STATE OF NEW JERSEY, PLAINTIFF,
v.
MICHAEL WHITEHEAD, DEFENDANT.
Superior Court of New Jersey, Law Division.
Decided May 10, 1978.*434 Mr. Leonard D. Ronco, Special Deputy Attorney in Charge, attorney for plaintiff, by Elaine Zamula, Special Deputy Attorney General.
Mr. Thomas Menchin, Deputy Public Defender, attorney for defendant, by Messrs. E. Carl Broege, and Michael J. *435 Marucci, Assistant Deputy Public Defenders; (Mr. Ezra D. Rosenberg, Assistant Deputy Public Defender, on the brief).
YANOFF, J.C.C. (temporarily assigned).
This opinion is written pursuant to R. 2:5-1(b) as an amplification of remarks made at sentencing on July 22, 1977 on an appeal from sentence in a homicide case. The trial resulted in a conviction of second-degree murder. The sentence imposed was a minimum of 10 years and a maximum of 20 years at New Jersey State Prison. Defendant is 19 years of age. As an adult this was his first offense. However, he had a substantial record as a juvenile, consisting of sustained petitions as to arson, assault and battery and robbery which I could take into account in assessing defendant's personality. State v. Ebron, 122 N.J. Super. 552 (App. Div. 1973), certif. den. 63 N.J. 250 (1973). He was under probationary supervision at the time of the offense. In the reasons given for imposing such a sentence pursuant to R. 3:21-4(e), I said:
I agree with the State that this was an heinous act, the reason for which we do not know and that a substantial term is required by way of punishment. It is also required as a general deterrent in the group of which the defendant is part of. [sic] It was clear during the trial that there was a conflict between two groups of families. I am convinced this sentence will become known to them and it will act as a deterrent. The sentence is below the maximum because of the defendant's youth. This is a case where the object of rehabilitation must yield to the necessity of punishment and deterrent [sic].
Were there no more, there would be no need for an opinion; an appellate court could exercise its judgment on consideration of the presentence report. R. 2:3-2; 2:10-3.
The fundamental consideration is the correctness of the sentence. However, at sentencing I used as an aid for the first time in this state "Sentencing Guidelines" which had been formulated for Essex County under a grant of the National Institute of Law Enforcement and Criminal Justice of the Law Enforcement Assistance Administration *436 (LEAA). This, it was urged, was not permissible and is the basis of the pending appeal.
After filing of notice of appeal, the Appellate Division made an order of remand for the purpose of taking testimony explaining the guidelines. At the hearing counsel for defendant were permitted to introduce evidence critical of the guidelines. See State v. Kunz, 55 N.J. 128 (1969).
I
History of the Project
The guidelines were developed for the purpose of reducing sentence disparity by making available to the sentencing judge as additional material the "going rate" for persons of similar characteristics for similar crimes. The basic technical premise of the guidelines is that not all factors (variables) which enter into a decision are equally important and that the salient variables can be selected by mathematical and statistical techniques.[1] The use of such techniques with voluminous data became feasible only with computer assistance,
In 1970 Professor Leslie Wilkins and Don M. Gottfredson (now Dean of the Rutgers School of Criminal Justice) published "Parole Guidelines," 28 CFR 2.20 (1970), which described a method of using such techniques in a revision of the operations of the Federal Parole Board. Suffice it that by use of such techniques the federal parole system has been revised, and is presently using a matrix based upon the significant factors which enter into the parole release decision[2]
*437 In 1973 a project financed by a LEAA grant was instituted to develop a remedy for sentencing disparity. The group participating consisted of statisticians, mathematicians and persons experienced in prison administration, with a steering committee composed of judges from various parts of the country.[3] The result was a report[4] (hereinafter "Feasibility Study") issued in October 1976 which stated that
The guideline system, in brief, takes advantage of and incorporates, the collective wisdom of experienced and capable sentencing judges by developing representations of underlying court policies. The system simultaneously articulates and structures legal judicial decision-making processes so as to provide clearer policy formulation, more cogent review and enhanced equity to criminal defendants everywhere. [at xiii]
The report states, in part:
(1) It is feasible to structure judicial discretion by means of sentencing guidelines: (a) the statistical wherewithal is neither excessively complicated, time-consuming, nor costly; and (b) conscientious judges across the country appear quite willing to adopt a guideline format.
(2) It is desirable to structure judicial discretion by means of sentencing guidelines: (a) totally unfettered judicial discretion and/or completely indeterminate sentencing are generally recognized *438 today as necessarily leading to inequities; (b) attempting to completely eliminate judicial discretion would lead to rigidity and/or circumvention of the law; and (c) it does not appear that any other presently available alternative would be just or as efficacious. [at xvi]
The technique advocated by the report is described as follows:
The guideline sentences were readily computed by giving assigned weights to particular aggravating and mitigating factors relating to pertinent characteristics of both the crime and the criminal, and locating those weights on a sentencing grid. The weights that resulted in an Offense Score (seriousness of the offense) were located on the Y axis and the Offender Score weights (prior record and social stability dimension) were located on the X axis. The cells of the grid contained the guideline sentence. By plotting the Offense Score against the Offender Score (much as one plots mileage figures on a road map), one is directed to the cell in the grid which indicates the suggested length and/or type of sentence. * * * [at xv]
It also states:
It is important to keep in mind that even when fully implemented, the guideline sentences are in no way intended to be binding, mandatory sentences. The judge as human decision-maker will still retain the discretion to override any suggested guideline. We are, however, suggesting that particularized written reasons be given when judges depart from the specific, narrowly drawn guideline sentence and later when the guideline model system becomes fully operational that judicial panels might perhaps be utilized in these more unusual cases. Moreover, the system we propose would feed back those departures into the data base used in constructing the guidelines, thus injecting a continuous element of self-improvement and regeneration into the guidelines. * * * [at xvi]
Following the Feasibility Study, guidelines were formulated for Denver County (Denver), Colorado; Cook County (Chicago), Illinois; Essex County (Newark), New Jersey, and Maricopa County (Phoenix), Arizona, using the principles and methods described.
The work in Essex County, directed by a team from the Criminal Justice Research Center and financed by a LEAA *439 grant, began in December 1976. It consisted of an analysis of 1205 presentence reports selected from cases sentenced in 1975 by a computer programmed to select a random sample. These 1205 cases comprised approximately one-fourth of all sentences imposed in Essex County in 1975. It was determined by the use of mathematical techniques that only 93 items had any significance in the sentencing decision (Data Collection Instrument EX. C-5). Thereafter, by use of mathematical and statistical techniques, with computer assistance, the significant factors and the crucial elements entering into the sentencing decision were ascertained. Tentative guidelines were formulated and validated by analyzing 500 current cases. With appropriate modifications, the final matrices were issued.
On July 9, 1977 the system was put into effect in Essex County. It was preceded by a letter to the New Jersey Law Journal, dated May 25, 1977, published June 9, 1977, 100 N.J.L.J. 508 (1977), which explained the system.
In the spirit of State v. Kunz, supra, the sentencing sheets, matrices and manuals used in conjunction with the guidelines were made available to both the prosecutor and defendant's counsel.
II
The Nature of the Problem
N.J.S.A. 2A:168-1 empowers a judge to suspend sentence in any case after conviction or plea of guilty or non vult and place the defendant on probation "for any crime or offense, except those hereinafter described,"[5] for a *440 period of not less than one year nor more than five years. The discretion thus entrusted to a sentencing judge ranges from a suspended sentence to the statutory maximum, except as limited by the right of appeal for abuse of discretion. See State v. Laws, 51 N.J. 494 (1968), cert. den. 393 U.S. 971, 89 S.Ct. 408, 21 L.Ed.2d 384 (1968); R. 2:10-3.
It is worth considering what guidance is provided by the basic principles set out by Chief Justice Weintraub in his classic statement in State v. Ivan, 33 N.J. 197 (1960):
The philosophical justification for `punishment' has divided men for centuries. Suggested bases or aims are (1) retribution, (2) deterrence of others, (3) rehabilitation of the defendant, and (4) protection of the public by isolation of the offender. Redmount, `Some Basic Considerations Regarding Penal Policy,' 49 Journal of Criminal Law, Criminology and Police Science 426 (1959). Today retribution is not a favored thesis, although some still claim a need to satisfy a public demand for vengeance. Perhaps it persists as an unarticulated premise in individual sentences. Present-day thinking emphasizes deterrence and rehabilitation. Few would permanently isolate the offender without regard to the nature of his crime upon a finding of incorrigibility. That course, however defensible in abstract theory, cannot be seriously considered until future behavior is predictable with substantial certainty. [at 199]
With customary perspicuity, he continued:
No single aim or thesis can claim scientific verity or universal support. Agreement can hardly be expected until much more is known about human behavior. Until then, the sentencing judge must deal with the complex of purposes, determining in each situation how the public interest will best be served. [at 201]
What is there in this statement of principles which can guide a judge in the exercise of his discretion? The difficulty in articulating a standard becomes apparent when one considers the objectives mentioned in Ivan, supra. When should an offender be isolated rather than rehabilitated? *441 And what weights should be afforded each of these considerations? Are there not cases in which giving primacy to rehabilitation will impair the object of deterrence or where the object of deterrence conflicts with rehabilitation?
Sentencing institutes among judges have confirmed the general impression that sentencing disparity is widespread. A 1974 study by the New Jersey Division of Correction and Parole revealed "wide discrepancies from one portion of the state to another in the lengths of incarceration to which offenders were sentenced for various crimes." Raymar, "Criminal Dispositions for New Jersey," 8 Seton Hall L. Rev. 1 (1976), at 22, n. 104. It is debatable whether appellate review has had much impact on the sentencing process. In each of the years 1974-75, 1975-76 and 1976-77, reversals on appeals from sentencing decisions have been at the rate of eight per cent of appeals from sentences.[5a]
Implicit in this is the conclusion to which common sense would lead, that the weights given to the sentencing principles outlined in State v. Ivan, supra, are far from uniform, and that the principles referred to there cannot guide a judge in sentencing.
Appellate courts have approved and disapproved sentencing judges' reliance upon one rather than another of the four justifications for punishment noted in Ivan, supra. In State v. Pickell, 136 N.J. Super. 340 (App. Div. 1975), a custodial sentence in a welfare fraud case was affirmed on the ground that deterrence was a legitimate objective of sentencing. But in State v. Harris, 70 N.J. 586 (1976), also a welfare fraud case, restitution as a condition of probation was vacated because "considering the offender and her worth to society in struggling to sustain her five children, the human cost of such deterrence in this instance *442 is too great." (at 596) In State v. Mpetas, 79 N.J. Super. 202 (App. Div. 1963) a sentence of three to five years for possession of three marijuana cigarettes was sustained. In State v. Ward, 57 N.J. 75 (1970) a two-to-three-year sentence for a similar offense was set aside and probation ordered; but cf. State v. Knight, 72 N.J. 193 (1976). In State v. Dunbar, 69 N.J. 333 (1976), a 15-to-20-year sentence for a youthful first offender on a second-degree murder conviction was sustained (Justice Pashman dissenting). Yet, in State v. Leggeadrini, 75 N.J. 150 (1977) in which a non vult plea to a first-degree murder charge by a 66-year-old offender resulted in a sentence of 25 to 30 years, the sentence was set aside and a sentence of 7 to 10 years imposed. In State v. Sherwin, 127 N.J. Super. 370 (App. Div. 1974), certif. den. 65 N.J. 569 (1974); cert. dism. Loughran v. N.J., 419 U.S. 801, 95 S.Ct. 9, 42 L.Ed.2d 32 (1974), the nature of the crime was said to make incarceration desirable for deterrence reasons. In others, as in Ward, supra, rehabilitation outweighed other considerations. Can it not be concluded that the objective selected is attributable to the social, economic and psychological background of the judge? See Ivan, supra, 33 N.J. at p. 201; Hogarth, Sentencing as a Human Process, 211-228 (1971).
The result in the individual case may be recognized as correct, but the principle stated is not a guide for decision in other cases.[5b]
Legal writing criticizing the unfettered authority of a sentencing judge is abundant: e.g., Frankel, "Lawlessness in Sentencing," 41 U. Cinc. L. Rev. 1 (1972); Davis, Discretionary Justice, 133 (1969); Morris, The Future of *443 Imprisonment, 45 (1975); Raymar, "Criminal Dispositions for New Jersey," supra; Singer, "In Favor of Presumptive Sentences," 5 Crim. Just. Q., 88 (1978), 24 Crime and Delinquency (1978).
Among the efforts to supply a remedy, that of the American Bar Association, in "Standards Relating to Sentencing Alternatives and Procedures" (1968), is notable. So too is the Model Penal Code (10 Uniform Laws Annotated, § 7.01 at 514-515 (1974)), which reads:
(1) The Court shall deal with a person who has been convicted of a crime without imposing sentence of imprisonment unless, having regard to the nature and circumstances of the crime and the history, character and condition of the defendant, it is of the opinion that his imprisonment is necessary for protection of the public because:
(a) there is undue risk that during the period of a suspended sentence or probation the defendant will commit another crime: or
(b) the defendant is in need of correctional treatment that can be provided most effectively by his commitment to an institution; or
(c) a lesser sentence will depreciate the seriousness of the defendant's crime.
* * * * * * * *
Subsection 2 reads similarly. However, sentencing judges were still left with generalities. For example, the guidelines formulated by the American Bar Association (ABA Standards Relating to the Administration of Criminal Justice, page 351, par. 2.1 (1974) state:
* * * * * * * *
(d) It should be recognized that in many instances in this country the prison sentences which are not authorized, and sometimes required, are significantly higher than are needed in the vast majority of cases in order adequately to protect the interests of the public. Sentences of twenty-five years or longer should be reserved for particularly serious offenses or, under the circumstances set forth in sections 2.5(b) and 3.1(c) (special term), for certain particularly dangerous offenders. For most offenses, on the other hand, the maximum authorized prison term ought not to exceed ten years except in unusual cases and normally should not exceed five years.
Little in this is of concrete assistance to a judge sentencing a specific defendant.
*444 The guidelines are an attempt to bring order and reason into the chaotic field of sentencing, furnishing the sentencing judge with a summary of what other judges have done which he can use in making his decision.
III
What the Guidelines Are and Are Not
There are four matrices, one each for crimes involving violence, property, drugs and a catch-all for miscellaneous offenses. Not included are cases of welfare fraud, because they result almost invariably in probation, and gambling cases, because only one judge sentences in such cases. Each matrix is formed by two axes, one an "offense" axis, the other an "offender" axis. The offense axis follows the statutory pattern, those with the least maximum sentence given the lowest number, and those with the greatest penalty ascribed the highest number. An exception is receipt of stolen automobiles (N.J.S.A. 2A:139-3) which, although bearing a ten-year sentence, rarely results in the imposition of incarceration. The offender axis takes account of the personal characteristics of the offender, basically prior juvenile and adult record, employment and drug addiction.
In each of the cells created by these axes are stated the number of cases in that cell of incarceration ("in"), probation ("out"), the median maximum sentences in months for the sample analyzed for the Youth Correctional Institute at Yardville, the median sentence in months for the Essex County Correctional Center, and the median maximum sentence for the New Jersey State Penitentiary, also in months. Figures for 25% leeway as to each sentence are given, 12 1/2% below and 12 1/2% above. A sample grid is annexed hereto.
A "Sentencing Sheet" (copy annexed) is prepared for each defendant which sets out the offense score, the constituent elements of the offender score, the total offender score, the guideline sentence, the actual sentence, and a *445 space for reasons given for deviation from the guideline sentence, if that occurs. The reverse of the Sentencing Sheet states explicitly that the guidelines are used by the judge only as an aid, and are not in any sense binding. They do not purport to describe a sentencing result. In the language of the Feasibility Study, supra:
The essentially descriptive nature of the guidelines must be emphasized. Based on recent practice in a given jurisdiction, the guidelines indicate the expected sentence for most cases and they identify the factors considered; they do not prescribe what the sentences or the criteria ought to be. [emphasis theirs, at xvii]
It is not expected that all cases will come within the guidelines. It is recognized that approximately 15% of the cases will be "exceptional" and cannot be expected to fall within the guidelines. The guidelines cannot possibly be considered as all-embracing.[6]
It is anticipated that the norms set out in the guidelines will change, as they did with changing perceptions as to marijuana between Mpetas and Ward, supra. It is expected, also, that additional data may alter or clarify the statistical picture. For that reason provision is made for periodic review of the guidelines. The request that judges give reasons for variation from the guidelines is intended to facilitate such review.
IV
Grounds of Attack on the Guidelines
The defense brief is a broad-based attack upon the use of the guidelines as promulgated in Essex County. It makes four major points: first, that the Essex County judges do not have the authority to promulgate sentencing guidelines; *446 second, that the guidelines contravene the sentencing policy; third, that use of the guidelines violates the standards of State v. Kunz, supra, and fourth, that the guidelines offend due process because they are irrational and deficient statistically. These points will be dealt with in order.
A. The argument that the Essex County judges do not have the authority to promulgate Sentencing Guidelines.
This contention proceeds from the premise that the judicial power under N.J. Const. art. VI, § I, par. 1, cannot be exercised in the abstract and can be exercised only where there is a justiciable controversy. It is argued also that under Art. VI, § II, par. 3 only the Supreme Court has rule-making power. Clearly, this is so. The question is whether the use of guidelines violates either of these principles. The history of the conduct of the court system in New Jersey since the adoption of the Judicial Article in 1947 demonstrates that the Judicial Article has not been a straightjacket preventing New Jersey courts from taking measures for the improvement of the functioning of the system. Pursuant to its rule-making power, Winberry v. Salisbury, 5 N.J. 240 (1950), the Supreme Court has promulgated rules of procedure, and the Administrative Office of the Court, under the authority of the Supreme Court, has issued directives controlling conduct of the courts. Examples in the criminal area are Pretrial Intervention Guidelines to be used in conjunction with R. 3:28 (PTI), see State v. Leonardis [II], 73 N.J. 360 (1977), and the administrative directive confining sentencing in gambling cases in each county to a single judge. See State v. DeStasio, 49 N.J. 247 (1967), cert. den. 389 U.S. 830, 88 S.Ct. 96, 19 L.Ed.2d 89 (1967). Of course, judges in Essex County do not have rule-making power;[7] however, the words "justiciable *447 controversy" do not confine judges to the decision of cases and prevent them from trying to improve the system.
In deciding whether it is desirable for judges in Essex County to attempt to make improvements in the system, one must bear in mind Mr. Justice Brandeis' historic statement in New State Ice Co. v. Liebmann, 285 U.S. 262, 311, 52 S.Ct. 371, 386, 76 L.Ed. 747, 771 (1931) (dissenting opinion):
To stay experimentation in things social and economic is a grave responsibility. Denial of the right to experiment may be fraught with serious consequences to the nation. It is one of the happy incidents of the federal system that a single courageous state may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country.
It would be pretentious to suggest that judges in Essex County are the equivalent of a state of the Union. Nevertheless, just as there is wisdom in permitting a state to experiment, so there is wisdom to permit latitude within the overriding framework of a state judicial administrative system. The Pretrial Intervention Program was authorized in 1970 in Essex County by the Supreme Court for just such reasons, and subsequently expanded to the entire State. Pressler, N.J. Court Rules, Comment, R. 3:28.
*448 There are many constitutional and technical requirements to which a sentence must conform at the risk of reversal, remand or modification,[8] none of which go to the heart of the matter: the decision whether to incarcerate and for how long. This is made plain by the Sentencing Manual for Judges (Administrative Office of the Courts 1976):
The somewhat amorphous theories of sentencing and the wide, almost undefinable, range of rationally relevant facts make the problem of justice to the individual and society difficult to determine.
* * * * * * * *
This problem, like a general discussion of theories of sentencing, arises herein only indirectly. The manual does not purport to develop criteria or standards, nor to state any other views of how the discretion should be exercised. [at 3]
Only the Supreme Court has the power to require judges to use guidelines. The guidelines are not an attempt to do what the Supreme Court has not done. They simply provide a sentencing judge with a condensed and organized overview of sentences imposed by colleagues in his county to use if he or she wishes. Some judges do not use the guidelines; some use them rarely. No penalty has attached or can attach to such judges. And even as to those who use the guidelines, there is no requirement that they adhere to them nor is there any binding requirement that they give reasons for departure from the guidelines. If a judge may give unmeasured and unmeasurable weight to such terms as "deterrence" or "rehabilitation," can it be said that he may not consider also a review of past sentencing practice?
*449 B. The argument that the Sentencing Guidelines contravene the sentencing policy of the New Jersey Supreme Court
Enough has been said and respectfully said to indicate that there is no clear picture of the entirety of the sentencing policy in this State. It is firmly established that the sentence should fit the offender as well as the crime. State v. Ward and State v. Leggeadrini, supra. The "offender" axis of the guidelines does precisely that. The data afforded by it comes from the presentence report. The presentence report itself contains only material deemed relevant to the sentencing decision. What is perceived to be nonessential is excluded. The guideline axis dealing with the offender focuses upon those aspects of the offender's personality which affect the sentencing decision. Much of what is in the presentence report is included in the offender's score in summarized version. Thus, legal status at time of events, prior juvenile incarcerations, prior adult convictions and prior adult incarcerations, all come from the "criminal record." Drug addiction and employment come from the "employment history" and "discussion of defendant's health." Upon inspection, without resort to mathematical techniques, it is intuitively apparent that these are factors which inevitably influence a sentencing decision. Appellate sentencing rulings show repeated reference to factors such as these. See, e.g., Williams v. New York, 337 U.S. 241, 69 S.Ct. 1079, 93 L.Ed. 1337 (1949); State v. Leggeadrini, supra; State v. Tirone, 64 N.J. 222, 230 (1974); State v. Green, 62 N.J. 547 (1973); State v. McBride, 127 N.J. Super. 399 (App. Div. 1974), aff'd 66 N.J. 577 (1975); State v. Ebron, supra; State v. Johnson, 67 N.J. Super. 414 (App. Div. 1961).
The selection made for the guidelines is more objective, being based on approved mathematical and statistical techniques which determine which items do in fact influence the *450 result.[9] The Public Defender's brief cites in support of its argument the Final Report of the Feasibility Study:
* * * [W]e were confident that there did exist an implicit policy formulation which acted as an underpinning for judicial decision-making in the sentencing area. Through careful analysis of present practice, we believed it possible to discover that implicit policy and make it explicit, thereby allowing in the future for clearer overall policy formulations, as well as more cogent review for individualized dispositions. We began with these premises, believe that we have checked them and found them valid, and continue to hold to them. [at 7]
This is entirely consistent with New Jersey sentencing policy. What it means is that the unarticulated, perhaps unconscious, considerations which influence a sentence can be exposed to scrutiny and, if necessary, modification and revision. The result is not concealment but further insight into the cognitive processes behind sentencing.
The guidelines used by the United States Board of Parole and those involved here are based upon the same mathematical and statistical principles.[10] Dr. Leslie J. Wilkins and Dean Don M. Gottfredson were among the experts who conducted both projects. The object of the Feasibility Study was to determine whether methods which had been used to develop parole guidelines could also be used for sentencing guidelines.[11] It is, therefore, illuminating to learn how the parole guideline system has been viewed by the courts.
*451 While the parallel between federal parole guidelines and sentencing guidelines is limited, on the issue involved here whether use of guidelines impedes individual consideration of the person involved both the federal parole system and our judicial system agree that each individual case must be considered separately and cannot be treated by rote. See State v. DeStasio, supra; U.S. v. Schwarz, 500 F.2d 1350 (2 Cir.1974); U.S. v. Hartford, 489 F.2d 652 (5 Cir.1974).[12] Federal cases have held almost uniformly that use of guidelines in parole decisions does not deny individualized consideration of each parole applicant: e.g., Garcia v. U.S. Board of Parole, 409 F. Supp. 1230, 1234 (N.D. Ill., E.D. 1976), aff'd 557 F.2d 100 (7 Cir.1977); Wiley v. U.S. Board of Parole, 380 F. Supp. 1194, 1197 (M.D. Pa. 1974); Battle v. Norton, 365 F. Supp. 925 (D. Conn. 1973). Contra, U.S. ex rel. Mayet v. Sigler, 403 F. Supp. 1243 (M.D. Pa. 1975), aff'd 556 F.2d 570 (3 Cir.1977) holding that as to youthful offenders, use of guidelines conflicts with the purpose of the Youth Corrections Act (criticized in Barr v. U.S., 415 F. Supp. 990, 995 (W.D. Okl. 1976). In Snyder v. U.S. Bd. of Parole, 383 F. Supp. 1153, 1155 (D. Colo. 1974) the court said:
There can be no question but that the adoption of guidelines for the exercise of parole discretion has important advantages in developing a consistent policy to aid hearing examiners in making their decisions. Fairness is perceived in dealing with similar persons in similar situations in similar ways. The use of these guidelines may be presumed to have a significant impact on reducing the effects of *452 disparate sentencing... Thus, the use of some standards may well be an important advance in criminal justice. [at 1155]
And in Wiley v. U.S. Bd. of Parole, supra, it was stated:
It should be emphasized that these guidelines were promulgated as just that merely guidelines and that the Board is free to render a decision either above or below the guideline range where in the opinion of the Board the circumstances warrant. [at 1197; emphasis theirs]
C. The argument that the Guidelines violate State v.B Kunz, supra
In Kunz defendant was convicted of receiving a stolen automobile, in violation of N.J.S.A. 2A:139-1. The presentence report "charged that a stolen vehicle ring, specializing in Cadillacs, was being operated by a named individual whose `contact in New Jersey' was the defendant and that the defendant had been involved with several identified purchasers of stolen automobiles in New Jersey." (55 N.J. at 131). The contents of the presentence report had not been disclosed to defendant prior to sentencing. The holding of the case was that the defendant should have been given an opportunity to review the presentence report prior to sentencing and the order made was that he be furnished with a copy of the report and "afforded fair opportunity to meet any prejudicial matter which may play a part in the sentencing." (55 N.J. at p. 144)
The requirements of Kunz are now embodied in R. 3:21-2. In the Tentative Draft Comments to the rule it was said:
The main argument in favor of disclosure is that the defendant should have the opportunity to meet what may be inaccurate and unreliable hearsay information in the report and also to correct prejudicial misapprehensions of the contents of the report by the sentencing judge. [Pressler, Rules Governing the Courts of the State of New Jersey, Tentative Draft Comments, at 425 (1977)].
Cases prior to Kunz involving claims that the judge's sentence was based upon erroneous assumptions of fact indicate *453 that the sentence will be considered defective only where the sentencing judge was influenced by a concrete erroneous statement of fact. In State v. Pohlabel, 61 N.J. Super. 242 (App. Div. 1960) the presentence report erroneously stated defendant's prior criminal record. In State v. Gattling, 95 N.J. Super. 103 (App. Div. 1967), certif. den. 50 N.J. 91 (1967), the trial judge, in imposing sentence, stated that it was well known that anyone who wanted to place a bet could see defendant, without substantiation in the presentence report. To the same effect see State v. Barbato, 89 N.J. Super. 400, 409 (Cty. Ct. 1965). But there is no such erroneous statement in the case at bar. The material used in the compilation of the guidelines and the manner in which the guidelines were formulated are at least reasonable.
Meticulous care has been exercised in Essex County to make certain that the guidelines are used in a manner conforming to both the letter and spirit of Kunz. Counsel are advised that the guidelines will be consulted. In the case at bar, the guidelines and the testimony upon which they were based were briefly but carefully explained to counsel. The manual used in preparing the matrices and a sentencing sheet based upon the appropriate matrix were supplied to counsel. Additionally, as already stated, the fact that the guidelines would be used was publicized in a letter to the New Jersey Law Journal in which the theory of the guidelines was explained. The basic Feasibility Report is in the public domain. Thus, defense counsel had all the material which the judge could possibly use in the sentencing. Kunz does no more than say that this is what should be supplied to defense counsel. Additionally, a hearing was held, also pursuant to Kunz, at which testimony was taken with respect to the formulation of the guidelines, and defense counsel were given an opportunity to challenge these matters on the record by both cross-examination and the production of witnesses. Kunz requires no more than this.
*454 The argument is that the guidelines do not disclose all of the characteristics of the individual cases which went into their formulation, and since they do not, this involves a violation of Kunz. However, Kunz makes no such ruling. Kunz requires only that the judge make available to a defendant all the material which he has.
There are many things which Kunz does not encompass. For example, it does not require the sentencing judge to tell counsel whether he considers rape a more heinous crime than manslaughter, or kidnapping than second-degree murder. Nor does it require the sentencing judge to reveal whether he considers age a mitigating or aggravating factor. A judge may consider an arrest record (State v. Green, supra) or a juvenile record (State v. Ebron, supra) in assessing the personality of a defendant, although it may not use it to infer guilt as to any underlying charge which defendant does not admit. There are many other things which a judge may use in assessing a defendant's personality, e.g. his employment record, family stability, social contacts, public activities. What weight is to be given to these is a matter for the sentencing judge. Kunz does not require that in weighing the arrest record the judge examine fully all the circumstances of the charges, nor does it specify that he must reveal what weight he gives to the arrest record. The same thing is true as to employment, family stability, et cetera.
Decisions involving federal parole hearings shed light on the problem. While grant of parole is an act of legislative grace, recent cases have declared that a prospective parolee is entitled to a modicum of due process protection. Wiley v. U.S. Board of Parole, supra at 1198-1201. Cf. Morrissey v. Brewer, 408 U.S. 471, 92 S.Ct. 2593, 33 L.Ed.2d 484 (1972). Thus, in the federal parole, as in our state system (Monks v. N.J. State Parole Bd. supra), an applicant for parole is entitled to a statement of reasons for refusal. Mower v. Britton, 504 F.2d 396 (10 Cir.1974). "Boiler plate" reasons, such as "Your release * * * would depreciate the seriousness of the offense committed," are *455 deficient. Garcia v. U.S. Board of Parole, supra, at 1238; Lupo v. Norton, 371 F. Supp. 156 (D. Conn. 1974); Billiteri v. U.S. Board of Parole, 385 F. Supp. 1217 (W.D.N.Y. 1974). In Barr v. U.S., supra, disclosure of a "high" offense severity rating and the salient factor score with the appropriate guideline range was considered a sufficient statement. No case has required the United States Parole Board to expose for scrutiny all the data which went into the parole guidelines.
D. Guidelines offend due process because they are irrational and deficient statistically
At the hearing technical criticism of the guidelines came from Dr. Edward Wolf, associate professor of statistics at Beirut [sic] (Baruch College) in the City University (of New York). Dr. Wolf is an expert in "regression analysis, variance [sic] multiple regression." His criticism is voiced in defendant's brief subsumed under the argument of failure to conform to Kunz. It belongs more properly under this heading.
The only cases cited in support of the contention that there is a violation of due process are Speiser v. Randall, 357 U.S. 513, 78 S.Ct. 1332, 2 L.Ed.2d 1460 (1958), which voices only the familiar rule that "beyond a reasonable doubt" is the standard of guilt in a criminal case, and two environmental protection cases, Ethyl Corp. v. Environmental Protection Agcy., 176 U.S. App. D.C. 373, 541 F.2d 1 (D.C. Cir.1976), cert. den. 426 U.S. 941, 96 S.Ct. 2662, 49 L Ed.2d 394 (1976) and Nat'l Resources Def. Council v. U.S. Nuclear Reg. Comm'n, 178 U.S. App. D.C. 336, 547 F.2d 633 (D.C. Cir.1976), cert. granted 429 U.S. 1090, 97 S.Ct. 1098, 51 L.Ed.2d 535 (1977), both of which deal with the role of a court in reviewing actions of an agency entrusted with the obligation of protecting the environment. Judge Bazelon's statement in Nat'l Resources Def. Council perhaps summarizes whatever pertinence those cases have to issues at bar:
*456 In order to determine whether an agency has lived up to these responsibilities, a reviewing court must examine the record in detail to determine that a real give and take was fostered on the key issues. This does not give the court a license to judge for itself how much weight should be given particular pieces of scientific or technical data, a task for which it is singularly ill-suited. It does require, however, that the court examine the record so that it may satisfy itself that the decision was based "on a consideration of the relevant factors." [178 U.S. App. D.C. at 348-349, 547 F.2d at 645-646, footnotes omitted]
The technical factors here are not nearly so complex as those in the environment cases. Nor are they so crucial since in the environment cases the agency has the power to block a projected project, whereas the guidelines are informational only. What Judge Bazelon's language makes clear is that courts should not substitute themselves for technical experts and not interfere where it is apparent that fair and reasonable consideration has been given to the pertinent material. Enough has been said to demonstrate that this has been done in the formulation of the guidelines.
"Due process," as applied to statutes, requires "only that a law shall not be unreasonable, arbitrary or capricious, and that the means selected shall bear a rational relation to the legislative object sought to be obtained." Robson v. Rodriquez, 26 N.J. 517, 522 (1958). Accord, Shepard v. Woodland Tp. Comm. and Planning Bd., 71 N.J. 230, 247 (1976). A fortiori, where the issue is whether a judge may use certain material to influence his mental processes, it is reasonable to conclude that the same principle applies. However, on analysis it is apparent that Dr. Wolf's technical criticism is without basis. It will be discussed to the extent that its elements can be isolated.
1. That the size of the sample used was too small and should have been 10,000 or more cases
It should suffice that the opinion of competent experts in the field the sample was sufficient, but the argument will be analyzed in the light of the record. It is reasonable to *457 infer that a larger sample might have created more confidence in the result, but there is no reason to conclude that the results have been more correct. Dr. Wolf came to his conclusion because some of the cells in the matrices had no cases in them, or a small number of cases. He was disturbed that the statistical picture of the sentencing grids was not logical because in many instances the cells showed that offenders with worse personal characteristics received less severe sentences than those with better personal profiles for offenses of the same statistical heinousness. Apparently, he came to the conclusion that the statistical work was unsound, because the sentence did not increase in regular steps as defendants' personal profile became worse. What was not understood was that the statistical picture was of things as they are, not as they should be; that it showed apparent widespread sentence disparity, a problem which the guidelines were designed to meet.
Dr. Wolf's conclusion as to the required number of cases was based on the theory that to make proper use of multiple regression[13] the statistical technique in which he is expert, it would require 30-40 cases per variable. Multiple regression, however, was only one of the techniques used. Arthur M. Gelman, an expert associated with the Feasibility Study from its beginning, testified: *458 An assortment of analysis was used in establishing these items of information. We use a combination of Pearson's Correlation analysis, multiple regression analysis and a series of tests using a Burgess or Standard Weighting System.[14]
Joseph M. Calpin, also associated with the Feasibility Study, and experienced in working with statistics in the field of criminal law, testified "the methodology used in this study is an off-shoot of the work done in the United States Parole Commission Pioneering Report." He testified as to the statistical methods used in the Essex County project, making reference to Pearson's Correlation" and "Burgess type scales."[15] Additionally, Dr. Wolf failed to consider a *459 factor which Dr. Wilkins made clear, that the matrices are to be used as a continuum, that the cells relate to one another, and that when it appears that the median sentence shown in a cell is not consistent with the adjacent cells, the adjacent cells must be consulted to ascertain the central sentencing tendencies.
2. That influential variables were omitted
This has reference to the offender axis in which 93 items deemed to be relevant were reduced by the statistical techniques to six or seven, depending upon the matrix involved. This attack asserts that the most meritorious aspect of the system is a defect.
The "Data Collection Instrument" contains 93 items, all of which come from the presentence report. Some of the items, such as "Variable 40 Victim's Addiction to Alcohol and Drugs" and "Variable 42 Employment Status of Victim," are, upon mere inspection, only peripherally relevant. Others, such as "Variable 43 Value of Property involved in Offense," or "Variable 44 Did Criminal Behavior Involve `Distribution' of a Drug?," are reflected in the statutory classification of the crime. The Methodology employed did no more than eliminate by formal method items which would probably have also been eliminated or relegated to a subsidiary status upon inspection and reflection. In any event, the information remains available to the sentencing judge in the presentence report. The reduction in the number of variables was done deliberately to clarify and improve the decision. Dr. Wilkins testified:
So sometimes by simplifying and indeed by reducing either relevant information, and I'm now quoting a Rand study, improves the decision itself. [sic]
Now that sounds rather odd that by in fact deleting relevant information you improve the decision, but that is in fact demonstrated to be so by both a large amount of experiments and certainly in relation to war games data.
*460 The items which were strained out by the methods employed, in most cases, on inspection would appear to have minor relevance. What has been retained are those which significantly affect the sentencing decision. This is precisely what was done in the federal parole system, to the satisfaction of Congress and the federal courts.
3. Failure to stratify by judges
Dr. Wolf suggested that it would have been better to classify sentences by judge, rather than by individual cases, and select the median judge as the standard. The problems with such a solution are obvious. How does one find a median judge? It is not likely that a judge who is median as to one crime also will be median as to other crimes. Nor is there intrinsic merit in the decisions of the hypothetical median judge. It would seem that a broad spectrum of sentencing decisions which show the going rate for crimes of a certain character and offenders of specified characteristics is more likely to demonstrate central sentencing tendencies.
Finally, the issue is not whether a better system is possible, but whether this system is so bad that its use for information purposes only impairs the defendant's due process rights.
4. Arbitrary grouping of offenses
The argument as to this is
Thus, the individual identity of each crime is camoaflaged [sic] behind "offense scores" that potentially bear no relationship to the aspects of the individual offenses that influence judges most. For instance, it was demonstrated at the hearings that the crimes of rape and "assault with intent to commit burglary" were grouped together in the same so called "generic" offense score. [Defendant's Brief, p. 42]
The hearing showed that attempts to make matrices for "crimes specific" were abandoned because there was not sufficient *461 data. Instead, a policy decision was made to classify crimes by the statutory penalties, with the exception of the charge of receipt of a stolen automobile, for which a sentence of incarceration is rarely imposed. Here, again, the issue is not whether a better system can be devised, but whether this system is so bad that it misleads and violates due process. As to this, it can hardly be argued that to adopt a statutory classification which has never been attacked for the ranking of the heinousness of crimes violates a defendant's due process.
5. That the Guidelines accomplish nothing
Dr. Wolf, referring to the decision as to whether to incarcerate (the "in-out" decision) said "Incarcerate everyone and the agreement would be 79 per cent with actually what occurred and so we're getting an improvement of six per cent." (It being calculated that 85 per cent of sentences imposed will fall within the guidelines.) From this he concludes that the guidelines have really accomplished nothing. Assuming the calculation correct, the conclusion is wrong. The guidelines are informative as to what judges have been doing. To point out that there is only a 6% variation between what the matrix shows and what it would show if everybody were incarcerated is without significance; certainly it does not prove the guidelines accomplish nothing.
The decision to use the guidelines as additional information on sentencing was made at the time defendant was sentenced. The reasons for that decision have been articulated above and expanded to meet issues raised by the brief filed on behalf of the defendant, primarily for the assistance of the appellate court. There is no need for an order.
See following page for appendix
*462 SAMPLE GRID VIOLENT ------------------------------------------------------------------------ | 12 | 24 | 22 | | 3 | 1 | 0 | | murders | murders | murders | | (1) 105.0 120.0 135.0 | (2) 126.0 144.0 162.0 | (0) | | (0) | (0) | (0) | | (3) 105.0 120.0 135.0 | (3) 157.5 180.0 202.5 | (4) 226.7 258.0 290.2 | 5 |-----------------------|-----------------------|------------------------| | (4) 63.0 72.0 81.0 | (11) 73.5 84.0 94.5 | (5) 105.0 120.0 135.0 | | (2) 6.1 7.0 7.9 | (1) 5.2 6.0 6.7 | (1) 10.5 12.0 13.5 | | (2) 105.0 120.0 135.0 | (6) 94.5 108.0 121.5 | (9) 84.0 96.0 108.0 | |-----------------------|-----------------------|------------------------| | 4 | 11 | 12 | | 9 | 6 | 4 | 4 | | (5) 52.5 60.0 67.5 | (2) 52.5 60.0 67.5 | | OUT | (6) 10.5 12.0 13.5 | (6) 13.1 15.0 16.9 | | | (0) | (4) 94.5 108.0 121.5 | |-----------------------|-----------------------|------------------------| | 8 | 9 | 13 | | 7 | 13 | 5 | 3 | | | (2) 52.5 60.0 67.5 | | OUT | OUT | (6) 10.5 12.0 13.5 | | | | (5) 52.5 60.0 67.5 | |-----------------------|-----------------------|------------------------| | 1 | 0 | 3 | | 1 | 2 | 0 | 2 | | | (1) 31.5 36.0 40.5 | | OUT | OUT | (2) 10.5 12.0 13.5 | | | | (0) | |-----------------------|-----------------------|------------------------| | 0 | 0 | 1 | | 5 | 2 | 4 | 1 | | | | | OUT | OUT | OUT | | | | | ------------------------------------------------------------------------ -1 0 1 construction ------------------------------------------------------------------------ | 21 | 31 | 11 | | 0 | 2 | 0 | | murders | murders | murders | | (0) | (0) | (0) | | (0) | (0) | (0) | | (5) 157.0 180.0 202.0 | (3) 231.0 264.0 297.0 | (1) 262.5 300.0 337.5 | 5 |-----------------------|-----------------------|------------------------| | (1) 178.5 204.0 229.5 | (4) 63.0 72.0 81.0 | (2) 52.5 60.0 67.5 | | (2) 15.7 18.0 20.2 | (1) 15.7 18.0 20.2 | (2) 34.1 39.0 43.9 | | (12) 63.0 72.0 81.0 | (21)105.0 120.0 135.0 | (6) 84.0 96.0 108.0 | |-----------------------|-----------------------|------------------------| | 25 | 21 | 11 | | 2 | 0 | 0 | 4 | (6) 63.0 72.0 81.0 | (8) 73.5 84.0 94.5 | (1) 73.5 84.0 94.5 | | (4) 10.5 12.0 13.5 | (2) 10.9 12.5 14.1 | (2) 13.1 15.0 16.9 | | (15) 73.5 84.0 94.5 | (10) 57.7 66.0 74.2 | (8) 57.7 65.0 74.2 | |-----------------------|-----------------------|------------------------| | 11 | 4 | 0 | | 0 | 0 | 0 | 3 | (0) | (1) 73.5 84.0 94.5 | (0) | | (5) 13.1 15.0 16.9 | (0) | (0) | | (5) 31.5 36.0 40.5 | (3) 52.5 60.0 67.5 | (0) | |-----------------------|-----------------------|------------------------| | 0 | 2 | 0 | | 1 | 1 | 0 | 2 | (0) | (1) 31.5 36.0 40.5 | (0) | | (0) | (0) | (0) | | (0) | (1) 262.5 300.0 337.5 | (0) | |-----------------------|-----------------------|------------------------| | 0 | 0 | 0 | | 0 | 0 | 0 | 1 | | | | | OUT | OUT | OUT | | | | | ------------------------------------------------------------------------ 2 3 4 Offender Score *463 SENTENCING SHEET VIOLENT OFFENDER ____________________ PROBATION DEPT. CASE NO. _____________________ JUDGE _______________________ DATE OF SENTENCING ___________________________ Offense(s) Convicted Of: ___________________________________________________ (title + statute number) ___________________________________________________ ___________________________________________________ --------- OFFENSE TYPE (MOST SERIOUS OFFENSE) | VIOLENT | --------- OFFENSE TYPE OFFENSE SCORE --------- Inter-Class Rank ____ = | | --------- OFFENSE SCORE OFFENDER SCORE A. Legal Status at Time of Offense ____ + 0 = Free 1 = Not free B. Prior Juvenile Delinquency Petition(s) Sustained ____ + 0 = None or one 1 = Two or more C. Prior Adult Incarcerations ____ + 0 = None 1 = One or more incarcerations D. Drug Addiction ____ · 0 = No use/not addicted 1 = Addicted ---------- E. Offender Status ____ = | | -1 = Part/full-time employment/school ---------- 0 = Unemployed/not in school OFFENDER SCORE GUIDELINE SENTENCE ________________________________ ACTUAL SENTENCE ___________________________________ REASONS (if actual sentence does not fall within guideline range): NOTE: The sentence information supplied herein is a result of a sentencing study project conducted with the assistance of staff from the Criminal Justice Research Center of Albany, New York. The project consisted of a study of actual case histories in Essex County and sentences imposed by the judges sitting there for the purpose of determining and analyzing the essential factors in the sentencing decision. A statistical analysis of this information was made. The end product produces in graph from the projected median sentences in the cases analyzed. The purpose of these graphs is to supply the judge with statistical information not previously available to him. They are intended as tools to aid the judge *464 in the exercise of his sentencing discretion, the graphs do not apply to gambling cases, or cases of welfare fraud, or false pretense charges relating to welfare. The information supplied herein is not binding on the sentencing judge in any sense, and, of course, has no official or authoritative status, sanction or approbation by any branch of government. Five copies of the Sentencing Sheet will be made, one for the Probation Office file, one for the Research Center, one for the judge's file, one for the prosecutor, and one for the defense counsel.
NOTES
[1] The techniques used in formulating the guidelines will be described in some detail infra.
[2] Gottfredson, Hoffman, Sigler, Wilkins, "Making Parole Policy Explicit," 21 Crime and Delinquency 34 (Jan. 1975). These guidelines are revised periodically. 42 Fed. Reg. 12043 (1977); Gottfredson, Wilkins, Hoffman and Singer, "The Utilization of Experience in Parole Decision-Making (1974) (prepared for National Institute of Law Enforcement and Criminal Justice, Law Enforcement Assistance Administration, United States Department of Justice); Hoffman and Beck, "Parole Decision-Making: A Salient Factor Score," 2 J. of Crim. Just. 195 (1974).
[3] The writer and Superior Court Judge John A. Marzulli, were members. We also were responsible for directing the implementation phase in Essex County.
[4] Wilkins, Kress, Gottfredson, Calpin and Gelman, "Sentencing Guidelines: Structuring Judicial Discretion" (1976) (prepared for a National Institute of Law Enforcement and Criminal Justice, Law Enforcement Assistance Administration, United States Department of Justice). Since the official version of the report was published in February 1978, a substitution of same has been made in the exhibits. Page references are to the official version (#XXX-XXX-XXXXX-X, Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402).
[5] The exceptions are limited: Murder in the first degree (N.J.S.A. 2A:113-4) in which a sentence of life imprisonment is required; State v. Funicello, 60 N.J. 60 (1972), cert. den. New Jersey v. Presha, 408 U.S. 942, 92 S.Ct. 2849, 33 L.Ed.2d 766 (1972); State v. Corbitt, 74 N.J. 379 (1977); and kidnapping, N.J.S.A. 2A:118-1, where the sentence must be not less than 30 years; State v. Rosenberg, 30 N.J. Super. 369 (App. Div. 1954); State v. Hampton, 61 N.J. 250 (1972); cf. State v. Brozi, 125 N.J. Super. 485 (App. Div. 1973), certif. den. 64 N.J. 501 (1974) as to indeterminate term for youthful offenders.
[5a] Annual Reports of the Administrative Director of the Courts, 1974-75, 1975-76, 1976-77, Table of Cases and Reversal Rate, pp. B-25 [1974-75], B-32 [1975-76] and B-32 [1976-77].
[5b] There are, of course, many areas in the law when a judge uses "principles, policies and * * * standards," rather than "rules" to reach a conclusion (Weinreb, "Law as Order," 91 Harv. L. Rev. 909, 913; however, they do not result in the acute, manifest problem of sentencing disparity, which should be mitigated, if it can be.
[6] Statements such as that in Singer, "In Favor of Presumptive Sentences Set by a Sentencing Commission," supra at 97, that the guidelines are prescriptive, are erroneous.
[7] The Public Defender's brief asserts that the guidelines do not have the imprimatur of the Chief Justice, Supreme Court, Administrative Office of the Courts, or the Assignment Judge of Essex County. This is true. However, it is also true that the Assignment Judge and the Criminal Assignment Judge of Essex County had full knowledge of the role played by the writer and Judge Marzulli in the basic feasibility study and in making the guidelines available for the Essex County judges; that the grant had been made for the implementation of the technique in Essex County; that the probation files in Essex County were being used to accumulate the necessary data, and that the judges sitting in criminal cases in Essex County were requested to make use of the guidelines on a voluntary basis. It is also true that all of these events transpired with the full knowledge and permission of the Administrative Office of the Courts. Although this does not appear in the record, this footnote is inserted to rebut a possible inference that those judges acted in an irresponsible and undisciplined manner.
[8] E.g., right to counsel at sentencing, State v. Jenkins, 32 N.J. 109 (1960); right of allocution, R. 3:21-4(b); presentence report, R. 3:21-2, except when convicted of first-degree murder, State v. Robinson, 139 N.J. Super. 58 (App. Div. 1976); separate sentence for each count must be stated, R. 3:21-4(e); State v. Quatro, 44 N.J. Super. 120 (App. Div. 1957), cert. den. 355 U.S. 850, 78 S.Ct. 73, 2 L.Ed.2d 60 (1957); a judge must sentence only for the crime of which defendant was convicted, State v. Marzolf, 152 N.J. Super. 47 (App. Div. 1977).
[9] The concept of the use of experience tables in the working of a parole system, or even in predicting delinquency and crime, is by no means new. It was considered by the U.S. Board of Parole in 1939, and by S. and E. Glueck of Harvard Law School in 1959. Newman, "Parole Making and the Sentencing Process," 84 Yale L.J. 810 (1970), at 872, n. 308.
[10] See note 2, supra; Feasibility Study, Supra. For comparison of parole and sentencing considerations see Monks v. N.J. State Parole Board, 58 N.J. 238, 243-244 (1971). A copy of the Federal Parole Guidelines may be found at 36-37 of the Feasibility Study.
[11] Feasibility Study, at 20.
[12] 18 U.S.C.A. § 4203(a) (i) specifically provides for "establishing guidelines." It was enacted in 1976 because "The promulgation of guidelines to make parole less disparate and more understandable has met with such success that this legislation incorporates the system into the statute." 2 U.S. Code Cong. & Admin. News, 352-353 (1976), cited in Geraghty v. U.S. Parole Comm'n., 429 F. Supp. 737 (M.D. Pa. 1977). Release on parole is by legislative grace, not of right, and the determination is wholly within the discretion of the Parole Board. U.S. v. Frederick, 405 F.2d 129 (3 Cir.1968).
[13] Ostle, Statistics in Research (2 Ed. 1963):
When we possess information on two or more related (or concomitant) variables, it is natural to seek a way of expressing the form of the functional relationship. In addition, it is desirable to know the strength of the relationship. That is, not only do we seek a mathematical function which tells us how the variables are interrelated, but also we wish to know how precisely the value of the one variable can be predicted if we know the values of the associated variables. The techniques used to accomplish these two objectives are known as regression methods and correlation methods. Regression methods are those used to determine the "best" functional relation among the variables, while correlation methods are used to measure the degree to which the different variables are associated. [Emphasis theirs]
[14] This quotation was incorrectly recorded; it has been corrected. The testimony continued:
* * * * * * * *
THE COURT: I'd like to have you define what these techniques are, Pearson's Corellation [sic] Multiple Regression, Burgess, did you use chi squared? [sic]
THE WITNESS: No we feel that Pearson's Corellation [sic] is a better measure of the association between the two variables looked at in the analysis.
THE COURT: Now will you tell us what these things are?
THE WITNESS: In simple language it's a degree of association between two items of information.
As X gets bigger does Y get bigger, does it get smaller, are there seemingly no influences on the size of Y based on the size of X, so that's basically what type of analysis Person's Corellation [sic] is.
[15] Burgess-type scales have long been used in the Federal Parole System. Pearson's Correlation is a linear correlation. As to the efficiency of this method as contrasted with Multiple Regression or "least squares coefficients," see Warner, "Estimating Coefficients in Linear Models: It Don't Make No Nevermind," 83 Psychol. Bull., No. 213-217 (1976); Davis and Corrigan, "Linear Models in Decision Making," 81 Psychol. Bull. No. 2, 95-106. For a layman's discussion of the statistical techniques used with the Parole Guidelines, see Wilkins, "The Problem of Overlap in Experience Table Construction Supplemental Report Three" (Davis, California: Parole Decision-Making Project, National Council on Crime and Delinquency Research Center, June 1973). Dr. Wilkins testified at the hearing that less sophisticated techniques produced better results with the kind of data involved.