concurring: I agree with the majority that under current Kansas Supreme Court precedent, Roger Shaw’s conviction of involuntary manslaughter while driving under the influence *1010of alcohol must be reversed due to an alternative means error. I also agree with the majority that the case should be remanded for a new trial only on the alternative means supported by sufficient evidence in the first trial. I write separately to express my view that an alternative means error, like almost every other kind of trial error, should be subject to harmless error analysis.
Our Supreme Court has stated die following rule of law governing alternative means cases in Kansas:
“In an alternative means case, where a single offense may be committed in more tiran one way, there must be jury unanimity as to guilt for tire single crime charged. Unanimity is not required, however, as to the means by which the crime was committed so long as substantial evidence supports each alternative means. [Citations omitted.]” ’ ” State v. Wright, 290 Kan. 194, 202, 224 P.3d 1159 (2010) (quoting State v. Timley, 255 Kan. 286, 289, 875 P.2d 242 [1994]).
I take no issue with applying this rule of law to determine whether an alternative means error has been committed. In fact, this rule of law appears to represent the prevailing view of courts in many states. See Annot., Requirement of Jury Unanimity as to Mode of Committing Crime Under Statute Setting Forth the Various Modes by Which Offense May Be Committed, 75 A.L.R.4th 91.
But in Wright, the Kansas Supreme Court went further than necessary by stating that an alternative means error is not subject to harmless error analysis. 290 Kan. at 205. The Supreme Court began the opinion by reaffirming the rule of law initially set forth in Timley, i.e., in an alternative means case, unanimity is not required as to the means by which a crime is committed so long as substantial evidence supports each alternative means. 290 Kan. at 202. But following a discussion of other alternative means cases in Kansas, the Supreme Court specifically disapproved of language in State v. Dixon, 279 Kan. 563, 604-06, 112 P.3d 883 (2005), which had allowed harmless error analysis in an alternative means case. 290 Kan. at 206. The Wright court determined that Timley and Dixon “ ‘simply cannot coexist.’ ” 290 Kan. at 205.
Timley actually contains very little analysis of the alternative means issue. In fact, the opinion only refers to the issue in order to explain die difference between an alternative means case and a *1011multiple acts case. 255 Kan. at 288-90. Timley stated the rule of law to determine whether an alternative means error has been committed, but the opinion never stated that the error could not be harmless. The Supreme Court later explained in Dixon that an alternative means error can be harmless. See 279 Kan. at 605-06. So in other words, Timley established the rule of law to determine whether an alternative means error has been committed; Dixon later clarified that an alternative means error sometimes can be harmless. I believe that the two decisions can peacefully coexist.
I begin by noting that the United States Supreme Court agrees that an alternative means error can be analyzed for harmlessness. In Hedgepeth v. Pulido, 555 U.S. 57, 129 S. Ct. 530, 172 L. Ed. 2d 388 (2008), a California jury convicted Michael Pulido of felony murder and, on his direct appeal, Pulido argued that the juxy instructions erroneously allowed the juiy to find him guilty of felony murder if he formed the intent to aid and abet the underlying felony before the murder or if he formed the intent only after the murder. The California Supreme Court found that the latter theory was invalid under California law, but it did not reverse the conviction because the court found that the error did not prejudice Pu-lido. After his direct appeal, Pulido sought habeas relief in federal court, and the federal ¿strict court granted the relief, finding that instructing the jury on the legally invalid theory had a “ ‘ “substantial and injurious effect or influence in determining the juiy’s verdict.” ’ ” 555 U.S. at 59. The State appealed, and the Ninth Circuit Court of Appeals affirmed the federal district court’s decision granting Pulido a new trial. But the Ninth Circuit court went further by agreeing with Pulido that instructing a jury on multiple theories of guilt, one of which is legally improper, is “structural” error that is not subject to harmless error review. See 555 U.S. at 59-60.
The State appealed to the United States Supreme Court. The Court first noted that the precedent relied upon by the Ninth Circuit based its structural error analysis on opinions issued prior to the Court’s conclusion in Chapman v. California, 386 U.S. 18, 87 S. Ct. 824, 17 L. Ed. 2d 705 (1967), that constitutional errors can be harmless. Hedgepeth, 555 U.S. at 60. The Court then pointed *1012out a series of post-Chapman cases in which it stated that certain types of instructional error are subject to harmless error analysis. 555 U.S. at 60-61. The Court emphasized that Neder v. United States, 527 U.S. 1, 11, 119 S. Ct. 1827, 144 L. Ed. 2d 35 (1999),
“makes clear that harmless-error analysis applies to instructional errors so long as the error at issue does not categorically ‘ “vitiate[e] all the jury’s findings.” ’ [Citation omitted.] An instructional error arising in the context of multiple theories of guilt no more vitiates all the jury’s findings than does omission or misstatement of an element of the offense when only one theory is submitted[, as happened in Neder].” Hedgepeth, 555 U.S. at 61.
The Court went on to state that drawing a distinction between alternative means errors and instructional errors such as those in Neder would be “ patently illogical.’ ” 555 U.S. at 61. The Court unanimously found that the error in Pulido’s case did not constitute structural error, requiring automatic reversal of his conviction. A majority of the Court also found that the Ninth Circuit had not sufficiently engaged in the proper analysis to determine whether the trial error had a substantial and injurious effect or influence in determining the jury’s verdict. 555 U.S. at 61-62. Accordingly, the Court remanded to the Ninth Circuit for proper application of the “substantial and injurious effect” analysis to determine whether Pulido was entitled to a new trial. 555 U.S. at 62.
Hedgepeth is distinguishable from Shaw’s case because the alternative means error in Hedgepeth was not based upon insufficient evidence. But the United States Supreme Court has held that an alternative means error based on insufficient evidence of one of the means does not provide an independent basis for reversing an otherwise valid conviction. In Griffin v. United States, 502 U.S. 46, 58-60, 112 S. Ct. 466, 116 L. Ed, 2d 371 (1991), the Court addressed the following issue: “[W]hether, in a federal prosecution, a general guilty verdict on a multiple-object conspiracy charge must be set aside if the evidence is inadequate to support conviction as to one of the objects.” 502 U.S. at 47. After a lengthy discussion of prior cases, the Court stated that it knew of no case in which it had set aside a general verdict “because one of the possible bases of conviction was neither unconstitutional. . . nor even illegal. . . *1013but merely unsupported by sufficient evidence.” 502 U.S. at 56. The Court further explained:
“Jurors are not generally equipped to determine whether a particular theory of conviction submitted to them is contrary to law — whether, for example, the action in question is protected by the Constitution, is time barred, or fails to come within the statutory definition of the crime. When, therefore, jurors have been left the option of relying upon a legally inadequate theory, there is no reason to think that their own intelligence and expertise will save them from that error. Quite the opposite is true, however, when they have been left the option of relying upon a factually inadequate theory, since jurors are well equipped to analyze that evidence [citation omitted], . . .
“[I]f the evidence is insufficient to support an alternative legal theory of liability, it would generally be preferable for the court to give ah instruction removing that theory from the jury’s consideration. The refusal to do so, however, does not provide an independent basis for reversing an otherwise valid conviction.” 502 U.S. at 59-60.
I have not reviewed the law in eveiy state to ascertain how many state courts apply harmless error analysis in alternative means cases. But Washington is noteworthy because Washington state courts follow the same rule of law governing alternative means cases that Kansas courts follow: In an alternative means case, unanimity is not required as to the means by which a crime is committed so long as substantial evidence supports each alternative means. State v. Kitchen, 110 Wash. 2d 403, 410, 756 P.2d 105 (1988); State v. Nonog, 145 Wash. App. 802, 811-12, 187 P.3d 335 (2008). Nevertheless, courts in Washington apply harmless error analysis to determine whether an alternative means error constitutes reversible error. In State v. Lobe, 140 Wash. App. 897, 167 P.3d 627 (2007), the defendant was convicted of, among other things, two counts of witness tampering, which may be committed by the alternative means of “attempting to induce a person to (1) testily falsely or withhold testimony, (2) absent himself or herself from an official proceeding, or (3) withhold information from a law enforcement agency. [Citation omitted.]” 140 Wash. App. at 902-03. On appeal, the defendant argued that his convictions should be overturned on jury unanimity grounds. The State conceded it had failed to present substantial evidence on each alternative means for each count but argued that the error was harmless.
*1014The Washington Court of Appeals applied harmless error analysis, stating that “where there are three alternative means of committing a crime and the jury is instructed on all three, either (1) substantial evidence must support each alternative means on which evidence or argument was presented or (2) evidence and argument must have only been presented on one means.” 140 Wash. App. at 905. In other words, the court reasoned that although substantial evidence was not presented on each alternative means, the verdict could be upheld if evidence was presented on only one means and jury unanimity was assured. Ultimately, the court found that the error in the case required reversal because it left too much doubt as to whether the jury relied on a means for which substantial evidence was not submitted. 140 Wash. App. at 906-07. However, the dissenting opinion would have affirmed tire conviction based on harmless error. 140 Wash. App. at 911-13 (Hunt, J., dissenting). See also State v. Allen, 127 Wash. App. 125, 130, 110 P.3d 849 (2005) (“But if one or more of the alternative means is not supported by substantial evidence, the conviction must be reversed unless this court can determine that the verdict was based on only one of the alternative means and that substantial evidence supported that alternative means.”).
Returning to Kansas law, generally, errors at any stage of trial proceedings are subject to harmless error analysis under K.S.A. 2011 Supp. 60-261, which provides:
“Unless justice requires otherwise, no error in admitting or excluding evidence, or any other error by the court or a party, is ground for granting a new trial, for setting aside a verdict or for vacating, modifying or otherwise disturbing a judgment or order. At every stage of the proceeding, the court must disregard all errors and defects that do not affect any party’s substantial rights.” (Emphasis added.)
Our Supreme Court has set forth the following test for determining whether an error is harmless:
“[B]efore a Kansas court can declare an error harmless it must determine that the error did not affect a party’s substantial rights, meaning it will not or did not affect the trial’s outcome. The degree of certainty by which the court must be persuaded that the error did not affect the outcome of die trial will vary depending on whether the error implicates a right guaranteed by the United States Constitution. *1015If it does, a Kansas court must be persuaded beyond a reasonable doubt that there was no impact on the trial’s outcome, i.e., there is no reasonable possibility that the error contributed to the verdict. If a right guaranteed by the United States Constitution is not implicated, a Kansas court must be persuaded that there is no reasonable probability that the error will or did affect the outcome of the trial.” State v. Ward, 292 Kan. 541, 565, 256 P.3d 801 (2011), cert denied 132 S. Ct. 1594 (2012).
. Not all errors are subject to this harmless error analysis. Structural error occurs in very limited circumstances where the error “ ‘affect[s] the framework within which the trial proceeds’ ” and thus defies analysis by harmless error standards. Boldridge v. State, 289 Kan. 618, 627, 215 P.3d 585 (2009) (quoting Arizona v. Fulminante, 499 U.S. 279, 309-10, 111 S. Ct. 1246, 113 L. Ed. 2d 302 [1991]). Errors which have been found to be structural include the denial of the right to counsel, the denial of the right to public trial, and the denial of the right to trial by jury. Boldridge, 289 Kan. at 627-28. But other errors of constitutional magnitude are simply errors in the trial process itself and are subject to harmless error analysis. See Neder, 527 U.S. at 8 (“ ‘[I]f the defendant had counsel and was tried by an impartial adjudicator, there is a strong presumption that any other [constitutional] errors that may have occurred are subject to harmless error analysis.’ ”).
This brings us to the question: Should an alternative means error be subject to harmless error analysis? An alternative means error presents a hybrid issue. The error is committed when the district court instructs the juiy on an alternative means of committing a crime that is not supported by substantial evidence. In this sense, an alternative means error is an instructional error. But the error implicates insufficiency of the evidence to the extent that the juiy may have convicted the defendant on at least one means of committing the crime that is not supported by substantial evidence.
The Wright court rejected harmless error analysis in alternative means cases with the following reasoning:
“ ‘[A] reversal mandated by Timley is a reversal for insufficient evidence. An insufficiency error cannot be harmless because it means the State failed to meet its burden of proving the defendant guilty beyond a reasonable doubt. This is a most basic guarantee of due process in criminal cases. [Citation omitted.]
*1016“ ‘The Timley super-sufficiency condition evolved for a good reason. It evolved because we recognized that we were allowing uncertainty as to how the State persuaded each juror. We were comfortable with this uncertainly — at that particular level of generality in the jury’s factfinding — only because we insisted on assurance that each juror’s vote was supported by a means for which there was sufficient evidence. Widrout that assurance, we are back to where we were before Timley. We have no guarantee drat die jury was unanimous at the level of factual generality drat matters most of all: guilt v. innocence.’ ” 290 Kan. at 205 (quoting Beier, Lurching Toward the Light: Alternative Means and Multiple Acts Law in Kansas, 44 Washburn L.J. 275, 299 [2005]).
In Wright, the Supreme Court reasoned that a reversal based on an alternative means error “is a reversal for insufficient evidence” and thereby cannot be harmless. 290 Kan. at 205. But this is an incomplete analysis of the issue. Granted, an alternative means error means that the State failed to present sufficient evidence to prove at least one of the alternative means of committing the crime. But this fact does not eliminate harmless error from consideration provided there was sufficient evidence to prove another means of committing the crime. The question then becomes: Based on the evidence presented at trial and the arguments made to the jury, does the record clearly establish that the jury found the defendant guilty of only the alternative means of committing the crime that is supported by sufficient evidence? If the answer to this question is yes, then we have a unanimous verdict supported by sufficient evidence and the conviction is valid.
I find it noteworthy that the Kansas Supreme Court’s treatment of alternative means cases is inconsistent with its treatment of multiple acts cases; A multiple acts error is committed when the prosecutor fails to elect a particular criminal act in which it will rely for a conviction, or the trial court fails to instruct the jurors that all of them must agree on the same underlying criminal act. Timley, 255 Kan. at 289. When a multiple acts error is committed, there is uncertainty whether the juiy found the defendant guilty of the same underlying criminal act. With a multiple acts error, like with an alternative means error, “ ‘[w]e have no guarantee that the jury was unanimous at the level of factual generality that matters most of all: guilt vs. innocence.’ ” See Wright, 290 Kan. at 205. But our Supreme Court holds that a multiple acts error is subject to harm*1017less error analysis. State v. Voyles, 284 Kan. 239, 252-53, 160 P.3d 794 (2007). If a multiple acts error is subject to harmless error analysis, then an alternative means error also should be analyzed for harmlessness.
The multiple acts issue and the alternative means issue are closely related; both issues implicate jury unanimity but in different ways. Thus, the two issues can be analyzed in a similar manner. Appellate courts can apply a modified three-step analysis for alternative means cases that is similar to the rubric our Supreme Court has adopted for multiple acts cases. See Voyles, 284 Kan. at 244-45. Step one: Do we have an alternative means crime? This is determined by examining the statutory definition of the crime to determine whether the statute creates two or more distinct ways of committing an offense. See State v. Schreiner, 46 Kan. App. 2d 778, Syl. ¶ 1, 264 P.3d 1033 (2011), petition for rev. filed December 5, 2011.
Step two: Assuming the defendant was convicted of an alternative means crime, was error committed? In an alternative means case, error is committed when the district court instructs the jury on alternative means of committing a crime and the State presents insufficient evidence to prove each alternative means. This is the test enunciated in Timley and Wright.
Step three: Assuming there was an alternative means error, is the error reversible? In a multiple acts case, the ultimate test for harmlessness is the “clearly erroneous” standard articulated by the Kansas Legislature in K.S.A. 22-3414(3). See Voyles, 284 Kan. at 252-53. “An instruction is clearly erroneous only if the reviewing court is firmly convinced there is a real possibility the jury would have returned a different verdict if the trial error had not occurred.” State v. Martinez, 288 Kan. 443, 451-52, 204 P.3d 601 (2009). This standard may be applicable in an alternative means case to the extent that the error is committed when the trial court gives a jury instruction containing an alternative means error and the defendant fails to object.
But generally, the burden to establish harmless error should be on the party who benefits from the error. Ward, 292 Kan. at 568-69. Also, the burden to establish harmless error in alternative *1018means cases should be set high to deter prosecutors and trial judges from routinely infecting jury instructions with alternative means that are not supported by the evidence. And as the Wright court stated, an alternative means error implicates “ ‘a most basic guarantee of due process in criminal cases.’ ” 290 Kan. at 205. So rather than applying the clearly erroneous standard that is used in a multiple acts case, it seems appropriate in an alternative means case to apply harmless error analysis under K.S.A. 2011 Supp. 60-261, including the test for determining harmless error articulated by the Kansas Supreme Court in Ward, 292 Kan. at 565.
Applying this three-step test to the facts herein results in the following analysis. Step one: We have an alternative means crime. Involuntary manslaughter while driving under the influence of alcohol (DUI) can be committed in three distinct ways: in the commission of DUI, in an attempt to commit DUI, or in flight from a DUI. Step two: Error was committed when the district court instructed the jury on an alternative means that was not supported by substantial evidence, i.e., involuntary manslaughter committed in flight from a DUI. Step three: The error was not reversible in this case because the jury never heard evidence concerning flight from a DUI and the prosecutor never argued or even mentioned this theory to die jury. Based on the evidence presented at trial and the arguments made to the jury, there is no reasonable possibility that any one of the jurors could have found Shaw guilty of involuntary manslaughter committed in flight from a DUI. Instead, die record clearly establishes that the jury unanimously found Shaw guilty of involuntary manslaughter committed in the commission of DUI, and this means of committing the crime is supported by sufficient evidence. The only reasonable conclusion is that the alternative means error did not affect Shaw’s substantial rights, meaning it did not affect the outcome of the trial. See Ward, 292 Kan. at 565.
I respectfully suggest that in Wright, the Kansas Supreme Court unnecessarily eliminated harmless error analysis from alternative means cases. This case provides a good example of why that ruling was a mistake. The only issue at Shaw’s trial was whether Aaron *1019Kichler’s death was caused by his speeding rather than by Shaw’s DUI. Shaw received a fair trial on that issue, and the jury unanimously returned a verdict of guilty, which is supported by sufficient evidence. Today, we are reversing Shaw’s conviction due to an error that did not impact the structure of his trial and certainly had no bearing on the outcome of the case. Remand for retrial under these circumstances is both unnecessary and unwarranted.