The Yale Law Journal

VOLUME
131
2021-2022
Forum

On “Confetti Regulation”: The Wrong Way to Regulate Gamified Investing

17 Jan 2022

On “Confetti Regulation”: The Wrong Way to Regulate Gamified Investing

abstract.“Gamified” investment apps like Robinhood use behavioral psychology to encourage frequent and often maladaptive trading activity. To address that problem, securities regulators may be tempted to regulate app design. Such an approach might involve bans on casino imagery, push notifications, confetti, or other aspects of the user experience. But that approach could draw the entire field of securities law into a techno-libertarian First Amendment thicket. This Essay describes the First Amendment litigation that regulators risk provoking, as well as the damage that they might do to the broader project of securities law. The Essay also proposes a strategy for regulators to avoid unnecessary litigation risk while still protecting consumers from the risks of gamified investing.

Introduction

Technology has made it easy to trade stocks and other speculative assets on mobile phones. Broker-dealers, market participants regulated under the securities laws, sponsor these apps. One popular app, Robinhood, offers attractive user-interface and user-experience design and salient contract terms—like no commissions for trading stocks—that are highly competitive in the market for “retail” or ordinary investor brokerage.1 Flashy graphics and frictionless trading have made it easier—and perhaps more fun—than ever before for ordinary people to trade stocks.

Robinhood’s zero-commission business model leads it to encourage substitute revenue sources, like encouraging clients to trade prolifically to maximize third-party compensation to the broker. To that end, these apps incorporate design features that are sometimes called “gamification”: behavioral prompts and flashy casino-like design elements that encourage unreflective or unconsidered decision making based on cognitive bias, imperfect rationality, and impulse.2 These “gamified” design elements include randomized “surprise stocks” that reward users for linking bank accounts and referring new users, push notifications hyping short-term volatility in “biggest mover” stocks, and (until recently) splashes of animated confetti to celebrate a trade.3 App developers point out that these features make investing more fun and approachable to nonprofessional individual investors—”retail investors,” as they are called within the industry.4 But by appealing to impulse rather than deliberation, the features promote patterns of risky trading that may not be in most retail investors’ best interests.5

Securities law subjects the financial intermediaries behind these apps to broker-dealer rules governing their communications with retail-investor clients. Now, regulators are asking how those rules might apply to gamified app design.6 A majority of the Securities and Exchange Commission (SEC) has expressed interest in regulating gamified app design, and the agency has requested information from the public on what it calls “digital engagement practices” in broker-dealer regulation.7 Massachusetts securities regulators have meanwhile sought to revoke Robinhood’s broker-dealer registration, alleging that “gamification” violates state-law fiduciary duties owed to clients.8 And the Financial Industry Regulatory Authority (FINRA), the self-regulatory organization for broker-dealers, signaled that its examination and risk-monitoring program is “increasingly focused” on “risks associated with app-based platforms with interactive or ‘game-like’ features.”9

For regulators, the concern is that gamification and other digital-engagement practices in zero-commission stock-trading apps may subtly influence investors’ behavior and preferences.10 This influence may not be transparent to users of these stock-trading apps, causing them to make decisions that are inconsistent with their preferences about how to achieve their financial goals—such as by encouraging maladaptive overconsumption of trades.11

In principle, there are many ways for regulators to both define the problem and address it through policy. For instance, the SEC’s concept of “digital engagement practices” includes not only “gamification,” but also backend practices like AB testing, machine learning, and other ways of finetuning algorithmic design.12 In other work, we address a fuller range of ways to frame and respond to behavioral design in retail brokerage apps.13

In this Essay, we focus on just one approach to the problem. Regulators might find it tempting to ban design features that they find objectionable, such as bursts of confetti after the successful execution of a trade, because of their plausible effect on the trading behavior of investors. We use “confetti regulation” to describe this kind of approach to the problem of behavioral design: command-and-control or prohibitory regulation of behavioral design features in brokerage apps. Such regulations might cover confetti splashes, push notifications, leaderboards, behavioral prompts, and the like.

Our Essay warns securities regulators away from confetti regulation, either through new rulemaking or through enforcement of existing law, for two reasons. The less important of the two is that confetti regulation could be hard to implement and justify, and risks devolving into a game of whack-a-mole: reacting to regulatory concerns as they pop up without planning for future concerns. The more significant problem is that confetti regulation would likely spur deregulatory efforts from technology attorneys who cast informational molehills as free-speech mountains.

Securities regulation is largely about controlling the terms that dictate how companies communicate with and provide information to the capital markets—yet it has somehow avoided serious First Amendment scrutiny for decades.14 But in recent years, the SEC “has lost a string of important appeals before the high court” concerning its enforcement program.15 In our view, it would be unwise for the agency to pursue regulatory strategies that would precipitate further deregulatory constitutional challenges. We are particularly concerned that Silicon Valley technology lawyers might set out to establish a First Amendment landmark decision on the “right to code,” and that along the way they might lay the groundwork to invalidate securities regulation itself at a more fundamental level.

The safer approach is to avoid making regulations about the software—by which we mean two things. First, regulators should avoid asserting direct control over “bad” software design, and instead focus on the business model that drives it. Second, regulators should justify regulatory action in terms of settled policies that are technology neutral. Predatory gamification might, for example, violate longstanding policies prohibiting brokers from putting their own interests in remuneration ahead of the retail customer’s interests, such as by “churning” customer accounts or recommending unsuitably large numbers of trades.16

In Part I, we begin by introducing gamification as the product of the modern stock-brokerage business model and discussing the potential social harms that it might generate. In Part II, we focus on one salient and superficially easy regulatory intervention: “confetti regulation,” or command-and-control style regulation of the aesthetic design of brokerage apps. We argue that policing brokerage app design in this way would raise hard line-drawing problems and, in any case, would plausibly be subject to heightened First Amendment scrutiny. In Part III, we highlight two implications of our argument. In an era when courts are engaged in constitutional deregulation, securities regulators might avoid confetti regulation to stave off unwelcome scrutiny of the securities laws’ information-control provisions. Instead, we suggest, regulators should consider framing gamification and other digital-engagement practices as old wine in new bottles: technologically mediated efforts to appeal to cognitive and behavioral tendencies that encourage self-directed clients to behaviorally churn their own accounts, maximizing revenue to the broker. We conclude by teeing up for future work components of a framework for assessing behavioral design against the securities laws’ goals.

I. gamification and retail investing

The business model of stock brokerage has changed significantly in recent years. Stockbrokers charge transaction-based compensation for providing financial-advisory services and market access to clients, including commissions for effecting their trades. These commissions were historically high, making active trading the domain of the wealthy and inaccessible to many ordinary investors.17 Several trends have disrupted this obstacle to active trading by ordinary investors: deregulation of fixed commissions and intermarket price transparency in the 1970s and 1980s,18 technological innovation in the 1990s,19 and the adoption of decimalized rather than fractional pricing with one-penny minimum tick size in the 2000s.20 The ensuing price wars among online discount brokers led many online discount brokers, including those with the biggest market share, to offer zero-commission trading by late 2019.21 Ordinary investors can therefore trade stocks without paying commissions to a broker.

But firms offering “free” services—particularly online services—typically do so by collecting revenues in ways that are less salient to the consumer.22 They may collect and analyze consumer data (e.g., social-media usage) for third-party consumption, or tease users into long series of microtransactions (e.g., unlocking new levels in Candy Crush).23 The story at zero-commission brokerages is much the same: these brokerages often sell clients financial advice, margin lending, net-interest income, and “payment for order flow” (PFOF).24 PFOF, in particular, appears to drive much of the gamification trend. It is something like a bounty system. Third parties want information about or access to retail investors’ trades.25 These third parties then paya broker (like Robinhood) to route the execution of those trades to them rather than elsewhere in the stock market.26

The PFOF/zero-commission business model gives investment-app developers every incentive to maximize user engagement with the product. In this respect, they are in the same boat as ad-financed social media or “free” phone games with in-app purchases—or slot machines, for that matter. This common incentive structure has led video slot machines, Facebook, Candy Crush, and Robinhood alike to use behavioral design to encourage habit formation and maximize time spent using a device.27 Robinhood famously splashed confetti across users’ screens upon execution of a trade or offered a virtual scratch-off ticket to those who had won some reward.28 In its request for information about digital-engagement practices, the SEC noted other examples of these kinds of design features, including “[s]ocial networking tools; games, streaks, and other contests with prizes; points, badges, and leaderboards; notifications; celebrations for trading; visual cues; ideas presented at order placement and other curated lists or features; subscriptions and membership tiers; and chatbots.”29

Lots of time spent “playing” a brokerage app is an undesirable outcome for most retail traders. Decades of research shows that in aggregate, retail investors perform worse the more actively they trade.30 Empirical models of retail-investor behavior attribute the persistence of underperforming active trading to different causes, including sensation seeking, overconfidence, and limited attention.31 Being distracted is not all that different from being duped if app-design features like push notifications, curated lists of securities, and leaderboards lead investors to trade more, or in different securities, than they would in the absence of these influences.32 Some recent studies have documented that Robinhood users engage in attention-induced trading in sets of securities that were more salient because they appeared on leaderboards within the app.33 These results indicate that “gamified” app design and other digital-engagement practices appeal to behavioral tendencies—and can even encourage trading in particular securities.

Studies like these raise troubling questions about the consumer-welfare implications of apps designed to stimulate frequent trading in stocks, exchange-traded funds, and cryptocurrencies by appealing to behavioral psychology. It seems likely that this kind of design offends a broader policy in securities law against brokers who put their own interest in transaction-based compensation ahead of the client’s by effecting or encouraging more trading than is in the customer’s best interest. We therefore agree with regulators who think gamified investing deserves regulatory attention.34 But we also think that the most intuitive approach—a simple ban on dangerous features35—would produce unintended consequences.

II. the easy case against confetti regulation

We see two reasons to avoid a regulatory strategy that focuses directly on app design. The first is that concepts like “gamification” and “behavioral design” are slippery and do not lend themselves well to line drawing. The likelihood that these features may occasionally be helpful or at least innocuous only complicates the line-drawing problem. Second, we expect that any law regulating software design directly will draw First Amendment challenges.

A. The Elusive Problem of Objectionable Confetti

Line-drawing issues will complicate any effort to regulate behavioral design. There are a few reasons for this. The first is that games in general are not identified by the presence of particular features or elements, but by a Wittgensteinian “family resemblance” to other games.36 One federal judge, characterizing “[t]he term ‘game’ [as] exceedingly vexed and difficult,” struck down a city ordinance that prohibited playing games in public spaces.37 The ordinance, she wrote, was “hopelessly vague and substantially overbroad, because there is no attempt to explain what is meant by ‘game,’ and because it prohibits a tremendous number of innocent and even desirable activities.”38 Any broad ban on “gamification”—a concept defined by a second layer of “family resemblance” to games themselves—would suffer the same difficulties.

Narrower definitions of gamification are perhaps possible, but these quickly run into problems of underinclusiveness that gambling regulators know well. “[N]o sooner is a lottery defined,” the North Carolina Supreme Court wrote in 1915, “and the definition applied to a given state of facts, than ingenuity is at work to evolve some scheme of evasion which is within the mischief, but not quite within the letter, of the definition.”39 So if one state’s definition of gambling revolves around a “game of chance,” for example, then gambling promoters will look for ways to introduce some trivial element of skill to the game. One recent hustle involves video-arcade machines that allow players to catch fish and other treasures in an unusually fast-paced and casino-styled koi pond. These “fish game tables” accept large-denomination bills and pay out occasional cash winnings.40 Chance (and AI) largely determines who wins and how much, but skill seems to play some small role—just enough, perhaps, to buy the business model a bit of time while the gambling regulators catch up to it.41

We suspect that securities regulators taking on the mantle of “gamification regulators” could easily find themselves in the same “whack-a-mole” situation: reacting to regulatory concerns as they pop up, but making little progress toward addressing future concerns.42 And when the mole can be reconfigured and adjusted—as when Robinhood replaced the “confetti” feature overnight with “new, dynamic visual experiences that cheer on customers through the milestones in their financial journeys”43—regulators will struggle all the more to update and define any ex ante regulations.

A second difficulty is that the kinds of gamification features that might be swept under a “confetti regulation” label are not always particularly objectionable. Confetti itself, for instance, might look crass compared to the financial industry’s staid aesthetic standards. But is it really the sight of confetti that leads users to trade in highly volatile stocks, assets, and cryptocurrencies, or is it the simple unadorned thrill of making big money off of a risky trade?44

Another example: consider confetti regulations that have attempted to control aesthetic design choices such as the colors in which information is presented. These design choices could plausibly change the salience of certain investing options, as evidence suggests that presenting financial data in red may subtly color investors’ perception of future risk and trading decisions.45 But if American traders have come to associate the color red with negative financial performance, should regulation try to sever that link because the presentation of information alters investor behavior?

Or consider further still push notifications, which present information in particular ways to increase its salience. Some push notifications might serve as calls to action by notifying a user that a particular stock is down more than five percent or that they have not yet traded in their new account (so won’t they check out a list of popular stocks?).46 But other push notifications seem more helpful or benign, such as those indicating that a good-til-canceled limit-order trade was executed or that a user has been logged out of their account after being idle for a certain period of time. There are other gray areas: many notifications are defaults subject to opting out, while others might require opting in.

Defining the scope of regulation is a well-understood problem, and these line-drawing issues complicate the ex ante rulemaking approach substantially.47 Ex post adjudication of principles-based rules, meanwhile, will remain subject to loud and influential, if not entirely persuasive, criticisms that the SEC is engaging in “regulation by enforcement.”48 And when regulators draw lines that are either fuzzy, misplaced, or informed by controversial science, they are likely to face challenges under the Administrative Procedure Act49 or the First Amendment. In our view, these are reasons to avoid regulatory techniques that are directly responsive to specific app design choices.

B. Unwanted First Amendment Attention

We are particularly concerned about the First Amendment challenges. Securities law is heavily concerned with regulating the flow of information—so much so that First Amendment scholar Fred Schauer once joked that it “would not be wholly inaccurate” to call the SEC the “Content Regulation Commission.”50 However, although it is full of “restrictions and requirements that in other contexts would set off a host of First Amendment alarm bells,”51 securities law has remained mostly sheltered from the searching First Amendment scrutiny that courts have applied in other contexts.52

The reasons for that shelter are unclear. What is clear is that the shelterlooks ever more anomalous amid the broader trend in favor of corporate speakers who brandish a “weaponized” First Amendment against profit-reducing regulations.53 And though we seriously doubt that a confetti ban would impair the freedom of expression in any normatively significant way, we think the same could be said for many more of the marketing regulations that courts have struck down as unconstitutional in recent decades.

Marketing at one time was not treated as First Amendment speech at all.54 The Supreme Court began to extend First Amendment protections to commercial advertising in the 1970s, and ultimately settled on an approach that required the government to satisfy intermediate scrutiny when regulating truthful, nonmisleading advertising for products and services that were not themselves illegal.55 In the past decade, however, the Court has appeared to inch toward treating advertising as fully protected. The Court’s 2011 opinion in Sorrell v. IMS Health drew significant attention for describing a limitation on the use of personal data for marketing purposes as a viewpoint-discriminatory law that targeted “speakers and their messages for disfavored treatment.”56

Labeling and disclosure requirements have recently come under particularly close scrutiny. In 2018, in NIFLA v. Becerra, the Supreme Courtheld that the First Amendment applies with full force to representations by professionals in highly regulated industries, and that even purely factual disclosure requirements can trigger strict scrutiny if they relate to “controversial” public policies.57 Here,the Court struck down a California law that required “crisis pregnancy centers” to provide patients with factual information regarding the availability of contraception and abortion services.58 Justice Thomas, writing for the majority, reasoned that

when the government polices the content of professional speech, it can fail to “preserve an uninhibited marketplace of ideas in which truth will ultimately prevail.” Professionals might have a host of good-faith disagreements, both with each other and with the government, on many topics in their respective fields. Doctors and nurses might disagree about the ethics of assisted suicide or the benefits of medical marijuana; lawyers and marriage counselors might disagree about the prudence of prenuptial agreements or the wisdom of divorce; bankers and accountants might disagree about the amount of money that should be devoted to savings or the benefits of tax reform.59

The Court’s hostility toward mandatory disclosures, together with its announcement that bankers’ “disagree[ments] about the amount of money that should be devoted to savings” are as sacred for First Amendment purposes as political debate, suggest that brokers and retail investors are well positioned to challenge the laws that govern their business.60 Justice Breyer underscored the implications of this turn, noting that the framework for professional speech set out in NIFLA, “if taken literally, could radically change prior law, perhaps placing much securities law or consumer protection law at constitutional risk.”61

For the SEC, the risk of constitutional deregulation extends beyond the Supreme Court to the D.C. Circuit. That court’s decisions are important and salient to the SEC, especially because of the agency’s programmatic interests in broker-dealer regulation.62 And the D.C. Circuit has been foreshadowing the possibility of closer scrutiny of securities law under the First Amendment since striking down the SEC’s “conflict mineral” disclosure rule in 2015.63 That case illustrated the stakes of First Amendment litigation risk in designing regulatory programs.64

In this environment, it seems unlikely that courts would extend to confetti regulation the kind of automatic deference securities regulations have received in the past. Any opinion invalidating such a regulation would mark the continuing erosion of securities law’s historically exceptional treatment under the First Amendment. Even an opinion that focused entirely on the speech status of software rather than the speech status of securities communications as such would demonstrate that the securities laws are vulnerable to First Amendment attacks.

It is easy to see what such a decision would look like. Suppose, for example, that the SEC adopted a rule prohibiting gamified design features—such as confetti, push notifications, and other “behavioral stimuli” that encourage trading—on the grounds that the prohibition was part of the broker’s duty of care, that it was in the public interest, and that it was for the protection of investors.65 Audiovisual content usually counts as speech, even if the message conveyed is ambiguous or thin.66 So, too, do videogames and software.67 Stimuli that are part of the user-interface design might therefore be characterized as being within the scope of First Amendment protection as well.

From here, once gamification is framed as falling within the First Amendment’s protection, it seems all too easy to challenge a ban on that speech as one that discriminated on the basis of content or even viewpoint. Confetti in an investing app might be read to endorse trading, or perhaps day trading, as a good thing—and a regulation banning confetti in trading apps but not in other apps might be said to single out the pro-trading “message” for suppression. Or, even more simply, a ban on displays of confetti might be read as a ban on depictions of confetti, which are a kind of content in their own right.68 If a court were to hold that confetti regulation is content or viewpoint discrimination, it would presumably apply strict scrutiny.69

Securities lawyers who are acoustically separated from the technology bar—and the techno-libertarian “Californian ideology” that surrounds it70—underestimate these admittedly formalistic and silly-sounding arguments at their peril. Whether under the First Amendment or Section 230 of the Communications Decency Act,71 it is routine in technology litigation to characterize controversies involving technology as implicating speech—often in abstract and unintuitive ways.72 This kind of litigation has produced holdings that computer source code is speech,73 that search results are akin to media editorial choice,74 and that an online marketplace is immunized as a “publisher” for purposes of third-party tort liability.75

It may have been reasonable at one time to expect these concerns to largely drop away if securities regulators were the ones dictating elements of software design.76 The SEC’s customary jurisdiction over securities-related information may have shaped courts’ and litigants’ views of the salience of the First Amendment and afforded the Commission a wider constitutional berth than, say, the Consumer Product Safety Commission would have if it tried to regulate videogame design. But we think it would be unwise for the SEC to expect that kind of solicitude today.

Confetti regulations’ novelty, combined with the definitional difficulties discussed above, will invite First Amendment challenges. Those challenges, in turn, may tee up opportunities for courts to confine the scope and strength of the SEC’s policy mission through constitutional deregulation. A court’s willingness to apply heightened scrutiny against a confetti regulation could invite more daring raids against the securities laws’ core information controls, such as the Quiet Period in initial public offerings.77 At worst, a court may condemn large swaths of securities law as paternalistic and incompatible with the First Amendment’s presumed market-fundamentalist commitments.78

III. implications

In this Part, we discuss the implications of our argument for regulatory interventions against gamification in stock-trading apps. Securities law should avoid attracting unwelcome attention by courts engaged in a project of constitutional deregulation. We therefore urge regulators to think of gamification—and other digital-engagement practices more broadly—in terms of well-grounded legacy doctrines like churning and the duty of quantitative suitability that go to reducing the conflicts of interest inherent in brokerage.

A. The Securities Laws and the First Amendment

In light of the First Amendment’s increasingly antiregulatory orientation where business interests are concerned, securities law’s historically light First Amendment coverage looks increasingly exceptional. “Securities regulation,” Roberta Karmel observed over thirty years ago, “is essentially the regulation of speech.”79 The days are gone when the D.C. Circuit might uphold disclosure requirements on the basis of “the federal government’s broad powers to regulate the securities industry.”80

What could happen if confetti regulation (or some other trigger) led courts to start treating the securities laws like other burdens on speech? In our view, robust expansion of the antiregulatory First Amendment to other traditional areas of economic regulation—like the securities laws—would be destabilizing and undesirable for its substantive effects on markets and its erosion of democratic control over the economy.

The securities laws use a number of prototypical regulatory tools like mandatory disclosure and restraints on fraudulent communications. But perhaps the most at-risk targets of constitutional raids are the securities laws’ restrictions on expressive and truthful commercial speech in the areas of professional advice and securities offerings. Regulation Best Interest (Reg BI), for example, codifies care and conflict-of-interest obligations of broker-dealers in making recommendations to retail customers.81 Meanwhile, the Securities Act of 1933 and its implementing regulations prohibit most truthful communications to prospective investors until the agency takes a triggering action on a registration statement.82 The general exception is when issuers comply with narrow content-based exemptions that purport to allow particular kinds of speech (as in the safe harbors during the Quiet Period before the effective date of a registration statement) or speech to a restricted audience (as in a private offering for which general solicitation is not allowed).83

In short, confetti regulation, as we have described it, would draw a potentially broad range of First Amendment attacks. Of course, the SEC could promulgate confetti regulations and seek to defend these in court. While it is risky business to predict what courts will do, the agency has had a poor track record in rulemaking and enforcement before the Supreme Court and the D.C. Circuit in recent years.84 With these courts attuned to perceived agency overreach, challenges to confetti regulation on First Amendment grounds might receive a welcome audience.

In one scenario, decisions vacating confetti-regulation rulemaking or enforcement proceedings on First Amendment grounds could erode courts’ historical recognition of the public interest in regulating speech in capital markets. That erosion would lead to sharper constitutional constraints on securities regulation’s disclosure and information-control provisions more generally.85 In another scenario, courts might reason about the First Amendment status of confetti in ways that would implicitly, perhaps inadvertently, elevate the First Amendment status of securities information. Suppose, for example, that an SEC-sympathetic court decided to upholdconfetti regulation as something akin to a time-place-manner law—the kind of law that does not discriminate against the content of any message, but merely regulates the mode in which the message is presented.86 That argument would nevertheless imply that there was a message in the underlying securities communications—and more to the point, that garden-variety securities communications lie within the realm of First Amendment protection.

First Amendment litigation may ultimately move the law in this direction no matter what. But securities regulators do have some control over the pace of change. Provocative incursions into the law of software will intensify the deregulatory barrage and accelerate the damage. We therefore suggest that regulators design policy with a goal of constitutional avoidance in mind—at least for now, while the First Amendment’s doctrinal trendlines look relatively threatening to economic policy.87

We are suggesting, in other words, that regulators kick the First Amendment can down the road. Regulators should address applicable harms from gamification through the familiar methods and techniques of securities law without creating a target-rich environment for these kinds of challenges and outcomes.

B. Behavioral Design and Regulatory Choice

What is left after confetti regulation is taken off the table? Securities law already offers rich doctrinal frameworks and normative principles for addressing potentially objectionable behavioral design in retail-investing apps. In our view, a pair of traditional doctrines—the prohibition against churning, and the “quantitative suitability” component of the broker’s duty of care—illustrate securities law’s normative concern that eliciting overtrading in a retail investor’s account is undesirable where it leads to capital losses or principal depletion. These doctrines are not specific to behavioral design, but they do capture a large share of what is troubling about it.

Churning occurs when a broker-dealer “seeks to maximize . . . remuneration in disregard of the interests of the customer,” such as where a broker with discretionary control over an account trades excessively to generate commission revenue.88 Zero-commission investment apps with gamification features promote the same kind of overtrading that was the core harm at issue in churning, but in a self-directed account. Even without commissions, the revenue model generates the same result: the broker maximizes PFOF revenue from other market intermediaries who want to trade against retail investors.89

Gamification can thus be understood as a means for the broker-dealer to maximize revenue by driving unsophisticated retail investors to overtrade. This strategy, which we call behavioral churning, exploits behavioral psychology to drive engagement with the platform, increasing consumption of high-volatility speculative trading in ways that produce a discreet but often sizable stream of revenue for the broker.. In this view, behavioral churning provides a framework not only for scholarly work in this area, but also for potential regulatory responses.90

Churning doctrine itself historically applied where brokers had discretionary control over trading in the client’s account. So to address the harm from behavioral churning in client-directed accounts, regulators might look to quantitative-suitability doctrine. The SEC codified this doctrine as a component of the broker’s duty of care under Reg BI.91 Under that duty, broker-dealers must have a reasonable basis to believe that a series of recommended transactions—considered together—is not excessive in light of the retail customer’s investment goals, and does not put the broker’s financial interests ahead of the customer’s.92 To the extent that gamification features fall within the definition of “recommendations” to retail customers, then the Reg BI duty of care would prohibit a business model that encourages behavioral overtrading to generate PFOF revenue without regard to whether that level of trading activity is in the customer’s interest.

Regulators and scholars would have to grapple with a number of objections to the quantitative suitability approach to behavioral churning. For instance, the Reg BI duty applies not to self-directed trades, but only to the broker’s recommendations.93 When can gamification objectively be understood as a kind of “recommendation”—a malleable concept roughly meaning a call to action that influences a trade decision—based on tailored and individualized advice?94 Some design features by their terms express a call to action, like a push notification sent to new users who had not yet traded in their account: “Choosing stocks is hard. Get started by checking which stock prices are changing the most.”95

Regulators have warned the brokerage industry about digitally mediated recommendations for decades. In 2001, FINRA’s predecessor issued a notice, approved by the SEC and having force of law, about online communications that would generally be “recommendations.”96 Two of the examples were “customer-specific . . . pop-up screen[s],” and lists of securities for which the broker makes a market.97 In this way, securities law has previously concerned itself with the antecedents of behavioral churning—and currently frames it as a recommendation in violation of the quantitative suitability component of the Reg BI duty of care.

But even this kind of theory implicates the First Amendment concerns we have articulated. Professional-advice speech like this is not obviously less “expressive” than a flurry of confetti, so constitutional risk remains a factor. A strategy of avoiding constitutional deregulation would counsel toward adopting or enforcing existing securities laws in ways that do not turn factually on the “speech” embodied in behavioral-design features.

Conclusion: Toward A Theory Of Behavioral Design And The Securities Laws

We have offered a preliminary sketch of the problem of gamification as behavioral churning. But it raises a number of theoretical, empirical, and regulatory-design implications. Given the scope of this Essay, we only briefly address them here.

We have assumed, as a normative matter, that it is appropriate to regulate behavioral design in zero-commission investing apps. There are other plausible theoretical justifications for doing so besides the “problem use” harm—such as their tendency to promote imprudent investing practices and their macroscale effects on asset allocation and market quality.98 But if behavioral churning is an adequate and settled basis for regulation, do these additional theories add, at the margin, any justificatory value or new objects for regulatory choice?

Gamification raises other important questions for securities-regulation theory. Consider two of the securities laws’ core aims: promoting competition and protecting investors.99 These aims are somewhat in tension. Competitive pressure may channel innovation toward attractive user-experience design that extracts a long stream of small payments on nonsalient product attributes.100 How should securities law weigh its normative goals with respect to that outcome?

In addition, some investors engage in maladaptively excessive trading as consumption of sensation or risk. But it does not necessarily follow that securities law should be designed to support (or hinder) that kind of trading. Rather, the desirability of regulatory interventions specifically targeted at retail trading behavior will depend on our view of the normative end goals of securities law’s “investor protection” regulatory mission.101 If those goals include encouraging responsible investing, regulators might even grow to appreciate prosocial or “white hat” gamification—akin to nudges that attempt to intervene in behavior with carefully designed defaults.102 SEC Commissioner Hester Peirce, for instance, has expressed optimism about that prosocial use of financial-regulatory technology.103

As a matter of regulatory design, we have focused only on one harm and one regulatory solution. The question of regulatory technique is more complex. Other factors besides the problem use harm may bear on the desirability of regulating gamification, given the trade-offs and constraints we have identified in this Essay. Might other harms be better addressed through other regulatory techniques?

In our preliminary view, the most politically salable and administratively simple approaches will tend to involve the greatest litigation risk from deregulatory constitutional challenges. At one extreme, banning PFOF would require rulemaking and probably inspire a consumer backlash by making the zero-commission model infeasible. But it would not provoke any conceivable First Amendment challenge. At the opposite end of the spectrum, regulators might bring enforcement actions under existing rules against firms that throw confetti following a trade. This technique could launch a whole quiver of not-quite-frivolous First Amendment arguments, some of which may well hit their mark.

Assistant Professors of Law, University of Nebraska College of Law. For helpful conversations, we thank Josh Braver, Jacob Bronsther, Eric Chaffee, Jake Charles, Jill Fisch, Gina-Gail Fletcher, Talia Gillis, Alan Kluegel, Guha Krishnamurthi, Da Lin, Lidiya Mishchenko, Alex Platt, Shalev Roisman, Barbara Roper, Steve Schaus, David Simon, Will Thomas; participants at the National Business Law Scholars Conference and Consumer Federation of America conference; and participants at the Junior Scholar Workshop Series and Research Accountability Group workshops. We especially thank Casey Dodge and Alan Dugger for their excellent research assistance. The authors acknowledge the support of McCollum summer research grants in writing this Essay.