Professors Henderson and Ribstein touch on two theoretical failures of the behavioralist movement which both reveal the prematurity of ‘behaviorally-informed’ regulatory proposals: the behavioralist assumptions that (1) behavioral biases theoretically necessitate, or at least enable, public intervention, and (2) governmental entities can net improve individual outcomes over the status quo of unfettered, if limited, human capabilities. I think both of these observations highlight the shocking dearth of theoretical exploration amongst behavioralists thus far. I want to focus on connecting these assumptions to the connection between behavioral economics and administrative regulation.
The first assumption calls for a comparative institutional analysis; the second at least invites one. Both are necessary – and neither sufficient – to justify behaviorally-informed regulations. Neither proposition, however, is self-evident. Professors Bar-Gill and Warren broadly claim that borrower irrationality gives rise to predatory lending practices, which yield sub-optimal outcomes where consumers borrow more than they otherwise might. Even assuming these biases into existence, however, it is unclear why, absent market power amongst lenders, these supra-competitive profits would not be competed away as price (and non-price) credit terms themselves. Bank of America, among others, offers a “keep the change”-style policy – a relatively painless way for a consumer to voluntary alter the default rule of their own expenditures to increase savings. Chase now peppers the airwaves with “Slate with Blueprint,” purporting to help a consumer navigate a myriad of bills, obligations, and interest rates to neatly tailor one’s own spending patterns. Unless behavioralists can demonstrate – or even articulate! – some reason to believe that firms (1) remain sufficiently rational to predate upon consumer biases, yet (2) despite this rationality will not compete the monopoly profits of those biases away – I echo several other contributors’ insights that the leap from behavioral quirk to public intervention is something of a non sequitur.
Yet I think that Professor Ribstein’s point addressing the second of these two assumptions is even more unsettling. Governmental interventions run a gamut of intrusiveness, both along actors and actions. Potential behavioral regulations may compel disclosures, reverse common-law default rules, or prohibit disfavored choices. Potential behavioral regulators could conceivably include, amongst others,the court system (through the common-law evolution of behaviorally-informed doctrines or the interpretation of behaviorally-informed statutes), state legislatures, state administrative bodies, Congress, or myriad administrative agencies. Behavioralists seem to prefer the latter – witness the CFPB and its towering regulatory superstructure. To the extent that we cannot trust markets for law when we cannot trust markets whatsoever, as Professor Ribstein points out, this is intuitively consistent. Yet I neither believe the Administrative Procedure Act – or administrative law, or the administrative process more broadly – is well-adapted to the assumption of pervasive irrationality among consumers (much less regulators), nor do I think behavioralism is prepared to implement the conceptual checks on the regulatory process that would flow from a consistent application of checks against behavioral irrationality.
By way of example, reference the Supreme Court’s position on equitable estoppel against administrative agencies. Hundreds of government agencies administrate various regulatory and benefits programs for myriad classes of individuals. By necessity these regulations are indecipherable to even many specialists; the agencies generally provide free advice upon inquiry. The Court finds this general advice a social good which, albeit subject to a meaningful chance of error, provides rational consumers of governmental services with a low-cost means of resolving low-value inquiries. We assume that individuals who have especially significant benefits or risks at stake on the interpretation of one rule or another will participate in the regulatory or adjudicative process more closely, or will otherwise rationally expend resources (say, on a tax lawyer) to ensure their conduct conforms to the law. Accordingly, an individual must demonstrate something approaching unequivocal statements by an authoritative government source and extreme, unjustifiable harm to a private entity as a result of ill-taken government advice in order to equitably estop a government agency or actor.
All of this, of course, assumes rational decision-making by individuals as a sine qua non. Even if we put aside the necessity of assuming rational actors within the government – no small task for a behavioralist! – it stands to reason that surely public actors, insulated from competitive pressures, must be called to account for the potential hyperbolic discounting effects the agencies have upon consumers of their services as well. In other words, the chances the Social Security Office’s free advice is in fact erroneous and will result in a deprivation of benefits, followed strictly, may actually be 1 in 5; however, a behavioralist’s optimism bias may suggest, for example, that individuals relying on that advice may assume it as 1 in 10. Or 1 in 100. Or 1 in 1,000. And so forth. To demand private firms hold consumers harmless for their irrationality is to ignore that behaviorally-informed administrative law would, or should (in a world of undivided light and perfect intellectual consistency) do so as well. Are behavioralists prepared to call for robust equitable estoppel against government agencies with superior information that provide information to consumers? This, of course, is merely a narrow, trivial example; the entire construction of the Administrative Procedure Act presumed rational – or super-rational – technocrats capable of “solving big problems;” to broach the possibility of individual irrationality at the regulatory level is both a new heresy and a very, very old one.
These inconsistencies are troubling. To the extent regulators and regulations flow despite – and not in response – to them, it leaves liberty-minded observers with an impression that behavioral economics wields a cudgel of irrationality to justify subordinating, rather than revealing, individuals’ true preferences. It is not historically atypical, however, to see a movement that stands to deeply undermine liberty shroud itself in the language of individualism and personal protections. That behavioralists yet feel compelled to do so – as unmodified “paternalism” remains a dirty word in a contemporary culture suffused with examples of it – should provide advocates of freedom with some small solace.
Chris: See this WSJ piece by Matt Ridley on the biases of regulators. As Judd points out, your claim about the benefits of bureaucratic decision-making are, as far as I know, not supported by empirics, and definitely not favorably compared to market solutions (like firms) to individual decision-making limits. The biases canvassed by Ridley in that article are as (more) likely reinforced by institutional pressures as mitigated by them.
Prof. Buccafusco —
Thank you again for the thoughtful reply.
However, the distinction between individual consumers and group regulators cannot bear the burden you demand of it here. You claim that the group regulatory structure is superior to private individuals’ decision-making processes in terms of minimizing the supply of socially wasteful errors. Let’s hold aside that there just simply isn’t any evidence in support of that proposition. You’ve missed the “compared to what?” question. The relevant for comparison are government groups (regulators) and groups in the market (firms). The key question is which of these two groups has the comparative advantage in mitigating decisions plagued by cognitive biases. Here, you offer a naked assertion that regulators will be more free of error than consumers but do not address firms and suggest that we have to wait until we have significant, replicated studies of group decision-making at the firm level. We do. Industrial organization and applied microeconomics have provided plenty of evidence on this score. We have no need to wait.
The statement “I think regulatory bodies are likely to have procedures that generate the latter” has two problems. The first is that this is an essentially empirical statement. My dichotomous rationality/irrationality assumption was meant to show that under either theoretical structure, behavioralism needs to demonstrate heterogeneous rationality to have policy relevance. You propose it as an assumption. So far as I can tell, you concede there are not significant, repeated studies demonstrating this.
Let’s, however, assume heterogeneous rationality. Assuming your selective irrationality — that individuals are irrational and groups are not. Again, this distinction merely underscores that the more appropriate institutional trade-off is not a question between agencies and individuals, but between agencies (involuntary public groupings) and firms (voluntary private groupings). And once again, there is a wide body of literature demonstrating how the former can misuse and misappropriate the regulatory process to capture private rents. There is also a wide body of literature demonstrating how the latter can respond through specialization to adapt to mistakes and incomplete information. The appropriate comparison is not whether agencies can correct for irrational individual behavior, but how well agencies can do so compared to private ordering through firms. I fail to see why we would default to assuming agency competence against a backdrop of a theory that assumes fundamental human irrationality.
Chris: You can’t assert as equivalent the effects of the market and the effects of group deliberation–or else perhaps you have solved the socialist calculation problem? Judd’s point, echoing Henry’s, echoing Armen Alchian’s, echoing Hayek’s is that the inexorable evolutionary force of the market is enormously powerful–among other things, incorporating and processing more knowledge than a single individual or group could possibly hope to possess and process. Yes, I’m sure group deliberation can have some de-biasing effect, just as I am sure that it can result in groupthink. But if we’re talking about relative institutional competence, I’ll take the market over bloated government bureaucracy any day. Meanwhile, while far from “perfect,” the market does not systematically make decisions about how to spend other people’s money, or about how other people should be allowed to behave. Whatever biases impair individual decision making, can there be any doubt that individuals are more biased in the direction of their own welfare than are even well-meaning and skilled technocrats deciding how to spend other people’s money?
Judd, your points may or may not be correct with respect to -firm- behavior. The distinction that I made was between individual consumer behavior and group regulatory behavior. I don’t believe that either of your homogeneous assumptions (pure rationality or pure irrationality) is warranted by the evidence. Rather, my point was that certain actors are more likely to be in positions where default, System I cognition will be overridden by more controlled System II cognition. I think regulatory bodies are likely to have procedures that generate the latter. A fuller understanding of firm behavior will have to wait until we have significant, replicated studies of group decision-making.
Professor Buccafusco —
Thank you for your comment; a few thoughts in return. You write the following:
“It is odd that many commentators in this symposium have suggested that market forces will serve to debias spur-of-the-moment individual decision-making while not believing that regulators will experience similar debiasing when engaged in deliberative, group decision-making.”
I do not think this is odd at all when one considers the economic effect of market incentives. Market incentives — specifically the profit motive — encourage efficient information-gathering, systematic re-evaluation of past choices, and specialization to reduce costs associated with cognitively biased decision-making. One can imagine traversing a labyrinth of regulations is an excellent example of such an activity. Voluntary organizations that arise through private ordering — that is, firms — have incentives to respond to and maximize their gains from these forces. Involuntary organizations that arise through political decision-making — agencies — do not. It is not that I (or I suspect others here) do not believe public agencies can behave rationally. Indeed, there is a wealth of public choice literature that demonstrates how public actors can — and will — respond vigorously to incentives, especially when they can internalize the benefits of those incentives and externalize the costs of their resulting policies.
The fundamental problem is that either the homogeneous assumption of rationality or the homogeneous assumption of irrationality cuts against the behavioralist agenda. If regulators are rational and behave rationally (as do individuals), the public choice literature more than adequately explains why agency decision-making is more likely to lead to sub-optimal results than private decision-making. If regulators and individuals are both irrational, BLE as of yet offers no policy implications to correct for these errors (and you appear to agree here) — or to even take account of them. If, as you suggest, institutions can behave rationally when individuals are nonetheless irrational, there is no reason, theoretical or empirical, to believe that rational institutions immune from dynamic incentives arising from the market (agencies) will learn faster or better than those subjected to those incentives (firms).
I’d like to address a very small part of Judd’s post, but, since it’s a point that has been raised by Stephen Bainbridge and Larry Ribstein, it seems worthy of comment. Judd writes, “Even if we put aside the necessity of assuming rational actors within the government – no small task for a behavioralist!…” Similarly Stephen and Larry point to a seeming irony in BL&E that regulators are likely to be affected by the same biases they are trying to correct, thereby undermining the legitimacy and likely success of the behaviorist project. These are potentially valid concerns, but I think they’re overstated.
On one hand, as Jeff Rachlinski et al.’s work shows, even sophisticated legal actors – in their studies, judges – are subject to a number of problematic heuristics and biases. On the other hand, many of these behaviors are likely to be mitigated by the context of legislative and regulatory decision-making. Almost all of the BL&E work focuses on individual decision-making rather than group decision-making (This is, in fact, one of its significant shortcomings.). When regulators have the opportunity, motivation, and information necessary to make decisions, there are good reasons to think they will be less likely to make snap judgments based on gut feelings.
The distinction between System I and System II decision-making is helpful here. One of the chief findings of behavioral economics is that in many cases people make automatic judgments based on more or less adaptive heuristics. These System I decisions are often the result of availability, representativeness, or emotional state. Behavioral economics does not say, however, that these are inevitable or unavoidable channels for decision-making. In the proper circumstances, people can shift from System I to System II cognition, allowing them to more rationally weigh evidence, consider alternative outcomes, etc. Much of the goal of BL&E is to consider low cost legal solutions to motivate System II cognition.
In many cases, administrative and legislative procedure will tend to promote System II decision-making. Regulators are less likely to be affected by heuristics because they will tend to have the opportunity, motivation, and information necessary for overriding their gut feelings. It is odd that many commentators in this symposium have suggested that market forces will serve to debias spur-of-the-moment individual decision-making while not believing that regulators will experience similar debiasing when engaged in deliberative, group decision-making. On the contrary, regulatory decision-making will generally have built-in procedures for mitigating the effects of heuristics while many individual decisions will not. I am not, of course, suggesting that all individual decision-making is irrational and all regulatory decision-making rational. Rather, I think that the concerns about heuristic cognition apply with differential force and thus create different levels of concern in different contexts.