|Home | About | Journals | Submit | Contact Us | Français|
To describe the key features of the first three Cash and Counseling demonstration programs and identify those which the programs have in common and those on which they differ.
Demonstration Cash and Counseling programs in Arkansas, Florida, and New Jersey.
Interviews with program staff and program materials and discussion.
Description and comparison.
The three Cash and Counseling demonstration programs have many common features; these arose because they were required to adhere to the basic tenets of the Cash and Counseling model and to federal regulations, to share the philosophy of consumer empowerment, and to work together to design their programs. However, their programs also differ substantially due to differences in state goals, Medicaid programs, and political environments. These variations must be taken into account to understand differences in the impact of the three programs.
In Arkansas, Florida, and New Jersey, Cash and Counseling was offered as an alternative to traditional, case-managed supportive services in which individuals with disabilities receive personal care from agency employees, who are previously unknown to them. While such agency care has important advantages, its recipients typically have limited choice about how and when their care is provided as well as by whom (Mahoney, Simone, and Simon-Rusinowitz 2000). Some traditional programs offer more than personal care services; access to these additional goods and services is controlled by case managers (Doty, Kasper, and Litvak 1996).
Obviously, a program's design may affect its results. Program design may affect the number of individuals who enroll, with possible effects on program efficiency (as well as on the ability to measure impacts). Design may affect the characteristics of enrollees, which, in turn, may affect program results as individuals may vary in their ability to benefit from the program. If a program faces opposition, its implementation may depend upon whether the program's designers are able to address the opposition's concerns, or at least avoid depending on the opposition for referrals, information, or staff resources. Designing methods to limit potential for exploitation of consumers may be critical to ensuring adequate quantity and quality of care. The financial success of a program may depend on the design of methods to control the cost of direct services (in this case, the cash allowance) and to administer the program.
The grant solicitation for states to design and implement Cash and Counseling programs laid out the basic tenets of such a program. The funders required that a program offer a flexible allowance that consumers could use to hire their choice of workers and to purchase other goods and services. Further, consumers were to be required to develop a plan for the use of the allowance to meet their needs for supportive services and then to manage their services under the plan. Consumers who were unable or unwilling to manage their services themselves were to be allowed to designate a representative, such as a trusted family member or friend, to help them. Further, programs were required to provide counseling (to assist consumers with planning and managing their services) and fiscal services (to assist consumers with payroll and bookkeeping).
The Cash and Counseling programs were expected to operate as demonstrations under Section 1115 of the Social Security Act. Thus, they were required to adhere to the federal regulations governing an 1115 demonstration, including those limiting eligibility and federal funding.
Consistent with these tenets and regulations, each of the three states made decisions that shaped its Cash and Counseling program. Their decisions were influenced by the philosophy of consumer empowerment, to which each was committed. Many decisions were shaped by the work of the Cash and Counseling National Program Office (NPO), and some were the joint work of all three programs in collaboration with that office. (The NPO was funded by The Robert Wood Johnson Foundation to assist the states with demonstration design and implementation.) At the same time, each state had somewhat different goals and operated in a different Medicaid and political environment. The result is that the three programs have many features in common, but each also has features that set it apart. Table 1 summarizes the key features of the three Cash and Counseling programs.
Providing background for the other papers in this volume, this paper describes the major program design decisions faced by the demonstration states, factors and issues shaping their decisions, and how each responded—giving rise to commonalities and differences in their programs. It also illustrates how such differences might lead to differences in program impacts.
The design decisions faced by the three demonstration states involved all major components of the Cash and Counseling model: choice of program to which Cash and Counseling was to be an alternative, eligibility and appropriateness, outreach and enrollment, permissible services under the allowance, determination of the amount of the allowance, and organizing and paying for counseling and fiscal services.
The primary data sources for this paper are on-site interviews with program staff and other stakeholders conducted about 18 months after each program began operating, as well as program materials and discussion with key program staff for the duration of the evaluation. The features of each program are described in detail in three reports (Phillips and Schneider 2002, 2003, 2004).
The central question the funders posed for the demonstration was: How does Cash and Counseling compare with traditional case-managed supportive services? States interested in participating in the demonstration were free to propose providing Cash and Counseling as an alternative to Medicaid personal assistance services, which is an optional Medicaid state-plan service; Medicaid home- and community-based waiver services; or programs funded by state general revenues. Any of these could be “cashed out” (i.e., considered in the calculation of the allowance), provided they were not consumer-directed programs.
The funders expected states to include elderly people whose functioning was impaired, as well as nonelderly adults with physical disabilities. A state was free to also include other population groups. As states decided which populations to serve, they were aware that the program's success might differ by population. When the solicitation for the demonstration was issued in 1996, younger adults with physical disabilities had long been advocating for consumer-directed care (DeJong 1979). However, the appropriateness of the consumer-directed model for frail elderly people was being debated at that time (Simon-Rusinowitz and Hofland 1993).
Arkansas and New Jersey offered the allowance in lieu of a single service, Medicaid state-plan personal care, to frail elderly adults and nonelderly adults with physical disabilities. Arkansas rejected also cashing out a home- and community-based waiver program due to concern that doing so would increase the opposition of providers of traditional services. Florida, which had no state-plan personal care benefit, decided to offer Cash and Counseling as an alternative to services under multiple Medicaid home- and community-based waivers to adults and children with developmental disabilities as well as to frail elderly adults and nonelderly adults with physical disabilities. The Florida waivers cover a broad array of services including homemaker and personal assistance services, environmental modifications, adaptive equipment, medical supplies, environmental modifications, therapy services not otherwise covered by Medicaid (such as occupational therapy), and case management (for elderly and younger adults with physical disabilities) or support coordination (for children and adults with developmental disabilities). Florida decided to cash out all waiver services except case management and support coordination.
The difference across the states in the range of services covered could lead to differences in the effects of Cash and Counseling. For example, the broader scope of services covered in Florida's existing program might lead to reduced effects on unmet need, relative to the other two programs.
Having identified the program(s) in which participants could chose the cash-out option, each state had to decide whether beneficiaries eligible for, but not receiving services from a cash-out program, would be eligible for Cash and Counseling. Expanding access to care to those not already receiving services carried the risk of inducing demand for services and hence increasing total program cost to the extent that beneficiaries who were not interested in traditional agency services found an allowance attractive. This risk was limited by a demonstration 1115 waiver cap on the proportion of Cash and Counseling enrollees who had not been receiving services in the cash-out program before their enrollment in Cash and Counseling. This cap was the average historical proportion of cash-out program participants who began receiving those services in a given year.
Only Arkansas decided to allow those eligible for, but not receiving services in the cash-out program to enroll in Cash and Counseling. One of Arkansas' major goals was to assess whether the Cash and Counseling model could serve people who were not being served by the traditional system, such as beneficiaries living in rural areas with no home care agencies. This feature made increased access to care and also increased state cost more likely in Arkansas than in the other two states.
The three demonstration states considered—but rejected—screening out eligible beneficiaries who might be inappropriate for Cash and Counseling. Concern about the possibility of abuse of the allowance by beneficiaries with a history, say, of substance abuse led them to consider screening. On the other hand, they were concerned about the feasibility of developing a valid screening process. If the process was not legally defensible and a beneficiary chose to contest exclusion from the program, a state might be held liable. In light of potential liability and mindful that exclusion would be inconsistent with the philosophy of consumer direction, all three states decided against implementing structured screening criteria. Instead, each program explained the rights and responsibilities of consumers and representatives and allowed eligible beneficiaries to decide whether to try it. However, due to concern that allowing “all comers” to participate increased the risk of abuse of the benefit and exploitation of consumers, each state empowered the program director to revoke the right of a consumer to participate in Cash and Counseling in the event of abuse or exploitation (but very few consumers had this right revoked).
Outreach to the community was necessary to generate interest in participation in Cash and Counseling because the model was relatively unknown and to meet the evaluation sample size targets. Each of the three demonstration states had to decide who would be responsible for outreach and for subsequent enrollment of interested beneficiaries. Responsibility for enrollment might be vested in providers of traditional services. Because such providers might couple outreach and enrollment with other tasks, such as professional assessment and care planning, this approach promised efficiency. On the other hand, if traditional providers were resistant to the Cash and Counseling approach, assigning this task to them might compromise objectivity and result in reduced enrollment.
Concerned that opposition from traditional providers would adversely affect outreach and enrollment, both Arkansas and New Jersey decided against assigning them responsibility for these tasks. Rather, Arkansas hired state employees and New Jersey hired contractors to conduct outreach and enrollment. In Florida, case managers (employed by agencies under contract to the state) and support coordinators (working as independent practitioners or as employees of agencies) already had responsibility for presenting all new services and programs to current recipients of waiver services. In light of its limited funding, Florida initially asked case managers and support coordinators to take responsibility for both outreach and enrollment for Cash and Counseling. However, when enrollment lagged far behind that of the other two states, Florida used grant funds to hire temporary state employees (outside civil service) to take responsibility for enrollment.
The funders of the demonstration did not envision a program that provided unfettered cash, but they insisted that states permit the allowance to cover a broad range of goods and services—essentially any good or service that helped a consumer to function more independently. Each of the states had to decide how to ensure that purchases were permissible while simultaneously permitting great flexibility.
The terms and conditions for the 1115 demonstration waiver allowed consumers to hire any relative, including legally liable ones (spouses and parents of minor children) but they were silent on whether consumers could hire their representatives as workers. Florida and New Jersey opted to allow legally liable relatives to be workers. Arkansas did not because it was concerned that doing so would be politically controversial. Because of an inherent conflict of interest, neither Arkansas nor New Jersey allowed consumers to hire their representatives as workers. Mindful of the situation of a single parent caring for a child with developmental disabilities, Florida decided to allow a representative to be a worker also. These differences in constraints on hiring legally liable relatives and representatives might lead to differences in program results across the three states. For example, allowing the hiring of representatives could potentially lead to greater abuse of the benefit in Florida than in the other two states.
While the development of purchasing plans was a tenet of the Cash and Counseling model, the three demonstration states developed different approaches for the approval of these plans. Arkansas compiled a list of goods and services that were clearly permissible. Approval by Arkansas state program staff was required for initial and revised purchasing plans containing any service not on the list; counselors had the authority to approve all other purchasing plans. Both Florida and New Jersey required that each initial and revised purchasing plan be reviewed by state program or district-level staff. Florida and New Jersey were aware that approval of purchasing plans could be delayed if a bottleneck developed at the state or regional level, with possible consequences for consumer satisfaction and other program outcomes. However, each of these two states wanted to retain control over the approval process at least until it had more experience with the Cash and Counseling program.
Each of the three demonstration states had to develop a procedure for determining the amount of the allowance. Each considered using the consumer's level of need (as assessed by a professional and embodied in a care plan) or prior Medicaid claims history. The states recognized that basing the level of the allowance on claims history was problematic. Those who have been enrolled in the traditional program for only a short time (or not at all) would not have a sufficient history on which to base the allowance. Moreover, because a beneficiary's condition or situation could change, prior expenditures for covered services might be inadequate to meet current need.
Arkansas and New Jersey decided to base the amount of the allowance on a beneficiary's professional care plan, while Florida decided to base the amount of the allowance either on claims history or a care plan. Florida used the care plan for beneficiaries who had been receiving services for only a short time and for those whose circumstances had changed materially. In practice, the amount of the allowance in Florida was calculated both from claims and care plans to determine whether there had been a material change.
The demonstration 1115 waiver required that each Cash and Counseling program be budget neutral, that is, cost no more per recipient per month than the program being cashed out. Budget neutrality may require that an “adjustment factor” be applied to the amount it would have cost to provide all the services in the care plan. This is because some portion of the goods and services in care plans typically are not delivered. A discrepancy between services planned and services delivered can arise, for example, because agency workers occasionally fail to come for scheduled visits or because consumers do not need or want scheduled visits (say, during hospitalization). A discrepancy can also arise if agencies are unable to hire sufficient workers to supply all the hours of care authorized in their clients' care plans. When there is a discrepancy between goods and services planned and goods and services delivered, failure to make an adjustment in the amount of the allowance essentially ensures that cost per recipient per month for Cash and Counseling services will exceed the comparable cost for recipients of services in the cashed-out program, thereby violating the budget-neutrality requirement.
To assess the need for adjustment, each of the three demonstration states collected care plans and claims for a sample of participants in the program(s) to be cashed out. Arkansas and Florida both adopted “adjustment factors.” The Arkansas factor ranged from 70 to 91 percent, depending on the home care agency that had previously provided the consumer's personal care. The Florida adjustment factor ranged from 83 to 92 percent and varied by population group. New Jersey did not apply an adjustment factor, the historical cost of services provided to sampled program participants in New Jersey was only slightly less than the cost of services planned for them. New Jersey was aware that this difference in the use of an adjustment factor could lead to differences in the cost of its Cash and Counseling program, relative to the other two states.
While assisting consumers with managing their own supportive services is a tenet of the Cash and Counseling model, the three demonstration states had to decide what counseling and fiscal services to offer and whether any were to be mandatory. Each state faced conflicting objectives with respect to counseling services; each needed to strike a balance between empowering consumers and protecting against abuse of the benefit and exploitation of consumers. States faced no conflicting objectives with respect to fiscal services, which were designed to support consumers in directing their own services, consistent with the purchasing plan. The states also had to decide how counseling and fiscal services would be organized, and how the cost of these services would be covered so that the program would be budget neutral.
The three demonstration states decided to offer very similar fiscal services. Each decided that fiscal services would be available to assist consumers in preparing worker paychecks, appropriate tax returns and other payroll forms, and processing invoices. However, consumers might receive the entire allowance and perform these fiscal functions themselves provided they demonstrated the necessary expertise (very few chose to). Each state also required that timesheets and invoices be compared with purchasing plans and that fiscal agent cut checks only for goods and services included in the plans.
To strike a balance between empowering consumers, on the one hand, and preventing abuse of the allowance, on the other, each state allowed consumers who chose to use the fiscal agent to receive a small fraction of the allowance in cash (up to 20 percent in Florida and up to 10 percent in the other two states) for incidental expenses that were not readily invoiced, provided the purchasing plan stated the amount to be paid in cash and the types of goods and services to be purchased with cash.
Each of the three demonstration states decided that counselors would visit consumers in their homes to assist with developing the purchasing plan and that counselors would assist consumers with hiring, training, and supervising employees. Each state mandated that counselors would periodically contact consumers to ensure their well-being. While each required monthly telephone contact, the required frequency for in-person visits varied. For example, Arkansas required quarterly (later semi-annual) visits, while Florida required a visit during the second and twelfth month after enrollment and annually thereafter.
Each state also had to decide whether or not counselors would be assigned professional responsibility for reassessing need and preparing revised care plans for participants in the Cash and Counseling program. Because revisions in a care plan's hours led to a revised allowance, making counselors responsible for these tasks carried the risk that counselors might act as consumer advocates and increase care plan hours—and hence allowances—over what they otherwise would have been. On the other hand, making counselors responsible for assessment and care planning might generate efficiencies because counselors were to visit consumers to carry out their other responsibilities.
To ensure objectivity, New Jersey gave the responsibility for reassessment to members of its Medicaid staff, who were not involved with Cash and Counseling consumers day to day. In contrast, Arkansas and Florida made counselors responsible for reassessment of consumers and care planning following reassessment. Arkansas and Florida were aware that Cash and Counseling counselors might be more likely than external Medicaid staff to approve more expensive care plans, which could increase program costs in their states, relative to New Jersey.
A demonstration state could choose to separate the counseling and fiscal functions, assigning them to different host organizations, or combine them in a single host. The major argument made for separation was that these functions require different expertise, while the major argument for combining them was that doing so could enhance efficiency by reducing the need for communication and coordination.
Each state also had to decide whether to have one entity serve the entire Cash and Counseling program (for counseling, for fiscal services, or both) or have multiple entities. A single entity might be better able to take advantage of economies of scale, but might not be able to cover the entire catchment area. Moreover, multiple entities would be required if consumers were to have an opportunity to choose among providers. These decisions about the structure of counseling and fiscal services interacted with the decision on whether to rely on the existing network of agencies providing case management to also provide counseling services since reliance on the existing network meant that their structure would carry over into Cash and Counseling.
As each state considered the organization of counseling and fiscal services, it also had to decide how to structure payment and accumulate funds to cover the cost of counseling and fiscal services, including whether consumers were to be charged for these services. The basic options for structuring payment were flat rates for specific tasks, hourly rates, and capitated payment (or some combination of these methods). Charging consumers ensured that costs were borne by those using the services. On the other hand, the states were concerned that consumers might not avail themselves of needed services if they were charged directly for them.
The three demonstration states decided initially to organize counseling and fiscal services differently. Arkansas was the only state to choose combine counseling and fiscal services in a single organization. Following a competitive bidding process, it contracted with multiple human services agencies, each provided counseling and fiscal services within a region of the state. New Jersey initially separated counseling and fiscal services. It signed memoranda of agreement with numerous human services agencies across the state to provide counseling services locally. Following competitive bidding, it contracted with another human service agency to provide fiscal services state wide. (Later, New Jersey combined counseling and fiscal services in the one entity state wide.) Florida was the only state to ask its existing networks of agencies offering case management and agencies and practitioners offering support coordination to also provide counseling. Funds were already committed for the payment of case management and support coordination, and diverting these funds to pay other entities to provide counseling would be a difficult political undertaking. In addition, the state believed that this approach ensured consumer access to providers who were knowledgeable about local service systems and afforded them continuity of care. Nevertheless, Florida was aware that many in the existing networks were not supportive of Cash and Counseling and that the program's implementation might suffer. After a competitive bidding process, Florida contracted with a human services agency to provide fiscal services state wide.
The three states also adopted different approaches to payment for and funding of counseling and fiscal services. Arkansas opted for a capitated payment covering both counseling and fiscal services with the monthly payment for a given consumer falling over time on the grounds that the need for counseling assistance fell as consumers mastered their responsibilities under Cash and Counseling. Before the amount of the allowance was determined, Arkansas (which paid a flat rate for each hour of personal care in their traditional program) set aside over $4 for each hour in a consumer's care plan to cover the cost of counseling and fiscal services.
To provide an incentive for expeditious assistance with the initial purchasing plan and thereby help control cost, New Jersey paid a flat rate for counseling to assist a consumer with the development of the initial purchasing plan. The state paid an hourly rate for counseling services thereafter, with an annual cap on counseling hours. New Jersey established a fee schedule for fiscal services, with some fees paid by consumers from their allowances (e.g., for each check cut) and some paid by the state (e.g., to complete the start-up paperwork for an employee). Before the amount of the allowance was determined, New Jersey set aside 10 percent of the expected cost of a consumer's care plan to cover the cost of counseling and fiscal services.
As indicated above, Florida funded the cost of counseling with the existing funding streams for support coordination and case management. Counselors for those with development disabilities received the same monthly per capita rate as support coordinators. Counselors for elderly and physically disabled adults were paid a fixed rate for each visit to develop the initial purchasing plan (with a cap on the number of such visits per consumer) and an hourly rate thereafter (with a cap on annual cost per consumer). Florida consumers were charged fees for the use of fiscal services, payable from their allowances, with the fees capped at $25 a month per consumer.
These differences in organizing and paying for counseling and fiscal services could lead to differences in results across the three demonstration programs. For example, if combining counseling and fiscal services resulted in improvements in efficiency, consumers in Arkansas' program might be more satisfied than those in the other states. Differences in payment methodology for counseling and fiscal services could result in differences in program cost across the states.
The Cash and Counseling demonstration programs of Arkansas, Florida, and New Jersey have many common features. These commonalities arise because the three demonstration states were required to adhere to the basic tenets of the Cash and Counseling model and to federal regulations, because they shared the philosophy of consumer empowerment, and because they worked together to design their demonstrations.
However, the three states were not required to implement a standardized intervention. A standardized intervention might well have been infeasible, given the differences in state goals, Medicaid programs, and political environments. These differences influenced decisions about key program features. Goals affected decisions about eligibility for Cash and Counseling. The Medicaid program affected which programs were selected to be cashed out. Political environment affected assignment of responsibility for outreach and enrollment and for counseling; services permissible for purchase with the allowance; and payment mechanisms for counseling and fiscal services.
Variation within commonality created healthy tension. Commonalities fostered cooperation among the three states, while the ability of each to determine key program features enhanced each state's accountability for the success of its program and intensified a spirit of friendly competition among them. As the three programs were implemented, there was tension between retaining a problematic feature long enough to resolve “start-up” problems and fully assess its potential and expeditious revision to improve program performance. Each of the three states learned from their own experience and that of the other states and revised some key features of their Cash and Counseling programs. Finally, having program variations within commonality created opportunities for the evaluation to learn more about factors that can influence program results. Broadly, the three programs may be viewed as replications, but the variations in key program features are substantial and are likely to be dimensions on which future Cash and Counseling (or related) programs will differ. Many of these variations must be taken into account to understand differences in impacts of the three Cash and Counseling programs on consumers, caregivers, and costs and are discussed in other papers in this volume.
We gratefully acknowledge the financial support and guidance for the demonstration and evaluation provided The Robert Wood Johnson Foundation and the Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services. We also would like to take this opportunity to thank the staff of the demonstration programs and the people of the three states who participated in the demonstration and evaluation.
Disclaimers: This article was prepared as part of the Evaluation of the National Cash and Counseling Demonstration, which was jointly funded by The Robert Wood Johnson Foundation and the U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation (ASPE). The views expressed here are those of the authors and do not necessarily reflect those of the foundation, ASPE, the Cash and Counseling NPO, the demonstration states, or the Centers for Medicare and Medicaid Services, whose waivers made this demonstration possible.