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.

Assessing the Appeal of the Cash and Counseling Demonstration in Arkansas, Florida and New Jersey

Publication Date

Stacy Dale, Randall Brown, and Rachel Shapiro

Mathematica Policy Research, Inc.

July 2005

This report was prepared under contract #HHS-100-95-0046 between the U.S. Department of Health and Human Services (HHS), Office of Disability, Aging and Long-Term Care Policy (DALTCP) and the University of Maryland. For additional information about the study, you may visit the DALTCP home page at or contact the ASPE Project Officer, Pamela Doty, at HHS/ASPE/DALTCP, Room 424E, H.H. Humphrey Building, 200 Independence Avenue, SW, Washington, DC 20201. Her e-mail address is:



Medicaid beneficiaries who have disabilities and qualify for assistance with such basic activities as eating, bathing, dressing, and fixing meals typically receive these supportive services through a Medicaid state plan, as personal care services (PCS), or through a Medicaid waiver program, as home- and community-based services (HCBS). By contrast, the Cash and Counseling Demonstration and Evaluation, implemented in Arkansas, Florida, and New Jersey, offered eligible Medicaid beneficiaries the opportunity to receive a monthly allowance to purchase supportive services as they saw fit. Cash and Counseling is intended improve beneficiaries' satisfaction and quality of life by increasing their control over supportive services.

When Arkansas, New Jersey, and Florida began implementing their five-year demonstration programs, in 1998, 1999, and 2000, respectively, they attempted to inform all eligible beneficiaries of the opportunity to participate through some combination of direct mailings, telephone calls, and home visits. The states' enrollment policies helped avert the enrollment of beneficiaries who wished to receive the monthly allowance but would not otherwise use PCS or HCBS (despite being eligible). Florida and New Jersey restricted enrollment to beneficiaries who were already using demonstration-covered services or, in New Jersey's case, had at least been assessed for such services. Although Arkansas enrolled beneficiaries not already using PCS, it required such beneficiaries to agree to pursue PCS from an agency in the event they were randomly assigned to the demonstration control group. In addition, the terms and conditions of the demonstration specified that ratios of new to continuing service users among demonstration participants were not to exceed historic benchmark ratios.

The evaluation set enrollment sample-size targets of 2,000 adults in each state (revised from 3,100) and of 1,000 children in Florida (revised from 1,550), which states expected to meet in about 12 months' time. Although the states eventually met or nearly met the sample-size targets, they took much longer than expected to do so. Arkansas ultimately enrolled 2,008 beneficiaries for the evaluation between December 1998 and April 2001 (29 months). New Jersey enrolled 1,755 beneficiaries between November 1999 and July 2002 (33 months). Florida enrolled 2,820 beneficiaries between June 2000 and July 2002 (26 months, although it met its target of 1,000 children in only 15 months).

Other states considering a Cash and Counseling program are likely to be interested in the number and types of eligible beneficiaries such a program would attract, factors that might deter some interested beneficiaries from participating, and whether offering the program might increase the total number of beneficiaries receiving the Medicaid PCS or HCBS benefit.



This report assesses the appeal of the Cash and Counseling demonstration by: (1) estimating the proportions of eligible beneficiaries that participated and comparing the characteristics of participants and nonparticipants; (2) describing beneficiaries' most common reasons for agreeing or declining to participate; and (3) examining whether the demonstration affected the number of beneficiaries accessing PCS or HCBS over time (that is, program flow).



Data from Medicaid claims for PCS in Arkansas and New Jersey and for HCBS in Florida were used to assess participation and program flow. Claims data were available for the 24 months before and after the first month of evaluation intake--December 1998 in Arkansas, November 1999 in New Jersey, and June 2000 in Florida. Participation and program flow were observed for all 24 intake months or, in the case of Florida, until the evaluation sample-size target was met for a particular age group.

In the participation analysis, beneficiaries were considered eligible for the demonstration if they had a claim for demonstration-covered services during the state's evaluation intake period, met age requirements in their state, and lived in a designated catchment area, if any was used. Beneficiaries were considered to be demonstration participants if they completed a baseline evaluation interview, regardless of random-assignment status. Participants and nonparticipants were compared on age distribution, race, sex, area of residence, mortality, mean monthly costs for PCS or HCBS, and whether they were using services when evaluation intake began. Logistic regression models were used to estimate the odds that beneficiaries participated in the demonstration as a function of their characteristics and PCS or HCBS costs.

In the program flow analysis, monthly ratios of new users of PCS or HCBS to all users were compared before and during evaluation intake. For each month, the denominator of the ratio was the number of beneficiaries who used PCS or HCBS and were old enough to enroll in the demonstration that month. The numerator was the subset of these beneficiaries who had no claims in any of the three preceding months.

Data from anonymous, pre-coded questionnaires were used to assess beneficiaries' reasons for agreeing or declining to participate in the demonstration. For participants, pre-coded reasons pertained to having flexibility and control over services. For nonparticipants, they pertained mostly to the responsibilities associated with consumer direction and satisfaction with current arrangements. Questionnaires were administered after beneficiaries spoke with an outreach worker about the demonstration by telephone or during a home visit, when the decision to participate, or not, was made. Questionnaires were completed by beneficiaries, family members, or outreach workers. States returned hard-copy questionnaires or electronic data files to Mathematica Policy Research, Inc. for analysis. We received questionnaires or data for 1,538 respondents in Arkansas, 4,669 in Florida, and 2,685 in New Jersey. We examined reasons for agreeing or declining to participate for each state, by age group and service use tenure.



Participation. The participants in our analysis samples represented fairly small proportions of the states' known eligibles. During evaluation intake periods of up to 24 months, participants represented 7.8 percent of all known eligibles in Arkansas, 8.2 percent of all known eligibles in Florida, and 6.3 percent of all known eligibles in New Jersey. The participation rate was notably highest among eligible Florida children, at 16.0 percent, and was achieved in 15 months. In Arkansas and New Jersey, somewhat larger proportions of nonelderly than elderly beneficiaries participated, whereas the reverse was true in Florida. The models used to predict the odds of participation suggested that, across states, the variables consistently associated with participation included service use tenure (beneficiaries not using PCS or HCBS when intake began were less likely than others to participate); mean monthly costs for PCS or HCBS (beneficiaries with monthly costs under $300 were less likely than others to participate); and mortality during the intake period (beneficiaries in their last year or two of life were less likely than others to enroll in the demonstration).

Reasons for Agreeing or Declining to Participate. Given a list of nine possible reasons for agreeing to participate in the demonstration, most respondents chose four or five. In all three states the four most common reasons for participating were to have greater control over the hiring of caregivers (selected by 66-88 percent of respondents, depending on the state), paying family members or friends (52-80 percent), obtaining care at more convenient times (63-75 percent), and receiving better or more care (68-74 percent).

Beneficiaries who declined to participate in the demonstration typically cited only one reason for their decision. In Arkansas the most commonly cited reasons were satisfaction with current care arrangements (cited by 46 percent of nonparticipants) and concern that the monthly allowance would not cover needed care (22 percent). Roughly eight in ten nonparticipants in Florida said that they were satisfied with their current arrangements. Substantial proportions of nonparticipants also indicated that they did not want to hire or fire workers (30 percent) or file payroll taxes or track expenses (32 percent). New Jersey nonparticipants overwhelmingly said that they did not participate because they were satisfied with their current arrangements (85 percent), and very few chose other reasons.

Beneficiaries' reasons for agreeing or declining to participate in the demonstration generally differed only slightly by age group or service use tenure.

Program Flow. In Arkansas and New Jersey program flow was quite stable during the 24 months before and after the first month of evaluation intake. Monthly ratios of new PCS users to all PCS users dipped and rose from month to month rather than climbing during the intake period, as would be expected if the demonstration had substantially contributed to the inflow of new users. However, the Arkansas program did temporarily suspend the enrollment of new users because it surpassed the historic benchmark ratio it used to monitor inflow. In Florida many beneficiaries not previously using HCBS began using services for the first time during the year before evaluation intake. Ratios of new service users to all users climbed during that period but then fell during most of the evaluation intake period.

The analysis of program flow was limited in two ways. We could not disentangle the effects of external events on aggregate program flow from the effects of Cash and Counseling. This was particularly limiting in Florida, where a class action lawsuit prompted the state to dramatically increase the supply of HCBS for beneficiaries with developmental disabilities during the year before evaluation intake. We also lacked information about why some demonstration participants who were randomly assigned to the control group did not receive Medicaid PCS during the observation period. This was a limitation particularly in Arkansas, whose target population included eligible beneficiaries who had difficulty accessing traditional services. However, given that states had difficulty meeting their enrollment targets and new service users were less likely than continuing users to participate in the demonstration, we conclude that the Cash and Counseling Demonstration had little effect on program flow during the observed periods in the three states.



The promise of Cash and Counseling--more choice and more control--is simple and compelling. Although the promise resonated clearly with many beneficiaries, it did not attract the participation of large percentages of those known to be eligible. Despite limitations in interpreting program flow trends, the demonstration also did not seem to attract many beneficiaries who were interested in the program allowance but would not use traditional PCS or HCBS.

The fairly low levels of participation may have resulted from challenges states faced in conducting outreach and enrollment activities, from beneficiaries being satisfied with their existing care arrangements or disinclined to assume new responsibilities, or simply from the programs' being new. To varying extents across states, barriers to participation seemed to include concerns about the adequacy of the program allowance, and reluctance to hire and fire workers and handle fiscal responsibilities. States might overcome such barriers in part by ensuring that beneficiaries are aware of the availability of counseling and fiscal services, and by fostering peer-support networks between active and prospective participants.

Older Adults
Location- & Geography-Based Data
State Data
Cash and Counseling Demonstration