dissenting:
I respectfully dissent from the majority opinion for the following reasons:
A. It misconstrues the concededly applicable Pennsylvania law governing the admissibility of evidence of intoxication and reviews the district court’s rejection of the defendants’ offer of proof of the blood alcohol content of plaintiff’s decedent for an abuse of discretion, rather than for an error of law.
B. Its statement that “we are required by our decision in Greiner v. Volkswagen werk Aktiengeselleschaft, 540 F.2d 85, 89 (3d Cir. 1976), to apply Fisher v. Dye in this appeal...” is not necessary to a decision in this case and raises difficult and substantial questions which were never presented to the district court.
C. Its references to the Federal Rules of Evidence are similarly unnecessary and raise the same difficult questions concerning whether those rules or state law would control had that issue been raised in the district court. Because the district court never considered or was asked to consider such Federal Rules, the majority’s implication that their application would have the same result as that reached in this case is simply speculation.
*332I.
This tragic collision occurred in the westbound lanes of the Schuylkill Expressway near Conshohocken, Montgomery County, Pennsylvania. The evidence showed that the roadway consisted of two lanes of travel in each direction with a steel guardrail separating the eastbound and westbound traffic. There is an uphill grade from east to west. For approximately two miles before the accident, the Geppert vehicle,1 a combination tractor-trailer, had been following the Rovegno vehicle, a dump truck, in the right-hand westbound lane. As the trucks began to move up the incline, the Rovegno truck slowed and, as the Geppert truck attempted to pass, the trucks collided with such force that both drivers lost control, the Geppert truck hit the guardrail and jacknifed, the trucks collided again, skidded across the right-hand lane, and overturned against the roadside embankment. Marshall Rovegno was killed in the accident.
The only eyewitness to the actual collision was the defendant Ward, who testified:
“Q Why don’t you describe the point you got on the Expressway to the point of the accident?
“A I came on the Expressway and traffic was normal, you know. I came on up until I got like, say, two miles from Conshohocken. It is a long grade there. We were coming up the grade and then it levels off. It was a truck right in front of me. All of a sudden, he started to slow down for no apparent reason. I started out into the left lane to go around him, because he was stopping so quick, you know. As I started past he just started to drift right over in my lane, so I tried to stop to get back around and, at that time, the collision.
“Q The impact between your truck and the other truck, where did that occur, what part of your truck and what part of his truck?
“A My right front fender and the left rear of his truck, about the wheel, I would say.
“Q At some point in the accident, did your right front wheel come off?
“A Yes.
“Q Where was your tractor located at the time of the impact?
“A In the left lane.”
N.T. (4/10/80) at 105-06, reprinted at 90a-91a (emphasis added). This account is consistent with the testimony given by the investigating state trooper whose analysis of the skidmarks at the scene indicated to him that the initial point of impact was in the passing lane. N.T. (4/9/80) at 33, reprinted at 17a.
Another truckdriver, Patrick Kinney, was following the Geppert vehicle. He contradicted Ward on the timing of the collision, stating that the Geppert truck had not yet finished changing lanes when the collision occurred. N.T. (4/10/80) at 87, reprinted at 72a. Because he was screened by the trailer of the Geppert vehicle, however, Kinney could not testify concerning the point of impact or the behavior of the Rovegno truck. Id. at 87, 89, reprinted at 72a, 74a.
Clearly, the plaintiff’s theory was that Ward had been negligent in beginning his passing maneuver too late, in overtaking Rovegno before he was completely in the passing lane, and in causing the collision. The defendants countered that Rovegno had contributed to the cause of the accident by negligently permitting his truck to drift over into the passing lane in front of the overtaking Geppert vehicle.2
Fundamental to the defendants’ argument on this point was their contention that *333Marshall Rovegno was intoxicated at the time of the accident.3 When the plaintiff’s counsel objected, the defendant offered to prove that a blood sample taken from the body of Mr. Rovegno disclosed a blood alcohol level of 0.158%. See Stipulation of Counsel at 112a-113a. Under Pennsylvania law, a blood alcohol level in excess of 0.10% raises a presumption of intoxication.4 Pennsylvania law also provides that driving while intoxicated is a criminal — rather than simply a traffic — offense.5 The defendant also proffered expert testimony on the effect of such a blood alcohol level:
“Dr. Aronson will testify, in his opinion, a person with a blood alcohol level of the level shown in Mr. Rovegno is unfit to drive, the effects of alcohol on judgments, coordination, and such, make him unfit to drive
N.T. (4/10/80) at 95, reprinted at 80a. The plaintiff’s counsel argued that, under Pennsylvania law, evidence of intoxication is inadmissible absent evidence of reckless, erratic or wild driving and that the facts of this case did not rise to that level.6 Based on this argument, the district court refused to admit the testimony.
The jury found both the plaintiff’s decedent and the defendant Ward to have been negligent and assessed the relative fault at 30% and 70%, respectively.7 In denying the *334defendants’ motion for a new trial, the district court stated:
“The Supreme Court of Pennsylvania has consistently held that proof of intoxication is only relevant where reckless or careless driving of an automobile is a matter at issue. Billow v. [Farmers] Trust Co., 438 Pa. 514, [266 A.2d 92] (1970). See also Miles v. Ryan, 484 F.2d 1255 (3d Cir. 1973). Defendants’ reliance on Greiner v. Volkswagenwerk, 540 F.2d 85 (3d Cir. 1976) is misplaced. In Greiner there was evidence of reckless driving sufficient to permit the trial judge to submit the alcohol issue to the jury. In the case sub judice there was no such ‘other evidence’ as excessive speed, recklessness or erratic driving on the part of Marshall Rovegno. Therefore submission to the jury of testimony as to intoxication would surely have been unfairly prejudicial. Defendants’ motion for a new trial on this ground is therefore denied.”
Rovegno v. Geppert Bros., Inc., et al., No. 79-4202, slip op. at 3 (E.D.Pa. July 11, 1980), reprinted at 193a. A subsequent trial on damages8 resulted in a verdict for the plaintiff in the amount of $191,363.00 in the wrongful death action and $18,630.00 in the survival action.9
This appeal followed.
II.
A. The Law of Pennsylvania
I believe that the district court erred as a matter of law in holding the evidence of intoxication inadmissible absent a showing of wild, erratic or reckless driving by the driver against whom the evidence is offered.10
I. The basic rule of Pennsylvania law
There is no dispute that the applicable rule of Pennsylvania law is that stated in Fisher v. Dye, 386 Pa. 141, 125 A.2d 472 (1956):
“[W]here reckless or careless driving of an automobile is the matter at issue, the mere fact of drinking intoxicating liquor is not admissible, being unfairly prejudicial, unless it reasonably establishes a degree of intoxication which proves unfitness to drive
Id. at 148, 125 A.2d at 476 (citations omitted).11 In my view, the first clause of this test states the principle of logical relevance: evidence of a driver’s intoxication is not admissible unless the reckless or careless driving of that person is a matter at issue.12 The remaining language restates the familiar rule that otherwise relevant evidence is *335inadmissible when its probative value is outweighed by its potential for prejudice, and applies that rule to this specific evidence: evidence of this kind is unfairly prejudicial “unless it reasonably establishes a degree of intoxication which proves unfitness to drive.” Greiner v. Volkswagenwerk Aktiengeselleschaft, 540 F.2d at 89; Billow v. Farmers Trust Co., 438 Pa. 514, 517, 266 A.2d 92, 93 (1970); Cusatis v. Reichert, 267 Pa.Super.Ct. 247, 249-50, 406 A.2d 787, 788-89 (1979).
In this case, the first part of the test is clearly satisfied. As noted above, the real crux of the dispute is the extent to which Rovegno contributed to the causation of the accident, thus making his “reckless or careless driving of an automobile . .. the matter at issue.” Fisher v. Dye, 386 Pa. at 148, 125 A.2d at 476.13 Thus, we reach the second part of the test, which in turn requires us to answer two questions. First, what means are available for “reasonably establishing] a degree of intoxication which proves unfitness to drive”? Second, was that degree of intoxication established here? The latter question is, as the majority correctly observes, a matter within the discretion of the district court which we will disturb only in limited circumstances. The former question, however, is a pure legal issue on which, as the majority correctly observes, our scope of review is plenary. Majority typescript op. at 4, citing Universal Minerals, Inc. v. C. A. Hughes & Co., 669 F.2d 98, 103 (3d Cir. 1981).14
I cannot agree that this language requires, as a matter of law, that “unfitness to drive” be proved through evidence of actual driving behavior. Neither, I suggest, would our late colleague Judge Hastie, who observed over 15 years ago that:
“Since the evidence establishes the decedent’s contributory negligence as a matter of law, judgment must be entered for the railroad. Thus, it is unnecessary to decide whether the trial court erred in excluding post mortem evidence of the alcoholic content of the decedent’s blood as evidence of intoxication. However, we do observe that the district court seems to have been persuaded that under Pennsylvania law eyewitness testimony of conduct and condition at the time in question is the only way in which intoxication may be proved in a civil case. We have examined the cases but have not been able to discover or infer such a rule. The decision below should not be relied upon as authority for any so comprehensive a limitation on the method of proving intoxication.”
Parker v. Reading Company, 363 F.2d 608, 610 (3d Cir. 1970) (dictum). See also Greiner v. Volkswagenwerk Aktiengeselleschaft, 540 F.2d at 89, quoting Commonwealth v. Cave, 219 Pa.Super.Ct. 512, 515, 281 A.2d 733, 734 (1971).15
*336The problem appears to be in applying the rule developed in cases concerning evidence of drinking to cases such as this involving scientific evidence of blood alcohol level.
2. Application of Pennsylvania law to evidence of blood alcohol level
As I see it, the source of confusion here is twofold. First, the language of the Fisher v. Dye test, which is repeated verbatim in the'subsequent cases, begins with the language “reckless or careless driving.” The evidence recited above, including that of the investigating state trooper, clearly put “careless driving ... in issue.” However, as noted above, the Fisher court used this language as part of its relevance test and not as part of its unfair prejudice test. Thus, Fisher cannot be said to require evidence of erratic driving in order to “reasonably establish a degree of intoxication which proves unfitness to drive.”
Second, it is quite easy, especially given the brevity of some of the opinions, to confuse the facts of the earlier Pennsylvania cases with the rule of law upon which they rely. In those cases, the evidence which was sought to be introduced was evidence that the driver in question had been drinking at some time prior to the accident. From that, the jury would be asked to infer both that the driver was intoxicated and that the intoxication was of a degree rendering him unfit to drive at the time of the accident. Morreale v. Prince, 436 Pa. 51, 52, 258 A.2d 508, 508 (1969); Fisher v. Dye, 386 Pa. at 148, 125 A.2d at 476; Cusatis v. Reichert, 267 Pa.Super.Ct. at 249, 406 A.2d at 788-90; Sentz v. Dixon, 224 Pa.Super.Ct. 70, 71-72; 302 A.2d 434, 435-37 (1973). See also Greiner v. Volkswagenwerk Aktiengeselleschaft, 540 F.2d 85, 88-89 (3d Cir. 1976). Cf. Miles v. Ryan, 484 F.2d 1255, 1257-58 (3d Cir. 1973) (admissibility of evidence of drinking by witness to impeach ability to observe and recollect). Manifestly, the most readily available evidence that the drinking rendered the driver unfit to drive is the fact that he was observed driving in a wild or reckless manner. That is the most commonly encountered factual scenario and, consequently, is characteristic of the majority of the decided cases. Because it is the most common, however, does not mean that erratic driving is the only permissible foundation for the admission of intoxication evidence. Parker v. Reading Company, 363 F.2d at 610 (dictum). This is notably so, in my view, where that evidence is of a higher quality than the “mere drinking” which concerned the Fisher court and requires less of an inferential leap for the jury to find unfitness to drive.
The reason why evidence of the driver’s physical behavior is the most common foundation evidence in these cases is because it is generally all that is available. Evidence of blood alcohol level, on the other hand, is the exception because it is generally available in only two situations: (1) where the driver’s behavior at the scene is such as to give the investigating officer cause to believe he is intoxicated so that he administers a breath test or arranges for a blood test; or (2) where, as here, the driver is killed in the accident and the blood alcohol level is determined during an autopsy. It is the former situation which has recently caused some degree of confusion in the analysis employed by Pennsylvania cases because of multiple types of proof and complicating factors, such as both evidence of drinking and elevated blood alcohol level, Cusatis v. Reichert, 267 Pa.Super.Ct. at 249, 406 A.2d at 789-90; evidence of drinking and erratic behavior and elevated blood alcohol, Couts v. Ghion, 281 Pa.Super.Ct. 135, 142-45, 421 A.2d 1184, 1188 (1980); problems in hypothesizing blood alcohol level at the time of the accident from the results of *337a test administered sometime later, id. at 145-46, 421 A.2d at 1190; Schwarzbach v. Dunn, 252 Pa.Super.Ct. 454, 461-62, 381 A.2d 1295, 1298-99 (1977); or evidence of blood alcohol level coupled with expert testimony that it would “affect” driving, Billow v. Farmers Trust Co., 438 Pa. at 517, 266 A.2d at 93.
This case, however, presents none of these complications. First, there is clear evidence of intoxication: the parties agree that the decedent’s blood alcohol level at the time of the accident was 0.158%. Under Pennsylvania law, a blood alcohol level in excess of 0.10% raises a presumption of legal intoxication and unfitness to drive.16 On the face of it, this objective evidence is of a higher quality, and thus presents less of a possibility of unfair prejudice, than subjective testimony of drinking, drunken behavior, or erratic driving. In addition, there are none of the problems encountered by the courts in Couts and Schwarzbach in relating the results of a later blood alcohol test back to blood alcohol content at the time of the accident — Mr. Rovegno’s unfortunate death has fixed precisely his blood alcohol level at the time of the accident. Finally, the blood alcohol level is supplemented by expert testimony that this level of alcohol constitutes a degree of intoxication rendering the decedent unfit to drive. This was the crucial point in Billow17 and complies precisely with the explicit requirement of Fisher.
3. Conclusion
I believe that we need look no further than Billow to conclude that the district court erred. Even reading that opinion to require something more than proof of a blood alcohol level in excess of the legal limit, to hold that this “something more” must be evidence of erratic driving or drunken behavior is to ignore the plain language. The Billow court stated:
“Crucial to appellants’ offer of proof was the testimony of a Dr. Robert McConaghie. . . .
“Dr. McConaghie’s opinion that a person with a blood alcohol content of .14 would be ‘affected’ in his driving falls short of1 the requirement that the evidence show ‘a degree of intoxication which proves unfitness to drive.’ ”
438 Pa. at 517, 266 A.2d at 93. In this case, the “requirement that the evidence show ‘a degree of intoxication which proves unfitness to drive’ ” has been explicitly and literally satisfied.
Further, I believe that the clear trend in Pennsylvania law is to recognize the reliability of scientific blood alcohol tests and to give effect to the statutory presumption of unfitness to drive with a blood alcohol level in excess of 0.10%.18 This is especially so in light of growing public outrage over alcohol related traffic deaths.19
B. Possible Applicability of Federal Law
Nowhere in these proceedings has it ever been suggested that Pennsylvania law does not control this case. That is all that needs to be said on the issue. Hamme v. Dreis & Krump Mfg. Co., No. 81-2174, slip op. at 5 n.4 (3d Cir. Apr. 12, 1982); Newark *338Morning Ledger Co. v. United States, 589 F.2d 929, 932 (3d Cir. 1976). Consequently, I view the majority’s observation that “we are bound by our decision in Greiner ... to apply Fisher v. Dye in this appeal . . .,” majority op. at 329, as unnecessary, possibly incorrect, and, at the very least, fraught with substantial and difficult questions.
All of our past decisions on this issue have either assumed or explicitly stated that state law controls the admissibility of evidence of intoxication in a diversity case. From my reading of these cases, however, the possible applicability of federal law — either decisional law prior to 1975 or the Federal Rules of Evidence thereafter — was never suggested to or considered by the court. See Greiner v. Volkswagenwerk Aktiengeselleschaft, 540 F.2d at 89; Gensemer v. Williams, 419 F.2d 1361, 1363 (3d Cir. 1970); Parker v. Reading Company, 363 F.2d at 610. The only contrary indication is Judge Kalodner’s comment in a footnote that the Pennsylvania rule applied only because “[tjhere is no Federal or state rule of evidence which would permit the testimony as to drinking. ...” Rosa v. City of Chester, 278 F.2d 876, 883 n.8 (3d Cir. 1960).
On the other hand, at least two courts of appeals have held that federal law controls this question. Ballou v. Henri Studios, Inc., 656 F.2d 1147, 1153 (5th Cir. 1981); Levitt v. H. J. Jeffries, Inc., 517 F.2d 523 (7th Cir. 1975). In Levitt, the Seventh Circuit held that the district court had erred in excluding evidence of intoxication based upon a more restrictive state rule similar to Fisher v. Dye, 517 F.2d at 525. Further, we have recently held that the Federal Rules of Evidence, and not state law, “govern the admissibility of documentary evidence in Federal diversity cases.” Pollard v. Metropolitan Life Ins. Co., 598 F.2d 1284, 1286 (3d Cir. 1979).
I believe that these cases, as well as a close analysis of what has been called broadly the “Erie doctrine,”20 casts substantial doubt on some assumptions concerning the law governing the admissibility of evidence in diversity cases which underly our previous decisions.21
C. The Result Under the Federal Rules of Evidence
For the reasons set forth in part II-B above, the majority’s references to the Federal Rules of Evidence are similarly unnecessary and troubling.
Any assumption that the district court would have reached the same result under rule 403 of the Federal Rules of Evidence is simply speculation.22
III.
In conclusion, given that 12 years have passed since Billow and in light of the clear tone of the Superior Court’s opinions in Schwarzbach, Cusatis, and Couts, I believe that the Pennsylvania Supreme Court would hold that relevant evidence of blood alcohol content, such as that contained in the offer of proof, is admissible without any further showing and would order a new trial in this case.
Without any such speculation, though, if the district court held, as I believe it did, *339that the proffered evidence was inadmissible as a matter of law absent evidence of erratic or reckless driving, then its decision is inconsistent with the plain language of Billow. At the very least, I would remand the case for reconsideration of the defendants’ motion for a new trial so that the district court may exercise its discretion within the proper legal framework.
Also, I suggest that the majority’s statement that we are bound to apply Pennsylvania law here and their references to the Federal Rules of Evidence are unnecessary to a decision in this case and raise substantial and difficult questions. I would hold simply that Pennsylvania law should control this case because no contrary suggestion was made to the district court.
. For ease of discussion, the vehicle driven by the plaintiffs decedent will be referred to as . the “Rovegno vehicle.” The vehicle owned by the defendant Geppert Brothers, Inc. and driven by its employee, the defendant Lewis J. Ward, will be referred to as the “Geppert vehicle.”
. See 42 Pa.Cons.Stat.Ann. § 7102(a) (Purdon Supp. 1981). Under Pennsylvania law, if the plaintiff is found to have been more negligent than the defendant, recovery is barred; to the extent that the plaintiff is less negligent than the defendant, but is still found to have contributed to the cause of the accident, his recovery is reduced by the percentage of his fault.
. At oral argument, counsel for the defendants put forth the theory that Rovegno had fallen asleep at the wheel. I note now, as the court properly did at that time, that such a theory is not proved on this record. Nonetheless, I observe that this is precisely the sort of theory which could have been properly and forcefully argued to the jury on these facts had the district court not improperly excluded the evidence of intoxication.
. 75 Pa.Cons.Stat.Ann. § 1547(d)(3) (Purdon 1977). This section provides in pertinent part:
“(d) Presumptions from amount of alcohol. —If chemical analysis of a person’s breath or blood shows:
* * * * * *
“(3) That the amount of alcohol by weight in the blood of the person tested is 0.10% or more, it shall be presumed that the defendant was under the influence of alcohol.”
. 75 Pa.Cons.Stat.Ann. § 3731(a)(1), (d) (Purdon 1977). These sections provide in pertinent part:
“(a) Offense defined. — A person shall not drive any vehicle while:
(1) under the influence of alcohol to a degree which renders the person incapable of safe driving;
******
“(d) Penalty. — Any person violating any of the provisions of this section is guilty of a misdemeanor of the third degree.”
. A partial transcript of that argument reads:
“THE COURT: What is it you are objecting to?
“MR. LINDSAY [counsel for the plaintiff]: Your Honor, there has to be some basis in fact for the admission of the results. Being an alcoholic, is presumption, which is rebut-table and has been rebutted by the defendant in his description of the accident. According to the cases, that is admissible if there is some reckless or wild driving involved. And there is absolutely no evidence whatsoever, at least at this point, of any erratic driving. The defendant says he was not wobbling — he even used the word himself. What happened has nothing to do with the blood alcohol, and I think it is not admissible. Therefore, I think the whole foundation has to be laid first before we can even get to this witness.
“MR. EMORY [counsel for the defendants]: Your Honor, I disagree with that.
Mr. Ward, in the testimony that was read in yesterday said Mr. Rovegno came across lanes for no apparent reason. We certainly feel that is a reckless condition.
The recent decision of the Pennsylvania Supreme Court, to the extent where recklessness is an issue, that the blood alcohol level is admitted.
******
“MR. LINDSAY: Your Honor, the testimony was this man very gradually moved over a very short distance, not any erratic driving. As a matter of fact, for two miles, this man was driving straight ahead and doing absolutely nothing. He came to a slight upgrade, he is carrying a load of bricks, and his vehicle slowed down because he is going uphill. There is no proof, no testimony, let alone proof, that he did anything that was erratic. He certainly did not do anything that was reckless.
The cases are replete with instances of recklessness: speed, being all over the highway.”
N.T. (4/10/80) at 94-95, reprinted at 79a-80a.
. In that the jury found Mr. Rovegno to have been 30% responsible without hearing the intoxication evidence, it is likely that the percentage would have been substantially higher, per*334haps barring all recovery, if the intoxication evidence had been admitted. See note 2 above. Therefore, I must conclude that the district court’s error was not harmless, even under the liberal provisions of 28 U.S.C. § 2111 (1976) and Fed.R.Civ.P. 61.
. The original trial was bifurcated on the issues of liability and damages. The first damage award was set aside and a new trial ordered on grounds unrelated to this appeal. Rovegno v. Geppert Bros., Inc., et al, No. 79 — 4202, slip op. at 3-4 (E.D.Pa. July 11, 1980), reprinted at 193a-194a. A second jury trial on damages resulted in this award.
. I agree with the majority that the damage award, although large, is not so outrageous as to “shock the judicial conscience.” Frankei v. Heym, 466 F.2d 1226, 1228 (3d Cir. 1972). If the comparative negligence figures are changed as the result of a new trial, the damage awards can be adjusted accordingly.
. A fair reading of the district court’s opinion, as quoted above, in conjunction with the arguments by the plaintiff’s counsel in opposition to the defendants’ offer of proof, see note 6 above, makes clear to me that this was the holding of the district court.
. In addition, a court must consider the application of the Pennsylvania statutes concerning driving while intoxicated. See notes 4-5 and accompanying text above. See also Schwarzbach v. Dunn, 252 Pa.Super.Ct. 454, 462, 381 A.2d 1295, 1298-99 (1977) (dictum).
. This aspect of the test meets the “floodgates” argument that every plaintiff and defendant will now try to prove that the other was intoxicated. To borrow a hypothetical which was used at oral argument, given a rear-end collision at a toll booth, the question of whether the driver in front is intoxicated simply is not relevant because that person’s driving could not have contributed to the causation of the accident and is, therefore, not in issue. See note 13 below.
. To prevent this requirement from becoming illusory by basing relevance on the mere allegation of negligence on the part of the driver against whom the evidence of intoxication is offered, it may be a component of the proper exercise of the trial court’s discretion under the second part of the test to require a threshold factual showing that there would be a basis upon which the jury could find negligence. If that is the case, 1 believe that the facts set forth above constitute, at the very least, such a threshold showing.
. Even if this was a question of discretion, I am somewhat puzzled by the majority’s citation to my opinion for the court in United States v. Alessandrello, 637 F.2d 131, 146 (3d Cir. 1980), cert. denied, 451 U.S. 949, 101 S.Ct. 2031, 68 L.Ed.2d 334 (1981), unless it is being used in a most general way. Alessandrello involved the question of whether the district court had abused its discretion in the method of conducting jury selection in a federal criminal case. See 637 F.2d at 134-37. It did not deal with the exercise of a court’s discretion under Fed.R.Evid. 403, as the majority implies, much less with evidentiary rulings in diversity cases where there may be an element of state law to be considered. Thus, when it comes to narrowly stating the controlling rule of law in this personal injury diversity case, Alessandrelio seems clearly distinguishable.
. The Superior Court’s language in Cave, which this court quoted with approval in Greiner, was as follows:
“It is true that the ‘mere’ drinking of intoxicating liquor is inadmissible to prove that a driver was under the influence of intoxicating liquor and unfit to drive an automobile. However, it is also true that if in addition to the drinking, facts are shown from which a conclusion reasonably follows that the driver *336was under the influence of intoxicating liquor, all the evidence, the drinking and the surrounding circumstances are admissible for the consideration of the trier of facts to determine whether or not the drinking was wholly or partly the cause of an accident for which he is being held responsible.”
219 Pa.Super.Ct. at 515, 281 A.2d at 734. To my mind, there is no way in which this language can fairly be said to be consistent with the narrow approach taken by the district court and the majority here.
. See notes 4-5 and accompanying text above.
. I agree with the majority that the defendants’ characterization of their offer of proof is not determinative. See majority op. at 330 n.4. My point is that the defendants made the Billow argument to the district court, which rejected it in favor of a requirement of erratic or reckless driving of some degree greater than that which the jury found to have been 30% of the cause of this violent and tragic accident. This, in my view, was the error.
. In Sehwarzbaeh, the Superior Court stated, albeit in dictum, that:
“The Legislature has expressly approved of the blood alcohol test as a means of determining whether a person is driving under the influence of intoxicating beverages. A blood alcohol content of .10 or above raises a presumption of intoxication. We feel that this presumption implies that a person with a blood alcohol content of .10 or above is unfit to drive. Otherwise why would driving under such conditions be unlawful?”
252 Pa.Super.Ct. at 462, 381 A.2d at 1298-99 (dictum).
. See, e.g., Man gets prison term in drunken-driving death, The Philadelphia Inquirer, Apr. 13, 1982, § B, at 1, col. 1.
. See Ely, The Irrepressible Myth of Erie, 87 Harv.L.Rev. 693 (1974); Wellborn, The Federal Rules of Evidence and the Application of State Law in the Federal Courts, 55 Tex.L.Rev. 371 (1977). Such an analysis really is less involved with the Erie decision than it is with reconciling Byrd v. Blue Ridge Rural Electric Cooperative, Inc., 356 U.S. 525, 78 S.Ct. 893, 2 L.Ed.2d 953 (1958), and Hanna v. Plumer, 380 U.S. 460, 85 S.Ct. 1136, 14 L.Ed.2d 8 (1965). It was just such an involved analysis which divided this court in Edelson v. Soricelli, 610 F.2d 131 (3d Cir. 1979) (applicability of Pennsylvania Medical Malpractice Arbitration statute in federal diversity case).
. Because it is not clear to me whether we have ever actually held — rather than simply assumed — that state law controls, it is equally unclear to me whether we are “bound” by Greiner or any other cases on this issue.
. It is also, in my opinion, highly questionable in view of the result reached by the Seventh Circuit in Levitt and the district court’s first reaction to this evidence before it considered the Pennsylvania cases. See N.T. (4/10/80) at 90, reprinted at 81a (“THE COURT: Why isn’t this a fact to be considered by the jury along with every other fact?”).