Does It Pay To Care? The Commodification of Care and Remuneration in The Market

PAGE 1

DOES IT PAY TO CARE? THE COMMODIFICATION OF CARE AND REMUNERA TION IN THE MARKET BY CRYSTAL MILLER A THESIS Submitted to the Division of Social Sciences New College of Florida in partial fulfillment of the requirements for the degree Bachelor of Arts in Economics and Gender Studies Under the sponsorship of Dr. Richard Coe Sarasota, Florida May, 2012

PAGE 2

Miller ii Acknowledgements: Dr. Richard Coe Dr. Patrick Van Horn Dr. Sarah Hernandez Dr. Duff Cooper Dr. Sue Mom' Miller Dr. Kevin Dad' Miller Dr. Gertrude, the Cat Doctor. All of my Dr. friends All of my friends

PAGE 3

Miller iii Table of Contents Title ................................ ................................ ................................ ......... i Table of Contents ................................ ................................ ................. iii Abstract ................................ ................................ ................................ v 1. Introduction ................................ ................................ ...................... 1 Rapid Growth of the Demand for Care Services ................................ .................... 1 Who Are Direct Care Workers? ................................ ................................ .............. 4 What does a Direct Care Worker Do? ................................ ................................ .... 6 Dir ect Care: Shape of the Industry ................................ ................................ ........ 8 Compensation of Direct Care Workers ................................ ................................ ... 8 A Wage Penalty for Caring? ................................ ................................ ................. 12 2. Useful Models for Understanding Wage Differentials ............... 14 The Standard Model ................................ ................................ ............................ 14 Supply and Demand for Care Giving ................................ ................................ ... 15 A Demand for Formal and Informal Care Giving ................................ ................ 17 Supply of Care Labor ................................ ................................ ........................... 21 Domestic Production of Care vs. Market Production of Care ............................... 25 Human Capital Theory and the Supply of Care Labor ................................ ............ 27 Intrafamilial Care Work Decisions and the Public Market for Ca re .................. 29 Care Worker Wage Differentials in the Human Capital Model ......................... 32 Crowding Hypothesis ................................ ................................ ................................ ...... 33 Dual Labor Market Theory ................................ ................................ ............... 35 3. Why Care? ................................ ................................ ....................... 38 Defining Care Labor ................................ ................................ ............................... 38 Contracts for Direct Care Services ................................ ................................ ........ 42 Direct Care Work as a Good or Service ................................ ................................ 43 How to Measure Care Services in the Market ................................ ...................... 45 The Costs and Benefits of Care Work ................................ ............................... 46

PAGE 4

Miller iv Interface between Theories of Care and Economics ................................ .............. 49 The Tradition of Care Labor ................................ ................................ ........................... 49 A L ink between Gender and Care Work ................................ ........................... 52 A Feminist Approach to Economizing Care ................................ ......................... 54 Feminist Constructionists ................................ ................................ ................. 55 Affirmative action feminists ................................ ................................ ............. 57 4. Data and Methods ................................ ................................ .......... 60 Overview of Select Variables ................................ ................................ ............... 61 Institutional Variables ................................ ................................ ................................ ...... 64 Human Capital Variables ................................ ................................ ................. 67 5. Results: ................................ ................................ ............................. 69 Statement of Econometric Model : ................................ ................................ ........ 69 Equation 3: Wage Estimation 2008 ................................ ................................ .... 72 Equation 4: Pooled Log Wage Estimation 1988 2008 ................................ ......... 74 6. Conclusion: ................................ ................................ ..................... 77 Appendix A : ................................ ................................ ........................ 82 References ................................ ................................ ........................... 86

PAGE 5

Miller v DOES IT PAY TO CARE? Crystal Miller New College of Florida 2012 ABSTRACT The growth in the supply of care laborers, a predomin antly female field, is stagnating at a time when there is a predicted increase in the demand for care labor. This gap in care will continue to widen unless care work is rewarded more fairly in the market. The goal of this thesis is to examine explanations for a care penalty in wages, and test empirically for factors which might exacerbate a wage penalty for participating in care work, including sex, race, marital status, education and experience. Samples from the years 1979, 1988, 1998 and 2008 were gath ered from the National Longitudinal Study of Youth 1979 and tests were performed with Statistical Analysis Software. A number of independent variables, those historically used to predict wages, were studied and their effect on the wages of care workers wer e tested and analyzed. The results supported the hypothesis. A wage penalty for participating in care labor was found to be statistically significant, even with controls for race, gender, education and experience. This is attributed to a cultural devaluati on of caring activities due to their association with altruism and mothering. Dr. Richard Coe

PAGE 6

Miller 1 1. Introduction Rapid Growth of the Demand for Care Services The Direct Care industry, encompassing positions such as personal aides, home health aides and cer tified nurses, is one of the fastest growing occupational fields in the United States. It has been reasonably suggested that demand for care will increase by 40% by the year 2050. 1 The reasons behind this rapid growth rate are many. The population of ind ividuals needing care (aged 70+) is increasing. The elderly live longer with diseases, and family size decreases, while child rearing is delayed and the divorce rate goes up. 2 In addition, the ability to geographically separate from family members is grea ter, and the costs diminished as technology allows for satisfactory alternatives to face face communication. All of these factors lead to an increased demand for contracted care services. The rise of professional care services has remarkable implications for the women concentrated in these professions, as well as those on the receiving end of this shift, the care recipients. What motivates care giving, and how is it different from other commodities? For some, care giving feels like a natural extension of inherent abilities, and lack of monetary compensation is exchanged for non pecuniary 1 PHI. Feb, 2011. Who are direct care work ers? Bronx, NY: Paraprofessional Healthcare Institute Quality Care through Quality Jobs, 3. 2 PHI. Nov, 2008. Direct care jobs and long term care: Untapped engine for job creation and economic growth

PAGE 7

Miller 2 rewards such as fulfillment, feelings of warmth, connection, and usefulness. 1 These non pecuniary rewards are sometimes called intrinsic rewards, psychic income, motivatin g factors or compensating differentials. Those participating in formal care giving are primarily women whose participation in the field is considered unskilled. It will be argued that care giving is not unskilled' labor, but rather undervalued' due to a cultural devaluation of caring activities, which results in disparate earnings in occupationally segregated occupations. Unskilled labor requires practically no training or experience to perform the job competently, while care labor does not yet require a degree or certification, it takes a certain skill set or a certain kind of person' to do it well. The rather general expectations about what care givers will be like, and why they will be performing care work, are derivatives from the old fashioned specialization in domestic labor by women. It is often assumed that altruism and maternal instincts guide the decision to enter into care labor; this is not always the case, though motivating factors certainly play a role. Motivations for care gi ving surface from welfare, affection as a moral consideration, a sense of responsibility, duty or obligation, intrinsic enjoyment, expectations of long term reciprocity, an explicit contract for reward, fear of retribution by threat. 2 Care giving can be t hought of as an 1 This is not to underplay the role that the media and tradition play in socializing gender roles. A utilitarian approach might suggest that because social forces inhabit us so powerfully, and affect our individual utility functions, than perhaps a broader look into how gender roles demonstrate themselves in the economy would be necessary. For more information on gender and socialization see Robin Leidner (1998). 2 J ochimsen Maren A. 2003. Integrating Vulnerability : On the Impact of Caring on Economic Theorizing." In Toward a Feminist Philosophy of Econo mics. Eds. Drucilla K. Barker, Edith Kuiper. London ; Routledge, 2003.

PAGE 8

Miller 3 imperfect commodity'; its bifurcation between the private and public sphere is unique, insofar as it can be executed in the home for little to no pay and in the market. 3 The disadvantages for engaging in feminine coded work are demonstrate d in the form of lower wages across the board for female employees than their male counterparts. 4 If this disadvantage, for not merely being a woman or a minority, but performing a function associated with femininity, is the case, we might expect to see lo wer wages for individuals choosing to go into care work. Particularly if the individuals working in direct care are mostly female, that is, regardless of the race or gender of the person participating, occupations will pay less when the occupation is or ha s been mostly women or minorities. It is clear is that there are no explicit ways to reward non self interested behavior in the market. This will become a larger problem when the demand for care is much greater than those willing to supply it. If there d oes not exist an undervaluation of care labor (relative to other low wage jobs) then it would be reasonable to expect that those individuals going into care labor retain their jobs, receive raises and promotions, live at the national average for the povert y threshold, and receive comparable pay for jobs of comparable worth. 3 Rajidner 1988 4 I will define feminine coded work' as either jobs employing mostly women (75% or more), or tasks involving what is traditionally thought of as women's work; like hom emaking, housekeeping, raising children, acting out concern for others mothering' etcetera ;

PAGE 9

Miller 4 Figure 1 : Fast Growing Occupations Predicted through 2018 Who Are Direct Care Workers? Women aged 25 44 compose the majority of paid caregivers. 5 In 2008, a low estimate of 3 million direct care workers were employed as Nursing Aides, Orderlies and Attendants, Home Health Aides and Personal Care Aides. 6 7 Low earnings and part time work, weak benefits and reliance on 5 PHI. Feb, 2011. Who are direct care workers?. Bronx, NY: Paraprofessional Healthcare Institute Quality Care through Quality Jobs, 3. 6 PHI. Nov, 2008. Direct care jobs and long term care: Untapped engine for job creation and economic growth. Bronx, NY: Paraprofessional Healthcare Institute Quality Care through Quality Jobs, 2. 7 Privately contracted care workers from the community are rarely counted in BLS su rveys, to account for this the Direct Care Alliance cites a figure 1.5 to 3 times larger than 3 million direct care workers accounted for in the BLS surveys.

PAGE 10

Miller 5 public welfare programs have define d many of those participating in direct care occupations, and have turned many others away from the industry. According to a 2010 brief published by the Paraprofessional Healthcare Institute (PHI), 89% of direct care workers are women, 30% are black and 16 % are Hispanic. Full time year round employees comprise about half of all formal care workers, only 52%. They annually earn a median income of $16,800, about half of the median annual U.S. populations earnings. 8 A little more than 45% have participated in college or pursued an advanced degree. 9 More than one quarter of all direct care workers have no health insurance and 46%, use public benefits such as Medicaid or food stamps. 10 In 2018, it will grow to be the largest occupational group in the U.S., just u nder retail salespersons. 11 Table 1 : Total Income for Care Workers as Compared to the National Poverty Level 2000 8 PHI. Nov, 2008. Direct care jobs and long term care: Untapped engine for job creation and econ omic growth. Bronx, NY: Paraprofessional Healthcare Institute Quality Care through Quality Jobs, 2. 9 PHI. Feb, 2011. Who are direct care workers?. Bronx, NY: Paraprofessional Healthcare Institute Quality Care through Quality Jobs, 3. 10 PHI. Feb, 2011. Who are direct care workers? Bronx, NY: Paraprofessional Healthcare Institute Quality Care through Quality Jobs, 3. 11 PHI. Feb, 2011. Who are direct care workers?. Bronx, NY: Paraprofessional Healthcare Institute Quality Care through Quality Jobs, 3.

PAGE 11

Miller 6 What does a Direct Care Worker Do? Direct care services cover a large spectrum of needs for the elderly and provide paraprofessional assistance to any disabled, chronically ill, cognitively impaired individual. Care services involve errands such as picking up medications, helping with daily living activities and primarily take place in a non institutional residence. 12 T he amount of time any individual requires care services varies according to the type of disability they have. Sometimes it is brief visits for those recovering from an injury or hospital visit, or long term assistance for those who will consistently need m ore than family and friends can sustain. 13 Direct care is care given under the direction of a regulated professional such as a Nurse or CNA, doctor or health agency, as opposed to a family member. Direct Care workers are Personal and Home Health Aides and P ersonal and Home Care Aides, Certified Nursing Assistants (CNA's) and Nursing Aides. 14 Home Health Aides, Personal, and Home Care Aides perform very similar duties, but differ in the daily regulation they have from professional staff, and slightly in the na ture of the tasks required. Home Health Aides generally work for an agency the government funds, and assist with light medical procedures such as pulse taking, checking respiration and with training, changing dressings and medication(s) administrations. Ho me health aides work under the supervision of professional staff and are required to communicate with them 12 Institutional residences are assisted living facilities and nursing facilities whereas non institutional residences are group living houses (not federally funded) and private homes. 13 Bureau of Labor Statistics. Home health and personal care aides : Occu pational outlook handbook : U.S. bureau of labor statistics in U.S Department of Labor [database online]. 14 Bureau of Labor Statistics. Home health and personal care aides : Occupational outlook handbook : U.S. bureau of labor statistics in U.S Department of Labor [database online].

PAGE 12

Miller 7 regularly on the status of the care recipient. Personal and home care aides can be either self employed or employed through a private agency. Persona l and home care aides often perform a variety of domestic tasks that are not associated directly with health (although home health aides do as well). Personal and home care aides may work independently of professional staff though it is likely, if they wor k for an agency, that supervision will occur, but with less regularity than home health aides. 15 The training and investment for these positions is required only to those working in Medicare and Medicaid certified nursing homes and home health agencies. 16 A h igh school diploma is not required, and much of the training occurs in the form of on the job training (OJT) from supervisors, experienced aides or professional staff. 17 In examining this fast growing niche of the service industry to determine whether a w age penalty exists for those who care, it will be important to understand some key terms associated with the direct care industry. I will define the following terms: care work which is separate from the market, and can occur informally by any individual outside of employment; care labor which is compensated through exchange and enforced by implicit or explicit contract; the care industry -health services, hospital services, educational and social services; direct care occupations which I will use to mea n home health aides, personal and home care aides, certified nurses, and nursing aides; direct care -physical and emotional support 15 Bureau of Labor Statistics. Home health and personal care aides : Occupational outlook handbook : U.S. bureau of labor statistics in U.S Department of Labor [database online]. 16 PHI. Feb, 2011. Who are direct care worker.? Br onx, NY: Paraprofessional Healthcare Institute Quality Care through Quality Jobs, 3. 17 Bureau of Labor Statistics. Home health and personal care aides : Occupational outlook handbook : U.S. bureau of labor statistics in U.S Department of Labor [database online]. Available from http://www.bls.gov/ooh/Healthcare/Home health and personal care aides

PAGE 13

Miller 8 usually taking place in a residence and applied from the care giver to the care recipient without a medium; as opposed to i ndirect care -work done which helps to sustain the care recipients standard of living or health but is not done to them (such as food delivery or transportation). In addition, there is formal (paid or contracted care) and informal care (unpaid or family c are.) 18 For the purposes of this paper, I will use formal care and paid care interchangeably, likewise with unpaid care or family care and informal care. Direct Care: Shape of the Industry I will attempt to define care labor in the following section. Firs t I will consider the nature of the paid versus unpaid labor and its traditional role in individuals' lives, and then outline the markets response to care work as a good or service, and finally, note care labors position in the economy. This lays the groun dwork for looking at the shape of the care industry and for applying relevant theoretical models. Compensation of Direct Care Workers The Direct Care Alliance has issued a number of policy briefs commenting on the status of the direct care industry. Some of the issues they highlight include the lack of coverage under the country's most basic labor laws. In a 2009 Supreme Court decision about inclusion of home and personal care 18 Unpaid care is closely associated with family care, though the two are by no means synonymous, unpaid care can be performed by neighbors or friends or other non relations. Unfortunately, there is a lacuna of data on privately contracted caregivers, while family caregivers are more accessible to document. Data presented on unpaid care giving may underestimate 5 10% of caregivers. For more information on this under ground economy of non kin care giving, see Judith C. Barker(2002).

PAGE 14

Miller 9 aides under the Fair Standards Labor Act, a companionship exemption, intended f or babysitters working on a casual basis, was applied to care workers employed by a third party. This means that care workers must forego access to minimum wage laws and overtime pay. In the hearing, Long Island Care at Home, LTD. Et al vs. Coke, Ms. Cok e argued for the right to back pay and the right to minimum wage and overtime compensation after living in her employers home. The Supreme Court ruled against Ms. Coke, affirming a policy position that direct care labor is low status work, in the words of Ms. Coke's Counsel: "Why should home care workers uniquely carry the burden of society's need for their services?" 19 A current amendment to the law has yet to pass through Congress. The Fair Home Healthcare Act, proposed by in Senate in 2007, would provide greater protections for home healthcare workers and help to define what exactly casual basis' means. That home care costs would increase, and limit access to private in home care is one of the unjustified concerns against the act. If wages in the home ca re industry rose, care workers would cease to leave the job so regularly, and would attract greater quality care and healthy growth in the industry. Quite the opposite of reducing access, it would distribute the costs of care from the health care agencies to the care workers themselves. 20 19 Source: Howes, Candace. The best and worst state practices in medicaid long term care | russell sage foundation in Direct Care Alliance [database online]. 2010 20 Pagac, Adrienne, and D resser, Laura. Better jobs for in home direct care workers | russell sage foundation in Direct Care Alliance

PAGE 15

Miller 10 Table 2 : Median Wages for Workers in the US Enhanced Federal Medicaid funding would also help to raise the minimum hourly wage floor for care workers. As diversity within the aging populatio n increases, the number of indivuals who rely on informal families or social connections for care and support likewise increases, at a time when government and other institutions largely define family' based on marriage and biological kin. 21 How can we sup port the aging Lesbian/Gay/Bi/Transsexual(LGBT) members of our society, when care is characterized by gaps in care, for one of the groups most prone to need contracted care? One DCA policy brief mentions the solidarity between direct care workers and the L GBT communities with regard to the undervaluation that takes place culturaly for both groups, and their shared need to be 21 McKenzie, Nancy. Supporting direct care workers in caring for aging lesbian, gay, bisexual, transgender individuals | russell sage foundation in Direct Care Alliance

PAGE 16

Miller 11 protected equally in the workplace and in civil life. 22 "1.4 3.8 million LGB Americans are reaching standard retirement age with an est imated 3.6 7.2 million projected to be 65+ in 2030." (DCA Policy Brief) The care industry needs to confront the possibility that whichever policy decisions are made will have a large impact on those with the highest level of health disparities and access t o care, the LGBT community. There must be efforts made to understand bias, prejudice, discrimination and their sources with regard to training on the part of home care agencies, especially as this group is most sensitive to changes in their environment an d the most likely to hire in home care. Table 3 : Numbers in American Long Term Care Number of Americans needing long term care services Nearly 13 million Number of nursing home residents 1.8 million Number receiving home and community based services (HCBS) 10.9 million Number of people receiving unpaid help from family and friends 9.8 million Number of people receiving HCBS who get paid help 1.4 million 23 Because Federal Medicaid statutes say that states must provide Medic aid long term care at an institutuional level to anyone over 21 who needs regular assistance with activities of daily living, many people who simply require 22 McKenzie, Nancy. Supporting direct care workers in caring for aging lesbian, gay, bisexual, transgender individuals | russell sage foundation in Direct Care Alliance [database online]. 2010 23 Howes, Candace. The best and worst state practices in medicaid long term care | russell sage foundation in Direct Care Alliance [database online]. 2010

PAGE 17

Miller 12 personal/home care servies are placed in nursing homes instead. 24 Arbitrary institutionalization doe s nothing for the quality of life for our aging citizens who require care. It is seen that receiving care in a residential setting is much preferred, as it contributes positively to health and wellness while dealing with the struggle of aging. 25 The policy implications include: reduce institutionalization; implement better spousal protection, organize single points of entry for persons seeking long term care services and establish policies to expand and stabilize the care work force, including mechanisms fo r collective bargaining by care employees. 26 A Wage Penalty for Caring? I will address criticisms, concerns and issues brought forth by economists, who verify the need for an economic dialogue with concepts of care. With the National Longitudinal Survey, I intend to run linear regressions based on a wage equation in order to look for a wage penalty for caring, using human capital investment models, in addition to bargaining/threat point models. I intend to find if there exists a penalty for being caring' in the market above and beyond already substantiated gender and human capital penalties. I will also derive summary statistics on the population working in care labor in order to affirm or deny similarities 24 Howes, Candace. The best and worst state practices in medicaid long term care | russell sage foundation in Direct Care Alliance [database onlin e]. 2010 [cited 4/20/2012 2012]. Available from http://www.russellsage.org/research/reports/howes medicaid 25 Using recovery act funds to improve direct care jobs and the quality of direct care services | russell sage foundation [cited 4/20/2012 2012]. Available from http://www.russellsage.org/research/reports/recove ry act direct care workers 26 Howes, Candace. The best and worst state practices in medicaid long term care | russell sage foundation in Direct Care Alliance [database online]. 2010 Available from http://www.russellsage.org/research/reports/howes medicaid

PAGE 18

Miller 13 between the National Longitudinal population and surveys done by other organizations on care labor. I analyze wages with regard to gender, race/ethnicity, educational levels, training, tenure, family size, poverty status, marital status residence and region. The wages for all years were adjusted to refl ect real 2008 dollars as extracted from the Consumer Price Index. To account for a wage penalty I created a dummy variable for those participating in care labor by separating out individuals with occupational codes for those professions. I looked at the c ohorts (N=12,686) in four different years, 1979, 1988, 1998 and 2008 in order to gain a longitudinal perspective. This paper studies the consequences of occupational segregation for those engaging in care labor which coincides with a dramatic rise in the f emale labor force participation rate, as well as a rise in the percent of women attaining higher education. The structure of the remainder of this paper is as follows. Section 1 introduced direct care labor, identifies participants and defined te rms. Section 2 reviews the standard models and its economic implications for the care industry. Section 3 is a survey of the literature on care work and its intersection with economics. Section 3 also helps to address the lacunae of theories of care in th e field, and forms hypotheses for why caring labor is economically disadvantageous. Section 4 provides an overview of the data and methods used for the empirical analysis. Section 5 presents results and concludes.

PAGE 19

Miller 14 2. Useful Models for Understanding Wage Differentials Economists have developed a number of ways to explain the decisions' individuals make as consumers and workers. It is not critical to fault the models and theories developed for their simplicity, but instead, be aware of the prescriptive fa ults of models. Models are like a blueprint, in the sense that they provide a simple view of how things ought to go; they are rarely dynamic and could not possibly answer or forecast how people act in reality. Instead, models and theories ought to increase our understanding of constraints and decisions made with scarce resources. Models aid to assess the validity of theories with empirical data. However, collecting data is often unglamorous, slow and costly. It also takes quite a while to test conclusions, especially concerning lifetime decisions made by individuals, and careful testing of theories should occur before application as a prescriptive tool. With all this in mind, economists have attempted to apply models to explain differences in wages between m en and women, and certain occupations over others. In this chapter, I review some of the standard economic models of wage determination. I will incorporate the important variables identified in these models into the wage regression in chapter 5. The St andard Model The factors I will study, first theoretically then empirically, include the demand for care services, the supply of labor and the wage influencing industrial and occupational characteristics. Each of these factors can be

PAGE 20

Miller 15 attributed to a diff erent economic model, which helps to illuminate the reasons for wage differentials. These models include the competitive supply and demand model, the Human Capital model, the Crowding Model and Dual labor market theory. Supply and Demand for Care Giving Supply and Demand for Care Giving Recent changes in domestic consumption have led to a rapid growth in demand for care labor, leading to estimates that the care industry will grow 50 percent by 2018, creating a demand for 1.1 million new positions in the next eight years. 27 The population of those needing care will continue to expand as individuals live longer with cancer, AIDs, cardiovascular disease, diabetes, dementia and obesity. 28 One estimate states that the population of those over 85 years old will grow from 4.2 million in 2000 to 2 0 million in 2050. 29 The future generation of retired, elderly will also witness an increase in geographical separation from their families, smaller families and delayed childrearing coinciding with an increase in the divorce rate. 30 All of these factors wor k towards isolating this population from traditional forms of family care and increasing the demand for contracted direct care. The supply and demand model helps to explain a general overview of what will happen to the market for direct care as demand in creases. Intuitively, the demand curve for care labor will shift outwards. However, it 27 Bureau of Labor Statistics. Home health and personal care aides : Occupational outlook handbook : U.S. bureau of labor statistics in U.S Department of Labor [databa se online]. 28 Pagac, Adrienne, and Dresser, Laura. Better jobs for in home direct care workers. Russell Sage Foundation in Direct Care Alliance [database online]. 2010 29 Pagac, Adrienne, and Dresser, Laura. Better jobs for in home direct care workers. Russ ell Sage Foundation in Direct Care Alliance [database online]. 2010 30 Pagac, Adrienne, and Dresser, Laura. Better jobs for in home direct care workers. Russell Sage Foundation in Direct Care Alliance [database online]. 2010

PAGE 21

Miller 16 is important to remember that the supply and demand model assumes a perfectly competitive market and constant movement towards a stable equilibrium. It is worth consider ing whether these terms can apply to the direct care industry and if it is necessary that they do. "An equilibrium exists when all persons willing to work at the going rate are able to find employment and all employers willing to hire someone at the going rate are able to find workers."(Blau et al, pp. 11) Firm equilibrium and market equilibrium refer to different markets. A perfectly competitive market assumes all firms are price takers (they respond to consumers' willingness to pay) and all firms have an identical product, all consumers have perfect information about prices and availability of substitutes and each firm has a relatively small share of the market. 31 Equilibrium occurs at the intersection of the supply and demand curve. In a competitive eq uilibrium, a firm will hire workers at a wage that is equal to their marginal product of labor, which in equilibrium will also be equal to the marginal revenue of labor. To fully comprehend derived demand suppose a cure for the diseases of aging is introdu ced into the population, which lowers the demand for direct care workers. This decrease in demand causes the firms who provide the caregivers to experience an inward shift of demand for the care workers. Low wages, or inability to find work, will dissuade individuals looking to enter the market, likewise some firms may have to exit the industry if demand remains low enough that their workers are not brining in the revenue to sustain their employees' wages. This 31 Jacobsen, Joyce P. The economi cs of gender Wiley Blackwell. 2005.

PAGE 22

Miller 17 standard model of supply and demand suggests t hat all perfectly competitive industries move toward equilibrium in the long run. Compensating differentials' is the term economists use to justify differences in wages for different occupations. This means that lower paid jobs ought to require less hum an capital investments, less prior training and labor to get into the industry and unpleasant jobs should make up for their unpleasantness with higher wages or more benefits. Economists assume that all agents are attempting to maximize their utility out of any given set of potentials; they will choose the actions from which they can derive the greatest happiness. Utility describes an individual's preferences for exchange situations, really, utility is the word economists use for how many utils' of satisfac tion an individual receives out of any given exchange, constrained by exogenous constants such as a budget or time restrictions. Perhaps, those entering care labor accept lower wages because the utility they receive from performing care is greater than the ir need for higher wages. A Demand for Formal and Informal Care Giving A Demand for Formal and Informal Care Giving The demand for caregivers has two aspects, a demand for informal caregivers, outside of the market, and a demand for formal caregivers. Both of these affect the demand from a firm, f or labor. When formal care giving demand is high, agencies are likely to demand more caregivers, that or increase the hours given to current caregivers. Often, a care recipient has a combination of formal and informal care givers. Another primary component in demand for care labor from health agencies is the marginal revenue product (MRP). The marginal revenue product theory of wages provides that

PAGE 23

Miller 18 a profit maximizing firm will hire additional laborers up to the point where the additional revenue the extra w orker brings in is equal to his or her wages. 32 The MRP of labor is the marginal product multiplied by the market price (in this case, the cost to the consumer for care services), if firms are perfectly competitive they will continue adding inputs up to th e point where the MRP of the input (labor) is greater than the market price of the input (wage). As is the case with health agencies hiring out direct care laborers, a firm using only one variable factor of production, means that variable factors' MRP cur ve is also the firms demand function for that input in the short run. The cost to an agency of providing a caregiver to a home is the wages of the caregiver and the initial training investment. In the case of health agencies, they reap a profit of 200% mor e than what they pay their direct care laborers. 33 This implies two things: that the demand for care labor will depend on the productivity of the worker, so that each firm is paying wages less than or equal to the revenue the caregiver generates. And, the d emand for the service (care) they provide will determine demand for caregivers in the labor market. This is called derived demand which refers to the response of the market for labor to the demand for their product in the goods and services market. This neoclassical model agrees with the classic supply and demand graphs below which show that wages are determined by the intersection of supply and demand, all else held constant. The supply curve shows the number of 32 Hamermesh, Daniel S. 1986. The demand for labor in the long run In ed. Orley Ashenfelter and Richard Layard, ed., 429. Handbook of Labor Economics. 33 Gross, Jane. When I needed help, part 2 NYTimes.com in The New York Times Company [database online].

PAGE 24

Miller 19 workers willing to work at all possible p rices. The demand curve shows the various amounts of labor they will hire at various prices by firms in this labor market in a given time period. 34 Figure 2 Supply and Demand for Labor for the Industry and the Firm Graph 1 Indus try Market for Labor Graph 2 Individual Firm Market for Labor Elasticity in the goods and services market is called price elasticity, and elasticity in the labor market is called own wage elasticity. The response of the market to changes in supply o r demand can be partially determined by examining the own wage elasticity of the supply and demand for labor. The marginal product of each worker produces the slope of the demand curve, while its position relative to supply, occurs from the derived demand. Own wage elasticity of demand refers to the sensitivity of the demand curve to changes in wages. Relatively high own wage elasticity of demand means that as wages goes up or down the firms demand for labor will cause a larger 34 Blau, Francine D., Marianne A Ferber,.; Anne E Winkler,. 2010. The Economics of Women, Men and Work. 6th ed. Upper Saddle River, NJ: Prentice Hall.

PAGE 25

Miller 20 movement along the curve, th an the change in wage. For example, imagine the labor market for production of refined oil. Demand for energy inputs have relatively inelastic curves in the goods market, meaning that individuals are not likely to drastically change the quantity demanded f or consumption in response to an increase in the price of oil. In the labor market, there is also inelastic demand, as wages for laborers go up, the movement along the demand curve will reflect a smaller percentage change in the quantity demanded, than the percentage change in wages. Meaning, it is more profitable for the firm to continue producing a certain quantity, than to decrease the amount of workers they pay. Suppose wages goes up by 10% and as a result, employment falls by 4%. Then the elasticity of labor demand is 0.4, which implies labor demand is inelastic since the percentage change in employment is less than the percentage change in wages. Figure 3 : Own Wage Elasticity of Demand 35 35 Blau, Francine D., Marianne A Ferber,.; Anne E Winkler,. 2010. Th e Economics of Women, Men and Work. 6th ed. Upper Saddle River, NJ: Prentice Hall.

PAGE 26

Miller 21 Goods or services that have high price elasticity are usually non necessary goods, or have many cheaper alternatives. The demand for labor in D2 in Figure 3 indicates that at wage P1, the firm with dema nd curve 2 is less likely to hire an additional worker; because labor demand is relatively inelastic at this point, changes in the price (wages of workers) will refer to a small change in the quantity of laborers demanded. The picture of labor for the firm in D1 at P1 (figure 3) has more elastic demand for labor than the firm in D2. If the demand curve for care labor has low own wage elasticity, an increase in the wages for care laborers, say, from P2 to P1, would create a minor movement along the demand cu rve. If the demand curve is steep, quantity labor demanded will exhibit a relatively smaller decrease in the quantity of workers demanded, than the change in wages. However, it will exhibit a major decrease in the quantity of laborers demanded if the deman d curve has a gentler slope, as in D1. Low own wage elasticity of demand will not warrant a substantial change in the quantity of care laborers demanded, as wages go up or down. The downward slope of the demand curve represents diminishing returns of margi nal products of labor. After a certain point, each additional labor performing with a fixed amount of capital will not be able to produce as much: imagine three people trying to work on one sewing machine. With too many inputs of labor, the marginal revenu e product of each additional worker diminishes greatly. Supply of Care Labor Supply of Care Labor The supply curve for labor is independent of the demand curve for labor. The supply of labor will depend on the individual's preferences for

PAGE 27

Miller 22 market goods over home goods. The inputs for consideration are leisure/home time or income; it is assumed that home goods only require the labor of an individual, and market goods, goods or services you can purchase, require income. In neoclassical models, individuals' tradeoff between lei sure and work, and the value of these goods is usually relative to the utility the worker gets out of market income versus home goods, additionally, the opportunity cost for engaging in one hour of leisure is the foregone wages for that hour of work. The o pportunity cost of home care has increased as women's opportunities in the job market have greatly increased in the last two decades. Some individuals have begun to receive compensation for the private in home care work they do for kin, if their medical pr ovider offers such a service. It is most common among those receiving Veterans Assistance, and this kind of compensation is usually reserved for special cases. 36 The two common effects of a wage increase that affect the supply of labor are the substitutio n effect and the income effect. The substitution effect takes place when a laborer chooses to work more as wages goes up, valuing the ability to purchase more market goods with the increased income more than leisure time. The income effect takes place when income goes up, and the ability to engage in more leisure time is preferred to earning more money, if the substitution effect dominates, the individual will choose to supply more labor, whereas if the income effect dominates, the individual will choose to supply less labor. It is difficult to predict which effect will dominate in a 36 National Alliance for Caregiving. 2011. Caregivers of veterans serving the homefront United Health Foundation, ht tp://www.caregiving.org

PAGE 28

Miller 23 given market and much of the forecast will depend on whether leisure time is a normal good. If an individual engages in the sorts of work she would regularly do with leisure tim e, the trade off is more ambiguous. Naturally, the neoclassical models rational agent prefers more money to less. In this case, the supply of labor is contingent upon the utility derived from working in ones' home, versus working for money care giving for someone else. This simplistic supply and demand model does not adequately capture the many complex factors involved in deriving the supply of labor. Additional factors include marital status, alternate sources of income, and the direct and indirect costs of working, family size, ambitions, ability, health and vertical job opportunities. Pricing of Direct Care Services Pricing of Direct Care Services How is the pricing of home care services devised? Recall in the above discussion of demand and supply in the market for care labor, the in creased wage rate needed to increase the supply of care laborers is determined from the derived demand for care labor. The Hicks Marshall Law of Derived Demand is useful for considering how wage increases will affect pricing of care services. First, an inc rease in the wage rate provides incentive for employers to substitute other factors for labor. In the care market this would be difficult to perform, as there are few reasonable substitutes for direct care that do not decrease the quality and effectiveness of care giving. When production costs rise, output decreases, employers hire fewer laborers and the price of the product (care) increases. The Hicks Marshall Law of Derived Demand asserts that own wage elasticity of demand will be high when the

PAGE 29

Miller 24 price elas ticity of demand for a product being produced with labor is high (remember that own wage elasticity occurs in the labor market and price elasticity occurs in the goods market). In addition, when there are close substitutes available, when the supply of oth er factors of production is highly elastic and when the labor costs are a large share of the total cost of production. Another important factor of determining the effect of wage increases on the demand for care services is to consider whether the demand fo r care services in the goods market is income elastic or not. Income elasticity of demand refers to the responsiveness of demand to an increase in income. And the income elasticity of care services will depend on whether consumers consider care services an inferior, normal, necessity, luxury or sticky good. Inferior goods are demanded less as income rises, such as fast food or cheap housing; for normal goods, demand increases as income increases, such as movie tickets or groceries. Necessity goods are nor mal goods whose income elasticity of demand is less than one, meaning an increase in income does not change demand by more than the increase in income: consider band aids or school supplies, potable water. Luxury goods are associated with an income elastic ity of demand greater than one, as income rises, the demand increases by much more than the increase in income. Luxury goods are often large investments that require a large income to consider. Sticky goods reflect no change in demand due to changes in inc ome, this might be due to imperfect information, or sticky preferences, indicating a response lag in demand for these goods with changes in income. If care work is a necessity good, the income elasticity of demand will be inelastic, indicating relatively

PAGE 30

Miller 25 i nelastic demand curves for labor, indicating a small increase in the wages of care workers will reflect a negligible decrease in the quantity of labor demanded. Domestic Production of Care vs. Market Production of Care Domestic Production of Care vs. Market Production of Care The domestic production of care inf luences the market production of care. Those willing to give care free are usually not considering this work volunteer' work, yet it cannot be said that every mother eagerly washes the dishes or responds to the cries of a disabled parent through pure enjo yment. One can say that they are responding to feelings of obligation and duty. 37 Another consideration is the internal investment benefits derived from performing a practiced skill. "The process of learning a skill often goes hand in hand with developin g a preference for exercising that skill, so that by selecting people with the skills for a job, employers unwittingly select more of those who would take the job for lower pay than others." 38 (Himmelweit, 44) The labor force participation decision is an important consideration as the majority of caregivers are women, and women supply the bulk of domestic inputs, in a way, already choosing to specialize and hone the skills required to give care, whether formally or informally. This pre market care giving seems to signal to employers that they will accept care giving positions for lower wages, as they might already experience a non pecuniary reward for performing care, having developed the skill as per social norms in the process 37 Volunteer work is defined as "tasks performed without direct reward in money or in kind, that mainly benefit others rather than the individuals themselves or their immediate family."(Blau et al, 57) 38 Paula England and Nancy Fo lbre demonstrate the preference formation as linked to human capital formation in a 1997 paper Reconceptualizing Human Capital' Presented at the annual meeting of the American Sociological association.

PAGE 31

Miller 26 of growing up. However, the cultural construction of intrinsic rewards prohibits a dynamic view of changing amenities associated with performing care work. If the intrinsic rewards for performing care work begin to diminish, (as with most service jobs, the novelty of performing serv ice work will decrease with time 39 ,) the worker might find the costs of changing professions too high, and the lack of vertical job opportunities in care professions give little opportunity for vertical job mobility. The advantage for women in care giving s hould give them a wage premium over men participating in the care industry, if indeed the status as woman's work can be applied to care labor in the market. Some of the tasks done around a house are non negotiable: the occupants need to be fed and cloth ed; there is a certain standard of maintenance/cleanliness for domestic living that can be quantified into hours of labor and capital. A high income elasticity of demand for a homemaker's services indicates that as income rises, families tend to pay others to do their housework; demand increases for homemakers services. Families with lower net income will spend a smaller percentage of their income on non essential services, for some, these non essential services include contracting for care. For others, hir ing a professional service to monitor a dependent is necessary, in order to be active in the labor market. For an analysis of wages for care services, I expect supply to be wage responsive; I expect the supply of caregivers to rise as wage levels increase. 39 See Steinberg et al. (1986) for an analysis of New York State civil service jobs, which show caring social skills have negative returns.

PAGE 32

Miller 27 Human Capital Theory and the Supply of Care Labor Human Capital theories consider in more detail the productivity of workers, attributing productivity, perceived and otherwise, in the market place to the investments individuals make to improve themselve s. These investments act as signals to employers, allowing them to distinguish applicants into potential productivity hierarchies. These signals include cognitive ability, education and labor market experience. Human Capital is any form of investment in on e's self that increases a person's well being. Human capital investments presumably lead to increased productivity in market or non market activities, or an increase in satisfaction with time spent in market and non market work (leisure) which usually mani fests in a higher return to the activities one chooses to engage in. The human capital model suggests that more goes into a career choice than availability of options. Women and men have different returns on investments. Women see a lower return on educat ional investments past high school explained by a preference for discrimination by employers as per the Human Capital Model. It is useful to consider, in calculating investments, the present value of the stream of payments generated by an investment. The formula below represents the present value of an investment, which demonstrates a diminishing return to investments over time. Equation 1 : Present Value for the Return on Investments

PAGE 33

Miller 28 PV = C + B 1 ( 1 + r ) + B 2 ( 1 + r ) 2 + ... + B T ( 1 + r ) T Where B= the opport unity cost of not entering the job market (as proxied by the difference in earnings annually between college and high school graduates) C=total costs associated with investment (incurred only while participating, therefore only in year 0) r = the discount r ate T=the amount of years The rate of return for an investment, graphically, shows the marginal utility for each additional year in the labor force. After a certain point, regardless of education and experience, there will be diminishing returns in wages from additional education. It is quite often the case that women see a lower rate of return on their investments than their male counterparts do. 40 Figure 4 : Lifetime Return to Education Investment Decision The values d erived from human capital investments divide into three categories: investment benefits to the individual, consumption benefits to the 40 Blau, Francine D., Marianne A Ferber,.; Anne E Winkler,. 2010. The Economics of Women, Men and Work. 6th ed. Upper Saddle River, NJ: Prentice Hall.

PAGE 34

Miller 29 individual, and external benefits to society. 41 Investment benefits to the individual garner higher wages in the future d ue to increased productivity; investment benefits have an exchange value, a quantifiable price. Adam Smith identified only two types of value, exchange value, what a good or service culls in the market, and use value, the value to an individual regardless of exchange value. 42 The investment benefits for becoming more productive are intuitive, an employer is more likely to hire an individual signaling a higher MRP, than a less efficient worker, giving market leverage to workers who invest in them(selves) and increase their productivity, as compared to those who do not. Microeconomists have further developed the values derived from a good or service with utility functions, which allow a consumer to maximize his or her consumption based on exchange values and u se values. Continuing on the value derived from investment, the external benefits to society, from an initial human capital investment, are the positive externalities of the investment. A well educated public is a positive externality of human capital inve stments. Human capital theorists rarely bother with the consumption benefit from investment, the benefit garnered being enjoyment via consumption of a good or service, the pleasure in being active, and the future pleasure of healthiness. Intrafamilial Car Intrafamilial Car e Work Decisions and the Public Market e Work Decisions and the Public Market for Care for Care 41 Strober, Myra H. "The Application of Mainstream Economics Constructs to Education: A Feminist Analysis," in Feminist Economics Today: Beyond Economic Man. Eds. Marianne A. Ferber and Julie A. Nelson. University of Chicago Press. 2003. 42 Strober, Myra H., "The Application of Mainstream Economics Constructs to Education: A Feminist Analysis," in Feminist Economics Today: Beyond Economic Man. Eds. Marianne A. Ferber and Julie A. Nelson. University of Chicago Press. 2003.

PAGE 35

Miller 30 Until the 1960's neoclassical economists had not focused on interfamily decision making preferring instead to consider the family as what Frances Wooley calls an economic "black box." 43 Gary Becker, one of t he University of Chicago's new home economists,' gave credence to joint utility functions, regarding human capital investments as dependent on the type of household they take place in. In his seminal work Treatise on the Family', Becker (1981) was the f irst to address intrafamilial decision making. However, several of his basic assumptions have been questioned. 44 Becker assumes a unitary model of the family, indicating that each family is an undifferentiated unit attempting to maximize household utility f unctions subject to an aggregate budget constraint, pooling resources and deriving one homogenous input, with a linear utility function tied from the output, to each spouse. One major attack is in consideration of one of his three auxiliary assumptions, th at the head of the unitary household be an altruistic male. 45 (Bergmann 1995, Pollak 1978, Nussbaum 1995, Sen 1998) However, this unitary view of the household neglects analysis of conflict and bargaining models, diminishes the unequal distribution of power /wealth/responsibility within families and provides no path for answering the disproportionate share of domestic labor often enacted by women and children. The human capital model, and Beckerian efficiency, are remarkable contributions to field of home' e conomics and have given much source material for feminist economists to begin to 43 Beneria, Lourdes. 2003. Economic rationality and glo balization: A feminist perspective. In Feminist economics today: Beyond economic man. eds. Marianne A. Ferber, Julie A. Nelson. 2nd ed., 115 133. Chicago: University of Chicago Press. 45 Jacobsen, Joyce P. The economics of gender Wiley Blackwell

PAGE 36

Miller 31 comment on the discipline at large. 46 The human capital model is important to include in our analysis of caring in the market because the variables affects supply side factors for female dominated professions. According to this model, there will be disparities in human capital accumulation between men and women, and women will be less efficient in terms of market production than their male counter parts. This hypothesis is not n ecessarily contrary to what specialization and comparative advantage analysis would suggest: that since women tend to specialize in domestic care, participating in the formal care market would make them more effective and thus increase their wages relative to men's. This thesis questions this conclusions by arguing that women will be more effective in giving care, however, the higher status assigned to jobs traditionally occupied by men, over the status of domestic chores, including that of care giving, a ffects the status of care giving in the public sphere, essentially lowering the wages for those doing women's work. The notion of 'devaluation by association' helps to explain why, perhaps, the importance of skills culturally coded as female, are not consi dered worth greater compensation. Females participating in care giving services should be paid more than men participating in care giving services, according to theories of comparative advantage, since women are often socialized to begin specializing in i nformal care services at a young age. However, compared to other occupations, the compensation for direct care services will be lower due to unacknowledged forms of human capital accumulation taking place, that 46 Becker is sometimes considered an economic imperialist' for extending the reach of economics into other academic disciplines, not always a welcome entrance. (Pollak 1988)

PAGE 37

Miller 32 is, domestic sphere specialization is not usu ally considered as a worthy economic investment, therefore, direct care occupations will be compensated according to standard views of human capital investment, lowering wages for those in care occupations. Care Worker Wage Differentials in the Human Capi Care Worker Wage Differentials in the Human Capi tal tal Model Model Wage differentials according to the human capital model occur from differences in investments in human capital, such as education, on the job training and work experience. It promotes the idea that the individual choices we make are dir ectly reflected in the wages we receive. Wage rates reflect returns to human capital investments. An investment in education is only worth making if the total cost (direct expense of education plus the wages foregone through absence from the labor market) is less than the gross benefits. (Refer to Figure 4) Those with higher education are presumably able to be workers that are more productive; those who are likely to stay in the labor force are rewarded for their experience with higher wages. Projections fo r years spent in the labor force are often lower for women. Employers might withhold training or promotions, citing a lack of commitment to the labor market due to absences for reproduction, pursuit of marriage and/or specialization in homemaking. If care labor were unskilled, than those with higher education would not go into it, as they are eligible for skilled labor, which should have higher returns. Using the standard model to explain wages would include the independent variables sex, age and experienc e. This

PAGE 38

Miller 33 theory is ambiguous about how long lasting the effects of participating in a female dominated profession will be. Crowding Hypothesis Barbara Bergmann's Crowding Model refers to the bottlenecks which occur when there are barriers to entrance for certain occupations and less desirable occupations become overcrowded due to rejections based on discrimination. 47 If discrimination in the market exists, then the standard model will not accurately respond to equilibrium points in the supply and deman d for labor. If employers have a preference for discrimination, they might reserve certain jobs for men, meaning that as more women enter the market and vie for these positions, very few of them get hired, crowding them into less prestigious occupations. T he following index of segregation neatly supplies evidence for occupational segregation. Table 4 Occupational Segregation in 2008 by selected Occupations 47 Bergmann, Barbara R. 1974. Occupational segregation, wages and profits when employers d iscriminate by race or sex. Eastern Economic Journal 1 (2) (Apr.): pp. 103 110, http://www.jstor.org/stable/40315472 Occupation Total employed (1,000) Percent of total Female Black Asian Hispanic Total, 16 years and over 139,064 47.2 10.8 4.8 14.3 Management, professional, and related occupations 51,743 51.5 8.4 6.1 7.3 .... Computer and mathematical occupations 3,531 25.8 6.7 16.1 5.5

PAGE 39

Miller 34 The wage levels in those occupations with an oversupply of labor will be lower due to the ea se with which replacement workers are willing to work for less. Health agencies and other care service employers have incentives that keep wages low and maximize their profits; if the supply for feminine occupations is artificially inflated due to crowding out of higher paid occupations, there should always be a worker willing to work for the wages offered. Here, occupational segregation does indeed imply lower wages. For a model to incorporate Bergmann's crowding hypothesis, wages must increase in tandem with demand if there are not barriers to entrance for higher paying occupations; if demand for care labor goes up and other female workers are readily available, wages will stay relatively the same. If turnover ....Architecture and engineering occupations 2,619 12.9 5.2 9.0 6.8 ....Community and social services occupations 2,337 64.2 19.3 3.3 9.8 ....Legal occupations 1,716 48.8 6.5 3.4 5.5 ....Education, training, and library occupations 8,628 73.8* 9.4 3.8 8.0 ....Healthcare practitioner and technical occupation s 7,805 74.3* 10.8 7.8 6.2 Service occupations 24,634 56.8 15.3 4.9 21.3 ..Healthcare support occupations 3,332 88.9* 25.5 4.1 15.2 ..Protective service occupations 3,289 21.4 17.8 2.4 13.3 ..Personal care and service occupations 5,024 78.3* 14.8 7.8 14.6 Sales and office occupations 33,433 62.9 11.3 4.2 12.6 ..Office and administrative support occupations 18,047 73.9* 12.5 3 .6 13.4 Natural resources, construction, and maintenance occupations 13,073 4.6 6.7 2.0 25.0 ..Farming, fishing, and forestry occupations 987 23.5 5.2 1.9 41.8 ....Graders and sorters, agricultural products 103 67.8 9.2 7 .3 50.3 ..Construction and extraction occupations 7,175 2.6 6.1 1.4 29.1 *Jobs with more than 70% Female work force.

PAGE 40

Miller 35 costs in the industry are low enough, the hi gh attrition rate will not be a boon to home care agencies and they will persist in profiting from hiring low skill employees and maintaining a preference for discrimination. It is unclear whether the pecuniary penalty for care work occurs because of an over inflated supply of labor due to institutional undervaluation of feminine work, or whether the intrinsic rewards act as an adequate compensating differential, or if care giving is viewed as a "natural activity that does not deserve or require enumerat ion." (Himmelweit, 46) Dual Labor Market Theory According to dual labor market theory (Dickens and Lang, 1985) for the most part, occupations can be split into two groups: "One with high wages, good working conditions, stable employment, rewards for edu cation and job experience and opportunities for advancement (primary sector) and one with low wages, bad working conditions, unstable employment, no rewards for education or job experience and no opportunities for advancement (secondary sector)." (Dickens and Lange 1985) Dickens and Lang show models telling imperfect representations of reality, meaning the markets responses are not necessarily moving towards competitive equilibrium. If barriers to entrance are being maintained via discriminatory preferenc es, then even the most basic competitive equilibrium, in which supply equals demand, cannot be reached without serious epistemological implications. How do employers determine what signals greater productivity? And if undervaluation of some traits over oth ers, regardless of their relevant efficiencies in the market plays a role in this decision, how our practical identities play a part ought to be considered. This

PAGE 41

Miller 36 empirical study of minorities and their corresponding occupational sectors finds that not eve ryone who is qualified for a primary job and wants one will be able to attain it. Moreover, minorities and women are involuntarily confined to secondary sector jobs, consistently. A preference for discrimination may be evident in dual labor market theorie s but also has even broader implications. The normative rigidity with which appearance based identity stereotypes perpetuate is evidence that our cultural institutions and knowledge systems are also entangled in a bias towards male traits over female trait s. It is unchanged that women and minorities are paid less for every dollar paid to a white male, and occupational segregation may very well work in tandem with dual labor market theories. If there is a penalty for caring, than it very well may manifest be cause of an undervaluation of care work that supports a wage premium for those not participating, those working in socially valued occupations will not suffer the costs of participating in woman's work. Gender essentialism means judging a person on many le vels based only on knowledge of gender stereotypes. It is a convenient demarcation signal for employers, and the fact that gender serves as a signal is not remarkable in and of itself. Surely, if the middling zone between the primary and secondary sectors continues to increase a sustainable direct care work force will dissolve, and contracting out for a large quantity of low paying care services will most likely not be able to answer a demand for quality care. The isolated patterns of mobility that allow so me individuals access to primary sector jobs, and deny others, have implications for how potential lifetime wages depreciate for those choosing to enter and stay in the care market. (Dickens and Lang, 1985)

PAGE 42

Miller 37 The Crowding Model and Dual Labor Market Theori es set the stage for an important consideration: is care work unskilled labor? Unskilled labor has evolved to become synonymous with homogeneity of the final product, meaning anyone could be generally as competent as anyone else at a given job could. Consi der burger flipping at a fast food restaurant, truck driving etcetera; these occupations require hardly any training, and the outcome of unskilled labor does not depend on the person doing it, so long as they do it at all. Hedonic pricing builds on the ide a that different characteristics of a good or service impact the pleasure of consumption differently. If care giving were unskilled labor, the pricing of the services would have little variance between individuals doing the labor. The low wages of the occu pations would be inherent to the position, in particular that mostly any unskilled worker would produce equally satisfactory results in providing support and care for a care recipient. One may attribute this to the undervaluation of care labor due to the m ethod of pricing based on culturally constructed ideas about care giving and its intrinsic rewards, or crowding effects in the market. Men's work in the public sphere has usually enjoyed higher status than women's domestic work in the family circle. 48 If t his assumption holds true, so will my hypothesis that engaging in women's work will retain its inferior status in the public sphere, evidenced by a wage penalty for exhibiting altruism in the market. 48 Blau, Francine D., Marianne A Ferber,.; Anne E Winkler,. 2010. The Ec onomics of Women, Men and Work. 6th ed. Upper Saddle River, NJ: Prentice Hall.

PAGE 43

Miller 38 3. Why Care? Defining Care Labor Care work has a long h istory as a female dominated occupation. 49 Its emergence as a profession solidified its position as a sex segregated occupation. 50 The content in the next chapter will require a basic understanding of gender roles. Most academic disciplines have accepted var ying distinctions between gender and sex: gender is culturally constructed; traits we align with femininity' have little to do with biological processes. Sex is determined by which reproductive organs dominate. 51 Gender is a way for individuals to categor ize the world and easily label and draw assumptions about the individuals they meet. Gender socialization begins at a very young age, and refers to the process through which the socially constructed differences between boys and girls are reinforced. Those claiming that gender roles are inherent expressions of our biological natures are generally accepted as mistaken. 52 49 Sen, Amartya. 1979. Personal utilities and public judgements: Or what's wrong with welfare economics. The Economic Journal 89 (355) (Sep.): pp. 537 558, http://www.jstor.org/stable/2231867 50 Nelson, Julie A. 2003. Separative and soluble firms: Androcentric bias and business ethnics In Feminist economics today: Beyond economic man. eds. Marianne A. Ferber, Jul ie A. Nelson. 2nd ed., 81 99. Chicago: University of Chicago Press. 51 Sex is determined by chromosomes, hormones, psyche, and external and internal genitalia Segen's Medical Dictionary. 2012 Farlex, Inc 52 Nelson, Julie A. 2003. Separative and solubl e firms: Androcentric bias and business ethnics In Feminist economics today: Beyond economic man. eds. Marianne A. Ferber, Julie A. Nelson. 2nd ed., 81 99. Chicago: University of Chicago Press.

PAGE 44

Miller 39 As of 2008, women continue to make up the majority of those working professionally in care labor. 53 Nancy Folbre defines caring work as that which presupposes a caring motive: "undertaken out of affection or a sense of responsibility for other people, with no expectation of immediate pecuniary reward ." (214, emphasis mine) I will use that definition for caring work which occurs regardless of t he boundaries of the market, and care labor will be labor undertaken with expectation of pecuniary reward, and potentially, out of affection or a sense of responsibility for others. 54 Members of families and friends enact the most common forms of care work, and unpaid care remains the primary form of care for many individuals who do not contract out for assistance in the home. 55 Caring work is associated with a bucket of feminist patter such as, "family labor" "unpaid labor" "reproductive labor" "sex affectiv e production" or "social reproduction". All of these terms refer to a "separate sphere", be it the home, the womb, the school. Care giving can occur within and throughout these spheres, in the family and outside, paid and unpaid. This is why the distinctio n between paid and unpaid work is not the same as caring and uncaring work. We can contract out for a caregiver through other means than traditional market exchange. Some jobs require emotional acting, for example, 53 PHI. Nov, 2008. Direct care jobs and long term care: Unt apped engine for job creation and economic growth. Bronx, NY: Paraprofessional Healthcare Institute Quality Care through Quality Jobs, 2. 54 Folbre, Nancy, and Nelson, Julie A. For love or money -or both? in JSTOR: The Journal of Economic Perspectives, Vol. 14, No. 4 [database online]. Autumn, 20002012]. Available from http://www.jstor.org/stable/2647078 55 England, Paula, and Nancy Folbre. 2003. Contracting for care In Feminist economi cs today: Beyond economic man. eds. Marianne A. Ferber, Julie A. Nelson. 2nd ed., 61 79. Chicago: University of Chicago Press.

PAGE 45

Miller 40 employees at a popular fast food restaura nt are trained to respond "[It] is/was my pleasure," when serving customers. 56 While others labor begrudgingly in social service offices, or at volunteer organizations, discourteously working to do good'. Caring actions and their motivations create the di stinction between caring motives and caring labor. The fast food service employee may not actually derive pleasure out of serving, as it is a scripted act of caring, meant to make the customer (here, the care recipient) feel cared for. While the social ser vice worker may be curt, over worked and impolite, the consequences of their actions in this field, speak to a caring motive, which does not equate to immediate consumption by those they interact with on the job. That care work can be performed externally and not be motivated by intrinsic altruistic preferences suggests that the motivations for a paid and unpaid care giver are important to consider for the future of long term care labor supply. Would it help to increase the much needed supply of caregivers if fewer individuals provided the service free? Intuitively, if more and more individuals are turning to the market for care services, the opportunity costs to those supplying the labor sans payment will increase, and they will become wage seekers. 57 This n eed not assume perfect self interest in the family and in the market, and in fact, altruistic preferences motivate some to 56 Leidner, Robin. 1999. Emotional labor in service work. In Emotional labor in the service economy. eds. R. J. Steinberg, D. M Figart, 61. Sage Publications. 57 When the opportunity cost of performing unpaid care work rises: there has been a shift in demand to market services, necessitating a relative increase in supply, if supply is relatively constant than wages should go up to attract more laborers. It can also indicate a supply side change, the opportunity cost for performing free labor may rise as the service becomes more rare and excludable. It may also indicate a greater efficiency at a task due to specialization or human c apital investments.

PAGE 46

Miller 41 care at a very high market price. Some economists reasonably express concern that since caring has been a private and even a purely s ocial service it has no place in the market, and is likely to be tainted through its commoditization. 58 Concepts of fairness in the distribution of domestic (unpaid care) labor are relevant, as supported by an application of specialization versus threat po int models in household time allocation between spouses. (Bittman et al, 2002) Gender constraints are placed on the distribution of labor in a household, regardless of the relative bargaining power of either spouse. The results that Bittman et al find do n ot lead to inherent biological expressions of preference, the specialization that results indicates a resistance to male participation in feminine tasks, not because they are unable to do them as a result of their sex. When the female partner reduces her h ours of unpaid work as her income rises, she can do nothing to increase her male partners' hours of unpaid work. In fact, when the woman's income rises above her male partner, her amount of unpaid work actually increases, as mitigation for gender deviance. 59 However, if demand rises and there is a lag in market response linked to time and consumer expectations, the wages for caregivers will not rise quickly enough to meet the increased demand. 60 Another distinction between paid and unpaid care services lies in the motivations of the caregiver herself, and these are often embedded within 58 Bittman, Michael, Paula England, Nancy Folbre, Liana Sayer, and George Matheson. "When Does Gender Trump Money? Bargaining and Time in Household Work" American Journal of Sociology 109.1 (2003): 186 214. 59 Bittman, Michael, Paula Engl and, Nancy Folbre, Liana Sayer, and George Matheson. "When Does Gender Trump Money? Bargaining and Time in Household Work" American Journal of Sociology 109.1 (2003): 186 214. 60 National Alliance for Caregiving. 2011. Caregivers of veterans serving the homefront United Health Foundation www.caregiving.org

PAGE 47

Miller 42 social norms. Long term reciprocal expectations, altruism and obligation or responsibility are motivating factors for informal care giving. It is important to note that these motivations also hold true for many engaging in formal care giving. The ease of transfer for motivating factors seems to indicate that those doing unpaid care labor will more readily withhold services if social pressure to fulfill informal care duties dec reases. 61 Contracts for Direct Care Services Contemporary labor is defined by the provision of a contract between employer and employee. However, I will be using the term contract' in a much broader sense. Contracts indicate two things, first, a condition al reward and second, a mutual expectation that fulfilling is essential for payment. By this definition, legal contracts are not necessarily those with paper trails. Many problems arise when verbal agreements between the intermediaries employing the caregi ver to are not communicating, or failing to bear witness to the effect of the care giving. While many agencies formalize agreements for contracting out their employees, often the care recipient and his or her family will have different expectations of how the care will manifest. The cost of receiving care services is very high, but the wages paid to care givers is very low. 62 This is the result of agencies paying care workers low wages yet charging two or three times more to the care recipient. 63 61 GROSS, JANE. Employer provisions for parental leave (EXCERPT), monthly labor review online, oct. 1989 in New York Times [database online]. New York. 62 Better Jobs Better Care. May, 2005. Family care and paid care: Separate worlds or common ground? Washington D.C.: Better Jobs Better Care, 5. 63 Gross, JANE. Home health aides: What they make, what they cost NYTimes.com December 30, 2008, 6:00 am2012]. Available from

PAGE 48

Miller 43 Direct Ca Direct Ca re Work as a Good or Service re Work as a Good or Service Care labor is not a category historically found in the census codes. Nevertheless, the social implications of caring' are deeply embedded in our ideas about selflessness and obligation, or responsibility to others; the marke t sphere attempts to separate social commitments from monetary compensation. All of these things contribute to an unclear understanding of how to categorize direct care services as a service. In economic terms, how do consumers decide whether or not to pu rchase care? The last 50 years have seen a rapid shift from the private to the public sphere in terms of what families and individuals are willing to pay for, and what they are willing to do themselves. 64 The comfort of recovering from injury in ones' home, in addition to high costs of hospital stays has contributed to an increased demand for direct care. Broadly, the commoditization of tasks usually performed in the domestic sphere has contributed to a growth in the labor market supply, and in particular, a rise in the female labor force participation rate. 65 What kind of good or service is care giving? The effects of care work are similar to that of a public good,' that is, non rival and non excludable. The positive externalities of care work can be rea ped by anyone, but when one person receives care, that same care cannot be enacted onto/by another. Care giving produces stable and healthy citizens more able to participate in http://newoldage.blogs.nytimes.com/2008/12/30/home health aides what they make what they cost/ 64 PHI. Nov, 2008. Direct care jobs and long term care: Untapped engi ne for job creation and economic growth. Bronx, NY: Paraprofessional Healthcare Institute Quality Care through Quality Jobs, 2. 65 Better Jobs Better Care. May, 2005. Family care and paid care: Separate worlds or common ground? Washington D.C.: Better Job s Better Care, 5.

PAGE 49

Miller 44 society in a beneficial way. The effects of a caring family are well documented to have positive influences on the future of an individual. The social externalities of care work have potential spillover effects onto all those who in the future will encounter the individuals receiving care (England and Folbre 1999, Wax 1999). James C oleman (1988) is a sociologist focusing on social costs diminishing because of good care, showing that unstable family life increases the costs to society in the form of trials and crime. Robert Putnam (2000) calls the positive externalities of care work s ocial capital. Social capital affects everyone participating in a community, it is good when those who need care are cared for, jobs are created, economic activity is uninhibited, and those able to participate in democratic activities are more able to make informed decisions, presumably. 66 Calculating the value of non market activities is a trying task. The labor inputs for consideration in valuation of non market activities are the hours and cost of the human and business capital invested into a task. The more rare and excludable a service, the more willingly an individual or firm will pay for that service. It is important to determine whether society is moving in a direction in which care services will become excludable. Often not explicitly compensated, emotional and caring work is difficult to compute. If occupational segregation does in fact lead to lower wages for those participating in female jobs, then no longer specializing in unpaid care giving would certainly make the paid service more rare and ex cludable, for market consumption. Excludable here meaning that not 66 Hochschild, Arlie Russell. 1983. The Managed Heart: Commercialization of Human Feeling. Berkeley: University of California Press.

PAGE 50

Miller 45 everyone can reap the benefits. Will access to care labor depend on income levels, will social provision increase through policy decisions affecting what Medicaid offers by way of in home c are provision? That care services become more excludable is not ideal, however it may lead to a fair restructuring of gender norms. In this way, the cultural expectation that women will be inherently welcoming and maternal, preferring to engage in what Ar lie Hochschild refers to as emotional labor, can be dismissed. Once the expectation that females perform their gender by performing care diminishes, inclusion in care service specialization could disentangle from gender roles. For men to begin to speciali ze in care giving as a viable occupation, the wages or social prestige of care giving would have to go up. How to Measure Care Services in the Market How to Measure Care Services in the Market Worker productivity in the direct care market is often unenforceable care givers work to alleviate the burdens of the care recipient's family, while also satisfying the needs of the care receiver. Anecdotally, care recipients discuss the desire for more time, more of a quality relationship with their care provider, however, the care worker must account for visiting many homes in a day. Worker productivity could be defined in terms of how many clients the care worker treats, or by the utility the care recipient receives from the paraprofessional training versus that of an untrained family member. The emotiona l output of a caregiver can be dynamic or static depending on the state of the client and her relationship with them. A caregiver contracted to complete four home visits in a day is a reasonable expectation, but does not always account for the instabilit y of

PAGE 51

Miller 46 health and care requirements. Some homes may obligate her to consistently perform tasks that are more difficult for, while the same pay rate is in place. If a client with whom she has forged a very close relationship requires her services once she has gone, she is more likely to feel the tug of an implicit emotional contract, rather than the explicit pecuniary contract, even though, perhaps her explicit contract will only compensate her for two hours in each household. These scenarios highlight the nat ure of care work as a complex good. 67 As is the case with customary pay per performance contracts (where compensation exists as a direct linear function of production), worker productivity is usually measured by the additional, or marginal product contribut ed per additional individual. While care workers' contracts remain tied to an agency model of compensation, quality of care services will remain ambiguous and wage compensation will not adequately reflect the skill level of what actually transpires. An in teresting tension arises with pay for performance contractual commitments and care laborers, specifically when they are more motivated to care less by care giving' more. An overworked care worker may seem very productive when engaging in this brand of "dr ive by" home care; however, it has negative impacts on consumers, providers and workers. The Costs and Benefits of Care Work The impact of care giving can have distinctive costs and benefits depending on whether it is a family, public, or priva te care provider. In the 67 J ochimsen Maren A. 2003. Integrating Vulnerability : On the Impact of Caring on Economic Theorizing." In Toward a Feminist Philosophy of Economics. Eds. Drucilla K. Barker, Edith Kuiper. London ; Routledge, 2003.

PAGE 52

Miller 47 case of family care provision, the benefits are clear. The care given by a mother to a child produces a nurtured citizen who is better prepared to contribute positively to those around her. While society does not shoulder the costs of being a good parent, they reap the benefits of the mother's emotional labor. This is, overall, beneficial for society in a way that does not threaten scarce natural resources. The perceived net cost is nothing. Many might assume that caring feelings p rovide enough motivation to do caring work sans monetary compensation, and the utility derived from caring for a family member is fair compensation. 68 However, there is a tipping point at which the amount of care expected of one person exceeds what they can comfortably give without compromising their pursuit of other life goals! 'Emotional labor,' which is often expected of female workers but almost never compensated, refers to the expectation of women to act a certain way, due to "inherent female feelings of warmth, empathy, and connectedness." 69 Often, emotional labor requires emotional acting, emotionally responding through gestures, facial expressions, words and messages of caring or compassion. Often, individuals are penalized for not acting on the job; in particular, women are expected act in ways society associates with maternal instincts. In one study, student evaluations of professors at a small college reflect differing expectations. Female professors received very harsh evaluations when holding str ict to deadlines for not being sympathetic, and if a female professor did not discuss her personal life, 68 Folbre, Nancy, and Nelson, Julie A. For love or money -or both? in JSTOR: The Journal of Economic Perspectives, Vol. 14, No. 4 [database o nline]. Autumn, 20002012]. Available from http://www.jstor.org/stable/2647078 69 Hochschild, Arlie Russell. 1983. The Managed Heart: Commercialization of Human Feeling. Berkeley: University of California Press.

PAGE 53

Miller 48 she was considered 'cold'. Meanwhile, the male professors engaging in the same behavior were considered respectable, 'not to be messed with', principle d or stoic. Many worry that paid care services will dehumanize and subvert the function of the family' as primary care and support. 70 Opposing this belief, it follows that where labor has access to the market, given that there is a market for that kind of labor, the industry grows. 71 Barbara Bergmann argues "anything that romanticizes housework and childcare is bad for women; these are forms of work that should be "industrialized", because they can be more performed efficiently' outside the home." 72 (Fol bre, p. 225) Deductively, were caring activities more acceptably outsourced, the economy would grow and quality of care would increase, as did quality of other goods once the production of them entered the market, such as the privatization of education, though issues of accessibi lity are important to consider. 73 As Nancy Folbre echoes, "there is a sharp division of views about whether markets, caring feelings, and caring activities are all at odds with 70 Folbre, Nancy. 1995. "Holding hands at midnight": The paradox of caring labor Feminist Economics 1 (1): 73 92. 71 Neo classical economics advocates for perpetual growth, so, growth in the market in a particular area is ve ry good for normal goods. This is why attempting to categorize care services as a good is important for future policy decisions regarding care labor. 72 Folbre, Nancy. 1995. "Holding hands at midnight": The paradox of caring labor Feminist Economics 1 (1): 73 92. 73 In America, the quality of schooling is directly related to its status as a public' or private' school, it is considered common knowledge, among some, that private' schools offer a better education. While those participating in a pu blic school education are not penalized, there exists a market for individuals willing to pay a higher price for what they consider a better education. Whether or not public or private schools actually offer a better education is up to debate, however, tea chers are paid more money at private schools, and higher quality teachers vie for these positions as a result.

PAGE 54

Miller 49 each other." 74 While it is nice to understand this div ision, it is up to debate whether to reconcile these odds, or whether instead to encourage and support the provision of caring work at a socially optimal output quantity, in which the marginal benefit of an additional compensated care laborer is equal to t he negative externalities' associated with such a task. 75 In this way, compensation will drive the quality and quantity of care provision up, while making the industry more attractive. Interface between Theories of Care and Economics The pooling of fem ale workers into several industries that are corollaries to the tasks we code as feminine remains a topic of debate among economists (Bergmann 88, Folbre 93, Blau 99). This is called occupational segregation. Unfortunately, feminist theorizing has only jus t begun to converge with theories in economics and a review of this discourse is necessary to understand its trajectory into the economic realm. Among many of the questions different economic camps proffer, "does occupational segregation imply discriminati on, if so how and why?" capture the spirit of inquiry for this exploration. The Tradition of Care Labor 74 Folbre, Nancy. 1995. "Holding hands at midnight": The paradox of caring labor Feminist Economics 1 (1): 73 92. 75 An externality is the consequence(s) of an action, market or non market, which are a derivative of that event. An example could be that pollution is a negative externality of fuel consumption.

PAGE 55

Miller 50 Many families remain the primary caregivers for older relatives. About one out of every four working adults is also caring for an older family membe r or friend. 76 "More than 65 million people, 29% of the U.S. population, provide care for a chronically ill, disabled or aged family member or friend during any given year and spend an average of 20 hours per week providing care for their loved one." Car egiving in the United States; National Alliance for Caregiving in collaboration with AARP; November 2009 This kind of informal care giving can be physically, emotionally and financially draining. 77 Some of the options for those seeking assistance giving c are to their relatives include adult day services, adult home help, assisted living, adult foster care, licensed and unlicensed assisted living, nursing homes and Hospice. 78 Deciding whether to hire out or do it oneself depends on ones' income. As income go es up, it may be reasonable to predict that consumption of care services will increase, if care services are a normal good. How do we make that decision? The costs of caring directly must be greater than the costs of hiring out. These costs apply directly to the care recipient if they handle expenses independently, or to the family member or loved one who handles these expenses. The consideration this third party gives to the 76 Fremstad, Shawn. Implementing the coverage provisions of health care reform: Wh at's at stake for direct care workers | russell sage foundation in Direct Care Alliance [database online]. 2010 [cited 4/20/2012 2012]. Available from http ://www.russellsage.org/research/reports/care work health care reform 77 National Alliance for Caregiving. 2011. Caregivers of veterans serving the homefront United Health Foundation, http://www.ca regiving.org 78 National Alliance for Caregiving. 2011. Caregivers of veterans serving the homefront United Health Foundation, http://www.caregiving.org

PAGE 56

Miller 51 costs and benefits of unpaid self provision of care versus paid provision of care services depends on four factors: ability, expense, relationship and burden. The ability of the contractor to perform the labor depends on his or her own health, geographic distance, skills, training, and willingness to learn how. The expense can come dire ctly from the care recipients' pocket, if they have money saved up or if the contractor has access to it, or from the third parties pocket. The total expense is equal to the opportunity cost of caring versus other market work, the initial investment in mat erials, moving costs and supplies for the home, plus the hours of labor given to the task itself. Often times, family members receive more internal or social benefit from caring for their loved ones, but this is contingent on the relationship they have wit h the care recipient. Direct relatives or those who live with the care recipient might find it appalling to hire an aide, or be more willing to hire given the greater quality of care. It would be more costly to hire out if caring for the care recipient w as worth' your time and money. The burdens that caregivers' shoulder are: stress, fatigue, emotional or verbal abuse from the care recipient and heavy lifting. 79 Sometimes these tasks are in addition to household chores and errands. Depending on the conven ience or inconvenience of performing these chores, the burden could be light or little. The third party who often does the contracting for paid services must weigh the pros and cons of the decision to contract out. The direct and indirect costs of shoulder ing the burden of a dependent family member or friend are numerous. However, the changing 79 National Alliance for Caregiving. 2011 Caregivers of veterans serving the homefront United Health Foundation, http://www.caregiving.org

PAGE 57

Miller 52 face of the home has helped to alleviate some of the tension between the decision to hire out or care for the dependent him/herself. Often an economic analysis takes place, rather than what historically was a familial duty. Hiring out for a care worker also helps to relieve the feelings of guilt and being a burden' to the family member or friend who provides the care services. 80 Sometimes, contracting for care beats th e benefits of performing traditional informal care giving for a loved one who needs it. A Link between Gender and Care Work A Link between Gender and Care Work Unpaid care work typically comes from female kin mothers, sisters, aunts and so on. Insofar as this labor gets done becau se of social and familial networks, many do not see the need or desire to begin compensating for what happens naturally. Often, employers have invisible prejudices against those performing childcare or domestic care, concluding that men are more motivated to work to provide income for their family, while a woman's family will act as a motivation, ostensibly that work provides a distraction away from their family. Paula England finds that the decrease in wage rates for care work can be partially explained by the negative perception of care work, which leads to a negative effect on the laborers' income, net of human capital, skill, demands of the occupations, and other controls. The negative perceptions of caring activities stem from care activities inextricab le association with women and their mothering roles. Farkas and England (1994) find that net of a job's sex composition, there is a negative penalty for 80 Himmelweit, Susan. 1999. Caring labor. In Emotional labor in the service economy. eds. R. J. Steinberg, D. M. Figart, 27. Sage Publications.

PAGE 58

Miller 53 working in a caring occupation. 81 When culturally based discrimination effects hiring practices, and few er high wage opportunities are afforded to minorities and women, these original discriminatory social cues are validated and a negative feedback loop persists. However, the relationship between our actions and our gender is not as rigid as we had thought. "notions of proper gender behavior are quite flexible, gender segregated service jobs reinforce the conception of gender differences as natural. The illusion that gender typed interaction is an expression of workers' inherent natures is sustained, even i n situations in which workers' appearances, attitudes, and demeanors are closely controlled by their employers." 82 Gender is used as a signal to privilege some people over others, and not necessarily some types of behavior over other types of behavior. Re gardless of what men or women are doing in the work place, most actions will be justified in terms of acting out our natural gender. I hypothesize the practice of using pre existing social cues to justify actions in the market has negative effects for thos e participating in care labor. Paula England (1980) also finds a negative return to wages for working in caring occupations for both men and women using a dummy variable for care. Kilbourne et al. (1994) along with Paula England, extend multiple regressio n analysis, and develop a scale to measure the amount of nurturant skill associated with different occupations. Kilbourne incorporates other occupational measures to account for care associations, including people skills' and demands for talking and hea ring. 81 Kilbourne, Barbara Stanek, Paula England, George Farkas,, K. Beron, and D. Weir. 1994. Returns to Skills, Compensating Differentials, and Gender Bias: Effects of Occupational Characteristics on the Wages of Women and Men. American Journal of Sociology 100:689:718. 82 Leidner, Robin. 1999. Emotional labor in service work. In Emotional labor in the service economy. eds. R. J. Steinberg, D. M. Figart, 61. Sage Publications.

PAGE 59

Miller 54 With all relevant controls in place, Kilbourne et al. find that working in an occupation with a higher score on the nurturant' scale decreases earnings for both males and females. 83 A Feminist Approach to Economizing Care Susan Himmelweit (2003) d efines a feminist approach to economics as fulfilling the following conditions: An exploration of difference within the approach is necessary, including those between men, women and minorities. The existence of difference must be the foundation of its app roach Must recognize that these differences are "structural that is, dependent on relationships between people in systematically different positions in society."(267) It must be able to explain changes that occur, including those that bridge or expand dif ferences between people. It must consider a broad enough purview of economics and economic change in order to take into account "all factors that have a significant impact on gendered behaviors within the economy."(267) I will consider below two of the f eminist approaches to economics that fit the conditions outlined above. A reasonable background on the dialogue 83 England, Paula and Lori Reid, and B arbara Stanek Kilbourne. 1996. The Effect of the Sex Composition of Jobs on Starting Wages in an Organization: Findings from the NLSY. Demography 33(4): 511 12

PAGE 60

Miller 55 between feminists and the economics discipline is necessary to comprehend the current status of paid care giving. Feminist Constructionists Feminist Constructionists Fe minist constructionists attempt to reveal inadequacies in the field due to gender biases, focusing on literature and models that neglect to repair themselves in light of gender difference. Charlotte Perkins Gilman was one of the first to write on the sexuo economic status quo before the 19th century had completed, and prophetically explicates the social conditions in which her contemporaries, women and minorities, continue under unfair duress in the centuries preceding her treatise and the centuries to foll ow. There has been a great deal of progress in rights movements into the 21st century and, indeed, conditions in developed countries are greatly improved from issues of servitude and domesticity and those born into white male privilege. However, it would b e a boon to the efforts of those championing equality to declare the fight for equal rights over. The injustices Gilman speaks to still occur with regularity, though much tempered. "Sex has been made to dominate the whole human world, -all the main avenu es of life marked "male," and the female left to be a female and nothing else." 84 A position of power is deceptively attractive; however, a reasonable study of feminism seeks not only the advancement of women, but also the advancement of a society by means of promoting equal access to opportunity. The underprivileged, under developed, over exploited are the demographic that many revolutionaries have spoken for and to, the Marxian labor class and the 84 Gilman, Charlotte. Women and economics : a study of the economic relation between women and me n Amherst, N.Y: Prometheus Books, 1994.

PAGE 61

Miller 56 marginalized other of feminism. The marginalized, other ed and uneducated who fill this position do so without any recourse. Economics provides an excellent tableau depicting living as making a living' and revealing inequalities in terms of efficiencies. Currency is often considered an economic ballot casting v otes wherever consumption occurs, the votes agents in a market cast, or do not cast, help economists identify patterns and trends in consumer consciousness. When a pattern emerges, for instance that one segment of the population consistently has access to higher wages, or a pattern in which another segment of the occupation are barred from high paying occupation, discrimination identifies itself as a market preference' for some over others (Ferguson 1989, Folbre 1994). This market preference for discrimina tion can be accounted for in both the standard neoclassical model and institutional models. However, the interpretation one chooses has important policy implications. An individual attempting to maximize utility, engaging in poorly compensated work, or even labor with no returns at all must by definition enjoy compensation in the form of greater utility (Folbre 1994). These are common standard applications to explain wage differentials under rational utility maximizing conditions. Neoclassical theory do es not allow for undervaluation, it only fails to reward under skilled laborers. Barbara Bergmann's' Crowding Model is a good answer to this dichotomy. When women are crowded into lower paying occupations, because higher paid occupations are being reserved for men, they increase the supply for these low wage occupations, pushing the prices down in a negative feedback loop, below that of competitive equilibrium. (Bergmann 1986, Pujol 1992)

PAGE 62

Miller 57 Affirmative action feminists Affirmative action feminists Affirmative action feminists attempt to improve the representation and status of women in economics, and not necessarily critique the economics discipline itself. Out of house hiring for care services does allow greater freedom for women in terms of financial independence. It also increases th eir threat point and therefore their interfamily bargaining power, which are both the advantages of the gradual shift from family to market in care services. Bittman et al (1994) conclude that despite the inability to increase men's unpaid care work in the home when women's income rises, the greater the income of the person in the feminine position, the more power she has in the relationship to leave it or bargain with those participating in it. 85 The market i.e. consumers and producers, will have to addres s new social norms in order to expand the care work field effectively. This may imply increased leisure time, as opposed to chore' time for families contracting out for care. The quality of care will rise because of the specialization of those working in the care industry. Training, benefits and vertical job opportunities will append the current dead end positions held by caregivers. Because care work, such as child rearing or disability care, is often not paid' work, there is no way to monitor the qualit y or productivity of the work. To retain one of Gary Becker's assumptions about the division of labor, the quality of care services could only increase if professionals who were 85 Bittman, Michael, Paula England, Nancy Folbre, Liana Sayer, and George Matheson. "When Does Gender Trump Money? Bargaining and Time in Household Work" American Journal of Sociology 109.1 (2003): 186 214.

PAGE 63

Miller 58 trained to do each job were given more compensation and training opportunities 86 Likewise, the shift of the family' into the market allows for more explicit employee/employer contracts which can support the transition from non paid emotional labor to compensated emotional labor. Being directly compensated for care giving will a llow women to be more ambitious in the market. Disadvantages of structuralizing family care include the lack of ability to give oversight by the one receiving care who may often be a child or a disabled person who does not (cannot) know to, or know how to, express disappointment in a care provision. Of course, improved regulations of these industries (health care, child care, senior services) could address this issue partially but not fully. Another disadvantage could be that making explicit the compensatio n received for a traditionally defined altruistic' or loving' act will be confusing at first and support structures should be created to avoid the dehumanization of care work and the initial feelings of rejection or abandonment a care receiver might batt le. Rewarding care in the market could have the effect of reinforcing the sexual division of labor. Many service care providers are ethnicities or in a low income bracket already. Would this shift not exacerbate that? A more complicated discourse on the pros and cons of care giving in the market has 86 In the movement from private to public, researchers have become more cognizant of what constitutes good care. For example, in the literature on child development, the old contrast between merely "custodial" day care "by strangers" in institutional centers versus "loving care in the home," has been replaced by a more careful analysis which concludes (to put it simplistically) that good care is good, and bad care is bad, wherever it takes place. Children on average seem to benefit, or at least not be hurt, when thei r mothers engage in paid work (Blau and Grossberg, 1992; Harvey, 1999). (Folbre)

PAGE 64

Miller 59 emerged among pro and anti market feminists. For example, some policies, which have the intended effect of encouraging women in the market, like paid parental leave or family allowance, might have the opposit e effect of making home seem a comfortable decisions. If care is undervalued, then a systematic decision on how caring ought to be valued needs to be had. (Noddings, Nelson, Folbre 1998?)

PAGE 65

Miller 60 4. Data and Methods The Natio nal Longitudinal Survey of Youth (NLSY79) is the data set used in this paper. It is panel data collected among 12,686 youths between the ages of 14 and 22 when the survey began in 1979. Interviews were conducted on an extensive range of variables annually thereafter and biennially beginning in 1994, resuming annual occurrence in 2002. The individuals interviewed are the cohorts of the original National Longitudinal Survey participants and, with cross sectional weights, comprise a representative sample of t he population. The most recent data available is from the 2008 survey. In this analysis, the sample is restricted to employed individuals, male and female, at 3 annual data packages, 1988, 1998 and 2008. For the purposes of this analysis, estimating a wa ge regression, it is necessary to exclude unemployed individuals whose wages are unavailable. In addition, I exclude those with more than five missing values for more than 2 years, to retain unbiased estimates. Out of this group I create a dummy variable for care workers, pooling registered nurses, private household childcare workers, health aides except nursing, nursing aides, orderlies and attendants according to the three digit occupational census codes provided on the NLSY file; they are listed in Appe ndix Table 2. Of these individuals, I eliminate those whose real hourly wages are greater than $1000 and less than $1.00. This is a justified measure because the hourly rate of pay is a computed variable derived from hours of work a week reported and self reported salary estimates. The outliers in this group may have come from measurement errors, data

PAGE 66

Miller 61 missteps or misconstrued calculations. In addition, I only include individuals who report working at least 25 hours a week. The sum of individuals reporting e ngaging in care work for any of the four years is 1,947, while the total sum of the employed population in all four years is 25,443. Table 5 shows the panels sample sizes for mean real current hourly rate of pay by year and care worker status. Overview of Select Variables I will be brief in my outline of the key independent variables; more details about their construction can be found in Appendix Table 1A. The wage, as mentioned earlier, is a constructed variable measuring hourly rate of pay at current oc cupation, according to hours worked a week and annual labor earnings. The wage rate is adjusted to 2008 dollars. Highest grade completed is restricted to a range of 0 20, where zero means ungraded, and the last value imputed for highest grade completed is interpolated onto missing values for proceeding years. Employment is restricted to those reporting a current hourly rate of pay. Table 5 : Mean Real Hourly Wages by Year and Care Worker Status Non Care Worker Care Worker All Mean Std. Dev. Median Wage. Mean Std. Dev. Median Wage. Mean Std. Dev. Media n Wage 1979 11.83 27.89 9.46 1979 6.70 4.712 4.45 1979 11.48 26.98 11.35 1988 16.29 21.28 14.56 1988 12.90 16.12 10.10 1988 16.13 21.07 16. 74 1998 19.81 21.58 17.10 1998 15.29 10.46 14.11 1998 19.55 21.12 21.59 2008 21.93 22.38 18.80 2008 16.45 12.07 14.00 2008 21.64 21.99 23.64 All 17.84 All 13.07 All 17.57

PAGE 67

Miller 62 In the Table 5, those reporting one for care work and zero for non care w ork separate the wages. The mean for non care workers increases from 11.83 in 1979 to 21.93 in 2008, an 85% increase in real hourly wage. For care workers the average wage increases by $9.74, a 145% increase in real hourly wages. There is a constant level of skewness for non care workers in wages over the years, indicating that as workers age, their wages relatively standardize, and that the disparity of wages remains the same over the years. This is supported by human capital accumulation theories and exp erience based raises. Care workers wages, however, increase in skewness, as they grow older, indicating disparate levels of human capital accumulation, in addition, the standard deviation for care workers increases over the years indicating an industry wit h very little wage stability. I find in my National Longitudinal Survey (NLS) sample in 2008, 77.2% of the entire population in the NLSY79 were employed from 1978 2008, 4.4% were unemployed and 17.6% were not in the labor force. Twenty five percent of al l women aged 18 44 were not in the labor force at all. This is a slightly larger percent of working women population, with the BLS make up of the population, in 2008, 59.2% of women were in the labor force, compared to about 75% of women in the NLSY79. The mean wages for care workers in 2008 is $13.07; this is higher than the national average according to the BLS for personal home care aides, $9.82. Part of the higher wages in the NLSY79 sample could be the inclusion of all direct care workers, including th e slightly higher paid registered nurses. In addition, my sample may have various measurement errors, due to self reporting wages. I also retain several egregious outliers: some of the individuals in my data are personally

PAGE 68

Miller 63 employed care workers, who can e arn up to $300 an hour, while those employed by agencies are more limited in the potential earnings they can make. In the future, it would behoove researchers to initially separate out individuals according to employment situation, and then proceed. As suc h, my wage data may report slightly higher wages than the national average. This will not serve as a barrier to analysis, as all fully employed workers are pooled, and higher wages for care workers as relative to the high wages of non care employees, shoul d make for a more moderate coefficient estimation. Initially one might be concerned about this procedure because the amount of people claiming care work is a small sum to begin, and as not all individuals respond to every variable and subsequently elimin ated from regression, the set may be diminished significantly. Measurement error due to small cell size should produce random error in measuring whether care work; thus, because random measurement error on an independent variable biases coefficients toward zero, the test of my thesis that working in the care sector has a negative effect on wages will be conservative. To gain a general sense of how wages are determined for various occupational groups, variables for work experience (Number of Jobs, Curren t Tenure and Training) are looked into. Previous literature finds that tenure and training are positively correlated to an increase in wages. 87 The largest problems with these data are a diminishing sample size as potential experience increases. The reasons for this include general attrition problems; only 80.9% of the original respondents remain in the year 2008, and of these, 87 England, Paula and Nancy Folbre. 1997. Reconceptualizing Human Capital. Paper presented at the American Sociological Association.

PAGE 69

Miller 64 some values are missing or incomplete due to changes in the survey and techniques. It is worth noting that the sample size in genera l, while varying, has over 6,000 observations in each year. In 1979, when the cohorts were aged 14 22, 6.9% of the employed population reported care work, in 1988, when the cohorts were 25 33, 5.2% of the employed population reported care work, in 1998, when the cohorts were aged 35 43, 5.8% of the population reported working in care work and in 2008, when the cohorts were 44 52, 5.3% of the population reported care work. This indicates a relatively steady composition of care workers as a percent of the population. In the descriptions of the variables below, I use a weighted percent to describe the distribution among the population. The National Longitudinal Survey Website supplies these weights, which help to accurately reflect the data in terms of the whole population. The real current hourly rate of pay for the total average working population rose from $11.48 in 1979 to $21.64 in 2008. Because of this increase in the real rate of pay, I find it still more telling to look at the median wages earned by care workers to determine the average rate of pay, as this is reflective of the normal rate offered by agencies to care workers. Institutional Variables Institutional Variables Table 6 : Gender Distribution of the Care workers in NLSY79 Gender by year # I n entire sample Weighted % of entire sample Average Wage for entire sample # Care workers Weighted % of care workers Average Wage of care workers Median Wage of care workers 1988 Male 5959 54.80 17.51 32 7.31 16.30 10.00 Female 5697 45.20 14.33 393 92.6 9 14.10 10.28 Male 6377 52.69 27.71 26 7.79 16.41 15.38 2008 Female 5948 47.31 18.29 335 92.21 16.45 13.88

PAGE 70

Miller 65 Nearly eight percent (7.79%) of care workers were men, while 92.2% of care workers were women. The average wage for men in the care industry wa s $16.30 in 1988 and $14.10 for women in the care industry, compared to $14.33 for women in the population and $17.51 for male non care workers. In 2008, the average wage of male care workers was only four cents less than female care workers, $16.41 and $ 16.45 respectively. This indicates there is little difference in pay assigned to both men and women working in this industry. Is it safe to assume that men do not get a wage premium for working in this field? Or rather, that with in care sex discrimination does not occur. The results support the hypothesis that care work is not undervalued due to sex discrimination, but rather gender discrimination via association in a very broad sense. Gender discrimination seems associated with the nature of the work, not for the individuals participating in it, as supported by previous studies. Table 7 :Race/Ethnicity of the Care workers in NLSY79 (2008 Race/Ethn icity in # in entire sample Weight ed % of entire sample Averag e Wage for entire sampl e # care workers Weighte d % of care workers Average Wage of care workers Median Wage for Care workers Hispanic 1793 6.06 13.71 76 6.57 11.31 8.64 1988 Black 2842 12.75 12.73 133 16.03 11.23 10.00 Non Hispanic/N on Black 7021 76.31 16.21 294 77.40 15.08 10.92 Hispanic 1933 6.56 20.38 69 6.49 14.65 10.00 2008 Black 3034 13.93 17.71 140 20.11 14.14 12.00 Non Hispanic/N on Black 7358 79.51 25.11 152 73.4 18.36 15.14

PAGE 71

Miller 66 The ethnicities of the cohorts were 7% Hispanic, 20% Black and 73% non Hispa nic and non Black in the care worker industry. This differs significantly in the 8% increase in black cohorts working in care than in the population. The average wages for Hispanic and Black care workers are $4.00 less than the average wage for non Hispani c non Black care workers, and $10.00 less than non Hispanic non black non care workers in the population. The median wages for care workers are more on par with industry standards, $9.82 according the BLS. Non Hispanic and non black care workers earn $3.14 more than black care workers and $5.14 more than Hispanic care workers. While this order was in place in 1988, the distances have widened in the 20 years between the surveys. Table 8 Poverty Statuses of the Care Workers in NLSY79 Poverty Status in 2008 # in entire sample Weighted % of entire sample Average Wage for entire sample # care workers Weighted % of care workers Average Wage of care workers Not in Poverty 5435 89.65 25.14 243 87.7 18.84 In Poverty 987 10.35 9.96 62 12.3 7.41 Those living in poverty are 12% of the care worker population, 2% more than the total population. Care workers which work more than 20 hours a week form 83% of the care worker population, while 92% of the total employed population works more than 20 hours a week. Only the care workers who work more than 25 hours a week were included in my sample.

PAGE 72

Miller 67 Human Capital Variables Human Capital Variables The human capital variables in my regression model include education, work experience 88 training and marital status. Table 9 : Selected Demographic Characteristics of Care Workers by Highest Reported Education Level, 2008 Care workers in 2008 Some High School High School Only Some College College Degree Post College Education Age (44 55) 11.54% 32.32 % 30.47% 13.44% 10.01% % Female 17.71% 31.67% 41.80% 7.63% 1.19% Race/Ethnicity Black 16.02% 33.76% 34.94% 8.89% 4.85% Hispanic 24.48% 27.15% 18.32% 8.98% 7.98% Non Black Non Hispanic 9.45% 32.40% 30.41% 14.87% 11.37% Wage Group 1 18.27% 44. 99% 20.47% 7.65% 4.84% 2 0% 15.31% 50.47% 21.11% 13.11% 3 0% 1.38% 46.08% 31.59% 20.95% 4 0% 0% 0% 0% 100% *Wage group 1= $1.00 15.00/hour 2=$15.00 $30.00/hour 3=$30.00 $55.00/hour 4=$55.00+ Almost 80% of all care workers graduated from high school, compared to 90% of the entire population. The wages of those graduating from high school and participating in care labor jump from less than $8.00 an hour to $11.00 an hour with a high school degree and with a few years of college, $18.00 an hour. Of those, care workers in the first wage group, more than half of all care 88 Here I use current tenure as a proxy for experience, however work experience is built into the model somewhat because the ages of the cohorts only differ by 8 years, the difference in potential work experience is about equal,

PAGE 73

Miller 68 workers earning less than $15.00 graduated at most from high school. However, almost 40% earning less than $15.00 have attained some level of college education. Table 9 adequately r eflects the assumptions of the human capital model, that those investing in higher levels of education will see a greater return in their wages. Table 10 : Training for the Care Workers in NLSY79 Vocationa l/Technic al Training in 200 8 # in entire sample Weighted % of entire sample Average Wage for entire sample # care workers Weighted % of care workers Average Wage of care workers None 6422 86.18 23.54 284 72.8 16.73 Yes 970 13.82 26.33 77 27.2 18.84 One of the key variables for i ncorporating the Human Capital Model, especially in terms of the labor force participation decision for women is the effect of marital status on wages. I hypothesize that marital status for men will have a positive effect on wages while marital status will have a more negative effect on wages for women than for men. It is unclear whether single men will receive a wage premium over married men in the market. From a human capital perspective, a single man has more opportunities to engage in his work place how ever, less motivation (no family to provide for).

PAGE 74

Miller 69 5. Results: Statement of Econometric Model: For an individual i in year t, I estimate equation 2, a standard OLS earnings regression 89 : Equation 2 : NLSY79 Wage Estimation LOGWAGE = C(1) + C(2)*CAREWORK + C(3)*MALEMARRY + C(4)*FEMMARRY + C(5)*FEMNOTMARRY + C(6)*CURRTENURE + C(7)*EDUC + C(8)NON BLACK NON HISPANIC + C(9)*BLACK + C(10)*RURAL + C(11)*SOUTH Care work is a dummy variable for individuals responding positively to care work occupational codes. Malenotmarry is a dummy variable for non married males, Femmarry and Femnotmarry are also dummy variables for whether married, grouped by gender. Currtenure is the current tenure of the individual in their current job by ye ar, this is approximated from the original variable, current tenure in weeks, by dividing each variable by 52 (the number of weeks in a year); Educ is the highest grade completed as of the interview date, it is a continuous variable. Black and Caucasian ar e dummy variables for the race/ethnicity of the respondent; this would have three response levels, Black, Hispanic and non Black non Hispanic. Rural determines whether the respondent lives in a rural or urban area, whether South is a dummy variable derived from regions which include North, North 89 Heckman, James J; Lochner, Lance J. ;Todd, Petra E., 2008. "Earnings Functions and Rates of Return," Journal of Human Capital, University of Chicago Press, vol. 2(1), pages 1 31.

PAGE 75

Miller 70 Central East and West. Studies have shown a more significant effect on wages for those living in the south. 90 These additional variables help to ensure that I do not spuriously attribute the effects of care worker st atus on a lower wage, in addition they are useful for determining whether wage effects are consistent with the Human Capital Model (education and tenure), or Bargaining Model (marital status) or merely discrimination (sex and race). Log wage regressions are a popular choice among economists for estimating earnings equations. Log wage estimations have a very useful interpretation. If any variable in the equation changes by one unit, then wage will change in percentage terms. "Specifically, Y will cha nge by § 100 percent, holding [all other variables] constant, for every unit that X increases." 91 The variables of interest are the dependent variable log wage, and the independent, explanatory variables the regressors. Log wage in a simple linear regression model, will be predicted according to the estimated ordinary least squares of the given response variables. The susceptibility of Ordinary Least Squares (OLS) models to omitted variable bias is one of its limitations. Not to mention that for panel data, t he significance tests, cannot track cross sections across time periods and thus do not account for observations not being independent. 92 The economic theory I utilize suggests tha t wages are linearly connected in their parameters (coefficients) to human ca pital investments, institutional and geographic factors. For the log lin 90 Tomaskovic Devey, Donald, Kevin Stainback, Tiffany Taylor, Catherine Zimmer, Corre Robinson, and Tricia McTague. 2006. Documenting desegregatio n: Segregation in American workplaces by race, ethnicity, and sex, 1966 2003. American Sociological Review 71 (4) (Aug.): pp. 565 588, http://www.jstor.org/stable/30039010 91 Studenmund, A. Using econometrics : a practical guide Boston: Addison Wesley, 2011. p. 223. 92 516 DEMOGRAPHY, VOLUME 33 NUMBER 4, NOVEMBER 1996

PAGE 76

Miller 71 regression model, I estimate the log wage as a function of marital status by sex, educational attainment, and current tenure at current job, rural living status, and region and care work status. 93 94 According to the hypothesis, in addition to the negative wage effects workers will receive for discriminatory reasons based on gender and race, and human capital deviances, there will be a wage penalty for engaging in work which does not f ollow the standard models ideal for self interest in the workplace, or for caring' as a mode of production. This refers back to the devaluation of caring work mentioned in chapter 3 and demonstrated by numerous studies (Kilbourne et al 1998; England et al 1994; England, 1992, Steinberg 1990.) The variables of most interest for this regression are whether care worker', whether married (male/female)' and educational attainment. I hypothesize that the wages of care workers will demonstrate a negative pecun iary penalty, but if direct care work is valued, the wages will respond in kind to increased demand and/or experience/skill attainment, demonstrating an insignificant coefficient for the earnings model. Here, work experience is proxied by age, represented in the pooled regression, since the ages of the cohorts only differs by +/ 8 years in each sample; and by current tenure. The log wage regression results, summarized in Table 10, are consistent with the hypothesis. The estimated equation estimation deri ved from the regression follows: 93 Wooldridge, Jeffrey. Introductory econometrics : a modern approach Mason, OH: South Western, Cengage Learning, 2009. 94 Thi s version of a wage regression is also called a Mincer earnings regression'. It is essentially a law of earnings determination, according to many empirical economists, also referred to as the Human Capital earnings function'. (Heckmann et al, 2003)

PAGE 77

Miller 72 Equation Equation 3 3 : Log Wage Estimation 2008 : Log Wage Estimation 2008 2008 2008 LOGWAGE = 1.265 0.120 CAREWORK** + 0.277 MALEMARRY** ( 0.047) (0.037) (0.024) 0.073 FEMMARRY** 0.070 FEMNOTMARRY**+ (0.024) (0.025) + 0.00038 CURRTENURE** + 0.092 EDUC* + 0.167 CAUCASIAN** (0.000) (0.003) ( 0.019) + 0.126 HISPANIC** 0.081 RURAL** 0.54NORTHCENT + ( 0.025) (0.020) ( .02 1), .071NORTHEAST +.10WEST ( .024) (.024) R 2 = 0.244 adjusted R 2 =0.243 N= 6544 A comparison group is useful to have in mind when discussing the results of a log lin regression. It is intuitive that the coefficients on these va riables represent their slope, all else held constant, recall the standard Y=MX+B formula from grade school? If the coefficients on the variables represent the slope for that variable, than the comparison group must be everything that variable is not. Sing le black male, non care worker, living in an urban area, not in the South, would be our comparison figure. If we enter zero in for all the variables, an unlikely circumstance, what remains is 1.265, which is the log wage for the single(urban, black, nort h residing, non care worker) male of our intercept, given any education and tenure level. This thought experiment does not allow inclusion of continuous variables such as education, or tenure. p** < .05 (t wo tailed test) Standard errors are located below the coefficients. NOTE: all variables are significant are at the .025 level.

PAGE 78

Miller 73 In 2008, whether married male, education, current tenure, Cau casian and Hispanic all demonstrate a positive effect on wages (for these ethnic coefficients, remember the base group is black individuals); care work, whether married female, and whether non married female, rural and south all demonstrate a negative effe ct on wages. Whether married or not, women receive a 7 7.3% less wages than single men, a married man receives a huge wage premium for having a spouse, while women see a large persistent negative wage effect whether married or not. This is consistent with household time allocation models, women spend more leisure time producing home goods, in particular, when married. Married women will have 1% less wages less than a single female and married males earn 34% more than married a single female. Married men al soe receive a wage premium over single men, earning 27% more wages. Education earns an individual about 9% more wages for every additional unit of education completed. This is obviously a rough estimation, as structural breaks exist with in the continuous variable of education. Current tenure has a positive wage return, with a 1.8% return to wages for every year maintaining the same job. 95 Care workers will earn 12% less wages, all else constant. 95 The return to current tenure is calculated by computing the product of the partial elasticity coefficient of current tenure, which is measured in weeks, by 52, which is an estimate of the weeks in a year.

PAGE 79

Miller 74 Equation Equation 4 4 : Pooled Log Wage Es : Pooled Log Wage Es timation 1988 timation 1988 2008 2008 LOGWAGE = 2.102 + 0.157 CAREWORK** + 0.219 MALEMARRY** ( 0.017) (0.018) (0.011) 0.095 FEMMARRY** 0.100 FEMNOTMARRY** + (0.011) (0.012) 0.00048 CURRTENURE** + 0.023 EDUC** + 0.198 CAUCASIAN** + (0.000) (0.001) ( 0.010) 0.090 HISPANIC** 0.118 RURAL** 0.046SOUTH** ( 0.012) (0.001) (.001 ) R 2 =0.1813 Adjusted R 2 =0.1810 N=10,175 A similar regression pooled f or the years 1988 2008 yields similar results. Some significant differences to notice include There is an increase in sex based marital status penalties for women, from a 7% wage penalty in only 2008, to a predicted 10% penalty in the span of time 1988 200 8. Changing household and marital norms shifting from the 1980's into the early 2000's could explain the increased female penalty in the longitudinal regression. Fewer women are leaving the workforce upon entering into marriage in 2008. The potentially low or no income status for women getting married in 1988 or earlier, is reflected in a greater penalty in the percentage of earnings observed for married women over 20 years, as shown by the regression. The care work penalty is consistent with the 2008 resul ts, over a pooled period, participating in care work will diminish wages by 15%. **p < .05 (two tailed test) Standard errors are in parentheses below the coefficients.

PAGE 80

Miller 75 Table 11 : Coefficients for Care Work for Regressions similar to those in Tables 9 and 10 except for changes indicated in first column. Coefficients to Care Work with Various Model Changes Care Work Coefficient Changed Variable Coefficients adj. R squared Education Lagged ( 3) 0.274 N= 3941 educ( 3)=0.056 0.115 Including Number of Jobs variable/ Education Lagged ( 1) 0.113 N=6544 educ( 1)=0.084 jobsnum= 0.005 0.218 1998 only 0.0983 N=6901 NA 0.201 1988 only 0.238 N=8137 NA 0.157 0.236 1979 only 0.395 N=4259 NA Table 11 reports only the coefficients on whether c are work' for a number of models in which small details where changed, to see whether the findings were robust. (Each row displays the results when I undertook only the variation from the model in table 9, year is only 2008 unless indicated; changes in pro cedure are not cumulative down the rows.) First, I lag education 3 periods, to the time unit that would correspond to 1979, this is when the cohorts were 14 22 and so may have a more equivalent distribution of highest grade completed. In the second variati on of the model, number of jobs is included as a potentially omitted variable and only one lags education. The next three variations of the model in table 9, are for different years in the sample, to corroborate the final regression in 2008 and the pooled regression in table 10. The year 1998 shows the least penalty for participating in care work, indicating a 9.8% impact on wages, all else constant. Overall, the results of the variations of analysis in table 11 show that, despite changes in the exact

PAGE 81

Miller 76 magn itude of coefficients, all coefficients for care work are negative and all are statistically non trivial.

PAGE 82

Miller 77 !"# $%&'()*+%& : The results in this paper strongly support the hypothesis that care labor and low wages go in tandem. Not because care is an occupat ionally segregated field or because it is unskilled labor, but rather because of a history of undervaluing care work, culturally reinforced through systematic occupational segregation. I find that care work has a negative impact on wages. Given these wage findings, one might wonder if workers would be better off withholding paid care services, until it is valued fairly in the market as a viable source of income. The first issue with fair valuation in the market' leads to questions of dependency and require ment. Should individuals who cannot afford care services in the market be denied them, left to explain their poor upbringing in court or else? However, this is not a legitimate option given that care services are often non negotiable. The supply of care l aborers must increase in the next ten years to meet the rapid growth in demand, and those with the biggest pockets and the greatest needs should be willing to pay more. This is not a long term solution and more research should be done into restructuring t he direct care industry, in tandem with the large scale changes planned for American health care industry. In Chapter 2, I summarized the economic foundations implicit in analyzing the structure of the care industry. While there is a gap between the amou nt of care demanded and the supply of care, much of this can be analyzed in terms of the relative elasticity's of the supply and demand in the labor market. Wage setting devices, such as compensating differentials, intrinsic rewards and marginal product we re reviewed. In terms of the

PAGE 83

Miller 78 market for care labor, some of these differentials may be explained due to gender bias, intrinsic rewards, public good classification or cultural devaluation of caring activities. Important models that help in understanding the supply of labor and the wage setting mechanisms include Human Capital Models and the labor force participation decision, compensating differentials, crowding models, and dual labor market theory. Chapter 3 gives an overview of the literature which suppl ements the theory discussed in chapter 2. A solid definition of formal care labor, as it is differentiated from informal care work, is important to qualify in terms of a discussion. Contracts for care services include implicit and explicit aspects, in whic h implicit emotional labor comes into play. This helps explain the difficulty in measuring care services, which are contingent upon what kind of good care labor is. A dialogue between feminist economists and theories of care help support my hypotheses that care work is culturally devalued because of its association with woman's work. Moreover, the link between gender and care services points unarguably towards its dominant agents, women. In formatting a model for analysis of care work, Susan Himmelweit lay s an important groundwork for a feminist model'; feminist constructionists and affirmative action feminists adhere to in economic literature on care work. In Chapter 4, I review the data and methods of the empirical analysis of the NLSY79 study, and cha pter 5 presents my results in terms of human capital variables, institutional variables and the penalty for caring in the market.

PAGE 84

Miller 79 The future of a stable, reliable industry for care services is contingent upon immediate policy focus on improving the condi tions in these occupations and reevaluating the values, we assign to caring activities. The amount of money allocated to programs such as Head Start, Medicaid, Medicare and other social programs should take into account that one of the fastest growing occu pations receives a heavy share of this money for the wages of the employees. It would be useful to have insight into the roles that unions, worker initiatives, and state regulation have on the effective standards of care and wage setting mechanisms for tho se doing care labor. The norms that once restricted women to performing caring occupations are weakening. The penalties for performing care are clear in a fiscal sense, but what about the penalties of continued self interested calculation on the future w ell being of the infirm? It is good to be skeptical of future commodification, which may very well raise the price and lower the quality of important care services. This does not mean that pay equity debates should lead to the conclusion that if women are paid less for nurturing jobs, than they should just look for higher paying jobs. Resistance to the commodofication is not a long term solution and as a society, we might do best to consider rewarding work fairly both in the public and private sphere. Betw een the two extremes of allowing the market to (under)value care, and denying care giving any monetary value at all, there are policies of deinstitutionalization, allowing individuals with disabilities to stay out of dehumanizing settings. However, researc hers have shown that deinstitutionalization, and the community care it requires, rely quite heavily on family and volunteer work, done primarily by women (Traustadottir

PAGE 85

Miller 80 1991). It could be perceived that increasing wages of direct care workers would increa se the costs of health care for consumers, however, this is not the case. Considering the larger scale health care required that neglect and accumulated health issues (left untreated) may cost, it would be a healthier and more sustainable choice to have re liable preventative care than costly reparative treatments. Two hours of home health care will cost $11,680.00 annually, as opposed to a nursing home which costs on average $50,005.00 annually. 96 An interesting study verifies that home health care is less c ostly than hospital based care. (Moalosi et al., 2003) Even though this study refers to TB patients, it is an equivalent kind of long term paraprofessional care, administered by the same direct care occupations discussed in this thesis. To answer the skep tics who decry a rise in the cost of health care, valuing care labor on a home and community based level would increase individuals long term sustainable health and decrease the much more costly hospital treatments and the need for emergency care. Policie s and wage dialogues should work towards supplying more support to individuals performing non market caring labor, which has positive benefits to society as a whole. This would be a good start in reversing a trend of impoverishment for those who provide va luable social capital. It is likely that individuals in caring occupations are concerned about the well being of their clients, and as a result, are not concerned only with earnings and leisure, as the traditional model of rational economic man might be. Suggestions for expanding notions of leisure and caring, rewards and income run up against the charge of nearly removing the home, and the 96 Data from: http://cgi.money.cnn.com/tools/elder_care/elder_care_cost_finder.html

PAGE 86

Miller 81 caring done in it, as a site of relation, pleasure, play, and growth. The reasons to label, and learn more about how these private sites affect our public lives, I would argue, is to potentially provide fuller and more accurate accounts about how individuals sustain themselves under a situation of scarcity of care and lead to more equitable policies, designed to help restructure caring work, to the benefit of those performing it. This overhaul of the ways we conceptualize dependency and relationships with patterns of exchange and reward, will require reconsideration into what we typify as work and how we remunerate it

PAGE 87

Miller 82 Appendix A : Table 12 Variable Construction Levels of Response ID THIS NUMBER WAS ASSIGNED ACCORDING TO THE RECORD NUMBER OF EACH RESPONDENT ON THE TAPE. THE FIRST RESPONDENT WAS ASSIGNED ID#1, THE SECOND WAS ASSIGNED ID#2, ETC. ANY EDUCATION WHAT IS THE HIGHEST GRADE OR YEAR OF REGULAR SCHOOL THAT YOU HAVE COMPLETED AND GOTTEN CREDIT FOR? 21 MARITAL STATUS MARITAL STATUS 5 RACE/ETHNICITY R'S RACIAL/ETHNIC COHORT FROM 3 SCREENER SEX SEX OF R 2 OCCUPATION OCCUPATIO N (CENSUS 4 DIGIT, 00 CODES) (ALL) 2000 CENSUS CODE FOR OCCUPATION ANY CAREWORK 2 VOCATIONAL/TECHNICAL TRAINING ATTENDED VOCATION/TEHCNICAL PROGRAM OR ON THE JOB TRAININGS SINCE LAST INTERVIEW? 2 HOURLY RATE OF PAY HOURLY RATE OF PAY JOB #1 ACTUAL DOL LARS AND CENTS ANY POVERTY STATUS POVERTY STATUS IN PRIOR YEAR, VARIABLE USES THE TOTAL NET FAMILY INCOME VARIABLE TO DETERMINE THE STATUS OF THE RESPONDENT'S HOUSEHOLD. 2 FAMILY SIZE FAMILY SIZE 11 REGION REGION OF CURRENT RESIDENCE, FOR VARIABLE CREAT ION SEE ATTACHMENT 104: GEOGRAPHIC REGIONS 4 URBAN/RURAL IS R'S CURRENT RESIDENCE URBAN/RURAL? 2

PAGE 88

Miller 83 Table 13 : Log Wage Regression 2008 Dependent Variable: LOGWAGE Method: Panel Least Squares Sample: 2008 2008 Perio ds included: 1 Cross sections included: 6544 Total panel (balanced) observations: 6544 Variable Coefficient Std. Error t Statistic Prob. C 1.265177 0.046933 26.95736 0.0000 CAREWORK 0.119766 0.037242 3.215934 0.0013 EDUC 0.091550 0.003230 28.34532 0.0000 MALEMARRY 0.276672 0.023617 11.71485 0.0000 FEMNOTMARRY 0.069798 0.024969 2.795387 0.0052 FEMMARRY 0.072831 0.024314 2.995483 0.0028 CURRTENURE 0.000382 1.99E 05 19.19905 0.0000 HISPANIC 0.125801 0.025241 4.983893 0.0000 CAUCASIAN 0.167077 0.019961 8.370130 0.0000 RURAL 0.080550 0.020905 3.853135 0.0001 NORTHCENT 0.053883 0.021635 2.490588 0.0128 NORTHEAST 0.070857 0.023894 2.965x3 0.0030 WEST 0.097862 0.023755 4.119624 0.0000 R squared 0.244638 Mean dependent var 2.815635 Adjusted R squared 0.243251 S.D. dependent var 0.743620 S.E. of regression 0.646885 Akaike info criterion 1.968689 Sum squared resid 2732.965 Schwarz criterion 1.982170 Log likelihood 6428.549 Hannan Qui nn criter. 1.973350 F statistic 176.2659 Durbin Watson stat 0.000000 Prob(F statistic) 0.000000

PAGE 89

Miller 84 Table 14 : Log Wage Regression 1988 2008 Dependent Variable: LOGWAGE Method: Panel Least Squares Sample: 1988 2008 Periods included: 3 Cross sections included: 10175 Total panel (unbalanced) observations: 22571 Variable Coefficient Std. Error t Statistic Prob. C 2.102224 0.017120 122.7962 0.0000 CAREWORK 0.157026 0.017876 8.783991 0.0000 EDUC 0.023 380 0.000876 26.69418 0.0000 MALEMARRY 0.219286 0.011262 19.47211 0.0000 FEMNOTMARRY 0.100252 0.011744 8.536063 0.0000 FEMMARRY 0.095262 0.011396 8.358970 0.0000 CURRTENURE 0.000488 1.29E 05 37.67877 0.0000 HISPANIC 0.089989 0.012246 7.348681 0.00 00 CAUCASIAN 0.197596 0.009768 20.22920 0.0000 URBAN 0.118202 0.009104 12.98337 0.0000 SOUTH 0.045819 0.008379 5.468591 0.0000 R squared 0.181380 Mean dependent var 2.710184 Adjusted R squared 0.181017 S.D. dependent var 0.649264 S.E. of regression 0.587569 Akaike info criterion 1.774842 Sum squared resid 7788.558 Schwarz criterion 1.778752 Log likelihood 20018.98 Hannan Qui nn criter. 1.776114 F statistic 499.8558 Durbin Watson stat 1.111720 Prob(F statistic) 0.000000

PAGE 90

Miller 85 Appendix Table 4: NLSY 2010: Opinions Survey for Individuals Reporting Care Work -----------------------------CAREWORK=YES The FREQ Procedure Frequency Percent Row Pct Col Pct Strongly Agree Disagree Strongly (DK) (REFUSE) Total agree Disagre e 1 918 2431 139 40 19 11 3558 13.17 34 .87 1.99 0.57 0.27 0.16 51.04 25.8 68.32 3.91 1.12 0.53 0.31 36.31 59.7 56.97 59.7 44.19 64.71 2 1610 1641 105 27 24 6 3413 23.1 23.54 1.51 0.39 0.34 0.09 48.96 47.17 48.08 3.08 0.79 0.7 0.18 63.69 40.3 43.03 40.3 55.81 35.29 Total 2528 4072 244 67 43 17 6971 36.26 58.41 3.5 0.96 0.62 0.24 100

PAGE 91

Miller 86 References National alliance for caregiving [cited 4/15/2012 2012]. Available from http:// www.caregiving.org/ Penalties for paid and unpaid care work | russell sage foundation [cited 4/15/2012 2012]. Available from http://www.russellsag e.org/awarded project/penalties paid and unpaid care work Using recovery act funds to improve direct care jobs and the quality of direct care services | russell sage foundation [cited 4/20/2012 2012]. Available from http://www.russellsage.org/research/reports/recovery act direct care workers (accessed 4/20/2012). Barker, Drucilla K., and Edith Kuiper, eds. Toward a feminist philosophy of economics L ondon ; Routledge, 2003. Becker, Gary. 1965. A Theory of the Allocation of Time.'' Economic Journal 75(299): 493 517. Beller, Andrea H. 1982. Occupational segregation by sex: Determinants and changes The Journal of Human Resources 17 (3): 371. Bener ia, Lourdes. 2003. Economic rationality and globalization: A feminist perspective In Feminist economics today: Beyond economic man. eds. Marianne A. Ferber, Julie A. Nelson. 2nd ed., 115 133. Chicago: University of Chicago Press. Bergmann, Barbara R. 1 974. Occupational segregation, wages and profits when employers discriminate by race or sex. Eastern Economic Journal 1 (2) (Apr.): pp. 103 110, http://www.jstor.org/stable/40315472 Better Jobs Better Care. May, 2005. Family care and paid care: Separate worlds or common ground? Washington D.C.: Better Jobs Better Care, 5. Bittman, Michael, Paula England, Nancy Folbre, Liana Sayer, and George Matheson. "When Does Gender Trump Money? Bargain ing and Time in Household Work" American Journal of Sociology 109.1 (2003): 186 214. Blau, Francine D., Marianne A Ferber,.; Anne E Winkler,. 2010. The Economics of Women, Men and Work. 6th ed. Upper Saddle River, NJ: Prentice Hall. Blau, Francine D., and Wallace E. Hendricks. 1979. Occupational segregation by sex: Trends and prospects The Journal of Human Resources 14 (2): 197. Bureau of Labor Statistics. Home health and personal care aides : Occupational outlook handbook : U.S. bureau of labor statistic s in U.S Department of Labor [database online]. [cited 4/15/2012 2012]. Available from http://www.bls.gov/ooh/Healthcare/Home health and person al care aides.htm#tab 1

PAGE 92

Miller 87 Cohen, Philip N. The gender division of labor: "keeping house" and occupational segregation in the united states in JSTOR: Gender and Society, Vol. 18, No. 2 [database online]. Apr., 2004 Available from http://www.jstor.org/stable/4149435 Coleman, James S. 1990. Foundations of Social Theory. Cambridge MA: Harvard University Press. Dagan, Hanoch. 1999. Takings and distributive justice. Virginia Law Review 85 (5 ) (Aug.): 741 804. Danby, Colin. 2004. Lupita's dress: Care in time. Hypatia 19 (4): 23 48. JSTOR DeVault, Marjorie L. 1999. Comfort and struggle: Emotion work in family life In Emotional labor in the service economy. eds. R. J. Steinberg, D. M. Figar t, 52Sage Publications, http://books.google.com/books?id=_gvXtgAACAAJ;; Dickens, William T., and Kevin Lang. 1985. A test of dual labor market theory. National Bureau of Econom ic Research, http://www.nber.org/papers/w1314 England, Paula. 2003. Separative and soluble selves: Dichotomous thinking In Feminist economics today: Beyond economic man. eds. Marianne A. Ferber, Julie A. Nelson. 2nd ed., 33 59. Chicago: University of Chicago Press. England, Paula. 1992. Comparable Worth: Theories and Evidence. New York: Aldine. England, Paula. 2005. Emerging theories of care work. Annual Review of Sociology 31 : pp. 381 3 99, http://www.jstor.org/stable/29737725 England, Paula and Nancy Folbre. 1997. Reconceptualizing Human Capital. Paper presented at the American Sociological Association. England, Paula, a nd Nancy Folbre. 2003. Contracting for care In Feminist economics today: Beyond economic man. eds. Marianne A. Ferber, Julie A. Nelson. 2nd ed., 61 79. Chicago: University of Chicago Press. England, Paula and Lori Reid, and Barbara Stanek Kilbourne. 19 96. The Effect of the Sex Composition of Jobs on Starting Wages in an Organization: Findings from the NLSY. Demography 33(4): 511 12 England, Paula, George Farkas, Barbara Stanek Kilbourne, and Thomas Dou. 1988. Explaining occupational sex segregation and wages: Findings from a model with fixed effects American Sociological Review 53 (4): 544. Ferber, Marianne A., and Julie A. Nelson. 2003. Feminist economics today: Beyond economic man Chicago: University of Chicago Press. Ferguson, Ann and Hennessy, Ro semary. "Feminist perspectives on class and work". [cited 4/15/2012 2012]. Available from http://plato.stanford.edu/entries/feminism class/ Folbre, Nancy, and Nelson, Julie A. For love or money -or both? in JSTOR: The Journal of Economic Perspectives, Vol. 14, No. 4 [database online]. Autumn, 20002012]. Available from http://www.jstor.org/stable/2647078

PAGE 93

Miller 88 Folbr e, Nancy. 1995. "Holding hands at midnight": The paradox of caring labor Feminist Economics 1 (1): 73 92. Forhan, Lisa. Minimum wage & overtime protection for all? | russell sage foundation in Direct Care Alliance [database online]. 2010 [cite d 4/20/2012 2012]. Available from http://www.russellsage.org/research/reports/overtime protection for care workers Franks, Peter, Carolyn M. Cla ncy, and Paul A. Nutting. 1997. Defining primary care: Empirical analysis of the national ambulatory medical care survey. Medical Care 35 (7) (Jul.): pp. 655 668, http://www.jstor.org/stable/3 766852 Fremstad, Shawn. Maintaining and improving social security for direct care workers | russell sage foundation in Center for Economic Policy Research [database online]. 2011 [cited 4/20/2012 2012]. Available from http://www.russellsage.org/research/reports/social security direct care workers Implementing the coverage provisions of health care reform: What's at stake for direct care wor kers | russell sage foundation in Direct Care Alliance [database online]. 2010 [cited 4/20/2012 2012]. Available from http://www.russellsage.org/research/r eports/care work health care reform Geisler, Charles C. 1995. Land and poverty in the United States: Insights and oversights. Land Economics 71 (1) (Feb.): 16 34. Gilman, Charlotte. Women and economics : a study of the economic relation between women a nd men Amherst, N.Y: Prometheus Books, 1994. Gross, JANE. Employer provisions for parental leave (EXCERPT), monthly labor review online, oct. 1989 in New York Times [database online]. New York, March 5, 2012, 8:05 am2012]. Available from http://www.bls.gov/opub/mlr/1989/10/art3exc.htm Gross, Jane. When I needed help, part 2 NYTimes.com in The New York Times Company [database online]. March 5, 2012, 8:05 am2012]. Available from http://newoldage.blogs.nytimes.com/2012/03/05/when i needed help pt 2/#more 12005 Gross, JANE. Home health aides: What they make, what th ey cost NYTimes.com December 30, 2008, 6:00 am2012]. Available from http://newoldage.blogs.nytimes.com/2008/12/30/home health aides what they make what they cost/ Hamermesh, Daniel S. 1986. The demand for labor in the long run In ed. Orley Ashenfelter and Richard Layard, ed., 429 Handbook of Labor Economics. Heckman, James J; Lochner, Lance J. ;Todd, Petra E., 2008. "Earnings Fu nctions and Rates of Return," Journal of Human Capital, University of Chicago Press, vol. 2(1), pages 1 31. Himmelweit, Susan. 1999. Caring labor In Emotional labor in the service economy. eds. R. J. Steinberg, D. M. Figart, 27Sage Publications, http://books.google.com/books?id=_gvXtgAACAAJ; Hochschild, Arlie Russell. 1983. The Managed Heart: Commercialization of Human Feeling. Berkeley: University of California Press.

PAGE 94

Miller 89 Hoffer th, Sandra L., and Duncan D. Chaplin. 1998. State regulations and child care choice. Population Research and Policy Review 17 (2) (Apr.): pp. 111 140, http://www.jstor.org/stable/40230180 Howes, Candace. The best and worst state practices in medicaid long term care | russell sage foundation in Direct Care Alliance [database online]. 2010 [cited 4/20/2012 2012]. Available from http://www.russellsage.org/research/reports/howes medicaid Jacobsen, Joyce P. The economics of gender Wiley Blackwell. 2005. J ochimsen Maren A. 2003. Integrating Vulnerability : On the Impact of Caring on Economic Theorizing." In Toward a Feminist Philosophy of Economics. Eds. Drucilla K. Barker, Edith Kuiper. London ; Routledge, 2003. Kalleberg, Arne L. 2000. Nonstandard employment relations: Part time, temporary and contract work. Annual Review of Sociology 26 : pp. 341 365, http://www.jstor.org/stable/223448 Kilbourne, Barbara Stanek, Paula England, George Farkas,, K. Beron, and D. Weir. 1994. Returns to Skills, Compensating Differentials, and Gender Bias: Effect s of Occupational Characteristics on the Wages of Women and Men. American Journal of Sociology 100:689:718. Leidner, Robin. 1999. Emotional labor in service work. In Emotional labor in the service economy. eds. R. J. Steinberg, D. M. Figart, 61. Sage Publ ications. McKenzie, Nancy. Supporting direct care workers in caring for aging lesbian, gay, bisexual, transgender individuals | russell sage foundation in Direct Care Alliance [database online]. 2010 [cited 4/20/2012 2012]. Available from http://www.russellsage.org/research/reports/lgbt care workers Meisenheimer II, Joseph R. "Feminist perspectives on class and work" October 1989, Vol. 112, No. 10 [cited 4/15/2012 2 012]. Available from http://plato.stanford.edu/cgi bin/encyclopedia/archinfo.cgi?entry=feminism class The services industry in the 'good' v ersus 'bad' jobs debate (EXCERPT), monthly labor review online, feb. 1998 February 1998, Vol. 121, No. 2 [cited 4/15/2012 2012]. Available from http://www.bls.gov/opub/mlr/1998/02/ art3exc.htm Moalosi, G., K. Floyd, J. Phatshwane, T. Moeti, N. Binkin, and T. Kenyon. 2003. Cost effectiveness of home based care versus hospital care for chronically ill tuberculosis patients, francistown, botswana The International Journal of Tubercul osis and Lung Disease : The Official Journal of the International Union Against Tuberculosis and Lung Disease 7 (9 Suppl 1) (Sep): S80 5. Nac. 2011. Accepting applications: Innovations in caregiving seed grant opportunity National Alliance for Caregiving (Fri, 18 Nov). 2011. Alzheimer's Caregiver's healthcare costs increase as person with dementia declines National Alliance for Caregiving (Tue, 22 Nov). 2011. Caregivers want web and mobile National Alliance for Caregiving (Thu, 07 Apr).

PAGE 95

Miller 90 Nakhn ikian, Elise. No excuse for abuse | russell sage foundation in Direct Care Alliance [database online]. 2011 [cited 4/20/2012 2012]. Available from http://www.russellsage .org/research/reports/abuse disabled National Alliance for Caregiving. 2011. Caregivers of veterans serving the homefront United Health Foundation, http://www.caregiving.org 2011. Double jeopardy for baby boomers caring for their parents Metlife Foundation, Nelson, Julie A. 2003. Separative and soluble firms: Androcentric bias and business ethnics In Feminist economics today: Beyond economic man. eds. Marianne A. Ferber, Julie A. Nel son. 2nd ed., 81 99. Chicago: University of Chicago Press. Noddings, Nel. 2003. Caring: A feminine approach to ethics & moral education Berkeley: University of California Press, Contributor biographical information http://www.loc.gov/catdir/bios/ucal052/2003042608.html Publisher description http://www.loc.gov/catdir/description/ucal0 42/2003042608.html Okamoto, Dina, and Paula England. 1999. Is there a supply side to occupational sex segregation? Sociological Perspectives 42 (4) (Winter): pp. 557 582, http://www.jstor.o rg/stable/1389574 Pagac, Adrienne, and Dresser, Laura. Better jobs for in home direct care workers | russell sage foundation in Direct Care Alliance [database online]. 2010 [cited 4/20/2012 2012]. Available from http://www.russellsage.org/better jobs home direct care workers Peterson, Janice, and Margaret Lewis. 2003. The Elgar Companion to Feminist Eeconomics Edward Elgar Pub. PHI. Nov, 2008. Direct care jobs and long term care: Untapped engine for job creation and economic growth. Bronx, NY: Paraprofessional Healthcare Institute Quality Care through Quality Jobs, 2. Feb, 2011. Who are direct care workers? Bronx, NY: Paraprofessional Healthcare Instit ute Quality Care through Quality Jobs, 3. Pierret, Charles R. The 'sandwich generation': Women caring for parents and children (EXCERPT), monthly labor review online, september 2006 September 2006, Vol. 129, No. 92012]. Available from http://www.bls.gov/opub/mlr/2006/09/art1exc.htm Polson, Diana. By our sides: The vital work of immigrant direct care workers | russell sage foundation in Direct Care Alliance [database online]. 20 11 [cited 4/20/2012 2012]. Available from http://www.russellsage.org/research/reports/immigrant care workers Radin, Martha. 1996. Contested Commodities Cambr idge, MA: Harvard University Press. Saunders, Lisa, and William Darrity JR. 2003. Feminist theory and racial economic inequality In Feminist economics today: Beyond economic man. eds. Marianne A. Ferber, Julie A. Nelson. 2nd ed., 101 114. Chicago: Unive rsity of Chicago Press. Sen, Amartya. 1979. Personal utilities and public judgements: Or what's wrong with welfare economics. The Economic Journal 89 (355) (Sep.): pp. 537 558, http://www.jst or.org/stable/2231867

PAGE 96

Miller 91 Smith, Peggie. Protecting home care workers under the fair labor standards act | russell sage foundation in Direct Care Alliance [database online]. 2009 [cited 4/20/2012 2012]. Available from http://www.russellsage.org/research/reports/protecting home care workers Steinberg, R.J. Lois Haignere, C. Possin, C.H. Chertos, and D. Treiman. 1986. The New York State Pay Equity Study: A Research Report. Albany: Center for Women in Government, State University of New York Press. Steinberg, R. J., and D. M. Figart. 1999. Emotional demands at work: A job content analysis In Emotional labor in the service economy. eds. R. J. Steinberg, D. M. Figart, 177Sage Publications, Steinberg, R. J., D. M. Figart, and American Academy of Political and,Social Science. 1999. Emotional labor in the service economy Sage Publications, Strober, Myra H., "The Application of Mainstream Economics Constructs to Education: A Feminist Analysis," in Feminist Economics Today: Beyond Economic Man. Eds. Marianne A. Ferber and Julie A. Nelson. University of Chicago Press. 2003. Studenmund, A. Using econometrics : a practical guide Boston: Addison Wesley, 2011. Toma skovic Devey, Donald, Kevin Stainback, Tiffany Taylor, Catherine Zimmer, Corre Robinson, and Tricia McTague. 2006. Documenting desegregation: Segregation in American workplaces by race, ethnicity, and sex, 1966 2003. American Sociological Review 71 (4) (Au g.): pp. 565 588, http://www.jstor.org/stable/30039010 Traustadottir, Ranneig. 1991. Mothers Who Care: Gender, Disability, and Family Life Journal of Family Issues 12:211 28. Tronto, Joan C. 1993. Moral boundaries: A political argument for an ethic of care New York: Routledge. Wooldridge, Jeffrey. Introductory econometrics : a modern approach Mason, OH: South Western, Cengage Learning, 2009.