Jones v. White

992 F.2d 1548

• Court: Court of Appeals for the Eleventh Circuit
• Decided: June 15, 1993
• Docket: Nos. 90-7129, 90-7295, 90-7394 and 92-6175
• Status: Published

Opinion

BIRCH, Circuit Judge:

In this consolidated appeal from denials of habeas corpus relief, we examine constitutional challenges to the Alabama Habitual Felony Offender Act (HFOA), which subjects a convicted felony defendant to life imprisonment without parole after the fourth Class A felony conviction. In these four cases, criminal defendants were sentenced under the HFOA. As habeas corpus petitioners, they allege that similarly eligible defendants were not so sentenced. The respective district courts concluded that the sentences did not result from an equal protection violation and denied habeas corpus relief. Following a thorough review of the records, we AFFIRM, although we base our reasoning for affirming the fourth case on grounds other than those relied upon by the district court.

I. BACKGROUND

A. Anthony Jones .

Petitioner Anthony Jones was tried and convicted of first-degree robbery in the Circuit Court of Macon County, Alabama on December 8, 1981. The prosecution appropriately advised Jones that it intended to proceed against him at sentencing as an habitual offender with evidence of four prior felony convictions. On January 20, 1982, Jones was sentenced as an habitual offender to a term of life imprisonment without the possibility of parole. His conviction was affirmed by the Alabama Court of Criminal Appeals on August 24, 1982. Jones v. State, 418 So.2d 955 (Ala.Crim.App.1982).

Jones then filed a petition for writ of habe-as corpus in federal court for the Middle District of Alabama. Among his grounds for relief was a claim that his habitual offender sentence was discriminatory. His petition was denied. Jones appealed to this court, which affirmed the district court’s denial of relief on all grounds except his recidivist sentence claim. We reversed the district court concerning Jones’s HFOA sentence claim by finding the assertion of an equal protection violation. Jones v. White, 811 F.2d 610 (11th Cir.1987) (per curiam). This court remanded the case to district court for an evidentiary hearing on Jones’s allegedly discriminatory sentence under the HFOA.

On remand, counsel was appointed for Jones and extensive discovery was authorized. On January 30, 1989, a magistrate judge conducted an evidentiary hearing on Jones’s claim that he was sentenced discrimi-natorily under the HFOA. Based on the evidence presented by Jones, the magistrate judge found that at least twelve persons sentenced in Macon County since the enactment of the HFOA had not received the correct or enhanced sentence. The magistrate judge noted that Macon County conducted only two terms of criminal court a year. Generally, cases subject to plea bargain agreements received early disposition in the court term. The prosecuting authorities in Macon County delegated the responsibility for obtaining information on defendants’ prior convictions to investigators. Because the investigators usually did not begin obtaining information on prior convictions until the approach of the criminal term, defendants who pled guilty in the early stages of the court term occasionally were advantaged. Information on prior convictions, often from other counties or states, was not available at the time the pleas were taken.

The magistrate judge observed that defendants who went to trial generally were subjected to the application of the HFOA. He specifically found that “[tjhere is absolutely no evidence, however, to suggest that persons are selected to be treated as habitual offenders because of their race or sex or religion or because they have chosen to exercise a particular constitutional right.” R2-99-10. With respect to Jones, the magistrate judge concluded that “[tjhere is no evidence to suggest ... that the decision to treat the petitioner as an habitual offender was motivated by a desire to punish the petitioner for exercising his right to trial by jury.” R2-99-15 n. 39. Therefore, the magistrate judge determined that, while the HFOA was not applied uniformly in Macon County, no evidence was presented to show that this result was based on any discriminatory reason violating a constitutional right. The magistrate judge accordingly recommended that Jones’s habeas corpus petition be denied.

The district court adopted the magistrate judge’s recommendation and denied Jones’s petition for a writ of habeas corpus. Jones appealed to this court the sole issue of whether he had been denied Fourteenth Amendment equal protection because of his claim of being selectively prosecuted under the HFOA. This court subsequently stayed Jones’s appeal pending the disposition of McLester v. Thigpen (McLester II), No. 87-T-00174-S (M.D.Ala. Feb. 24, 1992), the fourth case in this consolidated appeal.

B. William H. Mardis

Petitioner William H. Mardis and James Meeks were charged with the armed robbery of a drugstore in Montgomery, Alabama on July 22, 1981. Narcotics and money were stolen. At trial, a store employee positively identified Mardis as one of the robbers, and the prosecution presented an oral inculpatory statement by Mardis.

Prosecution witness Rhonda Thornton testified that she was with Mardis and Meeks when they reconnoitered the drugstore before the robbery, and that Mardis described his participation in the robbery to her and showed her drugs that he had taken from the store. Mardis’s counsel impeached Thornton’s credibility by showing that she was under indictment for selling cocaine, a drug-charge unrelated to the robbery; that she was codefendant Meeks’s girlfriend; and that she was a heavy drug user at the time of the robbery. At trial, Thornton was a suspect in several robberies and had a pending drug charge lodged against her. The prosecution agreed to accord Thornton youthful offender status in exchange for her testimony against Mardis. Although prior to and at the beginning of trial Mardis’s counsel requested that the prosecution produce any potentially impeaching evidence, the prosecution did not inform Mardis of this agreement.

Mardis was convicted of first-degree robbery in the Circuit Court of Montgomery County and sentenced under the HFOA to life without parole. His conviction was affirmed by the Alabama Court of Criminal Appeals. Mardis v. State, 423 So.2d 331 (Ala.Crim.App.1982). He filed two post-conviction petitions in the Alabama state courts. Each petition was denied, and the denials were affirmed by the Alabama Court of Criminal Appeals.

Mardis then petitioned in federal court for the Middle District of Alabama for a writ of habeas corpus. He presented the same two issues that he pursues in this court: the prosecution failed to disclose impeachment evidence concerning its agreement with Thornton and the HFOA was applied to him diseriminatorily at sentencing in violation of the Eighth and Fourteenth Amendments. A magistrate judge held an evidentiary hearing on March 6, 1990.

The magistrate judge determined that the prosecution’s failure to disclose the favorable treatment of Thornton was harmless error in view of the eyewitness identification and Mardis’s confession directly implicating him in the robbery. Regarding Mardis’s claim of discriminatory application of the HFOA, the parties stipulated to the factual records in the Jones case and in McLester’s first habe-as petition, both from the Middle District of Alabama. The factual and statistical presentations made in the evidentiary hearings in those cases were supplemented to the Mardis record. Mardis premised his claim that individuals who plead guilty are less likely to receive sentences under HFOA than those who elect to go to trial solely on statistical evidence. The magistrate judge concluded that Mardis had not met his burden of proof to show that the HFOA intentionally was applied in a discriminatory manner. He recommended that Mardis’s habeas corpus petition be denied. The district court adopted the magistrate judge’s recommendation and denied Mardis’s petition for writ of habeas corpus.

We briefly address Mardis’s claim that the prosecution failed to disclose Thornton’s favorable treatment. Under the harmless error standard, the crucial inquiry is whether the prosecution’s failure to- disclose potential impeachment evidence affected the judgment of the jury as to credibility. See Giglio v. United States, 405 U.S. 150, 154-55, 92 S.Ct. 763, 766, 31 L.Ed.2d 104 (1972). In a similar case concerning the prosecution’s failure to disclose the promise of a reward to a witness, this court concluded that the jury was unaffected because the witness’s credibility had been exposed to substantial impeachment by other evidence. McCleskey v. Kemp, 753 F.2d 877, 884 (11th Cir.1985) (en banc), aff'd, 481 U.S. 279, 107 S.Ct. 1756, 95 L.Ed.2d 262 (1987). Considering the overwhelming evidence of guilt, including eyewitness identification and Mardis’s confession as well as the jury’s exposure to substantial impeachment evidence concerning Thornton, we conclude that any error in the prosecution’s failure to disclose its agreement with Thornton was harmless.

Mardis’s appeal to this court was consolidated with Jones and Garrett. These three cases were stayed pending disposition of McLester II. Accordingly, we will incorporate our discussion of Mardis’s claim that the HFOA was applied diseriminatorily in sentencing him with our analysis of this issue relating to all of the cases involved in this appeal.

C. Larry Wayne Garrett

Petitioner Larry Wayne Garrett was indicted in the Circuit Court of Tallapoosa County on first-degree charges of receiving stolen property and burglary. When the case was called for trial on December 2,1985, and before the jury was selected, Garrett decided to plead guilty, pursuant to his counsel’s advice. During the plea proceeding, Garrett was informed of the rights that he would be relinquishing by entering a plea of guilty. Additionally, Garrett and his counsel completed an “Ireland Form,” which recites the rights surrendered by entering a guilty plea.

The plea proceeding transcript reveals that Garrett clearly and specifically was informed that he could be treated as an habitual offender, a category subjecting him to a potential life sentence without possibility of parole:

THE COURT: Okay. Have you ever been convicted of anything previously?

THE DEFENDANT: Yes, sir.

THE COURT: Okay. And tell me about those offenses.

THE DEFENDANT: I was convicted of receiving stolen property, and buying and receiving, theft of property the second and third degree cases and charges.

THE COURT: Second and third degree cases?

THE DEFENDANT: Yes, sir.

THE COURT: Okay. And what effect does that have under the Habitual Offender Law?

THE DEFENDANT: Your Honor, I understand the Habitual Offender Law when you are convicted more than twice. The third time throws you under. So I understand—

THE COURT: Okay. What is the specific effect of the convictions that he has?

MR. MARTIN [the prosecutor]: Judge—

THE COURT: Well, is there anything that will cause a mandatory sentence of life without parole?

MR. MARTIN: Judge we have notice of prior felonies. I want to be candid with the Court. There is some question as to identity. There [are] some legal questions of it. There may be as many as six.

THE COURT: Okay. Mr. Garrett, you are entitled to a separate sentencing hearing. It could be that your guilty plea will result in a mandatory sentence of life without parole; however, prior to that sentencing hearing, you and your attorney will have the opportunity to investigate all of your prior convictions to see if it is regular on its face, see if you were represented by a lawyer, see if there is anything else that would keep that — those prior convictions from being considered. Now, if they are legal convictions and if you were represented by a lawyer, I mil'impose sentence based on those prior convictions. And, if the appropriate sentence under the law, which I will study at that time, is life without parole, I will give you life without parole — if that’s what the law calls for. I’m going to abide strictly by the law in imposing sentence. And the State will be given the opportunity to present evidence of prior convictions at the sentencing hearing. Do you understand that?

THE DEFENDANT: Yes, sir.

THE COURT: Okay. And you wish to proceed with the guilty plea anyway?

THE DEFENDANT: Yes, sir.

Ex. 3-28-30 (Transcript of Plea Proceedings, Dec. 2, 1985) (emphasis added).

Prior to his sentencing, Garrett moved to withdraw his guilty plea. At the sentencing hearing on January 13, 1985, the court heard argument on Garrett’s motion to withdraw his plea. Because Garrett claimed that he had been misled by his retained attorney in pleading guilty, the court inquired of Garrett and his counsel. Garrett’s responses were not specific; he essentially represented that his attorney had encouraged him to plead guilty. Garrett’s attorney, Thomas Radney, explained that, based on his investigation that revealed two eye witnesses,¹ Garrett would have been convicted by a jury and he had no possible defense. Based on Garrett’s previous testimony at his plea proceeding² and the evidence presented to the court at sentencing, the trial judge refused to set aside Garrett’s plea.³ Because the state provided certification of Garrett’s four previous burglary convictions in Coosa County, the judge sentenced Garrett to life without parole as an habitual felony offender.⁴

Garrett’s conviction was affirmed on direct appeal by the Alabama Court of Criminal Appeals, and the Alabama Supreme Court denied his petition for writ of certiorari. Garrett then filed a petition for writ of habe-as corpus in the Middle District of Alabama. He claimed that his guilty plea was not knowing and intelligent, that his lawyer was ineffective, and that he was prosecuted selectively under the HFOA.

Garrett’s case was referred to a magistrate judge, who conducted two evidentiary hearings for which Garrett had appointed counsel. At the evidentiary hearings, Garrett testified that he had been informed that he would receive a sentence of fifteen years of imprisonment in exchange for pleading guilty. There is no other testimony or evidence in the record to support this allegation. To the contrary, the deputy district attorney who prosecuted Garrett testified that no offer of fifteen years was made to anyone in the case.

Garrett’s counsel testified that the state district attorney had offered to recommend that Garrett not be sentenced under the HFOA if he would identify his accomplice. He explained that Garrett rejected this offer. The district attorney testified, however, that he never offered to overlook the HFOA, and that the only offer made was life without possibility of parole.

At the evidentiary hearing, Radney, Garrett’s retained attorney for the plea proceeding, explained that he advised Garrett not to go to trial because the state’s evidence, including his prior felony convictions and the two eyewitnesses, was “overwhelming.” Ex. 3-18 (Testimony of Thomas Radney). Rad-ney conceded that he “couldn’t come up with any defense to plausibly put before a jury.” Id. at 19. He also testified that he did not tell Garrett at any time that he would not be sentenced as an habitual offender.

Garrett additionally presented computerized data from the Alabama Department of Corrections to show that the HFOA is not always implemented in Tallapoosa County. The parties supplemented this information by jointly filing a document showing fifty-three individuals who, although eligible, were not sentenced as habitual offenders. Garrett offered no explanation concerning the reasons that these fifty-three offenders escaped the strictures of the HFOA.

In his recommendation, the magistrate judge concluded that Garrett’s plea was knowing and voluntary. He found no basis for Garrett’s contention that his attorney forced him to enter his guilty plea. The magistrate judge also found Garrett’s ineffective assistance of counsel claim to be without merit. He concluded that the performance of Garrett’s counsel was not deficient at the sentencing proceeding, despite Garrett’s attempt to withdraw his plea. Although Garrett contended that he was in an adversarial relationship with his counsel at sentencing, the magistrate judge found that Garrett failed to show prejudice.

Garrett argued that he was sentenced selectively under the HFOA because he refused to reveal the name of his accomplice. While the magistrate judge found that Garrett had demonstrated that other offenders in Tallapoosa County had not been sentenced as habitual offenders, he determined that Garrett had failed to show an impermissible or discriminatory purpose. The magistrate judge specifically concluded that Garrett had failed to prove that he was sentenced unconstitutionally as an habitual offender.

Accordingly, the magistrate judge recommended that Garrett’s petition for writ of habeas corpus be denied. The district court adopted the magistrate judge’s recommendation. On appeal to this court, Garrett has pursued his arguments that his plea was not knowing and voluntary, that his counsel was ineffective, and that he was sentenced dis-criminatorily in violation of equal protection under the Fourteenth Amendment and the right to be free from cruel and unusual punishment under the Eighth Amendment.

At this point, we will address Garrett’s claims concerning the constitutionality of his guilty plea and ineffectiveness of counsel. In analyzing the constitutionality of a guilty plea, a reviewing court must determine that the plea represents “a voluntary and intelligent choice among the alternative courses of action open to the defendant.” Hill v. Lockhart, 474 U.S. 52, 56, 106 S.Ct. 366, 369, 88 L.Ed.2d 203 (1985). Additionally, the Supreme Court has determined that “the representations of the defendant ... [at a plea proceeding] as well as any findings made by the judge accepting the plea, constitute a formidable barrier in any subsequent collateral proceedings. Solemn declarations in open court carry a strong presumption of verity.” Blackledge v. Allison, 431 U.S. 63, 73-74, 97 S.Ct. 1621, 1629, 52 L.Ed.2d 136 (1977). This court has concluded that “[a] reviewing federal court may set aside a state court guilty plea only for failure to satisfy due process: If a defendant understands the charges against him, understands the consequences of a guilty plea, and voluntarily chooses to plead guilty, without being coerced to do so, the guilty plea ... will be upheld on federal review.” Stano v. Dugger, 921 F.2d 1125, 1141 (11th Cir.) (en banc), cert. denied, — U.S. -, 112 S.Ct. 116, 116 L.Ed.2d 85 (1991).

During Garrett’s plea proceeding, the state court judge carefully explained to Garrett the consequences of his plea. Garrett acknowledged that he understood that, by pleading guilty, he subjected himself to the possibility of being sentenced to life without parole under the HFOA. See Smith v. State, 494 So.2d 182, 182 (Ala.Crim.App.1986) (“If the defendant has prior felony convictions, he must be advised of the proper sentence to which he was subject under the Habitual Felony Offender Act.”). We find no coercion for Garrett to plead guilty by his attorney. Radney explained to Garrett prior to the plea proceeding the same consequences that the judge ascertained that Garrett understood before he pleaded guilty. Garrett was well awai’e of his prior felony convictions. Additionally, his contention that he was offered a fifteen-year sentence is not evidenced in the record. We find no merit to Garrett’s claim that his guilty plea was not knowing and voluntary; the record shows that Garrett understood the rights that he was waiving and the possibility of a life sentence without parole.

Garrett advances his ineffective assistance of counsel claim on two bases. First, he contends that Radney did not investigate properly Garrett’s prior felony convictions. Second, Garrett argues that Radney did not familiarize himself sufficiently with the HFOA to advise Garrett competently regarding the consequences and effect of his guilty plea.

The Supreme Court has held “that the two-part Strickland v. Washington test applies to challenges to guilty pleas based on ineffective assistance of counsel.” Hill, 474 U.S. at 58, 106 S.Ct. at 370. Under the Strickland test, Garrett initially must show “that counsel’s representation fell below an objective standard of reasonableness.” Strickland v. Washington, 466 U.S. 668, 688, 104 S.Ct. 2052, 2064, 80 L.Ed.2d 674 (1984). While it need not be errorless, counsel’s advice “must he within the realm of competence demanded of attorneys representing criminal defendants.” Stano, 921 F.2d at 1151. Providing meaningful advice on the options available to a defendant obligates counsel “to make reasonable investigations or to make a reasonable decision that makes particular investigations unnecessary.” Strickland, 466 U.S. at 691, 104 S.Ct. at 2066. Hill modified the second part of the Strickland test in the context of a guilty plea. Garrett “must show that there is a reasonable probability that, but for counsel’s errors, he would not have pleaded guilty and would have insisted on going to trial.” 474 U.S. at 59, 106 S.Ct. at 370.

Radney, Garrett’s retained counsel, was aware of Garrett’s prior felony convictions, which would result in his being sentenced as an habitual offender. Nevertheless, given the two positive eyewitness identifications and his inability to locate Garrett’s alleged alibi witness, Radney advised Garrett to plead guilty. In good faith, Radney determined that he could not present a colorable defense to a jury.

Regarding Radney’s performance at sentencing, Garrett has failed to show that his counsel did not require the state to meet its burden of proof. Not only did the prosecution give Garrett notice of its intention to proceed under the HFOA, but also introduced certified copies of four previous felony convictions. Since the state proved Garrett’s prior felony convictions, Radney could not prevent the court from sentencing Garrett as an habitual offender.

Although Garrett gained no advantage in his sentence by pleading guilty, he has not shown that the sentence would have been different had he been convicted at trial. He presented no evidence to demonstrate that his counsel’s assessment of the strength of the evidence against him was not correct, or that he would not have been convicted had he gone to trial. If convicted at trial, then Garrett’s sentence under the HFOA would have been the same, since it would have been predetermined by his extensive record of documented prior felony convictions. Thus, he was not prejudiced by his guilty plea. Garrett’s responses to the state court judge’s questions at his plea proceeding clearly evidence that he was aware that he would be sentenced under the HFOA. Therefore, Garrett has failed to demonstrate that his counsel’s advice was deficient as being objectively unreasonable.

Accordingly, Garrett’s first two claims have no merit. We consolidated and stayed the Jones, Mardis and Garrett eases until disposition of the McLester II case. Therefore, we reserve our analysis of Garrett’s alleged discriminatory sentence under the HFOA until after the following discussion of the McLester II ease.

D. Terry Wayne McLester

Petitioner Terry Wayne McLester was convicted after a jury trial for robbing a grocery store, the details of which are contained in our adjudication of his first habeas corpus petition. McLester v. Smith (McLester I), 802 F.2d 1330, 1331 (11th Cir.1986). Because McLester had seven prior convictions for burglary in the second degree,⁵ the state court trial judge sentenced him to life imprisonment without parole under the HFOA for his first-degree robbery conviction. The Alabama Court of Criminal Appeals upheld the conviction and sentence, McLester v. State, 423 So.2d 286 (Ala.Crim. App.1982), and the trial court’s dismissal of his petition for writ of habeas corpus or, alternatively, petition for writ of error eoram nobis, McLester v. State, 460 So.2d 870 (Ala. Crim.App.1984).

McLester then filed a petition for writ of habeas corpus in federal court for the Middle District of Alabama. The district court referred his petition to a magistrate judge, who recommended that McLester’s habeas corpus petition be denied. The district court adopted the magistrate judge’s recommendation. On appeal, this court affirmed the district court. McLester then filed a second habeas corpus petition, alleging that the application of the HFOA violates equal protection. As analyzed subsequently, we consider MeLester’s second habeas corpus petition to be an abuse of the writ pursuant to Rule 9(b) of the Rules Governing Section 2254 Cases in the United States District Courts.

The district court referred the second petition for habeas corpus relief to a magistrate judge, who conducted an evidentiary hearing on April 19-23, 1991. This evidentiary hearing provides the factual basis for this habeas corpus petition, and incorporates portions of the evidentiary records in the Jones, Mardis and Garrett cases, which have been consolidated with this case on appeal for the common issue presenting the putative unconstitutional application of the HFOA. Therefore, summarizing the testimony and other evidence presented is significant because it encompasses and analyzes the application of the HFOA statewide as opposed to specific counties used in Jones, Mardis and Garrett.

Generally, McLester called witnesses to testify concerning the problems in applying the HFOA evenly and fairly by evidence of the difficulties encountered in documenting prior felony convictions. The circuit clerk and registrar for Montgomery County explained that Montgomery County has used computerized docket sheets since 1984, and that it is one of sixteen Alabama counties to use computerized docket sheets. Locating felony convictions prior to 1984 in Montgomery County must be accomplished by a manual review of court records. Additionally, case action summaries do not reflect a plea bargain based on a reduced felony charge. If a convicted criminal defendant is sentenced to a term of incarceration in the state penitentiary, then a transcript of record is sent to the Alabama Department of Corrections. In contrast, the transcript of record for a convicted defendant, who receives a probationary sentence, is not sent to the Department of Corrections unless the defendant already is incarcerated on another conviction. When a convicted defendant receives a suspended sentence, the Montgomery County clerk’s office notifies the Department of Corrections even though the defendant does not serve time in the state penitentiary system.

A program analyst for the Alabama Department of Corrections testified concerning the information contained in the Department’s records. He confirmed that the records did not show probation or county jail time served. The Department of Corrections is advised only when a convicted defendant receives time in the state penitentiary. Additionally, if a defendant was convicted previously in federal or out-of-state cases, this information would not be evidenced in the Alabama Department of Corrections records. The program analyst stressed that the records of the Department of Corrections showed the number of Alabama incarcerations and not the number of prior convictions.

Several criminal defense attorneys practicing in Montgomery County also testified regarding their experiences with sentences where the HFOA was not applied. Prosecutors have negotiated plea bargains when problems were anticipated in obtaining certified out-of-state prior convictions that would have mandated application of the HFOA. In one instance, a convicted defendant received less than a life sentence by stipulating to one prior felony conviction, when his out-of-state convictions should have caused him to be sentenced to life without parole under the HFOA.⁶ Occasionally, prosecutors could not show prior felony convictions at sentencing because they were unable to obtain certified copies of prior convictions or use youthful offender convictions, or they could not show representation by counsel during the proceeding resulting in conviction.⁷

One attorney testified that he had waived a presentenee investigation report, where a plea bargain was reached based on an agreement with the prosecutor regarding the number of prior convictions. The attorney conceded, however, that the state invoked the HFOA in the “vast majority” of the cases in which the HFOA applied. R3-199. Another attorney explained that he did not believe that he was obligated to reveal to the prosecution the existence of prior felony convictions, which would necessitate a sentence under the HFOA, if the prosecution was aware of only one or two of these convictions. Accordingly, the convicted defendant would be sentenced on the basis of a lesser number of convictions.

Several district attorneys also testified at the evidentiary hearing. The district attorneys were aware that the HFOA was mandatory and that their offices had the responsibility to cheek the criminal background of a convicted defendant prior to sentencing and to use reasonable diligence to obtain admissible evidence of prior convictions. All the district attorneys were experienced prosecutors who shared similar difficulties in complying with their obligation to acquire evidence of defendants’ prior convictions, particularly those from other jurisdictions.

Michael Godwin, District Attorney for the Twenty-first Judicial Circuit, encompassing Escambia County, testified that his office routinely required the law enforcement investigating agency filing a felony report to provide his office with a “rap” sheet from the Criminal Justice Information Center, the state component of a national computer system. Additionally, his office used files from probation officers and requested the jail or sheriffs office to assist in acquiring FBI rap sheets. As opposed to taking a plea, a pre-sentence investigation is conducted before imposition of sentence. Godwin explained that FBI rap sheets occasionally showed arrests and indictments, but not ultimate or incorrect dispositions.

Obtaining certification of prior convictions could be problematic even within Alabama.⁸ Acquiring attestations and certifications outside of Alabama presented many recurring frustrations.⁹ On occasion, other jurisdictions had no record of information found on a rap sheet. Godwin’s understanding of reasonable diligence is using telegraphic or telephonic communications to request evidence of prior convictions, and he neither believes that the HFOA requires that he send personnel out of state to search through files, nor does he have the resources and budget to perform such investigations:

[I]t is a situation of stepping over dollars to pick up dimes. We can’t send someone all over the country and spend three or foui’ weeks on a case that we know what it is worth; we know that we have got these priors, we know that the defendant will admit it and we have got questions. You know, I don’t think any district attorney can just go to the end of the earth. It depends on the severity of the case. We don’t disregard that. But you just ...

You know, it is like any case. I would like to prosecute and convict every criminal in the state. But we don’t have the means to do that all the time. So, reasonable depends on a lot of things. It ... depends on your caseload, depends on ...

We have got seven capital murder cases in our county right now. We are going to have a lot less reasonable diligence in the next few months than I may have had three months ago. I mean, it is just a matter of manpower. We are going to pursue it as hard as we can, but on a case-by-case b[asisj. There are going to be some cases that may fall through the cracks. There are going to be some eases we are going to plea bargain that may not be the best thing in the world. But we are going to do the best we can.

R3-189-90. Godwin also explained that many of the prior convictions in Escambia County arise from nolo contendere pleas in Florida, which are not usable in Alabama regardless of adjudication and sentence, as proof of prior convictions for enhancement under the HFOA. Ex parte Jenkins, 586 So.2d 176, 177 (Ala.1991).

Godwin stated that his office had never disregarded the HFOA. While conceding that a charge might be reduced for plea bargaining purposes, Godwin testified that he had a duty under the HFOA to present prior convictions for sentencing. He stated that he was not aware of any Alabama district attorney who treated the HFOA “casually.” R3-171. Godwin testified that the HFOA, supported by Alabama district attorneys, is serving a useful purpose as an effective tool to prevent violent, serious felonies by recidivists:

I see numerous cases where people have committed serious crimes over and over and over again — I am talking about rapes, murders, burglaries, first-degree armed robberies — and they are back out on the street. And I think this act is designed to keep that type offender locked up. And certainly for a violent offender, I don’t think any would object, as far as district attorneys, that that is not ... or disagree that that is effective.

R3-171-72.

The District Attorney for the Tenth Judicial Circuit, comprising Jefferson County, testified that he considers the seriousness of the current charge against a convicted defendant in determining whether to expend resources to go beyond routine requests for evidence of prior convictions. He explained that “[y]ou may have difficulty proving a prior, one prior or two priors or three or whatever out of the number that you have. But that does not mean we are not applying. We are applying the Habitual Offender Act as far as we can prove it.” R3-211. The Chief Assistant District Attorney for Calhoun and Cleburne Counties testified that the prosecutors in that office would be more likely to use their resources to obtain evidence of prior convictions in a case involving a more serious Class A rather than a Class C felony.¹⁰ Without objection, McLester introduced into evidence a partial transcript of the testimony of Garrett’s counsel in the Garrett case consolidated in this appeal. This testimony concerned the attorney’s understanding that the district attorney, who allegedly offered not to prove Garrett’s prior convictions necessitating application of the HFOA if Garrett revealed the name of his accomplice, made similar offers in other cases.

Thomas Sorrells, the former district attorney of Houston County, prosecuted McLes-ter and testified at the evidentiary hearing. It was his policy to invoke the HFOA without exception. He explained, however, that a prosecutor did not have proof of a prior conviction for the purpose of the HFOA if he did not have a certified prior conviction, or if the record of the prior conviction did not show that the defendant was represented by counsel. Since his office disposed of two thousand felonies a year, resources might further limit efforts that could be expended in obtaining prior convictions in other jurisdictions. Overall, however, Sorrells stated that the HFOA was beneficial to the citizens of Alabama by reducing violent crime by recidivists. Sorrells also testified that the “nondiscretionary feature” of the HFOA made it “effective.” R5-502. He explained that “the nondiscretionary feature of the act is very important; that you can’t do it to one person and not do it to the other person; that you have got to be fair as far as you can be, to the best of your ability.” Id. at 504. Sorrells specifically testified that he invoked application of the HFOA in McLester’s case not to single him out, but because he understood that he had no discretion under the mandatory aspect of the statute.

Statistical evidence also was presented at the evidentiary hearing. The Alabama Department of Corrections provided the parties with a computer data base, consisting of a substantially complete record of Alabama inmates incarcerated for felony convictions, excluding death penalty inmates, beginning in 1981, when the Department of Corrections began keeping its computer file, until mid 1990, the date that the computerized information was compiled. This information did not show the number of prior felony convictions, prior out-of-state convictions or Alabama county incarcerations, whether a conviction had been reversed on appeal, whether a sentence was served as a youthful offender, or whether an inmate was serving a life without parole sentence. The Department of Corrections made no attempt to verify records received from Alabama circuit courts and conducted no ongoing verification process concerning the data that it received.

Within the data base obtained from the Department of Corrections, the parties noted specific problems precluding the potential application of the HFOA.¹¹ First, the data contains no record of felony convictions that did not lead to incarceration in an Alabama penitentiary. Second, there generally was no indication of felony convictions outside of Alabama. Third, the dates of offenses were not listed, preventing determination of whether the offense was committed prior to the time that the HFOA was enacted. Fourth, the data indicated only the part of a split sentence representing incarceration in an Alabama institution. Statistical expert Dr. Chester Palmer testified that the overall effect of these problems in the data base was that prior convictions for inmates were un-dercounted.

In an attempt to resolve these described problems, the parties decided to review a random sampling of criminal files and entered into a stipulation agreeing to accept the results as a fair representation.¹² Palmer prepared a computer program designed to determine whether convicted defendants were sentenced correctly under the HFOA. He found the overall error rate in the random sample for application of the HFOA to have been 28.2 percent.¹³

Following his review of the complete computer data base and the random sample, Palmer formulated several conclusions: (1) the error rate for sentencing Class B offenses was higher than that for Class C offenses, and probably higher than that for Class A offenses,¹⁴ (2) the error rate for sentencing offenses with more than one prior probably was higher than that for sentencing offenses with one prior, and (3) the error rates for sentences given in Houston County and Madison County probably were lower than the error rate for the entire state. Where convicted defendants with two or more prior convictions were not sentenced under the HFOA, Palmer did not conclude that the HFOA was not considered. R4-274. Rather, he found that the HFOA was applied incorrectly because the court was not made aware of all of the prior convictions. Id. at 274-75. Overall, 30.1 percent of the sentences were too short, or not in accordance with the HFOA. R4-270. Palmer conceded that “Statistical analyses do not ordinarily show cause. They show patterns.”¹⁵ R4-314.

Dr. John Sloan, an expert in statistics and criminology, participated in the selection and examination of the random files.¹⁶ After deleting files for various reasons, Sloan analyzed 307 of the 400 randomly selected files. For the period 1981 to 1990, he determined that the error rate for application of the HFOA to the sentencing of eligible convicted defendants increased with the seriousness of the crime.¹⁷ Id. at 342. For all classifications of felonies, he also found that, as the number of prior convictions increased, there was a greater probability that an eligible convicted defendant would not be sentenced under the HFOA.¹⁸ Sloan determined that 28 percent of the overall sentences were short, or not in compliance with the HFOA. Id. at 344. Therefore, in approximately 72 percent of sentencings for eligible defendants, the HFOA was applied appropriately. Id. at 368. Expressed another way in his written report, Sloan states that “more than 1 out of 4 felons who loere eligible for enhanced sentences failed to receive them during the 10 year period. The percentage of those failing to receive enhanced sentences may be as high as 33%.” Petitioner’s Ex. 21 at 7. Sloan observed that the error rate of application of the HFOA had decreased during the latter half of the decade, indicating that the state was improving in applying the HFOA. R4-355-56.

In conjunction with his statistical analyses, Sloan testified about the purposes of the HFOA as a recidivist statute. After reviewing the legislative history of the HFOA, Sloan concluded that the Alabama legislature designed the statute for the dual purposes of deterrence and selective incapacitation.¹⁹ Id. at 363-64. Because perception of being sentenced as an habitual offender can deter the potential felony recidivist, Sloan concluded that “misapplication of the Act cuts much more to the deterrence component of the intent of the legislature than it does to the incapacitation.”²⁰ Id. at 366. Given the 28 percent of the cases where the HFOA is not applied appropriately, Sloan testified “[t]hat leaves a 72 percent rate of certainty,” and “when the certainty of punishment begins to approach ... 75 to 80 percent in favor of receiving that punishment, then deterrence definitely works.” Id. at 368.

Sloan distinguished disparate and discriminatory sentences for the court:

[Disparity would be where we have two individuals who have been convicted of the same offense and they receive different sentences.

Discrimination would be where we have two individuals who receive ... who have committed the same offense and they receive different sentences because of either the personal characteristics of one of the offenders, say, he or she is black or female or whatever; there is some invidious distinction being made based on race or class or gender or country of origin or something like that, that would be discrimination.

The other example is where we would have two offenders who have committed the same offense, say they are both black or both white or both women or something like that, and they would get different sentences because of their ... the background in the case, say____ That is disparity. But not necessarily discrimination.

Id. at 391-92 (emphasis added). Although he found irregularities in the implementation of the HFOA in Alabama,²¹ Sloan did not find sufficient evidence to justify invalidating the operation of the recidivist statute that had been effective for ten years. Id. at 403. His analyses showed trends based on data consisting of length of sentence terms, number of prior convictions, and dates; he did not review the case files examined for personal characteristics, which would have been necessary to reveal discriminatory application of the HFOA. Id. at 421-22. In his written report, Sloan conceded that “[p]ossible conclusions, such as the Act is ‘racist’ or ‘sexist’ in its application, are beyond the scope of the current analyses.” Petitioner’s Ex. 21 at 11.

The purpose of the evidentiary hearing-held in this case was to provide a factual basis upon which the magistrate judge could determine whether McLester’s equal protection rights under the Fourteenth Amendment had been violated by the application of the HFOA. Following the evidentiary hearing regarding this claim, the magistrate judge entered an extensive report recommending that the district court deny McLes-ter’s petition for habeas corpus relief. The district court adopted the recommendation of the magistrate judge, and denied McLester’s petition for writ of habeas corpus. On appeal, MeLester pursues his equal protection claim, contending that state officials intentionally manipulated the HFOA so that it was not applied, failed to apply the HFOA through ignorance or deliberate indifference, or did not apply the HFOA mandatorily because of failure to implement stringent administrative procedures to insure its application.

II. DISCUSSION

A. Abuse of the Writ

Reviewing McLester’s first petition for ha-beas corpus relief, a previous panel of this court determined that his sentence as an habitual offender under the HFOA did not violate the Eighth Amendment since it was not disproportionate under the three-part test of Solem v. Helm, 463 U.S. 277, 290-92, 103 S.Ct. 3001, 3010-11, 77 L.Ed.2d 637 (1983). McLester I, 802 F.2d at 1332-33. Additionally, the court considered McLes-ter’s claim that the district court abused its discretion by denying his discovery request for “the names, prior criminal records, and the sentences of all persons convicted of the highest degree of murder, sodomy, kidnapping, and rape since July, 30,1979, who were sentenced to less than life without parole.” Id. at 1334. MeLester contended that the requested data would show that “a large number” of these individuals “had received lesser sentences than he in spite of the fact that they are second, third, and fourth time habitual offenders.” Id. In MeLester I, we found that the records of the Alabama Department of Corrections were made available to MeLester, and that manual examination of thousands of prisoners’ files would be too burdensome to require.

That panel stated that “[n]o contention was made attacking the constitutionality of the sentence as applied to MeLester based on the state’s stipulation and evidence in the record showing that Alabama has lacked uniformity in application of this supposedly mandatory recidivist statute.” Id. The panel further commented that “[sjince the statute is mandatory, the state of Alabama has a serious problem regarding its recidivist statute that we will be requested to address in some other case or in further proceedings in this case when the issue is squarely presented.” Id. at 1333 n. 2. Apparently actualizing this statement, McLester filed under 28 U.S.C. § 2254 the subject habeas corpus petition alleging as new issues that his sentence of life without parole denied him equal protection under the Fourteenth Amendment.

The Supreme Court has emphasized that “[ujnless a habeas petitioner shows cause and prejudice, see Wainwright v. Sykes, 433 U.S. 72, 97 S.Ct. 2497, 53 L.Ed.2d 594 (1977), a court may not reach the merits of ... new claims, not previously raised which constitute an abuse of the writ, McCleskey v. Zant, 499 U.S. -, 111 S.Ct. 1454, 113 L.Ed.2d 517 (1991).” Sawyer v. Whitley, - U.S. -, -, 112 S.Ct. 2514, 2518, 120 L.Ed.2d 269 (1992). Where a petitioner was aware of the factual and legal basis of a claim not presented when the first federal habeas petition was filed, he must demonstrate that the failure to present the claim in the prior proceeding was not because of intentional abandonment, deliberate withholding, or inexcusable neglect under the abuse of the writ doctrine. Funchess v. Wainwright, 788 F.2d 1443, 1445 (11th Cir.) (per curiam), cert. denied, 475 U.S. 1133, 106 S.Ct. 1668, 90 L.Ed.2d 209 (1986).

McLester was sentenced as an habitual offender on February 26,1982. Both the district court and this court granted McLes-ter’s motions to supplement his second habe-as petition, McLester II, with the entire record from McLester I. If the evidence from his first habeas petition supports his second habeas petition, then the issue of the discriminatory application of the HFOA could and properly should have been raised in his first habeas petition. McLester had his opportunity to address the constitutionality of the HFOA in his first habeas petition, but he failed to do so. Additionally, in his second habeas petition, McLester has attempted to demonstrate the existence of a discriminatory pattern or practice that existed at the time that he was sentenced by evidence that has occurred predominately in the years since his sentencing. In our judgment, McLester has not met the threshold requirement of cause and prejudice, and raising his constitutional challenge concerning the application of the HFOA in his second habeas petition is an abuse of the writ. His second habeas petition should have been denied on that basis, and the district court should not have proceeded to address his equal protection claim relating to the application of the HFOA.

McLester apparently interpreted the prior panel’s remarks, or dicta, concerning a constitutional challenge to the HFOA in his first habeas petition before this court as a suggestion or invitation that he pursue a second habeas petition. In his certificate for probable cause, which was granted, McLester states that “Petitioner requests the opportunity to pursue the Eleventh Circuit Court of Appeals invitation to ‘squarely present’ his equal protection claim.” R2-62-2. In reviewing McLester’s second petition for habe-as corpus relief, the magistrate judge considered himself bound by our former panel’s commentary to proceed to the merits of the constitutionality issue rather than to require a showing of cause and prejudice, and thereby ignored the dictates of Rule 9(b). Additionally, the magistrate judge, citing other cases pending both in the district courts and in this court on the constitutionality of the HFOA, appears to recognize an exception to Rule 9(b) for claims that are of special interest to the court system at a particular time, justifying McLester’s failure to raise his constitutionality claim when he brought his first habeas petition.

Significantly, the district court in its final order adopting the magistrate judge’s reasoning stated:

The court would simply add that, because it agrees that the petition for writ of habeas corpus is due to be denied on the merits, the court does not reach respondents’ abuse-of-the-writ claim. Indeed, it could be argued that because the petitioner’s claim lacks merit, he has not satisfied the prejudice prong of the cause-and-prejudice requirement to defeat abuse-of-the-writ defense.

R2-61-1-2 (citation omitted). By reviewing the merits of this second habeas petition before determining cause and prejudice, the district court transposed the order of review and acted contrary to Sawyer and McCles-key. The magistrate judge found that the costs and time expended in preparing for the evidentiary hearing justified proceeding to the merits, or a poor use of judicial resources would have occurred. Had this second habe-as petition been reviewed properly, the time and expense of the preparation for the evi-dentiary hearing would have been prevented.

Dicta from this court is not authority for district courts to circumvent Supreme Court mandates or federal rules of procedure. Additionally, “there should be a presumption that judges mean to do no more than to decide the case before them.... [F]or law-of-the-circuit purposes ... [the review of any precedent] ought to focus far more on the judicial decision than on the judicial opinion.” New Port Largo, Inc. v. Monroe County, 985 F.2d 1488, 1500 (11th Cir.1993) (Edmondson, J., specially concurring) (emphasis added). Mere speculation that this court may be “requested to address” a claim in “further proceedings” should not be considered as an invitation to ignore Rule 9(b). McLester I, 802 F.2d at 1333 n. 2. The federal courts have an ongoing policy in determining habeas corpus cases to remind state courts that, in order for state procedural rules to be considered independent and adequate state grounds for not reaching the merits of a claim, the state courts consistently must apply their own rules. It would be curious indeed for federal courts to disregard the preclusionary effect of Rule 9(b).

We want to be explicit so that an inappropriate signal is not sent to future habeas corpus petitioners or to district courts. We cannot approve and we discourage the review process and procedure that transpired in the district court in McLester’s second habeas petition. His second habeas petition should have been denied as abuse of the writ, and the district court should not have proceeded to a consideration of the merits.

B. Record Supplementation on Appeal with Evidence Not Considered by the District Court

Although we have decided that McLester II is an abuse of the writ, this court consolidated the other three habeas petitions in this appeal to resolve the issue of whether the HFOA is constitutionally applied based on the extensive evidentiary hearing that occurred in McLester II.²² Consequently, we address the threshold question that generally is answered in the negative: Can the record in a case on appeal be supplemented with the evidence or record in another case not considered by the district court? In a limited situation, such supplementation is appropriate.

“This court’s inherent equitable powers allow it to supplement the record with information not reviewed by the district [court], though this power is not often exercised.” Cabalceta v. Standard Fruit Co., 883 F.2d 1553, 1555 (11th Cir.1989); see Dickerson v. Alabama, 667 F.2d 1364, 1367 (11th Cir.) (“While federal appellate courts do not often supplement the record on appeal with evidence not reviewed by the court below, it is clear that the authority to do so exists.”), cert. denied, 459 U.S. 878, 103 S.Ct. 173, 74 L.Ed.2d 142 (1982); see also First Alabama Bank v. Parsons Steel, Inc., 825 F.2d 1475, 1487 (11th Cir.1987) (“This Court has the discretionary power to supplement the record on appeal, even to include evidence not reviewed by the court below.”), cert. denied, 484 U.S. 1060, 108 S.Ct. 1015, 98 L.Ed.2d 980 (1988). We have not allowed supplementation when a party has failed to request leave of this court to supplement a record on appeal or has appended material to an appellate brief without filing a motion requesting supplementation. Cabalceta, 883 F.2d at 1555; Ross v. Kemp, 785 F.2d 1467, 1474-75 (11th Cir.1986). By their joint motion, supplementation of the Jones, Mardis and Garrett records on appeal with the McLester II eviden-tiary hearing, including the testimony and exhibits, as well as the decision in that case had been properly requested. We granted the joint motion, which was not opposed by the state.

The Supreme Court has left to the discretion of the circuit courts when to consider matters for the first time on appeal to be determined on the facts of individual cases: “We announce no general rule. Certainly there are circumstances in which a federal appellate court is justified in resolving an issue not passed on below, as where the proper resolution is beyond any doubt, or where ‘injustice might otherwise result.’ ” Singleton v. Wulff, 428 U.S. 106, 120-21, 96 S.Ct. 2868, 2877, 49 L.Ed.2d 826 (1976) (deciding issues raised for the first time on appeal) (citations omitted). In determining whether to supplement the record in a given case, this court has established three considerations:

(1) whether “acceptance of the proffered material into the record would establish beyond any doubt the proper resolution of the pending issue,”; (2) whether remand of the case would be contrary to the interests of justice and judicial economy; (3) whether the inherent judicial powers of the court in habeas corpus actions dictate supplementation.

Cabalceta, 883 F.2d at 1553 (quoting Ross, 785 F.2d at 1475); see Singleton, 428 U.S. at 121, 96 S.Ct. at 2877.

We are persuaded that the supplementation of the records of the other three cases included in this consolidated appeal satisfies all three considerations. The evidence presented in the McLester II eviden-tiary hearing, the analysis of the Alabama Department of Corrections data for a ten-year period, enables us to determine whether there has been a violation of equal protection in Alabama’s application of its mandatory HFOA. See Dickerson, 667 F.2d at 1367 (“[T]he proper resolution of the substantive issues in this case, when viewed in the context of all of the relevant historical facts, is beyond any doubt.”). The state concedes the importance of the resolution of the equal protection issue involved in this consolidated appeal to the functioning of Alabama’s criminal justice system:

These cases [Jones, Mardis and Garrett] all involve the application of Alabama’s habitual felony offender act, § 13A-5-9, Code of Alabama (1975). Counsel for the Appellants in the three above cases requested the cases be stayed pending the resolution of this case [McLester II] as this ease developed more detailed information concerning the habitual offender act than had been developed at the trial level in these prior cases. The McLester case was tried over a three day period and the record on appeal exceeds 500 pages of trial testimony. Included in the trial testimony is testimony from two expert witnesses concerning the findings from several statistical studies on the issues presented. Because of the extensive testimony taken in this case and because the issue is of extreme importance to the criminal judicial system within the State of Alabama, [the state sought permission to exceed the page limitations in their brief so that] the testimony [can] be fairly summarized and the issues properly addressed.

Appellees’ Motion for Permission for Brief to Exceed Page Limitation at 2-3 (emphasis added).

Although the time, effort and money expended in the McLester II evidentiary hearing should not have been spent because the second petition for habeas corpus relief in that case was an abuse of the writ, the hearing did take place. The same evidence, applicable to all four cases in this consolidated appeal or future cases raising the issue of the constitutional application of the HFOA, should not be lost or repeated in another case, which would constitute a waste of expert and judicial resources. The “one dis-positive issue in this appeal raises a pure question of law” which requires evidentiary documentation to determine application of the HFOA; these statistics would be the same in these or any other potential cases on this issue. Dickerson, 667 F.2d at 1367.

Additionally, these consolidated cases are habeas cases, and we have unique appellate powers in the context of habeas corpus actions under 28 U.S.C. § 2254(a). Id. at 1368 & n. 7. Because these cases, consolidated by this court for determination of the common issue of the constitutional application of the HFOA, present the restricted circumstances in which evidence not considered by the district court may be considered on appeal, we supplemented the records of the Jones, Mor-áis and Garrett cases with the evidentiary hearing in McLester II. Although unstated, this reasoning appears to be implicit in this court’s granting the motions of the respective appellants to consolidate and hold their cases in abeyance pending the decision and appeal in McLester II.

C. Constitutionality of the Application of the HFOA

The Jones, Mardis and Garrett habeas petitions were consolidated on appeal to determine the common issue of whether the HFOA is being applied in Alabama to violate the equal protection rights of convicted defendants at sentencing. Our analysis of the constitutionality of the HFOA will include all aspects of the equal protection issues raised by the various petitioners. The referenced facts derive from the evidentiary hearing conducted in the McLester II case because this evidence is applicable to all of the petitions and the reason for which the first three cases were consolidated with McLester II.

Effective July 30,1979, the Alabama habitual offender statute at issue in this consolidated appeal provides:

(a)In all cases when it is shown that a criminal defendant has been previously convicted of any felony and after such conviction has committed another felony, he must be punished as follows:

(1) On conviction of a Class C felony, he must be punished for a Class B felony;

(2) On conviction of a Class B felony, he must be punished for a Class A felony;

(3) On conviction of a Class A felony, he must be punished by imprisonment for life or for any term of not more than 99 years but not less than 15 years.

(b) In all cases when it is shown that a criminal defendant has been previously convicted of any two felonies and after such convictions has committed another felony, he must be punished as follows:

(1) On conviction of a Class C felony, he must be punished for a Class A felony;

(2) On conviction of a Class B felony, he must be punished by imprisonment for life or for any term of not more than 99 years but not less than 15 years;

(3) On conviction of a Class A felony, he must be punished by imprisonment for life or for any term of not less than 99 years.

(c) In all cases when it is shown that a criminal defendant has been previously convicted of any three felonies and after such convictions has committed another felony, he must be punished as follows:

(1) On conviction of a Class C felony, he must be punished by imprisonment for life or for any term not more than 99 years but not less than 15 years;

(2) On conviction of a Class B felony, he must be punished for life in the penitentiary;

(3) On conviction of a Class A felony, he must be punished by imprisonment for life without parole.

Ala.Code § 13A-5-9 (1982). The primary goals of the HFOA are “ ‘to deter repeat offenders and, at some point in the life of one who repeatedly commits criminal offenses serious enough to be punished as felonies, to segregate that person from the rest of society for an extended period of time.’ ” Lidge v. State, 419 So.2d 610, 614 (Ala.Crim.App. 1982) (quoting Rummel v. Estelle, 445 U.S. 263, 284, 100 S.Ct. 1133, 1114-45, 63 L.Ed.2d 382 (1980)), writ denied, 419 So.2d 616 (1982).

Discussing the HFOA in McLes-ter’s first petition for habeas corpus relief, we explained that subsequent felony eonvic-tions trigger operation of the statute and that the enhanced penalty is “determined by the number and class [Class A, B or C] of prior felonies as well as the class of the felony which invokes the imposition of the statute.”²³ McLester I, 802 F.2d at 1332. We also have “reaffirmed our view that the Alabama statute [HFOA] is constitutional insofar as it augments sentences for a particular conviction because of past criminal convictions.” Williams v. Johnson, 845 F.2d 906, 911 (11th Cir.1988). The HFOA does not violate the Eighth Amendment because it does not produce a sentence that is grossly disproportionate to the severity of the crime. Seritt v. Alabama, 731 F.2d 728, 730-37 (11th Cir.), cert. denied, 469 U.S. 1062, 105 S.Ct. 545, 83 L.Ed.2d 433 (1984); see Thompson v. State, 462 So.2d 777, 781 (Ala. Crim.App.1984) (“The Habitual Felony Offender Act does not constitute cruel and unusual punishment, is not violative of due process of law and is not an ex post facto law.”).

We further have recognized that the HFOA is mandatory. McLester I, 802 F.2d at 1332. “ ‘The word “must” as it is used in this section leaves no discretion with the court as to whether a repeat offender is to be punished under the statute.’ ” Id. (quoting Watson v. State, 392 So.2d 1274, 1276 (Ala. Crim.App.1980), writ denied, 392 So.2d 1280 (Ala.1981)). Concomitantly, the prosecutor, as an officer of the court, has no discretion about whether to inform the court of a convicted defendant’s prior convictions about which he is aware so that the court can sentence the individual in accordance with the clear mandate of the statute.²⁴ Maye v. State, 472 So.2d 688, 690 (Ala.Crim.App. 1985); Aplin v. State, 421 So.2d 1299, 1302 (Ala.Crim.App.1981), overruled on other grounds, Chambers v. State, 522 So.2d 313 (Ala.1987); Gray v. State, 455 So.2d 163, 167 (Ala.Crim.App.1984); see McClaren v. State, 500 So.2d 1325, 1326-27 (Ala.Crim.App.1986) (“The Act [HFOA] must be applied as the law of the state by district attorneys, who have an affirmative duty to inform the courts of the existence of prior felony convictions.”).

Because the statute does not specify, the proof necessary to show previous convictions under the HFOA has evolved since its enactment.²⁵ In the early 1980’s, prior convictions were proved by introducing a certified copy of the minute entry that established that the defendant was represented by counsel or that he waived his right to counsel at the time of his conviction. Thatch v. State, 397 So.2d 246, 252 (Ala.Crim.App.), writ denied, 397 So.2d 253 (Ala.1981). Prior felony convictions properly were proved by a certified copy of the minute entry showing a valid, prior felony conviction, or by an admission of the defendant. Thomas v. State, 435 So.2d 1319, 1323 (Ala.Crim.App.1981), rev’d on other grounds, 435 So.2d 1324 (Ala.1982). A certified copy of bench notes was insufficient to prove a prior conviction or to enhance a sentence under the HFOA.²⁶ Id.

By the mid 1980’s, a defendant’s admission of a prior felony conviction was a proper method of proof under the HFOA. Faircloth v. State, 471 So.2d 485, 493 (Ala.Crim.App. 1984), aff'd, 471 So.2d 493 (Ala.1985); Travis v. State, 484 So.2d 1148, 1150 (Ala.Crim.App. 1985). When the defendant admitted a prior felony conviction, the necessity for notice under the HFOA was obviated, and the state was not required to prove that the defendant was represented by counsel or had waived counsel during the prior criminal proceedings. Pickens v. State, 475 So.2d 637, 639-40 (Ala.Crim.App.1985); Wilson v. State, 428 So.2d 197, 201 (Ala.Crim.App.1983). Subsequently, it was held that a prior felony conviction was not proven properly where there was no judgment or minute entry indicating conviction of the prior charge. Esters v. State, 480 So.2d 615, 616-18 (Ala.Crim.App. 1985).

Effective on July 23, 1987, the Alabama legislature codified the proper methods for proving a prior conviction for purposes of enhancement under the HFOA. In pertinent part, the statute provides as follows: “Certified copies of case action summary sheets, docket sheets or other records of the court are admissible for the purpose of proving prior convictions of a crime, if the prior conviction is otherwise admissible under the laws of this state.” Ala.Code § 13A-5-10.1 (Supp.1992). After this statute was enacted, Alabama courts held that prior convictions could be proved by a certified minute entry, a certified judgment entry, or by the defendant’s admission of the prior conviction. Brooks v. State, 520 So.2d 195, 200 (Ala.Crim. App.1987).

Evidence of prior convictions submitted pursuant to section 13A-5-10.1 must meet Alabama attestation requirements. The proper certification for documents from Alabama counties requires attestation by the custodian of the public documents and an official seal or certification by the record keeper.²⁷ Proving prior felony convictions from another jurisdiction requires the attestation of the clerk and seal of the court, plus a certificate from the judge, chief judge or presiding magistrate judge that the attestation is correct.²⁸ Consequently, proof of pri- or felony convictions failing to comply with the requirements for attestation from another jurisdiction cannot be considered for sentence enhancement under the HFOA. Webb v. State, 539 So.2d 343, 351 (Ala.Crim.App. 1987); see Carlisle v. State, 533 So.2d 645, 649 (Ala.Crim.App.1987) (This case was remanded for new sentencing because a conviction exhibit did not contain the authenticating certificate by a judge, chief judge or presiding magistrate.); McBride v. State, 480 So.2d 619, 620 (Ala.Crim.App.1985) (Certified copies of judgment entries from another state to enhance the sentence of an habitual offender were improper where the prosecution failed additionally to produce certificates from a judge, chief judge or magistrate judge verifying that the attestations were correct.).

In addition to determining proper proof of prior felony convictions, Alabama prosecutors and courts have been confronted with a statutory anomaly for drug convictions. Until September 30, 1988, Alabama statutes relating to the unlawful use or distribution of controlled substances were codified at Title 20 of the Alabama Code. In 1988, drug-related offenses were transferred to Title 13A, the criminal code of Alabama. Because the HFOA mandated application upon the proof of prior conviction “of any felony,” Alabama courts through 1986 considered all prior felony convictions to come within the purview of the HFOA, regardless of their origin. Gwynne v. State, 499 So.2d 802, 809 (Ala.Crim.App.1986); see Esters, 480 So.2d at 617-18 (A conviction in another state that would be a felony conviction in Alabama and all court martial convictions under the Uniform Code of Military Justice are prior convictions for purposes of the HFOA.); Chambers v. State, 418 So.2d 948, 950 (Ala.Crim. App.1982) (Three prior guilty pleas constituted separate felony convictions under the HFOA.).

In 1987, however, the Alabama Supreme Court decided that, because the Controlled Substances Act provided its own penalties, the punishment provisions of Title 13A, including the HFOA, would not apply to drug offenses. Ex parte Chambers, 522 So.2d 313, 316 (Ala.1987); see Ex parte Brannon, 547 So.2d 68, 69 (Ala.1989) (A defendant convicted of felony drug possession with prior convictions that were not drug related could not be sentenced under the HFOA.). In 1988, the Alabama legislature transferred the provisions of the Controlled Substance Act to Title 13A, §§ 13A-12-201 to 284. The legislature clarified that “[t]he provisions of this Act are to be included in the Code of Alabama 1975, as a part of Title 13A, “Alabama Criminal Code”, and all provisions of Title 13A, including the Habitual Felony Offender Act, are applicable thereto.” 1987 Ala.Acts, No. 87-603 § 11. Because article 5 of chapter 12 of Title 13A, the transferred sections of the criminal code relating to drug offenses, became effective on September 30,1988, conduct occurring after that date is governed by that article and is subject to enhancement under the HFOA. Drug offenses that occurred before September 30, 1988, are controlled by preexisting law.

Thus, the acquisition of admissible proof of prior convictions for sentencing purposes under the HFOA differs according to the time period in question. The proof or showing of prior felony convictions, stated simply in the HFOA, is actually a complicated, intricate process. It is dependant upon the cooperation and assistance of others for attestation, and potentially fraught with obstacles for the most conscientious prosecutor and sentencing judge in view of the volume of criminal cases handled by Alabama courts.

Given the mandatory application of the HFOA and the problems encountered by prosecutors and courts in obtaining admissible evidence of prior felony convictions, we turn to the equal protection issue presented to us in this consolidated appeal. Collectively, the petitioners assert a claim of selective prosecution or disparate sentencing treatment in violation of equal protection. Rickett v. Jones, 901 F.2d 1058, 1061 (11th Cir.1990). They specifically contend that they were sentenced to life without parole under the HFOA, while other similarly situated or equally eligible convicted defendants were not given the same mandatory HFOA sentence.

To show selective prosecution, each petitioner has the “heavy burden” of meeting two requirements. Owen v. Wainwright, 806 F.2d 1519, 1523 (11th Cir.1986) (per curiam), cert. denied, 481 U.S. 1071, 107 S.Ct. 2466, 95 L.Ed.2d 875 (1987).

First, he must make a prima facie showing that he has been singled out for prosecution although others similarly situated who have committed the same acts have not been prosecuted. Second, having satisfied the initial burden, he must then demonstrate that the government’s selective prosecution of him has been constitutionally invidious. The invidiousness requirement is met if the defendant establishes that the government’s selective prosecution is motivated by constitutionally impermissible motives such as racial or religious discrimination or his exercise of constitutional rights.

Id. (citations omitted); accord United States v. Gordon, 817 F.2d 1538, 1539 (11th Cir. 1987) (per curiam), vacated in part on other grounds, 836 F.2d 1312 (11th Cir.), cert. dismissed, 487 U.S. 1265, 109 S.Ct. 28, 101 L.Ed.2d 979 (1988). An evidentiary hearing on the issue of selective prosecution is not automatic; such a hearing is conducted “only where a defendant presents facts sufficient to raise a reasonable doubt about the prosecutor’s motive.” Owen, 806 F.2d at 1523.

We have no quarrel with the respective district courts’ concluding that the petitioners met the first part of the selective prosecution test.²⁹ While the petitioners were sentenced under the mandatory requirements of the HFOA, they have demonstrated that other similarly eligible convicted defendants were not. Therefore, the petitioners presented sufficient facts to create a reasonable doubt concerning the motive behind prosecutors’ not pursuing sentencing under the HFOA for those similarly eligible convicted defendants to warrant an eviden-tiary hearing on this issue. See Rickett, 901 'F.2d at 1065 (“[T]he officials simply failed to consult sources which they knew possessed the relevant information” concerning prior convictions of a similarly eligible codefend-ant. (Johnson, J., concurring)). Isolated instances of sporadic application of the HFOA were presented in the Jones, Mardis and Garrett cases. We, however, have determined that random applications of the HFOA, likely attributable to negligence or error, do not evidence selective prosecution or a constitutional violation. Id. at 1061. This consolidated appeal is distinguished from Rickett by the evidentiary hearing conducted in McLester II. Based on analysis of the Alabama Department of Corrections data for a ten-year period, the expert testimony and documentary evidence revealed that approximately 30 percent of eligible convicted Alabama defendants were not sentenced under the HFOA.

Even if we recognize that Jones, Mardis and Garrett have shown disparate treatment in their sentencing as habitual offenders, we cannot conclude that they have demonstrated that the reason they were so sentenced and similarly eligible convicted defendants were not was because of prosecutors’ invidious discrimination or a constitutionally impermissible motivation. Such a showing is necessary to satisfy the second part of the requisite test to show selective prosecution. “Intentional and purposeful discrimination is essential to a selective prosecution claim.” Id.

In these consolidated cases, no allegation has been made that the HFOA is facially discriminatory. Other than a suspect or quasi-suspect category, if petitioners were challenging a classification facially created by the statute, then “they could prevail by showing that the classification is not rationally related to a legitimate state purpose.” E & T Realty v. Strickland, 830 F.2d 1107, 1112 (11th Cir. 1987), cert. denied, 485 U.S. 961, 108 S.Ct. 1225, 99 L.Ed.2d 425 (1988). Petitioners also have not claimed that their prior convictions were not valid or that they individually should not have been sentenced under the HFOA. Significantly, there has been no allegation by any petitioner that his sentence imposed under the HFOA was based on a constitutionally impermissible classification. Cf. Wayte v. United States, 470 U.S. 598, 608 n. 10, 105 S.Ct. 1524, 1531 n. 10, 84 L.Ed.2d 547 (1985) (“A showing of discriminatory intent is not necessary when the equal protection claim is based on an overtly discriminatory classification.”). They claim only that the HFOA is unconstitutional because it is being applied unevenly. See E & T Realty, 830 F.2d at 1112 (Plaintiffs “are alleging that defendants unequally applied facially neutral legislation.”).

It is a “settled rule that the Fourteenth Amendment guarantees equal laws, not equal results.” Personnel Adm’r v. Feeney, 442 U.S. 256, 273, 99 S.Ct. 2282, 2293, 60 L.Ed,2d 870 (1979). To prevail on an equal protection claim that a facially neutral statute is being applied unequally, “intentional or purposeful discrimination” must be shown. Snowden v. Hughes, 321 U.S. 1, 8, 64 S.Ct. 397, 401, 88 L.Ed. 497 (1944); E & T Realty, 830 F.2d at 1112-13. Discriminatory purpose implies that the decisionmaker selected a course of action “because of" its detrimental effects on an identifiable group. Wayte, 470 U.S. at 610, 105 S.Ct. at 1532; E & T Realty, 830 F.2d at 1114. Error, mistake in judgment or arbitrary administration in applying a facially neutral statute does not violate equal protection. E & T Realty, 830 F.2d at 1114.

During his expert testimony, Palmer conceded that the statistical analyses showed only patterns, and not causes. Sloan also stated that any determination regarding a discriminatory motivation behind the administration of the HFOA was beyond the scope of his analyses. A similar admission was made by the expert who prepared statistical analyses of capital sentencing based on race concerning an allegedly unequal application of a facially neutral Georgia statute. McCleskey v. Kemp, 481 U.S. 279, 308, 107 S.Ct. 1756, 1776, 95 L.Ed.2d 262 (1987). Stressing the essential burden that a habeas petitioner alleging an equal protection violation has of proving discriminatory purpose with a consequent discriminatory effect on him, the Court explained that the petitioner “must prove that the decisionmakers in his ease acted with discriminatory purpose.” 481 U.S. at 292, 107 S.Ct. at 1767. The instances in which the Court has recognized constitutionally impermissible discrimination evidenced by statistical data showing discriminatory impact are rare and blatant. We note that the Court cited two exemplary cases, Gomillion v. Lightfoot, 364 U.S. 339, 81 S.Ct. 125, 5 L.Ed.2d 110 (1960) and Yick Wo v. Hopkins, 118 U.S. 356, 6 S.Ct. 1064, 30 L.Ed. 220 (1886), wherein the analyses showed that the discriminatory impact was virtually 100 percent in contrast to the 30 percent nonapplication of the HFOA resulting from statistical analyses of the prison data involved in this appeal. McCleskey, 481 U.S. at 293 n. 12, 107 S.Ct. at 1767-68 n. 12.

In McCleskey, the Court also stated that statistical evidence is more meaningful when the decisionmakers have the opportunity to explain disparities. 481 U.S. at 296, 107 S.Ct. at 1769. At the evidentiary hearing implicated in this appeal, state prosecutors testified concerning the reasons that some eligible convicted defendants were not sentenced under the HFOA. They explained that they may be aware of a convicted defendant’s prior convictions that would make him eligible for a mandated sentence under the HFOA, but be unable to prove the prior convictions in an admissible manner at sentencing, as required by the Alabama statutes and courts.

The methods of proving prior convictions in Alabama changed during the time period surveyed by the prison data. Additionally, in his final report concerning the application of the HFOA, Sloan attributed the state’s difficulty in proving prior felony convictions to incomplete or missing documentation of prior conviction records with the result that an eligible convicted defendant would not receive the sentence to which he was entitled under the HFOA:

[F]iles on offenders (e.g., prosecutor, circuit court, and probation files) were often massing, incomplete, or simply incomprehensible. In many instances, unless all the files associated with a case were available, it was difficult to determine the number of prior convictions, the actual sentence received, and other information on the case. Thus, a lack of standardized, organized and coherent records on offenders may significantly hinder the State’s ability to determine if an offender had prior felony convictions. Clearly, this may have been related to the short sentences uncovered by the analyses.

Petitioner’s Ex. 21 at 12 (emphasis added). The failure to apply a recidivist statute because of lack of knowledge of prior convictions or the unavailability of the requisite documentation at sentencing does not violate equal protection, although subsequent studies and statistics may reveal previous convictions. Oyler v. Boles, 368 U.S. 448, 456, 82 S.Ct. 501, 506, 7 L.Ed.2d 446 (1962).

Aside from the problems of acquiring knowledge and proof of prior convictions within Alabama, obtaining such convictions from other jurisdictions is complicated by the proof requirements of attestation as well as certification and the necessity of relying on other individuals working under the pressures of their own criminal justice systems to supply this information. The constraints of distance, finances and limited personnel severely restrict the state’s ability to acquire previous convictions from other jurisdictions. Consequently, the prosecutors explained at the evidentiary hearing that they would allocate their time and resources to the more serious felony offenses, although these inquiries were subject to the aforementioned practical impediments.

The Supreme Court has recognized that our criminal justice system accords prosecutors broad discretion as to whom they prosecute, and that this discretion is ill-suited to judicial review. Wayte, 470 U.S. at 607, 105 S.Ct. at 1530. Prosecutorial discretion, however, is not unlimited and cannot violate constitutional restraints. 470 U.S. at 608, 105 S.Ct. at 1531; see United States v. Petit, 841 F.2d 1546, 1554 (11th Cir.), cert. denied, 487 U.S. 1237, 108 S.Ct. 2906, 101 L.Ed.2d 938 (1988) (“The government possesses broad discretion in determining whom to prosecute, subject to constitutional constraints prohibiting the exercise of such discretion based on race or other invidious grounds.”). When reviewing the alleged selective application of a similar West Virginia recidivist statute, the Supreme Court concluded that “the conscious exercise of some selectivity in enforcement is not in itself a federal constitutional violation.” Oyler, 368 U.S. at 456, 82 S.Ct. at 506.

The conclusions drawn from the statistical evidence reviewed in this appeal fail to account for practical variables explained by the Alabama prosecutors and acknowledged by the expert witnesses. Therefore, a convicted defendant who is eligible for a particular sentence under the HFOA might not receive an enhanced sentence because the prosecution cannot prove his prior convictions. As the statistical analyses admittedly were compiled, such an individual would be listed as receiving an inappropriate or “short” sentence under the HFOA, and would be included among the 30 percent of individuals who were sentenced improperly. Because the statistical evidence provides only bare numbers and does not account for the problems encountered by Alabama prosecutors in proving prior convictions, the conclusions are of limited usefulness and fail to reflect sentencing realities, or the reasons for instances when the HFOA was not applied.

The testimony at the evidentiary hearing did not substantiate the concern that Alabama is administering the HFOA in a “careless and haphazard fashion.” Rickett, 901 F.2d at 1067 (Johnson, J., concurring). On the contrary, it showed that prosecutors strive to achieve the mandate of the HFOA by diligent efforts to obtain prior convictions. Their endeavors occasionally are thwarted by the inability to obtain proof of prior convictions in Alabama and from other jurisdictions. Moreover, the statistical evidence gives no indication that a discriminatory motive was the reason that the state did not produce prior convictions at sentencing to enable the judge to impose a sentence under the HFOA. “Even though the statistics in this case might imply a policy of selective enforcement, it was not stated that the selection was deliberately based upon an unjustifiable standard such as race, religion, or other arbitrary classification” to support a finding of denial of equal protection. Oyler, 368 U.S. at 456, 82 S.Ct. at 506. We conclude that the statistical evidence, for which the Jones, Mardis and Garrett cases were consolidated, “is clearly insufficient to support an inference” that Alabama prosecutors have “acted with discriminatory purpose” in not acquiring proof of prior convictions under the HFOA during the ten-year period analyzed. McCleskey, 481 U.S. at 297, 107 S.Ct. at 1770.

“Apparent disparities in sentencing are an inevitable part of our criminal justice system.” 481 U.S. at 312, 107 S.Ct. at 1778. While not perfectly applied, the HFOA appears to be serving its intended legislative purposes of deterrence and selective incapacitation in approximately 70 percent of Alabama sentencings, as shown by the statistical data analyzed for the purpose of determining the application of the HFOA. See E & T Realty, 830 F.2d at 1114 (Because no law can be enforced uniformly and without exceptions, the Constitution does not require perfect application.). The same statistical analy-ses give no indication that the instances of nonapplication are for constitutionally impermissible reasons. On the basis of the statistical evidence placed into this record on appeal at petitioners’ request to establish their claim of unconstitutional application of the HFOA, we conclude that petitioners have failed to establish that they are the victims of sentencing in violation of equal protection or that the HFOA is being unconstitutionally applied. “The equal protection clause requires no more than that state decisionmak-ers applying a facially neutral stat[ut]e not intentionally discriminate.” E & T Realty, 830 F.2d at 1114.

III. CONCLUSION

For the reasons discussed herein, we conclude that petitioners Jones, Mardis and Garrett have failed to show that the HFOA is being applied in violation of equal protection; their petitions for habeas corpus relief properly were denied. McLester’s second habeas petition is an abuse of the writ because he should have raised his allegation of selective enforcement when he filed his first habeas corpus petition. Although the district court should not have proceeded to address his equal protection claim, the court correctly denied relief. Accordingly, the denial of ha-beas corpus relief in these four consolidated cases is AFFIRMED.

1 . Mr. Downing, the burglary victim, testified at an evidentiary hearing in this case that he received a call that a van was behind his house. Downing and another individual went to Downing’s home, where they met Garrett emerging from the house with some of Downing's property. Garrett put a gun to Downing's head and threatened to kill him.

2 . At Garrett’s plea proceeding, the court thoroughly and carefully inquired concerning his satisfaction with his retained attorney’s representation:

THE COURT: Okay. Now, are you satisfied with the performance of your lawyer in this case?

THE DEFENDANT: Yes, sir, I am.

THE COURT: Okay. And you've had the benefit of counsel?

THE DEFENDANT: Yes, sir.

THE COURT: Was he appointed, or was he paid?

THE DEFENDANT: Paid.

THE COURT: And you totally understand his advice, and you don't have any reservations about the fact that he has given you sound legal advice?

THE DEFENDANT: Yes, sir.

THE COURT: Do you wish to consult with another attorney at this time?

THE DEFENDANT: No. No, not at this time.

THE COURT: You feel like you have all of the advice that you need in order to make this decision from your paid counsel?

THE DEFENDANT: From him, yes, sir, I do.

THE COURT: And do you waive any right that you might have under the Constitution of the United States based on any allegation that you weren’t represented by competent counsel in this matter?

THE DEFENDANT: Yes, sir, I waive it.

THE COURT: And that means that when you get to the penitentiary, you're not going to be writing all of those writs up here saying, I didn't have a competent lawyer. Do you understand that you are giving up that right, and we are adjudicating that now — because, if there is any question about it, we would rather go on out in front of the jury and try the case and let the jury decide whether you're guilty or not. Do you have any reservations at all in that department about your lawyer?

THE DEFENDANT: No, sir, I don't.

THE COURT: Okay. The Court specifically finds that the defendant is intelligent, alert — he has answered questions knowledgeably, and he has knowingly and consciously agreed that he has been represented by competent counsel in this case. And he has waived any possible right that he has to allege incompetency of counsel in this matter.

Is there anything that you feel like your lawyer should have done that he hasn’t done?

THE DEFENDANT: No. No, sir.

THE COURT: Okay. Is there anything that he has done that you feel like he should not have done?

THE DEFENDANT: No. I would have to say no, because being, you know—

THE COURT: You don’t have to do anything except tell the truth.

THE DEFENDANT: Well, okay. I will say no because of the fact that there is nothing that I can go on to say what he should have done, because I think he gave me a fair job on it.

Ex. 3-31-33 (Transcript of Plea Proceedings, Dec. 2, 1985) (emphasis added).

3 . At the sentencing proceeding, the trial judge explained to Garrett his reasons for refusing to allow him to withdraw his plea:

I was sitting there [at the plea proceeding], and I told you at that time that I didn't care whether you pleaded guilty or not. And I had a jury sitting out there. And we had, at the expense of the State of Alabama, brought you there for a trial. But you didn't want a trial, and you wanted to plead guilty. And you were represented by counsel, and I believe he's paid counsel. And you pleaded guilty. And now you are wanting me by the time the sentencing comes around to set that aside.... And, incidentally, you pleaded guilty and freely and voluntarily gave facts that would be admissible in any case that ever comes up in which you would [be] tried that would show your guilt. And you weren’t under any coercion or anything else from the State of Alabama to do that. And why would we set aside your guilty plea and then let the State read into evidence the transcript of your guilty plea, because there's a[n] [ad]mission against interest, I think, if there’s ever been one. And there is no way that you could keep that from being admitted into evidence.

Ex. 3-42-43 (Transcript of Sentencing, Jan. 13, 1985).

4 . At the sentencing proceeding, the prosecutor supplied the court with certified judgments of conviction in the four felony cases from Coosa County and a list of ten convictions from other counties. While Garrett's attorney Radney stated that he could not refute the four Coosa County convictions, he persuaded the trial judge not to consider the ten other convictions in sentencing Garrett because the state had not proved them. For purposes of the HFOA, Alabama law requires the prosecution to prove or establish prior convictions by attested, certified copies of those convictions.

5 . In January, 1979, McLester pleaded guilty to seven burglaries, which occurred in a single episode of burglarizing seven stores in a shopping mall. McLester I, 802 F.2d at 1331.

6 . This instance where the HFOA was not applied was described by an attorney who testified that, for the ten years preceding his testimony, he had participated in between 500 to 1,000 felony cases, and many times the HFOA was involved.

7 . The Supreme Court has held that the Sixth Amendment guarantees the accused "that he need not stand alone against the State at any stage of the prosecution, formal or informal, in court or out, where counsel’s absence might derogate from the accused’s right to a fair trial.” United States v. Wade, 388 U.S. 218, 226, 87 S.Ct. 1926, 1932, 18 L.Ed.2d 1149 (1967) (footnote omitted); see Swicegood v. State, 565 So.2d 1206, 1209 (Ala.Crim.App.1990) (For a prior conviction to be used for the purpose of enhancement under the HFOA, the convicted defendant must have been represented by counsel at trial.).

8 . One district attorney testified that he had experienced difficulty in locating records of prior convictions in some Alabama counties as recently as the preceding five to ten years because of differences in record keeping.

9 . Godwin testified concerning out-of-state convictions:

[O]ftentimes you just can't get people to respond. If you have got a case out of Detroit, New York or someplace like that and they show on a rap sheet, you can request until you are blue in the face and you may not get a response. It is just that simple.

R3-164.

10 . This district attorney testified that he had never had a case where he was aware of prior convictions and did not seek application of the HFOA. In his office, if a plea agreement had been executed and additional prior convictions were discovered, then the plea agreement was withdrawn. He explained that rural Alabama counties might take up to a month to respond to a request for prior convictions, while a metropolitan county generally responded within a week. In his circuit, judges would grant at least one continuance to allow time to obtain documentation of prior convictions when difficulties arose.

11 . An investigator with the Montgomery County District Attorney's Office testified at the eviden-tiary hearing concerning his checking a random sample of the data provided by the Department of Corrections against actual county records. The investigator found the Department of Corrections information to be “riddled with errors.” R5-457. These errors included youthful offender cases, the wrong charge, and split sentences, or a shorter sentence than was given being shown. Additionally, the investigator, who previously had been an investigator in the Attorney General’s Office, testified that he had assisted prosecutors in various counties in obtaining certified criminal histories from other jurisdictions and that "it is quite a lengthy process.” Id. at 466.

12 . Under the stipulation, the following records were excluded in drawing the sample: sentencing dates prior to January 30, 1981; drug-related offenses sentenced before 1988, when drug offenses were included under the HFOA; incarcerations for protection; sentences where either the offense or the date of the sentence was unrecognizable or invalid; sentences and convictions outside of Alabama; and sentences under the Alabama Youthful Offender statute.

13 . The random sample consisted of two classifications: 300 Type I, offenses for which the corrections data indicated that the HFOA should have been applied because of prior convictions, and 100 Type II, offenses for which the corrections data showed that the HFOA should not have been applied because no prior felony convictions were indicated. R4-277. The number of cases in each group was selected to produce an error rate of approximately 5 percent, which the parties agreed was an acceptable rate of variance. Id. at 278. Combining Type I and Type II cases resulted in an overall error rate of 28.2 percent. Id. at 293.

14 . Specifically concerning Class A offenses, Palmer found that the numbers were so few for each year that it was difficult to .evaluate patterns, which, even if they were found, would be unreliable. R4-310-11.

15 . Palmer further admitted that "I have presented a tabulation. I have not presented reasons, nor can I, from the information I have available.” R4-313.

16 . In his report concerning his analysis of the sample studied, Sloan explained the verification of records performed in each test case: "(1) the date of conviction and county in which the conviction occurred; (2) the felony class of the conviction offense for the test case; (3) the length of sentence received in the case; (4) the extent of prior convictions on the offender's part; and (5) whether the sentence received was within guidelines set down by the Habitual Offender Act.” Petitioner's Ex. 21 at 6.

17 . Sloan concluded that "about 32% of the sentences of recidivist Class A felons were short; nearly 48% of the sentences for Class B felons were short; and 24% of the Class C felons received short sentences. It appeared that between 1981-1990, those convicted of more serious felonies were more likely to receive short sentences. In other words, it appeared that the seriousness of the conviction offense (coupled with the number of priors) was inversely related to the length of the term of sentence. Class A and B felons (with priors) were more likely to receive short sentences than were Class C felons with prior convictions.” Petitioner's Ex. 21 at 10.

18 . The following percentages of failure to apply the HFOA to eligible convicted defendants occurred in the classifications of felony crimes: Class A, 43 percent (two priors), 44 percent (three priors); Class B, 36 percent (two priors), 45 percent (three or more priors); and Class C, 29 percent (two priors), 29 percent (three or more priors). R4-347, 349. In his report, Sloan concluded that "[ojverall, prior convictions of offenders appeared to be related to receiving short sentences, but in the wrong direction: as the number of priors increased, so did the percentage of short sentences." Petitioner's Ex. 21 at 10.

19 . Sloan views the HFOA as a policy statement, where "the legislature has deemed that recidivist offenders constitute a danger to the people of the state of Alabama and one way to deal with them is to deter and incapacitate them” through "a statute that will create a rule for the criminal justice system to follow in these cases." R4-371-72. He defined deterrence as "a theory that says that we can control the amount of crime that exists in society by punishing those who violate the criminal law.... [I]f the pain or the amount of punishment that is inflicted on someone outweighs the benefits from doing that particular activity, then they will not choose to engage in that.... [I]f people perceive in large numbers that ... the certainty of punishment, for whatever reason, is very low, then the deterrent aspect of the criminal justice must break down." Id. at 361, 367. Selective incapacitation “is simply where you identify an individual who has a high rate of offending, and you selectively incapacitate him. That is, you lock that person up, you imprison that person for an extended period of time. With the rationale being that while this person is incarcerated they cannot engage in any future criminality.” Id. at 362.

20 . Sloan clarified that "[i]f offenders perceive that the likelihood of their receiving enhanced punishment is, say, one out of three or one out of four, or whatever it might be in reality, that will certainly reduce the certainty component of the deterrence aspect of the habitual offender statute.” R4-365-66.

21 . Sloan expressed concern that the selective incapacitation component of the HFOA was being defeated inasmuch as his analyses showed that as the severity of the offense increased, so did the likelihood of receiving a short sentence:

I think that I am in agreement with the district attorneys that generally people would say it is more important to selectively incapacitate those who engage in serious crimes of violence,, such as Class A offenses, or serious property offenses such as in the Class B category, than it would be to incapacitate someone who is an habitual bad check writer [Class C offense].

R4-369. Considering that the purpose of the HFOA is to punish or deter habitual offenders, Sloan found it to be problematic that Class C recidivists, "who, at least theoretically might pose the least danger to society, are being sentenced correctly most of the time,” while "the B's and the A's, on the other hand, receive shorter sentences in a much greater percentage of the cases.” Id. at 373. Accordingly, the results of his analyses showed "that the intent of the legislature is being turned inside out here.” Id.

22 . Having consolidated the Jones, Mardis and Garrett cases, this court granted on March 15, 1991, their joint motion to stay their consolidated appeal until the occurrence of the McLester II evidentiary hearing, which had been postponed. The joint motion represented that the McLester II evidentiary hearing would provide expert testimony and evidence concerning the operation of the HFOA which had not yet been considered by this or any other court. On April 12, 1991, we granted a joint motion to hold the consolidated appeal in these three cases in abeyance until disposition of McLester II and ordered status reports apprising us as to when a decision might be expected in McLester II. We held these consolidated cases in abeyance until McLester II was appealed to this court on March 3, 1992, and then consolidated all four cases for oral argument on November 16, 1992. By granting joint motions based upon the representation that the McLester II decision and evidentiary hearing would provide relevant information for our determination of the constitutionality of the operation of the HFOA, this court implicitly, if not overtly, indicated that we intended to include the evidence from McLester II to resolve the three consolidated cases in order to have an informed statistical basis with the benefit of expert testimony to rule on the question of whether the HFOA is being applied in violation of equal protection.

23 . Under the HFOA, several offenses arising out of a common event or occurrence arc regarded as separate offenses for enhancement under the HFOA, although, in another jurisdiction, the offenses would be treated as one conviction or concurrent sentences were imposed. Manning v. State, 568 So.2d 327, 330 (Ala.Crim.App.1990), post-conviction relief denied, 580 So.2d 120 (Ala. Crim.App. 1991); see Parker v. State, 516 So.2d 859, 865 (Ala.Crim.App. 1987) (”[T]he Habitual Felony Offender Act refers to convictions, not to sentences.”). All felony convictions, even those that occurred before the HFOA became effective and regardless of their origin, are to be considered in determining the correct sentence under the HFOA. Lidge, 419 So.2d at 614. The determination of whether an out-of-state conviction should be used to enhance punishment under the HFOA is based upon Alabama's definition of the offense, and not on whether the foreign jurisdiction treats the crime as a felony or misdemeanor. Mitchell v. State, 579 So.2d 45, 51-52 (Ala.Crim. App.1991), cert. denied, 596 So.2d 954 (Ala. 1992).

24 . We are troubled by allegations in these consolidated cases that prosecutors have offered to plea bargain by not apprising the sentencing court of a convicted defendant’s complete history of felony convictions. This information is necessary for the HFOA to function as the legislature intended. Without a complete felony conviction history, a habitually convicted defendant, to whom the statute is directed, will not be sentenced under the HFOA, particularly when a sentence of life without parole is dictated. By offering such a bargain, prosecutors are exercising discretion where they have none. This alleged conduct violates the mandate of the HFOA as well as a breaches prosecutors’ ethical obligations as officers of the court. Alabama courts have reiterated that a prosecutor must inform the sentencing court of prior felony convictions of which the prosecutor is aware. Miliner v. State, 414 So.2d 133, 135 (Ala.Crim.App.1981).

25 .The phrase "when it is shown” that a convicted defendant has previous felony convictions in the HFOA has been interpreted to mean to prove or apprise the court in a evidentiary admissible manner, a responsibility which is incumbent upon prosecutors, and not discretionary. Miliner, 414 So.2d at 135.

26 . In Jefferson County, however, a prior conviction could be proved by a certified copy of a trial docket sheet. Lidge, 419 So.2d at 615 n. 2. Accordingly, in Jefferson County, certified copies of trial docket sheets are equivalent to minutes of the court. Anderson v. State, 480 So.2d 64, 65 (Ala.Crim.App. 1985).

27 . Exemplifications or copies of records ..., or parts thereof, ... which are kept in any public office in this state shall be proved or admitted as legal evidence in this state by the attestation of the keeper of said records ... that the same are true and complete copies of the records, ... or parts thereof, in his custody, and the seal of said keeper thereto annexed, if there is a seal (if there is no official seal, there shall be attached to such attestation the certificate of the clerk), and the seal of the circuit, district or municipal court of the proper county where such keeper resides that such attestation is genuine and made by the proper officer.

Ala.Code § 12-21-67 (1986).

28 .The record and judicial proceedings of the courts of any state or territory or of any [country subject to the jurisdiction of the United States] shall be proved or admitted in any other court within the United States by the attestation of the clerk and the seal of the court annexed, if there is a seal, together with a certificate of the judge, chief judge or presiding magistrate that the said attestation is in due form.

Ala.Code § 12-21-70 (1986).

29 . We acknowledge the view that convicted felons with three provable prior felony convictions are not necessarily similarly situated with convicted felons with three improvable prior felony convictions for the purpose of applying the HFOA at sentencing. This conclusion would terminate the two-pronged test for selective prosecution relating to recidivist sentencing with the first prong. Theoretically, however, “singling out” could occur in either category of convicted felons with provable or unprovable prior felony convictions. Such a determination could create issues as to why prior convictions were pursued for some convicted felons and not for others. That would return us to a prosecutorial intent or motivation analysis, which is the second prong of the selective prosecution test. Therefore, we progress to the second prong, which is not debatable, to resolve this equal protection issue concerning the application of the HFOA in Alabama.