THE PEOPLE, Plaintiff and Respondent,
v.
WILLIAM MICHAEL LEAHY, Defendant and Appellant.
Docket No. S035250.
Supreme Court of California.
October 27, 1994.*590 COUNSEL
Ronald Y. Butler, Public Defender, Carl C. Holmes, Chief Deputy Public Defender, Thomas Havlena and Alan J. Crivaro, Deputy Public Defenders, for Defendant and Appellant.
Linda F. Robertson and John T. Philipsborn as Amici Curiae on behalf of Defendant and Appellant.
Michael R. Capizzi, District Attorney, Maurice L. Evans, Chief Assistant District Attorney, Wallace J. Wade, Assistant District Attorney, Kathleen M. Harper and Donald Clarence, Deputy District Attorneys, for Plaintiff and Respondent. *591 Gil Garcetti, District Attorney (Los Angeles), George M. Palmer, Acting Chief Deputy District Attorney, Brentford J. Ferreira, Deputy District Attorney, John J. Meehan, District Attorney (Alameda), Rockne P. Harmon, Deputy District Attorney, Kent S. Scheidegger and Mark Daniel Ankcorn as Amici Curiae on behalf of Plaintiff and Respondent.
Daniel E. Lungren, Attorney General, George Williamson, Chief Assistant Attorney General, Gary W. Schons, Assistant Attorney General, Frederick R. Millar, Jr., Deputy Attorney General, Harvey M. Grossman, McCutchen, Doyle, Brown & Enersen, David M. Heilbron and Richard B. Ulmer, Jr., as Amici Curiae.
OPINION
LUCAS, C.J.
INTRODUCTION
The issues addressed by the Court of Appeal in this case were (1) whether the results of a horizontal gaze nystagmus (HGN) field sobriety test are admissible in the absence of a Kelly/Frye foundational showing, that is, foundational evidence disclosing general acceptance of the test within the relevant scientific community (see People v. Kelly (1976) 17 Cal. 3d 24 [130 Cal. Rptr. 144, 549 P.2d 1240] [hereafter Kelly]; Frye v. United States (D.C. Cir.1923) 293 F. 1013, 1014 [54 App.D.C. 46, 34 A.L.R. 145] (hereafter Frye)), and (2) whether a police officer without scientific expertise is qualified to give an opinion concerning the results of the HGN test. The Court of Appeal answered both questions in the negative and reversed defendant Leahy's convictions for driving under the influence of alcohol (Veh. Code, § 23152, subd. (a)) and driving with a blood-alcohol level in excess of .08 percent (id., subd. (b)).
We granted review and asked for supplemental briefing on the question whether the Kelly/Frye standard for admitting the results of new scientific techniques should be modified following the United States Supreme Court's decision in Daubert v. Merrell Dow Pharmaceuticals, Inc. (1993) 509 U.S. [125 L. Ed. 2d 469, 113 S. Ct. 2786] (hereafter Daubert)], holding that Frye was abrogated by rule 702 of the Federal Rules of Evidence (28 U.S.C.). As will appear, after reviewing such additional briefing and argument, we conclude that the Kelly/Frye formulation (or now more accurately, the Kelly formulation) should remain a prerequisite to the admission of expert testimony regarding new scientific methodology in this state. We further *592 conclude, consistent with the Court of Appeal's conclusions herein, that the HGN test is a "new scientific technique" within the scope of Kelly, and that the trial court improperly admitted police testimony regarding that technique without first requiring compliance with Kelly.
Accordingly, we affirm the judgment of the Court of Appeal reversing defendant Leahy's conviction on the ground of failure to comply with Kelly, and direct a remand to the trial court for a Kelly hearing in accordance with our opinion.
FACTS
The following facts were taken largely from the Court of Appeal opinion. Although that court also resolved a consolidated appeal in an unrelated case (People v. Tatar), we confine our discussion to the appeal of defendant William Michael Leahy.
Leahy was stopped after a police officer observed him driving a car traveling 55 miles per hour in a 25-mile-per-hour zone. Leahy's face was flushed, his eyes were red and watery, his speech was slurred, his balance was unsteady, and he exuded the odor of alcohol. Despite these relatively conventional indicia of intoxication, the officer was not convinced defendant was under the influence of alcohol. The officer decided to give defendant some field sobriety tests.
Defendant passed two such tests, the "internal clock" test and the "alphabet" test. He was also required, however, to take the HGN test. As a recent appellate decision explains, "Nystagmus is an involuntary rapid movement of the eyeball, which may be horizontal, vertical, or rotary. [Citation.] An inability of the eyes to maintain visual fixation as they are turned from side to side (in other words, jerking or bouncing) is known as horizontal gaze nystagmus, or HGN. [Citation.] Some investigators believe alcohol intoxication increases the frequency and amplitude of HGN and causes HGN to occur at a smaller angle of deviation from the forward direction. [Citation.]" (People v. Ojeda (1990) 225 Cal. App. 3d 404, 406 [275 Cal. Rptr. 472].)
In the present case, the officer believed that defendant failed the HGN test, and accordingly he arrested defendant. A subsequent intoxilyzer breath test revealed a .10 percent blood-alcohol level. Accordingly, defendant was charged with driving under the influence of alcohol (Veh. Code, § 23152, subd. (a)) and driving with a blood-alcohol level in excess of .08 percent (id., § 23152, subd. (b)). Defendant made a motion in limine to bar evidence of the HGN test result based on Kelly, supra, 17 Cal. 3d 24, and Frye, supra, 293 F. 1013. The trial court denied the motion.
*593 The trial court determined that "the Kelly-Frye rule is inapplicable because the nature of this test isn't a specific test for the determination of alcohol; it is only a symptom that the officer is testifying to ... in the same manner as he might be testifying to a symptom of slurred speech or a person's face turning red, or ... bloodshot eyes. It would be on that type of reasoning that it would be allowed in." Thus, the arresting officers were permitted to testify, over continuing objection, concerning the relationship between the HGN test and defendant's intoxication.
COURT OF APPEAL DECISION
The Court of Appeal focused primarily on the question whether police officers were competent to testify regarding HGN test results in light of the Kelly/Frye rule. The court assumed the continued vitality of the Kelly/Frye formulation, and concluded that the HGN test was indeed a scientific technique requiring expert foundational testimony. (Compare People v. Ojeda, supra, 225 Cal. App.3d at pp. 407-409 [allowing police officer to testify regarding HGN test results as indicating defendant's intoxication], with People v. Williams (1992) 3 Cal. App. 4th 1326, 1332-1335 [5 Cal. Rptr. 2d 130] [disallowing officer's opinion testimony, based on HGN test, that suspect had consumed alcohol], and People v. Loomis (1984) 156 Cal. App.3d Supp. 1, 5-7 [203 Cal. Rptr. 767] [disallowing police officer's opinion testimony, based on HGN test results, regarding suspect's blood-alcohol level].)
The Court of Appeal agreed with the analyses of the Williams and Loomis courts and concluded "[i]t was error to admit HGN evidence as either lay or expert testimony without a proper scientific foundation. The usual field sobriety tests are grounded in common knowledge, i.e., that intoxicated persons will often demonstrate lack of concentration, judgment, balance, and coordination. HGN is not. [¶] Consequently, at least as the law of California currently stands, it will be error in the event of any retrial to permit such evidence as the basis of an opinion concerning intoxication without a Kelly-Frye foundation, i.e., proof of general acceptance of HGN in the scientific community. [Footnote omitted.]"
DISCUSSION
We will first discuss the background leading to adoption of the Kelly/Frye formulation and then outline some of the considerations which militate for or against its retention. We next will address the Daubert decision and its effect on that formulation. As will appear, we conclude, consistent with the views of both parties herein, that Daubert affords no *594 compelling reason for abandoning Kelly in favor of the more "flexible" approach outlined in Daubert. Then we will turn to the question whether the trial court erred in permitting the officer to testify regarding the results of the HGN test given to defendant, and whether that error requires reversal of the judgment against defendant.
1. People v. Kelly Our unanimous 1976 Kelly decision (17 Cal. 3d 24) involved the admissibility of voiceprint evidence produced by a technique used to identify voices by spectrographic analysis. That case first set forth certain "general principles of admissibility" of expert testimony based on new scientific techniques, including the following "traditional" two-step process: "(1) [T]he reliability of the method must be established, usually by expert testimony, and (2) the witness furnishing such testimony must be properly qualified as an expert to give an opinion on the subject. (See Evid. Code, §§ 720, 801....) Additionally, the proponent of the evidence must demonstrate that correct scientific procedures were used in the particular case. [Citations.]" (Kelly, supra, 17 Cal.3d at p. 30, italics in original; see also People v. Diaz (1992) 3 Cal. 4th 495, 526 [11 Cal. Rptr. 2d 353, 834 P.2d 1171], and cases cited.)
Kelly next considered the appropriate test for determining the reliability of a new scientific technique. We recognized that one possible approach would be to leave questions of admissibility to the discretion of the trial court in the first instance, "in which event objections, if any, to the reliability of the evidence (or of the underlying scientific technique on which it is based) might lessen the weight of the evidence but would not necessarily prevent its admissibility." (17 Cal.3d at p. 31.)
We rejected the foregoing approach, however, and confirmed allegiance to the "germinal" Frye decision, supra, 293 F. 1013, and its formulation, as follows: "Just when a scientific principle or discovery crosses the line between the experimental and demonstrable stages is difficult to define. Somewhere in this twilight zone the evidential force of the principle must be recognized, and while courts will go a long way in admitting expert testimony deduced from a well-recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs." (Id. at p. 1014, italics added and quoted in Kelly, supra, 17 Cal.3d at p. 30.)
In Kelly, we explained that prior California cases, and "most" sister state cases, have followed Frye and "assigned the task of determining reliability of the evolving technique to members of the scientific community from which the new method emerges." (Kelly, supra, 17 Cal.3d at p. 31.) We *595 observed that the Frye standard holds several advantages, including (1) assuring that those persons most qualified to assess the validity of a scientific technique would have the determinative voice, (2) providing a "minimal reserve of experts" to critically examine each technique in a particular case, (3) promoting uniformity of decision based on finding a consensus in the scientific community, and (4) protecting the parties by its "essentially conservative nature." (Ibid.)
Expounding on the last factor, we noted that "[s]everal reasons... support a posture of judicial caution in this area. Lay jurors tend to give considerable weight to `scientific' evidence when presented by `experts' with impressive credentials. We have acknowledged the existence of a `... misleading aura of certainty which often envelops a new scientific process, obscuring its currently experimental nature.' [Citation.]" (17 Cal.3d at pp. 31-32.) We further noted that once a trial court has admitted evidence derived from a new technique and the decision is affirmed on appeal in a published opinion, it will become precedent controlling subsequent trials. (Id. at p. 32.) Thus, it is essential that the decision to admit such evidence is carefully considered.
For all the foregoing reasons, in Kelly we deemed the more cautious Frye formulation preferable to simply submitting the matter to the trial court's discretion for decision in each case. We now review that conclusion in light of a recent decision by the United States Supreme Court.
2. Daubert v. Merrell Dow Pharmaceuticals, Inc. In Daubert, supra, 509 U.S. [125 L. Ed. 2d 469], plaintiffs sued the manufacturer (Merrell Dow) of the anti-nausea drug Bendectin, alleging its ingestion caused birth defects. Merrell Dow moved for summary judgment, submitting an affidavit from an expert epidemiologist, Dr. Lamm, who stated that none of the 30 pertinent published studies had ever found Bendectin capable of causing malformations in fetuses. Rather than directly contest Dr. Lamm's factual statements, plaintiffs responded with their own experts' declarations to the effect that their unpublished studies, and reanalyses of published studies, indicated a link between Bendectin and fetal malformation.
The federal district court granted Merrell Dow's motion for summary judgment on the basis that the scientific theories of plaintiffs' experts failed to meet Frye's "general acceptance" standard. (See Daubert v. Merrell Dow Pharmaceuticals, Inc. (S.D.Cal. 1989) 727 F. Supp. 570, 572-576.) The United States Court of Appeals for the Ninth Circuit affirmed. (See Daubert v. Merrell Dow Pharmaceuticals, Inc. (9th Cir.1991) 951 F.2d 1128.) That court ruled that expert opinion based on methodology diverging "significantly from the procedures accepted by recognized authorities in the field ... *596 cannot be shown to be `generally accepted as a reliable technique,' ..." (Id. at p. 1130.) The Court of Appeals rejected plaintiffs' reanalyses as "unpublished, not subjected to the normal peer review process and generated solely for use in litigation." (Id. at p. 1131, fn. omitted.) The high court granted certiorari in light of "sharp divisions among the courts regarding the proper standard for the admission of expert testimony." (See Daubert, supra, 509 U.S. at p. [125 L.Ed.2d at pp. 477-478].)
The Supreme Court observed that although the Frye standard has been the "dominant" one in the 70 years since its formulation, it has come under "increasing attack of late." (509 U.S. at p. , and fn. 4 [125 L.Ed.2d at p. 478].) Indeed, according to the high court, "The merits of the Frye test have been much debated, and scholarship on its proper scope and application is legion." (Id. at p. , fn. omitted [125 L.Ed.2d at p. 478].) The court, however, found it unnecessary to resolve the debate. Instead, the court concluded that the Frye test had been superseded by adoption of the Federal Rules of Evidence in 1975. (See 28 U.S.C.A.)
Rule 402 of the Federal Rules of Evidence provides that "All relevant evidence is admissible, except as otherwise provided [by law]. Evidence which is not relevant is not admissible." Rule 401 defines "relevant evidence" as evidence having "any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence." As the high court notes, the rule's standard of relevance "thus is a liberal one." (509 U.S. at p. [125 L.Ed.2d at p. 479].)
Additionally, the Supreme Court in Daubert relied on rule 702 of the Federal Rules of Evidence, governing expert testimony: "If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise." As the court explained, "[n]othing in the text of this Rule [rule 702] establishes `general acceptance' as an absolute prerequisite to admissibility. Nor does [Merrell Dow] present any clear indication that Rule 702 or the Rules as a whole were intended to incorporate a `general acceptance' standard. The drafting history makes no mention of Frye, and a rigid `general acceptance' requirement would be at odds with the `liberal thrust' of the Federal Rules and their `general approach of relaxing the traditional barriers to "opinion" testimony.' [Citations.] ... Given the Rules' permissive backdrop and their inclusion of a specific rule on expert testimony that does not mention `general acceptance,' the assertion that the Rules somehow assimilated Frye is unconvincing." (509 U.S. at p. [125 L.Ed.2d at p. 480].)
*597 The high court in Daubert concluded that Frye's "austere standard" was "absent from and incompatible with the Federal Rules of Evidence, [and] should not be applied in federal trials." (509 U.S. at p. [125 L.Ed.2d at p. 480], fn. omitted.) In the following portion of its opinion, the court expounded on the screening functions federal courts should perform under rule 702 to assure that "any and all scientific testimony or evidence admitted is not only relevant, but reliable." (509 U.S. at p. ___ [125 L.Ed.2d at p. 480].); see also id., at pp. ___-___ [see also id., at pp. 480-485].) We now turn to an analysis of the California statutory provisions governing the admissibility of expert evidence and address the question whether Daubert's holding provides an appropriate analogy supporting the abrogation of the Kelly formulation in this state.
3. California Provisions on Admissibility Daubert, dealing with federal law, offers at most only persuasive authority to assist us in determining whether to reconsider our Kelly decision. We first review the California statutory provisions to determine whether the Kelly standard is "absent from and incompatible with" (see Daubert, supra, 509 U.S. at p. ___ [125 L.Ed.2d at p. 480]) the California Evidence Code provisions regarding the admissibility of expert testimony.
The California Evidence Code was enacted in 1965, and the pertinent provisions have not been amended since that time. Section 350 (further statutory references are to the Evidence Code unless otherwise indicated) provides that "[n]o evidence is admissible except relevant evidence," and section 351 provides that "[e]xcept as otherwise provided by statute, all relevant evidence is admissible."
Section 210 defines "relevant evidence" as "evidence, including evidence relevant to the credibility of a witness or hearsay declarant, having any tendency in reason to prove or disprove any disputed fact that is of consequence to the determination of the action." The foregoing provisions seem substantially equivalent to the "liberal" admissibility provisions of federal law (see Fed. Rules Evid., rules 401 and 402, 28 U.S.C.) cited in Daubert.
Sections 720 and 801 are the California provisions regarding admissibility of expert testimony. Section 720, subdivision (a), provides that "A person is qualified to testify as an expert if he has special knowledge, skill, experience, training, or education sufficient to qualify him as an expert on the subject to which his testimony relates...." Subdivision (b) of that section provides that "[a] witness' special knowledge ... may be shown by any otherwise admissible evidence, including his own testimony."
Section 801 permits an expert to state an opinion that is "(a) Related to a subject that is sufficiently beyond common experience that the opinion of an *598 expert would assist the trier of fact; and [¶] (b) Based on matter (including his special knowledge ...) perceived by or personally known to the witness ..., whether or not admissible, that is of a type that reasonably may be relied upon by an expert in forming an opinion upon the subject to which his testimony relates, unless an expert is precluded by law from using such matter as a basis for his opinion." (Italics added.)
(1) In 1982, as a result of the passage of an anticrime initiative measure (Prop. 8), the voters adopted a constitutional "Truth-in-Evidence" provision (Cal. Const. art. I, § 28, subd. (d)), stating that "relevant evidence shall not be excluded in any criminal proceeding...." In People v. Harris (1989) 47 Cal. 3d 1047, 1093-1095 [255 Cal. Rptr. 352, 767 P.2d 619], we upheld a trial court ruling excluding polygraph evidence proffered by the defense. In doing so, we rejected the argument that adoption of the foregoing constitutional provision eliminated restrictions on polygraph test results. Thus, we stated that "[a]doption of section 28(d) did not, however, abrogate generally accepted rules by which the reliability and thus the relevance of scientific evidence is determined. (See People v. Kelly[, supra,] 17 Cal. 3d 24....) The reliability of a scientific technique, therefore, is determined under the requirement of Evidence Code section 350, that `[n]o evidence is admissible except relevant evidence,' a provision necessarily incorporated by reference into Evidence Code section 352, which is expressly preserved by section 28(d). [¶].... [¶] ... The applicable rule in this state, one unaffected by ... the adoption of section 28(d), is that the admissibility of evidence based on a new scientific technique is determined under the principles of Frye v. United States[, supra,] 293 Fed. 1013, 1014...." (People v. Harris, supra, 47 Cal.3d at p. 1094.)
4. Survival of Kelly Sections 720 and 801, in combination, seem the functional equivalent of Federal Rules of Evidence, rule 702, as discussed in Daubert. Although some language in section 801 is broad enough to include a Frye standard of general acceptance (matter "of a type that reasonably may be relied upon by an expert"), nothing in these sections expressly establishes general acceptance as an absolute prerequisite to admissibility, and nothing in the legislative history leading to adoption of the Evidence Code indicates that a general acceptance standard was intended. To paraphrase Daubert, the drafting history makes no mention of Frye, and a rigid general acceptance requirement might seem at odds with the liberal thrust of the Evidence Code and its general approach of relaxing the traditional barriers to "opinion" testimony.
(2) Thus, were we approaching the question afresh, without the benefit of our analysis and holding in Kelly, supra, 17 Cal. 3d 24, we might reasonably conclude, by analogy to Daubert, that the framers of the Evidence Code *599 did not intend to adopt a general acceptance standard. But Kelly was decided in 1976, 11 years after the adoption of the Evidence Code and its provisions governing the admissibility of expert testimony. (See Fishback v. People (Colo. 1993) 851 P.2d 884, 890 [retaining Frye on similar analysis].) We were presumably well aware in Kelly that the Frye "general acceptance" standard arguably was "absent from and incompatible with" the preexisting California statutory provisions. Nonetheless, we concluded otherwise and found Frye compatible with those provisions, which we cited in our opinion. (See Kelly, supra, 17 Cal.3d at p. 30.) No significant relevant developments have occurred in this state since Kelly was decided to justify abandoning its conclusions. Thus, the principle of stare decisis appears applicable here, unless the criticism of the Frye doctrine cited in Daubert and in other sources persuades us otherwise.
Both defendant and the People herein urge us to retain our Kelly holding. Defendant encourages us to preserve the "cautious" and "conservative" approach we deliberately chose in Kelly. The People (represented by the Orange County District Attorney) agree, stating, "there must be some standard test for the trial courts to follow to minimize the influx of what has come to be known as `junk science' into the courtrooms of this state. The Kelly/Frye test, while not perfect, has acted to keep such unreliable evidence from the jury." Amici curiae Los Angeles District Attorney (LADA), Criminal Justice Legal Foundation (CJLF), California Public Defenders' Association, and Product Liability Advisory Counsel have also submitted briefs supporting the retention of Kelly.
LADA takes the position that for us to adopt Daubert would "ignore over forty years of precedent based upon policy considerations which have not changed, and without any compulsion to do so from the California Legislature or from the voters." In the view of CJLF, "California should retain the Kelly rule as an important safeguard against untested and unproven scientific theories." The other two above mentioned amici curiae are in general agreement.
The foregoing position, however, is not unanimous. Other amici curiae (namely, the State Attorney General and the Appellate Committee of the California District Attorneys Association (ACCDA)) urge us to employ the Daubert analogy and either discard, or substantially alter, our Kelly holding. It seems apparent from their briefs, however, that the primary concern of these amici curiae involves the interpretation and application of Kelly in the context of the deoxyribonucleic acid (DNA) cases (see, e.g., People v. Barney (1992) 8 Cal. App. 4th 798 [10 Cal. Rptr. 2d 731]). These amici seem more concerned with the narrow issue of the exclusion of DNA test evidence, an issue not involved here, than with the general question of the retention of Kelly.
*600 Amicus curiae ACCDA asserts that Kelly is unclear as to the proper standard for proving general acceptance by the scientific community. (Compare People v. Shirley (1982) 31 Cal. 3d 18, 56 [181 Cal. Rptr. 243, 723 P.2d 1354] [opposition by scientists "significant either in number or expertise"] with People v. Guerra (1984) 37 Cal. 3d 385, 418 [208 Cal. Rptr. 162, 690 P.2d 635] [unanimity of opinion unnecessary if use of technique supported by "clear majority" of the scientific community].) Because this issue may arise on remand, we address it briefly in this opinion (see post, at pp. 611-612).
Additionally, ACCDA suggests it is unclear how to apply Kelly's requirement (see 17 Cal.3d at p. 30) that correct scientific procedures be employed in the case. (Compare People v. Farmer (1989) 47 Cal. 3d 888, 913 [254 Cal. Rptr. 508, 765 P.2d 940] [careless testing affects weight of evidence not admissibility], with People v. Barney, supra, 8 Cal. App.4th at pp. 823-824 [Kelly's "`correct scientific procedures'" requirement "is not merely a question of weight but is an element of the Kelly-Frye admissibility determination"].) ACCDA urges us to modify our Kelly rule to clarify the foregoing uncertainties.
Amicus curiae Attorney General echoes the concerns of ACCDA, and asserts that Kelly/Frye (1) emphasizes "nose counting" of the scientific community rather than focusing directly on the reliability of the challenged technique, thereby excluding "demonstrably reliable, highly probative evidence, as it has in the case of DNA evidence," (2) is premised on an unrealistic model of a "relatively monolithic scientific community," (3) improperly defers to scientists the legal question of the admissibility of evidence; and (4) unduly penalizes crime victims, their families and friends by excluding relevant evidence of guilt.
The Attorney General also observes that the Kelly/Frye test is uncertain in various respects, such as (1) whether, and to what extent, it applies to expert testimony (see People v. McDonald (1984) 37 Cal. 3d 351, 372-373 [208 Cal. Rptr. 236, 690 P.2d 709, 46 A.L.R. 4th 1011]), (2) whether there must be general acceptance of the scientific technique itself or merely the scientific principle underlying it, (3) whether "probability estimates" (e.g., in DNA cases) are subject to Kelly/Frye, (4) what is the "relevant" scientific community, and (5) what constitutes "general acceptance" (e.g., a simple majority, a consensus, or a significant minority).
The Attorney General further suggests that the supposed benefits of Kelly/Frye (e.g., minimizing undue influence on juries, leaving scientific questions to qualified scientists, avoiding multiple suits, and promoting *601 uniformity of decision) can be alternatively achieved by giving the trial courts broad discretion to decide admissibility questions, subject to vigorous cross-examination of experts, full presentation of contrary evidence, and careful jury instructions on the burden of proof. According to the Attorney General, the DNA cases demonstrate that Kelly/Frye does not guarantee against multiple litigation or conflicting decisions.
As with amicus curiae ACCDA, the Attorney General urges us to clarify or "fix" our Kelly rule in the following respects, should we elect not to abandon it entirely:
First, we should make clear that "general acceptance" does not require unanimity, a consensus of opinion, or even majority support by the scientific community. The Attorney General argues that general acceptance should mean that the technique "is accepted by other well credentialed scientists outside the testing laboratory. ..." (Italics in original.)
Second, according to the Attorney General, we should clarify that Kelly does not apply to "probability estimates" such as those given in DNA cases, and we should hold that Kelly is inapplicable to the "particular standardized protocol" used to perform the test in the case at hand, rather than to the "fundamental validity" of the technique itself.
Finally, the Attorney General suggests we should adopt an abuse of discretion standard of appellate review, rather than permit a de novo review at the appellate level. (See People v. Barney, supra, 8 Cal. App.3d at pp. 810-811 [general acceptance of DNA testing subject to "limited de novo review" on appeal].)
In sum, rather than raise fundamental defects in our approach in Kelly, these amici curiae appear more concerned with clarifying or modifying our Kelly standard for use in future cases. But most of the suggested modifications or clarifications do not relate to issues presented in this case, and do not directly affect our decision whether or not to retain Kelly. Accordingly, we decline to address them here except insofar as they overlap with general criticisms of the Kelly/Frye rule as discussed below. Amici curiae's various suggestions for improving or "fixing" the Kelly rule may well have merit, but the present case is not a good vehicle for addressing them. Likewise, we will not permit amici curiae to convert this relatively narrow HGN case into a vehicle for reviewing the DNA cases. We turn, then, to the commentators critical of the Frye formulation itself.
The critics of Frye focus primarily on its conservative nature: As previously noted, the doctrine contemplates an undefined period of testing and *602 study by a community of experts before a new scientific technique may be deemed "generally accepted," thus delaying the admissibility of evidence derived from the technique. (See, e.g., Black, A Unified Theory of Scientific Evidence (1988) 56 Fordham L.Rev. 595, 636-637 [Black]; Giannelli, The Admissibility of Novel Scientific Evidence: Frye v. United States, a Half-Century Later (1980) 80 Colum. L.Rev. 1197, 1223-1224, 1248 [Giannelli]; Hanson, James Alphonzo Frye is Sixty-Five Years Old; Should He Retire (1989) 16 Western St. U.L.Rev. 357, 367-368 (Hanson); Note, Leading Cases, 101 Harv.L.Rev. 119, 125-127 (Harvard Note); Daubert, supra, 509 U.S. at p. ___, fns. 4, 5 [125 L.Ed.2d at pp. 478-479], and articles cited.) These commentators observe that conceivably, a reliable, readily provable, technique could nonetheless remain unknown and untested by the relevant scientific community, thereby indefinitely delaying its use in the courtroom. (Hanson, supra, pp. 367-368.) One commentator observes that to the extent Frye excludes favorable defense evidence, it may be constitutionally deficient. (Harvard Note, supra, pp. 125-127.)
But we acknowledged the foregoing basis for criticism in Kelly, noting that "[s]ome criticism has been directed at the Frye standard, primarily on the ground that the test is too conservative, often resulting in the prevention of the admission of relevant evidence [citations]." (17 Cal.3d at pp. 30-31.) We nonetheless concluded that "there is ample justification for the exercise of considerable judicial caution in the acceptance of evidence developed by new scientific techniques." (Id. at p. 31.) As we have previously indicated, Kelly set forth at length its reasons for adopting a more cautious approach to the admission of new scientific evidence. Nothing occurring in the years since Kelly was decided requires us to reconsider that conclusion.
Indeed, as late as 1989 we acknowledged that "[t]he courts are willing to forego admission of such techniques completely until reasonably certain that the pertinent scientific community no longer views them as experimental or of dubious validity. This all-or-nothing approach was adopted [in Kelly] in full recognition that there would be a `"considerable lag"' between scientific advances and their admission as evidence in a court proceeding. [Citation.]" (People v. Stoll (1989) 49 Cal. 3d 1136, 1156 [265 Cal. Rptr. 111, 783 P.2d 698].)
As previously noted, amicus curiae Attorney General criticizes Kelly on the additional ground that it requires the trial judge to defer admissibility questions to a "nose count" of scientists rather than allowing the judge or jury to directly confront the issue of the reliability of the evidence. But, as we observed in Kelly, it may be preferable to let admissibility questions regarding new scientific techniques be settled by those persons most qualified to assess their validity. (See 17 Cal.3d at p. 31.) Some commentators *603 have questioned the ability of trial judges, often wholly unschooled in scientific areas, to evaluate highly technical scientific material. As the Harvard Note, supra, observes, "The Frye rule ensures that judges and juries with little or no scientific background will not attempt to resolve technical questions on which not even experts can reach a consensus...." (101 Harv. L.Rev. at p. 127, fn. omitted.)
Moreover, the criticism of Frye has been balanced by favorable commentary and widespread support. As we previously observed, both the defendant and the People herein (through the Orange County District Attorney) endorse Kelly as an essential check on the admissibility of new scientific techniques. Moreover, most critics of Frye acknowledge the test has several positive aspects and operates well in many situations. (See Hanson, supra, 16 Western St. U.L.Rev. at pp. 361-363, 458-460; Harvard Note, supra, 101 Harv.L.Rev. at p. 127; Black, supra, 56 Fordham L.Rev. at pp. 637-638.)
For example, the Hanson article lists the following positive features of the Frye doctrine: "(a) The time and cost of proving and establishing new scientific principles are transformed from a courtroom ... to a laboratory.... [¶] (b) Each court thus need not confront the principle de novo, and may draw from other court decisions, the applicable literature, ... and if offered, additional expert testimony.... [¶] (c) Presumably a battery of well-qualified scientific and medical personnel are available to support the principle...." (Hanson, supra, 16 Western St. U.L.Rev. at pp. 458-459.)
The Harvard Note is likewise predominantly favorable toward Frye. The author states that, "[d]espite the[] criticisms, however, the Frye rule plays a vital role in the trial process.... [¶] Although the Frye test may be difficult to apply and at times may exclude relevant evidence, it has proven its value for over sixty years. It has prevented justice from becoming a matter of amateur guesswork based on unreliable techniques and has helped to assure that determinations of guilt or innocence are not influenced by the vagaries of pseudoscience. [Frye is]... a wise rule that contributes greatly to the integrity of the criminal process." (Harvard Note, supra, 101 Harv. L.Rev. at p. 127.)
Even Professor Giannelli, Frye's severest critic, acknowledges that Frye's requirement of a "special burden" as a prerequisite to admissibility of a new scientific technique has merit, and that if a "relevancy" test is chosen as an alternative (as in Daubert), special steps should be taken to assure the reliability of the evidence, including requiring the prosecutor to prove the validity of the technique beyond a reasonable doubt. (Giannelli, supra, 80 Colum. L.Rev. at pp. 1248-1250.)
*604 Despite the criticism of Kelly/Frye, this court has had numerous occasions to review and apply the doctrine, and has done so without apparent difficulty or critical comment. (See, e.g., People v. Diaz, supra, 3 Cal.4th at pp. 525-528 [tissue sample analysis]; People v. Pride (1992) 3 Cal. 4th 195, 238-239 [10 Cal. Rptr. 2d 636, 833 P.2d 643] [hair comparison analysis]; People v. Stoll, supra, 49 Cal.3d at pp. 1152-1161 [standardized personality tests]; People v. Bledsoe (1984) 36 Cal. 3d 236, 245-251 [203 Cal. Rptr. 450, 681 P.2d 291] [rape trauma syndrome]; People v. Shirley, supra, 31 Cal. 3d 18, 51-54 [hypnotically induced testimony].)
Additionally, we observe that the Legislature has had ample opportunity to amend the Evidence Code provisions to abrogate or modify the general acceptance standard that Kelly found implicit within them. The Legislature has made frequent amendments to the expert testimony provisions of that code since its adoption in 1965. (See, e.g., §§ 730-731 [appointment of experts], 795 [hypnosis testimony], 811-814 [market value evidence], former 892 [blood test evidence; later repealed].) Legislative failure to amend sections 720 or 801, although not conclusive, may be presumed to signify legislative acquiescence in our Kelly decision. (See People v. Escobar (1992) 3 Cal. 4th 740, 750-751 [12 Cal. Rptr. 2d 586, 837 P.2d 1100]; Bailey v. Superior Court (1977) 19 Cal. 3d 970, 977 and fn. 10 [140 Cal. Rptr. 669, 568 P.2d 394]; Estate of McDill (1975) 14 Cal. 3d 831, 837-839 [122 Cal. Rptr. 754, 537 P.2d 874].)
In sum, Kelly sets forth the various reasons why the more "conservative" Frye approach to determining the reliability of expert testimony regarding scientific techniques represents an appropriate one. Daubert, which avoided the issue of Frye's "merits," presents no justification for reconsidering that aspect of our holding in Kelly. Thus, we conclude that the Kelly formulation survived Daubert in this state, and that none of the above described authorities critical of that formulation persuades us to reconsider or modify it at this time. Accordingly, we turn next to the question of the application of that standard to the facts of the present case.
5. HGN Test Is a New Scientific Test Under Kelly (3) After reviewing the briefs and applicable cases, we conclude the Court of Appeal correctly decided that the HGN test was a "new scientific technique" within the scope of the Kelly formulation, and for that reason proof was required of its general acceptance by the scientific community.
a. California HGN Decisions The California decisions involving the HGN test do not address whether that test involves a "new scientific technique" within the scope of Kelly. (See People v. Williams, supra, 3 *605 Cal. App.4th 1326; People v. Ojeda, supra, 225 Cal. App. 3d 404; People v. Loomis, supra, 156 Cal. App.3d Supp. 1.) Instead, these cases consider the general admissibility of police testimony regarding the results of HGN tests in particular cases, a matter we discuss in part 6 hereof, post.
Other decisions make clear that Kelly is applicable only to "new scientific techniques." (See People v. Webb (1993) 6 Cal. 4th 494, 524 [24 Cal. Rptr. 2d 779, 862 P.2d 779]; People v. Stoll, supra, 49 Cal. 3d 1136, 1155-1156; Kelly, supra, 17 Cal.3d at p. 30.) The People argue that HGN testing is neither "new" nor "scientific." We disagree.
b. HGN Testing Is a "New" Technique As we stated in Stoll, supra, 49 Cal.3d at page 1156, "Kelly/Frye only applies to that limited class of expert testimony which is based, in whole or part, on a technique, process, or theory which is new to science and, even more so, the law. "(See also People v. Clark (1993) 5 Cal. 4th 950, 1018 [22 Cal. Rptr. 2d 689, 857 P.2d 1099] [allowing blood spatter evidence based on technique predating Kelly decision and derived from matters of common knowledge].)
In Stoll, supra, we observed that the standard psychological testing at issue in that case was "not new to psychology or the law.... [¶] California courts have routinely admitted defense expert opinion analogous to the one offered here, with no suggestion that Kelly/Frye applies.... [¶] ... Moreover, ... diagnostic use of written personality inventories ... has been established for decades. Modern courts have not resisted reference to these tests." (49 Cal.3d at pp. 1157-1158.)
The People observe that HGN testing has been used by law enforcement agencies for more than 30 years. A 1988 annotation confirms that "[t]he horizontal gaze nystagmus (HGN) test has been in use for 30 years, but it has not been widely applied in the United States until recently." (Annot., Horizontal gaze nystagmus test: use in impaired driving prosecution (1988) 60 A.L.R. 4th 1129, 1131, italics added and fn. omitted.) This annotation observes that the HGN symptom was first recognized in the 1960's in connection with barbiturate use, but the earliest court decision cited by the annotation as addressing the admissibility of HGN test results was decided in California in 1984 (People v. Loomis, supra, 156 Cal. App.3d Supp. 1), and that case disallowed the evidence.
In determining whether a scientific technique is "new" for Kelly purposes, long-standing use by police officers seems less significant a factor than repeated use, study, testing and confirmation by scientists or trained technicians. Unlike the psychiatric tests involved in Stoll, supra, 49 Cal.3d at pages *606 1157 to 1158, HGN testing has been repeatedly challenged in court, with varying degrees of success, in this and other states, and accordingly its courtroom use cannot fairly be characterized as "routine" or settled in law. (See Annot., supra, 60 A.L.R. 4th 29, and cases cited; People v. Williams, supra, 3 Cal.4th at p. 1336, and cases cited.) Our survey of decisions from states that recognize the Frye standard indicates that, despite wide controversy regarding the admissibility of HGN evidence for purposes of establishing probable intoxication, most courts agree that HGN testing is subject to that standard.
Given the recent history of legal challenges to the admissibility of HGN test evidence in this and other states, it seems appropriate that we deem the technique "new" or "novel" for purposes of Kelly. To hold that a scientific technique could become immune from Kelly scrutiny merely by reason of long-standing and persistent use by law enforcement outside the laboratory or the courtroom, seems unjustified.
c. HGN Testing Is a "Scientific" Technique In Stoll, we also observed that, by reason of the potential breadth of the term "scientific" in the Kelly/Frye doctrine, the courts often refer "to its narrow `common sense' purpose, i.e., to protect the jury from techniques which, though `new,' novel, or `"experimental,"' convey a `"misleading aura of certainty."' [Citations.]" (49 Cal.3d at pp. 1155-1156.) According to Stoll, a technique may be deemed "scientific" for purposes of Kelly/Frye if "the unproven technique or procedure appears in both name and description to provide some definitive truth which the expert need only accurately recognize and relay to the jury." (Id. at p. 1156, italics added.)
Defendant herein contends that the HGN test, both in "name" ("horizontal gaze nystagmus") and "description" (observation of involuntary jerking of eyeball induced by ingestion of alcohol) fits Stoll's definition of a scientific technique that provides the officer, and thus the jury, with supposedly "definitive" proof of intoxication. As noted in People v. Ojeda, supra, 225 Cal. App.3d at page 408, quoting from another case, "`[T]he principal obstacle to the admissibility of the horizontal gaze nystagmus test may be its pretentiously scientific name.'" A jury might be unduly swayed by HGN evidence solely by reason of its technical nomenclature.
Additionally, the HGN test, as described by the arresting officer in the present case, appears to provide to the jury a "definitive truth" wholly apart from its imposing name. The arresting officer in this case was asked what he "generally" did to determine whether or not a suspect was intoxicated. He responded, "Well, I I check their nystagmus, that's a big indicator for me." *607 He continued by noting that the HGN test usually was one of the first sobriety tests he performed.
The arresting officer in the consolidated appeal of Michael Tatar, the transcript of which we take judicial notice, was even more authoritative regarding HGN testing. According to the officer, he has "always been right" whenever he checked his HGN test results with subsequent blood-alcohol test results.
Thus, the "aura of certainty" emanating from the officers' description of HGN tests was unmistakable. A jury could be unduly and unjustifiably impressed by the confidence the testifying officers showed regarding the HGN procedure.
Defendant cites many cases from other states holding that HGN testing is a "scientific" technique for purposes of Frye analysis, and that, accordingly, expert testimony from persons other than the officers who administer the HGN tests is required to sustain its admissibility. (See State v. Barker (1988) 179 W. Va. [366 S.E.2d 642, 645]; State v. Superior Court (Blake) (1986) 149 Ariz. 269 [718 P.2d 171, 179-181] [hereafter Blake]; State v. Borchardt (1986) 224 Neb. 47 [395 N.W.2d 551, 559]; State v. Cissne (1994) 72 Wash. App. 677 [865 P.2d 564, 568] [hereafter Cissne]; State v. Witte (1992) 251 Kan. 313 [836 P.2d 1110, 1115-1116] [hereafter Witte]; Malone v. City of Silverhill (Ala. Crim. App. 1989) 575 So. 2d 101, 105; People v. Vega (1986) 145 Ill. App. 3d 996 [99 Ill. Dec. 808, 496 N.E.2d 501]; State v. Reed (1987) 83 Ore.App. 451 [732 P.2d 66, 68]; Com. v. Apollo (1992) 412 Pa.Super. 453 [603 A.2d 1023, 1026]; Com. v. Miller (1987) 367 Pa.Super. 359 [532 A.2d 1186, 1189].)
For example, in Witte, supra, the Kansas Supreme Court cited 11 decisions from other states holding that HGN testing is a "scientific" test requiring Frye compliance. (836 P.2d at p. 1115.) As stated in Witte, "The HGN test is distinguished from other field sobriety tests in that science, rather than common knowledge, provides the legitimacy for HGN testing. [Citations.] Certain reactions to alcohol are so common that judicial notice will be taken of them; however, HGN testing does not fall into this category. [Citation.] HGN test results are `scientific evidence based on the scientific principle that consumption of alcohol causes the type of nystagmus measured by the HGN test.' [Citation.] HGN evidence could have a disproportionate impact on the jury's decisionmaking process because of the test's scientific nature and because the jury may not understand the nature of the test or the methodology of its procedure. [Citations.]." (Ibid.)
Thus, we conclude that HGN tests involve a "new scientific technique" that is required to meet Kelly's general acceptance test.
*608 6. Police Officer Testimony Is Insufficient to Establish General Acceptance (4) The People concede that "most states have determined the [HGN] test is a scientific test" under Frye. They nonetheless insist that testimony by the officers administering the HGN test is sufficient to establish Kelly compliance. They rely on our statement in People v. Stoll, supra, 49 Cal.3d at page 1157, that "absent some special feature which effectively blindsides the jury, expert opinion testimony is not subject to Kelly/Frye." The People observe that the officers who administered the tests are subject to cross-examination, and that the jury is not likely to accord their opinions regarding HGN test results the same weight as a scientist's opinion on the same subject. According to the People, field sobriety tests such as the HGN test merely "rely on the experience of a police officer in observing the performance of persons who may or may not be intoxicated.... There is nothing more scientific about the HGN test than the other field sobriety tests."
The People also rely on language in People v. Ojeda, supra, 225 Cal. App.3d at page 408, that "police officers who deal daily with intoxicated persons become expert at recognizing the physical effects of intoxication.... This does not make them scientists." The People cite cases from other states holding either that the HGN test is not a "scientific" test or that the testimony of police officers regarding administration of the HGN test is sufficient compliance with the Frye standard. (See City of Fargo v. McLaughlin (N.D. 1994) 512 N.W.2d 700, 706-707; State v. Garrett (1991) 119 Idaho 878 [811 P.2d 488, 491-493]; State v. Sullivan (S.C. 1993) 426 S.E.2d 766; State v. Bresson (1990) 51 Ohio St. 3d 123 [554 N.E.2d 1330, 1334]; State v. Armstrong (La. Ct. App. 1990) 561 So. 2d 883, 885-887; see also State v. Clark (1988) 234 Mont. 222 [762 P.2d 853, 856-857] [allowing HGN evidence by police officer; Frye test not adopted in state]; State v. Murphy (Iowa 1990) 451 N.W.2d 154, 156-157] [same]; Whitson v. State (1993) 314 Ark. 458 [863 S.W.2d 794, 798] [same]; Emerson v. State (Tex. Crim. App. 1994) 880 S.W.2d 759, 768-769 [same]; State v. O'Key (1993) 123 Ore.App. 54 [858 P.2d 904, 907-908] [same].)
The foregoing decisions, however, do not explain how police officers are competent to establish general acceptance of HGN testing in the scientific community, or how they are qualified to relate the scientific bases underlying the nystagmus test. As stated in People v. Williams, supra, 3 Cal. App.4th at pages 1333 to 1334, "We will assume that [officer] Vega's training and experience qualified him as an expert to administer the nystagmus test and observe signs of nystagmus. Being qualified to attribute the observed eye movements to a particular cause, however, is a far different matter.... [¶] Vega's opinion that appellant was under the influence of alcohol, to the *609 extent it was based on the nystagmus test, rests on scientific premises well beyond his knowledge, training, or education. Without some understanding of the processes by which alcohol ingestion produces nystagmus, how strong the correlation is, how other possible causes might be masked, what margin of error has been shown in statistical surveys, and a host of other relevant factors, Vega's opinion on causation, notwithstanding his ability to recognize the symptom, was unfounded. It should have been excluded."
Consistent with both the weight of authority and the cautious, "conservative" nature of Kelly, we conclude that testimony by police officers regarding the mere administration of the test is insufficient to meet the general acceptance standard required by Kelly.
The People urge, however, that we take judicial notice of the various decisions (e.g., Blake, supra, 718 P.2d at pp. 179-181; People v. Buening (1992) 229 Ill. App. 3d 538 [170 Ill. Dec. 542, 592 N.E.2d 1222, 1227]; State v. Garrett, supra, 811 P.2d at pp. 490-493; State v. Hill (Mo. Ct. App. 1993) 865 S.W.2d 702, 704-705) and published studies concluding that HGN testing meets the Frye standard. But the conclusions of those decisions and studies are by no means unchallenged, for there appears to exist substantial opposing authority. (See Witte, supra, 836 P.2d at pp. 1119-1121, citing numerous articles and studies; Cissne, supra, 865 P.2d at p. 568, and fn. 5.)
Witte, decided in 1992, suggests that if the Arizona Supreme Court in Blake had been aware of the contrary authority and evidence, it might not have held that HGN testing satisfied Frye. (836 P.2d at p. 1121.) The 1994 Cissne decision likewise observed that "[a]s Witte noted, research and articles critical of HGN testing appeared after some jurisdictions concluded that HGN testing satisfied Frye standard[s]." (865 P.2d at p. 568.) Cissne concluded, "[w]e decline the State's invitation to follow those few jurisdictions that have concluded that HGN testing meets the Frye standard. The trial court must evaluate, weigh and consider the conflicting evidence before determining whether the test is novel, and, if it is novel, whether it is reliable...." (Id. at p. 569.)
Additionally, we note that several decisions from other states have refused to resolve the Frye issue on appeal by reference to scientific studies and articles not introduced at the defendant's trial. (See Com. v. Miller, supra, 532 A.2d at p. 1190; People v. Vega, supra, 496 N.E.2d at pp. 504-505; State v. Barker, supra, 366 S.E.2d at p. 646; State v. Reed, supra, 732 P.2d at p. 69.) Although theoretically we could attempt to weigh and evaluate the merits of the conflicting authority, it seems more appropriate to remand this issue to a trial court for an evidentiary hearing, as several other decisions *610 have suggested. (E.g., State v. Barker, supra, 366 S.E.2d at p. 646; Cissne, supra, 865 P.2d at p. 569; Witte, supra, 836 P.2d at p. 1121; see also People v. Brown (1985) 40 Cal. 3d 512, 533-535 [220 Cal. Rptr. 637, 709 P.2d 440] [explaining our reluctance to determine on appeal general acceptance of electrophoresis technique on basis of available scientific literature, without "an adequate future trial record made with the help of live witnesses qualified in the applicable scientific disciplines"].)
As stated in Witte, "The reliability of the HGN test is not currently a settled proposition in the scientific community.... Before this court rules on whether HGN evidence satisfies the Frye admissibility requirements, a trial court first should have an opportunity to examine, weigh, and decide disputed facts to determine whether the test is sufficiently reliable to be admissible for any purpose in Kansas." (836 P.2d at p. 1121.) We agree with that analysis and, accordingly, we deny the People's application for judicial notice.
7. Limited Remand Appropriate (5) In accord with our foregoing discussion, we conclude that the Court of Appeal was correct in reversing both counts of defendant's conviction, for no evidence was elicited to show that HGN testing had been generally accepted in the scientific community, and the error was clearly prejudicial. As the Court of Appeal concluded, "it is reasonably probable a different result would have been achieved on both counts but for the HGN evidence [citation]. Leahy passed the other field sobriety tests. His blood-alcohol level was said to be .10, only .02 over the statutory maximum; and criminalists generally concede a margin of error of at least plus or minus.01. Also, the passage of time between arrest and test can account for even more than that, up or down. About 90 minutes went by between arrest and Leahy's encounter with the intoxilyzer. Depending on when he had his last drink and a host of other factors, his blood-alcohol level could have risen or fallen several points in that time."
We accept, however, the People's suggestion that an entire retrial of the case may be unnecessary. Instead, we will direct the Court of Appeal to reverse defendant's conviction and remand the case to the trial court for a Kelly hearing in accordance with our opinion. If, at the conclusion of the hearing, the trial court concludes there is sufficient basis to admit the HGN testimony previously presented, the court should reinstate the judgment without reintroducing such testimony. If the trial court determines the HGN evidence is inadmissible under Kelly, the court should order a new trial if the People so elect. If the judgment of conviction is reinstated, or a new trial ordered, appellate review will be available to the respective parties regarding the trial court's ruling, limited to any new issues not resolved in this opinion. *611 (See People v. Collins (1986) 42 Cal. 3d 378, 394-395 [228 Cal. Rptr. 899, 722 P.2d 173] [remand procedure for determining prejudice from errors arising under People v. Castro (1985) 38 Cal. 3d 301 [211 Cal. Rptr. 719, 696 P.2d 111]; judgment on remand subject to appeal by both parties]; cf. People v. Pizarro (1992) 10 Cal. App. 4th 57, 95-96 [12 Cal. Rptr. 2d 436] [Kelly/Frye remand].)
Defendant objects to this limited remand procedure, contending that even if Kelly were deemed satisfied, the question would remain whether police officers were qualified to testify regarding the HGN results. (6) We reject this contention. Once it has been shown that HGN testing is generally accepted in the scientific community, no reason exists why police officers should be deemed unqualified to administer and report the results of those tests. Thus, in future cases, once the Kelly standard has been met, as reflected by a published appellate precedent, the prosecution will not be required to submit expert testimony to confirm a police officer's evaluation of an HGN test. Of course, nothing would prevent the defendant from challenging that evaluation with expert testimony of his own.
8. Establishing General Acceptance (7) Because the issue may arise on remand, we briefly comment on one aspect of the Kelly formulation that some amici curiae have found troublesome. In Kelly, quoting with approval from Frye, we indicated that a new scientific technique must be "`sufficiently established to have gained general acceptance in the particular field to which it belongs.'" (17 Cal.3d at p. 30, italics removed.) We further suggested it was "questionable whether the testimony of a single witness alone is ever sufficient to represent, or attest to, the views of an entire scientific community," and we observed that "[i]deally, resolution of the general acceptance issue would require consideration of the views of a typical cross-section of the scientific community, including representatives, if there are such, of those who oppose or question the new technique." (Id. at p. 37, italics added.)
Thereafter, in People v. Shirley, supra, 31 Cal.3d at pages 55 and 56, we indicated that "considerations of judicial economy" would permit a court to scrutinize "published writings in scholarly treatises and journals" in lieu of live testimony. We added that the burden of showing general acceptance lies with the proponent of the evidence to show a "scientific consensus" (italics added), and that "if a fair overview of the literature discloses that scientists significant either in number or expertise publicly oppose [the technique] as unreliable, the court may safely conclude there is no such consensus at the present time." (Id. at p. 56.) We concluded in Shirley that "it ... appears *612 that major voices in the scientific community oppose the use of hypnosis to restore the memory of potential witnesses...." (Ibid., italics added.)
We do not read Shirley as modifying or abandoning Kelly's insistence on ascertaining, if possible, whether the technique has become generally accepted by a "typical cross-section" of the relevant scientific community. Shirley's requirement of a "scientific consensus" by "major voices in the scientific community" seems entirely consistent with Kelly in this regard.
A few years later, in People v. Guerra, supra, 37 Cal. 3d 385, we expounded briefly regarding Kelly's general acceptance test, noting that the Frye test does not demand "absolute unanimity of views in the scientific community.... Rather, the test is met if use of the technique is supported by a clear majority of the members of that community." (37 Cal.3d at p. 418, italics added.) Guerra likewise seems quite consistent with both Kelly and Shirley. Kelly itself, at one point, described the Frye test as calling for "substantial agreement and consensus in the scientific community." (See 17 Cal.3d at p. 31.) Of course, the trial courts, in determining the general acceptance issue, must consider the quality, as well as quantity, of the evidence supporting or opposing a new scientific technique. Mere numerical majority support or opposition by persons minimally qualified to state an authoritative opinion is of little value under the foregoing cases.
9. Summary and Conclusions We conclude as follows:
(a) Our Kelly doctrine survived Daubert and continues to represent the standard by which new scientific techniques should be measured before evidence derived therefrom may be admitted in court.
(b) HGN testing is a new scientific technique requiring compliance with Kelly.
(c) Kelly contemplates appropriate expert testimony and evidence showing that HGN testing is generally accepted by a typical cross-section of the relevant scientific community. Until compliance with Kelly is demonstrated, testimony of police officers who merely administer HGN tests is, by itself, inadequate for this purpose.
(d) "General acceptance" under Kelly means a consensus drawn from a typical cross-section of the relevant, qualified scientific community.
The judgment of the Court of Appeal reversing defendant's conviction is affirmed, and the Court of Appeal is directed to remand the case to the trial *613 court for a Kelly hearing in accordance with our opinion. If, at the conclusion of the hearing, the trial court concludes there is sufficient basis to properly admit the HGN testimony previously presented, the court should reinstate the judgment. If the trial court determines the evidence is insufficient to properly admit the testimony presented, then the court should order a new trial, if the People so elect. If the judgment is reinstated, or a new trial ordered, appellate review will be available to the parties regarding the trial court's ruling, limited to any new issues not previously resolved in this opinion.
Mosk, J., Kennard, J., Arabian, J., George, J., and Werdegar, J., concurred.
BAXTER, J.
I respectfully dissent. I do not join the majority in its conclusion that a horizontal gaze nystagmus (HGN) test administered as a field sobriety test by a police officer is based on a "novel scientific principle" and may not be relied on by an experienced police officer in support of a conclusion that a suspect is intoxicated unless it is shown to meet the Kelly/Frye criteria. (People v. Kelly (1976) 17 Cal. 3d 24 [130 Cal. Rptr. 144, 549 P.2d 1240]; Frye v. United States (D.C. Cir.1923) 293 F. 1013 [54 App.D.C. 46, 34 A.L.R. 145.) Assuming that conclusion is true, however, it does not follow that the trial court erred in admitting the opinion of Officer Van Achen and his reasons for reaching that opinion.
The majority hold that this conviction should be reversed without regard to, or even mention of, the Evidence Code provision that governs the admission of the testimony of Officer Von Achen that, in his opinion, defendant was under the influence of alcohol while driving an automobile. Evidence Code section 803 mandates exclusion of opinion evidence "based in whole or in significant part on matter that is not a proper basis for such an opinion." The testimony of Officer Von Achen was not based wholly or in significant part on the results of the HGN test. His observation of defendant's reaction to that test was only one of several equally or more informative factors which led him to the conclusion that defendant was intoxicated.
Further, as a threshold issue, it is doubtful that evidence of an officer's observations of a suspect's response to the HGN field sobriety test is properly described as scientific evidence. If those observations are scientific evidence because the test is one used in the medical profession, however, there is no reason that the opinion of a trained officer that those observations, with others, may not be relied on by a police officer in support of an opinion that a suspect is intoxicated. That opinion is admissible without subjecting all of the factors which support it to the former Kelly/Frye test. *614 While members of the medical profession believe that the test is an accurate indicator of intoxication, and thus the test may be termed "scientific" for that reason, administration and interpretation of the results of the test do not involve application of the type of scientific procedure and analysis to which the Kelly/Frye test should be applied.
In short, moving a finger in front of a suspect's eyes and observing the reaction of the eyeballs is something a trained police officer can do in the field. One need not be a rocket scientist to be able to observe and assess the reaction of the suspect. If a police officer is able to administer other field sobriety tests and offer an opinion based on the suspect's performance that the suspect is intoxicated, he can also be trained to administer the HGN test and should be able to testify that, based on his observations and experience, persons who perform as the suspect did on those tests are intoxicated.
The majority's reason for concluding that an officer's observations of the reaction of a person to whom the officer has administered the HGN test should be equated with sophisticated scientific techniques such as DNA testing, homologous lucocytic antibodies (HLA) typing, or other scientific procedures to which the Kelly/Frye requirements have been applied in the past, does not withstand examination.
Moreover, assuming that an HGN field sobriety test is a scientific procedure, I do not agree that evidence regarding the test is admissible only if the Kelly-Frye criteria are met. Section 28, subdivision (d) of article I of the California Constitution (section 28(d))[1] mandates admission of evidence which the Kelly-Frye test excludes. The majority offer no persuasive reason for retaining a procedure which makes admissibility of concededly relevant evidence turn on the opinion of a segment of the scientific community. The result is contrary to the command of section 28(d), and the reason easing the burden on trial judges to decide hard questions is constitutionally impermissible.
Implicit in the majority opinion is a conclusion that California trial judges are less competent than judges of the federal court to determine the admissibility of scientific evidence. State judges, they hold, should not be called *615 upon to determine whether scientific evidence is sufficiently reliable to be relevant and thus admissible in the courts of this state. No matter that it is the responsibility of a judge to make such decisions or that the California Constitution mandates that all relevant evidence be admitted in criminal cases. Because our judges should not be called upon to make hard decisions, they must abdicate their responsibility and leave the decision to a "consensus" of the scientific community. Whatever justification there once was for such a rule, that justification evaporated when section 28(d) was added to article I of the Constitution.
I
IS EVIDENCE BASED IN PART ON THE HGN TEST SUBJECT TO KELLY/FRYE?
A. Nature of the Evidence.
To put the issues in this case in their proper perspective, one must know what evidence was offered at defendant's trial and the basis of the trial court's order overruling defendant's objection to admission of the evidence about the HGN test administered to defendant.
Officer Michael Von Achen of the Huntington Beach Police Department was the arresting officer. He testified regarding the symptoms of intoxication he had observed before deciding to place defendant under arrest. Defendant's reaction to the HGN test was only one of those symptoms. Officer Von Achen did not testify that this test alone established that a person was under the influence of alcohol, or that the test could establish the percentage of blood alcohol.
Von Achen was not a novice in the field of driving under the influence (DUI) arrests. He had more than nine years of experience, before which he had attended four and one-half months of police academy training. That training included the procedure for administration of field sobriety tests and observation of a suspect's reactions to the tests. The tests included a "walk-the-line" test, a "one-leg balance" test, and a "finger-to-nose" test. The training had been followed by supervised field experience in which drunk driving arrests were emphasized. Von Achen had also received training on how to administer the HGN test and had received 15 additional hours of training in DUI investigation.
During Von Achen's nine and one-half years as a police officer, he had stopped "thousands" of suspected drunk drivers, of whom approximately five hundred to six hundred had then been arrested. In determining whether *616 a driver is under the influence of alcohol, Von Achen considered the HGN test reaction important because, if it appeared from that test that a person was probably not under the influence, he would release the person without doing the other tests. He did not form an opinion that the person was under the influence on the basis of the reaction to the HGN test alone. His training and experience taught him, however, that if a person is highly intoxicated, the eyes will not be able to follow the finger on a smooth track, but will constantly be catching up to it in a "jerking" motion. If Von Achen found that response to be present, he would administer additional tests.
Von Achen stopped defendant's car after he observed the car make a turn onto Pacific Coast Highway at a fairly fast rate of speed and then drift across to the center and accelerate quickly. The car was traveling 55 miles per hour in a 25-mile-per-hour construction zone. He noticed that defendant's eyes were very red and watery and his face flushed. Defendant's speech was thick and slurred. Defendant said he had had two beers. His balance was unsteady when he stepped out of the car.
After questioning defendant about when he had last eaten, slept, or had anything to drink and whether he had any illness or had used medication or drugs, and defendant explained that he had just had a cast removed from a leg in which he had pulled ligaments, Von Achen noticed the odor of alcohol on defendant's breath and person. He then attempted to have defendant perform a field sobriety test, the "modified attention" test. Defendant attempted to do that test which involves putting heels and toes together, hands down at the sides, head back with the back arched a little bit, and eyes closed. In that position he was to estimate thirty seconds and then open his eyes. Defendant swayed from front to back two to three inches and opened his eyes after thirty-seven seconds.
Von Achen then administered the HGN test. He observed that defendant was not following his finger smoothly. At that point he formed the opinion that defendant was under the influence of alcohol. He then administered the "ABC" test. Defendant recited the alphabet correctly but in thick and slurred speech. Von Achen did not administer other field sobriety tests because defendant had had an injury to his leg.
Von Achen formed his opinion that defendant had been driving under the influence of alcohol on the basis of the appearance of his eyes and face, the manner in which he was driving, his response to the questions he was asked, his thick and slurred speech, as well as his observations of defendant's performance on the tests administered to defendant.
*617 B. The Kelly/Frye test.
Next, an understanding of the evidence to which the Kelly/Frye rule has been applied and the reasons for the rule is essential. First, and most importantly, the Kelly/Frye rule applies only to expert opinion testimony. In California the admission of expert testimony is governed by statute. (Evid. Code, §§ 801-805.) Experts are allowed to testify regarding matters that are beyond the competence of the jury. The opinion will be helpful, however, only if it is validly drawn from appropriate data. (Mendez, Cal. Evidence (1993) p. 316.) Evidence Code section 801 establishes the criteria for admission of expert opinion testimony. "If a witness is testifying as an expert, his testimony in the form of an opinion is limited to such an opinion as is:
"(a) Related to a subject that is sufficiently beyond common experience that the opinion of an expert would assist the trier of fact; and
"(b) Based on matter (including his special knowledge, skill, experience, training, and education) perceived by or personally known to the witness or made known to him at or before the hearing, whether or not admissible, that is of a type that reasonably may be relied upon by an expert in forming an opinion upon the subject to which his testimony relates, unless an expert is precluded by law from using such matter as a basis for his opinion."
The expert is permitted to state the reasons for his opinion and the factors on which it is based. (Evid. Code, § 802.) If the opinion is "based in whole or in significant part" on matters that are not a proper basis for the opinion, the court may exclude the testimony unless the expert can state an opinion after excluding the improper matter. (Evid. Code, § 803.)
In Frye v. United States, supra, 293 F. 1013, from which this state's Kelly/Frye rule is derived, the court was called upon to determine the admissibility of expert testimony that is based on a new scientific technique. The test in question was a "systolic blood pressure deception test," which might be described now as a form of lie detection test. It was based on a theory that systolic blood pressure rises when a subject experiences fear, rage, or pain, and that because a person who consciously lies or conceals facts about, or guilt of, crime fears detection when under examination, there will be a rise in blood pressure corresponding to the extent of the subject's fear and attempt to control that fear. The defendant in a murder prosecution had offered an expert to testify to the result of this test. The trial court sustained the prosecutor's objection and excluded the evidence.
The circuit court upheld the ruling after holding that expert testimony on a scientific principle or discovery is admissible only when it has "crosse[d] *618 the line between the experimental and demonstrable stages.... Somewhere in this twilight zone the evidential force of the principle must be recognized, and while courts will go a long way in admitting expert testimony deduced from a well-recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs." (293 F. at p. 1014.)
Notwithstanding the appellation "Kelly/Frye," this court approved the Frye formulation 10 years before the Kelly decision. In Huntingdon v. Crowley (1966) 64 Cal. 2d 647 [51 Cal. Rptr. 254, 414 P.2d 382], it was applied to new techniques for blood typing. The court held that the trial court had correctly followed the rule stated in Frye, saying: "This is also the rule in California. (See, e.g., People v. Jones (1959) 52 Cal. 2d 636, 653 [343 P.2d 577] [`truth serum']; People v. Carter (1957) 48 Cal. 2d 737, 752 [312 P.2d 665] [`lie detector' test]; cf. People v. Williams (1958) 164 Cal. App.2d Supp. 858, 860-862 [331 P.2d 251] [`Nalline' test].)" (64 Cal.2d at p. 654.)
In People v. Kelly, supra, 17 Cal. 3d 24, this court reaffirmed its approval of the Frye test as applied to the admissibility of evidence of a "new and emerging technique of speaker identification by spectrographic analysis, commonly described as `voiceprint.'" (Id., at pp. 27-28.)
None of the "scientific" tests or methods at issue in those cases is in any way comparable to the HGN test as administered by a police officer in the field. The test results and expert opinion based on them were aptly described by the court in People v. Williams (1958) 164 Cal. App.2d Supp. 858, 860-861 [331 P.2d 251], quoting Wigmore: `When the testimony thus appearing to the ordinary layperson to lack a rational basis is founded on observations made with esoteric methods or apparatus ... the method should be explained by the witness, and if it be vouched for as accepted in his branch of learning, it suffices to admit his testimony.' Wigmore on Evidence, third edition, section 659."
Here, there is no esoteric method or apparatus. A layperson serving on the jury can readily comprehend the nature of the test and there is no mystery to the means by which the results are evaluated when a police officer testifies based on his training and experience that the subject's reaction indicated that the subject was intoxicated. The answer might be different if the officer purported to be able to determine the blood-alcohol level of the suspect on the basis of an HGN test (see People v. Ojeda (1990) 225 Cal. App. 3d 404, 409 [275 Cal. Rptr. 472]; People v. Loomis (1984) 156 Cal. App.3d Supp. 1, 6 [203 Cal. Rptr. 767]), but there is nothing in the manner in which a simple *619 description of a subject's reaction to the HGN is relayed to a jury that carries an "undeserved aura of certainty." (People v. Stoll (1989) 49 Cal. 3d 1136, 1156 [265 Cal. Rptr. 111, 783 P.2d 698].) In Stoll, this court emphasized the limited applicability of the Kelly/Frye test. "Kelly/Frye only applies to that limited class of expert testimony which is based, in whole or part, on a technique, process, or theory which is new to science and, even more so, the law.... [¶] The second theme in cases applying Kelly/Frye is that the unproven technique or procedure appears in both name and description to provide some definitive truth which the expert need only accurately recognize and relay to the jury. The most obvious examples are machines or procedures which analyze physical data. Lay minds might easily, but erroneously, assume that such procedures are objective and infallible." (Ibid.) "[A]bsent some special feature which effectively blindsides the jury, expert opinion testimony is not subject to Kelly/Frye." (Id., at p. 1157.)
We concluded in People v. Stoll, supra, 49 Cal. 3d 1136, 1157, that the expert psychological testimony offered in that case "raises none of the concerns addressed by Kelly/Frye. The methods employed are not new to psychology or the law, and they carry no misleading aura of scientific infallibility." (Italics omitted.) The same may be said for the HGN test as administered by a police officer in the field as one indicator of possible intoxication. The test is not new. It has been in use since the 1960's and was described in medical literature in 1958. (Malone v. City of Silverhill (Ala. Crim. App. 1989) 575 So. 2d 101, 103, revd. on other grounds, 575 So.2d (Ala. 1990); Aschan, Different Types of Alcohol Nystagmus (Sweden 1958) 140 Acta Otolaryngol Supp. 69.) It is recommended by the National Highway Traffic Safety Administration to aid in determining if a driver is intoxicated. (City of Fargo v. McLaughlin (N.D. 1994) 512 N.W.2d 700, 703.) Under a contract from the National Highway Traffic Safety Administration to develop the best possible field sobriety tests, Dr. Marcelline Burns of the Southern California Research Institute concluded that the HGN test is the best single index of intoxication. (State v. Superior Court (1986) 149 Ariz. 269 [718 P.2d 171, 173].)
The majority conclude that we should nonetheless deem the HGN test to be new and wait even longer for the scientific community to do further research and achieve a consensus. To the majority the "long-standing use by police officers seems less significant a factor than repeated use, study, testing and confirmation by scientists or trained technicians." (Maj. opn., ante, at p. 606.) I disagree. There is no reason to believe that after 30 years the scientific community has any interest in further studies of HGN. For laboratory purposes there are other more accurate tests of blood-alcohol level. The HGN phenomenon was once of interest, but unless it has practical *620 use outside of field sobriety testing by law enforcement, the majority assumption that further testing will be done by the scientific community seems unfounded.
More importantly, however, the long-standing use by police officers is extremely significant because that is the "laboratory" in which the correlation between HGN and blood-alcohol level has been established. The majority ignore the fact that when an arrest for DUI is made on the basis of field sobriety tests, a blood, breath, or urine test follows. The correlation between DUI arrests in which the HGN reaction was observed by the officer who determined that the suspect was under the influence of alcohol and a subsequent breath, blood, or urine confirmation that the suspect's blood-alcohol level exceeds the legal limit is the only testing that should be necessary.
I see no distinction with respect to their complexity, sophistication or scientific "aura" between the HGN test and other field sobriety tests administered to persons suspected of DUI violations. The majority conclude that the HGN test is a scientific procedure in part because the "jury might be unduly swayed by HGN evidence solely by reason of its technical nomenclature." (Maj. opn., ante, at p. 607.) If so, the name can be changed. Certainly the majority does not mean to hold that the name of a procedure determines whether it is a scientific test subject to the Kelly/Frye requirements. There is no relevant scientific community available to reach a consensus on a nonscientific matter simply because the name of the procedure sounds impressively scientific.
If the HGN test had originated with law enforcement officers rather than the scientific community, and was named simply the "follow the finger test," and an officer testified that in the officer's experience a person under the influence of alcohol or drugs has difficulty in smoothly following a moving finger with his eyes, we would not have this question before us at all. Even if the scientific community does not agree as to the reliability of an HGN test for determining blood-alcohol levels, a police officer who has administered the test many times and has observed a correlation between subsequent tests for blood-alcohol level and the HGN test results is certainly able to testify, based on his experience, that a jerky movement of the eyeballs when the test is administered is a symptom of intoxication. A jury would have no difficulty in comprehending the nature of the test. Indeed, we allow lay opinion testimony regarding both drug and alcohol intoxication because laypersons are sufficiently familiar with the symptoms that they are able to identify them. A layperson may not be aware that the jerking eyeball movement is another symptom, but he or she can easily comprehend the nature of the *621 police officer's inquiry when the test is described. As used in the field by a police officer, this test is not a new scientific technique surrounded by a "`misleading aura of certainty'" and jurors are not likely to give it "considerable weight" as scientific evidence presented by "`experts' with impressive credentials." (People v. Kelly, supra, 17 Cal. 3d 24, 31-32.) "The gaze nystagmus test, as do the other commonly used field sobriety tests, requires only the personal observation of the officer administering it. It is objective in nature and does not require expert interpretation. Objective manifestations of insobriety, personally observed by the officer, are always relevant where, as here, the defendant's physical condition is in issue." (State v. Nagel (1986) 30 Ohio App. 3d 80 [506 N.E.2d 285, 286].)
The Supreme Court of Minnesota recently held that the HGN test does meet the Frye criteria, but also noted that it is not a new test. It distinguished the question of whether the test itself is a reliable indicator of the presence of drugs from the question of whether an officer may testify regarding an opinion formed on the basis of administration of that test and others to a suspect. (State v. Klawitter (Minn. 1994) 518 N.W.2d 577.)
The court held that the evidence had been properly admitted even though there was not a consensus as to its utility in determining drug impairment because the HGN test was only one element of a twelve-step protocol supportive of a conclusion based on the entire protocol that drug impairment was present. The arresting officer had testified regarding the factors in the protocol that led him to the opinion that the defendant was under the influence of cannabis. The defendant had challenged the reliability of the protocol, but the trial court ruled that the officer could testify concerning the observations made during the protocol and his opinion based on those observations.
The Minnesota Supreme Court upheld the trial court's ruling. It reasoned: "[P]roperly viewed, the protocol ... is not itself a scientific technique but rather a list of the things a prudent, trained, and experienced officer should consider before formulating or expressing an opinion whether the subject is under the influence of some controlled substance. [¶] ... [¶] Only the tests for horizontal and vertical nystagmus and for convergence are out of the ordinary, but they can hardly be characterized as emerging scientific techniques. Nystagmus and convergence have long been known, and the tests contemplated by the protocol have been in common medical use without change for many years. The tests are simple and do not require the use of complicated equipment.... [¶] ... [¶] ... It is not contended, however, that the presence or absence of nystagmus is determinative of the presence of drugs, but only that nystagmus, when it is present, may be an element *622 supportive of a conclusion of drug impairment based on the elements of the protocol, taken as a whole....
"The other point of the defense challenge goes to the police officer's competence to draw conclusions based on the protocol. Basically, however, following the protocol does not involve any significant scientific skill or training on the part of the officer. Drug recognition training is not designed to qualify police officers as scientists but to train officers as observers. The training is intended to refine and enhance the skill of acute observation which is the hallmark of any good police officer and to focus that power of observation in a particular situation.... [T]he protocol ... dresses in scientific garb that which is not particularly scientific.... Moreover, calling the officer an `Expert' [is misleading and] tends to lend weight to the officer's opinion. [¶] ... The real issue is not the admissibility of the evidence but the weight it should receive, and that is a matter for the jury to decide without being led to believe that the evidence is entitled to greater weight than it deserves.... [W]e conclude that opinion testimony based on nystagmus testing is admissible if a sufficient foundation has been laid for the opinion expressed and provided that the trial court, when requested, gives an appropriate cautionary instruction." (518 N.W.2d at pp. 584-586, fns. omitted.)
The Minnesota high court is not alone in concluding that use of a suspect's reaction to the HGN test as an indicator of intoxication is not use of a scientific test subject to the Frye criteria as it is no more scientific than other field sobriety tests. (Whitson v. State (1993) 314 Ark. 458 [863 S.W.2d 794, 798]; State v. Edman (Iowa 1990) 452 N.W.2d 169, 170; State v. Murphy (Iowa 1990) 451 N.W.2d 154, 157; State v. Garris (La. Ct. App. 1992) 603 So. 2d 277, 282; State v. Nagel, supra, 30 Ohio App. 3d 80 [506 N.E.2d 285, 286]; State v. Sullivan (S.C. 1993) 426 S.E.2d 766, 769.) Most recently the North Dakota Supreme Court, in City of Fargo v. McLaughlin, supra, 512 N.W.2d 700, reviewed the scientific literature and cases and concluded that expert testimony regarding HGN is not necessary.
"We begin our analysis by noting that the underlying scientific basis for HGN testing that intoxicated persons exhibit nystagmus is undisputed, even by those cases and authorities holding the test inadmissible without scientific proof in each case. State v. Superior Court, 718 P.2d at 177, 181; State v. Witte [(1992) 251 Kan. 313] 836 P.2d at 1112; State v. Garrett [(1991)] [119 Idaho 878] 811 P.2d at 491; State v. Cissne [(1994) 72 Wash. App. 667], 865 P.2d 566; Rouleau, Unreliability of the Horizontal Gaze Nystagmus Test, 4 Am.Jur. P.O.F.3d 439, 446 (1989). It is generally accepted that a person will show a greater degree of nystagmus at higher levels of *623 intoxication, and that a properly conducted HGN test can identify nystagmus. See State v. Superior Court, 718 P.2d at 181; State v. Witte, 836 P.2d at 1112-1113. We take notice of these physiological facts, and conclude that it is unnecessary to require expert testimony of these widely accepted principles.
"These principles comprise the only `scientific' components of the HGN test. The officer, based upon his training in these principles, observes the objective physical manifestations of intoxication, and no expert interpretation is required. State v. Murphy, 451 N.W.2d at 157; State v. Nagel, 506 N.E.2d at 286. The test requires simply that the officer observe the subject's eyes following a moving object." (City of Fargo v. McLaughlin, supra, 512 N.W.2d at p. 706.)
The North Dakota high court also relied on the view of the Idaho Supreme Court in State v. Gleason (1992) 119 Idaho 878 [844 P.2d 691, 695], which it quoted (512 N.W.2d at p. 707): "[The officer's] testimony relating to the HGN test results was not offered as independent scientifically sound evidence of [the defendant's] intoxication. Rather, it was offered and admitted for the same purpose as other field sobriety test evidence a physical act on the part of [the defendant] observed by the officer contributing to the cumulative portrait of [the defendant] intimating intoxication in the officer's opinion. [¶] In sum, we agree with those cases that stress that HGN test results are admissible in conjunction with other field sobriety tests. [Citations.] Where the officer has conducted other field sobriety tests in addition to the HGN test, the results of the HGN test are but one of the physical observations to support the officer's opinion that the accused was intoxicated."
The Supreme Court of Arkansas has reached the same conclusion. No Frye-type hearing is necessary both because it is not a novel scientific procedure and because evidence of an officer's observations of a subject's reaction to the test is not scientific evidence to which the Frye test is applicable: "HGN testing has been in existence for 35 years and officers are trained in this technique at the State Police Academy.... [T]he results of nystagmus testing for purposes of showing some indication of alcohol was not novel scientific evidence requiring a preliminary inquiry. Our opinion, however, might well be different had the officer attempted to quantify blood alcohol content based solely on the HGN test. [¶] In sum, we hold that the results of the HGN test were relevant to show alcohol consumption in conjunction with the results of other field sobriety tests performed." (Whitson v. State, supra, 863 S.W.2d 794, 798.)
There is simply no reason to restrict testimony such as that given in this case on the basis that the police officer's opinion was based on scientific *624 evidence that did not satisfy the Kelly/Frye test. That test should not be applied.[2] The trial court did not apply it[3] and we should not.
Officer Von Achen did not testify that the HGN test established that defendant had alcohol or any specific percentage of alcohol in his system at the time of arrest. He testified that defendant's reaction to the test and the other factors described above led him to form the opinion that defendant had been driving under the influence of alcohol. I would affirm the judgment of the trial court for that reason alone.
II
ARTICLE I, SECTION 28(d)
An equally important reason why this judgment must be affirmed is that the Kelly/Frye rule excludes relevant evidence which this court does not *625 have the power to render inadmissible by the perpetuation of a court-made rule of evidence.
The court has repeatedly recognized that Frye excludes relevant evidence. The court did so in People v. Kelly, supra, 17 Cal. 3d 24, 30-31, and again in People v. Stoll, supra, 49 Cal. 3d 1136, 1156. Yet section 28(d) commands: "[R]elevant evidence shall not be excluded in any criminal proceeding." The intent of the electorate when this provision was added to the Constitution by initiative in 1982 was to overturn judicially created exclusionary rules which keep relevant evidence from the trier of fact. (People v. Harris (1989) 47 Cal. 3d 1047, 1081-1082 [255 Cal. Rptr. 352, 767 P.2d 619]; In re Lance W. (1985) 37 Cal. 3d 873, 888-889 [210 Cal. Rptr. 631, 694 P.2d 744].)
The majority hold that what will now be simply "Kelly" continues to be an appropriate means to ensure that evidence is sufficiently reliable to be admissible under Evidence Code section 350.[4] The majority concede that the California Evidence Code provisions governing opinion testimony are sufficiently broad that they might be interpreted to permit admission of evidence that does not meet the Kelly/Frye standard. They "presume," however, that the court was aware of that possible breadth when Kelly was decided and nonetheless concluded that the Frye test of admissibility was compatible with those provisions. (Maj. opn., ante, at p. 599.)
It is, of course, necessary to engage in that presumption since the Kelly opinion does not address the question. The majority fail to note or acknowledge that Kelly is not the decision in which the Frye test was adopted. Indeed, the Kelly opinion itself states: "We have expressly adopted the foregoing Frye test.... (Huntingdon v. Crowley (1966) 64 Cal. 2d 647, 653-654....)" (Kelly, supra, 17 Cal. 3d 24, 30.) Huntingdon v. Crowley, supra, while decided in 1966, involved a claim of error in a pre-1965 trial, however, and itself relied on cases decided before the Evidence Code was adopted in 1965. (64 Cal.2d at p. 652, fn. 4.) Thus, this court has never actually or presumptively considered whether the Frye test is compatible with the Evidence Code provisions governing admission of expert testimony.
The majority assert, in support of adherence to the Frye, now Kelly, test, that there have been no significant developments since Kelly was decided to justify abandoning the Frye test. Therefore, the majority conclude, principles of stare decisis apply and the Frye, now Kelly, test is still to be applied *626 regardless of whether, contrary to the command of section 28(d), the result is to exclude relevant evidence.
The electorate which adopted section 28(d) as part of Proposition 8, "The Victims' Bill of Rights," in 1982 may be surprised to learn that it was not a significant development or that a majority of this court does not regard it as such.[5] I do. I believe it is incumbent on this court to reconsider the question of admissibility of expert scientific evidence in light of both the Evidence Code provisions and the command of section 28(d) that all relevant evidence be admitted unless barred by the federal Constitution or sections 352, 782, or 1103 of the Evidence Code.
While the Kelly/Frye test may be "appropriate," it is overinclusive and thus is not permissible under section 28(d). The ramifications of the court's holding today extend far beyond the erroneous application of Kelly to Officer Von Achen's testimony in this case. It will cause the exclusion of relevant evidence based on new scientific discoveries, whether proffered by the defense or the prosecution, until a sufficient number of scientific experts in the field take enough interest in the subject to enable the party offering the evidence to satisfy a court that there is a consensus among experts in that field as to the reliability of the evidence. While the rule is lauded as promoting uniformity in the admission of evidence, that uniformity will be of no benefit to the defendant who is convicted and sentenced to death when probative and even crucial evidence is excluded because the scientific technique through which it was developed is "new" and, although the developer may be well respected in the field, the technique has not yet attracted enough attention to achieve a "consensus" among members of the "relevant" scientific community. That uniformity may also deny the prosecution the use of highly probative evidence of guilt in the absence of which a murderer may go free.
The majority's lack of confidence in the ability of trial judges and juries to weigh the reliability of scientific evidence is doubly ironic because prosecutors themselves accept as dispositive reliable evidence that has not been held admissible under the Kelly/Frye criteria. DNA evidence, for instance, was recently accepted by the San Diego County District Attorney as proof of the innocence of a man convicted of rape and confined in prison for 10 years. (Perry, DNA Test Frees Inmate After 10 Years, L.A. Times (Sept. 29, 1994) p. *627 A-3, col. 2.) Under the rule perpetuated today by the majority, the election to rely on such evidence is a one-way street. Had the defendant offered that evidence at trial, and the prosecutor objected, the evidence would have been excluded unless the court concluded that it satisfied Kelly/Frye, and the defendant would have remained in prison awaiting the elusive consensus of the scientific community.
The Kelly/Frye test may exclude relevant evidence for many years after the new technique is developed even though a substantial body of experts agrees that the technique is reliable. That may happen here. Although the HGN test has been in use for 30 years, and the literature establishes that it is accepted as reliable by numerous experts whose qualifications to make that judgment are unimpeachable, other experts disagree. On remand, therefore, the trial court may exclude the evidence since the majority hold that the HGN test is still a new technique if the court concludes the test is not yet accepted as reliable by a "consensus" of experts.
Section 28(d) and the Evidence Code place no such limitation on the admissibility of evidence, however. Instead, Evidence Code section 351 mandates: "Except as otherwise provided by statute, all relevant evidence is admissible." The only limitation on the admission of relevant evidence found in the Evidence Code, absent a specific exclusion (see, e.g., Evid. Code, §§ 351.1, 352.1, 900 et seq., 1100 et seq.) preserved by section 28(d)), is that the evidence be relevant (Evid. Code, § 350) and that its probative value not be substantially outweighed by a danger of undue prejudice or necessitate undue consumption of time. (Evid. Code, § 352.) The majority concede that the California statutory requirements governing admission of evidence mirror those of the federal government and acknowledge the conclusion of the United States Supreme Court in Daubert v. Merrell Dow Pharmaceuticals, Inc. (1993) 509 U.S. ___ [125 L. Ed. 2d 469, 113 S. Ct. 2786] that those statutes make relevancy the sole criterion for admission. Yet the majority nowhere explain the authority for their conclusion that this court may continue to order exclusion of relevant evidence because the evidence does not meet the Kelly/Frye criteria.
We look in vain for the source of such judicial power only to be referred back to the Kelly decision which, we are told, explains why the Frye approach to determining reliability of expert testimony about scientific techniques is appropriate. Unfortunately, Kelly offers no insight into the source of judicial power to exclude relevant evidence. In fact, Kelly acknowledges that the admission of evidence based on new scientific techniques could be left to the trial court "in which event objections, if any, to *628 the reliability of the evidence (or of the underlying scientific techniques on which it is based) might lessen the weight of the evidence but would not necessarily prevent its admissibility." (Kelly, supra, 17 Cal. 3d 24, 31.) The court explained that the primary advantage, and its basis for mandating use, of the Frye test, was that it is "conservative." (Ibid.)
Neither the Evidence Code nor section 28(d) empowers the court to exclude reliable evidence simply because it has not been generally accepted in the relevant scientific community. Assuming that evidence might be so unreliable as to pose a danger that its prejudicial impact might substantially outweigh its probative value, those provisions do not authorize the court to delegate its responsibility to determine admissibility of evidence under Evidence Code section 352 to a segment of the scientific community. Nor is the majority justified in their reliance on legislative silence to justify the continued exclusion of relevant evidence on the basis of Kelly. Section 28(d) mandates admission of all relevant evidence. In People v. May (1988) 44 Cal. 3d 309, 320 [243 Cal. Rptr. 369, 748 P.2d 307], the author of the majority opinion himself recognized that it would "wholly frustrate this intent were we to hold that [a judicially created rule of evidence] survived merely because the Legislature had acknowledged the existence of judicial rules...."
The command of section 28(d) could not be clearer. Section 28(d) has eliminated all judicially created exclusionary rules "to ensure that all relevant evidence be admitted. That purpose cannot be effectuated if the judiciary is free to adopt exclusionary rules that are not authorized by statute or mandated by the Constitution." (In re Lance W., supra, 37 Cal. 3d 873, 889.) It is both unnecessary and futile to search for evidence of an intent in the electorate to abrogate the Kelly/Frye test specifically.
Therefore, assuming that it was based on a new scientific procedure, the sole basis on which Officer Von Achen's testimony could have been excluded in this case was if the HGN test or the testimony regarding its administration and outcome was perceived to be so unreliable that its probative value was substantially outweighed by its potential for prejudice. (Evid. Code, § 352.) That was not established by defendant and the majority does not suggest that the record would support exclusion on that basis.
Moreover, as noted earlier, even assuming that the officer's opinion regarding defendant's intoxication was based in part on the HGN test and those results were improper evidence, the opinion was admissible under *629 Evidence Code section 803 since it was not based in significant part on the results of that test.[6]
I would reverse the judgment of the Court of Appeal.
NOTES
[1] All references to the Constitution are to the California Constitution.
Section 28(d) of article I states: "Except as provided by statute hereafter enacted by a two-thirds vote of the membership in each house of the Legislature, relevant evidence shall not be excluded in any criminal proceeding, including pretrial and post conviction motions and hearings, or in any trial or hearing of a juvenile for a criminal offense, whether heard in juvenile or adult court. Nothing in this section shall affect any existing statutory rule of evidence relating to privilege or hearsay, or Evidence Code, Sections 352, 782 or 1103. Nothing in this section shall affect any existing statutory or constitutional right of the press." (Italics added.)
[2] There is a substantial body of opinion in states which consider the HGN test to be a scientific test subject to Frye that its reliability has been established and the evidence of its results may be admitted without expert opinion to establish its reliability in each case. (See Emerson v. State (Tex. Crim. App. 1994) 880 S.W.2d 759; State v. Garrett (1991) 119 Idaho 878 [811 P.2d 488, 491]; People v. Buening (1992) 229 Ill. App. 3d 538 [170 Ill. Dec. 542, 547-548, 592 N.E.2d 1222, 1227-1228]; State v. Armstrong (La. Ct. App. 1990) 561 So. 2d 883, 887; State v. Bresson (1990) 51 Ohio St. 3d 123 [554 N.E.2d 1330, 1334]; Anderson v. State (Tex. Ct. App. 1993) 866 S.W.2d 685, 686; Howard v. State (Tex. Ct. App. 1987) 744 S.W.2d 640, 641.)
For a comprehensive review of the literature regarding the HGN test see State v. Superior Court, supra, 718 P.2d 171. The Arizona Supreme Court concluded on the basis of that literature and evidence offered in that case that the HGN test is recognized as the most reliable of the roadside field sobriety tests. (Id., at pp. 173, 177.) That court concluded that the test meets the Frye standard because: "The evidence demonstrates that the following propositions have gained general acceptance in the relevant scientific community: (1) HGN occurs in conjunction with alcohol consumption; (2) its onset and distinctness are correlated to BAC [blood-alcohol content]; (3) BAC in excess of.10 percent can be estimated with reasonable accuracy from the combination of the eye's tracking ability, the angle of onset of nystagmus and the degree of nystagmus at maximum deviation; and (4) officers can be trained to observe these phenomena sufficiently to estimate accurately whether BAC is above or below.10 percent." (Id., at p. 181.)
[3] Prior to the admission of Officer Von Achen's testimony in which he explained his reasons for his DUI opinion, defendant had moved under Evidence Code section 402 to exclude evidence regarding the HGN test on the ground that there was no scientific acceptance of the test under the Kelly/Frye criteria as a valid test for detecting the presence of alcohol in a person's system. He also objected, relying on People v. Loomis, supra, 156 Cal. App.3d Supp. 1, and People v. Ojeda, supra, 225 Cal. App. 3d 404, on the ground that the reliability of the testing method had not been shown in the scientific community. The trial judge overruled the objection, ruling that Kelly/Frye was inapplicable because the test was not a specific test for the determination of the presence of alcohol. That conclusion was correct as the officer testified only about a symptom of alcohol in the system just as he did when testifying about slurred speech, a red face, and bloodshot eyes.
[4] One expert has criticized the court for "erroneously" assuming in People v. Harris, supra, 47 Cal. 3d 1047, that unreliable evidence is not relevant evidence. (See Mendez, California Evidence, supra, at p. 322.)
[5] Their surprise may be compounded since only two years ago we observed that the voters adopted Proposition 8, of which section 28(d) was a part, "to make significant substantive and procedural changes in California criminal law." (People v. Wheeler (1992) 4 Cal. 4th 284, 291 [14 Cal. Rptr. 2d 418, 841 P.2d 938].)
[6] Of course, since they are not expressly exempted, the Evidence Code limitations on admission of expert testimony themselves may have been abrogated by section 28(d).