Fernandez v. Cook

870 P.2d 870, 217 Utah Adv. Rep. 3, 1993 Utah LEXIS 103, 1993 WL 264683

• Court: Utah Supreme Court
• Decided: July 12, 1993
• Docket: 910395
• Status: Published

Opinions

HALL, Chief Justice:

The State appeals the grant of a petition for habeas corpus and the denial of the State’s motion for new trial. We reverse.

On May 9,1986, a jury convicted Anastacio Fernandez,' Jr., of two counts of rape of a child in violation of Utah Code Ann. § 76-5-402.1. At trial, the State’s primary witness was Fernandez’ daughter (“P.F.”), who stated that from the time she was eleven years old to the time she was thirteen years old, Fernandez had sexual intercourse with her “a lot.”¹ She testified in detail of two specific incidents when her father had intercourse with her: one occurring on October 30, 1983, and one on April 18, 1985. On cross-examination, P.F. admitted that she did not get along with her father and that she preferred living with her mother and without her father.

The State also presented the testimony of Dr. Kent Gibbs, the physician who examined P.F. subsequent to her accusations that Fernandez had intercourse with her. Dr. Gibbs examined P.F. when she was fourteen years old, approximately one year after the last time Fernandez allegedly had intercourse with her. Dr. Gibbs testified that after performing a pelvic examination on P.F., his findings were inconclusive as to whether she had had intercourse.² Dr. Gibbs further testified that a girl similarly situated who regularly engaged in intercourse a year before being examined would “not necessarily” possess a different physical makeup than P.F.’s.

Dr. Gary Sazama, a clinical psychologist, also testified on behalf of the prosecution. Dr. Sazama testified that after a two-hour interview with P.F., he concluded that she suffered from post-traumatic stress disorder, which is characteristic in rape and molestation victims. On cross-examination, Dr. Sa-zama admitted that stress could also result from fabricating a sex abuse story and anticipating a courtroom confrontation. Dr. Saza-ma later opined that P.F.’s particular condition did not result from pretrial stress.

Fernandez was represented by Clint Jud-kins at trial. Judkins presented the testimony of Fernandez and Arselia Fernandez (Fernandez’ wife and P.F.’s mother). Fernandez denied all of P.F.’s -accusations that he had had intercourse with her. He also testified that he and P.F. argued and that he thought her “story” was fabricated to get him out of the house. Consistent with the testimony of Fernandez and P.F., Arselia Fernandez testified that P.F. and her father did not get along and that P.F. was happier when he was gone. Arselia Fernandez claimed that P.F. had never discussed her father’s alleged misconduct with her. She further testified that she did not believe that the alleged rapes occurred and that P.F. was not always truthful.

The jury resolved the dispute in the evidence against Fernandez. He was sentenced to two consecutive minimum mandatory terms of fifteen years to life. Fernandez appealed the sentence and later, on order by this court, was resentenced to two fifteen-year terms to run concurrently instead of consecutively. Fernandez then filed a pro se appeal from the new sentencing order. That appeal was subsequently dismissed on motion of Fernandez’ present counsel.

In April 1988, Fernandez filed a petition for habeas corpus in the Third Judicial District Court. He sought a new trial, arguing that he was denied his constitutional right to effective assistance of counsel and that he was further denied the constitutional right to trial by an impartial jury. The district court dismissed the petition, finding that the issues could have been raised on direct appeal and that there were no unusual circumstances that would justify failure to do so. Fernandez appealed the district court’s ruling. This court reversed and remanded, holding that Fernandez’ failure to raise the claim of ineffective assistance of counsel on appeal did not preclude him from raising it in a habeas corpus proceeding where the alleged ineffective counsel had represented Fernandez on .direct appeal and did not raise the issue.³

On March 23, 1990, the district court held a hearing on Fernandez’ petition for habeas corpus. Among those testifying was Dr. Gibbs. In addition to presenting his findings of the examination performed on P.F., Dr. Gibbs testified that before trial, Judkins had not spoken with him or subpoenaed his records. He also stated that based on the examination, he had written the investigating detective advising that from his findings, he “would be surprised if she was regularly sexually active.” Additionally, based on a hypothetical posed, Dr. Gibbs testified that he would expect an easier examination, one without discomfort, for an eleven- to thirteen-year-old girl who claimed to have had intercourse one hundred times. However, consistent with his prior trial testimony, Dr. Gibbs stated that he could not determine conclusively whether P.F. had had sexual intercourse.

Fernandez testified at the hearing that he had discovered a tape recording wherein P.F. said that she wished Fernandez was back in jail. Fernandez claimed that Judkins knew of the tape but failed to question him about it at trial. Furthermore, Fernandez contended that Judkins was not prepared at trial and that despite their several meetings, Judkins never discussed trial strategy with him.

Judkins also testified at the hearing. Jud-kins testified that he had practiced law for nineteen years before becoming a circuit court judge in 1989 and that during that time, he had extensive experience in criminal law on the side of both the defense and the prosecution. Relative to his representation of Fernandez, Judkins stated that he prepared for trial by reviewing the prosecution’s file, the video tape of P.F.’s police interview, and Dr. Gibbs’ post-examination letter and by conducting a number of interviews. Jud-kins stated that a great amount of research was not necessary because he had “extensive background” in this particular area, representing others similarly situated, and therefore had knowledge of the law and defenses available. At the time of trial, Judkins felt that he was adequately prepared to defend Fernandez. Concerning any conduct or decisions that were being called into question, Judkins claimed that he made decisions and proceeded in a manner that in his judgment would best serve Fernandez’ interests.

Finally, Fernandez called as a “legal expert” Lynn Brown, an attorney with extensive experience in criminal defense. Brown testified as to what he or any reasonable attorney would or would not have done in preparation for Fernandez’ defense. All or most of the questions posed and the responses elicited referred indirectly to what Jud-kins should or should not have done, intimating that Judkins had rendered ineffective assistance.

The district court concluded that Fernandez had received ineffective assistance of counsel, denying him the constitutional right to counsel, and ordered that the ease be retried. The court later entered findings of fact and conclusions of law. The State appealed the district court’s decision on Fernandez’ habeas corpus petition but later sought and obtained a dismissal of the appeal. Back before the district court, the State contended that the findings of fact and conclusions of law were in error and should be set aside. In addition, the State argued that there was insufficient evidence to justify the district court’s decision and moved for a new trial on Fernandez’ habeas corpus petition.⁴ The State’s motion came before the district court for hearing, at which time the court granted Fernandez’ motion in limine precluding P.F., Wendy Sanders, Michelle Sanders, Brenda Riggs, and Shelly Spack-man from testifying at the habeas corpus trial.⁵ The district court subsequently denied the State’s motion for a new trial on the habeas corpus petition.

Before this court, the dispositive issue is whether the district court was correct in granting Fernandez’ petition for habeas corpus based on the legal conclusion that Fernandez was denied effective assistance of counsel.

The applicable standard of review for claims of ineffectiveness of counsel is set forth in Strickland v. Washington.⁶ In Strickland, the United States Supreme Court observed that the principles that apply in addressing ineffective assistance claims, on direct appeal apply in habeas corpus actions as well.⁷ Strickland held that claims of ineffective assistance of counsel involve a mixed question of law and fact.⁸ Thus, “where a trial court has previously heard a motion based on ineffective assistance of counsel, reviewing courts are free to make an independent determination of a trial court’s conclusions.”⁹ However, the findings of fact will not be set aside unless clearly erroneous.¹⁰

In the recent case of State v. Thurman,11 we again addressed the appropriate standard of review of trial court findings of fact as distinguished from the legal conclusions drawn therefrom. We reiterated that while deference is given to underlying findings of fact, which will not be set aside unless clearly erroneous, ultimate conclusions of law are to be reviewed for correctness.¹²

Fernandez claims that he was denied the right to counsel under the Sixth Amendment.¹³ The Sixth Amendment states in pertinent part, “In all criminal prosecutions, the accused shall enjoy the right to ... have the assistance of counsel for his defence.”¹⁴ The guarantee to assistance of counsel has been held to mean “ ‘the right to effective assistance of counsel’ ”¹⁵ In determining whether a criminal defendant has been denied effective assistance of counsel, we rely on the two-part test established by the United States Supreme Court in Strickland.¹⁶

A defendant claiming ineffective assistance of counsel must first demonstrate that counsel’s performance was deficient such that counsel “was not functioning as the ‘counsel’ guaranteed ... by the Sixth Amendment.”¹⁷ This requires a showing “that counsel’s representation fell below an objective standard of reasonableness.”¹⁸ The defendant must then show that his or her defense was prejudiced by counsel’s deficient performance to the extent that “counsel’s errors were so serious as to deprive the defendant of a fair trial, a trial whose result is reliable.”¹⁹ Prejudice is made out if the defendant demonstrates that “there is a reasonable probability that, but for counsel’s unprofessional errors, the result of the proceeding would have been different. A reasonable probability is- a probability sufficient to undermine the confidence of the outcome.” ²⁰ To prevail, a defendant claiming ineffective assistance of counsel must meet both parts of the Strickland test.²¹

After ruling that Fernandez had been denied effective assistance of counsel, the district court entered findings of fact and conclusions of law. A number of the district court’s “findings of fact,” however, are in reality conclusions of law.²² When reviewing a lower court’s findings of fact and conclusions of law, we disregard labels attached to them and look to the substance.²³ Therefore, we review findings of fact that are in actuality conclusions of law for correctness and not under the clearly erroneous standard.²⁴

The district court concluded that counsel “failed to properly prepare for trial and prosecute the case vigorously.” In support of this conclusion, the district court found that counsel did not personally interview Wendy Sanders, Michelle Sanders, Shelly Spackman, or Brenda Riggs, all potential witnesses who had knowledge of the rape allegations, but instead directed Arselia Fernandez to interview these witnesses. The results of these interviews were not made known at trial. Furthermore, the district court found that counsel did not consult Dr. Sazama about his interview with P.F. and his conclusions, nor did he obtain an independent opinion as to the validity of Dr. Sazama’s conclusions so as to rebut or minimize the effect of his testimony-

The district court also drew the following conclusions: Although counsel stated in his opening statement that he would prove fabrication as a defense to the charges of rape, he had not prepared to prove such a defense, and counsel had not prepared to cross-examine P.F. and other witnesses. The district court ultimately concluded that counsel’s lack of preparation constituted deficient performance such that the trial was prejudiced and that Fernandez was therefore denied the constitutional right to effective assistance of counsel.²⁵

On appeal, the State asserts that the district court erred in concluding that Fernandez had been denied effective assistance of counsel. The State contends that counsel’s trial preparation was reasonable, that eoun-sel’s fabrication defense was as compelling as the available evidence would permit, and that the acts and omissions being attacked by Fernandez were tactical decisions. Therefore, counsel’s conduct fell within the wide range of reasonable professional assistance. Furthermore, the State argues that Fernandez failed to show that prejudice resulted from counsel’s allegedly deficient performance.

Fernandez, on the other hand, maintains that counsel was ineffective in many aspects of his representation. Specifically, Fernandez contends that counsel failed to allow himself the necessary time to conduct pretrial investigation²⁶ and consequently did not adequately prepare for trial, considering the gravity of the charges and the complexity of possible medical issues and legal defenses. According to Fernandez, counsel’s lack of preparation included failure to interview several potential witnesses and to undertake independent medical research of the physical trauma an eleven- to thirteen-year-old girl would experience if intercourse had occurred over one hundred times as P.F. alleged. Moreover, Fernandez claims that no effort was made to effectively cross-examine P.F. about the long-standing feud between her and her father to support the theory that she fabricated the story that her father raped her. Fernandez also claims that counsel did not effectively cross-examine Dr. Gibbs regarding the “outrageous” number of times the alleged intercourse took place and the fact that Dr. Gibbs’ examination and medical findings were inconclusive as to whether she had had intercourse.

Additionally, Fernandez contends that counsel performed deficiently when, in his opening statement, he misled the jury in stating that he was prepared to prove that P.F. fabricated her allegations of rape and then not introducing evidence of the longstanding feud between Fernandez and his daughter. Finally, Fernandez claims that counsel’s performance was deficient when counsel failed to conduct formal discovery and request the names of witnesses the prosecution anticipated calling at trial.²⁷ Had he done so, Fernandez claims, counsel could have performed independent research on post-traumatic stress disorder and would have been prepared to effectively cross-examine Dr. Sazama concerning his diagnosis.²⁸

Many of Fernandez’ claims are based on what appear to be tactical or strategic decisions made by trial counsel. It is well established that trial tactics and strategies are within counsel’s prerogative and are generally left to counsel’s professional judgment.²⁹ Counsel is given “wide latitude” when making tactical decisions.³⁰ Accordingly, we will not question strategic decisions unless there is “no reasonable basis” for the decision.³¹

For example, Fernandez contends that counsel did not effectively cross-examine P.F. concerning both the number of times the alleged rapes took place and the feud between her and her father to support the fabrication defense. Fernandez also questions counsel’s decision not to cross-examine Dr. Gibbs concerning his findings and P.F.’s allegations regarding the numerous times that intercourse occurred in light of his testimony that his findings were inconclusive. Notwithstanding Fernandez’ contentions, it is reasonable to assume that counsel did not pursue questions surrounding the number of times the alleged rapes occurred due to the potential for inflammatory impact.

Concerning Dr. Gibbs’ findings, Fernandez has not demonstrated what a more extensive cross-examination would have uncovered. Moreover, it is difficult to imagine that information much more beneficial to Fernandez’ defense could have been obtained in that Dr. Gibbs testified that his examination was inconclusive and he could not determine whether P.F. had engaged in intercourse. In short, the foregoing claims involve counsel’s decisions as to the line of questioning. The decision of whether to pursue a certain line of questioning is entrusted to the judgment of counsel.³²

Also included as a strategic decision is counsel’s decision to call or not to call a certain witness.³³ Fernandez claims that counsel was deficient in not interviewing and later calling several witnesses, in particular witnesses with knowledge of the rape allegations who could possibly be used to impeach P.F. Counsel had Arselia Fernandez interview several witnesses. Fernandez has not shown what, if any, information might have been discovered had counsel interviewed the witnesses personally. From all indications, the information already obtained would have been harmful to Fernandez if used. In fact, Fernandez precluded those witnesses from testifying at the habeas corpus hearing, fearing that they would simply parrot P.F.’s testimony and corroborate her allegations. Moreover, counsel testified that he found nothing from the information obtained indicating a way to impeach P.F.’s testimony. He further stated that such testimony, in his opinion, would have been “devastating” to Fernandez’ defense. If counsel believes that pursuing certain investigations would be fruitless or harmful, we will not question that decision in the absence of a showing that counsel’s belief was unreasonable.³⁴ To do so would be to second-guess counsel’s trial strategy.³⁵

Finally, counsel’s manner of presenting a defense, in this case the fabrication defense, goes to the heart of trial tactics. To support the fabrication defense, counsel presented testimony from both Fernandez and Arselia Fernandez that P.F. and her father had a tumultuous relationship and that P.F. was not always truthful. In light of the available evidence, we agree with the State and the conclusion reached by the judge who presided over the original jury trial that counsel presented a more than adequate defense. It ultimately became a question of credibility, one which the jury found against Fernandez. Counsel’s unsuccessful defense of Fernandez does not equate with deficient performance.³⁶ Accordingly, counsel performed reasonably in his presentation of Fernandez’ defense.

In any event, even assuming that trial counsel was in some way deficient in his performance, Fernandez still must satisfy the prejudice requirement of the Strickland test to obtain relief. We therefore move to the second prong of the Strickland test. As the Court stated in Strickland:

[A] court need not determine whether counsel’s performance was deficient before examining the prejudice suffered by the defendant as a result of the alleged deficiencies .... If it is easier to dispose of an ineffectiveness claim on the ground of lack of sufficient prejudice, which we expect will often be so, that course should be followed.³⁷

Accordingly, without conclusively determining whether counsel’s performance was so deficient as to fall below a reasonably professional standard, we pursue the inquiry of whether counsel’s conduct was sufficiently prejudicial so as to deprive Fernandez of a fair trial. In doing so, we observe that Fernandez bears the burden of proving his claim of ineffective assistance of counsel.³⁸

While Fernandez asserts that counsel’s performance fell below an objective standard of reasonableness in many respects, he fails to address the Strickland prejudice requirement. With regard to all but one of his claims of counsel’s deficiencies, Fernandez is silent as to the required showing of prejudice. The one time Fernandez does address possible prejudice resulting from counsel’s errors, he simply states, “[The trial] cannot be relied upon as having produced a just result.” This is merely rephrasing that which must ultimately be shown to satisfy the second prong of the Strickland test but is clearly insufficient to affirmatively demom strate a reasonable probability that the trial result would have been different if counsel had not performed deficiently.³⁹ Fernandez’ claims of ineffectiveness, therefore, become only speculation. On many occasions, this court has reiterated that proof of ineffective assistance of counsel cannot be a speculative matter but must be a demonstrable reality.⁴⁰

We recognize that in some instances prejudice is presumed.⁴¹ Strickland stated that prejudice is presumed when there has been actual or constructive denial of counsel, when the prosecution has interfered with counsel’s assistance, and when counsel is burdened by a conflict of interest.⁴² The instant case does not present a circumstance where prejudice can be presumed. Therefore, Fernandez must affirmatively prove it.⁴³

A thorough review of the trial record prompts our conclusion that Fernandez has failed to satisfy the Strickland prejudice requirement of showing that but for counsel’s acts or omissions there is a reasonable probability that the result at trial would have been different. On the contrary, the fabrication defense was squarely presented to the jury and was as compelling as the evidence would permit. Accordingly, because Fernandez has failed to show that he was prejudiced by counsel’s performance, we hold that he was not denied effective assistance of counsel.

We therefore reverse and vacate the order of the district court granting Fernandez’ petition for writ of habeas corpus.

DURHAM and ZIMMERMAN, JJ., concur. HOWE, Associate C.J., concurs in the result.

1 . In an earlier video-taped interview, P.F. told Roberta Hardy of the Cache County Child and Family Support Center that the rapes occurred three times a week or seven days in a row. Later, in an interview with the examining psychologist, she claimed that her father raped her several times a week, sometimes seven days in a row. At the preliminary hearing, P.F. testified that intercourse occurred “more than 6 more than 30 around 50" times. At the habeas corpus hearing, Dr. Kent Gibbs, the examining physician, testified that the investigating detective claimed that P.F. reported that intercourse occurred one hundred times.

2 . Dr. Gibbs testified that the pelvic examination showed no evidence of lacerations and that the hymenal band was present entirely throughout the vaginal opening. Dr. Gibbs stated, however, that the hymenal membrane that covers the vaginal opening was not intact. He explained that only a completely intact hymenal membrane, which provides no opening to the vagina, has any relevance or bearing on whether sexual intercourse has occurred. Dr. Gibbs also testified that a one-finger examination showed the vaginal opening to be tight and that insertion of a small speculum caused P.F. discomfort.

3 . Fernandez v. Cook, 783 P.2d 547, 550 (Utah 1989).

4 .The State presented a detailed, three-page affidavit from Judge VeNoy Christoffersen, the judge who presided over Fernandez’ original jury trial, in which Judge Christoffersen stated that Judkins performed in a reasonably professional manner in defending Fernandez and that Judkins’ conduct did not prejudice the trial result.

5 . P.F. had told Wendy Sanders, Michelle Sanders, Brenda Riggs, and Shelly Spackman of the alleged rapes by Fernandez.

6 . 466 U.S. 668, 104 S.Ct. 2052, 80 L.Ed.2d 674 (1984).

7 . Id. at 697-98, 104 S.Ct. at 2069-70.

8 . Id. at 698, 104 S.Ct. at 2070.

9 . State v. Templin, 805 P.2d 182, 186 (Utah 1990) (citing Strickland, 466 U.S. at 698, 104 S.Ct. at 2070; Laws v. Armontrout, 863 F.2d 1377, 1381 (8th Cir.1988), cert. denied, 490 U.S. 1040, 109 S.Ct. 1944, 104 L.Ed.2d 415 (1989); State v. Crestani, 111 P.2d 1085, 1089 (Utah Ct.App.1989)).

10 . Id.

11 . 846 P.2d 1256 (Utah 1993).

12 . Id. at 1272.

13 . Fernandez does not assert that he was denied effective assistance of counsel under the Utah Constitution or that the Utah Constitution’s guarantees to counsel are different from those provided for in the United States Constitution. Our analysis is therefore limited to the federal Constitution. See State v. Tyler, 850 P.2d 1250, 1253 (Utah 1993); Templin, 805 P.2d at 185.

14 . U.S. Const, amend. VI.

15 . Templin, 805 P.2d at 186 (quoting McMann v. Richardson, 397 U.S. 759, 771 n. 14, 90 S.Ct. 1441, 1449 n. 14, 25 L.Ed.2d 763 (1970)).

16 . See supra 466 U.S. at 687 note 6, 104 S.Ct. at 2064 note 6.

17 . Strickland, 466 U.S. at 687, 104 S.Ct. at 2064.

18 . Id. at 688, 104 S.Ct. at 2064.

19 . Id. at 687, 104 S.Ct. at 2064.

20 . Id. at 694, 104 S.Ct. at 2068.

21 . Tyler, 850 P.2d at 1254.

22 . Specifically, findings of fact 1,’2, 6, 7, 8, 9, 10, and 12 draw legal conclusions based on the test elicited in Strickland.

23 . State v. Rio Vista Oil, Ltd., 786 P.2d 1343, 1347 (Utah 1990); Zions First Nat’l Bank, N.A. v. National Am. Title Ins. Co., 749 P.2d 651, 656 (Utah 1988).

24 . Zions First Nat'l Bank, 749 P.2d at 656 (citing Scharf v. BMG Corp., 700 P.2d 1068, 1070 (Utah 1985)).

25 . However, the district court concluded that Fernandez was not prejudiced by either counsel's failure to conduct formal discovery or the fact that the matter went to trial within twenty-three days of the arraignment.

26 .Although the district court concluded that Fernandez was not prejudiced by the fact that the matter went to trial within twenty-three days of the arraignment, see supra note 25, he reasserts on appeal that counsel allowed the matter to proceed to trial too rapidly and consequently could not properly prepare for trial.

27 . Although the district court concluded that Fernandez was not prejudiced by counsel's decision not to conduct formal discovery, see supra note 25, he reasserts on appeal that counsel’s failure to request names of potential witnesses to be called by the prosecution resulted in counsel’s being unprepared to conduct effective cross-examination.

28 . On appeal, Fernandez expands his argument by including claims that were neither raised at the habeas corpus hearing nor addressed by the district court in its findings of fact and conclusions of law. Because the claims were not asserted below, we do not address them. See State v. Anderson, 789 P.2d 27, 29 (Utah 1990) (citing State v. Van Matre, 777 P.2d 459, 462 (Utah 1989)).

29 . Tyler, 850 P.2d at 1256 (citing State v. Wood, 648 P.2d 71, 91 (Utah), cert. denied, 459 U.S. 988, 103 S.Ct. 341, 74 L.Ed.2d 383 (1982)).

30 . Strickland, 466 U.S. at 689, 104 S.Ct. at 2065.

31 . Tyler, 850 P.2d at 1256 (citing State v. Sammons, 156 Ariz. 51, 749 P.2d 1372, 1377 (1988)).

32 . State v. Speer, 750 P.2d 186, 191 (Utah 1988).

³³

34 . See, e.g., Strickland, 466 U.S. at 691, 104 S.Ct. at 2066.

35 . See State v. Malmrose, 649 P.2d 56, 58 (Utah 1982) (quoting State v. McNicol, 554 P.2d 203, 205 (1976)), where we stated, "[C]ourts will not second guess [the] 'legitimate exercise of judgment as to trial tactics or strategy.’ ”

36 . See Tyler, 850 P.2d at 1258.

37 . 466 U.S. at 697, 104 S.Ct. at 2069.

38 . State v. Bullock, 791 P.2d 155, 159 (Utah 1989), cert. denied, 497 U.S. 1024, 110 S.Ct. 3270, 111 L.Ed.2d 780 (1990); Codianna v. Morris, 660 P.2d 1101, 1109 (Utah 1983) (citing McNicol, 554 P.2d at 204).

39 . See Strickland, 466 U.S. at 693, 104 S.Ct. at 2067-68, where the Court stated, "[A]ctual ineffectiveness claims alleging a deficiency in attorney performance are subject to a general requirement that the defendant affirmatively prove prejudice.”

40 . Tyler, 850 P.2d at 1254; State v. Colonna, 766 P.2d 1062, 1066 (Utah 1988); Codianna, 660 P.2d at 1109; Malmrose, 649 P.2d at 58; State v. Gray, 601 P.2d 918, 920 (Utah 1979); McNicol, 554 P.2d at 204.

41 . Strickland, 466 U.S. at 692, 104 S.Ct. at 2067.

42 . Id.; see also Tyler, 850 P.2d at 1258 n. 43.

43 . See Strickland, 466 U.S. at 693, 104 S.Ct. at 2067-68.