Developmental Problems in a Regional Data System for Corrections

DEVELOPMENTAL PROBLEMS IN A REGIONAL DATA SYSTEM FOR CORRECTIONS Gordon P. Waldo and Theodore G. Chiricos or researcher and administrator alike, the dearth of valid and reliable data within the system of criminal justice, has long been a source of confusion and irritation. As early as 193 1, the Wickersham Commission proposed a remedy in the form of a “complete body of statistics covering crime, criminal justice and penal treatment”.’ (US. National Commission on Law Observance and Enforcement, 1931: 3). As late as 1967, such a data system did not exist, and the President’s Commission on Law Enforcement and the Administration of Justice was still considering a proposal for a AUTHORS NOTE: The development of the “Southeastern Correctional and Criminological Research Center, ’’ which has organized the regional data system, was supported by Ford Foundation Grant 67-584.

GORDON P. WALDO is the Director of the Southeastern Correctional and Criminological Research Center at the Florida State University. His major areas of interest are in criminology, the sociology of deviance, methodology, and social psychology. He has been appointed by the governor of Florida to the Task Force on Corrections, Probation and Parole of the Florida Interagency Law Enforcement Planning Council. THEODORE G . CHIRICOS is Assistant Director of the Southeastern Correctional and Criminological Research Center at the Florida State University. He is also Assistant Professor of Sociology and Assistant Professor of Criminology at the same university. His areas of research include criminology, juvenile delinquency, deviant behavior, social problems, and minority group relations. Dl51

[2161 CRIMINOLOGY I NOVEMBER 1970

“National Criminal Justice Statistics Center” (Lejins, 1967). While enthusiasm for a national criminal justice data system persists, little has occurred to hasten its realization.’ Little will occur, it seems, until some of the specific details and difficulties attending the implementation of such a system are considered. This paper attempts several tentative steps in that direction by reporting a number of operational problems encountered in the development of a regional data system within one segment of the criminal justice systemadult institutional corrections. In so doing, an effort is made t o anticipate some of the problems that could be expected on a much larger scale should a national criminal justice data system be initiated. In addition, the problems discussed should provide further insight into the shortcomings of existing data systems such as Uniform Crime report^,^ National Prisoner Statistics and the recently developed Uniform Parole Reports ( 1968) and Uniform Probation Reports (1968). While the problems discussed in this paper should be of immediate interest t o planners and administrators of local, regional, or national criminal justice data systems, their implications are much broader. Specifically, these problems should raise a number of questions in the minds of researchers who use data collected in existing systems, or who anticipate the use of data from regional or national criminal justice data systems in the future. The Southeastern Regional Data System (SRDS) represents a cooperative venture in which the states of North Carolina, South Carolina, Georgia, and Florida are collecting uniform information at the level of adult institutional correction^.^ This regional effort represents a direct response t o the recognized need within corrections for “integrated information systems which can form the basis of more accurate social accounting and furnish a foundation for evaluative research” (Joint Commission on Correctional Manpower and Training, 1967: 65).

Waldo and Chiricos / REGIONAL DATA SYSTEM [2171

Many of the operational difficulties encountered in the development of this system are those which would be common to any “horizontal data system.” In this paper the term “horizontal data system” refers to the uniform collection of information at one level of the criminal justice process, across geographic and political boundaries. The FBI’s Uniform Crime Report represents such a system at the level of police activity. National Prisoner Statistics is a horizontal data system in adult institutional corrections which cuts across more geographic boundaries but collects far fewer items than the Southeastern Regional Data System discussed here. Uniform Parole Reports and Uniform Probation Reports are horizontal systems that are in the developmental stage. In addition, it should be noted that a “vertical data system” involves the uniform collection of information at all or several levels of the criminal justice process within a limited geographic and political boundary (i.e., municipal, county, or state government). Such a system allows one t o follow an individual case from initial police contact through court and correctional disposition. Clearly, a national criminal justice data system which includes all elements of the criminal justice process for all local and state political units in the United States would reflect both the vertical and horizontal data system models.

PROBLEMS ENCOUNTERED

Problems encountered in the development of a regional data system for adult institutional corrections may be organized into three broad categories. The first set of problems was generated by organizational or structural differences among the Departments of Correction in the four states. The second set of problems arose from operational difficulties encountered in the pursuit of uniformity in the

[2181 CRIMINOLOGY 1 NOVEMBER 1970

data system. The third group of difficulties stems from cultural variations among the four states in their responses to criminal behavior. It became apparent that. such variation could help t o determine what kind of inmate will compose the data system population. The problems selected for discussion neither exhaust all logical possibilities nor cover all that were encountered. They are, however, representative of the difficulties which confronted the SRDS and may be anticipated in other regional data systems or proposed national data systems. STRUCTURAL AND ORGANIZATIONAL DISCREPANCIES

The four states within the SRDS differ substantially with regard to size of inmate population, number and type of institutions, types of offenders handled, and other organizational characteristics which could effect the data collection process. It should become clear that each of these structural variables bears directly or indirectly upon the others, with all of them having an additive effect upon the data collection process. Of principal importance when considering structural differences is the disparity in jurisdictional authority over various types of offenders which is exercised by the several Departments of Correction. As shown in Table 1 , the Department of Correction in State “A” receives less than one-half of the state’s convicted felons. In that state, the county public work camps-which have no political or legal relationship with the Department of Correction-exercise a “first choice” on felons and receive all misdemeanants. Thus, the proposed horizontal data system was forced to exclude all misdemeanants and approximately half of the felons in State A because of the overlap in state and county authority. By contrast, the Department of Correction in State “B” receives all felons and all misdemeanants with sentences exceeding thirty days. Such ubiquitous jurisdiction contri-

Waldo and Chiricos / REGIONAL D A T A SYSTEM [2191

TABLE 1

STRUCTURAL CHARACTERISTICS OF THE DEPARTMENTSOF CORRECTION IN FOUR STATESa Structural Characteristics Type o f Offender Received From Courts Average End o f Month Inmate Population-1 968 Total Inmate Admissions-1968 Were All Inmates Processed through A Single Receptior Center in 1969? Number of State Correctional Institutions-1 969 Percentage of Inmates Released through Parole-1 965

State A

50% o f felons no misdemeanants

State B

State

C

all felons all misdeall felons meanants someb rnisdewith 30 days meanants

State D mostC felons n o misdemeanants

2,400

9,990

4.857

7,635

1,292

13,005

5,099

3,318

Yes

27

0 38.9

The data for each state refer t o information compiled Only for t h e respective state Departments of Correction and does not include inmates sentenced t o county public work camps or t o municipal, federal, or military prisoners. b. The Department of Correction in State C receives a small number of misdemeanants (900 in 1968) who have been sentenced t o "6-12 months in a State Penitentiary." c. Some felons in State D may be sentenced-by the j u d g e l o county jail or county work camp. At the present time t h i s alternative i s seldom used. a.

butes directly t o State B's larger inmate population and turnover. In effect, the counties in State B have a contract with the Department of Correction for a number of inmates t o maintain the highway system. These inmates remain under the legal control of the Department of Correction. For purposes of uniformity among the four states, misdemeanants from State B were excluded from the data system. In State "C" the Department of Correction handles all felons plus a small number of misdemeanants who receive

[2201 CRIMINOLOGY /NOVEMBER 1970

sentences ranging from “six to twelve months in a state penitentiary.” The Department of Correction in State C has a legal relationship with the county work camps but has no real control over them. The situation in State “D” is very similar to that in State C. The Department of Correction has jurisdiction over most felons but no misdemeanants. One statute in State D permits the sentencing judge to commit certain felons t o a county jail or a county work camp if their sentence is less than one year. Other statutes provide that a sentence for a few specific offenses may be served either in county or state institutions-regardless of sentence length. In practice, however, these judicial options are rarely invoked. Clearly, variations in jurisdictional authority make it difficult for such a horizontal data system to achieve a common population denominator-even when the system is limited to four states. A larger horizontal data system, such as the National Prisoner Statistics, is able t o make even less of a claim to communality in its population base. However, the actualization of a national criminal justice data system could alleviate some of this difficulty. Since a national criminal justice data system would be a vertical, as well as a horizontal data system, it would include county and city correctional institutions. Thus, many of the misdemeanants and other felons who are not part of the population in a strictly horizontal data system (i.e., SRDS) would be included. Because of its vertical nature, such a national system would automatically overcome the kinds of jurisdictional problems discussed here. Problems of jurisdiction are related to the issue of inmate population size. For example, the average end-of-month inmate population under the control of the four Departments in 1968 ranged from 2,400 in State A, to almost 10,000 in State B. Table 1 also shows that the total number of admissions handled by the four Departments in 1968 ranged from 1,292 to 13,005. The importance of size in determining the structure, operation, and personnel allocation of complex organizations

Waldoand Chiricos / REGIONAL DATA SYSTEM [2211

is well known (Anderson and Warkov, 1961; Terrien and Mills, 1955; Mott, 1965: 38-73). Indeed, a fundamental characteristic of social organizations is that the larger the system, and the greater the number of points interacting within that system, the easier it is for “deviance” to occur (Mott, 1965: 189 and 264-273). In this case deviance refers t o unauthorized variations in the data collection process which result either from misunderstanding, carelessness, or purposeful distortion of information. Thus, for purposes of a data system, the more institutions participating, the less certain one may be that the procedures employed and the data being collected are, in fact, uniform and reliable. The ideal of centralized processing of inmates is certainly more difficult t o achieve in states with large inmate populations and rapid inmate turnover. Limitations of institutional capacity and shortage of personnel in “reception and diagnostic centers” may demand-in states with large inmate populations-that inmate processing and data collection occur at more than one point in the state correctional system. Thus, it is not surprising to note that State B, which admitted the most inmates of the four states in 1968, processed (1) no misdemeanants and some felons through one institution; (2) all youthful felons through a second institution; (3) the remaining felons and misdemeanants through some 74 additional institutions. Until recently, only State D was attempting to process all Department of Correction inmates through a single institution. State A has processed all Department of Correction inmates through a central reception center since late 1968. Previously, the state’s reception center handled only youthful offenders, with other inmates being processed by individual institutions. State C opened a reception and diagnostic center in mid-1969 with the intention of receiving all inmates through the institution. Prior to 1969, inmates processed by the Department of Correction were received through sixteen widely

[2221 CRIMINOLOGY I NOVEMBER 1970

dispersed institutions. While State C’s Reception Center is currently functioning, it will not be sufficiently staffed to process all received inmates until 1970. Without uniformity in the processing of inmates, uniformity of data collected at the point of inmate admission cannot be assured. Such assurance is even less likely when differences among fifty states-rather than four-must be accommodated. The discrepancy among the four states in number of institutions under the jurisdiction of the Departments of Correction is also closely tied t o the question of inmate population size and the problems of control over data collection. As seen in Table 1, the number of institutions ranges from a low of 1 1 in State A t o a high of 77 in State B. Clearly, if one attempted an information system that went beyond admissions data it would require an elaborate system of coordination and cooperation among the many individual institutions. The problems of reliability and uniformity would, of course, be greatly compounded by such a system and must not be discounted by proponents of larger data systems. PROBLEMS I N THE DATA COLLECTION PROCESS

Ironically, one of the foremost impediments t o the development of a truly uniform regional data system was the fact that each of the participating states entered the regional system with some existing process for gathering information. From the outset, the four states differed in terms of: (1) which items of information were collected at admission; (2) which format was used for the collection of common items; (3) what type of personnel gathered or verified the data; (4) what attempts were made to verify the collected data; and (5) what type of institution was involved in the datagathering process. Obviously, a major reason for introducing regional uniformity to the data collection process is the desire t o overcome these very discrepancies so that meaningful comparisons across political and geographical boundaries may be

Waldo and Chiricos I REGIONAL DATA SYSTEM [2231

attempted. However, because cooperation and accommodation are crucial to the success of such a regional venture, each of the participants must be convinced that changes made in their existing process of data collection are worth the price of uniformity. The establishment of this conviction was probably the single most important step in the development of the regional system, and it would be a necessary first step in the establishment of any national data system. Clearly, this conviction has not developed in all fifty states even at the level of institutional corrections. Indeed, if one is to judge from the frequency of “unavailable data,” the reporting by states to National Prisoner Statistics is, occasionally, most erratic. On the regional level, a series of conferences was necessary before the four states agreed to a limited core of items and coding formats to be used in the SRDS. This set of 33 items, which more than doubled the number of admission items collected by National Prisoner Statistics, presumed substantial modification of existing data collection processes. Table 2 gives some indication of the magnitude of change required in the state data systems before the uniform regional system could be implemented. The table lists the 33 desired items and indicates simply whether each state was not collecting the item prior to the uniform system, or if it was collecting it in a different format. A necessary change in a state’s existing data system is indicated by an “x” in either column of Table 2. A blank in both columns suggests that the state was not required to make a change in its data collection to accommodate the particular item. It will be noted, for example, that for the 33 items in the uniform system, State A was required to make 24 changes in its existing data system; State B agreed t o make 28 changes; State D, 30; and State C, 31. This frequency of adaptation was necessary even though an attempt was made to keep the changes to a minimum by adopting the coding format of one of the four states whenever possible.

I

P

N

h)

-

1. State reporting 2. Date current sentence begins 3.Primary offense 4. Secondary offense 5. Sentence for primary offense 6. County o f sentencing 7. County o f birth 8. Date o f birth 9. Race 10. Sex 11. Marital status 12. Religious preference 13. Religious participation 14. Education 15. Weekly pay (before arrest) 16. Employment when arrested 17. Main occupation

Item

TABLE 2

X

X

No Prior Collection o f Item

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

Prior Collection in Different Format

X

X

X

X

No Prior Collection of Item

State D

X

X

I

X

X

X

No Prior Collection of Item

Prior Collection in Different Format

State C

X

I

X

X

X

No Prior Collection o f Item

Prior Collection in Different Format

State B

X

X

Prior Collection in Different Format

State A

ADMISSION ITEMS IN SCCRC DATA SYSTEM BY PRIOR COLLECTION STATUS IN FOUR STATES

X

I

x

X

x

x

x

x

x x

2

X

x

x

x

z

x x

I

f

x x x x

x

x

x

x

x x

x

x

x

x x

x

x x x x

x

x

x

x

x x

x

x

x

x x

x

x

x x

x

x

Ix

x

X

X

g

X

x

x x

x

12251

[2261 CRIMINOLOGY I NOVEMBER 1970

It must be understood that modification of an existing data collection system inevitably generates a series of problems for the participating states. These problems are encountered at every administrative level-from budgetary to clerical. In the SRDS the scope of these problems was not fully anticipated by the cooperating states or by the staff of the Research Center. Probably the simplest problem was the composition and printing of new data collection forms which included-at a minimum-33 items of information for the

SRDS. Perhaps the most difficult adjustment for the participating states has been the necessary modification of routine for the personnel collecting admission data. An expected and understandable resistance to these changes had to be overcome. Even the slightest change required the consideration, approval, and cooperation of everyone from the director of a state’s Department of Correction to the clerk or correctional officer immediately responsible for the collection of data. In effect, these personnel required retraining, or at least reorientation to a new set of items, or to old items in new formats. Further, the importance of a uniform data system in general had to be impressed upon clerical staff before they were willing to accept specific coding format changes. As noted earlier, the number and geographic location of institutions collecting data greatly complicated the process of communicating these ideas to widely dispersed personnel. In this regard, it should be emphasized that the level of confidence placed in the reliability and validity of the collected data is directly related to the level of understanding and enthusiasm shown by the data collection personnel. The problem of modifying existing data systems was increased by the fact that in many instances the four states had been collecting the same items but were recording them in dissimilar formats. As shown in Table 3, State A had previously coded an inmate’s race into four categories; State B combined race with sex into ten categories; State C did not

Waldo and Chiricos / REGIONAL DATA SYSTEM [2271

use a computerized records system and therefore did not have any specific coding format; and State D collapsed race with sex into four categories. For this item, the code used by State A was adopted as the SCCRC code, and the remaining states made the necessary coding changes. Another example of coding differences is given by the method of recording an inmate’s occupation. State A coded this information into 9 categories; State D used 87; and State B, 478. The code used by State B was adopted as the SCCRC code for this item. This type of problem was repeated for a number of additional items, and, for each discrepancy in the coding process, changes were required in the coding format and, subsequently, in the frame of reference used by the data collector. In some cases the changes could be made by collapsing categories in a computer program. However, this required that the coding format using the least number of categories be adopted as the SRDS format. To the extent that it was feasible, this was done in an attempt t o reduce the modifications required of the states. Further operational discrepancies among the states were found in: (1) the type of personnel who collect and process the data; and (2) the kind of equipment available for that processing. State A, for example, uses inmates to collect and TABLE 3

PREVIOUS CODING FORMAT FOR RACE IN FOUR STATES State A

State B

State C

State D

1=White P=Negro 3=lndian 9=Other

l=White male 2=Colored male 3=lndian male 4=Oriental male %Other male 6=White female 7=Colored female 8=lndian female 9=Oriental female +Other female

State C was not using a computerired record system and therefore did not have a specific coding format.

1=White male 2=White female 3=Colored male 4=Colored female

[2281 CRIMINOLOGY

I

NOVEMBER 1970

process data, the latter being accomplished with the aid of a “second generation” computer. State B also uses a second generation computer, but it is programmed and operated by free personnel. The data for State B are collected by a variety of free personnel, including custodial, clerical, and professional. State C has no data processing equipment and uses clerical and professional staff to collect data in the reception center. Prior to the existence of that center, custodial and clerical staff handled this task. State D uses professional staff to collect and process data, but it does not have access to a computer at the present time. With such vast differences in the personnel and equipment used in the data collection process, the degree of assurance or confidence that may be placed in the validity and reliability of the data will vary from state to state. Ideally, a uniform data system would employ the same type of personnel and, in a general sense, the same level of equipment in data collection and processing throughout the system. Attempts made to check or verify the collected data also vary from state to state. States A and B make no attempt to verify the information collected a t admission. State C is trying to verify some of the data by writing t o an inmate’s parents, relatives, former school or employer, or any other likely source that may provide answers. However, staffing limitations at the state’s new reception center have cut into the process of verification. State D probably makes the most rigorous attempt at verification-but even here, the care with which this is pursued varies with staff availability. Verification usually suffers when other administrative tasks demand personnel. A national data system must be concerned with the type of personnel used in collecting data and the degree of uniformity attained in the verification process if comparability across states is desired. Still another operational difficulty was met in attempts t o establish a single uniform identification number to be used for all inmates received across the four states. Each state had

Waldo and Chiricos / REGIONAL DATA SYSTEM [2291

been using a different procedure to generate an I.D. number. However, these numbers would not accompany an inmate who was reincarcerated in any one of the other three states. Not wanting to develop a totally new number, the possibility of using either a social security or FBI number was considered. The social security identification was rejected because all inmates did not have a number and, more importantly, some inmates had several. An inmate’s FBI number was agreed upon as the SRDS identification number. However, this mode of identification is not without difficulty. It frequently occurs that the FBI does not have a prior fingerprint record for an offender at the time he is received and fingerprinted by a state’s Department of Correction. Thus, on occasion, when the Department of Correction requests a “rap sheet’’ from the FBI, it receives a form indicating that this inmate has no police or court record and, on that account, a number has not been assigned by the FBI. This is apparently caused by the failure of an arresting agency or the courts to send fingerprints to the FBI at the time of arrest or conviction. In a few cases it may be explained by a delay in the receipt of the prints, so that the request from the Department of Correction is processed pn’or to a number being assigned for an arrest. Whatever the reason, the FBI number provides something less than a totally satisfactory means of identification, and this problem will demand a suitable resolution before a national criminal justice data system is possible. CULTURAL VARIATIONS IN RESPONSE TO CRIME

Cultural variations in the response to crime exist between states as well as within states and may be measured by criminal statutes and correctional policies for the several states. These official responses are termed “cultural” inasmuch as they may be considered to reflect the unique character of a state’s concern for the problem of crime. The more severe the statutes and policies, the more one may

[231 CRIMINOLOGY I NOVEMBER 1970

assume a state’s determination to reduce crime through incarceration. Such response patterns could help determine the kinds of people who are brought into the correctional system, the manner in which they are handled, and the length of their imprisonment-all of which could affect a correctional data system. Several indices of variable official response are given by: (1) statutory provisions for criminal sentences, (2) actual sentencing practices, (3) correctional implementation of judicial sentences, and (4) the differential use of probation and parole. Operationalizing these indices, it is possible to compare the states with regard to minimum and maximum sentences provided for specific crimes under law (Table 4); median sentence received for specific offenses (Table 5 ) ; median time served for these same offenses (Table 5); and the percentage of a state’s inmate population released on parole (Table 1). Table 4 gives the minimum and maximum sentences prescribed by law for a selected group of crimes. The man who commits a forcible rape in State C may get a minimum sentence as low as one year; the same man committing the same offense in State B must receive at least a life sentence! Aggravated assault, it will be noted, brings a maximum sentence ranging from one to ten years within the four states of the data system. These types of differences would necessarily be multiplied in any horizontal data system of larger scope, and would greatly complicate interstate comparisons of inmate populations. Perhaps of greater importance to a data system in corrections are the formal and informal norms that affect the judicial and correctional implementation of statutory provisions. As shown in Table 5 , the median sentence received for specified offenses varies widely from state to state. For example, it would appear that sentences received in State B are generally less severe than in the other states-except with regard to sex offenses. At the same time, however, it seems

TABLE 4

Max: death Min: 5 years

Death; or life if jury recommends mercy

Rape

Kidnapping

Fine at discretion o f court or 1-7 years

Max: 2 years Min: 6 months

Making, issuing worthless checks $100 or more

Escape (felon 1st)

~

Max: life Min: 5 years

Burglary

~~~~~~~~

Max: 1 year Min: up t o $200 fine or 3 mos. or both

Aggravated Assault

State A

~~~~

Max: 2 years Min: 6 months

in jail

$50 fine, 30 days

Life

Max: death Min: life

Max: death Min: life

Max: 10 years Min: 4 months

State B

~~

~~~~~~~~~~~

Max: 10 years Min: 1 year

Max: 5 years Min: 1 year

~~

Max: 5 years Min: 12 months or less-max. fine $1000

~

10 years or less, except for ransom, then death

Max: death Min: term of years

Max: life Min: 1 year

Max: 5 years Min: up to 1 year or up to $1000 fine or both

State D

Max: misdemeanor

~~~~~~~~~

Max: 20 years Min: 1 year-except when for ransomthen life

Max: death Min: 1 year

Max: 20 years Min: 1 year

Max: 10 years Min: 1 year

State C

MINIMUM AND MAXIMUM SENTENCES PROVIDED BY LAW FOR SELECTED CRIMES IN FOUR STATES

12321 CRIMINOLOGY / NOVEMBER 1970

that formal and informal implementation of these sentences by the correctional system irregularly reduces these interstate discrepancies. It will be noted with regard to burglary, that the maximum discrepancy among the four states in actual time served is 3.2 months. For other offenses these maximum differences are greater, reaching a high of 53.5 months for sex offenses. One additional cultural variation that relates to the development of a horizontal data system in corrections is the differential use of probation and parole within the four states.’ Although most correctional authorities view the expanded use of probation and parole as a favorable development in the treatment of offenders, a latent function of this expansion has been a gradual change in the characteristics of institutional populations. The increased use of probation and parole not only tends to decrease the prison population, but it decreases it in a selective manner. For example, since probation and parole tend t o select the “best risk cases,” the incarcerated inmate is much more likely to be a “hard core case.” If one state has a more lenient policy concerning probation or parole for a particular type of crime, this leniency is likely to be reflected in the proportion of the state’s institutional population serving time for that offense. Figures on the differential use of probation in the four states are not readily available, but figures for parole can be obtained from National Prisoner Statistics (1964: 5 1). As shown in Table 1 , the use of parole in 1965 was six times as frequent in State C (56.5%) as in State A (9.4%). States B and D were approximately equal in the use of parole, releasing 37.6% and 38.9% respectively. It seems prudent t o predict that the incarcerated populations of the four states, particularly A and C, will differ at any time simply as a result of the differential use of parole. These discrepancies in statutory provisions, sentencing practices, institutional policies, and the use of probation and parole do not impede the development of a horizontal data

40.3 73.1 98.6 32.3 21.6

-

109.2

Med Sen

44.7 8.7 17.7 39.2 21 .o 21.2 16.0

Med Time

State A

240.0 18.1 24.9 56.0 170.0 23.5 19.7

Med Sen Rcvd 1960

49.6 17.1 24.8 32.5 43.7 18.0 16.0

Med Time Served 1964

State B

~

Life 47 .O 46.8 90.0 120.0 55.8 42.0

~~~

46.9 15.4 16.4 29.7 35.6 20.9 16.7

Med Time Served 1964

State C Med Sen Rcvd 1960

~~~~

240.0 35.7 40.3 74.2 86.8 42.9 35.5

b. National Prisoner Statistics, State Prisoners: Admissions and Releases, 1964 (Washington, U.S. Government Printing Office), p. 52.

59.6 19.8 22.7 33.9 34.8 23.7 17.7

1964

Served

Med Time

State D Med Sen Rcvd 1960

~~~

a. National Prisoner Statistics, Characteristics of State Prisoners, 1 9 6 0 (Washington, U.S. Government Printing Office), p. 50.

Larceny

Homicide Auto Theft Assaults Robbery Sex Offenses Burglary

item

TABLE 5

MEDIAN TIME RECEIVED AND MEDIAN TIME SERVED FOR SELECTED CRIMES IN FOUR STATES

[2341 CRIMINOLOGY I NOVEMBER 1970

system, but they could add to the problem of interpreting empirical differences between the states with regard t o inmate characteristics. For example, could one expect the inmate population of State A t o be “underrepresented” in the category of “auto theft” simply because convicted auto thieves in that state tend to spend less time in prison? Conversely, might it be anticipated that convicted murderers would be “overrepresented” in State D at any one point in time because convicted murderers in that state spend more time in prison? At the outset, one cannot specify precisely how these variations will affect the data, but as a minimum precaution, these interstate differences should be acknowledged with the intention of analyzing the data in light of them. Unfortunately, these problems of interpretation will likely be compounded in any data system which includes additional states and additional points within the criminal justice system.

CONCLUSIONS

The implementation of the Southeastern Regional Data System has raised a variety of operational and interpretive problems-most of which could be expected in any attempt to develop a limited horizontal data system. Extrapolating the difficulties encountered by the SRDS to larger horizontal systems such as National Prisoner Statistics, Uniform Crime Reports, Uniform Parole Reports or Uniform Probation Reports, raises important questions concerning the reliability and “uniformity” of those systems. Indeed, almost all of the operational problems experienced by the SRDS also beset National Prisoner Statistics. For National Prisoner Statistics, however, these problems loom much larger because that system encompasses twelve times as many states. A projection of these kinds of difficulties t o other areas of the criminal justice system prompts a serious concern for the

Waldo and Chiricos / REGIONAL DATA SYSTEM [2351

immediate feasibility of a proposed national criminal justice data system. At a minimum, the experiences of the SRDS suggest that advocates of a national data system: (1) should not overestimate the speed with which diverse agencies within one state and across fifty states can be brought t o accept and implement the idea of a national uniform data system and the desirability of specific items of information; (2) should not underestimate the delicacy of the political effort required to insure that these diverse elements can be integrated in such a system; (3) should not minimize the cost in money and field-based manpower that will be required to realize the goals of uniform, valid, and reZiabZe data collection and reporting; (4) must not ignore the serious problems of interpreting data that are generated by a criminal justice system which encompasses enormously diverse definitions of what criminals are, and how they should be handled. I t is hoped that further insights into the problems of operationalizing data systems will be provided by those implementing horizontal data systems in other areas of the criminal justice process. In addition, those working with vertical data systems (i.e., SEARCH) will certainly add to this knowledge as it relates to the police, courts, jails, and noninstitutional corrections. Without this kind of information, we are not likely to know the most feasible approach to developing a national criminal justice data system. Although a national data system for the entire criminal justice process is desirable and likely to be attempted within the next decade, it may be more appropriate, operationally, to pursue limited regional data systems that could be combined nationally at a later date. The limited regional systems should be able to resolve the kinds of difficulties reported in this paper before a meaningful national system is sought. The ultimate goal will be more readily achieved if the several regional attempts employed identical models, objectives, and procedures for data collection. It is at this level-the mapping out of objectives and procedures-that

[2361 CRIMINOLOGY I NOVEMBER 1970

national rather than regional coordination seems desirable. Such coordination would allow the resolution of operational problems at a reduced (i.e., regional) level of complexity, while insuring the final compatibility of regional elements for a national data system. Despite the multitude of difficulties cited, it is encouraging to note that a regional data system for adult institutional corrections has begun t o function in the Southeast. Further encouragement is found in the fact that a latent function of the process of planning and developing a regional information system has been the generation of an unprecedented amount of interaction, communication, and cooperation among the participating Departments of Correction. This interaction has, itself, given rise t o several specific research projects which are regional in scope. It is hoped that this spirit of cooperation can be transferred to other elements of the criminal justice system in the Southeast and to other attempts at regional data systems. NOTES 1. For several discussions of criminal statistics see: Sellin and Wolfgang (1964);Cressey(1957); Beattie(1959);andChilton(1968). 2. One notable exception is the recently funded “Project SEARCH” (System for Electronic Analysis and Retrieval of Criminal Histories). The fourteen-month project is expected to demonstrate the value and feasibility of a “national statistics service” and the “instant retrieval” of criminal history data by police, courts, and correctional systems. States involved in the project are: Arizona, California, Maryland, Michigan, Minnesota, and New York. California was named the coordinator state of the project which, if successful, could provide the basis for a national criminal justice statistics and information center planned by the Law Enforcement Assistance Administration (LEAA). 3. For an analysis of the Uniform Crime Reports see: Lejins (1966). 4. With the Director of the Department of Correction in each participating state on its Board of Directors, the Southeastern Correctional and Criminological Research Center (SCCRC) was established at The Florida State University. The purpose of the Center is to bring a regional focus to research and administrative problems shared by institutional corrections in North Carolina, South Carolina, Georgia, and Florida. The original impetus for the SCCRC came from an informal meeting between the Directors of the several state Departments of Correction. The Ford Foundation agreed t o fund the SCCRC for three years. The Center’s

Waldo and Chiricos / REGIONAL DATA SYSTEM 12371

staff and the four cooperating states are developing a uniform data collection system for adult prisoners which begins with a minimum of 33 items of information collected for each inmate at the point of admission to the correctional system. This information is to be collected in the same format in each of the four states and channeled to the SCCRC. 5. The Probation and Parole Departments are separate legal entities from the Department of Correction in all four states. However, the precise relationship between these agencies differs in each state.

RE FER ENCES ANDERSON, T. R. and S. WARKOV (1961) “Organizational size and functional complexity: a study of administration in hospitals.” Amer. SOC. Rev. 26 (February): 23-28. BEATTIE, R. (1959) “Sources of statistics on crime and corrections.” J. of the Amer. Statistical Assn. (September): 582-592. CHILTON, R. J. (1968) “Persistent problems of crime statistics,” pp. 89-95 in S. Dinitz and W. C. Reckless (eds.) Critical Issues in the Study of Crime: A Book of Readings. Boston: Little, Brown. CRESSEY, D. R. (1957) “The state of criminal statistics.” National Probation and Parole Assn. J. (July): 230-241. Federal Bureau of Investigation (n.d.) Crime in the United States, Uniform Crime Reports. Washington, D.C.: Government Printing Office. Federal Bureau of Prisons (1964) National Prisoner Statistics, State Prisoners: Admissions and Releases. Washington, D.C.: Government Printing Office. GOTTFREDSON, D. M., K. B. BALLARD, Jr., P. S. VENEZIA, and E. A. WENK (1968) Uniform Parole Reporting: One Year of Experience. Davis, California: National Council on Crime and Delinquency Research Center. Joint Commission on Correctional Manpower and Training (1967) Research in Correctional Rehabilitation. Washington, D.C.: Government Printing Office. LEJINS, P. (1967) “National crime data reporting system: proposal for a model.” Appendix C of President’s Commission,Task Force Report: Crime and its Impact-An Assessment. Washington, D.C.: Government Printing Office: 178-206. _ _ _ (1966) “Uniform crime reports.” Michigan Law Rev. (April): 1001-1030. MOTT, P. E. (1965) The Organization of Society. Englewood Cliffs, New Jersey: Prentice-Hall. SELLIN, T. and M. E. WOLFGANG (1964) The Measurement of Delinquency. New York: John Wiley. TERRIEN, F. W. and D. L. MILLS (1955) ‘The effects of changing size upon the internal structure of organizations. Amer. SOC.Rev. 20 (February): 11-13. United States National Commission on Law Observance and Enforcement (1931) Report on Criminal Justice. Washington, D.C.: Government Printing Office. VENEZIA, P. S. and A. W. COHN (1968) Uniform Probation Reports: A Feasibility Study. Probation Management Institutes, National Council on Crime and Delinquency.