sector where work-life programs and environments. policies are more common and otten relatively generous. To reallocate costs of traditional benefits to more closely meet the needs of a majority of today's workers as well as organizational objectives. To recruit and retain female employees. To attract employees in low-wage, unskilled operations. FAST FOOD AND RETAIL NON-PR FESSIONAL SERVICES To set the firm apart as unique and as R To increase commitment and reduce the "employer of choice" in this arena. absenteeism and turnover. To create an image of employee support in order to enhance recruitment in low-wage unskilled positions. for example, although firms hiring professiomls may need to offer more work-life support in order to attract and retain talented employees, these programs make little sense in most retail and fast food settings, with low wages and many part-time and/or contingent workers. It is through careful strategic analysis, therefore, that firms should measure the costs and possible benefits before embarking on this kind of activity. Im lication for Future Researc It is important to note the limitationsof this research project. One limitation is the reliance on self-reports from the owner/managers of the firms in both the initial survey and the in-depth interviews. We did not attempt to verify, through other means, the existence of the programs and policies described. Further, this study did not attempt to measure tangible effects of differing configurations of work-life programs and policies; it simply documents whether or 39 not they are present. Finally, our scoring system, as in previous studies, did not attempt to discriminate between difTerent types of work-life programs, in terms of cost or scope, in measuring overall responsiveness. For example, the same weight was assigned to the presence of an on-site child care center as to a child care referral service. The former seems to represent a much higher level of commitment to providing work-life support by a given employer, but this difference is not reflected in our scoring scheme. Further research is needed to refine the measurement of work-life responsiveness, assigning a greater value to more costly initiatives. Also, in response to the apparent need of practitione rs for concrete evidence of the effects of work-life programs and policies, future research should focus on measurement of tangible improvements in organizational performance resulting from their implementation, as well as more intangible factors such as reduced stress and burnout and higher levels of employee commitment. 40 APPENDIX I ORGANIZATIONAL CHARACTERISTICS OF COMPANIES IN PHASE I SAMPLE. MACRO Size 20 - 49 employees 72% 50 - 99 employees 13% 100 - 249 employees 8% 250+ employees 8% Industry Type Wholesale/Retail Trade 42% Mining, agriculture, forest, fishing, construction, manufacturing, transportation, communication 25% Finance, insurance, real estate, business gt professional services 16% Health, education, other services 17% Organizational Type Independently Owned 74% Subsidiary 26% Organizational Health Doing better than competitors 64% Doing same as competitors 36% MICRO % Female 0-25 35% 26- 50 28% 51-75 21% 76- 100 16% % Employees under age 40 0-25 8% 26-50 24% 51-75 28% 76 - 100 40% 41 '/o Part-time Employees 0-25 59'/o 26 - 50 16'lo 51-75 13'/o 76- 100 13'/o '/o Professional, administrative or managerial 0-25 80'/o 26-50 15'/o 51-75 2'lo 76- 100 3'lo 42 REFERENCES Auerbach, J.D. (1988). In the business of child core: Employer initiatives and working women. New York: Praeger. Bureau of National Affairs, Inc. (1990). Work and Family Programs; A Growing Benefit for Small Companies, in The BNA Special Report Series on Work and Family. Washington, D.C. (January). Friedman, D. (1991). Linking Work-family Issues to the Bottom Line, Report Number 962. New York: The Conference Board. Galinsky, E., Friedman, D., & Hernandez, C. (1991). The Corporate Reference Guide. New York: The Families and Work Institute. Hofferth, S. L., Bayfield, A., Deich, S., & Holcomb, P. (1991). The tVational Child Care Survey: l990. Washington, D.Cx The Urban Institute Press. Grover, S. L. & K. J. Crooker (1995) Who Appreciates Family Responsive Human Resource Policies: The Impact of Family-Friendly Policies on the Organizational Attachment of Parents and Non-parents. Personnel Psychology, 48(2), 271-288. Katz, M. 8c C. Piotrkowski (1983). Correlates of family role strain among employed black women. Family Relations, 32, 331-339. Lambert, S. L. (1990). Processes Linking Work and Family: A Critical Review and Research Agenda. Human Relations, 43, 239-257. Lobel, S. A. (1991).Allocation of Investment in Work and Family Roles: Alternative Theories and Implications for Research. Academy ofManagement Journal, l6, 507-521. Morgan, H. & F. Milliken (1992). Keys to Action: Understanding Differences in Organizations'esponsiveness to Work and Family Issues. Human Resource Mnnagement, 3I, 227-248. Morgan, H. & Tucker, K. (1991). Companies that care: The most family friendly companies in America, what they offer ond how they got that way. New York: Simon & Schuster. Society of Human Resource Management. (1992) Work and family survey report. Alexandria, Va. Spence, J. T., Deaux, K. & Helmreich, R. L. (1985). Sex roles in contemporary society. In Handbook of Social Psychology. eds. G. Lindzey & E. Aronson New York: Random House, 149-178. U.S. Bureau of the Census (1991). County Business Patterns, annual. U S. Departmentof Commerce(1993). Statistical Abstract ofthe United States . Washington, DC: Bureau of Census. Vanderkolk, B. S. & A. A. Young (1991). The Work and Family Revolution: How Companies Can Keep Employees Happy ond Business Profitable. New York: Fact File, Inc. 43