Apples, Oranges and Aardvarks

Call it the fallacy of incommensurate balancing. There’s probably a simpler phrase—and I’d use that if it carried the same meaning, which is balancing or comparing attributes which aren’t of the same type. The law is gorged with this fallacy. A common example is the weighing judges are expected to do in the classic discovery motion: burden verses relevance (or utility). It’ll cost $10,000 to produce the documents: is it worth it? Well, they relate to a possibly material issue; and there may [or may not] be another way to prove the issue. There’s maybe $200,000 at stake; or not. So, what’s the answer? The problem of course is that burden and relevance aren’t on the same continuum. We might say we’re comparing apples to oranges, but really, it’s apples to aardvarks.

The same fallacy lies at the heart of almost every multifactor test. A motion to continue the trial: that’s about 18 factors (CRC 3.1332(c)). It compares the need of one party for e.g., more discovery, or to get a new lawyer up to speed, with the prejudice to the other side, e.g., an expert’s scheduling problems, or just the built-in cost of delay. At least in venue motions based on the convenience of the witnesses you can indeed compare the number of material witnesses on each side and the degree to which they will be inconvenienced. But in many motions, such as granting leave to amend, to continue a hearing, or allow more discovery in response to a summary judgment or an anti-SLAPP motion, the parties’ interests are entirely incommensurate.

A common battle has this structure: one side invokes the venerable rule that the merits matter, and the other invokes adherence to procedure. An opposition is late, or exceeds the permitted length; shall it be stricken and its argument forfeit? The arguments are ships passing in the night: one party insists, correctly, the rules be followed (or why bother?); the other, correctly, says the merits must be heard.

Let’s take another set of examples: “sliding scale” tests. A motion for preliminary injunction depends on three sets of balancing: the balance of harms if the injunction issues or not; the likelihood of plaintiff prevailing, and then a final balancing of the results of the first two. Comparing the likelihood of prevailing isn’t problematic, because if the plaintiff is likely to prevail then the defendant isn’t. The balancing of harms, however, sometimes does pose problems because the types of harms can be so different - one side might lose a home, the other side unable to fulfill a father’s dying wish that the house be sold. How do we tell if one side has “more” harm than the other when the harms aren’t of the same kind? But it’s the final balancing which is classically incommensurate: on one hand we have a certain likelihood of prevailing (a little all the way to virtually certain) versus the extent to which one party’s harm exceeds the other’s. What kind of a comparison is that, really?

Unconscionability (which comes up all the time in petitions to compel arbitration) works in a similar way: there is some (a little? a lot?) of procedural unconscionability in an arbitration agreement (e.g., contracts of adhesion, or not enough time to review the document), which we compare to substantive unconscionability (e.g., non-mutual terms). A little of one and a lot of the other may result in a declaration of (overall) unconscionability. There are so many ways in which contracts can be unconscionable that the comparison of a “little” of one type verses a “lot” of the other isn’t obvious. (I’ll get to a recent development on this below.)

Notice that all these issues are matters of discretion, at least to some extent. (On unconscionability there’s case law both that it’s a matter of discretion and that it’s not.) There are enumerated factors to be considered; trial judges will be reversed if they don’t know they have discretion, or if they obviously didn’t even consider the pertinent factors. In re Marriage of Tara & Robert D., 99 Cal. App. 5th 871, 880 (2024). The factors structure the trial court decision, and they provide the basis for appellate review. The factors seem to set forth the rule of decision: all you need to know are the factors and the facts of the case.

But if these are matters of discretion, different judges might decide the same issue differently and still be right. That’s why appellate judges are supposed to affirm discretionary decisions they disagree with (as long as the trial decision is within certain limits). But if that’s true then we must conclude that the factors which appear to be criteria for decision are not the criterion. If different results are consistent with the same facts, then there must be some other rule or factor which does dictate the result.

If the factors don’t specify the answer, what does? And what role, really, do the factors have?

The real role of all these factors is simply to orient the judge (and the lawyers trying to convince the judge). The factors state the sort of decision to be made, they list the things to think about. Multifactor tests map the field of play; they don’t tell you where to throw the ball.

Now, in some areas, despite a multifactor test, there’s a buried rationale, a golden thread, which tends to drive decisions. In trademark law, for example, three (of at least 8) factors are often isolated as the most important, GoTo.com, Inc. v. Walt Disney Co., 202 F.3d 1199, 1205 (9th Cir. 2000). In some cases, the overarching notion of unfair competition drives the decision. In the unconscionability realm, a recent case from our Supreme court unearths guiding principles as it discusses whether or not to sever unconscionable terms and send the case to arbitration anyway. The Court tells us to ask whether “the central purpose of the contract is tainted with illegality,” and whether the “stronger party engaged in a systematic effort to impose arbitration on the weaker party not simply as an alternative to litigation, but to secure a forum that works to the stronger party’s advantage.” Ramirez v. Charter Commc’ns, Inc., 2024 WL 3405593, at *19 (Cal. July 15, 2024). In many areas, such as continuances, defaults, and considering late briefs, there’s an underlying value, often decisive, of hearing issues on the merits. In other areas, too, we might read years of cases to get a sense of the throughput, a vector, to find an underlying rationale.

For the rest, we may never find a golden thread which shapes the implementation of a multifactor test. Then what? Justice Scalia famously objected to multifactor tests because they seem unpredictable and unfair. “The Rule of Law as a Law of Rules,” 56 U. Chi. L. Rev. 1175, 1182 (1989); June Med. Servs. L. L. C. v. Russo, 591 U.S. 299, 348–49 (2020) (Roberts, C.J., concurring); Lexmark Int’l, Inc. v. Static Control Components, Inc., 572 U.S. 118, 136 (2014) (“unpredictable and at times arbitrary results”) (unanimous opinion by Justice Scalia).

But unpredictability isn’t always unfair. Unpredictability inhabits the trial courts—why else try cases?—but we don’t conclude that results in all those cases are unfair.

Confronted with incommensurate factors, we do what do we in our everyday lives, where we face these sorts of conundrums all the time. How do we spend our limited time? How do we spend money? We use our common sense, built by experience. There are often conflicting answers, but it doesn’t necessarily follow that any is wrong.

The idea that a legal test can on its own decide every case is the true fallacy. Just as no contract can provide for every eventuality, so too no legal test can determine every case. As Holmes told us long ago, the “distinctions of the law are founded on experience, not on logic.” Oliver Wendell Holmes, Jr., The Common Law (1881). The law is not an abstract realm; it’s made by humans, for humans, and administered by humans. That’s just the deal.