Kevin N. Wright, Ph.D. School of Education and Human Development State University of New York at Binghamton

Introduction
Capacity
Purpose of Assessment
Methodology
Results
Conclusions

SEVERAL STATE AND federal correctional agencies have developed performance measurement systems. The success of these efforts led the Association of State Correctional Administrators (ASCA) to initiate activities to develop a national outcome-based performance measurement system. ASCA identified measures that reflect the most crucial aspects of correctional management, developed indicators to assess each measure, and specified which of the operational definitions for each indicator to allow for application across jurisdictions. For ASCA to continue, it must support the development of capacity within correctional agencies to participate in a national reporting system.

This research assessed the current capacity of departments of corrections to participate in a national reporting system. A survey of departments found that almost all prison systems possess the capacity to measure and report about the aggregate prisoner population under supervision and individual characteristics of that population. Most prison agencies collect information about escapes, homicides and suicides. Beyond these measures, a significant disparity exists among departments regarding their capacity to measure and report on all other indicators.

To achieve its goal of advancing the commitment of adult institutional corrections to performance-based management, ASCA must invest in the appreciation for and commitment to the philosophy and practice among correctional professionals.

Performance measures assess the results achieved from activities performed within an organization. By routinely measuring performance, an agency can begin to chart its achievement in accomplishing desired outcomes. Such information can then be used to improve or enhance services, to evaluate management practices and to support continuous improvement. Performance-based management has become increasingly popular in prison administration over the past decade. The Federal Bureau of Prisons led the way among correctional agencies in adopting performance-based management. The Bureau developed its Key Indicator and Strategic Support System (KISSS) in which vital information about prisoners, staff, finances and health services was provided to central office and facility executives. States such as Florida, Minnesota and North Carolina embraced performance-based management as part of statewide initiatives to implement the practice across state agencies.

What can be learned about the quality of performance through comparisons of an organization to itself over time is limited. When an organization can examine its practices against those of other similar organizations, it can begin to assess its approaches and managerial styles against alternatives used elsewhere to determine what works best. Areas needing improvement and best practices can be identified. Enhanced possibilities for making decisions rationally and objectively are available with cross-organizational comparisons.

Currently, sources of cross-jurisdictional information about prison performance are few. Two organizations compile information about prisons and prisoners from departments of correction across the nation. While it is not their primary purpose, the two sources provide some information about prison performance that can be used for cross-jurisdictional comparisons and analysis. The Criminal Justice Institute, Inc., a for-profit consulting firm located in Middletown, Connecticut, produces the Corrections Yearbook annually. Data for the publication are supplied by state and federal prison systems in response to a mailed questionnaire. CJI (2002) notes that it relies upon its respondents for the accuracy and currency of the information it reports. Most of the data consist of statistics about numbers of prisoners and prisons; however, a few outcome indicators of performance are presented in the Yearbook, including escapes, assaults, homicides and suicides.

Interestingly, most outcome indicators are reported as frequencies rather than rates. Since some jurisdictions have larger numbers of prisoners than others, and thus a larger at-risk population, cross-jurisdictional comparisons are inappropriate. The number of incidents must be divided by number of prisoners to obtain a rate per individual; only then can the jurisdictional data be compared.

Some interesting anomalies can be found in reviewing the frequency tables provided by CJI (2002). In the 2001 volume of the Corrections Yearbook, the Alaska prison system reported 44 escapes from secure facilities, Kentucky had 69 and Louisiana had 39. All other states and the federal system had four or fewer escapes, with most having none (p. 33). Only two conclusions seem possible from these data. Either Alaska, Kentucky and Louisiana are extremely lax in attending to security and protecting the public, or these three jurisdictions define "escape" very differently from all other jurisdictions.

Similarly, in the same volume, Pennsylvania reported that only 16 assaults on staff were committed by its 36,000 prisoners during 2000; whereas, its neighbor to the west, Ohio, listed a whopping 467 assaults committed on staff by its slightly larger population of 45,000 prisoners. Once again, one is left with two options: either the Pennsylvania prison system is doing an extraordinary job of preventing attacks on staff in comparison to Ohio, or the two departments are counting assaults differently.

The second source of information that can be used for some comparisons of cross-jurisdictional prison performance is the Bureau of Justice Statistics, an entity of the federal government located within the Department of Justice. Every few years, BJS conducts a national survey of individuals incarcerated in state and federal facilities. Since data collection in the BJS survey is not left to individual jurisdictions but is conducted by trained interviewers supervised by agency staff, one can be more confident of cross-jurisdictional comparability than with the CJI information. However, most of the BJS survey information concerns prisoners rather than prison performance. Only a few questions, such as whether an individual has been injured as the result of an attack by another prisoner, whether an individual self-identifying as having a mental health or drug and alcohol problem is receiving treatment, and whether the respondent is participating in a prison program, can provide a basis for comparing outcomes across jurisdictions. Unlike the Corrections Yearbook, where information is nicely summarized annually across jurisdictions, one would have to examine spreadsheets containing the data to assess how departments are doing using the BJS data.

Since these are the only significant and regularly collected cross-jurisdictional data, and given their limitations, the opportunity for comparing performance among prisons or prison systems is extremely limited. Yet, compelling reasons exist for seeking a cross-jurisdictional assessment of performance. One reason is to clear up confusion that arises from inappropriate comparisons. All prison systems collect information about performance. Unfortunately, little uniformity exists in defining measures and collecting information. While comparisons among prison systems should not be made, reporters, legislators and budget analysts regularly attempt to do so. This practice can lead to erroneous conclusions. Second, while the control and punishment of criminal behavior is clearly delegated to be a state function, this does not preclude the establishment of a set of clearly articulated values about what constitutes quality incarceration by correctional professionals. National performance standards would establish objective norms against which agencies could measure their performance. Third, the nation's prison systems vary considerably with respect to practices, resources and management. A national performance measurement system would allow for cross-jurisdictional assessments of best and most efficient practices. Fourth, performance measurement recognizes good practice and identifies agencies that need improvement. As such, a national reporting system would support quality improvement and encourage jurisdictions to consider how they should perform in the future. And, finally, the establishment of a national system and the request that all state agencies participate would encourage jurisdictions to begin developing the infrastructure for performance- based management.

Recognizing these needs, the Association of State Correctional Administrators (ASCA), the national professional organization of chief executive officers of the nation's prison systems, set as its highest priority the development of a national performance measurement system. ASCA's interest in performance measurement dates back to the mid-1990s, when it began to work with the Urban Institute to determine the availability and use of common data elements among departments of corrections. In a national survey of state and federal prison agencies, the Urban Institute found that jurisdictions varied considerably in the degree to which they collected various data elements and the extent to which those data were available electronically. Importantly, departments do not collect the information in exactly the same manner (Urban Institute, 1998: 98).

ASCA set about designing a national performance measurement system in August 1999, when its Executive Committee endorsed a resolution to create a subcommittee to develop measures that correctional departments could use to track, compare and assess progress toward meeting identified objectives. Over the next two years, that subcommittee worked to develop an initial performance measurement model. The group identified eight standards of correctional performance that represent the most important elements of institutional correctional processes and selected four of those eight standards for further specification. The subcommittee then chose indicators for each of the four standards, each sufficiently detailed and comprehensive to capture the crucial aspects of the over-reaching area of correctional performance. The final task in developing the performance measurement model was the specification of definitional rules for measuring each indicator that would be sufficiently detailed to assure comparability across jurisdictions; that would account for important cross-jurisdictional differences in mission, legal structures and organizational arrangements; and that would produce information applicable across jurisdictions. If jurisdictions follow these rules, then data will be reliable and valid. The problems noted above with the CJI data will be avoided.

A total of 43 indicators for the four performance standards comprise the ASCA model. Because of the detail of the counting rules, the fully specified model is lengthy. (A copy can be obtained from the ASCA headquarters.) Each indicator is expressed as a rate per population unit. Rates are necessary so that outcomes can be compared across jurisdictions. If actual occurrences were simply counted without converting them into rates, large states would generally show higher frequencies than smaller agencies.

I describe one indicator here to illustrate the specification contained in the measurement definitions. That measure is prisoner assaults on staff. Any one of several indicators could have been chosen for prisoner assaults on staff, but the subcommittee selected the following definition:

Stock population on June 30 is used as a proxy for average daily population as it is an easier measure for prison systems to measure.

For the numerator, one must consider whether all aggressive acts toward staff are to be counted or only those involving injury and weapons use. If the latter definition is used, what constitutes a weapon? Should the throwing of body fluids be included? What is an injury—a cut or bruise or more serious medical issues that require the attention of a health care professional? Who is included as a staff member—any civilian in the facility, including volunteers, student interns, repair personnel representing a private vendor, contract workers, etc., or only paid employees of the department of corrections? Counting rules are specified to address these issues. The rules selected for the numerator of prisoner-on-staff assaults are as follows:

1. A staff member is defined as an individual who is employed by the DOC or facility on a full-time, part-time, or contractual basis, and/or other individuals performing correctional services, e.g., volunteers or interns. Civilians who are attacked, but are not paid staff should not be counted. These individuals might include visitors, truck drivers, service personnel repairing equipment in the facility and construction workers employed by contractors who have projects within the facility.
   
   
2. A single incident may have more than one victim; count the number of victims.
   
   
3. Count all injuries that require medical attention and treatment.
   
   
4. The fact that an assault has taken place does not have to be substantiated by the disciplinary process; however, there must be sufficient evidence that the injury resulted from an attack and not an accident.
   
   
5. To be counted, the assailant must be incarcerated.
   
   
6. Do not include prisoners housed in other states or held in privately run facilities.
   
   
7. In the unified system, if the assailant is known, to be counted that individual must be convicted of a crime and serving sentence greater than one year.

The selection of these particular rules was somewhat arbitrary. For example, in the definition of staff, contract workers could have been included. What is important is that an agreed upon plan for data collection has been specified. If departments of correction collect information as specified, then the outcome measures will be comparable across jurisdictions.

With this model, ASCA now possesses the basic structure to proceed with the development of a national performance measurement system. For ASCA to continue its endeavor, it must engage in two additional tasks. The Association must acquire a data platform into which departments can begin to report performance data. Simultaneously, ASCA must support the development of capacity within correctional agencies to participate in a national reporting system. Development of capacity involves building the infrastructure that will allow for continuous performance monitoring.

back to top

Capacity

Historically, correctional agencies developed information systems to support operational activities. Early on, all information was maintained in paper files. With computerization, agencies began to capture some of the information contained in written records and automate it. For the most part, these data files were developed and designed to support operational decision-making, so they tended to be specific and functional.

Most departments of corrections maintain several record systems, each associated with a specific functional area. Departments keep records on prisoners, including demographics, offense history, current offense, sentence, and perhaps, information about substance abuse, education, mental health and so forth. Departments automate some but not all of this information. Agencies also keep information about significant events that occur in the day-to-day operations of prisons. When a significant event occurs—an assault, for example—staff members write a report or series of reports describing the event. Many systems have designed forms for recording the information that is then automated. Departments of corrections also record information about staff. These data are often maintained by human resources personnel and vary in the degree to which they are automated. Prison medical units record information about the physical and mental health of prisoners as well as information about substance abuse and the treatment that prisoners undergo.

The most frequent use of prison records is in day-to-day decision making. Personnel considering transferring a prisoner may review his or her records to determine time served or time-remaining, program needs or history of disruptive behavior while incarcerated. Because different functional areas are involved in creating and collecting information about prisoners, prisons' automated information systems tend to draw from these different data sources to create an integrated data record for each prisoner.

Another use of prison records is in report writing. Here, rather than focusing on an individual, assessments are conducted on aggregate populations. Prison administrators may be interested in the number of prisoner-on-prisoner assaults that occur each month. A lawsuit could generate a need to know the percentage of the prison population receiving some form of substance abuse treatment. Here, individual data must be scanned and aggregate patterns tallied.

As an organization moves from operational- oriented management to strategic- and performance-based management, the need for information to support decision making increases. The same is true if the agency decides to participate in a national performance indicator system. A department must enhance its distinctive operational databases—prisoner information, disciplinary records, personnel data, medical records, etc.—to create integrated data systems that can be queried to support decision making. In particular, integrated systems that allow for performance monitoring must be designed.

The capacity of a department to participate in a national performance measurement system is directly linked to the development of its data systems. Departments that have integrated, performance-based information systems can begin to participate and report in a national system rather quickly and with minimal staff time and expense. Departments that still employ operational databases will find such participation more difficult. The ability to query databases and report on performance is referred to as capacity. Capacity to participate in a national performance measurement system is determined by the data collected, the automation of those data, and the ability to query the information to obtain measures of indicators of performance.

back to top

Purpose of Assessment

The topics contained within the ASCA model represent performance issues that prison executives consider to be among the most important aspects of managing prisons well. For this reason, one would expect most departments of correction to collect most of the information included in the model. However, ASCA decided to pursue a national reporting system because different prison systems collect data in very dissimilar ways, using unique definitions and data collection rules. For that reason, cross-jurisdictional comparisons are inappropriate and not valid.

The purpose of this assessment was to determine to what extent the Nation's departments of corrections are prepared to participate in a national performance measurement reporting program. In conducting this assessment, I used the ASCA model with its own unique set of definitions and counting rules. However, I suspect that the findings would be the same whatever the model is.

It was expected that most departments of correction would use some but not all the counting rules for most indicators. Some systems might also employ rules that were not specified in the ASCA model. In other cases, agencies could measure an indicator completely different from the way it is specified as an ASCA indicator. For example, some state prison agencies use an incident-based measure of prisoner-on-prisoner assaults rather than definitions consisting of numbers of assailants and victims as selected by ASCA.

After determining the extent to which an agency measures a particular indicator using the counting rules, the next concern was to determine what would be required to collect the information as it has been specified. It was expected that for some indicators, a department might need only to make minor changes that involve minimal staff time and expense. The information necessary to report rates as specified could already be collected, and staff would need only to write computer code to capture and report the information as desired. In other cases, beginning to collect the data could pose a major task for an agency requiring a policy change, considerable expensive data collection, and the retraining of operational staff as to the value and utility of the information.

Because any national reporting system that is eventually designed and implemented will require the input of automated data, the third topic examined in this assessment was the extent to which information regarding each key indicator is automated. If the data are collected, but not available in a machine-readable format, then the assessment needed to determine what was required to automate the data for a prison agency.

The assessment of prison agencies' capacity to participate in a national performance measurement system, thus, proceeded with three questions:

1. Does the department collect the information on a key indicator as specified by the definitions and counting rules?
   
   
2. If not, what would be involved in beginning to collect the information as defined and specified?
   
   
3. Are the data automated? If not, what would be involved in automating the information?

These three questions needed to be answered for each key indicator contained in ASCA performance measures model. The following sections describe the methodology used for this task and the findings of the assessment.

back to top

Methodology

Procedure: Because of the level of detail necessary about how data are collected and which, if any, of the counting rules are applied in obtaining the information, written questionnaires were inadequate for this task. Rather, interviews were conducted with those intimately knowledgeable about the separate departments' information systems. These interviews were conducted by telephone by two researchers working with the ASCA Subcommittee.

Respondents received a copy of the model via e-mail and were told that a member of the project team would contact them. During an initial e-mail and/or telephone contact, the assessment was explained, a time for an interview was scheduled, and respondents were told that they would be asked the three questions outlined above for each data element.

At the time of the survey, 52 distinctive data elements comprised the model (several elements were revised based upon information obtained from the survey). Interviews generally lasted approximately one hour but ranged from 35 minutes to 2 hours. During the course of the telephone sessions, respondents were asked the three questions listed above for each of the 52 data elements. The respondents' answers were recorded on data sheets.

Sample. A census of all state prison systems and the federal prison system was initially thought to be necessary to determine capacity among the Nation's prison departments to participate in a performance measurement program. All ASCA members (with the exception of city and foreign departments of corrections) were contacted and asked to nominate an individual to participate in the assessment. As nominations came in, nominated individuals were contacted to arrange an interview. Because of the breadth of the ASCA model, some departments had representatives from different functional units present during the interviews. Respondents included research, information technology, and/or planning staff.

After 17 interviews were completed, response patterns were being repeated by new interviewees with little new information about capacity being obtained. The subcommittee decided that data from all jurisdictions were not required. The group, however, saw value in including a few additional states in specific geographic locations and with specific characteristics to assure that those surveyed represented the array of departments across the nation. Personnel from seven additional agencies were interviewed, resulting in a total sample of 24 departments.

Analyses. The interviews produced a wealth of rich detail regarding prison systems' data collection practices and information technology systems. Two techniques were employed to synthesize this information to gain an overall perspective of capacity. First, a coding system was developed to summarize response patterns. For each data element, respondents' answers were classified into four categories: 1) collecting as specified, 2) collecting but recoding needed, 3) collecting but not automated, and 4) not collecting. The individual who conducted the interview assigned the data elements into one of the four categories. To ensure reliability of these data, the classifications were emailed back to respondents, and they were asked to confirm that the representation was accurate.

The second set of analyses involved qualitative assessments. The two people who conducted the telephone interviews reviewed field notes looking for patterns as to the ability of departments to provide the desired performance information. Each reviewer worked independently and then the two compared their assessments to assess and enhance reliability of their evaluations.

back to top

Results

Quantitative Assessment. The summary results of the telephone surveys are provided in Table 1. By scanning the first column of results, Collecting as Specified, one can determine to what extent the Nation's departments of corrections are currently prepared to participate in a national reporting system if it were implemented today. Recall that all key indicators have both a numerator and denominator. The numerator contains substantive information about the topical area of the indicator (e.g., number of escapes from secure facilities), and the denominator includes information that adjusts for population size (e.g., number of prisoners held in secure facilities). A quick scan of the "Collecting as Specified" column reveals that agencies are collecting data about the denominators (i.e., information about the prisoner population). Still, even here, 20 percent of the departments, could not provide information about basic characteristics of their prisoner populations.

Regarding the numerators, the actual substantive information of interest, the findings are far less encouraging. For a few indicators, almost half the departments are collecting information as specified (i.e., 46 percent for escapes from within secure facilities and 38 percent for prisoner-on-staff homicides). But, for many other indicators, the percentage of departments collecting the information as specified is extremely low. Only 17 percent of the departments currently use the same definition of prisoner-on-prisoner attacks as the ASCA model. Only 8 percent of the systems can report, as specified, when a prisoner has sexually attacked a staff member. Likewise, only 8 percent can report major disturbances that occur each year as it has been defined.

Reviewing the percentages of departments that can provide information regarding the various key indicators provides a partial explanation of capacity. One also needs to examine how many departments cannot provide information as specified. The last two columns, Collecting Not Automated and Not Collecting, identify departments that will require substantial time and effort to begin measuring performance as it is defined by the ASCA model. By adding the percentage of the two columns together, one obtains an indication of how many departments will require considerable effort to begin reporting.

For the first standard, Public Safety, about 20 to 25 percent of the prison systems have no automated information on the key indicators or are not collecting the information currently. For example, 21 percent of the agencies fall into these categories for escapes from within secure facilities, 25 percent for escapes from outside secure facilities and 21 percent for returns to prison for a new conviction.

However, the second standard—Institutional Safety—is where one observes considerable under-capacity to participate in a reporting system. Approximately half of the departments lack automated data or do not collect information about:

1. Victims of Prisoner-on-Prisoner Assaults,
2. Prisoner-on-Staff Assaults,
3. Prisoner-on-Prisoner Sexual Assaults,
4. Prisoner-on-Staff Sexual Assaults,
5. Staff Sexual Misconduct,
6. Staff Homicides, and
7. Disturbances.

Few departments collect and automate information about the number of hours of substance abuse and mental health treatment and assessment provided by staff as originally specified in Standard III. As noted above, this finding was one of the motivating factors behind the development of new indicators for this standard.

A noticeably different situation regarding departments' capacity emerges when reviewing the last two columns of Table 1 for Contextual Information. Only one department reported that it does not collect these data currently.

The final column of Table 1 includes departments that are collecting the information outlined in the key indicator, but would require additional computer programming to report the information as specified. Approximately one third of the departments fall into this category. In some cases, the recoding effort would involve minimal staff time and effort and could be easily accomplished, while in other cases, substantial work would be required for a state to begin to provide the information as specified.

Qualitative Assessments. A distinction can be made between departments with and without integrated, performance-based systems. This survey of correctional departments' abilities to report key indicators found that integrated systems are able to begin participating in a national reporting system relatively quickly and easily. Since the ASCA model identifies what correctional executive administrators view as the most important elements of prison practice, departments with their own performance measurement models generally include the same standards and indicators. There may be definitional differences between the ASCA model and the departments' models, but because these agencies have developed integrated information systems, only minor modifications will be needed to measure the indicators as specified. Based upon the interviews conducted in this assessment, it appears that about 20 to 25 percent of the departments have integrated, performance-based information systems.

For the remaining departments, whose information systems were not fully integrated, participation in a reporting system varies across the indicators. Most systems have relatively well-developed databases for prisoner records. Consequently, they can provide information about the denominators of the key indicators—prisoner population on a particular day, segments of the prisoner population on a particular day (i.e., male prisoners, female prisoners), and prisoner population housed in secure facilities on a particular day. Departments can also provide information about the characteristics of the prisoner population, or the information detailed in the contextual information key indicators.

To "level the playing field" across correctional systems, the counting rules specify that unified systems, departments that house non-convicted individuals and misdemeanants, should only include data for felons serving one year or longer. (Non-unified systems should include data on all convicted felons.) Many of the unified systems report that segmenting the population according to this criterion would be complicated and problematic. Many offenders have multiple and complex sentences that make such computation extremely difficult if not impossible. Consequently, many unified systems are not able to provide data even for the prisoner population and characteristics indicators.

Departments with non-integrated data systems face the greatest difficulty with the substantive areas specified in Standards II and III, institutional safety and substance abuse and mental health. Almost all departments keep track of the numbers of prisoner and staff homicides, prisoner suicides and results of drug testing. Beyond these measures, departments with non-integrated data systems will struggle. Most departments maintain incident-based records of significant events. These databases cannot be queried to produce information as it has been specified regarding prisoner-on-prisoner and prisoner-on-staff assaults, sexual assaults, or disturbances. Their automated records lack the necessary detail to respond to the definitions and counting rules of the ASCA key indicators. Beginning to collect and record the information would require most departments to make significant modifications in how they take information from incident reports. Data collection instruments and/or computer screens would need to be redesigned. Operational staff would have to be trained to record the new information. And, databases would have to be reformed to include new information fields. For most departments, these changes would be substantial.

Any of the indicators involving information about staff—prisoner-on-staff assaults, sexual assaults of staff and staff sexual misconduct—would also pose a significant challenge for most departments. Among the departments with operational data systems, staff records tend to be maintained by human resources. The information is not available in a form that allows for the specifications regarding staff victimization and behavior in the ASCA model. A department may be able to determine from its prisoner records that a staff member was victimized, but often will not be able to identify the gender of the staff member.

As described in the quantitative section above, the original indicators developed for the substance abuse and mental health standard proved problematic and were re-specified. Even with the changes, information regarding substance abuse and mental health treatment tends to be recorded and maintained by the health services unit of the department of corrections. Few departments have automated this information. Furthermore, health records currently used would be extremely difficult to automate as information is kept in traditional hospital-type files (jackets). Most departments would have to design a completely new record-keeping system to produce the information outlined in the ASCA model.

back to top

Conclusions

This survey found that almost all prison systems possess the capacity to measure and report on the aggregate prisoner population under supervision and individual characteristics of that population. Most prison agencies collect information on escapes, homicides and suicides. Beyond these measures, a significant disparity exists among departments regarding their capacity to measure and report on all other indicators.

Departments can be divided into two general groups. Approximately 20 to 25 percent of the Nation's prison agencies have strategic management systems that combine traditional operational databases to create integrated information systems to monitor performance. These departments are positioned to begin participating fully in a national reporting system. They may need to make minor changes in how information is collected and retrieved, as well as writing new computer code to compile information exactly as it is specified by the key indicators and their counting rules.

The one exception to the above conclusion is in the area of substance abuse and mental health measurement. Because the records are collected by health units of prison systems, even the integrated information systems often lack detailed information to respond to the ASCA indicators in this area.

The remaining 75 to 80 percent of the Nation's departments of corrections, beyond the measures identified above, have more limited capacity to measure and report on the remaining ASCA key indicators. Automated disciplinary records are incident-based and lack detail about weapons used, injuries sustained, and victims. Agencies are unable to co-mingle staff data with prisoner information. Departments simply do not record information about disturbances in the detail necessary to count and report incidents as specified. To a large extent, the reason for the lack of capacity is that departments still maintain separate operational databases. For some systems, the only prisoner records are automated; disciplinary, staff and medical records are not automated.

Respondents had an interesting reaction to the survey. Even though they were told that the study was intended to assess capacity to report on the ASCA indicators, many were concerned that they would have to begin reporting in the near future. They were trying to figure out how they could produce the information. This response could pose a potential problem in how data are produced. If departments simply pass responsibility off to middle-level central office staff, they likely will approach the task of data production as a discrete problem to be solved, whereas ASCA's goal is to advance performance-based management. Furthermore, a discrete problem orientation may not produce information about the key indicators that is reliable, valid and adheres to the definitions and counting rules specified. If this were to occur, the new performance monitoring system would suffer the exact same problems as exist with the Corrections Yearbook.

Capacity development involves more than convincing directors to participate in a national performance measurement system. Strategic management includes analyzing current conditions within the organization and its environment, evaluating alternative courses of action, devising strategies for performance improvement, taking risks, being creative and sustaining a continuous process that accumulates experience and redirects practice and decision making in light of future goals. Development of the agency's capacity to utilize a performance measurement system will be an important part of transforming an organization from a rule-based to a results-based management practice. The change from an operations-oriented to a strategic-oriented organization is a lengthy, expensive and staff-intensive effort.

If ASCA's goal is to advance the commitment of adult institutional corrections to performance- based management, it must invest in the appreciation for and commitment to the philosophy and practice. This is much broader than a willingness to participate in a particular reporting system. Directors in general and particularly new directors can be introduced to the topic at ASCA meetings. However, significant advancement will only occur if experienced correctional executives who have led their agencies through such transformations assist officials from other agencies considering and willing to undertake such change.

As an agency undertakes performance-based management, it will necessarily need to integrate its operational information systems and to create a management information system. Most prisoner record systems tend to be adequate, but considerable development work is needed in other areas. Critical incident databases need to be enhanced to include more detail about the perpetrators, victims, weapons, and injuries sustained. Disciplinary information, often not contained in a database, must be coded and linked back to significant events. Staff data must be linked to prisoner data. And finally, health records across systems must be evolved into performance-measures databases. Substantial investments in hardware and software are required to accomplish these tasks. Obtaining or reallocating state resources for these investments may be difficult for some agencies given the current revenue problems faced by most states. Thus, alternate sources of funds are essential. Equally important is the training of staff at all levels within the organization to use performance data and analysis for decision-making. Considerable planning is required if this action is to be successful.

Once an agency is able to produce performance data that are used internally to improve correctional practice, then it is ready to begin participating in a national reporting performance measurement system.

back to top