Wisconsin Department of Corrections Division of Community Corrections
Preliminary Agent Workload Study

I. INTRODUCTION

II. WORKLOAD STUDY BACKGROUND

III. WORKLOAD, CASE MANAGEMENT AND PUBLIC SAFETY IN PROBATION AND PAROLE

IV. WORKLOAD MEASUREMENT METHODS IN PROBATION AND PAROLE

V. THE WISCONSIN DIVISION OF COMMUNITY CORRECTIONS WORKLOAD STUDY

A. Participants
B. Selection of Supervision Cases/Investigations for Study
C. Duration of Time Recording
D. Division's Supervision and Investigation Standards

VI. WISCONSIN DIVISION OF COMMUNITY CORRECTIONS'WORKLOAD STUDY FINDINGS

A. Agent Hours Available for Supervision
B. Investigation and Case Supervision Findings

VII. RECOMMENDED CHANGES TO CCSD WORKLOAD ACCOUNTING

APPENDIX

I. INTRODUCTION

In 2000, the Wisconsin Department of Corrections contacted the National Council on Crime and Delinquency (NCCD) to conduct a workload study of probation and,parole agents for the Division of Community Corrections (DCC). Similar workload studies were initiated for the Division's unit supervisors, program assistants, and program assistant supervisors. The agent workload study was initiated as part of a memorandum of understanding between the state and the Wisconsin State Employees Union (WSEU). The memoranda recognized that both the community supervision program and the work performed by agents in the field had changed considerably since the last (1989) workload study. Consequently, the agent workload study has two primary objectives:

1. Update the Division's Case Classification and Staff Deployment (CCSD) system which is used to estimate the number of agents required to adequately supervise offenders placed on probation and parole. This task involved estimating the agent time required, on average, to supervise offenders, conduct pre‑sentence investigations, and perform other investigative tasks according to agency standards for adequate performance. The study recognizes that public safety is enhanced when probation and parole agents have the time necessary to perform the correctional mission in compliance with the standards employed by the Division.

2. The study findings also permit the Division to estimate future agent workload demand more accurately, deploy available staff resources more efficiently, and ensure equitable  distribution of workload capacity across regions of the state.

During the last 15 years, NCCD has conducted approximately 80 similar workload studies of adult correctional. agencies.

II. WORKLOAD STUDY BACKGROUND

An increasing workload burden is one of the critical problems facing America's probation and parole agency managers. During the last decade, the number of offenders supervised by probation or parole agencies has grown from 3.2 million to more than 4.5 million, an increase of more than 40%.¹  While each correctional agency operates ina different environment, increased national demand for community supervision is attributable I to several factors, including increased use of probation as a sentencing option, diversion of offenders from crowded institutions, higher arrest and conviction rates, and longer probation terms.

Wisconsin has experienced a similar trend. The number of probation and parole cases has grown from 20,000 in the early 1980's to 31,000 in 1990. Wisconsin currently has 64,000 offenders, and 1999‑2001 trends indicate that this population is growing. In 1990, the Division employed 550 probation and parole agents. Currently there are nearly 1,200 agents actively supervising offenders in Wisconsin's communities.

In addition to the sheer growth in the State's probation and parole caseload, Wisconsin's DCC, like many probation and parole agencies, has implemented new programs and assumed new duties that change the nature of the work agents must perform. Since completion of the last workload study in 1989, for instance, the DCC has developed statewide intensive supervision programs for drug offenders and sex offenders, as well as electronic monitoring, day reporting and random drug testing. It has also implemented a Domestic Violence Program, as well as Neighborhood Supervision, enhanced supervision and absconder programs in selected areas of the state. Additional responsibilities have also been assumed, including:

• Supervision of 17 year olds (1996);
• Sex offender registration and supervision (1993); and
• The collection of supervision fees from offenders (1996).
  

The Division now collects approximately $15 million in court fees and $5 million in supervision fees from offenders each year. The administrative burden of accounting for these collections has grown considerably.

The changes to the nature and type of the work performed have a direct impact on the probation and parole agents who perform community supervision, as well as the staff who support them. Many agents now work evenings and weekends to enhance public safety. Agents also receive more training. The Division initiated Agent Basic Training for all new staff in 1991. In addition, specialized supervision programs require additional agent training and often increase the planning and case support activities of agents. One major change in the staff resource allocation was the application of the Fair Labor Standards Act to probation and parole staff that established strict enforcement of a 40‑hour week. A recent memorandum of understanding (MOU 14) between the state and the WSEU recognized these revised expectations. The agency has also employed new technologies to increase the operational efficiency of staff by computerizing forms, developing computerized case record keeping systems and increasing the use of email, pagers, and cell phones. These changes underscore the need for the current workload study.

III. WORKLOAD, CASE MANAGEMENT AND PUBLIC SAFETY IN PROBATION AND PAROLE

In many jurisdictions, the increase in the number of offenders placed on community supervision has not been accompanied by a proportionate increase in the agents who supervise offenders in the. field. When a correctional agency does not have adequate staff, agent contact with offenders may be reduced to a level at which community supervision is ineffective. One possible consequence of agency understaffing is not well understood: If correctional agencies fail to maintain credible supervision standards, public safety may be jeopardized. A brief historical review of the relationship between offender risk to the community, case contact standards and workload in community corrections makes this point clear.

In 1981, the United States Department of Justice=s National Institute of Corrections (NIC) adopted a case management system called the Case Classification and Staff Deployment (CCSD) as a national "model" for managing probation or parole supervision. This approach to case management was the product of a three year research and demonstration effort funded by the Wisconsin State Legislature and managed by the Wisconsin Department of Corrections. In effect, the Wisconsin Department of Corrections developed what became known as the NIC model system, and it was the first correctional agency to implement this innovative management tool.

During the 1980's, NIC provided extensive training and technical assistance to over, 100 state and local correctional agencies that implemented the CCSD system. It remains the most widely used case management system for community corrections in the United States. The system was developed with a clear objective: To reduce criminal offending, and thus improve public safety, by more efficiently managing existing correctional resources. When an offender enters probation or parole, an actuarial risk assessment procedure is employed to estimate the probability, or risk, that the offender will commit a new offense during community supervision. If the probability of a new offense is very high, the ‑offender is assigned to a high standard of supervision that involves frequent agent contact and increased surveillance. If risk is low, the offender is assigned to a much lower supervision standard that requires less time to meet.

The CCSD system is explicit in its assumption that the most critical resource community correction agencies manage is the time agents have.available to supervise offenders. Given the unprecedented growth in state probation and parole caseloads during the 1980s and 1990s, efficient staff deployment and adequate staffing have become areas of concern for both correctional agency managers and elected officials. The CCSD system has proven to be an extremely useful management tool because it matches the level of supervision each offender receives with the risk the case poses to public safety. For instance, correctional agencies that employ the system, as does Wisconsin DCC, adopt differential case supervision standards. These standards range in intensity from one face‑toface contact every three months by an agent for a minimum risk offender to weekly face‑to‑face contacts. including a home visit for a high risk offender.

The operating assumption of the NIC model case classification system is simple. Given that agencies have limited staff resources for supervision, it makes sense to supervise high risk offenders, who are most likely to re‑offend, much more closely than low risk offenders. This strategy, which is central to the CCSD system, is based on research conducted in Wisconsin that observed the impact of different supervision levels on criminal behavior. The Wisconsin study indicated that criminal activity among high risk offenders was reduced by nearly 50% when they received close supervision that included multiple monthly contacts with probation agents.²

This study demonstrated that correctional agencies could reduce crime and enhance public safety by systematically providing close supervision to high risk offenders. It also established that this result could only be obtained when high risk offenders were accurately identified and adequate staff resources were available to provide effective community supervision. Consequently, NIC encouraged correctional agencies to accurately assess their staffing needs by conducting workload studies that estimate how much time agents require to meet clear supervision standards based on offender risk.

IV.WORKLOAD MEASUREMENT METHODS IN PROBATION AND PAROLE

The method NIC recommends for estimating the number of agents necessary to perform effective community supervision assumes that the agency has:

1. Risk assessment procedures for identifying high and minimum risk cases; and

2. Has adopted case contact standards based on risk for supervising offenders.

Wisconsin's CCSD includes both elements. For instance, DCC requires that a high risk offender receive four face‑to‑face contacts per month versus one contact every three months for a minimum risk supervision case. In effect, an agent may require four or five times longer to supervise a high risk case than a minimum risk case. The Division has established similar standards for other case types and offender investigations. The CCSD workload management system estimates agency "workload demand" for cases based on risk profile and the applicable case contact standard. It also employs an objective measure of the time agents have available to supervise offenders when estimating future staffing needs. This approach to agent workload recognizes that adequate staffing is necessary to consistently meet supervision standards for all offenders.

The methodology NCCD employs for conducting correctional workload studies conforms to the CCSD system recommended by NIC. The studies employ a prescriptive, case‑based methodology. For instance, case time estimates are established by asking agents to record, under actual field conditions, the time required to supervise a sample of cases assigned to different risk levels during a month. Agents are asked to meet all applicable agency performance standards when serving these cases. They also record the time they spend performing other duties such as case support, unit training or management assignments, and community activities. The workload value or time required to perform each of the major agency supervision or, investigative tasks is derived by observing cases that met or exceeded minimum agency performance requirements.

For example, the study estimates the time an agent requires to supervise a high risk offender each month when he or she contacts the offender face‑to‑face four times and makes at least one home visit. The agent is also expected to respond to non‑routine case demands, such as violations or new criminal offenses, and to adequately document his. or her case activity. In effect, the study focuses on the case that must be supervised at a standard defined as adequate by the correctional agency given the offender's risk to the community.

NIC supports a prescriptive approach to workload measurement because it ensures that supervision can be performed in a manner that can have a positive impact on public safety.  Theoretically, an agent could supervise a high risk case without contacting the offender every month or verifying employment or residence. Obviously, this would take far less time than the current DCC supervision standards for high risk offenders, but this low level of effort is unlikely to reduce criminal activity.

A similar example may be drawn for pre‑sentence investigations. Agents could conduct these investigations without checking the offender=s prior criminal history, investigating the circumstances of the offense, or contacting victims. Again, this approach would take far less staff time and require fewer agents to perform than a thorough investigation that carefully assesses each offender's background. Judges, however, would make poorly informed sentencing decisions that could adversely impact public safety. As these examples illustrate, the purpose of employing clear performance standards when estimating agency workload is to ensure that adequate staff resources are available to protect the public.

V. THE WISCONSIN DIVISION OF COMMUNITY CORRECTIONS WORKLOAD STUDY

Conducting a workload study is a labor‑intensive process, requiring substantial planning and preparation with management and field staff. In August 2000, NCCD staff began working with a DCC planning committee to design this workload study. The committee consisted of representatives from DCC management, agents, and program assistants or supervisors from several DCC units. The agent workload study group met on several occasions to review the duties of the position, identify case types, design data collection forms, and test study procedures.

Each agent participating in the workload study received training to familiarize them with the purpose of the study, as well as instructions on how to record time on sample cases during the two month study period. Agents from the workload planning committee attended each training session to answer questions from the participants.

A. Participants

Forty‑four units were randomly selected across the state, representing both rural and urban areas. Approximately 575 agents present for duty during the study period from the sampled units were involved in the workload study. Each agent recorded the time he or she spent (including travel) on a sample of their supervision cases or investigations that met or exceeded agency standards during a 60‑day period during March and April 2001. Time spent on general case support and non‑case related administrative activity times was also recorded. No DCC supporting staff, such as sergeants and service coordinators, were included in the agent study. Program assistants and program assistant supervisors participated in their own workload studies and thus did not record time for this one. Unit supervisors helped implement the agent workload study but did not record time.

B. Selection of Supervision Cases/Investigations for Study

Each agent with an active caseload was asked to record time for three or four randomly selected cases of offenders currently receiving supervision plus a new intake case assigned to them during the study. Pre‑sentence court investigations, as well other types of offender investigations chosen for time recording, were the first ones assigned to each agent by unit supervisors during the study, The sampling procedures employed were designed to ensure that participants could not choose cases or investigations assigned to them for time recording.

C. Duration of Time Recording

Supervision case time was recorded for two separate one‑month periods. Investigations were recorded from the time of assignment until completion. New intake supervision cases were tracked for a one‑month period beginning the day the case was assigned to the agent.

D. Division Supervision and Investigation Standards

In developing workload estimates, the study employs data from sample cases in which DCC supervision or investigation standards were met or exceeded by the recording agent. The result is an estimate of how long it takes an agent to meet the division's' minimum standard for adequate performance. The supervision or investigation standards used were those currently in effect at the time of the study. Wisconsin Division of Community Corrections employs a risk classification system in its case management system and differential contact standards for each of five supervision levels: minimum, medium, maximum, high, and intensive supervision (alternately referred to as High Risk Sex Offender. supervision). For instance, high and intensive supervision cases are contacted by the assigned agent face‑to‑face at least four times per month and receive a home visit. Maximum supervision cases receive two face‑to‑face contacts each month and one home visit. Medium supervision involves one monthly face‑to‑face contact. Minimum cases receive face‑to‑face contacts once every 90 days. Agents are also respon sible for documenting case activity, managing offender treatment plans and responding to special circumstances as needed. The Division has similar standards for pre‑sentence investigations (PSI), full and partial social investigations, pre‑parole investigations (PPI) and special bulletin notices (SBN). Detailed standards appear in the appendices. Study findings described below are based on analyses of the time agents recorded for sample cases that meet applicable agency standards. Unit supervisors reviewed all cases to establish that standards were met and additional edit checks were performed by NCCD.

VI. WISCONSIN DIVISION OF COMMUNITY CORRECTIONS WORKLOAD STUDY FINDINGS

The presentation that follows reviews the major workload findings. The analyses of the time agents have available to perform supervision and investigative work is followed by a description of the workload values estimated for the supervision cases and investigations examined during the study.

A. Monthly Hours Available for Agents

Based upon a 40‑hour workweek, full‑time agents are currently employed for 173.3 hours each month. The current workload study estimated the time agents spent performing case support or administrative tasks that reduces the time they have available to supervise offenders or complete investigations. Results of the study, indicate that agents required, on average, 13.1.hours per month to perform administrative activities, including staff training (6.4 hours), participating in unit or regional meetings and planning committees ‑ (6.1 hours), providing public information in the community (0.5 hours), and other administrative activities (0. 1 hours).

Agents also perform a variety of case support activities that require them to, assist with supervision cases or investigations that are not assigned to them. For example, one agent may need to substitute for another by meeting with an offender or performing some other supervision activity (2.8 hours). For security reasons, an agent may perform backup coverage by accompanying another agent who is conducting an offender home visit or transporting an offender (4.6 hours). Agents spend an average of 2.1 hours per, month meeting formally or informally with co‑workers about their cases. Agents also spend an average of 0.3 hours coordinating or developing DCC programs, 0.3 hours facilitating or participating in offender group work, and 0.4 hours doing other case support. These kinds of case support tasks, which are common in correctional agencies, required 10.5 hours per month to perform. See Table 1. (See the Appendix for activity definitions.)

Table I

The combined administrative and case support time estimate is 23.6 hours per month.  Deductions must also be made for vacation, sick time, holidays, and break time when authorized by the agency. These deductions are generally described in the following manner:

Vacation Leave: The mean vacation time available to agents during a particular calendar or fiscal year. The data are obtained from the division's personnel records.

Sick Leave: The mean amount of sick leave used by agents during the same calendar or fiscal year as used to compute vacation time available. This figure excludes agents on extended approved sick leave.

Holidays: The number of holidays authorized by the agency per calendar year.

Break Time: This time is set by contract and averages 0.25 hours per half‑day shift. Break time is calculated by dividing the number of remaining monthly available hours (i.e., after subtracting vacation, sick leave, and holiday) by four to obtain the number of half‑days on work status. The result is multiplied by 0.25 hours to find the amount of break time expected while an agent is at work.

When leave and break time, plus case support and. administrative time are subtracted from 173.3 monthly work hours, agents have 117.2 hours each month to supervise offenders and complete investigations,

In addition, a further adjustment must be made to time available per month to accurately reflect the entry‑level agents currently assigned to the division. During their first year, new agents must complete a training program that averages 30 hours per month.³ They are not available to supervise cases or complete investigations during training. At present approximately 9.5% of the agents working for the division are in this training status.⁴ Time available is adjusted to 114.4 hours to recognize the reduced availability of DCC agents in training status.⁵

Table 2

B. Investigation and Supervision Case Findings

Investigation workload hours observed during the study are shown in the figure below. The workload hours shown in the figure were estimated by averaging the time agents recorded to complete investigations that met applicable DCC standards. For instance, agents recorded 12.9 hours for full pre‑sentence investigations (PSI) that met standards. Social full investigations required 4.3 hours and social partials 2.7. Pre‑parole investigations (PPI) and special bulletin notices (SBN) took 2.6 and 10.2 hours respectively. New case intakes required 3.3 hours. ‑Revocation completion times were observed in Milwaukee's specialized unit and averaged 5.0 hours.

Similar findings for classified cases are shown below. Again the workload hours shown are drawn from sample cases in which the agent met or exceeded applicable DCC supervision standards during the month they were observed. As might be expected, the times vary significantly by case type because high risk cases are contacted more frequently by agents. For instance, agents required 5.8 hours each month to meet standards for a case supervised at the High Risk Sex Offender (HRSO) standard. Minimum risk cases, which agents supervised at a much lower standard, required only one hour. Administrative case s were not studied because their supervision is contracted. Interstate compact cases are supervised according to the risk level assigned by CCSD and the times shown are applicable to them.

VII. RECOMMENDED CHANGES TO CCSD WORKLOAD ACCOUNTING

Based on the study findings, NCCD recommends the following changes to the current CCSD workload accounting system. These changes impact both the budget projection procedures employed by Division administrative staff and the ledger system used in field units to track each agent's workload burden.

First, the supervision, investigation and agent time available estimates from this study should be used to update the current CCSD system. This will more accurately reflect the time agents require for completing investigations and supervising offenders. It will also more accurately reflect the time they have available to perform these tasks after their case support and administrative duties are fully accounted for. In addition, CCSD should be revised to permit full crediting of pre‑parole investigations currently completed by agents in the field and to reflect additional workload credit for completing revocations. The current ledger system does not fully account for these agent tasks. Revocation work could be credited to the ledger when the 414 is served and PPI work can be credited in the same manner as a social, investigation (i.e., during a 30‑day time frame). A recent memorandum of understanding (MOU 14) between the state and the WSEU has already recognized the need for many of these changes. This study confirms many of the issues raised in the MOU.

An additional recommendation concerns the use of workload 'points' to estimate each agent's monthly workload burden. The point system was established several years ago as a unit of measurement in the Division's unit ledger. It has been used to manage workload in field units and has been referenced in, bargaining agreements between labor and management. Since one workload point is equivalent to 35 minutes, it can easily be converted into workload hours. I Conversion to hours seems to offer several advantages. It would simplify the CCSD workload accounting system, make it easier to maintain and probably improve the accuracy of unit workload estimates. An important additional benefit is that the use of workload hours should be more easily comprehensible to all the parties involved in the Division's workload accounting operation, This includes agents in the field, WSEU representatives, DCC managers and legislative staff.

1. Bureau of Justice Statistics, Correctional -P=IaOons in the U.S., Wasbb ton, D.C., U.S. Department of Justice, August 2001.

2. Eisenberg, Mchael and Gregory Markley, "SonrtWng Works in Conmmnity Supervision," Federal Probation, Vol. 5 1, No. 4, 1987. Bairct Heinz and Bemus, A Two‑Year Follow‑Up of the Wisconsin Case Classification Project, American Correctiona Association Monogr (1981).

3. New agents must complete 360 hours of training including 280 hours of classroom time and 80 hour of on thejob orientation. Over the course of a year, this training time will average 30 hours per month.

4. Approximately 120 agents complete ABT during a year. This is'9.5% of the total allocated agents (1,265).

5. The adjusted hours available per agent computation equals (I l7.2*.905)+(87.2*.095). New agents have 30 less hours a month (thus 87.2) and comprise 9.5% of allocated agents.

Appendix