Scholars, it sometimes seems, are no less susceptible to fads than are eighth graders. Untold numbers of books about doleful vampires fill the e-readers of tweeners, while hundreds of ambitious young scholars across all fields of knowledge search for the research equivalent of Edward and Bella. And like all trends, scholarly fads can be overplayed, even when the underlying insights motivating the trends remain relevant and interesting.
In legal scholarship, and especially in the area of tax law, the trend for the last ten years or so has been to use “behavioral law and economics,” in which scholars attempt to analyze legal and policy questions through a lens that has been enhanced by the insights of behavioral psychology. For example, people actually respond to prices being expressed as $3.99, rather than $4.00, in ways that go far beyond the possible pecuniary importance of a penny. The mysteries of the human mind are endlessly interesting and important.
At some point, however, we find ourselves with too much of a good thing. Call it an application of the old economic concept of diminishing marginal returns, or simply the basic insight that wells can run dry. Whatever it is, however, it appears to be time to conclude that behavioral law and economics has gone from a useful approach that could be carefully applied to relevant questions, to being a hammer that sees everything as the next possible nail.
Why should anyone other than law professors care about this question? Because legal writing, much more than any other body of scholarly work, finds its way into the public discourse—in discussions about Supreme Court decisions, Congressional confrontations, and our understandings of how laws work (and fail to work). When legal scholars are using an intellectual approach more because it is popular among other scholars than because it is useful, public understanding suffers.
In this article, I will explain the basic idea behind behavioral law and economics (which I will refer to as BLE), and how it has been overplayed. I will then use the example of the Affordable Care Act to show how attempts to use BLE have failed, and how that failure exposes the fundamental limits of using BLE in legal scholarship.
Advances in Legal and Economic Scholarship: The Beginnings of Trends, and When the Writing Is on the Wall
Few people who attended law school in the last generation or so could have missed the news that “Law & Economics” has had an outsized impact on legal scholarship. Although there is actually little evidence that courts have used the conclusions of law and economics scholars in crafting legal decisions (outside of a small number of areas, the most prominent being antitrust law), legal scholars in large numbers enthusiastically jumped on the law-and-economics gravy train.
And it really is a gravy train, in the sense that writing in the area of law and economics was correctly viewed as a direct route to high-profile publications, tenure, recruitment to elite law faculties, and so on. As much as we would like to imagine that every young scholar follows her intellectual curiosity wherever it might take her, the fact is that it is highly risky to play in a different sandbox from all of the other kids. Small-bore innovation is rewarded, mostly because it simply allows other scholars to start from a similarly narrow set of assumptions in assessing the younger scholar’s work.
Scholarship in all fields, therefore, is likely to be highly path dependent, continuing in certain directions long after the time when anyone might bother to remember what got it all started. In the case of Law & Economics scholarship, the original idea was to use the supposedly rigorous results of economic scholarship (which, of course, was itself strongly affected by research fads) to enhance traditional studies of legal doctrine. Asking how people would respond to legal rules—how, for example, potential criminals would respond to the “price” of crime, in the form of likely punishments—fed decades of legal scholarship.
When that line of research was played out, the question was what would come next. Adding behavioral insights seemed like a very good idea. And for some time, it was.
The Behavioral Angle: Treating People as Fallible Humans, Not Rational Maximizers
Stripping out the bells and whistles, the basic Law-&-Econ argument has always been that scholars should assume that people respond “rationally” to incentives. For example, if the Securities and Exchanges Commission changes the disclosure requirements for equity offerings, then it would be foolish to assume that profit-motivated companies would not adjust their actions in ways that might defeat the goals of the regulators.
Even people who trade billion-dollar portfolios, however, are still people. It became increasingly obvious that understanding how people behave in real-world situations required us to imagine that people do not always coolly calculate costs and benefits. “Hot” stocks reach prices unsupported by any underlying sensible valuations, and entire financial markets become infected by combinations of fraud, incompetence, and wishful thinking. (The 2008 global financial crisis shows just how damaging that combination of all-too-human behaviors can be.)
Enter behavioral economics. There is no doubt that scholarship is enhanced by allowing for the possibility that people are not acting in the way that old-fashioned economics textbooks assumed. People exhibit myopia, undue optimism, herd behavior, and so on.
Some of the arrows in the behavioral economics quiver seem quite powerful. For example, under a concept called “anchoring,” people with very little information about the value of an item can be influenced by providing all-but-random numbers as references. For example, if you ask people to put a dollar value on a new car, their answers will be all over the map. But if you hint that the price should be $40,000, even for a car that is worth much less, their guesses will cluster around $40,000. Car dealers use this insight in negotiations all the time.
Similarly, the notion of “confirmation bias” seems to describe people’s behavior, in predicting that people will pay special attention to evidence that supports what they think they know, while ignoring other evidence. This phenomenon is now commonly used to describe the consumption of political news, especially in the right-wing “echo chambers” on cable TV.
The problem is that the number of areas in which people’s biases are widely shared, and in which those biases skew people’s behavior in predictable ways, is distressingly small. Part of the problem is that empirical studies testing various psychological approaches is limited to very small experiments, from which some scholars have attempted to draw large-scale conclusions. The larger issue, however, is that behavioral concepts are so broad and open-ended that they can lead in virtually any direction.
The Affordable Care Act and Labeling the “Mandate” as a “Penalty” or a “Tax”: Why Does It Matter, and What Does Behavioral Law and Economics Have to Do With It?
Just over a year ago, the most talked-about Supreme Court case was the decision to declare the Affordable Care Act (the ACA, or “Obamacare,” as some call it) a valid exercise of Congress’s taxing power. In a Verdict column at the time, I described why Chief Justice Roberts’s decision was correct, because it recognized the form-over-substance nature of the arguments by those who said that it was not a tax simply because Congress did not call it a tax.
Of course, even those who are not familiar with behavioral law and economics would recognize that how something is labeled can have a profound effect on the way that people respond to it. For example, we generally find that people make different decisions when a price is described as “$5, plus a $1 premium if you want the deluxe version,” as opposed to “$6, with $1 discount if you want the standard version.” There is no difference between the two pricing schemes, but plenty of research exists demonstrating that people respond differently, depending upon how something is labeled. The problem is that the research does not tell us that people will respond systematically in one direction.
This insight was picked up last Fall in an article in The Atlantic by my GW Law colleague Professor Naomi Schoenbaum, a rising star in the areas of employment law, family law, and gender. Professor Schoenbaum carefully explained how labels matter in behavioral law and economics arguments, and she then described how people might react differently to the label “tax” as opposed to “penalty” in the ACA.
How might people respond differently? As opposed to calling something a “penalty,” which conveys social disapproval, “[c]alling something a tax . . . makes the matter appear more of a choice: buy health insurance or pay the government for failing to do so–we don’t care.” After discussing other ways in which behavioral biases might affect the way people think about the ACA’s mandate as a tax rather than a penalty, she summarized how one set of behavioral considerations could cause people to respond systematically in one direction: “With the tax frame . . . individuals [would be] less likely to buy health insurance.”
Professor Schoenbaum’s ultimate point was that, going forward, the Obama Administration should pay special attention to how it labels its arguments in health care—and, indeed, in all of its public statements. Her argument was thus well supported by pointing out how some behavioral concepts can matter deeply in public policy, by possibly changing people’s behavior.
What Professor Schoenbaum did not need to point out is that there are other cognitive concepts that cut in the other direction. In fact, one can also easily use BLE-inspired labeling notions to argue that the tax label would cause people to be more likely to buy health insurance under the ACA.
For example, we certainly know that people fear the IRS’s reach and power. Without that underlying social suspicion of the tax agency’s power, we would not see the continuing spectacle of Republicans continuing to push the non-scandal that emerged earlier this year—even though it became clear months ago that there is nothing more to it than we knew when it first came to light in the Spring.
Labeling something a tax, therefore, can make it something that people will go far out of their way to avoid. “Buy insurance, or you’ll be subject to an extra tax,” can be understood as, “If you do not buy insurance, the IRS will pay you a visit.” Rather than viewing the choice as a simple matter of paying the tax to avoid buying the insurance, some people might view a tax consequence as especially worrisome.
In addition, we know that many taxpayers happily pay accountants and lawyers more in fees than they save in taxes, simply because they want to avoid paying taxes. (An acquaintance of mine hated paying taxes so much that she gave away money to her relatives every year—not because she wanted them to have the money, she said, but just because she could do it tax free.)
In short, there seem to be as many just-so stories supporting the idea that the tax label will cause people to be more likely to buy health insurance as there are stories supporting the conclusion that they would be less likely to do so. It could also be that there is no difference at all in how people react, if the same number of people respond in one direction as the other.
Yet the public discussion of the ACA case found some people asserting vehemently that the tax label was somehow an assault on people’s freedom, because “behavioral theories prove” that people will buy more insurance under one label than the other. The fact is that the state of knowledge in BLE is not up to that task.
The ACA Example Is Emblematic of the Larger Problem: Behavioral Explanations Are Generally Too Open-Ended to Assist in Legal Analysis
There is also a broader problem that this example helps us to understand. We have now reached the point where there are so many “well-known cognitive biases” in BLE that anyone can reach virtually any conclusion, by relying on a carefully chosen set of behavioral principles. And as I noted earlier, the state of empirical research is such that the “reality test” is too often unavailable.
This problem, moreover, is not new to BLE. A generation or more before behavioralism became faddish in economics, the big new idea was “game theory,” which purported to be a way to move beyond the old-fashioned approach that had dominated economics for decades. After several decades of game theoretic research had been produced, however, Franklin Fisher, a renowned MIT microeconomist, was unimpressed. In an important review article, he first described the state of knowledge before game theory became popular, and then described the state of knowledge after game theory had become all the rage. Upon completing his description of the post-game-theoretic state of knowledge, he then noted wryly that “these are the very words I used in describing the pre-game-theory state of” the relevant economic theory.
This is, I fear, exactly what has happened with behavioral law and economics. Before BLE became popular, only the most narrow legal scholar would have denied that people are more complicated than the standard rational actor approach assumes. Every good legal scholar should have known that labels matter (else law professors would not endlessly revisit the form-versus-substance and formalism-versus-modernism debates), and that we must carefully construct laws to take into account people’s likely responses. Today, every good scholar knows that labels matter, and that we must carefully construct laws to take into account people’s likely responses.
At some point, the winds of change will blow through legal scholarship. Based on what I am seeing from young tax scholars, BLE might be well on its way to becoming passé. More importantly, however, everyone should be especially skeptical when they hear someone claim that “we know” something by virtue of behavioral law and economics. All we really know is that life is still complicated, and it takes careful study to understand the law and predict its effects.