UNITED STATES COURT OF APPEALS FOR THE SECOND CIRCUIT
decided: March 7, 1983.
LARRY FRAZIER, PETITIONER-APPELLANT,
JOHN R. MANSON, COMMISSIONER OF CORRECTIONS, RESPONDENT-APPELLEE; KENNETH SCHAFFER, PETITIONER-APPELLANT, V. CARL ROBINSON, WARDEN, RESPONDENT-APPELLEE
Appeals from a judgment of the United States District Court for the District of Connecticut, T. Emmet Clarie, Judge, dismissing prisoners' consolidated petitions for writs of habeas corpus claiming that Conn. Gen. Stat. § 18-71, which increases the "good time" credit for inmates sentenced after October 1, 1976, discriminates against prisoners sentenced prior to that date and violates the Equal Protection Clause.
Mansfield and Meskill, Circuit Judges, and Neaher,*fn* District Judge.
MANSFIELD, Circuit Judge:
Larry Frazier and Kenneth Schaffer appeal from a judgment of the United States District Court for the District of Connecticut (T. Emmet Clarie, Judge), affirming the dismissal of their consolidated petitions for writs of habeas corpus. The petitions claim that Conn. Gen. Stat. § 18-7a, which increases the good time credit per year for inmates sentenced on or after October 1, 1976, discriminates against prisoners sentenced prior to that date, including petitioners, in violation of the Equal Protection Clause of the Fourteenth Amendment. We affirm.
Frazier was originally sentenced on July 3, 1975 by the Connecticut Superior Court in Fairfield County to a term of not less than 54 nor more than 108 years, following his conviction of several criminal counts. On December 8, 1976, he was resentenced to a term of not less than 30 years nor more than 60 years.
On July 3, 1973, Schaffer was sentenced to a term of 17 years to life by the Superior Court after his conviction for murder. Both appellants are currently serving their sentences and the Connecticut Commissioner of Corrections is computing good time credit on their sentences based on the provisions of Conn. Gen. Stat. § 18-7 which applies to persons sentenced before October 1, 1976.*fn1
While appellants were serving these sentences, the Connecticut legislature enacted Conn. Gen. Stat. § 18-7a,*fn2 which applies only to prisoners sentenced on or after October 1, 1976, its effective date, and provides more good time credit per year than its predecessor, § 18-7. A prisoner sentenced before October 1, 1976 who maintains a good conduct and work record receives under § 18-7 110 days of good time for each year of the first five years of a sentence and about 140 days for each subsequent year. On the other hand, a prisoner sentenced on or after October 1, 1976, with a similar conduct and work record, receives 120 days of good time for each year of the first five years of a sentence and 180 days for each subsequent year. Thus, the effect of Connecticut's non-retroactive enlargement of good time benefits is to subject two persons convicted of the same offense and sentenced to the same maximum term to different periods of incarceration, depending on whether they were sentenced before or after October 1, 1976. Frazier and Schaffer argue that this statutory classification arbitrarily discriminates against persons sentenced before October 1, 1976, and thus violates their right to equal protection of the law as guaranteed by the Fourteenth Amendment.
Each appellant exhausted his available state remedies by prosecuting unsuccessful appeals in the Connecticut courts. The Connecticut Supreme Court applied a "rational basis" test to the statutory classification and rejected the equal protection claim. Frazier v. Manson, 176 Conn. 638, 410 A.2d 475 (1979).*fn3 The Court found that the purpose of § 18-7a was a rational one, i.e., "to consolidate good time statutes and to eliminate multiple systems of computing and crediting good time," thereby eliminating disparity in good time eligibility based on the type of sentence imposed or the institution in which an inmate might be incarcerated. However, the court did not expressly explain how this or any other rational purpose was served by discriminating in the amount of good time credits between persons sentenced before and after October 1, 1976. It merely cited two decisions of the Rhode Island Supreme Court holding that retroactive extension of enlarged good time benefits would amount to encroachment by the legislature on the sentencing authority of the judiciary. See Opinion to the Governor, 91 R.I. 187, 162 A.2d 814 (1960), and Mastracchio v. Superior Court, 98 R.I. 111, 200 A.2d 10, cert. denied, 379 U.S. 852, 13 L. Ed. 2d 55, 85 S. Ct. 96 (1964). In addition, the Connecticut Supreme Court found that " 'the seeming inequity in fixing a cut-off date is outbalanced by the factors of reliance and burden on the administration of justice which argue for prospective application only. Stovall v. Denno, 388 U.S. 293, 300-01, 87 S. Ct. 1967, 18 L. Ed. 2d 1199 (1967). '" 410 A.2d at 481 (quoting Mirenda v. Ulibarri, 351 F. Supp. 676, 677 (C.D. Cal. 1972)).
After rejection of their equal protection claim by the Connecticut Supreme Court, Frazier and Schaffer filed their present federal petitions pursuant to 28 U.S.C. § 2254, which were consolidated and referred to Magistrate F. Owen Eagan. He recommended that they be dismissed on the ground that the nonretroactivity of § 18-7a could be rationally justified as furthering the legitimate state purpose of avoiding a legislative modification of a judicial sentence in contravention of Connecticut's constitutional separation of powers. Judge Clarie approved the magistrate's recommended ruling.
Legislation that does not employ suspect classifications or impinge on fundamental rights must be upheld under the equal protection clause of the Fourteenth Amendment when the legislative means are rationally related to a legitimate government purpose. Clements v. Fashing, 457 U.S. 957, 102 S. Ct. 2836, 2843, 73 L. Ed. 2d 508 (1982); Schweiker v. Wilson, 450 U.S. 221, 230, 67 L. Ed. 2d 186, 101 S. Ct. 1074 (1981); Buffalo Teachers Federation, Inc. v. Helsby, 676 F.2d 28, 29 (2d Cir. 1982) (per curiam). Appellants argue that since good time credits affect the length of incarceration they impinge on a protected liberty interest and as such should be analyzed under the strict scrutiny standard that requires a compelling state necessity to sustain a discriminatory classification. We disagree. In McGinnis v. Royster, 410 U.S. 263, 35 L. Ed. 2d 282, 93 S. Ct. 1055 (1973), the Court inquired into the constitutionality of bestowing good time credits on some prisoners but not on others and concluded that "the determination of an optimal time for parole eligibility elicited multiple legislative classifications and groupings, which . . . require only some rational basis to sustain them." Id. at 270. In accord, Doyle v. Elsea, 658 F.2d 512, 518 (7th Cir. 1981); United States ex rel. Sero v. Preiser, 372 F. Supp. 660, 671 (S.D.N.Y.), aff'd in part and remanded in part, 506 F.2d 1115 (2d Cir. 1974), cert. denied, 421 U.S. 921, 95 S. Ct. 1587, 43 L. Ed. 2d 789 (1975); cf. Marshall v. United States, 414 U.S. 417, 422, 38 L. Ed. 2d 618, 94 S. Ct. 700 (1974); see also Haag v. Ward, 632 F.2d 206, 208 (2d Cir. 1980) (per curiam); Dillard v. LaVallee, 559 F.2d 873, 874 (2d Cir.), cert. denied, 434 U.S. 999, 98 S. Ct. 641, 54 L. Ed. 2d 495 (1977). While good time credits do affect the length of incarceration, implicating a protected liberty interest, cf. Wolff v. McDonnell, 418 U.S. 539, 556-57, 41 L. Ed. 2d 935, 94 S. Ct. 2963 (1974), the state's equally important interests, including promotion of discipline and rehabilitation, militate against a compelling necessity test, McGinnis v. Royster, 410 U.S. at 274-76; Abrahams v. Rodgers, 691 F.2d 87, 88 (2d Cir. 1982).
Appellants attempt to distinguish McGinnis on the ground that it dealt with good time credits relevant to computing minimum terms and parole eligibility, whereas here the good time computation affects maximum terms and discharge. However, the McGinnis court knew that the good time differences at issue there affected the time of release of prisoners to the community, resulting in longer incarceration for those in county jail custody without good time credits than for those in state prison. Id. at 267; Doyle v. Elsea, supra, 658 F.2d at 518.
In applying the rational basis test to determine whether the discriminatory classification under attack serves some legitimate purpose, we look first to expressions by the legislature on the subject. However, we need not necessarily stop there since "the search for the 'actual ' or 'primary ' purpose of a statute is likely to be elusive." Michael M. v. Sonoma County Superior Court, 450 U.S. 464, 469-70, 67 L. Ed. 2d 437, 101 S. Ct. 1200 (1981); McGinnis v. Royster, supra, 410 U.S. at 276-77. If there are no articulated purposes or if those expressed do not appear to be pertinent or exclusive, it is our function, through examination of the surrounding circumstances, including the legislative history of the statute, and application of logic, to decide whether the legislative classification was a "deliberate, considered choice" rather than "the result of inadvertence or ignorance," Schweiker v. Wilson, 450 U.S. 221, 235-36, 67 L. Ed. 2d 186, 101 S. Ct. 1074 (1981), and, if so, whether the legislature "could have concluded rationally" that the discrimination was justifiable and "could rationally have distinguished" between the classifications under attack. McGinnis v. Royster, supra, 410 U.S. at 274. In short, we then search for "a legitimate purpose that we may reasonably presume to have motivated an impartial legislature." United States Railroad Retirement Board v. Fritz, 449 U.S. 166, 180-81, 66 L. Ed. 2d 368, 101 S. Ct. 453 (1980) (Stevens, J., concurring). It may logically be inferred from a considered choice that the legislature acted to further some particular purpose. Schweiker v. Wilson, supra, 450 U.S. at 236.
Applying these principles here, the articulated purpose of § 18-7a, to eliminate multiple systems of computing and crediting good time that existed prior to 1976 because of different good time standards for different institutions in the state, while certainly a legitimate state purpose, does not justify the creation of a dual system for calculating good time based on date of sentence. The denial of the new good time standards to pre-October 1976 prisoners is irrelevant to the legislature's expressed purpose, which would have been achieved with equal effectiveness if the new standards had been made retroactive. Yet the legislature deliberately chose to exclude from the enlargement of good time benefits all persons sentenced prior to October 1, 1976. The enacted statute expressly states that it applies only to "person[s] sentenced to a term of imprisonment, on and after October 1, 1976." Conn. Gen. Stat. § 18-7a(a). Furthermore, the legislative history, though sparse, confirms the deliberateness of the nonretroactivity provision. See, e.g., Conn. Gen. Assembly, Raised Committee Bill No. 5125, § 3 ("This act shall not apply to those persons sentenced to a term of imprisonment prior to the effective date of this act. Said persons shall continue their eligibility for such credit as was provided for prior to the effective date of this act").
Absent any evidence that the legislature lacked a legitimate reason for the distinction or acted for an unlawful purpose, we may explore other rational justifiable bases. The state offers the justification advanced by it to the Connecticut Supreme Court and accepted by that court, i.e., that retroactivity would impose an excessive administrative burden on it. See Stovall v. Denno, 388 U.S. 293, 300, 87 S. Ct. 1967, 18 L. Ed. 2d 1199 (1967); Jackson v. Alabama, 530 F.2d 1231, 1238 (5th Cir. 1976); Mirenda v. Ulibarri, 351 F. Supp. 676, 677 (C.D. Cal. 1972). We find this contention unpersuasive. Since this case was heard in the state and federal courts upon stipulated facts without an evidentiary hearing, we are presented with no actual findings as to the extent of the administrative burden. The sole support offered by the state is an affidavit by one Donald Parker, a research analyst for the Department of Corrections, indicating that there were approximately 4,000 persons under sentence in Connecticut when § 18-7a became effective. However, this untested document makes no attempt to attest to the actual burden of recalculating the good time that would be earned under the new liberal terms of § 18-7a if it applied to the 4,000-odd pre-October 1976 inmates in Connecticut jails. Given today's computer facilities, which Connecticut should be able to put at its disposal, we may reasonably infer that " any supposed administrative inconvenience would be minimal. '" Williams v. Illinois, 399 U.S. 235, 245, 26 L. Ed. 2d 586, 90 S. Ct. 2018 (1970) (quoting Rinaldi v. Yeager, 384 U.S. 305, 310, 16 L. Ed. 2d 577, 86 S. Ct. 1497 (1966)). The constitutional imperatives of the Equal Protection Clause cannot be satisfied by mere conjecture as to administrative inconvenience. See Williams v. Illinois, supra, 399 U.S. at 245.
Notwithstanding the inadequacy of the administrative convenience rationale offered by the state, the legislature's classification may be justified by a legitimate discernible purpose which was suggested by the Connecticut Supreme Court and relied upon by the district court -- the avoidance of encroachment by the legislature on the judiciary's sentencing authority. The legislature could reasonably have relied upon the fact that in imposing sentence a judge frequently takes into account existing systems of computing good time. It could then have determined that the more liberal method of awarding good time credits under § 18-7a should not be made retroactive for the reason that if it had been in effect prior to October 1, 1976, longer sentences would have been imposed initially upon appellants and other similarly situated prisoners. We therefore conclude that it was rational for the Connecticut legislature to avoid a retroactive legislative modification of judicial sentences.
We need not decide whether the Connecticut legislature's reluctance to intrude on judicial sentencing powers is constitutionally mandated under Connecticut's separation of powers provision, Conn. Const. art. II, nor under art. IV, § 13, which vests the pardoning and clemency powers in the executive and not the legislature. See Comment, Today's Law and Yesterday's Crime: Retroactive Application of Ameliorative Criminal Legislation, 121 U. Pa. L. Rev. 120, 145-47, 150 (1972). Nor is our decision affected by the Connecticut legislature's grant on two occasions of other benefits to prisoners retroactively. See Conn. Gen. Stat. §§ 18-98a and 18-98b. To our knowledge, these statutes have not been challenged as violative of Connecticut's constitutional separation of powers. Moreover, even if the legislature might have granted some benefits retroactively it was not required to do so when it appeared to have a legitimate reason for denying them, as in this case. For the same reason we are unpersuaded by the decision of the Massachusetts legislature to confer enlarged good time benefits on prisoners sentenced prior to an act's effective date, for time served after that date, apparently also without challenge under Massachusetts constitutional separation of powers provision. See Comerford v. Commonwealth of Massachusetts, 233 F.2d 294, 295 (1st Cir.), cert. denied, 352 U.S. 899, 1 L. Ed. 2d 90, 77 S. Ct. 141 (1956).
We are not persuaded to deny rationality to the statute here under attack merely because the Assistant Attorney General of Connecticut has disavowed the legislative encroachment rationale for the challenged classification, apparently because he fears that such a ruling would jeopardize the legislature's power to grant retroactive relief when appropriate. The fact that retroactivity may be permissible under other circumstances does not preclude the legislature from denying it where a rational basis exists for such action. While it is reasonable to accord some deference to the executive's view of legislative intent in order to support a statute against constitutional challenge, we need not reject a plausible, legitimate purpose that rationally supports a statutory classification simply because, for tactical reasons, the state agent charged with enforcement of the statute disdains that purpose. We are particularly reluctant to do so when it appears that the highest court of the state impliedly subscribes to the rationale. Frazier v. Manson, supra, 410 A.2d at 480 (citing decisions of Rhode Island Supreme Court in Mastracchio v. Superior Court, supra, and Opinion to the Governor, supra).
"There is no requirement that two persons convicted of the same offense receive identical sentences." Williams v. Illinois, supra, 399 U.S. at 243. A legislature may prospectively reduce the maximum penalty for a crime even though prisoners sentenced to the maximum penalty before the effective date of the act would serve a longer term of imprisonment than one sentenced to the maximum term thereafter. Similarly, the legislature may award enlarged good time benefits prospectively only when such prospective application rationally furthers a legitimate state purpose. See Jackson v. Alabama, 530 F.2d 1231, 1238 (5th Cir. 1976); Jones v. Cupp, 452 F.2d 1091, 1093 (9th Cir. 1971); Comerford v. Commonwealth, 233 F.2d 294, 295 (1st Cir.), cert. denied, 352 U.S. 899, 1 L. Ed. 2d 90, 77 S. Ct. 141 (1956); United States ex rel. Sero v. Preiser, 372 F. Supp. 663, 670-71 (S.D.N.Y.), aff'd in part and remanded in part, 506 F.2d 1115 (2d Cir. 1974), cert. denied, 421 U.S. 921, 95 S. Ct. 1587, 43 L. Ed. 2d 789 (1975); Mirenda v. Ulibarri, 351 F. Supp. 676, 677 (C.D. Cal. 1972); Mastracchio v. Superior Court, 98 R.I. 111, 200 A.2d 10, 11, cert. denied, 379 U.S. 852, 13 L. Ed. 2d 55, 85 S. Ct. 96 (1964).
The order of the district court dismissing the writs of habeas corpus is affirmed.