Recently, in the course of amateur trading, I heard about the exchange-traded fund BOXX. BOXX is a new professionally managed fund that implements a sophisticated trade with financial derivatives known as a box spread. If implemented properly, box spreads are essentially risk-free trades that mimic the returns of buying government treasuries (i.e., lending money to the government).
For a retail investor like myself, there is no reason to attempt a box spread. It’s much simpler to just buy treasuries. However, the BOXX fund offers a unique advantage: whereas the gains from treasuries are taxed as short-term income, the gains from BOXX are taxed at a considerably lower rate, so long as the investor holds his position for over a year. This is undoubtedly the reason the fund was created.
I imagine that the legislators who created the tax laws for investment gains did not foresee this application of a box spread packaged in an exchange-traded fund. It is clearly not in the spirit of the law for someone to be able to pay a lower tax rate on an investment that is essentially identical to another one that is taxed as ordinary income. But writing tax laws for investment is complex, and people have incentives to exploit loopholes to lower their tax burden. Lawmakers are thus in a constant arms race with investors to stamp out these loopholes while trying to avoid creating new ones.
In situations like this, it’s natural to wonder why we have such precise rules in the first place. Wouldn’t it be more effective to operate on the basis of more general principles, such as the principle that, if two investments are virtually identical, they should be taxed similarly?1
While I’m not an expert on tax law, I suspect there are plenty of good reasons to use precise rules in this domain. The financial system is a nuanced and delicate apparatus that relies on predictability to function well. Rules keep the system as predictable as possible, even if the fine print of some laws induce unreasonable or arbitrary behavior. Moreover, it would be an enormous challenge to try to enforce abstract principles about the financial system. How many people can explain the mechanics of a box spread and the mathematics of arbitrage that ensure that the strategy offers returns similar to treasuries? Very few judges — let alone juries — would possess the specialized knowledge required to evaluate such claims. Lastly, codifying rules ensures a kind of fairness that limits the number of people punished for misunderstandings. This is even more important in criminal law, where the punishment is typically much more severe.
Nevertheless, could there be domains where the advantages of using more general principles outweighs the costs? I witnessed one first hand.
Jury duty
I always wanted to serve on a jury. Somehow, it took until my early 30s to receive a summons. Intent on finding a spot in a trial, I did my best to signal interest and availability to the judge. During voir dire, I kept my mouth shut, which I guess is an effective way to seem impartial. I was thrilled to have the opportunity to spend a few days in an environment that I had read and heard so much about.
On its face, the case in Massachusetts Superior Court was pretty unremarkable. A delivery driver who had slipped and fallen in an unlucky way on an icy driveway was suing the homeowners for failing to maintain a safe property. The homeowners, who weren’t home at the time, alleged that they acted responsibly given the weather conditions.
The case had some surprising elements, though. The driveway on which the driver slipped was the neighbor’s driveway. The driver had walked up the homeowners’ two outdoor flights of stairs to reach their front door, but then supposedly decided that it was too dangerous to walk back down these steps. Instead, he cut through the homeowners’ front yard and jumped over a small brick barrier to the neighboring property. Unlike the homeowners’ stairs, the driveway that the driver walked down lacked handrails, and he slipped on black ice. (This driveway was also, suspiciously, nearer to the driver’s truck than the homeowners’ residence.)
An overwhelming amount of information was presented in the three-day trial — much of which seemed irrelevant. For example, an expert witness for the plaintiff presented copious details about leg anatomy and the nature of the driver’s injury. Props were introduced to approximate the size of the package the driver would be holding and the dexterity he would need to grab the handrail while walking with it. I learned far more than I would have liked about the driver’s marital life.
There was essentially only one type of evidence that mattered, which was ironically introduced by the plaintiff. Immediately after the driver sustained his injury, his mother came to pick him up and bring him to the hospital. When she arrived, she took several pictures of the homeowners’ property and the neighbor’s driveway. As a result, we jurors did not need to rely on too much verbal testimony to paint a picture of the scene. This turned out to undermine the plaintiff’s case.
Practical challenges of deliberation
My favorite part of the trial was the final morning, when the judge delivered instructions for the jury. This is a part of trials that you don’t typically see dramatized on television, but spoke very much to my academic interests. The judge asked us to consider causation, negligence, even free will! It was much more like an experimental philosophy vignette than I expected. Among the questions we were asked to consider: Was the driver forced to take the alternative path back to his truck? Did he have free will when he made that decision? Was it a negligent decision? More negligent than the homeowners’ lack of care of their property? What would reasonable homeowners have done in this situation?
While the details were fascinating to me, they also proved problematic for our deliberations. One juror joked that she felt like she was retaking the SATs. And as the jury foreperson, I was responsible for leading the group to a consensus.
There were several issues. First, much of the language was tough for us to understand. Even the lawyers seemed to use this language in inconsistent ways. For example, in closing arguments, the plaintiff’s lawyer argued that the driver lacked free will when he abandoned the homeowners’ staircase for the neighbor’s driveway because he was forced into an unsafe situation, and he could foresee that he had a high risk of injury if he had gone down the homeowners’ stairs. In contrast, the defense’s lawyer argued that the mere fact that the driver could have grabbed the handrail and walked down the stairs but instead chose to go far out of his way to hop off the property proved that he made a free decision and that the homeowners were not responsible for the injury. The lawyers seemed to be operating under two different conceptions of “free will” in service of their own goals. Obviously, the driver made some kind of conscious decision to abandon the stairs, and the relevant question was whether the staircase was sufficiently risky to justify this decision. To my naive ears, the discussion of free will seemed superfluous and distracting.
Another more personal challenge for me as foreperson was encouraging the jurors to consider key hypothetical situations. For example, to assess whether the homeowners exercised reasonable care of their property by not treating the snow and ice on their stairs, I thought — perhaps erroneously — that it would be helpful to disentangle questions about the driver’s decisions from those about the homeowners’ conduct. Specifically, after reading through our instructions carefully, I was convinced that we could answer “no” to all questions about the homeowners’ responsibility if we could agree to the answer of a single question in one hypothetical scenario in which the homeowners were fully aware of the weather conditions and the state of their property, but nevertheless decided that they did not need to take any steps to treat the stairs. If we could reach consensus that the homeowners would be reasonable under these circumstances, then a fortiori, the homeowners would be reasonable in the more realistic situations we were considering. The driver’s decisions would also not matter for the sake of assigning blame. But when I tried to lay out this logic, some of the other jurors started arguing about how my hypothetical scenario was not a plausible one. Of course, these jurors were correct regarding plausibility, but I was trying to make the point that sometimes it’s helpful to imagine a scenario that’s unlikely to be true in order to simplify a question. (At least one other juror, who happened to use logic for his work as a machinist, understood what I was trying to say.) I felt like I could have done a better job laying out my reasoning, and in any case, our discussion seemed to gloss over most of the logical concepts we were tasked with discussing.
The reasonableness of “reasonableness”
Despite my frustrations with the deliberation process up to this point, we all reached a consensus in less than an hour: the homeowners were not responsible, and we could fill out “no” to all questions. Our decision had little to do with interpretations of causation, free will, and so on, as laid out in the jury instructions. There was essentially only one concept that turned out to be relevant: reasonableness. Based on the pictures of the property that the plaintiff’s lawyer presented in her opening statement, we all developed strong intuitions that the homeowners’ conduct — leaving their property as is that morning — was reasonable. The pictures showed barely any snow or ice, and the stairs looked safe, especially if someone held on to the railing. Meanwhile, the path that the driver took back to his truck seemed eminently unreasonable for someone who was supposedly worried about falling on ice. In the end, it was these gestalt impressions that won us over.
Should it be concerning that the jury instructions played only a minor role in our deliberations? I left the courtroom with mixed feelings. On one hand, I had high confidence that even if we did follow the instructions in a more careful fashion, we would have ended up in the same place. Moreover, our decision depended upon a kind of intuitive expertise that the jurors had about what it means to be a responsible homeowner. On the other hand, I worried that other trials may be less clear-cut. In those cases, how can we trust that the jury is properly interpreting the judge’s directives?
It seems that there are two broad strategies to addressing these issues. One is to try to improve the language of jury instructions to make concepts like causation and free will more legible for the jury by, say, introducing better examples and distinctions. Indeed, I am applying to law school with hopes of working on this. But another more radical approach is to try to dispense with many of these esoteric philosophical concepts altogether — at least in certain trials. Are they really making the system fairer, or are they merely adding confusion and noise?
I am certainly not enough of an expert on the law to give a cogent justification for keeping or dispensing with this type of language. But I did come away with more respect for the holistic “reasonableness” approach than I had initially. Before the trial, it would have seemed insane to me to think that these types of lawsuits could hang on people’s vague impressions of what is reasonable. Now, however, I see a major advantage of the vagueness: it prevents the exploitation of loopholes that can occur in other parts of the law, such as in my example of the BOXX fund.2 Indeed, imagine trying to write down precise rules for what constitutes appropriate homeowner behavior. Not only would this be impractical for both lawmakers and homeowners who would need to internalize a long list of rules, but we would also probably find that the rules did an imperfect job of matching our sense of fairness. Additionally, some homeowners might figure out how to comply with these rules in a malicious way that causes injury, and other responsible homeowners might be punished for silly technicalities.
There is also a specific advantage of using reasonableness as a standard: the concept, by its very nature, deters exploitation. Somebody who tries to follow the law in a way that is not in the spirit of the law is, almost by definition, not behaving reasonably. The notion of reasonableness incorporates both the actor’s intentions and her behavior. A list of rules could do this, too, but perhaps not so elegantly.
Of course, when the process by which a judgment is reached is more opaque, there are also more opportunities for bias to creep in. Would we have found the homeowners’ conduct as reasonable if they weren’t seemingly educated and affluent? There are also more opportunities for randomness to cause trouble. Imagine enforcing basic traffic laws on the basis of what a “reasonable” driver would do. This would be far more chaotic than simply relying on speed limits and other signage.
Clearly, when a few well-constructed rules can capture most of the relevant space of possible violations, the precision of rules is preferable to vague principles. In these cases, we can also adjust rules to better deter violators. For example, if people are driving too fast, we can raise the speed limit. But we don’t know what the optimal set of rules would be for something like responsible home ownership, and even if we did, there would likely be edge cases that irresponsible people would exploit. In these complex domains, people’s automatic pattern detectors work remarkably well at distilling a high-dimensional space into a one-dimensional judgment of reasonableness, and the impenetrability of people’s introspection about the causes of these judgments acts as a kind of cipher against exploitation.
In contrast, what should we call the kind of prescriptions about causality, negligence, and free will that we were tasked with assessing as jurors? At least in the trial I witnessed, they seemed to suffer from the worst of both worlds. They sounded like well-defined rules in the sense that someone might believe that there is a fact of the matter about whether an action caused another action and so on. Yet they lacked the objectivity of verifiable rules like speed limits or temporal cutoffs for long- versus short-term investments. Unlike reasonableness, people generally do not have much experience judging whether an action was “freely” made, except in very obvious cases. I suspect this sometimes causes more harm than good, even if these legalistic concepts are unavoidable in certain legal contexts.
I am certainly not the first person to ask this question. There’s a rich legal literature about the pros and cons of “rules” versus “standards.”↩︎
Since my jury duty experience, I have since learned that this is an often cited reason for preferring standards over rules. This wonderful book lays out the logic in great detail.↩︎
