Executive Summary ( 96-01)
January 12, 1996
Like many other parts of the nation, Minnesota has experienced increasing rates of violent crime in recent years. But, compared with other states, Minnesota has tended to rely less on prison and more on community correctional services to respond to crime problems. This is one reason that Minnesota had more adults on probation in 1994 per 1,000 population than all but four states. As a result, probation services play an important role in Minnesota's criminal justice system.
The 1995 Legislature asked the Legislative Auditor's Office to study a particular aspect of probation services. Specifically, we were asked to conduct "a weighted workload study. . . based on uniform workload standards and level of risk of individual offenders," and to recommend a method of probation funding that could be implemented in fiscal year 1997. (Minn. Laws (1995), Ch. 226, Art. 1, Sec. 11, Subd. 3.) In subsequent discussions with legislators, including members of the Legislative Audit Commission who attended a roundtable discussion on this topic, we learned of broader concerns which caused us to modify the focus of our study somewhat. Therefore, our evaluation asked:
To what extent are there regional variations in the number of persons on probation, the length of probation, and the services provided by probation offices?
To what extent do probation agencies measure the outcomes of their services?
To conduct our study, we surveyed all state and local agencies that administer probation services in Minnesota. In addition, we visited 13 of these agencies and spoke by phone with staff of most of the others. We also spoke with national experts about offender classification, and we contacted officials in all 50 states to discuss their classification practices.
We learned that there are currently no uniform, statewide methods of categorizing offenders by risk in Minnesota, and it would take a significant amount of time for service providers to implement them. As a result, we concluded that a "weighted workload study. . . based on. . . the risk of individual offenders" cannot be done at this time. Instead, in this report we provide the Legislature with (1) options for allocating state probation funds that could be implemented in fiscal year 1997 but which are not based on individual risk assessments, and (2) suggestions for how to implement a statewide method of offender risk assessment, if the Legislature believes that this should be a longer-term goal.
Minnesota has a complicated system for providing probation services, with a mix of state and local service providers. In fact, a recent national review of probation practices singled out Minnesota and three other states as having "particularly complex combinations of responsibility for probation services." (LIS, Incorporated, State and Local Probation Systems in the United States: A Survey of Current Practice (Washington, D.C.: National Institute of Corrections, July 1993), 7.) In the 31 counties that participate in Minnesota's Community Corrections Act (CCA), county staff provide all juvenile and adult probation services. These counties have 16 separate administrative agencies, and they account for 71 percent of Minnesota's population and include most of the Twin Cities metropolitan area. In 24 other counties, the Department of Corrections provides all probation services. In the remaining 32 counties, there are 25 separate county probation agencies that provide probation services in combination with the Department of Corrections, but the division of responsibilities is not set forth clearly in state law.
Probation services are funded primarily by a combination of state and county revenues. First, counties that participate in the Community Corrections Act are eligible for state block grants, which can be used for probation and other correctional purposes. These grants pay for about 19 percent of correctional expenditures in CCA counties statewide, although this ranges from 12 percent in Hennepin County to 79 percent in Rock/Nobles counties. (Probation agencies representing 9 of the 31 CCA counties received more than 50 percent of their funds from state CCA grants in 1994.) County property taxes pay for most of the correctional expenditures in these counties that are not funded by state CCA grants. Second, the state reimburses half of county expenditures for the salaries of adult misdemeanant and juvenile probation officers in non-CCA counties. Third, the Department of Corrections receives a state appropriation for the full cost of providing probation services to adult felons in non-CCA counties. Finally, the 1995 Minnesota Legislature appropriated $14.5 million for the 1996-97 biennium to help service providers throughout Minnesota reduce their probation caseloads.
Service providers spent an estimated $20.5 million in state funds for probation in 1995. (We estimated that service providers spent at least $34 million in county funds for probation in 1995, although we did not have complete 1995 data on county spending.) As we examined state spending for probation in individual counties, we computed expenditures per capita to adjust for differences in county populations. There was considerable variation in per capita spending in 1995 among counties, but CCA counties--as a group-- spent significantly less in state funds per capita than non-CCA counties as a group ($3.59 compared with $6.83). In our view, the spending difference largely reflected the varying methods used to provide state funding to CCA and non-CCA counties. For example, probation services for adult felons in non-CCA counties are entirely funded by the state, but these services are paid for with a combination of state and county funds in CCA counties. (In non-CCA counties, as a group, state funds paid for more than two-thirds of total probation spending in 1995. In CCA counties, as a group, state funds paid for less than one-fourth of total probation spending.)
Each year, probation agencies provide the Minnesota Department of Corrections with information on the number of offenders they are supervising. Agencies reported that on December 31, 1994 they supervised 97,318 persons on court-ordered probation, plus another 1,929 offenders on parole or supervised release from prison. Minnesota had 24.6 adults on probation per 1,000 adult residents in 1994, compared with a national rate of 15.4 adult probationers per 1,000 adult residents. Minnesota's rate was exceeded only by those of Texas (30.2), Delaware (29.2), Washington (28.3), and Georgia (27.3). Most of Minnesota's growth in probation over the past decade reflected growth in probation for adult misdemeanants and gross misdemeanants, with lower rates of growth in the number of adult felons and juveniles on probation.
In addition, we found that:
The number of adults on probation per 1,000 population ranged from 2.9 in Kittson County to 52.4 in Pine County. Likewise, the number of juveniles on probation per 1,000 persons aged 12 to 17 ranged from 4.9 in Sibley County to 127.7 in Meeker County.
To a considerable degree, such variation reflects differences in the willingness of Minnesota courts to use probation, especially for adult misdemeanants and juveniles. For example, probation officials in northwestern Minnesota told us their courts prefer to place relatively few adults on probation for misdemeanors--partly reflecting the courts' desire to make the most effective possible use of their limited probation resources. In contrast, two counties elsewhere in Minnesota had more than 30 misdemeanants on probation per 1,000 population in 1994. Likewise, 22 counties had no more than one juvenile on probation for "status offenses" in 1994, while six had more than 20 such offenders per 1,000 population aged 12 to 17. One county had no juvenile property offenders on probation, while another had 67 such offenders per 1,000 population aged 12 to 17.
We considered whether variation in the use of probation might reflect differing crime rates. However, for both adults and juveniles, we found relatively weak relationships between counties' number of arrests per 1,000 population and their number of probationers per 1,000 population. We did find a strong relationship between counties' number of felons convicted per 1,000 population in 1993 and their total number of felons on probation per 1,000 population at the end of 1993.
Legislators asked our office to conduct a study of probation services partly because they wanted to know more about the components of probation in Minnesota. We found that:
We learned that agencies vary in the following ways:
The frequency with which they meet with offenders. For example, 3 of 50 service providers estimated that they meet every month with all offenders that have been placed on probation for non-felony crimes against persons; in contrast, 10 service providers told us that they meet monthly with 25 percent or fewer of these types of offenders.
The nature of their contacts with offenders. Ten of 47 probation agencies that serve juveniles told us that at least 80 percent of their meetings with juveniles occur at the probation office. In contrast, nine probation agencies have more than half of their meetings with juveniles at schools or homes, which likely requires more time spent traveling.
In addition, all service providers conduct "one-on-one" meetings with certain adult and juvenile offenders, but some providers use alternative approaches for those considered low risks for reoffending. For instance, 12 of Minnesota's 50 service providers conduct periodic group meetings as a way of maintaining or increasing staff contacts with offenders, and one service provider requires several hundred offenders to contact the probation office periodically through an electronic kiosk.
The number of presentence investigations (PSIs) ordered by the courts and the time devoted to each. For example, 3 of the 25 agencies that serve felony offenders told us that their courts ask for PSIs on no more than half of their convicted felons, despite the fact that these investigations are required by law. According to probation officials, the number of PSIs conducted and the thoroughness with which they are prepared largely reflect the expectations of the courts they serve.
The duties for which probation officers are responsible. In some counties, probation agencies use their own staff to provide counseling services, diversion programs, drug testing, and mediation services in family disputes; in other counties, these services might be provided by private or other public agencies. In some counties, probation officers collect payments of fines and restitution from offenders; in other counties these tasks are handled by court administrative staff. Some service providers use volunteers to assist with supervision, investigations, or administrative duties, but most do not.
By highlighting service variations, we do not necessarily mean to suggest that probation services should be uniform throughout Minnesota. Many variations exist because probation officials have tailored their services to meet the expectations of their courts and the communities they serve. Service variations may also result from differences in probation agencies' workloads and their ability to pay for services. But we think it is important for legislators to recognize that variations in probation services complicate the task of developing a uniform funding formula that accurately reflects the time that agencies spend working with offenders. For example, even if a study were to document the statewide average amount of time per month that service providers devote to a certain category of offenders, it is likely that many service providers would vary from this average considerably.
Probation agencies differ not only in the services they provide but also in the way they determine which offenders should receive particular levels of supervision. Corrections experts usually recommend that probation offices use formal methods of classifying offenders to determine those who need high levels of supervision and those who do not. We found that service providers in nearly all Minnesota counties have formal methods for classifying adult felons on probation, but such methods are not used as often for adult misdemeanants and juveniles. For example, of the 32 counties in which probation responsibilities are split between county and state employees, county staff in only 6 classify adult misdemeanants in a systematic way. (In contrast, the Department of Corrections classifies misdemeanants in a standardized way in all 24 counties where it provides all probation services, and staff in most CCA counties use formal methods to classify misdemeanants.)
Of those Minnesota agencies that use formal classification practices, all but one assess offenders based on their risk of reoffending. (Several probation offices supplement their risk assessments by using general "needs" assessments to identify offenders for special types of services or interventions. Also, one large county classifies offenders based on number of prior offenses and the severity of the most recent offense.) However,
This contrasts with the practices of 42 states that have statewide classification systems for their adults on probation and 16 states that have statewide classification systems for juveniles. Unlike Minnesota, most of the states with uniform, statewide classification approaches have probation systems in which services for all offenders are provided by state employees.
We also found that only one service provider in Minnesota has "validated" its risk classification instrument in the past five years--that is, examined whether offenders classified as "high risk" do, in fact, reoffend at higher rates than other offenders. Without validating risk assessment instruments, probation agencies cannot be certain that they are managing offender risks effectively. The lack of validation studies reflects the fact that:
For example, probation offices usually do not determine the recidivism rates of the offenders they supervise, nor do most determine the total percentage of restitution owed to crime victims that has been collected from offenders.
In addition to asking for "a weighted workload study. . . based on uniform workload standards and level of risk of individual offenders," the Legislature asked us to recommend a method for allocating probation funds in fiscal year 1997 that would use these standards and risk assessments. Our study found that Minnesota service providers differ in the methods they use to classify offenders; there is not consensus on the "best" method of classification. If the state adopted a single, uniform method of offender classification, it would take a significant amount of time for service providers to implement the system and collect data on the risk of individual offenders that could be used as a basis for fund allocations.
Thus, while our report identifies the steps required to implement a funding system based on a uniform method of risk assessment (which we call Option 1), it cannot be implemented in fiscal year 1997. In fact, if the Legislature believes that Option 1 should be implemented, we think that fiscal year 1999 would be the earliest possible time that it could be used to make funding allocations. (Given the time required to adopt and implement a uniform instrument, it would be at least calendar year 1998 before service providers could produce information on one year's worth of offenders who entered probation. It would be preferable to have at least two years of data to use for purposes of fund allocation.)
Consequently, we evaluated other options that the Legislature could use to allocate probation funds, starting this year. For example, to address the Legislature's stated interest in a funding approach based on uniform workload standards, we examined two funding options (Options 2A and 2B) that would incorporate workload standards recommended in 1994 by Minnesota probation officials.
Option 2A would use "maximum caseload" standards recommended by the legislatively-established Probation Standards Task Force to weight data on probation caseloads reported annually by service providers to the Department of Corrections. Service providers would not be guaranteed a level of state funding that would enable these standards to be achieved, and service providers would not be required to comply with the standards. But these standards provide a basis for estimating the relative amount of time and effort that various types of offenders should require, in the judgment of the task force.
Some service providers have expressed concerns about the reliability of the probation caseload data collected by the department, which would be used in Option 2A to compute funding allocations. We share these concerns, especially due to differences throughout Minnesota in the proportions of adult misdemeanants and juveniles placed on probation and in the types of probation services provided to them. In addition, allocating funds based on actual probation caseloads might create incentives for offenders to be placed or kept on probation, perhaps in cases where this would not be appropriate.
For these reasons, we examined another approach (Option 2B) that would use a second set of "maximum caseload" standards recommended by the Probation Standards Task Force. By applying these standards to the number of persons entering probation, rather than the number on probation at a given time, this approach might at least reduce financial incentives to keep offenders on probation. Like Option 2A, this option relies on workload standards that seem to have considerable support among Minnesota probation officials, but which may need further scrutiny.
Finally, Option 3 would allocate probation funds statewide using the Community Corrections Act formula, presumably with revisions recommended by a legislative work group in 1995. (FIt would be possible to allocate new probation funds using the existing CCA formula, as the Legislature mandated for fiscal year 1996. But the 1995 Legislature created a work group to recommend a "new formula that is more fair and equitable" (Minn. Laws (1995), Ch. 226, Art. 1, Sec. 11, Subd. 3), apparently reflecting legislative concern about the adequacy of the existing CCA formula.) This approach would allocate funds based on measures of arrests, court cases, convictions, population, and county ability to pay--not based on the number of persons on probation. Option 3 would eliminate incentives to place or keep people on probation, but its measures (and their weights) would not necessarily reflect the workloads of probation agencies or the risks of the offenders they serve. For example, the measures of county arrests and court cases in the proposed formula are not closely related to the rates of probation use in Minnesota counties.
Our report discusses each option in detail, and we offer general guidelines for selecting one. However, we think the choice of an option will depend on legislators' goals for a funding formula and their interpretations of the options' fiscal impacts. None of these options would allocate funds in a way that perfectly reflects the risks of offenders or the workloads of individual probation agencies, but we think they represent plausible funding approaches that the Legislature should consider.
Table 1 shows each option's potential fiscal impacts on Minnesota's three categories of service providers. If used to allocate caseload reduction funds appropriated by the 1995 Legislature, Options 2A and 3 would result in CCA counties, as a group, receiving a percentage of new funding that would be slightly larger than their proportion of the state's 1995 population. All of the options would provide CCA counties with a percentage of new funding substantially larger than their estimated share of 1995 state-funded probation expenditures in Minnesota. The counties in which the Department of Corrections provides all services would, as a group, receive a percentage of state funding under Options 2A and 2B that is less than their percentage of the state's population; under Option 3, they would receive a percentage of funding larger than their percentage of the population. As a group, the counties in which county and state employees both provide probation services would receive a percentage of funding under Options 2A and 2B that is larger than their share of the state's population; under Option 3, they would receive a smaller share of funding.
If these options do not adequately address the Legislature's goals for probation funding, they can be modified. For example, if legislators believe that it is particularly important to use state funds to reduce caseloads for persons convicted of felonies or domestic assaults, the "weights" assigned to these offenders could be increased. Or, legislators may wish to incorporate a measure of county ability to pay into Option 2A or 2B.
Regardless of the funding options selected, we think there are several actions that the Legislature should take to improve accountability for Minnesota probation services. We recommend:
First, for purposes of allocating new state probation funds, the Legislature should clarify whether the Department of Corrections should approve the spending plans of service providers in addition to allocating funds to counties in accordance with the adopted funding formula. To receive fiscal year 1996 caseload reduction funding, service providers were required by the department to show that their planned expenditures met criteria set forth in law. If the 1996 Legislature would like to continue this state review before new funds are allocated, it should explicitly require this procedure in law. Second, the Legislature should consider how decisions about fund allocations should be made in those counties where probation services are provided both by county and Department of Corrections employees. Because responsibility for probation services is split in these counties, the allocation of a county's caseload reduction funds between the providers could be the subject of disagreements.
Although not required for the purpose of allocating future caseload reduction funds, we think the Legislature should also consider clarifying in law which service providers are responsible for various categories of offenders. The law does not indicate which service providers should supervise gross misdemeanants in counties where both state and county employees provide probation services, and practices vary throughout Minnesota. Also, although the Department of Corrections is the sole provider of probation services to adult felons in non-CCA counties and receives a state appropriation for this purpose, the law does not appear to assign exclusive responsibility to the department for this population. Finally, although county staff in all non-CCA counties provide services to adult misdemeanants, their authorization for this responsibility is in the state's juvenile code (Minn. Stat. §260), and this is a source of possible confusion.
In addition, we think that it is important to track the outcomes of community correctional services to (1) help decision makers evaluate past funding choices and make new ones, and (2) enable service providers to validate risk-based classification instruments. We recommend:
The Legislature should direct the Department of Corrections to establish an outcome measurement task force to recommend by January 1997 statewide probation outcome measures, along with procedures for collecting outcome data. These recommendations should be subsequently reviewed by Minnesota's Criminal and Juvenile Justice Information Policy Group. The Legislature should require service providers to periodically report information to the Department of Corrections about the reoffending rates of adult offenders.
Even if the Legislature decides not to allocate state probation funds using a uniform, statewide classification instrument, we recommend that:
As we have noted, there is very limited statewide information available on Minnesota probationers, and the accuracy of this information has been questioned. We recommend that:
A uniform set of offender information could help service providers to more readily obtain information on offenders who have been on probation in other counties, and it could also be used to construct or validate classification instruments and produce statewide caseload information. Our report suggests types of information that the Department of Corrections may wish to require.
Finally, we recommend that:
If the Legislature clarifies the state's goals for probation services, we think that policy makers and service providers will have a stronger basis for selecting appropriate outcome measures and workload standards.
The Program Evaluation Division was directed by the Legislative Audit Commission to conduct this evaluation in June 1995.
For a copy of the full report, entitled "Funding for Probation Services," (96-01),136 pp., published on January 12, 1996, you may use our order form (request report #96-01). Alternatively, please call 651/296-4708, e-mail our office at Legislative.Auditor@state.mn.us, or write to Office of the Legislative Auditor, 658 Cedar St., St. Paul, MN 55155.
Staff who worked on this project were Joel Alter (project manager), Jan Sandberg, and Michael Blumfield.