Summary
holding that election error is analyzed under the harm standard applicable to constitutional errors
Summary of this case from McQueen v. ThalerOpinion
No. PD-1576-05.
September 13, 2006.
Appeal from the 23rd District Court, Brazoria County, K. Randall Hufstetler, J.
Joseph W. Varela, Houston, for Appellant.
Jeri Yenne, Criminal District Atty., David Bosserman, Assistant District Atty., Angleton, Matthew Paul, State's Atty., Austin, for State.
KELLER, P.J., delivered the opinion of the Court in which WOMACK, KEASLER, HERVEY, HOLCOMB, and COCHRAN, JJ., joined.
What kind of harm analysis is appropriate when a trial court errs by refusing a defendant's request to require the State to elect from among multiple offenses? How is it to be employed? The Court of Appeals found harm under a constitutional-error harm analysis and reversed the conviction. We find that the Court of Appeals erred in the manner of its analysis, and we reverse its judgment.
I. BACKGROUND A. Trial
Appellant was indicted for aggravated sexual assault of a child. The victim was appellant's niece, who was six years old at the time the abuse occurred and seven years old at the time of trial. During the year 2002, while the child was six, appellant lived with the child, her mother, and the mother's boyfriend. Appellant slept on the couch in the child's bedroom. In response to the prosecutor's inquiries, the child related a sequence of events that occurred every time appellant sexually assaulted her. Each time, appellant would undress himself and remove the child's underwear. He would touch her "private parts" with his hand and then touch her "private parts" with his "private parts." The child's testimony indicated that, by "private parts," she meant genitalia. According to the victim, this sequence of events occurred one hundred times. Except for one time during the day, this activity always occurred at night. The child could not remember the last time she was sexually assaulted, and she gave no further details regarding this activity.
A nurse reported possible abuse to Child Protective Services (CPS). On July 10, 2002, a case worker for CPS went to the child's home to interview the family, but the mother's boyfriend barred entry. The case worker returned the next day with a co-worker and some police officers, and she spoke to the child and the child's mother. While the case worker was in the home, she saw appellant sleeping on the couch in the child's bedroom. She spoke to him as well and then called her supervisor. After speaking to her supervisor, the case worker had the child taken into CPS's custody. She also instructed the police officers to take the comforter from the child's bed to test for evidence of semen.
On July 18, 2002, the victim was taken to a pediatrician for an examination. The examination revealed that the hymen was completely absent from the "five o'clock" to "eight o'clock" positions. The pediatrician remarked that this condition was uncommon even in child abuse victims and constituted "clear evidence of penetrating vaginal trauma" stemming from multiple incidents of abuse. However, the pediatrician could not say when the hymenal injury occurred. The examination also revealed two bumps that looked like the beginning of genital warts, but there were no indications of injury to the labia, the vestibule, or the clitoris. The child was examined too late for semen to be present.
Semen stains were found on the comforter, but the condition of the stains were such that there was no way of knowing how long they had been present. Appellant's own testimony suggests that he admitted to law enforcement that the semen stains were his, because he had slept in the child's bed when she was not at home.
The interchange occurred during cross-examination as follows:
Q. And you told Officer Worsham that your semen would be on die comforter because you slept on there, is that correct?
A. I guess.
During redirect, appellant explained that the child was not there when he slept in her bed. Outside the presence of the jury, the State elicited testimony that one of the semen stains was in sufficient condition for DNA comparison, and that the DNA test showed a match with appellant's DNA profile. But the trial court suppressed evidence of appellant's DNA profile due to inaccurate information contained in the warrant used to obtain appellant's blood sample. Consequently, we do not consider the DNA evidence in our analysis.
When the child testified that the sexual assaults had occurred one hundred times, defense counsel requested a limiting instruction. At the end of the State's case-in-chief and at the close of the evidence, defense counsel requested that the State be required to elect which instance of sexual assault it would rely upon for conviction. Defense counsel also requested that the jury charge be modified to reflect an election for one offense. These requests were denied. In denying appellant's request at the end of the State's case-in-chief, the trial court commented: "I will deny your motion because I have no earthly idea what to do with the election in the indictment."
B. Appeal
The Court of Appeals held that the trial court erred in failing to grant appellant's election request. Deciding that the error was constitutional in nature, the court further found that it was not harmless beyond a reasonable doubt. In arriving at this latter conclusion, the appellate court relied upon two of its prior cases: Phillips v. State and Farr v. State . Phillips gave four reasons for the election rule:
Dixon v. State, 171 S.W.3d 432, 434-435 (Tex.App.-Houston [14th Dist.] 2005).
Id. at 435-437.
130 S.W.3d 343 (Tex.App.-Houston [14th Dist.] 2004), aff'd, 193 S.W.3d 904 (Tex.Crim.App. 2006).
140 S.W.3d 895 (Tex.App.-Houston [14th Dist.] 2004), aff'd sub. nom., Phillips v. State, 193 S.W.3d 904 (Tex.Crim.App. 2006).
Dixon, 171 S.W.3d at 436-437.
(1) to protect the accused from the introduction of extraneous offenses,
(2) to minimize the risk that the jury might choose to convict, not because one or more crimes were proved beyond a reasonable doubt, but because all of them together convinced the jury the defendant was guilty,
(3) to ensure unanimous verdicts, that is, all of the jurors agreeing that one specific incident, which constituted the offense charged in the indictment, occurred, and
(4) to give the defendant notice of the particular offense the State intends to rely upon for prosecution and afford the defendant an opportunity to defend.
See Phillips, 130 S.W.3d at 349.
See Phillips, 130 S.W.3d at 349.
Of the four underlying reasons cited by Phillips, the Court of Appeals emphasized the requirement of juror unanimity, which it believed was undermined in the present case because, without an election, "some [jurors] may have selected a nighttime offense, while others may have chosen the daytime assault." In passing, the appellate court also suggested that the refusal to require an election deprived appellant of notice of what offense the State intended to rely upon and exposed him to the risk that the jury would convict in the absence of proof beyond a reasonable doubt.
Id. at 436.
Id. at 436 n. 6.
In a concurring opinion, Justice Frost gave her opinion that the election error was harmful even under the standard for nonconstitutional errors. In a dissent, Justice Hudson concluded that the error was harmless, even under the constitutional standard, because none of the four purposes were frustrated by the failure to grant the election.
Id. at 438-439 (Frost J., concurring).
Id. at 440-442 (Hudson J., dissenting).
C. Discretionary Review
In its petition for discretionary review, the State contends that the Court of Appeals made two mistakes: (1) finding harm under the facts of the case, and (2) using the standard of harm for constitutional, rather than nonconstitutional, errors.
II. ANALYSIS
One of our seminal election cases conducted a harm analysis, without explaining whether or not the underlying error was constitutional. In O'Neal v. State, the defendant's stepdaughter testified that the defendant "had been having sexual intercourse with her on a regular basis since she was five years old." The stepdaughter "then gave a detailed account of an act of intercourse that occurred on or about April 24, 1984." In addition, the defendant's daughter, who slept in the same bed, "testified that she had witnessed numerous acts of intercourse between appellant and the complainant over the past three or four years." The defendant moved to elect both at the end of the State's case and at the close of all the evidence, but the trial court did not grant the motion until the latter point in time, and this Court found this delay in granting the motion to be error. In finding the error harmless, this Court emphasized that the testimony of both the complainant and her stepsister focused on the act that occurred on April 24, 1984, and consisted of a detailed account of the incident. Under the circumstances, the State's case-in-chief evidence "clearly gave notice to the appellant which act of intercourse the State would rely upon for conviction," and so the defendant was not harmed by the late election.
746 S.W.2d 769, 770 (Tex.Crim.App. 1988).
Id.
Id.
Id. 771.
Id. at 772.
Id.
Id. at 773.
Unlike in O'Neal, the complaining witness in this case did not testify about one distinct, detailed incident. Rather, she described the manner in which appellant sexually assaulted her and said that it occurred numerous times. Consequently, all of the incidents presented in the case were presented with equal specificity, and, except for the fact that one incident occurred during the day, none of the incidents were distinguished in any manner from each other. O'Neal did not, however, purport to hold that error was harmless only under the circumstances before it; the Court simply found the error in that case to be harmless for the reasons it outlined, without attempting to determine how other situations should be treated. We turn, then, to the Court of Appeals's handling of the harm analysis in the present case.
Recently, on discretionary review from the court of appeals's decision in Phillips, this Court has held that election error, at least insofar as it implicates notice and jury unanimity, is analyzed under the harm standard applicable to constitutional errors. This Court also approved as accurate the four purposes cited therein as underlying the election rule. We turn, then, to those purposes in addressing the question of harm.
Phillips v. State, 193 S.W.3d 904 (Tex.Crim.App. 2006).
Id. at 913-914.
The first purpose may be easily dispensed with: appellant was not entitled to be protected from the admission of evidence of extraneous sexual offenses committed by him against the child. Article 38.37 permits the admission of evidence of these offenses to show the previous and subsequent relationship between appellant and the child victim. It may be argued that protection against the introduction of extraneous offenses includes protection against impermissible uses via a limiting instruction. Even if that is so, a limiting instruction would have been of no practical use in the present case, where no attempt was made by the child to meaningfully distinguish the various offenses.
TEX. CODE CRIM. PROC, Art. 38.37, §§ 1(1), 2(2). The State filed a notice of intent to introduce extraneous offenses that included appellant touching the victim's sexual organ with his hand and his sexual organ "on numerous occasions in Brazoria County Texas within the last six years." See Art. 38.37, §§ 3.
With regard to the second purpose, we see no risk that the jury found appellant guilty of an offense that was not proven to its satisfaction beyond a reasonable doubt. The "multiple offenses" were all recounted by the same source — the child. This case is not concerned with evidence of different activities from different sources that a jury might perceive to "add up" to the defendant being guilty even though no individual offense was proven beyond a reasonable doubt. Moreover, the child complainant did not testify about a number of varied incidents with differing details that might have incrementally added to the idea that the defendant must have done something to provoke the plethora of stories about his activities. Rather, the child articulated one sequence of events and merely answered that this sequence happened one hundred times, with all but one of those instances occurring at night. The child was either credible in giving this unified account or she was not. Whether the sequence of events was alleged to have occurred one, ten, fifty, or one hundred times does not by itself impact the believability of the child's story.
Of course, the child's allegation of multiple instances is made more credible by its consistency with the medical evidence. But that fact mitigates against harm rather than in favor of it. Because the medical evidence very strongly pointed to multiple incidents of sexual trauma, appellant was faced, not with a risk that the jury would convict despite the State's failure to prove any particular offense beyond a reasonable doubt, but with the overwhelming likelihood that the jury would believe that multiple offenses had been proven beyond a reasonable doubt. Fortunately for appellant, the jury in this case could convict him of only one, and the State's failure to elect means the remaining offenses will be jeopardy-barred.
See Ex parte Goodbread, 967 S.W.2d 859 (Tex.Crim.App. 1998).
We likewise perceive no risk that the present case led to a non-unanimous verdict. The only distinguishing detail among the one hundred offenses is that one occurred during the day, while all the others happened at night. The difference is the result of a single line of the child's testimony — in fact, the use of a single word, "daytime," as denoting one of those one hundred incidents. But there is simply no basis in the record for the jury to believe that one incident occurred during the day but that none occurred at night. The night-time scenario being typical (ninety-nine out of one hundred), it is obvious from this record that anyone who believed the complainant's allegations in any respect would believe that sexual assaults occurred at night.
The dissent contends that "with one hundred undifferentiated incidents to consider as evidence of one charged offense, there is a distinct danger that the jurors will fail to reach a unanimous verdict . . . but . . . could convict on as many as twelve different incidents." Dissent at 739. But it is meaningless to say the jurors in this case could have disagreed about which incident was the basis for conviction. The offenses were described as identical and any one of them could have occurred on any given day during the time period appellant stayed with the victim's family. Aside from the day/night distinction discussed in the body of this opinion, what precisely would the jurors be disagreeing about? Clearly, however, the jurors unanimously agreed that appellant committed at least one sexual assault, at night, sometime during the year, in the manner described by the complainant.
Finally, we reject the notion that appellant was deprived of adequate notice. As we have already discussed, the only distinction made between the incidents is that one occurred during the day. But "time is not [usually] a material element of an offense," and in some cases "it may be impossible for the State to know precisely, or even approximately, when the charged offense occurred." Especially where young children are involved, we have cautioned that courts cannot impose unrealistic expectations regarding proof of when an offense actually occurred: "[I]t is not often that a child knows, even within a few days, the date that she was sexually assaulted. And, the younger the child, the greater the possibility" that she will be uncertain about the timing of the offense. Without more, a variance in the time of day is not a basis for claiming lack of notice. And in this case, there is no more.
Garcia v. State, 981 S.W.2d 683, 686 (Tex.Crim.App. 1998).
Sledge v. State, 953 S.W.2d 253, 256 n. 8 (Tex.Crim.App. 1997).
The dissent claims that we err in relying upon Garcia for the proposition that the failure to elect in this case did not deprive appellant of notice. However, we cite Garcia, in part, to answer the Court of Appeals's suggestion that appellant needed notice of whether the offense occurred during the day or at night, and the dissent does not explain why our citation to Garcia fails in that purpose.
The dissent does suggest that notice requires enabling the defendant "to differentiate [the incident relied upon for conviction] from all other, uncharged-but-admissible incidents," dissent at 738, but the dissent does not offer a reason for imposing such a requirement. When a young child is simply unable to differentiate between the incidents other than to say they occurred at different points in time, and no other evidence in the case does so, any attempt by the State to differentiate incidents would be wholly arbitrary. While we may require such an attempt as part of a generally applicable rule, and in an abundance of caution, we cannot say that the failure to differentiate in such a situation truly deprives the defendant of notice, when the State has given all the notice it can.
Further, there is no real doubt regarding which option the State would pick if forced to choose whether the offense relied upon was committed during the day or at night: it would pick the night, a characteristic shared by ninety-nine percent of the offenses about which the victim testified. And as previously discussed, there is no remotely significant risk, under the facts of this case, that a jury would convict appellant without believing that he committed an offense at night.
We are confident beyond a reasonable doubt that the error in failing to require an election did not contribute to appellant's conviction or punishment. Consequently, we reverse the judgment of the Court of Appeals and remand the case to that court for further proceedings consistent with this opinion.
COCHRAN, J., filed a concurring opinion in which HOLCOMB, J., joined.
PRICE, J., filed a dissenting opinion in which MEYERS and JOHNSON, JJ., joined.
It is a truism that bad facts make bad law.
The facts in this case are bad. They are also depressingly familiar: a young child is repeatedly molested by an authority figure — usually a step-parent, grandparent, uncle or caregiver; there is (or is not) medical evidence of sexual contact; and the child is too young to be able to differentiate one instance of sexual exposure, contact, or penetration from another or have an understanding of arithmetic sufficient to accurately indicate the number of offenses. As in this case, "he did it 100 times." The real gravamen of this criminal behavior is the existence of a sexually abusive relationship with a young child, male or female, marked by continuous and numerous acts of sexual abuse of the same or different varieties. This scenario plays itself out in Texas courtrooms every day.
Yet, as is evidenced by this case, current Texas law does not easily accommodate the prosecution of generic, undifferentiated, ongoing acts of sexual abuse of young children. This is because our penal statutes are intended to prosecute a person who commits one discrete criminal offense at one discrete moment in time. Our criminal procedures are intended to protect a defendant from being tried for being a "bad" person who acts in conformity with a criminal character propensity. Our rules are intended to give the defendant advance notice of precisely what criminal act he is alleged to have committed and when it occurred. Our state constitution requires the jurors to make a unanimous decision on the occurrence of one specific criminal act. The law focuses the advocates, judge, and jurors on whether the person charged is guilty of this one, very specific criminal act that he is charged with having committed.
We are headed for a train wreck in Texas law because our bedrock procedural protections cannot adapt to the common factual scenario of an ongoing crime involving an abusive sexual relationship of a child under current penal provisions. The general legal principles are being stretched beyond recognition and common logic in what appears to be a futile attempt to accommodate both (1) the defendant's rights to a specific verdict for one specific criminal act and (2) the simple fact that the criminal conduct at issue is not really one specific act at one specific moment. Our bedrock legal doctrines are becoming ever more convoluted, contradictory, and inconsistent as a result. Worst of all, these convolutions, contradictions, and inconsistencies affect not only Texas law as it applies to the present fact scenario, but they leak out into other factual scenarios and the trial of other offenses, affecting trials of all varieties.
The procedural laws and legal principles adversely affected include: due-process notice, statutes of limitation, grand-jury indictment, specificity of the indictment, counts and paragraphs within an indictment, the prosecution's duty to elect a specific act upon which it will rely, the admissibility of extraneous offenses, jury charge application paragraphs, the requirement of a unanimous verdict, and double jeopardy protections.
Perhaps the Texas Legislature can address this conundrum and consider enacting a new penal statute that focuses upon a continuing course of conduct crime — a sexually abusive relationship that is marked by a pattern or course of conduct of various sexual acts. Such a statute would have advantages and disadvantages for both the prosecution and defense, but it might well assist in preserving our bedrock criminal-procedure principles of double jeopardy, jury unanimity, due-process notice, grand-jury indictments, and election law.
In the meantime, I join the majority opinion in this case, although I find much merit in the arguments of the dissent.
It is debatable whether the complainant testified to one discrete occurrence, and one hundred additional, uncharged occurrences, or one hundred incidents of sexual assault that occurred in exactly the same manner every time (except for one daytime incident). This ambiguity stems from the State's mode of questioning, which asked what "would" the appellant do, rather than what "did" the appellant do. If we consider the testimony as a description of a single incident, combined with general testimony of one hundred other occurrences, O'Neal v. State clearly controls, and the error at failing to elect was "harmless" in that there was no danger that the defendant or the jury would be genuinely confused as to the incident upon which the State was relying to convict.
746 S.W.2d 769 (Tex.Crim.App. 1988).
Nevertheless, the majority does not read the record in this way. Neither did the court of appeals. Instead, the majority reads the complainant's testimony as a description of one hundred occurrences of the same conduct and says that "all" of the one hundred occurrences were presented with "equal specificity" — which is to say, with almost no specificity at all. The majority then asserts that the appellant "was not entitled to be protected from admission of evidence of extraneous sexual offenses committed by him against a child" because Texas Code of Criminal Procedure Article 38.37 "permits the admission of these offenses to show the previous and subsequent relationship between appellant and the child victim." It is true that Article 38.37 allows for the admission of uncharged misconduct for relevant purposes. But it does not allow for admission of uncharged misconduct for any and all purposes, and it remains incumbent on the State to prove up a discrete offense for which it will rely to convict, and to elect the incident it will rely upon to convict when it proves up more than one. Article 38.37 in no way permits the State to prove a single indicted count by presenting unelected, undifferentiated evidence of one hundred incidents.
Op. at 734.
Id., at 734.
Further, I disagree that Garcia v. State stands for the proposition that the failure to elect in this case was harmless because the State was not bound by the date alleged in the indictment. Garcia does not address harm resulting from the failure to elect. Garcia merely stands for the proposition that the State is not bound to the specific date of an indictment, and that the "on or about" language in an indictment has the primary purpose of establishing that the State is not barred by the applicable statute of limitations. Thus, the State need not prove an exact date in order to convict.
981 S.W.2d 683 (Tex.Crim.App. 1998).
Id., at 686.
This proposition in no way, however, allows the State to prove the indicted charge by presenting an array of undifferentiated, uncharged misconduct, without electing to proceed upon a specific, discrete incident. The State must still provide evidence of a discrete incident (even if the date is uncertain) to prove any given charged offense beyond a reasonable doubt and, when evidence of other occurrences is admitted (say, under Article 38.37), it must still provide the defendant with notice by electing the incident it will rely upon for conviction, to differentiate that incident from all the other, uncharged-but-admissible incidents. To hold otherwise is an invitation for the State to throw multiple, non-specific accusations of misconduct at a defendant in hopes of convicting him, not because it has proved a discrete incident beyond a reasonable doubt, but because it has proved enough amorphous misconduct that the jury believes he is a criminal in a more general sense.
Further, with one hundred undifferentiated incidents to consider as evidence of one charged offense, there is a distinct danger that the jurors will not only fail to reach a unanimous verdict in convicting the defendant, but that they could convict on as many as twelve different incidents. Whatever the jeopardy implications, clearly such a verdict would meet neither the unanimous jury guarantee of the Texas Constitution, nor the "substantial majority" requirement of the Sixth Amendment. Error of this kind vitiates the entire jury verdict, calling into question whether the appellant received the jury trial guaranteed by the Sixth Amendment at all. Accordingly, I do not agree that the trial court's error in failing to require an election when the State sought to prove a single count with evidence of one hundred, unelected, undifferentiated incidents is harmless error.
TEX. CONST., ART. V, § 13.
See Johnson v. Louisiana, 406 U.S. 356, 362, 92 S.Ct. 1620, 32 L.Ed.2d 152 (1972).
U.S. CONST, amend. VI; see Sullivan v. Louisiana, 508 U.S. 275, 279-80, 113 S.Ct. 2078, 124 L.Ed.2d 182 (1993) (charging error relating to definition of burden of proof not subject to harm analysis because it wholly deprives defendant of the jury verdict the Sixth Amendment entitles him to).
I respectfully dissent from the judgment of the Court to reverse the court of appeals.