About 1.2 million individuals receive noninstitutional supportive services through state Medicaid plans or home- and community-based waiver programs (LeBlanc et al. 2001; Kitchener and Harrington 2001).1 Under state plans, services are largely restricted to human assistance with personal care and homemaking and must be provided by licensed home care agencies. These agencies recruit, train, schedule, and supervise the staff who assist beneficiaries. Waiver programs may also offer other disability-related goods and services, but their coverage is often limited and someone other than the beneficiary (namely, a case manager) decides whether they are needed. This system of care, while adequate for many recipients, has been criticized for over-medicalizing services and not being flexible enough to meet recipient needs effectively. Moreover, a perennial shortage of workers providing this help, stemming from its low pay, poor benefits, and physically and emotionally demanding nature, worsens when the economy is strong and is likely to deepen as the U.S. population ages and demand for workers increases.
In contrast to these traditional service models, states are increasingly offering Medicaid beneficiaries the opportunity to obtain personal care from individual providers (Velgouse and Dize 2000). This alternative has come to be known as consumer-directed care, since beneficiaries who use individual providers assume the employers role of hiring, managing, and possibly terminating their workers (Eustis 2000). An expanded model of consumer direction would allow beneficiaries to manage not only their human assistance but also other covered supportive goods and services. Consumer direction is based on the premise that personal care does not require the intervention of medical professionals because it is low tech and nonmedical. Rather, consumers should be empowered to make informed choices about their assistance and provided with supports to take control of it (Benjamin 2001; Stone 2001; Eustis 2000; Stone 2000; and Doty et al. 1996).
For people with disabilities, supporting choice and control over personal assistance has the potential to better meet individual needs and resonates strongly with basic American values, as re-affirmed in the Administrations New Freedom Initiative. Moreover, consumer direction could enlarge the pool of personal care workers if consumers can hire family or friends to help them. Finally, consumer-directed care might be less costly because agencies would no longer be responsible for hiring, training, and supervising workers (Stone 2000; and Eustis 2000).
Nevertheless, consumer-directed personal assistance in a publicly funded program, like Medicaid, raises many concerns. These include (1) how to ensure care quality and establish accountability for adverse consumer outcomes; (2) how to ensure the benefit is used appropriately; (3) how to ensure workers are trained adequately and treated fairly; (4) whether family members who might otherwise help without pay may be hired as workers; and (5)whether consumer direction should be available to individuals with cognitive deficits (Benjamin 2001; Kane and Kane 2001; Kapp 2000; and Doty et al. 1996). Despite these concerns, in 1999, there were an estimated 139 publicly funded consumer-directed personal assistance programs in the United States (Flanagan 2001).
Cash and Counseling Model
Cash and Counseling is an expanded model of consumer-directed supportive services in that it provides a flexible monthly allowance that beneficiaries--as consumers--may use to hire their choice of workers, including family members, and to purchase other goods and services (as their state permits). Cash and Counseling requires consumers to develop plans showing how they would use the allowance to meet their personal care needs and provides counseling and fiscal assistance to help them plan and manage their responsibilities. Consumers who are unable or unwilling to plan and manage their care themselves may designate a representative, such as a family member, to help them or to do it for them. These features are meant to make Cash and Counseling adaptable to consumers of all ages and with all types of disabilities.
With funding from The Robert Wood Johnson Foundation and the Office of the Assistant Secretary for Planning and Evaluation of the U.S. Department of Health and Human Services, the Cash and Counseling Demonstration and Evaluation was implemented as a voluntary demonstration in three states--Arkansas, Florida, and New Jersey.2 Because the Medicaid programs and political environments differed considerably across the demonstration states, they were not required to implement a standard intervention, but had to adhere to basic Cash and Counseling tenets, as summarized above. Because of such differences, Mathematica Policy Research, Inc. (MPR) is evaluating each program separately.
Cash and Counseling Evaluation
MPRs evaluation of the three Cash and Counseling demonstration programs addresses four broad questions: (1) Who participated in Cash and Counseling? (2) How was it implemented? (3) How did it affect consumers and their caregivers? (4) How did it affect public costs? To estimate impacts on consumers and their caregivers and on public costs, the evaluation randomly assigned interested, eligible Medicaid beneficiaries to receive either Cash and Counseling benefits (the treatment group) or personal assistance services (PAS) as usual from Medicaid-certified agencies (the control group). It then is comparing the groups outcomes, based on responses to telephone interviews and Medicaid and Medicare claims data. In addition, the evaluation will look at beneficiaries reasons for participating or declining to participate in Cash and Counseling, based on questionnaires filled out when the beneficiary made this decision. It will also use beneficiary-level Medicaid PAS data to investigate trends in state PAS use for indirect evidence of demonstration-induced demand for PAS.3
This report, which addresses the second broad evaluation question, describes the implementation of the Arkansas demonstration, IndependentChoices, by considering:
- What were the major goals and features of IndependentChoices?
- How well did consumers manage the responsibilities of the program?
- How did consumers take advantage of the increased flexibility the program offered?
- How did consumers like the program? For what types of consumers did it appear to work best, and for whom did it work less well?
- What lessons does IndependentChoices offer policymakers and program developers about consumer direction?
Sources and Methodology
Data and information for this report come from (1) on-site discussions with program staff conducted in March 2000;4 (2) a mail survey of program counselors conducted in May and June 2000; (3) telephone interviews with consumers when they enrolled in the demonstration, and four months and nine months later; and (4) data the IndependentChoices program supplied. The report also incorporates the insights of National Program Office staff.
In addition to describing the implementation of IndependentChoices from the perspective of state program staff, this report includes a primarily descriptive examination of consumer and counselor interview responses and program data. We used regression analysis, however, to assess whether certain consumer characteristics affected their experiences and satisfaction with the program, by controlling for other factors on which consumer groups might differ. The regressions included a set of explanatory variables drawn largely from baseline interview data. The report presents selected statistics but references additional tables in the appendix.
Characteristics of IndependentChoices Consumers
Arkansass Cash and Counseling demonstration program was open to elderly and nonelderly adults eligible for Medicaid PAS and enrolled beneficiaries for its evaluation from December 1998 to April 2001. (Beneficiaries continued to enroll in the demonstration after April 2001, but were not part of the evaluation.) The program provided treatment group members, referred to in the remainder of this report as consumers, a monthly allowance based on the number of hours of PAS for which they had been assessed. It also provided, at no direct cost to consumers, supportive services: counseling (for example, to help consumers develop an allowance spending plan) and optional bookkeeping services (for example, to pay and withhold taxes for workers hired with the allowance). The IndependentChoices allowance and support services were offered to 1,004 consumers who had been randomly assigned to the evaluation treatment group. The typical consumer was elderly (age 65 or older), white, female, and living with someone other than a spouse. Just over a third of consumers reported living in a rural part of the state. Nearly all had paid or unpaid help with personal care and household activities when they enrolled in the program; however, about two-thirds reported they needed more help. Slightly less than half had been receiving publicly funded home care for more than a year when they enrolled in IndependentChoices. Many consumers were generally satisfied with the paid help they had been receiving; however, nearly a third were not (Table 1).
In addition to presenting data for all IndependentChoices consumers, we also compare the program experiences of elderly and nonelderly consumers. We hypothesized that elderly and nonelderly consumers might differ in their ability to manage the allowance, informal care resources, attitudes toward personal assistance, or desire to control that assistance.
| TABLE 1: Consumer Characteristics at Enrollment
| Age 18 to 64
|Age 65 to 79
|Age 80 or Older
| Race Self-Identified as White Only
|Race Self-Identified as Black
|Self-Identified as Some Other Racea
| Lived Alone
|Lived With Spouse Only
|Lived With Others
| Lived in Rural Area
| Had Paid or Unpaid Help With Personal Careb
|Had Paid or Unpaid Help With Household Activitiesc
|Needed More Help With Personal Careb
| Needed More Help With Household Activitiesc
| Had Been Receiving Public Home Care For 1 Year or Mored
| Was Satisfied Overall With Paid Help and Related Goods
|SOURCE: Age and sex come from IndependentChoices program records. All other data come from MPR baseline interviews conducted with IndependentChoices participants between December 1998 and April 2001.
NOTE: Table includes responses for 1,004 consumers in evaluation treatment group.
- Only 12 of the 1,004 consumers considered themselves of Hispanic (or Latino) origin when asked in the baseline interview.
- Personal care includes bathing, transferring from bed, eating, and using the toilet. Questions refer to week before baseline.
- Household activities include light housework, yard work, meal preparation, and shopping. Questions refer to week before baseline.
- Public home care includes help at home from someone who was paid by Medicaid or some other public program.
As a basis for understanding differences (if any) in program experiences between elderly and nonelderly consumers, we examined differences in the two groups at enrollment. While elderly and nonelderly consumers were similar on most dimensions at enrollment, they did differ in a few. Elderly consumers had less formal education, were less likely to be new to publicly funded home care, and were more likely to be satisfied with current PAS (Appendix Table A.1 and Table A.2). In addition, proxy respondents to the baseline interview were much more common for elderly consumers than nonelderly consumers (57 versus 25 percent).
Of the 1,004 consumers, 924 responded to the evaluations four-month interview and 885 to the nine-month interview. The baseline characteristics of the responding consumers were nearly identical to those of the original 1,004 enrollees. As for the baseline interview, proxy respondents for the follow-up interviews were much more common for elderly than nonelderly consumers.5
Characteristics of IndependentChoices Counselors
IndependentChoices counselors were surveyed about a year and a half after the program started operating. Five had backgrounds in social work, one was a nurse, and one was a speech pathologist. Five were college educated. All had worked for the program for between 8 and 21 months at the time of they were surveyed (Appendix Table A.3).