Number 2 in Series, "Governing-for-Results and Accountability"
The views expressed are those of the author(s) and do not necessarily reflect those of the Urban Institute, its board, its sponsors, or other authors in the series.
This is the second of several reports resulting from the Urban Institute's Governing-for-Results and Accountability Project, an examination of performance management that looks at strategic planning, program budgeting, and citizen involvement in the states. To date, the Urban Institute team has conducted extensive discussions in North Carolina, Minnesota, and Florida and has interviewed a number of agency directors from other states. More states will be visited throughout 1999 and 2000 to broaden the base of information about this emerging, important topic.
The term governing-for-results refers to the system of spending and measuring results of spending whereby a government, both the executive and legislative branches, focuses decisions and activities on the end benefits, as well as costs, of government activities and actions. This means maximizing the quality of services and programs to the government's citizens and developing proper ways to measure end outcomes of those services and programs, especially in the opinion of customers.
Performance contracting, long used in such government services as highway maintenance and solid waste management, is becoming increasingly attractive to state human services agencies. Often frustrated by declining performance, rising costs, or both, they want to pay for results, not activities. Six agencies that have adopted performance contracting are described here: Oklahoma's Community Rehabilitation Services Unit, North Carolina's Division of Social Services, Illinois's Department of Children and Family Services, Florida's Department of Children and Families, Minnesota's Refugee Services Section, and Maine's Department of Human Services.
These agencies arrived at performance contracting by different routes. Maine and Florida were directed by the state legislature to begin the process; implementation in Oklahoma, Illinois, and North Carolina began from within the agency; and Minnesota was influenced by federal performance measurement requirements.
Performance contracting models vary by amount and timing of payments, the extent to which incentives and disincentives are offered service providers, the frequency of providers' reports on performance, and the extent to which providers are involved in developing performance indicators. The six examples described here fall into two categories:
- Performance contracting that ties payments to outcomes. Oklahoma, North Carolina, and Illinois have such contracts.
- Performance contracting that contains performance specifications and outcome targets. Florida, Minnesota, and Maine have such contracts.
Performance Contracting That Ties Payments to Outcomes
These contracts make payments to providers at specific milestones or upon documentation of the final outcome or end result.
The Community Rehabilitation Services Unit of the Oklahoma Department of Rehabilitation Services (DRS) began its performance contracting system, the Milestone Payment System, in 1992. The state now pays contractors at predefined milestones, when increments of predefined outcomes have been completed. Contractors submit bids for the average cost per client of the following accomplishments, the last four of which are client outcomes:
- Determination of consumer needs (10 percent of bid);
- Vocational preparation completion (10 percent of bid);
- Job placement (10 percent of bid);
- Four weeks' job retention (20 percent of bid);
- Job stabilization (10 to 17 weeks of employment with fewer than two support contacts per month, depending on the type of disability) (20 percent of bid); and
- Consumer rehabilitated (stabilization plus 90 days) (30 percent of bid).
Providers are paid a percentage of the total average cost of closure per client at each of the six milestones. They bill DRS when they achieve individual clients' milestones. Incentives for creaming are reduced by making higher payments to providers for serving difficult-to-place clients. Clients and their families are involved in setting objectives for employment during the assessment phase. Customer and employer satisfaction after the job stabilization milestone is required for the final payment.
The program reports that providers' costs per placement declined 51 percent between 1992, when performance contracting began, and January 1997, when the transition was complete. The last group of contractors to switch to performance contracting showed an additional 12 percent decline.
The average number of months clients spent on waiting lists was decreased by 53 percent, from 8.14 months under the hourly system to 3.85 months under Milestones, and the average number of weeks spent in assessment declined from 12.1 to 9.9, a drop of 18 percent. The number of individuals who never got a job fell by 25 percent.
The system faced some challenges, including developing operational definitions of outcomes for making payments and getting contractor staff to adopt a philosophy of working for client outcomes.
For more information, see www.onenet.net/~home/ milestone.
The Division of Social Services in North Carolina's Department of Health and Human Services initiated statewide performance contracts with private adoption agencies in 1995. Public agencies must work within county boundaries to recruit and prepare adoptive parents, so private agencies are necessary to work with adoptive parents throughout the state. In 1998, the state had three private contractors; in 1999, it had four.
The contracts are funded with a combination of federal and state dollars. Once funding is exhausted, usually early in the fiscal year, the private agencies can participate in the state's Special Children Adoption Fund. This fund was created in 1997 to help both private and public agencies increase adoptions of children with special needs.
Private adoption agencies bill the state on a monthly basis when they have achieved three specific outcomes:
- Placement for the purpose of adoption (60 percent of average placement cost);
- Decree of adoption (20 percent of average placement cost); and
- One year of intact placement after the decree of adoption (20 percent of average placement cost).
The state worked with the private agencies to design the payment schedule.
In fiscal year 1993-94, there were 261 adoptions in North Carolina. In FY 1995-96, when the state entered into performance contracts with three private adoption services, the number increased to 364. In 1996-97 there were 631 adoptions and in 1997-98, 603 adoptions. The special needs adoption coordinator attributes the decline in the 1997-98 number to three factors: (1) children who were easier to place were adopted earlier, in 1996-97; (2) there is a backlog of cases awaiting termination of parental rights; and (3) the children now waiting for placement have special needs.
For more information, contact Esther High, Coordinator, Special Needs Adoption, at (919) 733-4622.
|Positive Attributes of Performance Contracting in Six State Agencies
Involves Contractors in Developing Performance Standards and Outcome Targets
|Allows a Hold-Harness Clause in First-Year Contracts
|Monitors Performance with Regular Reporting
|Provides Feedback on Performance and/or Requires Corrective Action Plans
|Can Show Performance Results Based on Data Reported by Providers
|Provides Comparative Performance Data to Contractors
The Department of Children and Family Services (DCFS) wanted to improve its performance in finding permanent homes for children in foster care. In 1997, it began performance contracting in its Relative Home Care caseload that tied incentives to promoting increased permanency rates and reducing instability rates for children. The state worked closely with providers in implementing the system. At the same time, the department improved the adoption services it provided to children, such as raising the amount available for reunification services from $600 per child to $8,000 per child.
The department increased the number of cases each caseworker (private or public) must serve per year from 25 to 33, but it held payment constant at 25 cases per year. This policy requires caseworkers to successfully place at least eight children annually through adoption, successor guardianship, or reunification with their families. If caseworkers do not place at least eight children successfully, they may find themselves working on 33 cases while being paid for 25. If the number of children per caseworker exceeds 30, the state may cancel a contract. If caseloads fall below 25 per caseworker, the agency continues to receive payments for the full 25 cases. Children coming into the system are not placed with agencies that are not performing.
DCFS regularly provides the public and private agencies under performance contracts with comparative data that rank each agency on performance. Agencies are also given the opportunity to first review, then reconcile their numbers if they disagree with the state's account.
In the first fiscal year, the number of placements statewide increased from 2,411 to 5,570; placements per caseload rose from about three to six. In the second fiscal year, the number of placements reached 9,503, or about seven placements per caseload. The majority of these placements have come from the performance-based contracts.
In the first two years of performance contracting, the Relative Home Care caseload declined by 41 percent. Based on this initial success, DCFS decided to expand performance contracting to the Traditional Foster Care caseload in 1998. In the first year, Traditional Foster Care also doubled the average number of placements per caseload from about three to six.
The disruption rate after placement is no higher now than it was before performance contracting. Due to the large number of placements, more children are requiring postadoption services. In response, the state is expanding such services.
For more information, contact Mike Shaver, Project Director, at (312) 814-4650.
Performance Contracting That Contains Performance Specifications and Outcome Targets
These contracts do not contain specific incentives or disincentives for performance. Rather, they are meant to encourage and improve performance by including performance specifications and setting targets. As a result, these contracts usually involve a collaborative relationship between agencies and their service providers.
The Refugee Services Section of the Minnesota Department of Human Services (DHS) began performance contracting with its "mutual assistance associations" in 1990. The performance contracts deal specifically with job placement. Funding and performance measurement requirements come from the U.S. Department of Health and Human Services (DHHS).
Each contractor receives a grant based on the number of clients it proposes to serve and the cost of placement per client. Contractors then submit a two-year work plan that identifies performance indicators with targets for all activities leading to job placement. They set quarterly targets for job placements, clients' average hourly wages, termination of clients' cash assistance due to earnings, job retention, and number of full- time placements receiving health benefits. Contractors regularly report information on job placements, retention, and related activities and receive a quarterly status report from the program. If performance is lower than 80 percent of the target, providers must submit a corrective action plan within 15 days and must implement the plan in the following quarter.
If performance does not improve in the next quarter, contractors are placed on probation. If there is still no improvement, they remain on probation and must report to the program on a weekly basis. Annual reviews determine each contractor's second-year budget based on the first year's performance.
Contractors are expected to find jobs paying above minimum wage for their clients and to attempt to upgrade the jobs of clients already placed; 90-day follow-up is required for all job placements, including upgrades. The program supervisor reports that contractors are often able to make placements for below their projected costs. After meeting their targets, they are expected to continue to make placements throughout the contract period.
Minnesota's program was rated one of the six top performers in the country in 1997 by DHHS. In five fiscal years (1995 to 1999), the program increased job placements annually from 591 to 1,136, with this year's target at 1,285, which the program expects to exceed. Providers also regularly exceed their targets for clients' average hourly wages, jobs with health benefits, and termination of cash assistance because of earnings. The supervisor believes her program is effective because of open communication between the state and contractors, the state's monitoring efforts, and the willingness of state staff to provide timely technical assistance.
For more information, contact Quy Dam, Supervisor, Refugee Services Section, at (651) 297-3210.
A 1998 statute required the Department of Children and Families (DCF) to base all its contracts with local service providers on performance. The legislation created district units that monitor contracts and report to district administrators.
The department has about 1,700 contracts with 1,200 providers. Performance specifications and targets are included in each contract. In 1996, a working group of district, program, and administrative staff set the same types of outcomes for similar providers across the state. Providers' performance targets are negotiated each year by the department and the provider.
Performance contracts are generally limited to one year because of the state's budget cycle; therefore, data may be received after termination of a contract. The department uses such data to address existing providers' performance.
The department encountered several challenges in the conversion to performance contracts. The first year required a lot of training-state personnel traveled to districts to explain the new contracting procedures, including how to write the contracts so as to include performance specifications. Matching the collective targets of the local contractors to overall agency targets also proved difficult. Local contractors serving populations that are not represented in state DCF agency goals were faced with measuring outcomes they believed were not within their control.
DCF is still working on standardizing its contracting procedures across provider agencies and has created a Contracted Client Services unit to help integrate contract management and monitoring. It has not yet seen changes in the outcomes of services, but there has been increased attention to outcomes within the agency. The legislature has provided 49 new positions during the past two sessions to support DCF's efforts. Staff report that the conversion to performance contracts has produced a positive change in the agency's culture.
For more information, contact Bob Fierro, Director, Contracted Client Services, at (850) 487-1373.
The 1994 legislative act requiring Maine's Department of Human Services (DHS) to make all its contracts performance-based included a provision for collaboration among people inside and outside the organization, including legislators, providers, consumers, and state personnel. The law also included a hold-harmless clause for first- year contracts. After that, performance data would be used in award decisions.
Every contract has to include performance goals, performance indicators, and a strategy for reaching the goals. DHS initially asked providers to develop performance indicators, then worked with them to decide on the most important ones. Task forces of eight persons-two DHS staff, two providers, two consumers, and two legislators-were established for each of the 43 service areas to develop goals and indicators for the contracts. Contract officers and program staff served as facilitators to the work groups, changing their roles from mostly administrative to partnering.
Originally, providers had to specify what strategies they would use to accomplish the goals set by the state and how they would measure the indicators established by the task forces. However, it became apparent to DHS that data collection for the indicators had to be standardized. Without standardization, contract officers had difficulty monitoring performance, particularly given the large number of indicators negotiated in the contract process. Further, many of the indicators lacked validity, adding to the difficulties of interpretation.
Standardization of indicators and data collection procedures was accomplished through lengthy negotiations between contractors and the state, including assistance from the University of Southern Maine's Muskie School of Public Service. The school also trains service providers in the development and use of performance indicators.
Performance contracting has caused DHS to change from a predominantly internal contracting process to one that relies upon partnerships with providers to determine the right outcomes to measure. As a result, decisions are no longer all made from the top down.
Maine does not yet have information on program results, but agency staff report that moving to performance contracting has created a new, outcome-oriented culture at DHS. More program and contract staff are thinking in terms of outcomes, and providers are requesting assistance with data collection procedures for information they need to improve local management practices.
The department has focused equally on the development of process and the development of outcome indicators for its service areas. Although the legislation deals with the impact of state programs on individuals, DHS has not yet reached that level of measurement through its contracts. The first set of performance data is due this fall, but staff at the Muskie School expect it will be another two or three years before the system is fully operational.
For more information, contact Bruce Clary, Chair, Master's Program in Public Policy and Management, Muskie School of Public Service, at email@example.com.
Performance contracting can be implemented in many ways, with variations in the design of incentive and payment systems. Systems that tie payments to outcomes tend to show results more quickly: Oklahoma, Illinois, and North Carolina are showing impressive gains. Minnesota, which does not tie payment to outcomes, has improved its results by requiring target setting and regular reporting from contractors. Maine and Florida, which initiated performance contracting in response to legislation, are attempting agency-wide implementation, and their results will bear watching.
Lessons Learned in Performance Contracting
- Work with service providers to formulate client outcomes, indicators, and targets. Try to standardize indicators and data collection procedures for particular service areas (Maine).
- Get input from providers on design of payment schedules (North Carolina).
- Gain commitment and trust from providers by including a hold-harmless clause for the first year (Maine).
- Monitor and report back to providers regularly on their performance and how it compares to that of similar providers (Illinois).
- Work closely with providers to identify difficulties in service provision and to learn about changes in the client population that will affect performance (Minnesota).
- Setting quarterly targets and monitoring them is a good practice for providers who are paid with a grant contract, especially in the absence of milestone payments (Florida).
- Consistent communication and feedback on performance help to promote accountability and maintain contractor motivation (Minnesota).
- In the absence of financial incentives for performance, public congratulations or rewards may be appropriate ways to encourage contractors with higher-than-average performance (Minnesota).
- Use past performance in deciding future awards (Oklahoma).
- Use data from providers to publicize internally and externally the results of programs. Include comparisons of results achieved by contractors providing similar services and include explanatory information where appropriate (Illinois).
- Behn, Robert D., and Peter A. Kant. (1999). "Strategies for avoiding the pitfalls of performance contracting." Public Productivity in Management Review 22 (4): 470-89.
- Gordon, Stephen B. (1998). " Performance incentive contracting: Using the purchasing process to find money rather than spend it." Government Finance Review 14 (4): 33-36.
- Martin, Lawrence L. (1999). "Performance contracting: Extending performance measurment to another level." PA Times 22 (1): 1,8.
- www.welfareinfor.org/contractperf.htm. The Welfare Information Network provides resources related to performance contracting in human services.
- Communicating with Citizens about Government Performance: Local Leadership and State Support
- Human Resource Incentive Systems to Increase Performance
- Training State and Local Government Personnel to Govern-for-Results
- Intergovernmental Performance Linkages
- Citizen Use of Performance Data
- Presenting Performance Data
The Urban Institute
2100 M Street, N.W.
Washington, D.C. 20037
Phone: (202) 833-7200
Fax: (202) 429-0687
Copyright © 1999
The Urban Institute initiated its Governing-for-Results and Accountability Project in 1998. Supported by the Smith Richardson Foundation and the Alfred P. Sloan Foundation, the project team includes Blaine Liner (Director of the State Policy Center and Project Manager), Harry Hatry, Elisa Vinson, Ryan Allen, Pat Dusenbury, and Scott Bryant. The National Conference of State Legislatures is assisting with the project. This series of interim reports will include snapshots and success stories of the states and localities that have made contributions to Governing-for-Results and Accountability.