PEOPLE v. SHELTON (1984)

Glenn Brinton Shelton appeals a judgment entered on a jury verdict of guilty of violating Vehicle Code section 23153, subdivision (a) (causing injury to another while driving under the influence).¹  Imposition of sentence was suspended and he was placed on probation for four years on condition he spend 120 days in county jail.   Appealing, he claims error in the giving of two jury instructions and in exclusion of statements made by the victim.

Shelton was driving northbound at dusk on February 21, 1982, when his car was hit by a motorcycle driven by Michael Bonney on which Geri Hodge was a passenger.   Both of the cycle riders were seriously injured.   Bonney was hospitalized for two months, Hodge for ten days.

Shelton testified he was making a left turn to cross the southbound lane of a divided highway.   He paused to permit pedestrians to cross, saw no oncoming headlamps and, as he began to proceed, his car was hit by the motorcycle.   Hodge testified she saw Shelton make the turn without pause.   Bonney swerved to avoid Shelton's car, but the car hit the front part of the motorcycle.   Nine-year-old Chip McDonald saw Shelton's car hit the front of the motorcycle, Shelton leave the vehicle and walk over to a liquor store.   Shelton testified he was going to telephone the police but did not as emergency vehicles had already been dispatched.

California Highway Patrolman Nezet arrived at the scene.   Shelton attempted to perform but failed a series of field sobriety tests and was arrested.   Chip McDonald watched the tests and testified Shelton was “wobbling ․ and walking all over the place,” that he walked a “crooked line” and almost fell.   His breathalizer blood alcohol test some two hours after the incident measured .20/.21.

The court instructed the jury in the words of CALJIC No. 12.61 (1982 Revision):²

“If the evidence establishes beyond a reasonable doubt that at the time of the cehmical [sic] analysis of the defendant's blood, breath, or urine there was 0.10 percent or more by weight of alcohol in the defendant's blood, you should find that the defendant was under the influence of an alcoholic beverage at the time of the alleged offense, unless from all of the evidence you have a reasonable doubt that the defendant was under the influence of an intoxicating beverage at the time of the alleged offense.”

Shelton says the instruction is constitutionally infirm in that it shifts the burden of proof of an element of the offense, driving a vehicle while under the influence of alcohol, from the People to him;  his conviction must be set aside as the People bear the burden of proving all elements of the offense beyond a reasonable doubt.   This argument was made and rejected in People v. Schrieber, 45 Cal.App.3d 917, 923, 119 Cal.Rptr. 812, and People v. Lachman, 23 Cal.App.3d 1094, 1097, 100 Cal.Rptr. 710.

 The alcohol-related carnage on the highways has been brought to public attention by law enforcement (see Annual Report of Fatal and Injury Motor Vehicle Traffic Accident (1977) Calif. Highway Patrol ³ ), organized private citizens (Mothers Against Drunk Drivers (MADD)) and the California Legislature (see § 23160 et seq., increasing penalties for persons convicted of driving under the influence of alcoholic beverages and drugs).   We approach the task of analyzing the legislative efforts mindful of the impact our decision may have on these concerns.   This statute, and those increasing the penalties, are expressions of public outrage at the death and injury on the highway, and are designed to combat the hazard drunk drivers pose to society (see People v. Duroncelay, 48 Cal.2d 766, 772, 312 P.2d 690;  see also South Dakota v. Neville, 459 U.S. 553, 103 S.Ct. 916, 920, 74 L.Ed.2d 748;  Taylor v. Superior Court, 24 Cal.3d 890, 898–899, 157 Cal.Rptr. 693, 598 P.2d 854).   In testing the constitutionality of a statute, too, a cardinal principle is that the legislative act is presumed to be constitutional, and unconstitutionality must be clearly shown and doubts resolved in favor of its validity (see 5 Witkin, Summary of Cal.Law (8th ed. 1974) § 43, p. 3281;  see also 16 Am.Jur.2d, Constitutional Law, § 137 et seq.;   and see Zidell v. Bright, 264 Cal.App.2d 867, 869, 71 Cal.Rptr. 111).

 In California, a defendant is “under the influence” of alcohol (within the meaning of § 23153) if “liquor has so far affected the nervous system, brain, or muscles of the driver of an automobile as to impair, to an appreciable degree, his ability to operate his car in the manner that an ordinarily prudent and cautious man, in the full possession of his faculties, using reasonable care, would operate or drive a similar vehicle under like conditions” (People v. Dingle, 56 Cal.App. 445, 449, 205 P. 705, construing the phrase “under the influence of intoxicating liquor,” as used in § 17 of the Motor Vehicle Act (Stats.1919, p. 214);  emphasis added).

Section 23155 reads:

“(a) Upon the trial of any criminal action, or preliminary proceeding in a criminal action, arising out of acts alleged to have been committed by any person while driving a vehicle while under the influence of an alcoholic beverage in violation of subdivision (a) of Section 23152 or subdivision (a) of Section 23153, the amount of alcohol in the person's blood at the time of the test as shown by chemical analysis of that person's blood, breath, or urine shall give rise to the following presumptions affecting the burden of proof:

“(1) If there was at that time less than 0.05 percent by weight of alcohol in the person's blood, it shall be presumed that the person was not under the influence of an alcoholic beverage at the time of the alleged offense.

“(2) If there was at that time 0.05 percent or more but less than 0.10 percent by weight of alcohol in the person's blood, that fact shall not give rise to any presumption that the person was or was not under the influence of an alcoholic beverage, but the fact may be considered with other competent evidence in determining whether the person was under the influence of an alcoholic beverage at the time of the alleged offense.

“(3) If there was at that time 0.10 percent or more by weight of alcohol in the person's blood, it shall be presumed that the person was under the influence of an alcoholic beverage at the time of the alleged offense.

“(b) Percent by weight of alcohol in the blood shall be based upon grams of alcohol per 100 milliliters of blood.

“(c) This section shall not be construed as limiting the introduction of any other competent evidence bearing upon the question whether the person ingested any alcoholic beverage or was under the influence of an alcoholic beverage at the time of the alleged offense.”

The constitutionality of the section 23155 presumption has been upheld by the California courts (People v. Schrieber, supra, 45 Cal.App.3d 917, 923, 119 Cal.Rptr. 812;  People v. Lachman, supra, 23 Cal.App.3d 1094, 1097, 100 Cal.Rptr. 710, both dealing with the identical presumption of former section 23126).   In Lachman, the court stated (at p. 1097, 100 Cal.Rptr. 710):

“A statutory presumption affecting the burden of proof in a criminal cause does not alter the People's duty to prove defendant's guilt beyond a reasonable doubt.   It merely allows proof of an ultimate fact by permitting that fact to be presumed from proof of a preliminary fact.   Whether the ultimate fact is proved by direct evidence or by a presumption which arises from proof of a preliminary fact, the defendant's burden of rebuttal remains the same:  he need only raise a reasonable doubt as to the sufficiency of the proof of the ultimate fact.  [Citations.]”  (See also People v. Schrieber, supra, 45 Cal.App.3d 917, 923, 119 Cal.Rptr. 812.)

As pointed out in Schrieber, the establishment by law of certain disputable presumptions in criminal cases is neither new nor uncommon (see Witkin, Cal.Criminal Procedure, § 344–346, pp. 336–339;  Witkin, Cal.Criminal Procedure, 1983 Supp., pp. 460–461;  United States v. Gainey, 380 U.S. 63, 66–67, 85 S.Ct. 754, 757, 13 L.Ed.2d 658).

 Use of a presumption in proof of the elements of a criminal case is a rule of evidence which the Legislature is authorized to establish (Tot v. United States, 319 U.S. 463, 467–468, 63 S.Ct. 1241, 1244–1245, 87 L.Ed. 1519).   The U.S. Supreme Court stated, however, the Fifth and Fourteenth Amendments set limits on the power of the Legislature to make the proof of one fact or group of facts evidence of the ultimate fact on which the guilt is based.

“Under our decisions, a statutory presumption cannot be sustained if there is no rational connection between the fact proved and the ultimate fact presumed, if the inference of the one from proof of the other is arbitrary because of lack of connection between the two in common experience.”  (Tot v. United States, supra, 319 U.S. at pp. 467–468, 63 S.Ct. at pp. 1244–45.)

The California cases have consistently held that limitation has not been violated in the case of the stated blood alcohol level creating the presumption the person is “under the influence” (People v. Schrieber, supra, 45 Cal.App.3d 917, 119 Cal.Rptr. 812;  People v. Lachman, supra, 23 Cal.App.3d 1094, 100 Cal.Rptr. 710).   Certainly it must be admitted there is a real and rational connection between the blood alcohol level and being under the influence.

 Probably no fact is more firmly established medically than that the ingestion of alcohol in any substantial quantity impairs one's ability to drive a vehicle (People v. Schrieber, supra, 45 Cal.App.3d 917, 920, 119 Cal.Rptr. 812;  see Alcohol and the Impaired Driver, AMA Committee on Medicolegal Problems (AMA Report 1970) pp. 27–34).⁴  The only question subject of debate is the degree of impairment to a particular individual occasioned by the particular quantity of alcohol ingested.   It is apparent different persons react differently to a given quantity and this invariably requires a case by case analysis.

The relationship between blood alcohol level and drunkenness has been the subject of numerous studies and some disagreement exists as to when a person is “drunk.”   We should caution the reader, however, in analyzing the results of these studies, “drunkenness” is not the same as “being under the influence,” the former being gross impairment and the latter being less debilitating (see People v. Dingle, supra, 56 Cal.App. 445, 452–453, 205 P. 705).   Those studies are, however, significant.   The AMA Report makes the following observations:

“A number of investigators have studied the correlation between the concentration of alcohol in the blood and the presence of acute alcoholic intoxication.5  Lambercier et al published a review of the correlation between drunkenness and the blood-alcohol level in 1946;  this review covers thirty authors and a total of 30,000 subjects.   Table 36 contains the results of a study of over 6,000 subjects by seven investigators in different parts of the world and relates the percentage of subjects diagnosed as drunk with the blood-alcohol level observed.

“All of these indicate that at levels of 0.15 percent w/v (150mg/100ml), or over, more than 50 percent of persons are grossly intoxicated.   A very few persons are drunk at blood-alcohol levels of 0.05 percent w/v (50 mg/100ml), while practically all people are drunk at levels above 0.35 percent w/v (350 mg/100 ml).   Persons with histories of long use of alcohol are less likely to show signs of gross intoxication at lower levels, since they have learned to control their behavior.   They deliberately attempt to conceal their intoxication and have a greater degree of tolerance than persons with less experience in repeated alcoholic consumption.   Despite the multitude of signs and symptoms, acute alcoholic intoxication can be at times confused with other conditions.   These are discussed in the chapter on Diagnosis of Alcoholic Intoxication.”

Eleven conclusions of these tests as applied to persons driving a vehicle are summarized in the AMA Report this way:

“The determination of alcohol in the blood is an important means of defense for the sober driver.   At levels as low as 0.08% two-thirds of the drivers on careful examination showed signs of marked impairment.”  (P.H. Andresen:  Traffic and Alcohol, Medicolegal J 18:98 (1950).)

“Drinking drivers may be classified as intoxicated solely on the basis of chemical tests.   Occasional discrepancies will be found which do not invalidate the significance of results for police work.”  (D.F. Bavis:  145 Drunken Drivers—A Blood and Urine Alcohol Study, J Lab Clin Med 25:823 (1940).)

“The probability of accident involvement increases rapidly at levels above 0.08% and becomes extremely high at levels above 0.15%.”  (R.F. Borkenstein, R.F. Crowther, R.P. Shumate, W.B. Ziel and R. Zylman:  The Role of the Drinking Driver in Traffic Accidents, Dept. of Police Administration, Indiana University (1964).)

“Clinical tests showed all drivers under the influence at 0.07%.”  (L.E. Gelin and G. Wretmark:  Alcohol och Korskicklighet:  Resultat av Praktiska Kooprov (Alcohol and Responsible Driving:  Results of Practical Driving Test), Svenski Lakartidn 48:3041 (1951).)

“The level of 0.15% used in defining intoxication should be lowered.”  (O. Gruner and W. Werner:  Untersuchung uber die Gefahrlichkeit des Alkoholbeeinflussten Kraftfahrers (Investigation on the Danger Presented by the Driver Under the Influence of Alcohol), Med Sachverst 53:73 (1957).)

“Blood-alcohol levels between 0.08 and 0.11% definitely diminish the ability to drive.”  (M.R. Lambercier and R.M. DuPan:  L'Intoxication Alcoolique Aigue et Les Accidents d'Automobile.   II.  Les Effets Physique de l'Intoxication Alcoolique Aigue.  (Acute Alcoholic Intoxication and Automobile Accidents.   II.  Psychological Effects of Acute Alcoholic Intoxication), Schweiz Med Wschr 76:395–8, 421–8 (1946).)

“Blood-alcohol levels over 0.05% significantly increased the probability of being involved in an accident.”  (G.H.W. Lucas, W. Kalow, J.D. McColl, B.A. Griffith and H.W. Smith:  Quantitative Studies of the Relationship Between Alcohol Levels and Motor Vehicle Accidents, Proceedings of the Second International Conference on Alcohol and Road Traffic, op. cit., pp. 139–142 (1953).)

“It is dangerous, even for experienced drinkers, to drive when the blood-alcohol level lies between 0.10 and 0.15%.”  (F. Naville, R. Herrmann and R. DuPan:  L'Intoxication Alcoolique Aigue et Les Accidents d'Automobile (Acute Alcoholic Intoxication and Automobile Accidents);  III.   Les Effets de l'Alcool Chez Les Automobilists (Effects of Alcohol on Automobile Drivers), Schweiz Med Wschr 76:446 (1946).)

“Concentration of above 0.05% is considered dangerous for road safety.”  (R. Pinto Ribeiro:  O Diagnostico Medico-legal de Embriaguez Alcoolica (the Medicolegal Diagnosis of Alcohol Intoxication), Med Cirurg, Porto Alegre 12:15 (1950).)

“Evidence of blood-alcohol concentrations of 0.05% and higher, together with evidence of driving error, is sufficient to identify drivers who may be presumed to be affected by alcohol.”  (H.W. Smith and R.E. Popham:  Blood Alcohol Levels in Relation to Driving, Canad Med Ass J 65:325 (1951).)

“Altered driving pattern occurs at levels between 0.05 and 0.10%.”  (M.H. Thelin:  Modalites Cliniques et Recherches de Laboratoire dans les Cas d'Ivresse Comme Causes d'Accident de la Circulation, Acta Med Lig Social 1:387 (1948).)

None of these tests resulted in a report concluding a blood alcohol level of .10 percent did not have significant effect on every individual driver.

Thus there is a rational connection between a blood alcohol level of .10 percent and the ultimate fact, “being under the influence” as an impairment to an appreciable degree of the ability to operate a vehicle in a manner like that of an ordinary prudent and cautious person in full possession of his faculties, using reasonable care under like circumstances.   The presumption thus created by the Legislature has a rational connection and is constitutional.

 Shelton argues the jury was instructed the presumption was mandatory.   While the statutory language suggests the presumption is mandatory by its use of the word “shall,” and it expressly affects the burden of proof, as applied in this case it was a permissive presumption.⁷  The trial judge told the jury if it found beyond a reasonable doubt the blood alcohol level was .10 percent or higher, it “should ” find the defendant was under the influence.   “Should,” however, is not mandatory in the same sense as “shall” or “must” is and, in addition, that portion of the instruction was immediately followed by this language, “unless from all the evidence you have a reasonable doubt that the defendant was under the influence of an intoxicating beverage at the time of the alleged offense.”   The instruction given as it was does not shift the burden of proof or call for a mandatory finding of guilt after establishing the facts giving rise to the presumed fact.   Rather, it leaves with the jury the ultimate decision based on the total evidence whether the defendant is under the influence.   They were instructed they could have a reasonable doubt as to his being under the influence no matter what the blood alcohol level might be.

While “should” suggests some duty or obligation, it carries with it the meaning there is an expected or a probable result.  “Should” is generally interchangeable with the word “ought” and suggests an expectation, fitness, propriety or expediency, but “should” is often said to be milder than “ought.”  “Must” or “shall,” on the other hand, implies an obligation which is mandated with some determination (see Webster's Third Internat. Dictionary, “ought,” p. 1599 for synonyms;  see also Funk & Wagnall's Standard College Dictionary (1974) “should,” p. 1243;  American Heritage Dictionary, “should,” p. 1199).

The instructions given make it clear the prosecution has the burden of proof beyond a reasonable doubt the driver was under the influence of the alcoholic beverage and if a reasonable doubt exists, there can be no conviction.   It does not shift the burden of proof, but at most requires the defendant to go forward in the production of evidence.

Shelton argues two recent United States Supreme Court cases, County Court of Ulster Cty. v. Allen, 442 U.S. 140, 99 S.Ct. 2213, 2215, 60 L.Ed.2d 777, and Sandstrom v. Montana, 442 U.S. 510, 99 S.Ct. 2450, 61 L.Ed.2d 39, and one recent California Supreme Court case, People v. Roder, supra, 33 Cal.3d 491, 189 Cal.Rptr. 501, 658 P.2d 1302, call for a review of this area of the law.   We have examined these cases and find nothing requiring a reversal of case law in California upholding the validity of the rebuttable presumption as applied in this case.

County Court of Ulster Cty. v. Allen, supra, involved a permissive presumption created by the Legislature of New York that presence of a firearm in an automobile is presumptive evidence of its illegal possession by all persons then occupying the vehicle.

The opinion differentiates between the mandatory presumption and the permissive presumption.

“To the extent that the trier of fact is forced to abide by the presumption, and may not reject it based on an independent evaluation of the particular facts presented by the State, the analysis of the presumption's constitutional validity is logically divorced from those facts and based on the presumption's accuracy in the run of cases.  [Citations.]

“․

“The trial judge's instructions make it clear that the presumption was merely a part of the prosecution's case, that it gave rise to a permissive inference available only in certain circumstances, rather than a mandatory conclusion of possession, and that it could be ignored by the jury even if there was no affirmative proof offered by defendants in rebuttal.   The judge explained that possession could be actual or constructive, but that constructive possession could not exist without the intent and ability to exercise control or dominion over the weapons.   He also carefully instructed the jury that there is a mandatory presumption of innocence in favor of the defendants that controls unless it, as the exclusive trier of fact, is satisfied beyond a reasonable doubt that the defendants possessed the handguns in the manner described by the judge.   In short, the instructions plainly directed the jury to consider all the circumstances tending to support or contradict the inference that all four occupants of the car had possession of the two loaded handguns and to decide the matter for itself without regard to how much evidence the defendants introduced.”  (County Court of Ulster Cty. v. Allen, supra, 442 U.S. 140, 159–162, 99 S.Ct. 2213, 2226, 60 L.Ed.2d 777;  fns. omitted.)

Ulster County upheld the permissive presumption as it was applied by the court, and we find nothing in that opinion to require a different result here.

In Sandstrom v. Montana, supra, 442 U.S. 510, 99 S.Ct. 2450, 61 L.Ed.2d 39, the court was dealing with the presumption “the law presumes that a person intends the ordinary consequences of his voluntary acts,” shifting the burden of proof of the necessary intent elements in a murder case.   The court there held the presumption was mandatory and would be improper if a reasonable juror could believe the presumption was mandatory.   Also, since the jury may have interpreted the instruction as shifting the burden of persuasion (burden of proof), even if it did not consider the presumption to be conclusive, it was constitutionally infirm.   Viewed as mandatory the presumption, as it read, denied the defendant the constitutional presumption of innocence in the case where specific intent had to be shown by the prosecution (id. at p. 523, 99 S.Ct. at p. 2459).   In any event, under the language of the presumption, the court held shifting the burden of proof to the defendant thereby requiring him to show the element of intent did not exist “by the preponderance of the evidence” violated the United States Constitution.  Sandstrom teaches in each instance we must look at the way the presumption was presented to the jury, and as we have seen, the instruction as given here did not have the mandatory or burden of proof shifting effect condemned in Sandstrom.

Finally, we examine People v. Roder, supra, 33 Cal.3d 491, 189 Cal.Rptr. 501, 658 P.2d 1302.   This case involved a presumption that a second-hand dealer who buys or receives stolen property under circumstances that should have caused him to make inquiry of the person from whom it was acquired that he had a legal right to sell it, has the burden of showing he made a reasonable inquiry or be presumed to have bought or received knowing it was stolen.   This case reviewing Ulster and Sandstrom states the threshold inquiry must be to determine whether we are dealing with a mandatory or permissive presumption, and in so doing we must look at the way it was presented to the jury.   It held a reasonable jury could well believe it was dealing with a mandatory presumption and one which relieved the prosecution of its burden of proving every element of the offense beyond a reasonable doubt (id. at pp. 503, 504, 189 Cal.Rptr. 501, 658 P.2d 1302).

Section 23155 creates the presumption, and under the authorities we are called upon only to determine how the presumption was presented to the jury.   The question then is whether a reasonable juror could have understood it to be mandatory, rather than permissive (People v. Roder, supra, 33 Cal.3d 491, 502, 189 Cal.Rptr. 501, 658 P.2d 1302;  County Court of Ulster Cty. v. Allen, supra, 442 U.S. 140, 158, fn. 16, 99 S.Ct. 2213, 2226, fn. 16, 60 L.Ed.2d 777).   If the jury was told the blood alcohol test result of .10 percent creates a mandatory presumption, the jury might have been improperly instructed but our opinion need not reach that issue.   We conclude the jury was instructed in a manner it could only believe it was a permissive presumption.

Unlike Roder where the jury was instructed it “shall” presume the ultimate fact from the facts giving rise to the presumption, the jury here was instructed they “should” find.   As we pointed out above, “should” connotes something less than a mandatory direction.   While “should” does not give a clear statement the presumption is rebuttable, when it is followed by the statement “unless from all the evidence you have a reasonable doubt the defendant was under the influence” it is clear the prosecution has the burden of proof of all elements and the jury is not bound by the presumption.

People v. Roder, supra, describes the value of this phrase as applied to its fact situation.

“[T]he instruction given by the court told the jurors that if they found the basic facts ‘then you shall presume that defendant bought or received such property knowing it to have been stolen unless from all the evidence you have reasonable doubt that defendants knew the property was stolen.’  (Italics added.)   This concluding clause—not contained in the statute—could have been interpreted by the jurors to mean that even if they found the basic facts to exist, they were not required to find in accordance with the presumption and could refuse to do so if ‘from all the evidence’—that is, the evidence presented by either the prosecution or the defense (see People v. Henderson (1980) 109 Cal.App.3d 59, 62 [167 Cal.Rptr. 47] )—they had a reasonable doubt that defendant knew the property was stolen.   Thus, this language—taken by itself—could be read to suggest that the ‘presumption’ of guilty knowledge was simply a permissive inference, which the jurors could, but were not required to, draw.”  (People v. Roder, supra, at pp. 502–503, 189 Cal.Rptr. 501, 658 P.2d 1302;  emphasis added.)

The court went on to state, however, as applied in that case it failed to clear the confusion created by the use of the words “shall presume,” a mandatory direction.   The confusion was noted by the fact the jury returned for some further instructions on how they should treat the presumption, and the trial court spoke in terms of shifting the burden to the defendant.   Here, on the other hand, we do not have the use of the word “shall,” nor do we have evidence of any confusion in the minds of the jurors as to exactly what kind of a presumption we are dealing with.

We believe a reasonable jury could only construe this instruction as creating a permissive presumption and, as applied, it is constitutional.

 Shelton contends there was error in the giving of an instruction that the law requires motorcycles manufactured after January 1, 1978, to be equipped with headlamps which automatically come on when the engine is started.   The instruction dealing with Bonney's vehicle may have relevance in a civil action but has very little relevance in this criminal prosecution.   Any error occasioned in giving the instruction, however, was not prejudicial (People v. Watson, 46 Cal.2d 818, 836, 299 P.2d 243).   Similarly, out-of-court statements made by witnesses relative to Bonney's contributing to the causation of the accident were properly excluded as irrelevant under Evidence Code section 352, but the asserted error was, in any event, not prejudicial (ibid.)   It is Shelton's violation of the law that is in issue, not Bonney's, and Bonney's role in causation of the accident has only minor relevance.

Judgment affirmed.

I respectfully dissent.

The bony fingers of Death steer the car of the drinking driver.   Our highways are littered with bodies broken by the bottle.   Cemeteries bear mute testimony to the need to prohibit the drinker from driving.   Knowing all this, I cannot endorse the suggestion “the end justifies the means” in this case.   The drinking driver, though he be a “loose cannon on our highways,” is still entitled to fundamental due process when prosecuted.

Here, we do not address misdemeanor drunk driving.   We review a felony conviction with potential for commitment to prison.  (Veh.Code, § 23180.)   I thus consider this case in the context of criminal due process requirements.   The instruction given here creates a mandatory presumption Shelton was under the influence of alcohol at the time of the alleged offense based upon a chemical analysis of his blood taken nearly two hours after his arrest.   In effect, the instruction shifts the burden of persuasion to Shelton.   He must create a reasonable doubt he was under the influence of an intoxicating beverage when stopped by the police officer.   Careful constitutional analysis reveals the instruction violates due process principles;  the People bear the burden of proving all elements of the charged offense beyond a reasonable doubt.   Shelton's conviction must be set aside.

I

The criminal violations with which Shelton is charged arise from Vehicle Code section 23153, subdivision (a).   That statute provides in pertinent part as follows:

“It is unlawful for any person, while under the influence of an alcoholic beverage ․ to drive a vehicle and, when so driving, do any act forbidden by law or neglect any duty imposed by law in the driving of the vehicle, which act or neglect proximately causes death or bodily injury to any person other than the driver.”  (Italics added.)

To prove Shelton was under the influence of an alcoholic beverage at the time of the alleged offense, the prosecution introduced the results of a blood alcohol test administered some two hours later.   In order to prove Shelton was under the influence of an alcoholic beverage when stopped and bridge the gap in the evidence created by the significant lapse of time between his arrest and the administration of the test, the prosecution relied on the presumptions created by Vehicle Code section 23155, subdivision (a) ¹ and Evidence Code section 607.²

As the majority points out, earlier California cases have held similar presumptions (under Vehicle Code section 23126, subdivision (a), the predecessor to section 23155) “neither unfair nor unconstitutional.”  (See People v. Lachman (1972) 23 Cal.App.3d 1094, 1097, 100 Cal.Rptr. 710;  DeWoody v. Superior Court (1970) 8 Cal.App.3d 52, 87 Cal.Rptr. 210;  People v. Schrieber (1975) 45 Cal.App.3d 917.)   However, these cases rely on the outdated United States Supreme Court decision in Leary v. United States (1969) 395 U.S. 6, 119 Cal.Rptr. 812, which states a statutory presumption does not infringe upon due process of law if “it can at least be said with substantial assurance that the presumed fact is more likely than not to flow from the proved fact on which it made to depend.”  (Id., at p. 36, 119 Cal.Rptr. 812.)   A more recent Supreme Court case has restated the standard by which the constitutionality of statutory presumption should be examined.   This new standard requires (1) a rational connection between the basic fact and the elemental fact to be proved, and (2) an inference so strong the elemental fact is proved, beyond a reasonable doubt, by proof of the basic fact.  (Turner v. United States (1970) 396 U.S. 398, 416, 90 S.Ct. 642, 652, 24 L.Ed.2d 610.)

Based on Turner, recent United States Supreme Court decisions (Ulster County Court v. Allen (1979) 442 U.S. 140, 99 S.Ct. 2213, 60 L.Ed.2d 777;  Sandstrom v. Montana (1979) 442 U.S. 510, 99 S.Ct. 2450, 61 L.Ed.2d 39) and a California Supreme Court case (People v. Roder (1983) 33 Cal.3d 491, 189 Cal.Rptr. 501, 658 P.2d 1302) have created a new framework in which courts should analyze the constitutionality of statutory presumptions.   The presumption in this case must be reexamined in light of these most recent high court decisions.³

II

The first step in the determination of the constitutionality of a statutory presumption is to determine the type of presumption created.  (Sandstrom v. Montana, supra, 442 U.S. at p. 514, 99 S.Ct. at p. 2454.)   The Supreme Court in Ulster County Court v. Allen, supra, 442 U.S. 140, 170, 99 S.Ct. 2213, 2231, 60 L.Ed.2d 777, outlined the various categories of statutory presumptions.  (People v. Roder, supra, 33 Cal.App.3d 491, 500, fn. 8, 189 Cal.Rptr. 501, 658 P.2d 1302.)   The primary categories identified by the court as the “touchstone of any constitutional analysis” are (1) “permissive inferences” and (2) “mandatory presumptions.”

A permissive inference allows, but does not require, the trier of fact to infer the elemental fact from proof of the basic fact and places no burden of any kind on the defendant.   A permissive inference leaves the trier of fact free to credit or reject the inference and does not shift the burden of proof.  (Ulster County Court v. Allen, supra, 442 U.S. at p. 157, 99 S.Ct. at p. 2225.)

The mandatory presumption is a stronger evidentiary device, however.

“[A mandatory presumption] may affect not only the strength of the ‘no reasonable doubt’ burden but also the placement of that burden;  it tells the trier that he or they must find the elemental fact upon proof of the basic fact, at least unless the defendant has come forward with some evidence to rebut the presumed connection between the two facts.”  (Ulster County Court v. Allen, supra, 442 U.S. 140 at p. 157, 99 S.Ct. 2213 at p. 2224, 60 L.Ed.2d 777.)

The mandatory presumption commonly runs aground on the rock of constitutional principles.   However, even mandatory presumptions may be constitutional in some situations.  Ulster County further subdivided mandatory presumptions into two categories:  (1) those that merely shift the burden of production to the defendant, following the satisfaction of which the ultimate burden of persuasion returns to the prosecution;  and (2) those that entirely shift the burden of proof to the defendant.  (Id., at pp. 157–158, fn. 16, 99 S.Ct. at p. 2225, fn. 16.)

A presumption affecting the burden of production merely fixes the risk of nonproduction on the defendant.   If the defendant produces any evidence tending to show the nonexistence of the presumed fact, the presumption disappears—the determination of that issue is made without the aid of the presumption.   A presumption which falls into the first category is not per se unconstitutional.

“To the extent that a presumption imposes an extremely low burden of production—e.g., being satisfied by ‘any’ evidence—it may well be that its impact is no greater than that of a permissive inference, and it may be proper to analyze it as such.  [Citation.]”  (Ulster County Court v. Allen, supra, 442 U.S. at p. 158, fn. 16, 99 S.Ct. at p. 2226, fn. 16.)

The second category of mandatory presumptions is that which affects the burden of proof.   Such a presumption fixes the burden of persuasion on the defendant.   It is not extinguished merely by the production of any evidence showing the nonexistence of the presumed fact but, rather, remains until the defendant persuades the jury to a specified requisite degree.   This type of presumption may be violative of constitutional principles.

“[S]ince the prosecution bears the burden of establishing guilt, it may not rest its case on a [mandatory] presumption, unless the fact proved is sufficient to support the inference of guilt beyond a reasonable doubt.”   (Ulster County Court v. Allen, supra, 442 U.S. 140, 167, 99 S.Ct. 2213, 2229, 60 L.Ed.2d 777.)

III

In the instant case, the jury was instructed as follows:

“If the evidence establishes beyond a reasonable doubt that at the time of the chemical analysis of the defendant's blood, breath or urine, there was 0.10 percent or more by weight of alcohol in the defendant's blood, you should find that the defendant was under the influence of an alcoholic beverage at the time of the alleged offense, unless from all of the evidence you have a reasonable doubt that the defendant was under the influence of an intoxicating beverage at the time of the alleged offense.”

This instruction (CALJIC No. 12.61 (1982 Revision)) embodies the statutory presumptions provided in Vehicle Code section 23155 and Evidence Code section 607.⁴  It is a classic example of the mandatory presumption described in Ulster County (442 U.S. at p. 155, 99 S.Ct. at p. 2223).   The instruction tells the jury they “should” find Shelton under the influence of an alcoholic beverage at the time of the alleged offense (the elemental fact), if they find beyond a reasonable doubt a blood alcohol level of 0.10 percent or more at the time of the test (the basic fact).   Contrary to the majority's assertion, “should” is a mandatory word which requires the presumption of the elemental fact from proof of the basic fact.  (See People v. Roder, supra, 33 Cal.3d 491, 506, fn. 15, 189 Cal.Rptr. 501, 658 P.2d 1302.) ⁵

The presumption described in this instruction clearly does not fall into the first category of mandatory presumptions discussed in Ulster County.   The statutes creating the presumption expressly provide that it shall affect the burden of proof.   The wording of the instruction imposes upon the defendant a burden to persuade the jury “to a reasonable doubt” he was not under the influence of alcohol at the time of the alleged offense.   Such a requirement is more than the “extremely low burden of producing any evidence” contemplated by the court in Ulster County.  (Id., at p. 158, fn. 16, 99 S.Ct. at p. 2226, fn. 16.)

The majority asserts the presumption merely “requires the defendant to go forward in the production of evidence.”   Noting the instruction includes the proscription “unless from all the evidence you have a reasonable doubt that the defendant was under the influence of an intoxicating beverage at the time of the alleged offense,” the majority reasons the presumption described does not shift the burden of proof or call for a mandatory finding of guilt based on the establishment of the basic fact.   The holding suggests the inclusion of the above-quoted language in the instruction cures any possible constitutional infirmity.   I disagree.

A presumption may not place upon a criminal defendant the burden of proving the nonexistence of the presumed fact by raising a reasonable doubt as to its existence.   Under our system of justice, a defendant is presumed innocent until proven guilty.   The prosecution must exclude any reasonable doubt as to his innocence;  defendant has absolutely no burden to produce any evidence.   As has been stated time and again, the defendant may literally “sit on his hands” and put the prosecution to its proof.  (See Coffin v. U.S. (1894) 156 U.S. 432, 15 S.Ct. 394, 39 L.Ed. 481;  United States v. Gainey (1965) 380 U.S. 63, 85 S.Ct. 754, 13 L.Ed.2d 658.)   Justice Jefferson, in his California Evidence Benchbook (2d ed.) section 46.4, pages 1714–1715, summarized the law as follows:

“The beyond-a-reasonable-doubt concept that is mandated to be applied in presumptions in criminal cases by Ulster County Court v. Allen, supra, would seem to render Evidence Code section 607, as literally read, a violation of due process of law ․ whenever proof of an element of the offense depends on the use of a presumption.   In such a case, the inference from the basic facts to the presumed fact must follow beyond a reasonable doubt.   This necessarily casts the burden of proof of such fact beyond a reasonable doubt upon the prosecution.   To place the burden on the defendant to raise a reasonable doubt would thus violate the Ulster County Court mandate.”

(See also People v. Burres (1980) 101 Cal.App.3d 341, 352–353, fn. 3, 161 Cal.Rptr. 593.)

IV

The majority notes there is a “real and rational connection between the blood alcohol level [of defendant] and being under the influence” and, on that basis, concludes the presumption ․ created by the Legislature ․ is constitutional.”   In so holding, the majority is apparently judging the constitutionality of the presumption in this case by the “rational connection” test first enunciated in the civil context by Mobile, J. & K.C.R. Co. v. Turnipseed (1910) 219 U.S. 35, 31 S.Ct. 136, 55 L.Ed. 78, and applied to a criminal prosecution in Tot v. United States (1943) 319 U.S. 463, 63 S.Ct. 1241, 87 L.Ed. 1519.

Both Turnipseed and Tot held due process requires a rational connection between the basic fact and the presumed fact before a presumption may be used to establish an essential element of the case.  (Tot v. United States, supra, 319 U.S. at p. 467, 63 S.Ct. at p. 1244.)   However, neither case explained how strong the rational connection must be.

In Leary v. United States, supra, 395 U.S. 6, 89 S.Ct. 1532, 23 L.Ed.2d 57, the court attempted to answer this question by stating a presumption is unconstitutional “unless it can at least be said with substantial assurance that the presumed fact is more likely than not to flow from the proved fact on which it is made to depend.”  (Id., at p. 36, 89 S.Ct. at p. 1548, italics added.)   However, it was not until Turner v. United States, supra, 396 U.S. 398, 90 S.Ct. 642, 24 L.Ed.2d 610, that something more definite than the minimum standard enunciated in Leary was provided.   Distinguishing Leary, the Turner court suggested the rational connection between the basic and the presumed facts should be judged by the “more exacting reasonable-doubt standard normally applicable in criminal cases.”  (Turner, supra, at p. 416, 90 S.Ct. at p. 652.)   The “overwhelming evidence” cited by the court in Turner created a new standard by which courts must judge the constitutionality of statutory presumptions in criminal cases.  (Murray & Aitken, “The Constitutionality of California's Under-the-Influence-of-Alcohol Presumption” (1972) 45 So.Cal.L.Rev. 955, 978.)   Under Turner, not only must there be a rational connection between the facts, but also an inference so strong that the presumed fact is proved, beyond a reasonable doubt, by proof of the basic fact.  (Ulster County Court v. Allen, supra, 442 U.S. 140, 167, 99 S.Ct. 2213, 2229, 60 L.Ed.2d 777.)   As interpreted by the California Supreme Court, due process principles allow the use of a mandatory presumption in criminal cases “only if the basic fact proved compels the inference of guilt beyond a reasonable doubt.”  (People v. Roder, supra, 33 Cal.3d 491, 499, fn. 7, 189 Cal.Rptr. 501, 658 P.2d 1302.   Italics added.)

In the past, courts have side-stepped the constitutional problems created by the use of the statutory presumptions in drunk driving cases by finding a “rational connection” between the defendant's test results and his intoxication.   These courts have further noted their “substantial assurance” of defendant's guilt to satisfy the requirements of Leary.   Assuming the defendant's blood alcohol level had peaked before arrest, the courts extrapolated to a higher result at the time of the offense by adding an amount equal to the standard hourly alcohol burn-off rate.⁶  Due process questions were treated perfunctorily by presuming “the greater the elapse of time, the greater the benefit to the defendant.”  (People v. Schrieber, supra, 45 Cal.App.3d 917, 119 Cal.Rptr. 812;  In re Martin (1962) 58 Cal.2d 509, 512, 24 Cal.Rptr. 833, 374 P.2d 801.)   The discrepancy in the prosecution's proof was overlooked because any error in the computations of defendant's intoxication caused by the practical limitations of timely administering the test was assumed to be in favor of the defendant.

Several recent studies have cast doubt on this assumption, however.⁷  Apparently, the scientific community now questions the accuracy of the methods used to determine blood alcohol level for purposes of trial.   Several of these studies go so far as to conclude it is impossible to prove beyond a reasonable doubt that a person was “under the influence” while driving his vehicle based solely on the results of a subsequently administered test.   These articles cite recent data which indicates blood alcohol can actually rise and fall several times during the body's consumption process;  this makes it impossible (under present testing methods) to determine, with any certainty, defendant's blood alcohol level during the post peak (i.e., “burn-off”) period.   As stated by the AMA Committee on Medicolegal Problems in Alcohol and the Impaired Driver (1970):

“While the value ․ determined [by extrapolation] is usually higher than that in the blood sample analyzed, this may not be true if the elapsed time period occurred during a rising phase of the blood alcohol curve.”   (Id., at pp. 22.)

In the instant case, nearly two hours elapsed between the alleged offense and the administration of the blood alcohol test.   It is therefore likely the test was administered during the volatile burn-off period.   Interestingly, the prosecution's forensic toxicologist only testified as to Shelton's blood alcohol level at the time of the test;  he was not asked to extrapolate the results to the time of the offense.   When the court inquired about the gap in the evidence, the prosecutor explained:  “I didn't cover it, Your Honor, because I don't want him to extrapolate to a lower number.”  (Italics added.)

V

Ulster County, Sandstrom and Roder, mandate the conclusion the instruction given in this case violates basic due process protections.   Proof beyond a reasonable doubt of blood alcohol test results in excess of 0.10 percent does not compel the conclusion the defendant was under the influence of alcohol at the time of the offense.   The instruction therefore impermissibly shifts the burden of proof of an element of the defense, driving a vehicle under the influence of alcohol, from the prosecution to the defendant.   Since the prosecution bears the burden of proving all elements of the offense beyond a reasonable doubt, giving the instruction in this case was error of constitutional dimension.

Instructional errors of constitutional dimensions have been held to require automatic reversal unless shown to be harmless beyond a reasonable doubt.   (Chapman v. California (1967) 386 U.S. 18, 24, 87 S.Ct. 824, 828, 17 L.Ed.2d 705;  People v. Bedolla (1979) 94 Cal.App.3d 1, 6, 156 Cal.Rptr. 171.)   Here, it cannot be said the error was harmless.   The results of Shelton's blood alcohol are, by their very nature, material evidence of his guilt or innocence on the drunk driving charge.  (See People v. Hitch (1974) 12 Cal.3d 641, 647, 117 Cal.Rptr. 9, 527 P.2d 361.)  “The value of such objective scientific data of intoxication to supplement the fallible observations of humans of behavior seemingly symptomatic of intoxication cannot be disputed.”   (People v. Sudduth (1966) 65 Cal.2d 543, 546, 55 Cal.Rptr. 393, 421 P.2d 401.   Italics added.)

I would reverse the judgment of conviction.

FOOTNOTES

1.   All statutory references are to the Vehicle Code unless otherwise specified.Section 23153, subdivision (a), reads:  “It is unlawful for any person, while under the influence of an alcoholic beverage or any drug, or under the combined influence of an alcoholic beverage and any drug, to drive a vehicle and, when so driving, do any act forbidden by law or neglect any duty imposed by law in the driving of the vehicle, which act or neglect proximately causes death or bodily injury to any person other than the driver.”

2.   The use note to CALJIC No. 12.61 (1982 Revision) states it incorporates statutory changes effected by the 1982 amendment of section 23155 effective as an urgency statute February 18, 1982, and is to be used where the offense occurred after February 18, 1982.

3.   Report of the California Highway Patrol for 1977 indicates of the 6,457 accidents in California involving a fatality, 2,183 (one-third) were alcohol-related, and of that number 1,535 drivers were known to have had blood alcohol levels of .10 percent or greater (ibid. p. 69).See also California Highway Patrol Final Report, DUI Reduction, Month Project, December 1980.

4.   The infallibility of certain of the instruments used in the testing procedure is not above question.   That, however, is a proper item of concern which should be addressed when the People attempt to prove the .10 percent level.   Since proof of the alcohol level must be beyond a reasonable doubt, the errors that occur in testing are a proper matter for the defense to argue.

5.   “Acute intoxication as defined in this section refers to impairment of normal function, changes in appearance and alterations in behavior which are the unmistakable signs of intoxication.   It does not refer to subclinical intoxication where more precise measurements must be used in order to identify the state.   The condition as defined here, then, is one of gross impairment and is synonymous with the lay term ‘drunk’.”   (Emphasis added.)

6.  L1-11TABLE 3: RELATION BETWEEN BLOOD-ALCOHOL LEVEL AND DRUNKENNESSC3-8Percent of Persons Diagnosed as DrunkS8Total0.000.0510.1010.1510.2010.2510.3010.351Personsto 0.05to 0.10to 0.15to 0.20to 0.25to 0.30to 0.35to 0.400.401Examined Investigator019508393 981001001942Widmark 238939799100-100950Schwartz 1018478390 9596931001000Jetter 10688192971001712Andersen 0465092100100100100100140Harger 014699094 94100100100100Prag 725498593 979810099750Hine 432628995 9899991006594 FN* Percent weight by volume (0.05 percent w/v=50 mg/100 ml,0.15 percent w/v=150 mg/100ml) of Blood Alcohol

7.   The CALJIC Committee amended CALJIC No. 12.61 (1982 Revision) to allow the section 23155, subdivision (a)(3), presumption to be treated as a permissive inference responsive to People v. Roder, 33 Cal.3d 491, 189 Cal.Rptr. 501, 658 P.2d 1302.  (CALJIC No. 12.61, Supp. Service, pamp. No. 1 (June 1983), Use Note.)

1.   Vehicle Code section 23155, subdivision (a) provides:  “Upon the trial of any criminal action ․ arising out of acts alleged to have been committed by any person while driving a vehicle while under the influence of an alcoholic beverage in violation of ․ subdivision (a) of Section 23153, the amount of alcohol in the person's blood at the time of the test as shown by chemical analysis of that person's blood ․ shall give rise to the following presumptions affecting the burden of proof ․ (3) If there was at that time 0.10 percent or more by weight of alcohol in the person's blood, it shall be presumed that the person was under the influence of an alcoholic beverage at the time of the alleged offense.”   (Italics added.)

2.   Evidence Code section 607 provides:  “When a presumption affecting the burden of proof operates in a criminal action to establish presumptively any fact that is essential to the defendant's guilt, the presumption operates only if the facts that give rise to the presumption have been found or otherwise establish beyond a reasonable doubt, and, in such case, the defendant need only raise a reasonable doubt as to the existence of the presumed fact.”

3.   Burg v. Superior Court (1983) 35 Cal.3d 257, 198 Cal.Rptr. 145, 673 P.2d 732 (1983) upholds the constitutionality of Vehicle Code section 23152, subdivision (b).   The decision is not dispositive of the issues before this court.Vehicle Code section 23152, subdivisions (a) and (b) describe wholly separate offenses.   Although analogous for some purposes, the two subdivisions of the statutes differ substantially in the context of this analysis.  (See Burg v. Superior Court, supra, at p. ––––, at p. 148 of 198 Cal.Rptr., at p. 736 of 673 P.2d.)

4.   See footnotes 1 and 2.

5.   The CALJIC Committee amended CALJIC No. 12.61 in June 1983 to provide a permissible inference.   The phrase “should find” is now optional and replaceable by “may but are not required to infer.”   The Committee made this change believing it to be necessary to comply with the constitutional principles enunciated in People v. Roder.  (See CALJIC No. 12.61.1, Supp. Service, Pamphlet No. 1 (June 1983) Use Note.)

6.   In People v. Flannel (1979) 25 Cal.3d 668, 160 Cal.Rptr. 84, 603 P.2d 1, Chief Justice Bird suggested an accurate determination of a defendant's blood alcohol content at the time of the offense requires adding to defendant's test results an amount equal to the standard alcohol hourly burn-off rate (0.018) multiplied by the time elapsed between the alleged offense and the test.  (Bird, C.J., concurring and dissenting at p. 689, 160 Cal.Rptr. 84, 603 P.2d 1.)

7.   Donnellan, “Driving With 0.10% Blood Alcohol:  Can The State Prove It?” (1982) 16 USFL Rev. 817, 818;  Radlow and Hurst, “Delayed Blood Alcohol Determinations in Forensic Applications” (1979) 2 Crim.Just. J 281;  Fitzgerald and Hume, “The Single Chemical Test for Intoxication, A Challenge to Admissibility” (1981) 66 Mass.L.Rev. 23;  Taylor, “Blood-Alcohol Presumptions:  Guilty Until Proven Innocent” (1978) 53 Cal. SBJ 170, 175.

COLOGNE, Associate Justice.

BROWN, P.J., concurs.