I write separately to memorialize my perception that this case exemplifies the unfortunate cloud of confusion that surrounds the “abuse of discretion” standard of review. Properly viewed, the trial court’s ruling here was not an exercise of discretion but an application of a rule of law. The trial court had no discretion to decide what the applicable law was or to determine its logical effect in light of the facts found. Its legal analysis was either correct or incorrect. Since this court’s power to decide questions of law is paramount to that of the trial court, we are entitled and indeed obliged to reverse any ruling that we find rests upon an error of law, provided of course the error was prejudicial.
At bottom the concept of “discretion” is one of latitude. It means that on certain types of issues, the trial court’s ruling will survive review even if the members of the reviewing court might have ruled otherwise. Such a standard has two proper functions. One is to shield rulings on issues that the trial court has presumptively superior competence to decide correctly. The most obvious and familiar example of this is where the court makes a finding of fact on conflicting evidence. Ordinarily such a finding is binding on appeal. If the issue was tried by a jury, this rule has a constitutional dimension, for the right to jury trial would be a hollow thing if judges were free to substitute their own findings for the jury’s. But even when the facts are tried by a judge there is a sound reason for appellate deference, which is that the trial judge is in the better position to determine the true meaning of conflicting evidence. This is easy to see in the case of live witnesses, whom the reviewing court has no opportunity to see or hear. It is more difficult to defend in the case of documentary evidence, which may be transmitted to the reviewing court and examined by it with a degree of care and competence seemingly equal to that of the trial court.
The second function of a discretionary standard of review may be best characterized as granting the trial court a kind of arbitral power, not unlike that of a baseball umpire, in the name of judicial economy. This function *1223recognizes that on certain issues, the trial court’s ruling should stand simply because the social cost of questioning it outweighs the private benefit of having a reviewing court substitute its views for those of the trial court. This will most obviously be true where the normative and policy considerations bearing on the issue stand in equipoise, such that reasonable minds obviously could and would differ over the correct result. Basically these are situations in which the law says that a litigant need only be allowed one shot at a favorable adjudication—before the trial judge—and unless that adjudication can be shown to be clearly wrong, it should stand. A deferential standard of review encourages parties to make their best case in the trial court and not to appeal merely on the chance that the reviewing court will feel differently about the dispositive issue.
It may be impossible to generalize about the kinds of issues that will fall within a trial court’s discretion. They are probably best viewed as a family of customs that have grown out of the judicial experience of the ages. Obvious examples may be found in the area of equitable remedies, where such questions as the balance of harms may be dependent on such a complex and debatable set of competing considerations that there is no social utility in second—guessing a decision, once it is properly made. In such situations the trial court is empowered, umpirelike, to simply call ’em like it sees ’em, without fear that an appellate court will overrule its “call” simply because, to it, the runner looked safe.
Every zone of discretionary latitude not resting on superior competence represents a grey area within which the correct outcome is doubtful or debatable and so the trial court possesses an arbitral power by a kind of default. A discretionary standard vests in the trial court a zone of autonomy, the dimensions of which may be rendered somewhat more definite by characterizing it as “broad” or “narrow.” Although the phrase “abuse of discretion” connotes a sense of moral opprobrium somewhat reminiscent of “miscarriage of justice,” it really means only that in making the ruling under scrutiny, the court strayed outside the allowed zone.
In no case is a discretionary standard a license to commit error. Whether the zone in a given setting be broad or narrow, it never extends to getting the law wrong. The law may be obscure; it may be uncertain in the sense that its application to a given situation is not squarely governed by precedent or statute; but it is not the kind of grey area in which the trial court enjoys the autonomy of an umpire. Indeed it is the job of courts, particularly appellate courts, to make it as black-and-white as they reasonably can, so that citizens may shape their conduct to conform to it. The governing law can therefore never be a question entrusted to trial court discretion. Nor, properly understood, can the application of law to a given set of facts. The trial court has the *1224primary authority to decide what the facts are, and in some cases may have the primary authority to make a normative judgment about them, but it never has primary authority to decide what rule of law applies.
The trial court’s ruling here, as I understand it, rested on the premise that the Department of Motor Vehicles (Department) had failed to establish that the lab report was prepared at or near the time of the blood-alcohol analysis it was offered to prove. This analysis rests on an erroneous legal premise, i.e., that the official records exception to the hearsay rule, on which the Department relied, required that the written report have been made at the time of the entry. The actual statutory requirement is that the “writing made as a record’ of the blood analysis have been “made at or near the time” of the blood analysis. The “writing made as a record” of the blood analysis was not the printed lab report offered in evidence, but the computer entry of which that printed report was a secondary or derivative record. The report was not objected to on the ground that it did not comply with the rules governing secondary evidence of writings. (See Evid. Code, §§ 1520-1567 & especially 1530 [official writings], 1552 [printed representations of computer data].) It was objected to on grounds of hearsay. In that context it constituted one of two layers of hearsay. The earlier layer was the computer entry “made as a record” of the blood analysis. That entry satisfied the contemporaneity requirement; the technician’s uncontested certification recited that “the recording of the analysis results was done at the time of the analysis.”
The date of the printout was relevant only to the second layer of hearsay, which was the lab report’s assertion that the earlier entry had been made. The report came within the official records exception for that purpose if it was made near the time of the “act, condition, or event” it recorded, i.e., the presence in the computer system of the entry it reiterated. The printed report was necessarily made while this condition existed, because it is presumably impossible to print out or otherwise retrieve information that is not present on the computer from which it is retrieved. The report was therefore created at or about the time of the condition it described. Since it also established that the underlying event—the blood-alcohol test—was recorded at or about the time of its occurrence, both hearsay layers came within the official records exception to the hearsay rule.
The trial court might have had discretion, at least in theory, to find that some of these facts were otherwise. Conceivably it could have refused to credit the technician’s averments. But it had no discretion to accept the averments and then sustain a hearsay objection. That was an error of law. It did not depend on any fact the trial court was peculiarly competent to find. It did not fall within any “grey area.” It is therefore not shielded by a deferential standard of review. I do not understand the majority opinion to *1225actually hold otherwise, but my colleagues feel constrained to reiterate hoary judicial recitals that the issue was entrusted to the trial court’s discretion. In my view such recitals can only perpetuate the confusion and perplexity already attending this subject.