dissenting.
The record reflects that previously, in a well written and reasoned opinion by Justice Reeves of the San Antonio Court of Appeals, that court correctly reversed the trial court’s judgment of conviction in this cause after it found that trial counsel for Kenneth Earl Jackson, henceforth appellant, was ineffective because he failed to correctly advise appellant of the consequences of electing the jury instead of the trial judge to assess appellant’s punishment. See Jackson v. State, 662 S.W.2d 74 (Tex.App.-4th Dist.1983). Also see that court’s opinion of Jackson v. State, 640 S.W.2d 323 (Tex.App.-4th Dist.1982). The effect of appellant going to the jury rather than the trial judge for assessment of punishment amounted to him receiving ten more years. Thereafter, a majority of this Court, in a well written and reasoned opinion by Judge Miller, correctly voted to affirm the judgment of the court of appeals. See Jackson v. State, 766 S.W.2d 504 (Tex.Cr.App.1985).
This case is now before this Court pursuant to a remand order from the Supreme Court of the United States, “for further consideration in light of Texas v. McCullough, 475 U.S. 134, 106 S.Ct. 976, 89 L.Ed.2d 104 (1986).”
I find that no real conflict exists between what this Court previously stated and held in this cause and what the Supreme Court stated and held in Texas v. McCullough, supra. Therefore, my vote is to affirm the judgment of the court of appeals, which reversed the trial court’s judgment of conviction. Because a majority of this Court erroneously votes to do otherwise, I respectfully dissent.
In its opinion the San Antonio Court of Appeals correctly found the following: “Because of counsel’s erroneous advice [to go to the jury rather than the trial judge for punishment], appellant made a decision which resulted in a sentence ten years greater than he would have received had he gone to the court for punishment. By going to the jury, appellant had everything to lose and nothing to gain since the minimum sentence which the jury could impose was the maximum the court could have imposed.”
This Court, when this cause was previously before it, found that “The record, including the State’s complete cross examination of appellant, his attorney, and the trial judge at the evidentiary hearing ordered by the court of appeals, clearly indicates that nothing except ‘the [judge’s] election reasons were considered in electing the jury in this case.' ” Also see what this Court stated in footnotes 4 and 5 of its majority opinion.
The issue here does not involve whether counsel was ineffective at the guilt stage of the trial; the issue instead is whether counsel was ineffective at the punishment stage of the trial. Thus, “[b]ecause applicant’s complaint deals with the punishment or sentencing phase of his trial ... the two-pronged test set out in Strickland is not applicable to the instant case. Thus we look to the standard used before the Supreme Court handed down Strickland ”, which was Ex parte Duffy, 607 S.W.2d 507 (Tex.Cr.App.1980). Ex parte Cruz, 739 S.W.2d 53, 58 (Tex.Cr.App.1987).
Under Ex parte Duffy, supra, in making the determination whether trial counsel was ineffective, the question that must be answered is whether counsel rendered reasonably effective assistance of counsel.
Notwithstanding what a majority of this Court held in Ex parte Cruz, supra, just a little over a year ago, the new majority of this Court, without even mentioning Ex parte Cruz, supra, invokes and applies to the punishment phase of this cause the amorphous Strickland v. Washington, 466 *524U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984), two prong test. The new majority, however, albeit erroneously, also concludes that trial counsel for appellant was effective counsel under Ex parte Duffy, supra.
Usually when a defendant challenges his trial counsel’s effectiveness, such is not based upon a single error. In this cause, however, the single error of counsel, in electing to have the jury rather than the trial judge assess punishment, “permeated the entire [punishment] proceeding and rendered counsel’s assistance ineffective.” (Page 511 of this Court’s original majority opinion.) This single error occurred because trial counsel for appellant, although having a “lock” under the law as it existed at the time on the minimum punishment that could be assessed, chose instead to advise appellant to go to the jury for punishment, which decision maker was not bound by the minimum punishment provided in the law that existed at the time. Under the law that then existed, and given the evidence that was then in existence, the trial judge could not have legally assessed appellant’s punishment at greater than the mimimum.
In this regard, in making the determination whether trial counsel for appellant was ineffective, it is necessary to look to the law that existed when he made his decision to advise appellant to have the jury rather than the trial judge assess his punishment, and not to what the law might be today, several years after the fact. I find that this Court’s original opinion by Judge Miller has correctly exhibited the law that existed when the jury assessed appellant’s punishment; therefore, I will not repeat all of what he has already stated.
The law that existed when appellant elected on retrial to have the jury rather than the trial judge assess punishment was clear. Under North Carolina v. Pearce, 395 U.S. 711, 89 S.Ct. 2072, 23 L.Ed.2d 656 (1969), no matter whether the trial judge or the jury assessed the first punishment, if the trial judge was called upon to assess the punishment at retrial, and he chose to increase the punishment from that which had previously been assessed, in order for his decision to stand it was necessary that the record reflect an explanation from him and evidence to support his decision to increase the punishment. Otherwise, absent any evidence that would warrant an increase in the punishment that had previously been assessed, there was a presumption that the trial judge acted vindictively. However, a jury at the retrial was not barred from assessing a greater punishment than that which had been assessed at the first trial. See Caffin v. Stynchcombe, 412 U.S. 17, 93 S.Ct.1977, 36 L.Ed.2d 714 (1973).
Thus, when appellant’s retrial occurred, and when he elected to go to the jury for punishment, if a defendant obtained a new trial, as occurred here, unless he thought he could “beat” the minimum possible punishment, no reasonably effective trial attorney would ever elect to have the jury assess the punishment on retrial. This was so because in that instance the defendant had a “lock” on the minimum punishment that could be assessed-if he elected to have the trial judge assess his punishment and there was no evidence that might reflect or indicate that appellant was guilty of any wrongful identifiable conduct since his first punishment was assessed.
I pause to emphasize that the trial judge in this cause did not assess appellant’s punishment on retrial; the jury did, and under Chaffin, supra, where the jury is called upon to assess the punishment on retrial, it is free to assess the punishment within the minimum and the maximum range that the law provides. Thus, the jury did nothing wrongful in this cause.
The Supreme Court in its remand order in this cause only stated that it was remanding this cause to this Court for it to reconsider its decision in light of Texas v. McCullough, 475 U.S. 134, 106 S.Ct. 976, 89 L.Ed.2d 104 (1986). I assume this to mean that the Supreme Court wants us to remake the determination whether counsel was ineffective in light of Texas v. McCullough, supra. Given how religiously this Court has adhered to North Carolina v. Pearce, supra, and if one subscribes to stare decisis, I am unable to understand *525how anyone could have divined that the Supreme Court would abolish in Texas v. McCullough, supra, the “presumption of vindictiveness” rule of law that it had created in North Carolina v. Pearce, supra.
The law at the time of appellant’s retrial was clear as crystal. If the trial judge was called upon to assess punishment on retrial, and there was no evidence to justify an increase in the punishment from that which had previously been assessed, and the trial judge nevertheless increased the punishment, his act of increasing the punishment caused a “presumption of vindictiveness” to exist. It is, however, probably incorrect to use the expression “presumption of vindictiveness” in this context. It is probably more correct to use the term “assumption” in this context rather than the term “presumption” because the term “presumption” is usually reserved in our law to indicate the product of some inference, whereas the term “assumption” is defined to mean that something is accepted as fact until the contrary is shown to exist. Thus, in this context, we should not infer from the mere fact that a trial judge assessed a greater punishment from that which had previously been assessed that he did so because he was vindictive. I believe in that context one is only saying that unless there is evidence that would justify his act of assessing a greater punishment, we assume he acted vindictively.
In this instance, there is nothing in the record that might reflect or indicate just exactly what the trial judge would have done when appellant's second punishment hearing was conducted, had he been elected at that time to assess appellant’s punishment. At the evidentiary hearing, he was simply called upon to speculate what he might have done had he been elected to assess appellant’s punishment. Appellant’s trial counsel was in the same boat, in that he, too, was called upon to speculate why he advised appellant to elect to have the jury rather than the trial judge assess appellant’s punishment. Therefore, it appears to me that this Court should decide the issue based on the record as it existed when the jury assessed appellant’s punishment, and not on how someone might view the situation from hindsight.
In Texas v. McCullough, supra, the Supreme Court abolished its “presumption of vindictiveness” rule that existed when appellant’s punishment hearing was conducted in this cause. This rule of law had previously been applied to the situation where a trial judge, without any evidence to support an increase in punishment on retrial, increased the defendant’s punishment from that which had previously been assessed. In that instance, the trial judge was presumed to have acted vindictively, and such act on the part of a state trial judge increasing the defendant’s punishment violated the Due Process Clause of the Fourteenth Amendment. In McCullough, supra, with a stroke of the pen, the Supreme Court abolished that rule of law. However, the Supreme Court found in that cause that there was record evidence that supported that trial judge’s decision to increase the defendant’s punishment on retrial. The Supreme Court found that even if it invoked and applied the former presumption of vindictiveness rule to the case, there was sufficient evidence in that record to overcome it.
Thus, it does not appear to me that the Supreme Court has done as much damage to this area of the law as some think or believe it might have done. I find nothing in Texas v. McCullough, supra, that would permit a trial judge, without any basis therefor, assessing a greater punishment on retrial than what had been assessed at the original trial. I find that the only difference between the law as it formerly existed and as it now exists under Texas v. McCullough, supra, is that an appellate court may not now invoke and apply a “presumption of vindictiveness” to a trial judge who assesses a greater punishment than that assessed at the first trial. Furthermore, I find nothing in Texas v. McCullough, supra, that the defendant, in order to win, must prove that the trial judge was in fact vindictive.
Thus, I believe that the principles enunciated in this Court’s opinions of, for example, Lechuga v. State, 532 S.W.2d 581 (Tex.*526Cr.App.1975); Ex parte Bowman, 521 S.W.2d 858 (Tex.Cr.App.1975); and Miller v. State, 472 S.W.2d 269 (Tex.Cr.App.1971), are still applicable to this kind of situation.
Judge Miller, on behalf of a majority of this Court, in footnotes 4 and 5 of his opinion for the Court in this cause, has clearly and correctly demonstrated the fact that there was no record evidence that had the trial judge been elected to assess appellant’s punishment at his retrial, he would have been justified in increasing appellant’s punishment from that which had previously been assessed.
The majority opinion, however, holds that there was sufficient “identifiable” conduct by appellant after his first trial and before his second trial that would have warranted the trial judge, had he been elected to assess the punishment, assessing a greater punishment than the minimum.
The majority opinion states that appellant’s failure to appear for a trial setting and was surrendered by the sureties on his bond are two acts of identifiable conduct. However, as Judge Miller correctly pointed out in footnote 5 of his opinion for the Court, such was not “evidence” because “same were never developed as a factual matter at appellant’s trial or at the eviden-tiary hearing ordered by the court of appeals. Nor has this ‘evidence’ been alleged or briefed as identifiable conduct before the court of appeals, or before this Court on the State’s petition for discretionary review, either by the locally affected district attorney or the State’s prosecuting attorney. As is obvious from the record, the parties to this protracted litigation do not rely upon this ‘evidence’ as identifiable conduct.”
The majority opinion also relies upon the fact that appellant had been charged with committing the offense of theft since the first trial. Of course, given the fact that in our criminal jurisprudence we assume that a person who has only been accused of committing a criminal wrong is not guilty, this is not evidence that would have warranted the trial judge at the retrial, had he been elected to assess appellant’s punishment, assessing a greater punishment than was assessed at appellant’s first trial. Furthermore, as Judge Miller pointed out in footnote 5 of his opinion for this Court, “the evidentiary value of appellant’s arrest, as it bears upon the issue of identifiable conduct, is nil. This is especially true since appellant denies committing the offense, has never been prosecuted for the crime, and no testimony was offered at the evidentiary hearing to even indicate appellant did commit the offense.” (Emphasis in the original.)
The majority almost concedes that the accusation for theft was no evidence of identifiable conduct because, as it points out, the theft charge was later dismissed by the State.
I find that Judge Miller correctly concluded from the record that was then before us that “evidence of [any] identifiable conduct [that can be attributed to appellant] ... is speculative, unfounded and outside the record.”
Furthermore, the record of this cause is absolutely clear that counsel did not rely upon, and may not have even known about the “identifiable” factors that the majority opinion relies upon. If counsel was unaware of the acts of “identifiable” conduct that the majority opinion relies upon, does this omission itself not also cause him to have been ineffective counsel, for failure to make a proper investigation of appellant’s case?
The only reason that trial counsel gave at the evidentiary hearing, in support of his decision to advise appellant that the jury rather than the trial judge assess appellant’s punishment, was that he believed that the trial judge would be harsher on him than a jury would be because he believed that the trial judge was then campaigning hard for reelection, apparently on some sort of “law and order” campaign, and that if he “socked” it to appellant in this cause, this would enhance his chances of being reelected. The record, however, is clear that counsel was mistaken that the trial judge was running for reelection at that time. Nevertheless, counsel does make it clear in his testimony that he believed that the trial judge would assess a *527greater punishment in order to improve his chances of getting reelected. Given the fact that the trial judge was not campaigning for reelection, in my view counsel’s testimony on the subject is totally valueless from an evidentiary standpoint.
Did the Supreme Court in Texas v. McCullough, supra, hold, either expressly or implicitly, that a defendant whose punishment on retrial was assessed by a trial judge at more than than what it had previously been assessed, and there was no evidence in the record to justify the increase, could never complain unless he showed that the trial judge was actually vindictive when he reassessed the punishment? Given the fact that the Supreme Court went ahead and addressed the issue in that cause in light of the “presumption of vindictiveness”, I do not believe so. Furthermore, I do not believe there is a single trial judge in the United States who would would be willing to testify in a court of law that when he assessed a defendant’s punishment on retrial he acted with a heart regardless of social duty and had a mind fatally bent on mischief.
It appears to me such a defendant in Texas may still establish a case under either the Fourteenth Amendment’s Due Process Clause or under the Due Course of Law Clause of the Texas Constitution, see Article I, § 10, Texas Constitution. If he establishes on the record that there is no evidence in the record to support the trial judge’s decision to assess a greater punishment from that which had previously been assessed, the lack of such evidence will constitute a violation of due process of law under the Fourteenth Amendment, or in Texas it will constitute a violation of the due course of law clause of the Texas Constitution. In this instance, there is no record evidence that might have justified the trial judge, if he had been elected to do so, to assess a greater punishment than that assessed at appellant’s first trial. Therefore, appellant is entitled to relief under either the Due Process Clause of the Fourteenth Amendment or the Due Course of Law Clause of the Texas Constitution.
Under Ex parte Duffy, supra, given the record evidence in this cause, appellant’s trial counsel clearly did not render reasonably effective assistance of counsel in advising appellant to go to the jury rather than the trial judge to assess his punishment in this cause. I further conclude that, even under the two prong Strickland v. Washington, supra, amorphous test, which this Court held in Ex parte Cruz, supra, is inapplicable in Texas to the punishment stage of a defendant’s trial, counsel’s performance was deficient; that such performance was prejudicial to appellant, and that but for counsel’s deficiencies appellant’s punishment would have been assessed at the minimum of 15 years’ confinement in the Department of Corrections, rather than ten more years than the minimum, or 25 years’ confinement in the Department of Corrections.
For all of the above and foregoing reasons, I respectfully dissent to the majority opinion finding that counsel was not ineffective under Ex parte Duffy, supra. I further dissent because I believe that Ex parte Cruz, supra, deserves a better burial than the one that the new majority in this cause gives it.