Skip to main content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.


The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Consumer Choice and the Frontline Worker

Publication Date
Dec 31, 1993

U.S. Department of Health and Human Services

Consumer Choice and the Frontline Worker

Pamela Doty U.S. Department of Health and Human Services

Judith Kasper, Ph.D. Johns Hopkins University

Simi Litvak, Ph.D. World Institute on Disability

Humphrey Taylor Louis Harris and Associates


PDF Version: (15 PDF pages)

This article appeared in the GENERATIONS issue "Frontline Workers in Long-Term Care" (Fall 1994, Volume 18, Number 3, pp65-70). For additional information about this subject, you can visit the DALTCP home page at or contact the office at HHS/ASPE/DALTCP, Room 424E, H.H. Humphrey Building, 200 Independence Avenue, S.W., Washington, D.C. 20201. The e-mail address is: The Project Officer was Pamela Doty.

The opinions and views expressed in this report are those of the authors. They do not necessarily reflect the views of the Department of Health and Human Services or other funding organizations.

Ever since states and localities seriously began, about 20 years ago, to develop home and community-based long-term-care programs, controversy has brewed over whether--or under what circumstances--it is preferable to provide in-home attendant services through “independent providers” or via federally certified or state-licensed home health and homecare agencies. The former are attendants hired on an individual basis, usually by the clients themselves. Independent providers are usually considered to be either employees of the service recipient (like traditional domestic workers) or “self-employed” contractors (akin to carpenters, plumbers, house painters, and other small business people).

Until recently, the debate over use of independent providers versus agency providers has been framed primarily in terms of trade-offs between cost and quality. Proponents of independent providers have argued that this mode is less costly and therefore allows programs to serve more clients and/or offer more hours of service to individual clients. Proponents of agencies have argued that formal organizational structures and processes are needed to ensure accountability. In this view, agencies are better at establishing safeguards against instances of financial fraud, poor quality care, abuse, neglect, and mistreatment of vulnerable clients, as well as any other accidental or negligent harm that could occur to either clients or attendants while services are being provided. Moreover, proponents argue, agencies can better protect public programs by assuming legal liability for these occurrences.

Recently, advocates from the disability rights and independent living movements have sought to recast this debate into one that pits “consumer direction” against medical or professional dominance. They argue that people with disabilities should be accorded the right to hire/fire, train, and supervise their aides and should either be permitted to pay the aides directly or participate in paying them (by endorsing checks, signing time sheets, etc.). In this view, the “dependency” of people with disabilities who require attendant services is reinforced when government programs or other third-party payers require that recruitment, training, supervision, and payment of attendants be carried out under the direction of medical personnel like registered nurses and/or under the auspices of certified or licensed home health or homecare agencies.


Use of independent providers--or at least the potential to access independent providers rather than being required to receive services from agency-employed attendants--is often employed as a rough proxy measure of the availability and prevalence of “consumer directed” services. The theory is that use of independent providers is more compatible with service recipients having a significant role in hiring and firing, training and supervising, as well as paying their aides.

A survey, carried out by the World Institute on Disability (Litvak and Kennedy, 1990), of all known federal/state or state-only financed programs for homecare as of 1988, found that 75 programs (42 percent of the programs on which information was reported) used independent providers. In many of these programs, use of independent providers was the predominant mode; however, some used independent providers, like local government employees, as an occasional alternative to services that were usually provided via home health or homecare agencies. Most programs that used independent providers (80%) allowed consumers to hire and fire their own aides, and half (52%) allowed consumers to train their aides, but fewer than half (44%) allowed consumers to pay their attendants.

A different approach to defining client direction was taken by the Commonwealth Fund Commission on Elderly Living Alone (1991). The commission sponsored a survey, conducted by Louis Harris and Associates, of a sample of older persons (aged 65 and over) receiving aide and attendant care financed through the Medicaid personal care services (PCS) option in three states (Michigan, Maryland, and Texas). The extent of client direction of PCS aides was operationalized by an “index of choice” consisting of five factors: Is the aide someone the client already knew? Does the client help decide the aide’s work schedule? Does the client sign the aide’s time sheet and/or paycheck? Does the client have responsibility for the aide’s job performance? Does the client--or a family member--participate in the hiring and firing of the aide? On each of these indicators, “yes” responses from clients scored 1 point, “no” responses scored 0--yielding cumulative scores for each client between 0 and 5. The purpose of the index was to determine whether and to what extent consumers who used independent providers actually had more choice and control--and with respect to which specific aspects of service provision--than did client who used agency providers. Clients in Michigan and Maryland used independent providers almost exclusively, whereas clients in Texas were required to receive services through home health agencies.

In Michigan, fully 70 percent of clients had higher scores (3, 4, or 5) on the index of choice, as contrasted with 21 percent of clients in Texas and 25 percent in Maryland. These findings clearly indicate that use of independent providers does not always guarantee a high degree of client direction. Among clients interviewed in the Commonwealth Fund Commission survey, only 25 percent in Maryland reported being involved in setting their aides’ work schedules as compared to 33 percent in Texas and 55 percent in Michigan. On the other hand, nearly two-thirds (62%) of Maryland clients reported signing time sheets or paychecks, a lower rate than in Michigan (78%) but a much higher rate than in Texas (14%).


Detailed case studies of Medicaid PCS programs carried out by the World Institute on Disability (Litvak and Kennedy, 1991), including case studies in Michigan, Maryland, and Texas, provide some explanations for the lower consumer choice scores among Maryland clients as compared to clients in Michigan, even though virtually all clients in both states used independent providers. Maryland was found to limit the range of consumer choice and control in several important ways. First, Maryland interpreted quite strictly the federal prohibition of the employment of “family members” as paid providers under the Medicaid PCS optional benefit. Accordingly, clients were prohibited from employing as independent providers not just close family members like spouses, parents, and children, but also aunts, uncles, cousins, step-parents, step-siblings, and in-laws. Only grandparents (but not grandchildren) could be employed.

In contrast, Michigan chose to interpret the federal prohibition as liberally as possible and banned only the employment of “legally responsible” relatives (i.e., spouses and parents of minors). As a result, nearly half (49%) of all independent providers in Michigan were relatives of the client. An additional 22 percent were friends, neighbors, or persons recommended by friends or relatives. In contrast, most independent providers in Maryland (82%), like most agency-employed aides in Texas (75%), were hired because they were recommended by an agency, caseworker, nurse, or other professional. Interestingly, Texas had less stringent prohibitions on the employment of family members than did Maryland and Michigan; officials in Texas maintained that if clients wanted family members, friends, and neighbors to be their aides, it was possible to arrange for them to be employed through agencies. In practice, however, it seems that this rarely occurred, although 13 percent of Texas clients in the Commonwealth Fund Commission survey (as compared to 5% among Maryland clients surveyed) reported that their aides were family members. Seventy-four percent of clients in Maryland and 64 percent of clients in Texas, as contrasted with 27 percent of clients in Michigan, said that they were not acquainted with their aides prior to the aides’ employment.

Another factor found to limit the range of client choice in Maryland was the use of registered nurse “case monitors” to provide for quality assurance and client protection, and to satisfy the federal requirement for nurse supervision of Medicaid personal care services. Some nurse case monitors were public health nurses directly employed by county governments who did case monitoring of Medicaid personal care services along with other duties; others were employed under contract to the state and did only case monitoring. The nurse case monitors were reported to be heavily involved in recruiting aides for clients--often individuals who were working or had worked in the past for other Medicaid personal care clients. When the state of Maryland originally established its Medicaid personal care optional benefit in 1980, the intent was to emulate Oklahoma, where clients recruited independent providers mainly from among their friends and neighbors. However, in Maryland, this pattern never developed. In contrast to Maryland, the Oklahoma model involved only the most minimal supervision by registered nurses (i.e., an annual record review and no face-to-face supervision). Several persons interviewed by the World Institute on Disability researchers suggested that the strong role of R.N. case monitors in Maryland discouraged the hiring of individuals from the client’s social circle in favor of aides known to and recommended by the nurses.

Client involvement in training and supervision of their independent providers was also circumscribed in Maryland. Maryland required that aides be trained and supervised by the R.N. case monitors who, in turn, were required to make supervisory home visits at least every 60 days. In contrast, Michigan permitted clients to train and supervise their own aides. The federal mandate for nurse supervision was met in Michigan through an annual record review.

Finally, different service authorization and aide payment systems across the three states created somewhat different behavioral incentives for workers, with different implications for consumers. Michigan considered aides to be employees of the client and endeavored to make this concept meaningful through the use of a “two-party check,” which, though issued by the state, required the client’s cosignature before it could be cashed by the aide. Furthermore, both independent providers in Michigan and agency-employed aides in Texas were paid by the hour. By contrast, independent providers in Maryland were paid daily “visit” rates, because for purposes of Social Security and unemployment insurance, this would permit them to be considered “self-employed” independent contractors. The Maryland program assigned service recipients to three levels of need, depending on their level of ADL (activities of daily living) impairment. Aides were paid $10 per day for Level I recipients, $20 per day for Level II recipients, and $25 per day for Level III recipients. These levels were supposed to correspond, roughly, to 2-hour, 4-hour, and 5-hour “visits.” Independent providers in Maryland were permitted to work for up to four Level I clients at the same time, or two Level II clients, or one Level II and one Level III client, if both clients shared the same residence. As a result, aides in Maryland who agreed to work for Level III clients were effectively restricted to part-time work, since they were generally prohibited from taking on additional clients, and their daily rate for providing service to Level III clients was based on an assumption of five hours of work. Not surprisingly, there were reports that Level III clients had greater difficulty recruiting attendants. Maryland’s payment method may also have created incentives for aides to rush through their assigned tasks since, although they had to show up daily, they were not actually required to work a set number of hours.


The Commonwealth Fund Commission survey found that across all three states in which Medicaid personal care services clients were interviewed, there were high overall levels of client satisfaction with aides (see Table 1). Across all three states, 94 percent of clients reported being either very or somewhat satisfied with their aides; only 5 percent reported that they were not very or not at all satisfied. When asked more specifically if there was any one thing about their aides that they would like to change, 81 percent of clients said “nothing,” and 8 percent said “improve housekeeping.”

Greater satisfaction with aides was related to the following:

  1. Clients’ overall life satisfaction (i.e., clients who were very satisfied with their lives in general were more likely to report being very satisfied with their aides).
  2. Having known the aide prior to the aide’s employment.
  3. Length of time the aide had been with the client (i.e., client satisfaction was much higher with aides who had been working for more than two years as compared to those working less than six months).
  4. Number of hours per week that the aide worked for the client (i.e., satisfaction with aides working more than 15 hours per week was greater).

There is evidence that clients did not report high levels of dissatisfaction simply because--as critics of consumer satisfaction surveys often allege--they were grateful just to have any services and fearful that complaints about poor service would only result in loss of access to needed assistance. Twenty-eight percent of clients surveyed reported that they had sought to change aides in the past and, of this group, 93 percent said that they were successful in getting a different aide. This suggests that if clients are satisfied with their aides, it is because they do not put up passively with aides who are unsatisfactory--regardless of whether aides are independent providers or agency employees.

However, clients whose scores on the choice index were lower tended to report less satisfaction with their aides than those whose scores were higher. Thus, 59 percent to 78 percent of clients whose scores on the choice index were 0 or 1 reported that they were very satisfied as compared to 96 percent to 100 percent of those with scores of 4 or 5. Satisfaction rates (very satisfied versus somewhat satisfied) were notably higher in Michigan than in either Texas or Maryland--in keeping with the higher choice scores in Michigan.

Respondents with higher scores on the choice index were also more likely to perceive their aides as having positive attributes. These included being “very concerned” about their well-being; more like a friend than an employee; someone with whom they could discuss a problem, who made them feel “very safe,” who was “always” reliable, and who had improved their quality of life “a great deal.”


Although the Commonwealth Fund Commission survey did not interview PCS aides, it provides some evidence--albeit indirect--of worker satisfaction in the form of lower reported absenteeism and turnover, associated with higher levels of consumer choice. While 73 percent of all clients surveyed reported no absenteeism by their aide during the previous month, the proportion reporting no absenteeism rose steadily from only 49 percent of clients scoring 0 on the index of choice to 97 percent of clients with choice index scores of 5. Similarly, 79 percent of all clients reported that they could always count on their aides to be at their homes on time, but the proportion of clients reporting such high reliability rose from 59 percent of clients scoring 0 on the index of choice to 100 percent of those scoring a perfect 5.


Of course, worker morale is generally believed to be, at least partly, a function of pay and benefits. In the three states included in the Commonwealth Commission survey, the World Institute on Disability case studies found little variation in wage rates, despite differences in provider mode (an hourly rate of $3.35-$6 for independent providers in Michigan, $3.35-$4.41 for agency employees in Texas, and a per visit rate that translated into roughly $5 per hour for independent provides in Maryland). Independent providers received no benefits in Michigan or Maryland. FICA withholding could be arranged if requested in Michigan. In Maryland, attendants were considered to be independent contractors responsible for paying their own taxes. Texas agencies routinely provided FICA withholding and paid the employer’s share as well as employment compensation; some provided workers’ compensation and transportation costs--but not vacation, sick leave, or health insurance. Thus, differences in pay and benefits by mode across the three states do not appear to provide a basis for explaining client-reported differences in employee performance.

However, from a national perspective, there is evidence that, on average, independent providers tend to receive lower pay and benefits than do agency providers. The World Institute on Disability’s 1988 survey of all known state and federal/ state-financed programs for personal attendant services found that the average hourly wage for independent providers across all programs was $4.82 ($3.51 for family members), with the number of benefits averaging 1.1. In contrast, the average hourly wage was $6.98 for employees of certified home health agencies ($6.60 for homecare agency employees), with 3.2 benefits. However, differences in payment rates by provider mode were strongly influenced by payment source. Medicaid home and community-based waiver programs tended to show greater differentials in terms of higher pay and benefits for agency employees versus independent providers. At the same time, Medicaid waiver programs also paid higher rates to both independent and agency providers than did Medicaid personal care option programs or other funding sources like the Social Services Block Grant, Title III of the Older Americans Act, and state-only revenues. It should be noted, however, that waiver programs tend to be much smaller in terms of numbers of clients served than are programs funded by the Medicaid personal care option; moreover, waiver programs serve individuals who are, on average, more disabled because they must be certified as needing a nursing home level of care.

Home health/homecare agencies that lobby or negotiate for higher rates often claim to be acting as advocates for adequate pay and benefits for direct care workers, and not merely to obtain higher profits. There is some anecdotal evidence to support such claims. Certainly, home health agencies, both individually and collectively through their trade associations, have much greater potential influence in this regard than do the thousands of unorganized clients who employ independent providers. However, agencies also typically charge overhead and profit margins that add 30 to 40 percent to the hourly cost of service--over and above wages and benefits paid to direct care workers. In contrast, the administrative costs to states of independent providers can be limited to payrolling and acting as “fiscal agent” for purposes of tax withholding and payment of unemployment and workers’ compensation. California’s In-Home Social Services program employs a computer management firm to be “fiscal agent” on behalf of clients who use independent providers, and these administrative costs amount to about 0.5 percent of the program’s annual budget. Even Maryland’s use of contract nurses to train and supervise personal care attendants only added 10 percent administrative costs to payments to independent providers. Thus, the unit costs to the state per hour of attendant service tend to increase significantly when services are provided through agencies, whether or not agencies also successfully negotiated higher wages and benefits for attendants. Unless state legislatures are willing to appropriate additional funds for home and community-based long-term-care budgets, program administrators must compensate--typically by raising eligibility cut-offs and serving fewer clients or by reducing the number of hours of service authorized per client.

These trade-offs are vividly illustrated by what happened when, in 1987, the state of Montana shifted from the use of government (county) employed attendants and independent providers to the use of a single contract home health agency (HHA) to provide Medicaid PCS. Initially, the hourly amount ($3.85) allocated for direct care workers by the state remained the same, but the contract HHA withheld 50 cents for taxes and benefits. The state paid the contract agency an additional $1.40 per hour for nurse supervision and administration, including 19 cents to cover the employer’s share of FICA. Some aides (who may previously have been avoiding their self-employment tax liability) protested this “pay cut.” The year after the new system went into effect, the contract agency determined that attendant wages were too low to retain attendants and that the amount allocated for administrative overhead was insufficient. The agency and the state negotiated a new contract that raised attendant wages by 50 cents per hour, and in subsequent years, wages were increased again. Administrative overhead payments to the contract agency also rose. By 1991, the state was paying the agency $7.75 per hour of service provided, of which $5.52 (70%) went to attendant wages, taxes, and benefits (including overtime), with the remainder allocated to such administrative costs as scheduling, nurse supervision, and training, along with profit margin. In order to accommodate these higher unit costs per hour of service while holding down overall Medicaid PCS expenditure increases, the state cut back on the hours of PCS that could be authorized per person. During the last year in which independent providers were used, the maximum per capita authorization was 70 hours per week; by 1991, the cap had been lowered to 40 hours per week.


In recent years, public programs that provide attendant services through independent providers have increasingly run afoul of the Internal Revenue Service, the Social Security Administration, and federal and state labor departments over issues of tax payment and withholding and provision of workers’ and unemployment compensation. It appears that states will find it more and more difficult in the future to sidestep these issues by classifying independent providers of in-home attendant services as “independent contractors.” This is because the IRS has estimated that as many as 3-4 million workers nationally (of whom only a fraction are in-home attendants whose services are paid for with public funds) are misclassified as independent contractors. The resulting loss of income, FICA, and FUTA tax revenue has been estimated at approximately $1.5 billion in 1989. The IRS further estimates that whereas only about 2 percent of workers characterized as employees regularly fail to file tax returns, approximately 28 percent of those characterized as independent contractors fail to do so.

A small minority of clients, mostly young adults who are severely disabled physically but not mentally, are quite willing, even eager to assume employer responsibilities--provided that the actual costs of the employer’s share of FICA as well as workers’ compensation and unemployment insurance coverage are fully reflected in cash payments from public programs. For most clients and family members, however, the paperwork burden would likely be confusing and rather frightening. Fortunately, states are finding that they can resolve their problems with the IRS and continue to use independent providers by agreeing to become, or to arrange for other organizations to become, “fiscal agents” for purposes of tax with holding while still maintaining the client as the attendant’s “employer of record” (Flanagan, 1994).

Requirements with respect to workers’ compensation vary across states. Only one state (New Hampshire) mandates workers’ compensation across all occupational categories. About 25 state programs now have some coverage for domestic workers. In some states, court rulings have exempted the state from responsibility for paying workers’ compensation or otherwise accepting liability for independent providers’ on-the-job injuries, on the grounds that independent providers are employees of the client, not the state, or are independent contractors. In other states, court rulings have established that the state must pay for insurance coverage or otherwise cover the costs of independent providers’ work-related injuries--but that this obligation does not make the independent provider an “employee” of state or county government for other purposes. Unions have attempted--so far unsuccessfully--to use workers’ compensation cases to win collective bargaining rights for nonagency-employed in-home attendants by claiming that they are really government employees subject to civil service rules and regulations. A law recently enacted (August 1993) in California gives the state the authority to develop a “public authority” with a board of directors made up of consumer representatives that could, among other things, act as the IRS recognized fiscal agent, purchase liability insurance, conduct background checks of independent providers on behalf of clients, and also bargain collectively with unions representing homecare attendants.

A larger concern to state officials in terms of potential financial exposure is tort liability for injuries inflicted by aides on clients. However, except for New York--where all aides are employed by agencies that assume all tort liability for them--there has been virtually no such litigation to date.


The results of the Commonwealth Fund Commission survey of client choice and satisfaction among older persons with disabilities receiving in-home attendant services financed under the Medicaid personal care services optional benefit suggest that higher levels of consumer satisfaction are associated with greater consumer involvement in hiring and firing, scheduling, supervising, and paying in-home attendants. It appears that client choice is facilitated by an independent provider mode of service provision with minimal professional supervision. Many advocates of improved pay and benefits for homecare workers have been skeptical of the independent provider mode of service provision, believing that frontline workers are better off as employees of home health or homecare agencies. At least some of these concerns about “exploitation” of aides who work as independent providers are likely to be alleviated in the future as increased regulatory oversight leads states to set up organizational structures to act as “fiscal agents” for tax and insurance purposes, while still allowing clients to perform the recruitment, training, and other day-to-day managerial functions of employers vis-à-vis their aides. Some of the newer organizational structures under discussion could even make it possible for independent providers to bargain collectively with public programs for higher wages and benefits. Research is still needed, however, to determine whether worker satisfaction, like client satisfaction, is promoted by “client-directed” models of service provision.

Pamela Doty is senior policy analyst, Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services, Washington, D.C. Judith Kasper, Ph.D., is associate professor, School of Hygiene and Public Health, Johns Hopkins University, Baltimore, Md. Simi Litvak, Ph.D., is research director, World Institute on Disability, Oakland, Calif. Humphrey Taylor is present and CEO, Louis Harris and Associates, New York.


Commonwealth Fund Commission on Elderly Living Along, 1991. The Importance of Choice in Medicaid Home Care Programs: Maryland, Michigan, and Texas. New York: Louis Harris.

Flanagan, S., 1994. Consumer-Directed Attendant Services: How States Address Tax, Legal and Quality Assurance Issues. For the Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services, Cambridge, Mass.: SysteMetrics/MEDSTAT Systems, Inc. []

Litvak, S. and Kennedy, J., 1990. New Models for the Provision of Personal Assistance Services. For the National Institute for Disability and Rehabilitation Research (NIDRR Grant Number G008720134). Oakland, Calif.: World Institute on Disability.

Litvak, S. and Kennedy, J., 1991. Case Studies of Six State Personal Assistance Services Programs Funded by the Medicaid Personal Care Option. For the Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services. Contract # HHS-100-89-0025. Oakland, Calif.: World Institute on Disability. []

TABLE 1: Consumer Choice and Satisfaction
For the survey sponsored by the Commonwealth Fund Commission in 1991, an “index of choice” was constructed, based on the following five factors:
  1. Did the homecare recipient know his or her aide before the aide began providing services?
  2. Does the recipient help schedule the time at which the aide comes to the home?
  3. Does the recipient sign the aide’s timesheet and/or paycheck?
  4. Does the homecare recipient have responsibility for the aide’s job performance?
  5. Does the recipient or a family member participate in the hiring and firing of aides?

Affirmative responses to any of these indicators earned one point each; negative responses earned zero points. Thus cumulative scores for each respondent ranged from 0 to 5.

  Sample Average State Index of Choices
Maryland Texas Michigan 0 Indicators 1 Indicator 2 Indicators 3 Indicators 4 Indicators 5 Indicators
Base 879 300 303 276 51 209 282 208 92 37
Very satisfied 84% 76% 87% 90% 59% 78% 85% 86% 96% 100%
Somewhat satisfied 10% 15% 9% 6% 16% 13% 10% 10% 3% --
Not very satisfied 4% 6% 2% 3% 14% 5% 3% 3% 1% --
Not at all satisfied 1% 3% 3% 1% 12% 4% 1% * -- --
SOURCE: Commonwealth Fund Commission on the Elderly Living Alone, 1991. * Because of the urban and rural quotes in each state and the selection of specific geographic areas, the term “sample average” not “total” is used. The “sample average” is not projectable to any universe.
Cash and Counseling Demonstration