Oklahoma Milestone Payment System

Dan O'Brien and Becky Cook

OVERVIEW

The Milestone system is a reimbursement method based on incentive payments for service outcomes. Under the Milestone system, everyone who plays a role in the service delivery process benefits. The consumers of the service benefit because they receive the best in support to obtain jobs of their choice. The service providers benefit because they receive fair compensation from the state agency for achieving results for their consumers. The state agency benefits because the payment system links funding to outcomes for its consumers.

The Milestone system pays the provider in increments when the consumer passes predefined checkpoints, or milestones, on the way to a predetermined outcome. It differs from fee-for-service funding because it reimburses the provider for the average cost of achieving the outcome, rather than for the time spent on the process. This innovative system has the potential to cut government spending, streamline service delivery, and, ultimately, to increase consumer satisfaction.

The Community Rehabilitation Services Unit of the Oklahoma Department of Rehabilitation Services developed the Milestone payment system. Through several state and federally funded programs, the Department of Rehabilitation Services provides a variety of opportunities for people with disabilities through vocational rehabilitation, education, independent living programs, library and other specialized services for the blind, and the determination of medical eligibility for Social Security disability benefits.

THE PROBLEM

No publicly funded service agency has adequate dollars to fund every worthy program or to meet the needs of every potential customer. Agencies make tough choices about the quality, scope, and extent of services for their customers. Faced with these tough decisions, our agency sought methods to improve outcomes while containing costs, stretching funds to reach more customers. One of the methods of cost containment at the agency’s disposal was our choice of provider reimbursement systems. Unfortunately, our previous reimbursement systems worked against the very goals we sought to accomplish they rewarded failure instead of success.

Generally, what is paid for gets measured, and what gets measured gets done. Since funding mechanisms tend to distort both the kinds of services delivered and how they are delivered, careful thought went into the funding design so that unintended consequences and perverse incentives were minimized.

Typical of many other public agencies, the Department of Rehabilitation Services used medical model fee-for-service payment systems, usually based on hourly rates for services. While paying for process, the Department recognized the hourly rate system did not reward successful achievement of outcomes for the customers of the service. In fact, hourly-based funding created perverse incentives that worked against the customer and agency goal of rapid placement in a job of the customer’s choice. Hourly billing bore an inverse relationship to consumer independence: the more independent the consumer, the lower the billing. The provider, induced by the payment system, was financially encouraged to provide billable hours rather than working for consumer independence.

The problem for the Department of Rehabilitation Services was to create a reimbursement system with incentives to achieve the desired outcome for the customer in the most cost effective, efficient way.

THE SOLUTION

The Milestone system is successfully solving many of the problems inherent in the previous reimbursement system. Providers are reimbursed for assisting persons with severe disabilities to gain competitive employment. The Milestone system creates financial incentives through increments of payment based on severity of the individual's disability at rates negotiated with the provider. Milestones differ from the hourly fee-for-service funding system by paying for the average cost of providing the service outcome rather than the cost of staff time spent on the process.

The Milestone payment system integrates Total Quality Management concepts with payment mechanisms. Total Quality Management (TQM) is defined as " . . . a continuous improvement process which involves participatory management and makes use of teamwork" TQM concepts are at the heart of the Reinventing Government and National Performance Review advocated by Vice President Gore. Total Quality Management defines quality as customer driven, involving employees at all levels, using a team approach to eliminate waste and continuously improve quality. The fundamental problem that Womack et. al., identified for business systems is identical for government-funded programs. They state that the ". . . problems are inherent in the system’s incentive structure and logic." The Milestone system rethinks the logic and incentive structure of the process for purchasing vocational services.

The systematic application of cooperative planning and teamwork is the cornerstone of the success of Milestones. Setting aside the traditional adversarial approach of American business and government, all parties approached decisions with the long-term health of the total system and the success of all contributors in mind. The short-term gains of individual players are subservient to the long-term goals. With Milestones, the goal was to establish a system that delivered a high quality product at a reasonable price with incentives for continuing quality improvement.

Critical to Milestones’ success is a commitment from all parties to create a payment system that meets customer needs while addressing the financial realities of both the state agency and its service providers. This process involves a commitment to collaboration and negotiation between the state agency and its providers.

State agency staff devoted significant effort in the year prior to the initial pilot and during each revision to achieving consensus with providers on the milestone definitions. This collaboration was a critical step in reducing both unanticipated consequences and resistance to change from the service providers. The definitions of individual milestones represent realistic implementation of the best practices in the field. The milestone definitions include objective quality standards derived from research and best practice data, creating practice and payment standards as well as quality benchmarks.

The key issue resolved in the design phase was the collaborative development of realistic operational definitions of the service outcomes. The outcome was then subdivided into incremental steps or "milestones" to which payment was tied. The first step involved state agency staff developing an outline of the Milestone concept. Then, a representative group of providers was convened to assist the staff in developing the detailed design. This process was repeated for each of the two primary populations served, persons with a mental illness and persons with a developmental disability. Each group has distinct needs, and therefore a different pattern of service. The planning committees were working groups with six to twelve members. Membership was held constant for the two to three-month planning phases. The committee met every two weeks with drafts of decisions mailed to members between meetings.

The initial questions to be answered were how the final outcome would be operationalized and how it would be verified. Decisions of the planning committee were made by consensus. At the outset, state staff set the parameters within which the committee had to work. The two parameters were that the Milestone system had to create financial incentives to reach the outcome, and that the final design had to represent a Win/Win/Win situation. To be stable the design had to accommodate the long-term needs of the consumer, the provider and the state agency. The next question presented to the committee was "If we were to stop you on the way to the final outcome and take a snapshot, where would we stop you and what would we be able to see?" Six logical payment points were defined where progress could be observed by the funding agency staff and a payment made.

Quality standards were articulated based on the simplest application of best practices in the field. Our approach to quality we are calling the Prop-Smart approach. The Prop-Smart approach to quality is borrowed from the field of anthropology. Cultures always have a proper and a smart way to perform any task. The proper way is the formal or "according to regulations" way to perform the task. The "proper" way typically puts more emphasis on effectiveness than efficiency. The "smart" way is the method that is most widely used and normally emphasizes ease or efficiency over effectiveness. Put simply the "proper" is what we say we do and the "smart" is what we really do. An example in job coaching is the job analysis which the job coach is "supposed" to do before the first day of work to make sure the job is a good match and the job training is structured to maximize learning. The "proper" way is taught in training and is very time consuming and very effective if done thoroughly and properly. The "smart" or easy way is what job coaches typically do which is to try the consumer on different jobs until one works and only write it up if somebody asks for it. We chose to try to marry these two approaches to achieve a consensus approach which while not perfect was "good enough", i.e., "satisficing". The "Prop-Smart" method uses consensus among the stakeholders to develop a workable combination of the "proper" and the "smart". This middle ground is used to build accountability into the milestone process to assure quality. Micro-audits are conducted by the Vocational Rehabilitation Counselor at the end of each Milestone to assure compliance with the consensus quality indicators.

Payment percentages were negotiated for each of the Milestones. Negotiations on payment percentages reached a compromise between difficulty of implementation for the provider and the state agency’s requirement that payments be weighted toward the ultimate outcome. If the providers had their wish, all the money would be paid up front, while the state agency would have wanted to pay 100% at the end. Neither solution would have worked in the long term.

The resulting Milestone payment structure was distributed to all current providers for comment. The revised structure was used to fund a pilot to test the design. The committee was called together to review the results of the pilot and redesign at the end of the first pilot. Implementation was then begun on a voluntary basis. Typically, the more entrepreneurial vendors converted to Milestone funding in the first round. Before mandatory conversion, the committee again reviewed the Milestone structure and made adjustments to correct problems. Subsequently all providers were required to convert to Milestone funding.

An example of outcomes defined for a Milestones contract is a six-milestone payment structure: (1) Determination of Consumer Needs and Career Planning - 10% of bid; (2) Job Placement - 10% of bid; (3) four weeks job training - 15% of bid; (4) 10 Week Job Retention - 20% of bid; (5) Job Stabilization - 20% of bid; and (6) Consumer Rehabilitated (Case Closure, Stabilization+90 days) - 25% of bid.

Each milestone definition includes quality outcome indicators to be accomplished before payment. For example, the quality indicators for the last two milestones, five and six, include verification that the consumer and employer are satisfied with the job placement, training and support provided. This creates a quality check by the most qualified evaluators, the recipient of the service and his or her supervisor. This information is submitted to the agency staff overseeing the service before payment. The system is customer-driven because the service recipient or his or her guardian must be satisfied before the final two payments, totaling 50 percent of the bid, can be paid. Increasing the percentage of the bid paid as the consumer moves toward the final outcome also reinforces the final outcome, a stable job.

FORCE FIELD ANALYSIS

The force field graphs illustrate key differences in the forces at work in employment of persons with severe disabilities, contrasting the "before" and "after" implementation of Milestones.

GOAL

Consumer Choice No Work History

GOAL

Staff members who value rapid placement of consumers in jobs of their choice create push in the right direction. Milestones strengthen this by tying provider income, staff evaluation and sometimes bonuses to these values. The opposing forces of Staff Values versus Staff Convenience present implications for a public agency’s choice whether to deliver services directly by using state employees or to contract out services to providers. The force field analysis helps demonstrate the nature of the problem and illustrates why contracting out services works better. Outside contracting creates valuable extra push in the direction of the consumer’s choice with fewer constraints on the manipulation of force vectors. The equation is shifted in a positive direction, by reversing the direction of the financial incentive vectors. This creates greatest difference by moving the provider income force vector from resisting to assisting the consumer. Under the hourly payment system, the provider generated more income by dragging the process out for the statutory maximum of 18 months. Under Milestones, the provider’s cash flow and income improve if the consumer gets a job of their choice as rapidly as possible.

The resistance of some state agency staff to getting out of the office and getting their hands dirty is difficult to counter in a state merit system. For merit system employees, the consequence of not doing the work effectively may be a transfer to an easier job. Needless to say, this makes it difficult to generate the necessary push toward the goal. A partial solution in state systems may be performance based budgeting. The agency or work unit’s future budget and staffing could be based on past performance. This does not provide the immediate pressure for productivity that Milestones do through external contracting. It may, however, be the only option in state systems where contracting out is difficult.

Consumers and their families create push toward the goal to the extent they are empowered. Under Milestones, the consumer and family are directly involved in the planning process and in asserting their satisfaction with the final outcome, creating greater positive push toward the goal.

THE TARGET POPULATION

The employment programs paid through the Milestone system target individuals with any type of severe disability who have been determined "eligible" for services according to federal definitions under the Rehabilitation Act of 1973, as amended. Individuals selected for Milestone employment programs need assistance in obtaining, learning, and/or keeping a job. The majority of these employment programs target individuals who have either never worked at all or never worked successfully. Although the majority of individuals currently served are those with primary disabilities of mental retardation or mental illness, many other types of severe disabilities are represented in the target population, including traumatic brain injury, cerebral palsy, dual sensory impairment, and other severe physical disabilities.

Scattered across the state among thirty providers, approximately one thousand individuals are served annually with a total budget of $4.8 million. The total number of potential clients for this service is difficult to estimate. Although the Rehabilitation Services Division of the Department of Rehabilitation Services serves a total of approximately thirty thousand consumers each year, not all of these individuals would be appropriate for Milestone-based employment services. Those who are served in Milestone programs range from individuals with the most severe disabilities for whom direct placement in a job, with supports, is the desired service to individuals with disabilities who can achieve competitive employment with lower levels of support.

A COMPARISON EXAMPLE:

The Players:

Vocational Rehabilitation Counselor: The VR Counselor is a rehabilitation professional, usually of Masters degree level, who is an employee of the state rehabilitation program. The counselor, who is knowledgeable about consumers with disabilities and their vocational needs, is responsible for assisting the consumer to determine and achieve a suitable vocational objective. The counselor works with the consumer and family, as appropriate, to devise a plan which will lead toward achievement of the vocational objective. The counselor has funds to purchase services for the consumers to assist them to achieve their objectives. The counselor is responsible for issuing authorizations to vendors for the milestone payments and for assuring that services are delivered for the consumer according to the quality indicators by the vendor before payment is made.

Consumer: A person with a disability who has been determined eligible for services by the VR Counselor according to federal guidelines. The consumer and family have the responsibility to set the job goal and must be satisfied with the service and support they have received before payment is made on the last two milestones.

Provider: A private or public vendor of services who has a contract with the Department of Rehabilitation Services to deliver specific services leading to employment of the consumer in a competitive job. For individuals with severe disabilities, the assessment, job placement, and training services are delivered by a "job coach," who goes with the individual to the job site until the work tasks are learned to the employers’ satisfaction. The Rehabilitation Counselor makes authorizations against their contract for specific milestones for individual clients.

John is a 24-year-old man with Down syndrome with an IQ of 42 and a history of mild heart problems. He graduated from high school at age 21, having completed a special education program for individuals with mental retardation. John, who receives SSI, now lives in a group home with other young men with disabilities. John’s parents are still actively involved with his life, and he continues to spend time with them. He is not able to speak intelligibly, except to those who know him well, and he does not read or drive.

Since graduating from high school, John has worked at the Sunny Day Sheltered Workshop, performing contract work for 37 to 60 cents per hour. The state pays Sunny Day a $5,000 slot per year to keep John in the workshop. His parents, not satisfied with this type of employment for their son, have requested a job in the community through Sunny Day’s new Supported Employment Program.

If John is to work in the community, he will need specialized support to get a job which fits his interests and abilities, and he will need assistance in learning the tasks of his job once he gets it. Since he does not communicate efficiently by speech, John will need some assistance in learning to communicate with his work supervisor and his fellow employees. John will also need a job within walking distance of his home.

John likes people, gets along with the other group home residents, and is enthusiastic about performing tasks he has learned, such as making his bed at home. John is noted for his love for pizza, one of the words he can communicate quite clearly. He tends to get excited and distracted if there is a lot of activity going on around him, and, as a result sometimes he forgets the sequence of the tasks he is to perform. It takes many repetitions of a task before John learns it, but once he learns it, he retains it. If John is not able to perform all of the tasks of a particular job, it may be necessary to negotiate a specialized position with an employer willing to allow John to perform only certain tasks of a job description.

John’s Service Under Hourly Funding:

Because there were many others on the waiting list for supported employment at Sunny Day, John had to wait eight months from the time of his parents’ request to begin services to lead toward community employment.

Once John rose to the top of the waiting list, he was referred to the Vocational Rehabilitation (VR) Counselor who determined John to be eligible for the service. The VR Counselor authorized for 200 hours of Supported Employment Assessment at $23/hr, and services were initiated. During the next twelve weeks, John was assessed extensively on several jobs. Reports of the assessments and mountains of papers documenting staff time were presented to the VR Counselor. During the assessments, John demonstrated ability and interest in working at a restaurant, preferably during the day shift of one of those near the group home. The cost of assessment was $4,600.

The VR Counselor authorized another 200 hours at $23/hr for job placement services. John was placed on the first job that became available on a cleaning crew on a night shift at a hospital. Unfortunately, the results of the earlier work assessments were not matched to the job selected for John, nor did they match his job choice. John was not able to walk to this job, and night hours were not his preference. John had trouble staying awake, and the group home staff were unreliable about getting him to work on time. As a consequence, John did not perform well and quit the job after two weeks. Total billing for the unsuccessful work attempt was $2024, which included 80 hours of training on the job and eight hours transportation time.

The provider continued billing hours of service on the original 200-hour authorization to develop a second job as a bus boy at a restaurant serving a large buffet. The restaurant was near John’s home, but the demanding pace of the restaurant and the profusion of customers asking things from him was confusing for John. John refused to go back after the second week. In this second unsuccessful work attempt, $2,576, or the remainder of the $4,600 authorized, was billed.

After two job failures, John was labeled "non-compliant" and was required by the provider to complete additional evaluations before being placed again. Feeling that he could not lose the $9,200 investment already made in attempts to place John successfully, the VR Counselor authorized another 200 hours for a third job placement.

After another six months’ wait, John was offered an opportunity to work at a pizza restaurant near his home. John’s tasks were to fold boxes for the carryouts, to clean the bathrooms and floors, and to take out trash and pick up trash in the parking lot. Although John loved this job, he required extensive training to learn the several tasks involved to the employer’s satisfaction. In addition to the 200 hours authorized, the provider requested another 46 hours.

After four jobs and a total billing of $14,858, the provider had found a job that pleased John and suited his skills and abilities. He settled into a successful job after 18 months in the program.

John’s Service Under Milestones:

Under milestones, John’s story is quite different. From the time he applies, John waits four months to begin services. The VR Counselor issues an authorization for Milestone One, Assessment/Needs Determination, at a fixed rate of $870. Once he starts the program, John progresses quickly through this phase in which his strengths, interests and support needs are identified. The work performed by the provider is focused on assessments that meet John’s expressed needs and interests.

Through the job coach’s discussion with John, his family, his counselor, and the group home staff, it is clear that John wants a day shift within walking distance of his home. Since John has never worked in a community job, the provider wants to make sure he gets a chance to try out more than one option. John’s favorite pizza restaurant nearby might be a potential employer. The job coach learns from the group home staff that John smiles and gives the restaurant manager a "high five" at his weekly visits to the restaurant. Several of the other restaurant staff know John by name. Based on these discussions, the job coach arranges three assessments, one with the pizza restaurant. The on-site assessments indicate John likes the work at the pizza restaurant best and evidences the ability to learn the tasks, although there is some concern over the amount of training time that will be required because there are several multiple-step tasks for John to learn.

The results are discussed with John, his family, and his VR Counselor. The provider presents the required assessment documents to the VR Counselor along with the bill for $870. The Counselor reviews the proposed support plan and documentation, pays the bill, and determines the severity of John’s disabilities evidenced during assessment meets the criteria for a higher level of payment for the remaining five Milestones.

The job is analyzed prior to John being offered the job, and the demands are compared with John’s assessments. After some negotiation, the pizza restaurant manager has agreed to the tasks John is to learn: to fold boxes for the carryouts, to clean the bathrooms and floors, and to take out trash and pick up trash in the parking lot. John is offered the job, and he and his parents accept. The VR Counselor issues authorization for Milestone Two, Placement, at $1050 (10% of the vendor’s bid). When John starts work, his job coach has devised a training plan with clearly defined objectives describing the job demands and techniques for learning/teaching each task. For the agency to be paid, the job must match the job goal identified in his vocational plan.

The VR Counselor reviews the verifying documentation, makes authorization for the remaining four Milestones, and the provider is paid for the placement milestone after his third day of work. The job coach begins to train John on the work tasks. After John has worked successfully for four weeks the job, the provider is paid $1575 (15% of bid), and again after ten successful weeks, the provider receives another payment of $2100 (20% of bid). The provider receives the remaining payments when John is stable on the job, $2100 (20% of bid). This is determined by John’s successful completion of 17 weeks on the job, satisfactory work ratings by the employer, and documentation that John and his family are satisfied with the job. The sixth and final payment is made three months after John was determined to be stable on the job and John and the employer are satisfied. The VR Counselor reviews the documentation and pays the final payment of $2625 (25% of bid) for a total cost of $10,500. John is successful, working independently with support from his work supervisor, and is satisfied after eight months in the program. John works 30 hours per week and has received one raise since he started work. He earns $5.50 per hour. John loves his paycheck and wants to work more hours. His SSI benefits have been reduced, but as a result of his earnings, John has more spending money, and he is a tax-paying citizen.

Chart 1. Comparison of Successful Employment Outcomes for John

*These figures represent the average cost in the final year of fee-for-service and the first year of Milestones

Milestones create a customer driven process. Consumers make rapid progress with higher quality standards and less wasted effort and less cost. Assessments are more focused on consumer needs and interests. Job placements are faster with better job matches as opposed to those jobs easiest to get. Since the agency is paid only once for each Milestone, more care is exercised. The payment system encourages the agency to find the right job the first time and to get John stabilized on his job quickly. The hourly system reinforced poor job matches and created financial incentives for lengthy training. As a result of the change to Milestones, John’s independence is encouraged rather than discouraged.

ONE PROVIDER’S STORY:

Chart 2: Comparing Fee-for-Service to Milestones

THE BIDDING PROCESS

The bidding process is critical to the success of the Milestone system. By allowing providers to bid based on their projected costs, the system assures that pricing will be adequate and competitive. This assures that the state agency can purchase the service, and that it will not pay too much.

A structured process guides providers in making a milestone bid. Providers develop a program budget, then estimate the numbers of individuals they can assist with the projected level of staffing. During the transition, these estimates are based on past performance under hourly funding. The final step in developing a bid is to project, based on historical data, the number of persons with disabilities who will complete each step or milestone in the process. This allows the provider to include in average cost calculations the projected number of "drop outs" at each milestone.

The bid formula multiplies the estimated number completing each milestone by the weighting factor for that milestone. The weighting factor is the number of units of 5% contained in that milestone. For example, a milestone which is paid at 10% of the bid has a weight of two (2 units of 5%). If eight consumers are projected to complete that milestone, the formula is 8 X 2 = 16. Each milestone represents a number of units of 5% that will be accomplished. The total units of 5% are added up, divided into the total budget for the program, and multiplied by 20 to get back to 100% (5% X 20 = 100%). The resulting figure is the bid per consumer. Chart 2 demonstrates the formula.

CHART 2: FORMULA FOR MILESTONES BID

BID COMPUTATION

1) $_______________ divided by _______________ = ____________

Total Program budget Total Units of 5% Cost/Unit of 5%

2) $____________ X 20 = _______________ per Consumer

Cost/Unit of 5% Bid Amount

This bid per consumer is a realistic projection of the cost of providing the predefined quality outcomes and includes the costs of dropouts. In other words, if the provider completes the milestones as projected, they will have adequate income to support the continuation of the service. This demands that service providers develop a reasonable business plan. Those who do as well as or better than projected are rewarded, and providers who fall below agreed upon projections suffer revenue shortfalls that must be made up by increasing outcomes or from reserve funds. Bids are submitted to the funding agency along with supporting documentation. Bids are evaluated equally on quality factors and costs, as compared to statewide average cost figures. During the mandatory conversion bids which were significantly above or below the average range were rejected. After the first 18 months under milestones projects were required to re-bid in order for the state to recoup the efficiencies which had developed. Bids were required to be below the average bid from the previous bidding process. This resulted in a further reduction in costs of about 9%. Bids in many rural areas of the state, where multiple providers are not available, are not truly competitive. These bids are evaluated on quality factors and compared against average statewide rates.

CORRECTING PROBLEMS

This system naturally rewards those who do what they contract to do and financially disciplines the providers who don’t meet contract projections. This replaces the cumbersome formal discipline system that could drag out for years. The Milestone system uses natural or automatic consequences that are immediate. This results in faster correction of failing performance without cumbersome legalistic, bureaucratic sanctions.

We found the providers grouped into three categories the "honest cheaters", the "try harders" and the competent. Each group responded differently to the incentives in each system. The "Honest Cheaters" were providers whose prime directive seemed to be "milk the system", they wanted to maximize income rather than provide the service. The milestone effect was most notable with this group. All of them lost interest after the transition to milestones. A before and after example for this group was one provider who earned $108,000 for one outcome in an entire contract year. In the first year of Milestones they earned only $6,500 for that one outcome and decided to give up the contract. The second group was the "try harders", they "try" but don’t succeed, they sincere but inept or incompetent. A typical before and after scenario saw the provider’s income go from $25,701 per outcome to $11,715 per outcome under Milestones. They were forced to become better managers and more outcome focused. The third group was the competent, they tend to do a good job regardless of the perverse incentives. The typical before and after picture shows a reduction from $10,033 to $9,640 per outcome under milestones. Overall, the biggest savings were in the transition year and the most significant improvement was in customer serviceas measured by speed of placement and quality of job match.

One of the weaknesses of an outcome-based payment system can be a tendency to create disincentives to serve the most difficult customer. Two anti-creaming elements are included in the Milestone system one to avoid punishing vendors who took a few risks and the other to reward those who sought out the more risky customers, the first is achieved by allowing for partial payments for dropouts and the latter through a higher rate for more "challenging" cases. The drop-out provision allows bidders to include the cost of dropouts in their bid calculations (see example of bid sheet Chart 2) , this avoids placing vendors in the impossible position of trying to predict with 100% accuracy who will succeed. Other outcome payment systems have neglected this element and created a "no-win" situation where the provider is guaranteed to lose money if they have even one drop-out. The second method is in line with the commitment to managing incentives rather than regulate behavior, we looked for a way to create financial incentives to serve the more difficult customer. The milestone system requires multiple bids based on the anticipated level of need/difficulty. The higher level of difficulty, "highly challenged", is reimbursed at a higher rate. The criteria for the "highly challenged" payment were developed with the prospective bidders by asking the question "what makes the process take longer, how could we objectively identify potentially time consuming and therefore more expensive cases?". A list was developed prior to the bidding process which was the consensus of the stakeholders on observable criteria for "highly challenged". This creates an incentive to find and serve the more difficult customer in order to increase agency income. In order to test the efficacy of these anti-creaming strategies, the innovators looked at before and after statistics for the most vulnerable and time consuming segment of the supported employment customers. The group most likely to be screened out, or "creamed" are the individuals with IQ’s below 40. Program statistics from FY91 show that before Milestones conversion 17% of the customers served had IQ’s below 40, the after statistics (FY98) show the same 17% with IQ’s below 40. The innovators concluded that these incentive based methods worked as expected.

OUTCOMES

The three most important measures of success are customer satisfaction (both employers and consumers of the services), reduced need for regulation and oversight by the state agency, and increased achievement of the core outcome, successful competitive employment.

The primary achievement of the Milestone system is creation of a self-regulating, customer focused, outcome driven service delivery system that resembles a business transaction. This involves several significant aspects. First, it shifts payment from process to outcomes. Second, it creates a consensus on outcome definitions between customer, agency and providers, incorporating the latest research and best practices. Finally, it creates a system that is self-correcting, since incentives are tied to quality results rather than process.

Opportunities for customer satisfaction checks come into play at several points in the Milestone process. For instance, if a provider places an individual on a job which he or she does not like or for some other reason does not want, the provider is not approved for payment for that milestone. This motivates the provider to please two major customers, the consumer of the service and his or her Vocational Rehabilitation Counselor, so that payment may be received. Since job satisfaction is a major component of success and the provider cannot be paid without achieving it, a major shift of emphasis toward quality as defined by the customer has resulted.

Formal customer satisfaction surveys are required for payment of the last two milestones. The state agency’s Counselors are instructed to withhold payment for these milestones (as much as 45% of the bid) unless there is evidence the consumer of services and employer are satisfied with the service. This has resulted in providers returning to the job site to correct problems, which they could have ignored under previous payment structures.

Time on waiting lists is reduced by an average of 53% because consumers are now more efficiently and effectively served. Consumers who previously had been perpetually "assessed" were now moved through the milestone process to employment and independence. The number of individuals who never got a job dropped by 25%. Under Milestones there was a 102% increase in the number of consumers succeeding in work and some provider’s report that 95% of consumers are succeeding at their first job. Providers report as much as a 100% increase in productivity in terms of consumers placed and stabilized in competitive employment.

Provider efficiency is improved by reducing the paperwork load required of the direct service provider. When hourly rates are the basis for payment, the process of documenting increments and categories of services provided must be meticulously tracked. Now, time formerly devoted to detailed documentation is redirected toward accomplishing the outcome for the consumer. The only information that must be documented by the provider and approved by the counselor is the proof that the outcome was achieved.

Effectiveness is improved by linking payment for the final milestone to the achievement of outcomes that have real meaning for the customer. For example, goal planning is reinforced by the requirement that the job achieved matches the customer's career goal. One of the best measures of effectiveness is whether the employer and the consumer of services are satisfied with the job placement and service. Before the last two milestone payments can be made, the employer and consumer of services must be satisfied.

COST BENEFIT ANALYSIS AND RESULTS

Milestones were implemented in a staged process of pilot programs. The first stage of the pilot program involved testing the concept itself by establishing new programs using Milestone funding. Two pilot programs were funded in October 1992. The second stage of the pilot involved conversion of the twenty-three existing Supported Employment programs to milestones from hourly contracts, which was completed in January 1997. The final stage will convert the remaining seventeen contracts to Milestones before the end of 1997.

The average number of months consumers spent on waiting lists was reduced from 8.14 months under hourly to 3.85 months under Milestones. This represents a reduction of customer months of waiting by 53% among the providers who have a waiting list. Consumers who previously had been perpetually "assessed" now move quickly through the milestone process to employment and independence. The average weeks spent in assessment before job placement dropped 18% from 12.1 to 9.9.

Efficiency is improved by reducing the paperwork load required of the direct service provider. Under an hourly rate system, every minute and category of service was meticulously tracked. Providers previously completed as many as 18,000 data entry screens each month to insure payment for every increment of service time, under Milestones they now complete less than 100 data entry screens each month. Now, time formerly devoted to detailed documentation is redirected toward accomplishing the outcome for the consumer, as much as 10% of total staff time. The only information documented by the provider and approved by the Counselor is proof that the outcome was achieved. The change to Milestones has resulted in a 33% overall reduction in the work hours required by record keeping.

One of the best features of the Milestones program is that providers generally like it better and feel it benefits the consumers they serve. Six months after conversion to milestones, a survey of direct service staff indicated that 81% believe the quality of services under Milestones is better or much better than under hourly.

The cost to the state agency is fixed in advance, making it possible to make accurate predictions of future expenditures. The average cost of the service outcome, competitive job placement was reduced by 51% from FY91 before the first pilot to FY98; bids dropped another 9% for FY99 and are expected to slowly drop as efficiency improves. Actual program outcomes among three early programs to convert, described in Chart 4, bear this out. The results of these providers for one year of operation indicate a 35% reduction in cost and over 100% increase in successful closures. While we do not expect the results to be so dramatic for the entire group of programs converting, the trend, as these results indicate, is expected to be positive overall.

Chart 4: Cost Comparisons, Before and After Milestones for 3 typical Providers

*Figures for FY 1995

Provider 1, the earliest to convert to Milestones, has two years of data available for comparison. The chart describes their first year of conversion from fee for service to Milestones. In their second year under Milestones, the program cost per success was $9,679 for 16 successful closures, indicating an ongoing trend toward lower cost and increased outcomes when compared to fee for service.

OBSTACLES

Obstacles include changing attitudes from process thinking to outcome thinking. Systems change is often resisted by those comfortable with the status quo. Others will fear the change because of the performance demand it places on systems that previously were only required to "try hard." It requires government agencies and providers to think like a business. Conversion from fee-for-service to Milestones can best be achieved through collaboration of the staff and providers effected by the change. This requirement for collaboration based on the principle of reciprocal obligation, may be difficult for state agency staff who take a more adversarial approach to service providers.

Reciprocal obligation is the key to a Win/Win/Win scenario. This change of attitude started with the funding agency, which consistently modeled integrity and transparency in the collaborative planning and implementation process. The providers’ initial mistrust was overcome through consistent adherence to the principle of reciprocal obligation, i.e., the state owed the provider adequate income in return for meeting the required outcomes. This is the only stable long-term basis for a business relationship that has the potential to improve productivity and quality simultaneously.

The Oklahoma Department of Rehabilitation Services' staff provided extensive training on the conversion process prior to Milestone implementation. Training included practical "how to" processes (for example, how to claim for Milestones, how to use the new computer-based client data system, case management, predicting individual milestone achievements, etc.). It also included how the change would affect the philosophical orientation of the agency. Despite the many hours of training provided the providers prior to converting to Milestones, there was still some last minute distress among providers at the time the change occurred, especially over the bid process. State agencies or others implementing a Milestone system should anticipate the problems that occur with any major conversion process, especially one that shifts philosophy as well as funds. Training and technical assistance in advance of the conversion is a requirement for success.

REPLICABILITY

The Milestone model of funding is adaptable to any public service with a definable outcome where service provision is contracted out to providers. The State-Federal Vocational Rehabilitation system, of which The Oklahoma Department of Rehabilitation Services is a part, has the advantage of having a clear outcome: competitive employment of people with disabilities. This system could certainly be used by the 80 state rehabilitation agencies (blind and general rehabilitation agencies). It has been recommended to the Texas Rehabilitation Commission by the Texas Comptroller as part of the Texas Performance Review, the Alabama Rehabilitation agency is replicating the Milestone payment system as well as nine other states and the Commonwealth of Australia. The U.S. House of Representatives passed, by a vote of 410-1, the "Ticket to Work and Self-Sufficiency Act of 1998" which includes the milestone payment system for a national vocational voucher system.

Many other federal and state programs desire the same employment outcomes for their customers. For instance, state agencies implementing Welfare-to-Work programs would be a key audience for this type of payment system. The Milestone system could easily be applied to managed care and charter schools where outcomes/results have been clearly articulated. The model could be the foundation of a voucher payment system. The milestone steps could each be vouchered separately, allowing the customer to move their voucher to a new provider at the end of any step. A voucher system built on a milestone foundation would have the advantage of a clearly defined product with integrated quality standards. A market based on the Milestone system would meet all three criteria for a competitive market, a clearly defined product, a well-informed consumer and easy entry to the market for suppliers. The development of a market for these services would empower the customer and increase service quality through competition fostered by customer choice.

The state staff person filling the gatekeeper role would issue vouchers. They would determine eligibility and level of need and issue a voucher for each milestone. The vouchers would need to have staged expiration dates or be issued in three sets. This would avoid obligating state funds long-term for dropouts. Vouchers could be used at any certified vendor. The state agency would establish simple standards for certification primarily designed to prevent unscrupulous vendors from abusing the consumers and payment system. If a consumer chose to change providers in the middle of the Milestone process, the new provider would have to accept the remaining vouchers as full payment or justify the need for repeating a milestone(s). As with the existing Milestone system, the state agency staff person would determine whether there was reasonable cause to issue a second voucher. If the consumer stayed with the same provider for the entire process, each Milestone would be paid once.

In Reinventing Government, Osborne and Gaebler list issues, which are relevant to the replication process. They are: watch out for creaming, anticipate powerful resistance, involve providers and employees in developing correct measures, regularly review and modify measures, and don’t use too few or too many measures. Creaming, serving only the easiest to serve, was overcome in Milestones by the multiple payment levels based on disability and the allowance for dropouts in the bid. Resistance was reduced by involving providers in the planning process and through successfully staging implementation. These steps allowed providers to have some control over the outcome, and to see others prosper under the new system. The terms of the milestones were reviewed with providers and revised after each iteration of the model, seven times in five years. The number of Milestones was balanced at six to seven, which use a small number of observable outcomes: job retention, wages, employer satisfaction, and consumer satisfaction.

THE INNOVATORS

The Milestone system was a 1997 finalist in the Harvard Kennedy School of Government/Ford Foundation Innovations in American Government Awards. It is also the subject of a Kennedy School of Government case study. It was a winner of the Pioneer Institutes Better Government Competition and a finalist in the Council of State Governments Innovations Awards both for 1997

.

The program was developed by the Department of Rehabilitation Services (DRS) Community Rehabilitation staff, Dan O’Brien (405) 951-3479 e-mail: deobrien@aol.com, and Becky Cook (405) 951-3474. Both can be reached by mail at: 3535 NW 58th St. Suite 500, Oklahoma City, OK, 73112. Detailed information on the methodology can be obtained at the DRS Milestones website at: http://www.onenet.net/~home/milestone.

Mr. O'Brien has a Masters Degree in Public Administration and extensive management experience as a provider of services and a small business owner. Ms. Cook has a Masters Degree in Psychology and extensive experience managing various provider-based services and related payment systems for the state agency. Their mutual desire to build a fair, streamlined reimbursement system that rewarded providers for achieving desired outcomes for consumers led to development of the milestone system.