https://cpes.org.uk/om/files/original/9a32a668d2d63f14ee2ad8b0854b7608.pdf c24f0a92d04ee69f288b9f58d847df13 Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource International Contributions to Labour Studies 3 Text A resource consisting primarily of words for reading. Examples include books, letters, dissertations, poems, newspapers, articles, archives of mailing lists. Note that facsimiles or images of texts are still of the genre Text. Text Any textual data included in the document International Contributions to Labour Studies (1993) 3, 91-100 Family structure, flexible employment, and labor market segmentation: evidence from a study ofthe temporary help industry June Lapidus* Introduction Two-thirds of all employees in the temporary help services (THS) industry in the United States are women. This paper addresses the question of whether or not the over-representation of women in this industry reflects a preference for flexible employment on the part of women workers. Asking this question provides an opportunity to explore the larger question of how gender operates in the labor market. THS industry observers can be roughly categorized according to whether they emphasize supply- or demand-based phenomena. Supply-based analyses accept as virtually axiomatic that rising female labor force participation rates have created a supply of workers faced with the competing demands of market and family labor. Industry observers such as Wayne Howe (1986) and Thomas Plewes (1988) of the Bureau ofLabor Statistics argue that since women with family responsibilities need flexible employment in order to balance these competing demands, women selfselect into the THS industry. The view expressed by Howe and Plewes is part of a more general literature in which wage and occupation differences between men and women are attributed to labor supply characteristics. Mincer and Polachek (1974) and Polachek (1981) argue that women's human capital investment decisions are heavily influenced by their expectations of intermittent labor force attachment over their life cycle. In this view, women self-select into occupations with wage structures that do not penalize intermittent attachment. These decisions are in turn responsible for the observed differences in occupational distribution for men and women. England (1982) refutes Polachek's argument by demonstrating that the characteristics of women in male-dominated occupations are similar to those of women in other occupations, suggesting that something other than human capital as determined by expectations of family formation is behind the sex-based occupational *Department of Economics, Roosevelt University, Chicago, Illinois, USA. I would like to thank Randy Albelda, Eileen Appelbaum, Rebecca Blank, Elaine McCrate, and Sam Rosenberg for comments on earlier drafts of this paper. Rick Edwards, Nancy Folbre, Patricia Greenfield and Jim Kindahl provided valuable assistance on the dissertation from which this research is drawn. 1052-9187/93/010091 + 10 $08.00/0 © 1993 Academic Press Limited 92 J. Lapidus differences. A parallel procedure can be used to test the hypothesized relationship between women's dual work and family responsibilities and the gender composition of the THS industry. If Howe and Plewes are correct, we would expect married women with children to be over-represented in the THS industry. They are not. Demand-based analyses of the temporary help industry separate industry growth from gender composition and treat the latter as exogenous. Mangum, Mayall and Nelson in a 1985 Industrial and Labor Relations Reveiw article situate the growth of temporary help in the development of dual internal labor markets. Appelbaum (1986) attributes industry growth to corporate downsizing and restructuring based on a ring and core model which creates (for the employer) a numerically flexible group ofworkers who can be easily dismissed in case of economic downturn. Golden and Appelbaum (1992) develop a model which tests demand, supply, and bargaining power as factors responsible for the 1980s boom in temporary employment. Their findings indicate that demand side and bargaining power forces drove the growth. The demand-based explanations do not try to explain the gender composition of the industry endogenously. Rather, they assume that forces in the economy govern the development of particular jobs and sectors, while exogenous power relations outside the sphere of the economy determine the allocation of jobs. This characterization holds for labor market segmentation theory (cf. Gordon et al., 1982), despite the fact that it developed in part in an attempt to understand why competitive forces did not act to eliminate race (and by extension gender) inequality in the labor market. Ultimately, by relying on a model in which the characteristics of the job rather than the worker determine labor market segments, it too treats gender as exogenous. Rubery (1989) and Beechey and Perkins (1987), in citing the work of the Cambridge Labour Studies Group, discuss the development ofa 'second generation' labor market segmentation theory. Their work is an attempt to go beyond the equation of supply-side phenomena with human capital and other self-selection models on the one hand and demand-based theories in which the characteristics of the worker are exogenous on the other. A study of the temporary help industry provides an ideal opportunity to extend the work of the Cambridge Group. After presenting empirical evidence suggesting that something other than choice based on family constraints is responsible for the gender character of the industry, I then inquire whether the growth of temporary help has been predicated upon the availability of a pool of low-wage women workers with few opportunities for more stable and renumerative employment. The temporary help industry The temporary help supply industry provides business enterprises with workers for temporary assignments. 1 The workers remain on the payroll of the THS firm. The industry is part of a larger class of employment relations in which there is an ex-ante understanding between employer and employee that the job is oflimited duration. 1 While THS employment shares some characteristics with part-time work, there are important differences. Almost three-fourths of female temporary help service employees report that they usually work 35 hours per week, and over two-thirds reported working 8 or more hours per day. Additionally, of those female THS employees who did work fewer than 35 hours, almost half did so for economic reasons (either slack work or they could only find part-time work). Family structure, flexible employment, labor market segmentation 93 This study focuses on the THS industry, a subset of what has come to be called contingent work, for several reasons. The particular form ofcontingent work offered by the THS industry is touted by the industry as ideal for women with family responsibilities. The industry affords almost daily flexibility, permitting the schedules of sick children and school calendars to be accommodated. The reduction in search costs for workers would appear to offer an advantage to mothers who may need to exit and re-enter the labor market frequently. Focusing this research on the THS industry rather than on the contingent labor market as a whole thus tests the supply-side theory on terms most favorable to it. The US Department of Commerce May 1985 Current Population Survey (CPS) asked respondents whether or not their salary is paid by a temporary help agency. It is the only national data base on the characteristics of THS employees; all data presented here are from that survey. However the CPS has important limitations, particularly with respect to sample size; only 160 female and 84 male THS employees. In addition, the structure of the CPS is such that wage data are only collected for the outgoing rotation which is only one-fourth of the households in the survey. Thus, wage data are only available for 34 women and 14 men. Table 1 depicts the demographic characteristics ofTHS employees by gender and compares male and female temporary workers to each other and to male and female non-temps, respectively. Two categories of variables are reported in Table 1: the human capital variables Age and Education and the family characteristics N child (number of children) and Marital (marital status). For women, the proportion married not separated is less for the sample of temporary workers than it is for other women in the labor force (44% for the former and 49% for the latter). While this difference is not statistically significant, the sign is the opposite of that predicted by supply-side theories. The number of children for THS and non-THS female employees is not statistically significant either, although the sign here is consistent with what would be expected by Plewes (1988), Howe (1986), and others. At first pass then, the family characteristics of female THS workers do not appear to differ from those of their non-temporary counterparts. The same is not true for men. The magnitude of the difference in marital status between male THS and non-THS workers (30% married not separated for the former and 73% for the latter) is much greater than the corresponding gap for women and is statistically significant at the 1% level. Male THS workers are also younger and less likely to have children than are other men, differences that are again significant at the 1% level. Occupational distribution Sixty-two percent of all female THS employees are in clerical occupations. Before turning to the empirical investigation of the supply-side hypothesis, the possibility that the over-representation ofwomen in the THS industry merely reflects women's concentration in clerical work must be considered. This argument hinges on there being something intrinsic to the nature of clerical work that makes it amenable to temporary help. For example, it is likely that jobs without firm-specific human capital requirements are more likely to be filled by temporary workers-or more 94 J. Lapidus Table 1. Demographic characteristics of THS employees and others THS EMPLOYEES AGE NCHILD EDUCAT MARITAL Malen=84 Femalen= 160 Men THSn=84 LFn=684 Women THSn=160 LFn=384 32 34 (1·20) 32·00 39·16 (4·75) 34·24 37·32 (2·65) 1-17 1·59 (2·56) 1·17 1·74 (3·69) 1·59 1·52 (1·38) 13·98 14·36 (1·23) 13·98 14·12 (0·65) 14·36 14·10 (1·30) 30% 44% (2·20) 30% 73% (8·14) 44% 49% (1·07) Data calculated from May 1985 Current Population Survey. Numbers in parentheses are t-statistics. LF=those in the labor force but not in the THS industry. generally, by workers without long-term attachments to the firm. If clerical work relies less on firm-specific skill than do other occupations, and women are more likely to be clerical workers, this may explain their over-representation in temporary work. There are two potential problems with reliance on the characteristics of clerical work to explain the gender composition of the THS industry. First, defining and categorizing skill is problematic.1 To conclude that the skill requirements of clerical work account for its dominance of THS employment it would be necessary to demonstrate that clerical work is superior to other occupations in this regard. I have found no evidence to support the conclusion that clerical occupations require less firm-specific human capital than, for example, welding. Yet in the case of predominantly male occupations such as welding, adjustment to short-term fluctuations in demand occur primarily through variations in overtime. The form of flexibility seems to depend, inter alia, on the gender of the worker.2 Second, time series studies to determine the impact of technological changes, in particular the introduction of microcomputers in the office, on the demand for or the skill level of clerical work, are indeterminate. The panel commissioned by the National Research Council to study the impact of technical change on clerical employment determined that the effect on skill level is inconclusive (Reskin and Hartmann, 1986). This is consistent with the findings of Cohn (1985) in a study of 1 See, for example, the discussion in Braverman (1974) passim, and the controversy surrounding job evaluation studies in the comparable worth debates. See also Howell and Wolff (1991) and Francis Green (1992) on the question of what constitutes skill. 2 Temporary clerical work has much in common with construction work. Both require a high level of general knowledge acquired off the job: typing, language and word processing skills gained in high school or secretarial school in the case of clerical work and skills taught in apprenticeship programs in the case of construction. Both involve a relatively short tenure at any particular job site as workers complete assignments and move among placements with changes in demand. But the similarities end here. Job assignments and pay scales are established via a union hiring hall in the case of construction and through temporary help firms in the case of clerical work. Perhaps 9 to 5 should establish a union hiring hall. Family structure, flexible employment, labor market segmentation 95 clerical work in two British Industries. Cohn finds the skill level to be remarkably stable over time. While an analysis of the secular growth in THS employment is beyond the scope of this paper, the timing of the industry's growth does not match findings of the National Research Council or of Cohn. Empirical analysis While the preferences of women workers for employment flexibility owing to their family responsibilities are not directly observable, it is reasonable to expect that if women were making decisions based on these responsibilities, married women with young children would be over-represented in temporary employment. To formally test the impact of family responsibilities on the likelihood of being a THS employee, I pooled the female temporary workers and a random sample of non-temporary women workers in the labor force, both from the CPS, to determine whether family characteristics could predict the likelihood of a woman being a THS employee. Similar procedures were followed for male workers, thus allowing for analysis along two dimensions: differences between temporary and non-temporary workers and between men and women. The model tested took the form: log[P/(1-P)] =a+ ~ 1 AGE+ ~2AGESQ+ ~3EDUC+ ~4MARITAL + ~5EARNERS + ~6AGECH1 + ~7AGECH2 +E. The independent variable takes on the value of 1 if the worker is a THS employee, otherwise it is zero. Note that this is a stratified sample in which the sampling rate was 100% for THS employees and 1% for the labor force as a whole. Only the constant term is affected by this procedure (Maddala, 1983: 90). The estimation results are reported in Table 2. The hypothesis that women are self-selecting into THS employment based on family characteristics is not supported by the results in Table 2. The coefficients on all variables except AGE and A GESQ are insignificant at the very generous test level of 10%. The sign on the marital status variable is opposite that predicted by the theory that women choose flexibility based on family type. The signs on the AGECH1 and AGECH2 variables are consistent with that view, but they are not significant. Since colinearity among the independent variables would make it more difficult to get significant t statistics (a difficulty that would favor my hypothesis), the model was run sequentially omitting those variables among which colinearity could be expected. None ofthe significance levels were changed by this procedure, suggesting that the research results are quite robust. The basis created by omitting variables would have pushed the estimates of the coefficients upward for the remaining variables and reduced the standard errors. The supply-side factors were not significant even under these circumstances. Another way to show that the THS workers are not self-selecting on the basis of family responsibilities is to compare them with women who are out of the labor force. This would allow us to determine whether the THS industry's flexible employment option is drawing women into the labor force who would otherwise not be working or seeking work. In a procedure parallel to that described above, the model was tested 96 J. Lapidus Table 2. Binary logit results from sample offemale temporary employees and a random sample of women in the labor force Dependent variable: Observations: Log likelihood: Variable y= 1 if worker is THS employee, otherwisey=O. 530 Degrees of freedom: 522 304·799 Coefficient Std. error t-Stat* P-value CONSTANT AGE AGESQ EDUC MARITAL EARNERS AGECHJ AGECH2 0·348141 -0·118069 0·001345 0·055648 -0·255239 0·339837 0·418461 0·166489 0·926609 0·040781 0·000494 0·042859 0·229875 0·271865 0·273442 0·275306 0·375715 -2·8905205 2·721207 1·298401 -1·110336 1·250020 1·530344 0·604742 0·707128 0·003789 0·006504 0·194149 0·266854 0·211292 0·125932 0·545350 Variables coded as follows: AGE=age ofrespondent. A GESQ =square of age. MARITAL= marital status, 1 if married not separated, else 0. EDUC =education, highest grade attended. EARNERS= 1 if there is another earner in the family, else 0. AGECHJ =I if there is a child between the ages ofO and 6, else 0. AGECH2 = 1 if all children are over 6 and at least one child is between 6 and 13, else 0. Source: Current Population Survey, May 1985. on a data set offemale THS employees and women not in the labor force. In this case, the presence of significant differences can be interpreted to mean the THS employees are not drawn from the same population as women not in the labor force. The education and age of children variables were significant at the 10% level. The industry does not appear to increase female labor supply. The lack of explanatory power in the family characteristic variables for women is particularly striking when compared with the results for men. Table 3 reports the results from a logit model in which probability of a male being a THS employee is the dependent variable. For men, the coefficient on MARITAL is the only significant coefficient and its sign is negative. While family responsibilities do not affect the likelihood of being a THS employee for women, they appear to decrease the likelihood for men. One possible explanation for this difference is that the cost to the employer of securing flexibility through utilizing a THS service is calculated as a percent ofthe worker's hourly wage and is paid to the THS firm, not to the worker. It is possible that this option is less viable in the case of married male workers who customarily command a higher wage in the labor market. The potential role of male/female wage differentials in explaining the concentration ofwomen in THS employment has important implications both theoretically and empirically. In a model in which men and women calculate the potential income from THS and non-THS employment each might be expected to jointly maximize their income and flexibility. Most THS industry researchers have focused on a presumed greater utility offlexibility for women to explain their participation in the Family structure, flexible employment, labor market segmentation Table 3. Probability of a male worker being employed by a THSfirm 97 Dependent variable= 1 if worker is THS employee, otherwise y=O Observations: 530 Degrees of freedom: Log likelihood: 193·239 Variable Coefficient Std. error t-Stat* 523 P-value CONSTANT AGE AGESQ EDUC MARITAL AGECHJ AGECH2 0·638982 -0·050089 0·000458 -0·027709 -1·6985637 -0·095094 -0·127076 1·116299 0·057142 0·000710 0·044439 0·369244 0·465432 0·482422 0·572411 -0·876557 0·644925 -0·623524 -4·600046 -0·204314 0·263412 0·567044 0·380727 0·518976 0·532940 0·000004 0·838108 0·792233 Variables coded as follows: AGE= age of respondent. AGESQ =square of age. MARITAL= marital status, 1 if married not separated, else 0. BDUC =education, highest grade attended. EARNERS= 1 if there is another earner in the family, else 0. AGECHJ = 1 if there is a child between the ages ofO and 6, else 0. AGECH2= 1 if all children are over 6 and at least one child is between 6 and 13, else 0. Source: Current Population Survey, May 1985. THS market. However, it is also possible that the income/flexibility trade-off is different for men and women owing to sex-based differences in the cost of flexibility as measured by foregone wages. The extremely small sample for which wage data are available precludes careful consideration of the relative income/flexibility trade-off for men and women. However, the limited data do support the theory that the costs to women of increased flexibility are quite low when compared with their other options in a sex-segregated labor market. Mean wages for female THS employees are identical to those for a random sample of female non-THS workers. The mean hourly wage of women THS employees is $5·96 an hour, as compared with $5·92 an hour for all women workers. The mean wages of female temporary and non-temporary workers are identical despite the fact that female temporary workers tend to have higher levels of education, so there may in fact be some trade-off between flexibility and wages. In order to estimate the effect of temporary work on the wages of female temporary employees, their predicted wage ifthese same women were employed in non-temporary work was calculated and compared with their actual wage. Following the procedure used by Blinder (1973), the coefficients for non-THS employees were combined with the means for THS employees to establish a predicted wage if the employees were not employed in the temporary sector. 1 1 By running a regression for non-temporary workers I am truncating the sample. This could eliminate low-wage workers from the sample and confound the structural effect ofbeing in the non-THS sector with the effect of having the unobservable characteristics of permanent workers. However, the conventional ways of correcting this bias have recently been shown not to be very good (Stozenberg and Relies, 1990). We do not have any really good ways of rectifying the problem. 98 J. Lapidus The following wage equation was estimated using ordinary least squares for a sample of female non-temporary workers: y=a+~ 1 AGE+~2NCHIL+~3MARITAL+~4EDUC+c. AGE, EDUC and NCHIL are coded as above. In order to eliminate union wage effects, the random sample was restricted to those jobs not covered by a union contract. The predicted value of mean wages for temporary workers, were they working in the non-temporary sector of the labor force, is $5·96 per hour, the same as their actual wage. Thus it appears that women temporary workers do not pay a wage penalty for the flexibility they receive. Again, these data are particularly interesting when contrasted with those for men. For male THS employees, the average hourly wage is $5·39; for non-THS workers it is $10·02. The predicted wage for male temporary workers of$7·60 per hour; a gap of $2·21 less than their actual wage. The cost of flexibility to a male worker is in stark contrast to the absence of such a cost for women. This difference, rather than the desire for flexibility due to the family responsibilities of female workers, may partially account for the gender composition of the THS work force. Implications for labor market theory The assumption with which this research was begun was that the over-representation of women in THS employment represents a cost (paid by women) for resolving the conflicting demands ofmarket work and family responsibilities. Itprovides flexibility, but the cost is foregone wages and benefits, job stability, and prospects for advancement. Thus, I expected women with family responsibilities to be more likely to be THS employees. Since this does not appear to be the case the question of why two-thirds of the industry's employees are women remains. As noted in the Introduction, both supply-side theories and labor market segmentation theory assume that the labor market opportunity set is created independently ofthe gender ofthe worker. They then differ on whether the assignment ofworkers to particular jobs represents rational choice on the part ofthe worker or the employer. The weaknesses of the former were addressed explicitly in the empirical model above. The implications of these findings for labor market segmentation theory are by necessity more speculative in nature. The categories ofprimary and secondary labor markets are problematic for categorizing both the THS industry and women. THS employment, by definition short-term work with high turnover, is used most heavily in precisely those industries usually associated with the primary labor market (Bureau of National Affairs, 1986). The growth ofthe temporary help industry can perhaps best be understood as an attempt by employers to circumvent the primary labor market without incurring the discipline problems usually associated with the secondary labor market (Lapidus, 1990). Albelda (1985), Albelda and Tilly (1990), and Dickens et al. (1988) all suggest that the duallabor market typology does not seem to provide a good characterization for women. Indeed, anomalies exist within the labor market segmentation literature itself. Gordon et al. (1982) argue that women are employed in the secondary labor market as the result of the slotting of workers into relatively good and bad jobs (determined by job, not worker, characteristics) based on social rankings in society as Family structure, flexible employment, tabor market segmentation 99 a whole. However, in describing the distribution ofemployment among labor market segments they report that 61·2% offemale workers in 1950 were in the independent primary or subordinate primary sector. Beechey and Perkins' (1987) note that the work of the Cambridge Labour Studies Group pays greater attention to supply-side factors. 'In the revised theory, labour markets are conceptualized as structured by the productive system, on the one hand, and by the system of social reproduction on the other' (Beechey and Perkins, 1987: 138). In other words, in addition to the technological and industrial characteristics cited by the labor market segmentation theorists, the process of job creation itself is in part a function ofthe gender of the worker. Given this assumption, the findings presented here are no longer surprising. Employment in the temporary help industry is two-thirds female not because particular groups of women workers choose these jobs as compatible with family responsibilities. Rather, they are female precisely because these jobs were created with women workers as the prototype. There are many explanations for the increase in the demand for temporary workers in the 1980s: increased uncertainty, declining union strength, increased competition placing pressure on firms to cut labor costs, and changing political alignments. However, converting the demand for temporary help into effective demand requires an available supply of workers. Women workers offer employers such a supply for two reasons. First, women's alternatives are jobs paying wages already depressed by a segregated labor market. Second, and perhaps more important, is an ideological consideration. Women workers are a group for whom it is possible to state a priorithat they are doing the work by choice. The failure of temporary work to provide job security or benefits is obscured by the assumption that women are not wholly dependent on their own wages. Conclusion In summary, the data do not support the hypothesis that women with family responsibilities are choosing employment in the temporary help industry. Rather, family characteristics fail to predict whether women workers will be employed by a THS firm. This is in contrast to the results for men. Since women are comparing temporary employment to other options in a sex-segregated labor market, the opportunity cost of flexibility is lower for women than it is for men. This conclusion is supported by the gap between actual and predicted wages for male, as contrasted with female, temporary workers. Using the framework of the second generation labor market segmentation school, the interplay between family and labor market can be addressed without reducing questions of labor supply to either the microeconomic level of individual decision making or to an exogenous status. It may be a fruitful approach for further research into the dynamics which produce a sex-differentiated labor market. References Albelda, R. 1985. Nice work ifyou can get it: segmentation ofwhite and black women workers in the post-war period, Review of Radical Political Economy, vol. 17, no. 3 Albelda, R. and Tilly, C. 1992. All in the family: family types, access to income, and family income policies, Policy Studies Journal, vol. 20, no. 3 100 J. Lapidus Appelbaum, E. 1986. Restructuring work: temporary, part-time and at-home employment, in Hartmann et al. (eds), Computer Chips and Paper Clips, Volume II, Washington, DC, National Academy Press Beechey, V. andPerkins, T. 1987.AMatterofHours: Women, Part-time Work and the Labour Market, Minnesota, University of Minnesota Press Blinder, A. 1973. Wage discrimination: reduced form and structural estimates, Journal of Human Resources, vol. VII, no. 4 Braverman, H. 1974. Labor and Monopoly Capital: The Degradation of Work in the Twentieth Century, New York, Monthly Review Press Bureau of National Affairs. 1986. The Changing Workplace: New Directions in Staffing and Scheduling, The Bureau of National Affairs, Rockville, Maryland Cohn, S. 1985. The Process of Occupational Sex- Typing: The Feminization ofClerical Labor in Great Britain, Philadelphia, Temple University Press Dickens, W., Friedberg, R. and Lang, K. 1993. Is the relation between gender and labor market segmentation changing? Unpublished paper, Berkeley, California, University of California Edlefsen, L. and Jones, S. 1984. Gauss Version 1.49B. Kent, WA, Aptech Systems, Inc England, P. 1982. The failure ofhuman capital theory to explain occupational sex segregation, Journal of Human Resources, vol. XVII Golden, L. and Appelbaum, E. 1992. What was driving the 1982-88 boom in temporary employment? American Journal of Economics and Sociology, vol. 5, no. 4 Gordon, D., Edwards, R. and Reich, M. 1982. Segmented Work, Divided Workers, The Historical Transformation of Labor in the United States, Cambridge, Cambridge University Press Green. 1992. On the political economy of skill in the advanced industrial nations, Review of Political Economy, vol. 4, no. 4 Grune, J. (ed.). Manual on Pay Equity: Raising Wages for Women's Work, Washington, DC, Conference on Alternative State and Local Policies Howe, W. 1986. Temporary help workers: who they are, what jobs they hold, Monthly Labor Review, vol. 109, no. 4 Howell, D. and Wolff, E. 1991. Trends in the growth and distribution of skills in the US workplace, Industrial and Labor Relations Review, vol. 44 Lapidus, J. 1990. The temporary help supply industry and the operation of the Labor Market, unpublished Ph.D. dissertation, Amherst, MA, University of Massachusetts Maddala, G. S. 1983. Limited Dependent and Qualitative Variables in Econometrics, Econometric Society Monograph No. 3, Cambridge, Cambridge University Press Mangum, G., Mayall, D. and Nelson, K. 1985. The temporary help industry: a response to the dual internallabor market, Industrial and Labor Relations Review, vol. 38 Mincer, J. and Polachek, S. 1974. Family investments in human capital: earnings of women, Journal ofPolitical Economy, vol. 82 Plewes, T. 1988. Understanding the data on part-time and temporary work, in US Department of Labor, Office of the Secretary, Women's Bureau, Flexible Workstyles: A Look at Contingent Labor, Washington, DC, US Government Printing Office Polachek, S. 1981. Occupational self-selection: a human capital approach, Review ofEconomics and Statistics, vol. 63 Reskin, B. and Hartmann, H. Eds. 1986. Women's Work, Men's Work: Sex Segregation on the Job, Washington, DC, National Academy Press Rubery, J. 1989. Labour market flexibility in Britain, in Green, F. (ed.), The Restructuring of the UK Economy, Hemel Hempstead, Harvester Wheatsheaf Stozenberg and Relies. 1990. Theory testing in a world of constrained research design, Sociological Methods and Research, vol. XVIII US Department of Commerce, Bureau ofthe Census. 1985. The May 1985 Current Population Survey, Machine-readable data file US Department of Commerce, Bureau of the Census. 1978. The Current Population Survey: Design and Methodology, Technical Paper No. 40, Washington, DC, US Government Printing Office  Dublin Core The Dublin Core metadata element set is common to all Omeka records, including items, files, and collections. For more information see, http://dublincore.org/documents/dces/. Title A name given to the resource Family structure, flexible employment, and labor market segmentation: evidence from a study of the temporary help industry Creator An entity primarily responsible for making the resource June Lapidus Source A related resource from which the described resource is derived International Contributions to Labour Studies 3, pages 91-100 Date A point or period of time associated with an event in the lifecycle of the resource 1993