| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
Correspondence: Address correspondence to A. E. (Ted) Benjamin, Department of Social Welfare, School of Public Policy and Social Research, University of California–Los Angeles, 3250 Public Policy Building, Los Angeles, CA 90095-1656. E-mail: tedbenj{at}ucla.edu
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Key Words: Home care • Service model • Family caregiver • Stress • Satisfaction
Home-Care Workforce
Concerns about the home-care workforce have several dimensions. First, there are not enough such workers, who, because of their low status, poor pay, and difficult working conditions, are not easily recruited to this work and, if they are, have very high turnover rates (Burbridge, 1993; Crown, Ahlburg, & MacAdam, 1995; Yamada, 2002). Second, too little is known about the quality of services provided, whereas we already know that workers receive little training, have heavy workloads, and can expect few tangible rewards for high performance (Carr & Kazanowski, 1994; Feldman, 1994). Third, working conditions are often difficult. Not only do workers receive little or no training, but they also receive little or no supportive supervision (Dawson & Surpin, 2001a), are vulnerable to job-related injuries (Gregory, 2001; U.S. Department of Labor, 2001), are often unclear about their specific roles and responsibilities, know little about client health conditions, and have little opportunity to make decisions about their work (U.S. General Accounting Office, 2001). Fourth, workers receive little recognition or respect for the work they do from their employers (including home-care agencies, clients, and their families) or from the general public (Stone, 2001). Consequently, home care fails to attract adequate numbers of capable people and fails to retain them if they enter the workforce.
Although studies of working conditions in long-term care have involved primarily nursing homes, research on home-care workers is growing. Better pay and benefits is regularly cited as one solution across settings, but more attention is being paid to other aspects of working conditions, including job design (Feldman, 1993; Feldman, Sapienza, & Kane, 1990). Strategies to improve recruitment and retention by redesigning work life include more relevant training, increased supervision and peer support, enhanced status, and improved career opportunities (Feldman, 1993).
Consumer Direction and the Workforce
Most research on home-care-worker issues has been confined to workers employed by agencies. The burden of work-life design has been assigned, in effect, to the home-care industry and to programs (most notably, Medicaid and Medicare) that finance home-based services. Recent interest in "consumer-directed" models of home-based services has added a new twist to this discussion about work-life conditions and their impact on worker recruitment and turnover. A growing number of states have adopted Medicaid- or state-funded programs in which eligible recipients with functional limitations are authorized to receive a cash allowance or a given number of service hours at home. They are thus able to recruit and hire their own workers and supervise their own services with modest or minimal agency involvement. In the largest study to date that has examined the impact of consumer direction, self-directing consumers consistently reported more positive outcomes than did agency clients (Benjamin, Matthias, & Franke, 2000; Doty, Benjamin, Matthias, & Franke, 1999). Differences between older and younger users of consumer-directed services were modest, suggesting that old age is not a significant barrier to self-direction of supportive services in the home (Benjamin & Matthias, 2001). Studies done in other program settings have yielded similar findings (Feinberg & Whitlach, 2001; Simon-Rusinowitz, Mahoney, Desmond, Shoop, Squillace, & Fay, 1997).
Empirically, we know far less about what consumer direction means for home-care workers. Labor advocates have expressed concern that structuring stronger recipient roles may come at the expense of those workers hired to provide services to them (Dawson & Surpin, 2001a). Historically, development and monitoring of consumer-directed programs have targeted consumer issues more than worker ones (Feldman, 1997). When compared with agency workers, consumer-directed workers may face several work-life issues. These include lower (often minimum-wage) pay; few or no fringe benefits; little or no training in the basics of home care; no supervision, back-up, or support from a supervisor or fellow employees; uncertain roles and workloads defined solely by the client without professional mediation; vulnerability to harassment, boredom, and unhealthy work environments; and few apparent opportunities for career advancement (Dawson & Surpin, 2001b; Wilner & Wyatt, 1998). To the extent that these conditions characterize consumer-directed programs, we would expect workers in these programs to experience more stress and less satisfaction than their agency counterparts.
More positively, home-care agencies may utilize the overhead they build into their reimbursement structure to create working conditions that encourage lower job-related stress and more satisfaction for workers. On average, agency workers may receive better pay and benefits, more formal training, more job supervision, and more peer support than consumer-directed ones. Their jobs seem to involve relatively clear role definitions and clear lines of accountability (through their agency supervisor). Agencies and supervisors provide at least some protections from harassment and from other forms of unhealthy working conditions. Although relatively limited, agency providers seem to have more opportunities for advancement along a career ladder than consumer-directed workers (Dawson & Surpin, 2001a; Feldman, 1997).
To test whether worker outcomes vary across service models that seem to shape work life in dissimilar ways, we conducted a study of workers providing home-care services under agency-based and consumer-directed approaches. We sought to understand whether and how these service approaches have consequences for home-care workers, whether other factors account for any differences in worker outcomes across models, and how the relationship of the worker to the recipient/consumer affects worker outcomes. The latter is important because many consumer-directed home-care programs permit recipients to hire family members as workers. We also considered the role of worker characteristics (age, gender, ethnicity, education, experience, other employment) and recipient case mix (daily activity limitations, cognitive status, and behavioral problems).
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Sampling and Survey Procedures
Sampling was done in two stages. First, a random sample of 1,095 IHSS recipients was surveyed for a companion study. This sample was drawn from a listing of all program recipients over the age of 18 in mid-1996 and was stratified by service model (consumer directed and agency) and by age (over and under 65). Data on recipients were collected through telephone interviews and the IHSS program database. To ensure competency for interviewing, recipients with severe cognitive impairments (i.e., on ratings of memory, orientation, and judgment) were excluded, while those with mild to moderate impairments were included in the sampling frame.
Second, workers were randomly sampled from the pool of workers serving sampled recipients. Based on power calculations, we targeted a worker sample of 500 (Benjamin et al., 1998). Reflecting the recipient sample, the worker sample was stratified by service model and age of the recipient. The resulting sample included 365 agency workers and 253 consumer-directed ones. The larger agency sample resulted from various technical factors, including adjustments we made to anticipated difficulties with agency-worker participation and a higher than anticipated response rate from agency workers.
Trained and experienced professional interviewers at UCLA's Institute for Social Science Research conducted computer-assisted telephone interviews between October 1996 and March 1997. (These interviewers had no preconceived ideas about work-life conditions in either home-based model.) Interviews were conducted in English, Spanish, and three Asian languages (Cantonese, Mandarin, and Vietnamese). The overall provider response rate (including those impossible to contact) was 62.7%, with a slightly higher rate for the agency workers (65.1% vs. 60.0%).
Measures
The telephone survey of workers included items about demographics, working conditions, relationships with recipients, and two worker-outcome areas: stress and satisfaction. Established measures were used where appropriate (Feldman et al., 1990; Leon, 1994); in some instances, they were adapted to the study population, objectives, or both. Worker-satisfaction measures addressed both the immediate service experience and broader perspectives on direct-service work (Dillard & Feather, 1991; Donovan, 1989; Eustis, Kane, & Fischer, 1993; Feldman et al., 1990; Grau, Colombotos, & Gorman, 1992; Hines, Durham, & Geoghegan, 1991; Leon, 1994; Roberts & Sarvela, 1990; Ross & Reskin, 1992). Items on worker stress and burden addressed relationships with the recipient and family members, work-related problems, worker reactions to demands of the job, and other themes (Berwick, Murphy, Goldman, Ware, Barsky, & Weinstein, 1991; Cafferata & Stone, 1989; Cantor & Chichin, 1990; Jones, Roth, & Jones, 1995; Stephens & Christianson, 1986).
Analysis
Factor analysis yielded 10 worker-outcome dimensions for stress and satisfaction. Stress dimensions included (a) concern about client safety, (b) issues with client's family, (c) client behavioral issues, (d) getting along with client, (e) worker emotional state, and (f) clarity of work role. Satisfaction dimensions included (a) work role satisfaction, (b) performance self-assessment, (c) career advancement, and (d) independence and flexibility.
We conducted all outcome analyses at the worker level using multiple regression based on the general linear model. Predictor variables included service model (consumer directed or professional agency), type of consumer-directed worker (family, nonfamily), worker demographics, and recipient case mix. For service model and type of consumer-directed provider, we used a contrast coding scheme that, unlike a dummy coding scheme, allows the excluded category to be changed to examine any two-way comparison of interest (Cohen & Cohen, 1983; Marascuilo & Serlin, 1988). In the first contrast, professional agency workers are compared with all consumer-directed workers. In the second contrast, paid family consumer-directed workers are compared with nonfamily consumer-directed workers. Unrelated workers known previously to the recipient, namely, friends and acquaintances, are considered nonfamily workers.
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
Working Conditions
There are substantial differences among worker groups in the proportions living with their client and how much choice the worker reports in how required tasks are done (see Table 1). There are notable differences between the professional agency and consumer-directed models (but not within consumer-directed family and nonfamily) on hours per week with the client, mean numbers of clients, wages, and amount of training and supervision. Agency workers earn more, have more clients, have fewer hours with any one client, and report receiving more formal training and supervision. This difference in overall formal training masks the amount of more informal training received by workers in the consumer-directed model, especially training by health professionals that is tailored to specific recipients (Doty et al., 1999).
Outcomes
We computed mean scores for each of the 10 outcome variables by service model and worker relationship. Without considering other intervening variables, differences by service model are statistically significant for 9 of the 10 outcome variables (Table 2). For three of the six stress dimensions, workers in the professional agency model report less stress, while on the other three, consumer-directed workers indicate lower stress. For two of the four satisfaction dimensions, agency workers report greater satisfaction; in turn, consumer-directed workers have higher scores on one satisfaction dimension. In comparing family with nonfamily workers, nonfamily workers fare better on two stress outcomes and family workers do better on one, involving the relationship with the recipient. Nonfamily workers are more satisfied on one satisfaction outcome.
|
|
Stress
On all six stress outcomes, worker differences between service models are statistically significant. When the effects of worker and recipient characteristics are taken into account, service model is a significant outcome predictor for three dimensions of worker stress. Being a consumer-directed worker is associated with more favorable outcomes on two stress dimensions: worker relationship with the recipient (being relatively clear about and comfortable with authority relationships with the recipient) and clarity of client–worker roles. However, being an agency worker is associated with more positive outcomes in terms of less worry about the client in the worker's absence. For two worker outcomes related to conflict with family members and recipient behavioral issues, model differences are not statistically significant.
For two of the stress dimensions, a familial relationship between recipient and worker is a significant outcome predictor. Being related to the recipient is associated with getting along well and experiencing few problems in the relationship with the recipient, whereas being unrelated is associated with a more positive worker emotional state.
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Given these limitations, how should we interpret our findings? Some observers concerned with the worker experience have tended to characterize consumer direction as undermining the traditional structure (i.e., training, supervision, and support) on which agency workers depend and thereby likely increasing job ambiguity and worker stress. Alternately, consumer-directed arrangements have been described as freeing workers from the rigidity of agency procedures and expectations and thus allowing them to be more responsive to recipients while enjoying a more intimate and satisfying work environment. Our findings suggest that both of these scenarios may be partially true but that, for most workers, the consequences of work-life differences between models are relatively modest. In California, agency workers (who are rarely friends or family) understandably seem more protected or insulated from worries about whether recipients are secure and safe when their workers are not around. Agency workers seem more able to leave their work behind and not assume responsibility beyond what is required during the working day. By contrast, consumer-directed workers cannot easily recast familial and friendship relationships into "professional" or "detached" ones.
Related workers seem to face additional pressures not encountered by other workers. Compared with nonfamily workers, paid family workers in our study have been serving the same recipient longer and are more likely to live with or near those they serve, more likely to assist with a wide range of tasks, and more likely to provide additional, nonpaid help. In effect, their job descriptions are more wide ranging and more demanding than those of other workers. It is no surprise that related workers are more emotionally involved with and may worry more about their familial recipient. Like any family caregiver, they may need support in managing the stress of caring work. These findings correspond to those reported by Tilly, Wiener, and Cuellar (2000) for four Western European countries engaged in consumer direction.
As suggested, consumer-directed workers are much more likely than agency ones to perform additional service tasks without pay (60.6% vs. 12.4%; data not shown). Within the consumer-directed model, related workers are much more likely to provide unpaid hours (78.7% vs. 42.5%) and to provide more of them. This echoes early findings from the Cash and Counseling Evaluation that family caregivers provide more care than nonfamily ones (Eckert, San Antonio, Siegel, Simon-Rusinowitz, & Mahoney, 2001). These data point to both the personal commitment of consumer-directed workers and the blurred boundaries between work and personal relationships common under this service arrangement. From a labor viewpoint, unpaid hours represent a problem; from a family perspective, these hours probably represent a natural by-product of family loyalties. Our open-ended interviews suggest that public payment for providing care may moderate the stress of caregiving for workers by conferring both tangible rewards and intangible legitimacy. This is not a new argument, for others have argued that public payment should be utilized to support informal care systems (Linsk, Keigher, Simon-Rusinowitz, & England, 1992; Simon-Rusinowitz, Mahoney, & Benjamin, 1998). Understanding the dynamics of paid family care is especially important since many recipients, especially older ones, when given a choice, are more likely to hire and be more satisfied with services from a related worker (Benjamin & Matthias, 2001; Doty et al., 1999).
Given the advantages typically attributed to agency employment (i.e., better wages and benefits, more formal training and supervision), it may seem surprising that satisfaction differences between the two worker cohorts would be so small. Focus-group discussion and informal interviews with workers in the consumer-directed model indicate that economic rewards are not the primary reason for doing supportive-services work. Possible dissatisfaction with pay and benefits may partly be offset by other less tangible rewards associated with interpersonal ties to the service recipient (as family member or friend) and with a flexible work situation substantially controlled by the recipient and worker. Research on nursing home workers has shown that both continuity of the service relationship and participation in decision making about work can contribute to employment satisfaction (Bowers & Becker, 1992). A sense of worker choice about how tasks are performed is associated with more satisfaction and (especially) with lower stress. Paradoxically, even in a consumer-directed model where recipient preferences are considered paramount, it seems important that workers retain a sense of choice about the work they do.
Moreover, our research suggests that consumer-directed workers receive considerable informal training from various sources, including family physicians, home-health nurses, and therapists, and that this training is recipient specific and thus targeted to specific care needs. From a worker perspective, the absence of formal (agency or academy) training is far less of an issue than expected; findings on service satisfaction from recipients support this view (Benjamin et al., 1998; Doty et al., 1999). Whatever the service model, adequate training and pertinent information on the recipient's condition are associated with more satisfaction and less stress.
If consumer-directed workers are not as dissatisfied as might be expected, then given their pay and benefit advantages, agency workers seem less satisfied than expected. We suggest several reasons for this. First, although the agency benefit advantage is clear, its impact is muted. Whereas most agencies offer some fringe benefits, most IHSS agency workers still do not have health insurance or sick and vacation leave. In our study, the majority of workers in the professional agency model had no benefits: Only 39% had health insurance, 37% had paid sick leave, and 37% had paid vacation leave (data not shown). Fewer than 3% of consumer-directed workers had these benefits. Second, agency workers are typically assigned to care for multiple recipients (a mean of 4.3) and are under agency pressure to allocate their time carefully among them. Our sense is that this frustrates some workers and that workers may distance themselves emotionally from the work and recipients, thereby reducing the satisfaction experienced. Third, physical geography also seems to matter. More agency workers report having transportation and scheduling problems in their work, because they provide services to several recipients across what can be large agency territories. In contrast, more consumer-directed workers live with their recipients and more live nearby (data not shown).
Two other characteristics of recipients and workers also seem important in understanding outcomes. Not surprisingly, providing services to recipients with behavioral problems is associated with lower satisfaction and more stress. In these circumstances, workers probably need additional information, training, and support to manage such problems. The impact of race and ethnicity is more difficult to sort out, but these are clearly salient for both workers and recipients (Benjamin et al., 2000). For example, ethnicity seems to be a much stronger predictor of satisfaction outcomes than either service model or worker–recipient relationship. White workers report more positive role perceptions than Latinos or Asians, whereas Asians and Blacks are more positive about the career benefits associated with their jobs. Consumer direction seems to play out in distinctive ways across racial and ethnic groups (Sciegaj, 2002), and much more research is needed to illuminate how cultural differences interact with different models for providing paid services in the home.
Over the years, as states have begun to embrace consumer direction, they typically have also expanded the discretion to hire to include employing friends and family members. Research a decade ago by Doty, Kasper, and Litvak (1996) argued that there was little evidence that agency workers were much, if any, better off than consumer-directed ones and that hiring family members as workers was good for the recipient and the worker. Our findings support the first contention, because we conclude that there is little difference in worker outcomes between the two models. Our study also supports the second, because on most measures of worker satisfaction and stress, related workers fare as well as others.
Preliminary results from the Cash and Counseling Demonstration and Evaluation lend support to our findings and suggest that they may be meaningful beyond California. The evaluation of this experiment in three states (Arkansas, Florida, and New Jersey), in which Medicaid-eligible individuals were randomly assigned to consumer-directed and agency-delivered personal assistance services, included analysis of worker experiences in both models. Preliminary findings from one state indicate that most workers in both models reported high levels of work satisfaction. Consumer-directed workers reported higher emotional strain, with family workers reporting the most. On most dimensions of work life, however, these workers reported no less stress than agency ones. Although consumer-directed workers received less training than agency ones, they still believed they were well prepared for their work (Dale, Brown, Phillips, & Carlson, 2003).
It is very likely that consumer-directed models will continue to be adopted by state Medicaid programs. As an added impetus, a second wave of Cash and Counseling funding recently has been announced. Yet worker issues remain. For example, policy makers and professional advocates continue to call for stronger quality-assurance mechanisms in consumer-directed programs (Tilly et al., 2000). Critics assume that the quality of services continues to be problematic and that recipient, family, and worker judgments about the quality of personal assistance services are inadequate in the absence of external, professional oversight. Because most such care is personal and supportive, rather than medical, we would argue with Doty and colleagues (1996) and others that recipient, family, and worker judgments are central to monitoring quality.
Critics also assume that in the absence of agency-style training and supervision, recipients are more vulnerable in the hands of consumer-directed workers. Most evidence suggests otherwise, indicating that initial and targeted training may be valuable to some recipients and workers, but that the training and supervision gap is overstated for personal assistance services. This perspective echoes that of other research that also emphasizes the high levels of satisfaction reported by consumers of care (Doty et al., 1996; Eustis et al., 1993; Foster, Brown, Phillips, Schore, & Carlson, 2003; Tilly et al., 2000). Additional research is needed to understand better how and when formal and informal training occurs within the consumer-directed model and which types are most consequential and to define more specifically which training needs should be addressed and how.
Recent developments in California highlight other issues that threaten to alter consumer-directed services in that state. In a very difficult fiscal climate, significant Medicaid (MediCal) cutbacks have been proposed, two of which directly affect the IHSS program. One proposes elimination of recently created "public authorities," public bodies established to represent worker interests, especially to improve wages and benefits, and to provide training and registry services. As yet, no systematic research has addressed the impact of these bodies on worker or recipient outcomes, and it is far from clear what consequences their elimination might have. To the extent that public authorities have stimulated improvements in wages and benefits for IHSS workers, this development may directly threaten worker well-being and the future capacity of the program to attract workers in a market characterized by chronic shortages of entry-level workers.
The second cutback proposal in California involves public payment to family members for IHSS services. Although the broader issue of payments to families has been hotly debated in many states, California has, for three decades, permitted program payments to any family member, including state-local payment to two categories of family members not reimbursable under federal Medicaid regulations (i.e., spouses and parents of minors). Although payment to other family members is not currently in jeopardy, the state is seeking to eliminate IHSS payment to spouses and parents of minors to force recipients to hire workers who are eligible for federal reimbursement. This makes short-term fiscal sense, but we know too little about the benefits of payment to family members for personal assistance work or the costs to the recipient and the family as a whole of ending payment for spouses and parents. Because hiring related workers seems especially appealing to older adults and to some ethnic and racial groups, we need research that examines paid service relationships within the family over time and that traces the positive and negative consequences of related caregiving. Random assignment of recipients (permitted to hire family member versus not) might permit more careful examination of the consequences for both consumers and workers of paying family members within consumer-directed programs.
|
Footnotes |
|---|
1 Department of Social Welfare, School of Public Policy and Social Research, University of California–Los Angeles.
2 Lewis Center for Regional Policy Studies, School of Public Policy and Social Research, University of California–Los Angeles.
Decision Editor: Linda S. Noelker, PhD
Received for publication August 27, 2003. Accepted for publication April 20, 2004.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. E. Benjamin, R. E. Matthias, K. Kietzman, and W. Furman Retention of Paid Related Caregivers: Who Stays and Who Leaves Home Care Careers? Gerontologist, July 1, 2008; 48(suppl_1): 104 - 113. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Leece Paying the Piper and Calling the Tune: Power and the Direct Payment Relationship Br. J. Soc. Work, June 17, 2008; (2008) bcn085v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. J. Stout, K. J. Hagglund, and M. J. Clark The Challenge of Financing and Delivering Personal Assistant Services Journal of Disability Policy Studies, June 1, 2008; 19(1): 44 - 51. [Abstract] [PDF]
|
|
|
|
|
|
S. Dale, R. Brown, B. Phillips, and B. L. Carlson How Do Hired Workers Fare Under Consumer-Directed Personal Care? Gerontologist, October 1, 2005; 45(5): 583 - 592. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
