¶1concurring.
¶2While I agree with most of what Judge Clinton has written for the majority, I feel compelled to write separately for a variety of reasons.
¶3These petitions for discretionary review were granted and consolidated in response to a perceived desire on the part of the bench and bar for additional guidance concerning application of Tex.R.App.Pro. 81(b)(2) to Rose error. As Judge Clinton’s opinion ably demonstrates, various courts of appeal have, in spite of similar fact patterns, decided this issue in widely disparate manners. Such a state of affairs has caused dissension between the courts of appeals and uncertainty on the part of everyone affected.
¶4*325Unfortunately, it is impossible to set out a step-by-step procedure that will automatically lead to an inerrant conclusion that the error was or was not harmful. Any harm analysis tends to be subjective and uncertain, but when a court is faced with Rose error, these inherent difficulties are exacerbated by the simple fact we are not privy to the jury’s deliberations.
¶5Judge Clinton identifies a number of factors which will affect an appellate court’s application of Rule 81(b)(2). These factors, briefly summarized, include: (1) what, if any, discussion of parole occurred during voir dire, (2) argument by counsel, (3) the existence of jury notes concerning application of the parole laws, (4) the actual sentence assessed (is it in accord with the facts of the case, does the number of years imposed suggest application of a one-third rule? etc.), (5) a deadly weapon finding, (6) the facts of the case, (7) prior convictions, (8) additional instructions concerning parole,
¶6It is important to stress that these factors are neither exhaustive nor universally applicable. Every case will present a different set of circumstances. The majority opinion should not leave the impression that a laundry list of these possible factors should be checked off each time a court undertakes an 81(b)(2) analysis. Such an approach would lead, inevitably, to the type of confusion which has pervaded this area since we decided Rose.
¶7Tex.R.App.Pro. 81(b)(2), which sets out the harmless error rule applicable to the instant case, states:
If the appellate record in a criminal case reveals error in the proceedings below, the appellate court shall reverse the judgement under review, unless the appellate court determines beyond a reasonable doubt that the error made no contribution to the conviction or to the punishment, [emphasis added]
¶8This standard for review is remarkably similar to standard, established by the United States Supreme Court in Jackson v. Virginia, 443 U.S. 307, 99 S.Ct. 2781, 61 L.Ed.2d 560 (1974), applied to sufficiency of the evidence cases, viz.
[T]he critical inquiry on review of the sufficiency of the evidence to support a criminal conviction must be not simply to determine whether the jury was properly instructed, but to determine whether the record evidence could reasonably support a finding of guilt beyond a resonable doubt. But this iquiry does not require a court to “ask itself whether it believes that the evidence at the trial established guilt beyond a reasonable doubt.” Instead, the relevant question is whether, after view the evidence in the light mos favorable to the prosecution, any rational trier of fact could have found the essential elements of the crime beyond a re-sonable doubt.
¶9Id. at 318 (citations omitted).
¶10In addition to an obvious similarity between these rules, an appellate court conducting a sufficiency review faces similar problems and temptations as a judge conducting a harm analysis. Such a judge is put in the position of trying to determine how the events of the trial affected the factfinder. While a judge conducting a *326sufficiency review may be tempted to substitute his or her opinion of the appellant’s guilt for that of the factfinder, the standard of review, by its very terms, precludes such an approach. Likewise, a harm analysis under Tex.R.App.Pro. 81(b)(2) seems to invite the judge to question whether the error might have affected his or her reasoning. Again, 81(b)(2) prohibits such a substitution of the appellate perspective for that of the factfinder.
¶11The notion of how to conduct an 81(b)(2) analysis was recently discussed by this Court in Harris v. State (Tex.Cr.App. No. 69,366 June 28, 1989) (not yet reported, pending on rehearing).
Although Rule 81(b)(2) has been cited innumerable times by this Court as well as the courts of appeals, beyond simply repeating the language of the rule in conclusory terms, we have failed to articulate a coherent standard for determining when an error is harmless. In other words, the harmless error rule is expressed in conclusory terms that implicated subjective concerns. What is absent from the rule is the objective standards that must be explored to reach a legally correct resolution. In this regard it must be emphasized that the function of an appellate court’s harmless error analysis is not to determine how the appellate court would have decided the facts, but to determine to what extent, if any an error contributed to the conviction or the punishment. The language of the rule dictates that a reviewing court’s repsonsibility transcends determining whether the conviction was correct.
¶12Id.,slip op. at 29-30. After setting out a number of factors that would suggest whether the error in Harris might have affected the jury, much has Judge Clinton has done in this case, the Court concluded:
General consideration having been set out, we are left only to provide a skeleton on which to place them. A procedure for reaching this determination should: first, isolate the error and all its effects, using the considerations set out above and any other considerations suggested by the facts of an individual case; and second, ask whether a rational trier of fact might have reached a different result if the error and its effects had not resulted.
¶14If this restatement of Rule 81(b)(2) is not helpful, it must be remembered that no harm analysis will be a simple, mechanical process, It, like the “outstanding reasonable hypothosis” standard applied in circumstantial evidence cases, is not a different standard. Instead it is a different way of looking at the same standard. See Carlsen v. State, 654 S.W.2d 444, 449 (Tex.Cr.App.1983) (opinion on rehearing).
¶15An 81(b)(2) analysis of Roseerror should not be more difficult than testing the sufficiency of a circumstantial evidence case. The notions of proof and relevancy applicable to one case should easily translate into the other context. Hopefully, the majority opinion will be read as a comprehensive resource to reveal how various courts have dealt with this issue rather than a prescription of how to conduct the analysis itself.
¶16. Even if we were able to listen to the jury deliberate, a discussion of the parole laws would not necessarily mean that the jury was affected by the discussion. Conversely, a jury could easily set a sentence, relying heavily on parole considerations, and never mention parole. The ultimate answer, whether the sentence was affected by the parole instruction, is locked within the minds of the individual jurors. And, even if we were to ask the jurors whether they were affected, many would probably be unable to answer definitively.
¶17. The majority opinion does state that the final paragraph of § 4 may never be considered as a "curative” instruction. I agree wholeheartedly. The curative instruction discussed in Rosedid not appear in the statute and went far beyond the statute's ambiguous admonition not to consider parole. See the majority opinion at note 23.