Indivisibilities in Technology Regulation
Lee Fennell’s Slices and Lumps: Division and Aggregation in Law and Life reveals the benefits of isolating configurations in legal analysis. A key characteristic of configurations, or “lumps” whether found or created, is that they are indivisible. To say a lump is indivisible is not to say that it is literally impossible to divide, but rather “that it is considerably less valuable when divided, or that it is expensive (perhaps prohibitively so) to divide successfully” (p 11).
This Essay will extend Fennell’s approach to indivisibilities to the context of technology regulation. Fennell discusses at least two types of indivisibilities in the book. I will call these indivisibilities of fact and indivisibilities of law. Indivisibilities of fact are facts about the world that make it difficult to divide up a resource in ways other than predetermined lumps. Indivisibilities of law are outcomes at law that are relatively “all-or-nothing.” Indivisibilities of both types are at play in current issues in technology regulation.
With respect to indivisibilities of fact, this Essay will discuss the example of indivisibility of privacy regulation. Some argue that piecemeal, sector-specific privacy regulation is the same as no regulation at all due to realities of the technosocial environment. This comes down to a debate about the degree to which the level of consumer privacy—a fact about the world—is indivisible.
With respect to indivisibilities of law, this Essay will discuss the example of consent in the law of adhesion contracts in the digital age. Whether there is consent is a binary distinction, with major implications at law. Some consumer advocates have argued that consent should be segmented into meaningful consent and less meaningful consent. But, perhaps, the concept of consent is indivisible. Whether or not consent can be understood as divisible—a characteristic of the law—has major implications for this area of law and policy.
This Essay will proceed as follows. First, I will distinguish between indivisibilities of fact and indivisibilities of law. Then I will discuss an example of each in the field of technology regulation. In discussing each example, I will suggest how evaluation of regulatory choices in response to technosocial change could be enriched by this distinction.
I. Indivisibilities of Fact and Indivisibilities of Law
Slices and Lumps centers around the broad concept of lumpiness, but this Essay will focus on one particular slice of lumpiness: indivisibility. Indivisibility is so fundamental to lumpiness that Fennell at one point calls indivisible a “synonym” for lumpy (p 12). I am less confident that the two are pure synonyms as used in the book, so for clarity purposes this Essay will use the term “indivisible.” To continue, indivisibility is a characteristic that makes it difficult to divide a resource. Fennell gives two broad reasons why a resource may be indivisible: expense and complementariness. First, a resource may be prohibitively expensive and difficult to divide. The example she gives is a child, in an allusion to the biblical tale of the judgment of King Solomon (p 9). Fennell also describes expense-based indivisibilities in employment contexts, observing:
[T]he fixed costs of hiring and training make it infeasible to recut the work into smaller servings and distribute it to more workers. For example, if each new hire requires the employer to lose money upfront for months or even years before positive returns can be realized, then the employer will want to incur that cost as few times as possible (p 121).
The second cause for indivisibility is that some resources have much more value when clustered together. So even though it may be procedurally simple enough to separate the resource into parts, few would wish to do that. Her primary example of this is a pair of shoes, which give more than twice as much value together as individually. Fennell goes on to give other examples of complementariness, noting that the “segments that make up a full bridge span are strongly complementary; subtract just one, and the bridge becomes useless. A partially fenced yard does no better than an unfenced yard at containing animals, a car with three tires drives no better than a car with no tires, and small and scattered patches of land are useless for large-scale development.” (p 11).
Indivisibilities can impact negotiating positions between individuals in society. When a provider is supplying an indivisible good or service, midway through the provision, the provider has increased bargaining power because while the provider has undisputedly provided value, the value to the buyer is low until the good or service is completely provided (p 21). She states:
More broadly, indivisibilities present the potential for contracting parties to apply leverage to each other. Renovations, auto repairs, medical procedures, and many similar services exhibit indivisibilities that make it difficult for consumers to readily switch to a competitor midway through. Information asymmetries may also make it difficult to know whether an announced change in price as the work progresses represents a strategic ploy to exploit the leverage provided by the lumpy situation or simply a response to new information that has been uncovered in the earlier phases of the work. (pp 22–23).
Fennell gives the example of James Gandolfini’s successful holdout in producing episodes of the Sopranos.
Fennell describes a series of ways that government or other dispute-resolving actors may overcome indivisibilities in resolving competing claims. These methods include joint custody of the good; converting the good into money, which can then be proportionally divided; “giving claimants [lottery] chances at the good that are proportionate to the strength of their claims”; or “giving the good to one claimant while compensating the others.” (p 10).
The book describes both indivisibilities of fact and indivisibilities of law. The difference between the two is the substrate upon which the indivisibility acts. To some extent, indivisibilities of law may be said to be reducible to indivisibles of fact: social, or even psychological, limitations on how humans will process concepts. The distinction is still useful, as the subsequent two sections will show, even though it may collapse at the margins in theory. While Fennell does not emphasize this distinction in the book or use this particular terminology, the distinction is implicit in her argument. In addition to the examples I highlight below, she refers to goods, services, and law all as types of “lumps,” implying that all of these things are subject to indivisibilities in the way they present. (pp 21–25).
In a dispute over which of two individuals who have claim to a homerun baseball, Fennell called the ball an “indivisible good.” (p 190). The ball’s indivisibility limited the ways in which the dispute between the two individuals could be resolved (p 10). Few collectors would want half of a baseball because it could no longer be thrown, would be difficult to display, and would present unique preservation challenges. As a result of the forgoing, the half-ball would have limited resale value. This is a paradigmatic illustration of indivisibility in the sense this Essay will use it: not literally incapable of being divided, but division is so undesirable that different solutions to disputes should be favored.
Fennell also speaks of “indivisible entitlements” at law (p 191). She notes that at several node-points in law, results are indivisible. She provides the examples of a defendant being guilty or not, liable or not, or subject to specific performance or not (pp 25, 190). The court’s result is indivisible. Saying a court found a defendant partially guilty with respect to single charge makes no more sense than to say a person is a little bit pregnant. She notes that in creating indivisible outcomes at law, the law can choose to create liability cliffs, legal rules where “small differences in behavioral inputs can yield large differences in liability, [in contrast to] a more modulated set of legal responses. This liability cliff question goes to the lumpiness of legal consequences. Where legal consequences are binary, a lot rides on what goes into the on-off judgment” (p 191).
Fennell’s analysis collapses indivisible liability outcomes and indivisible legal consequences. Yet separating the two is instructive as to how the legal system handles indivisibilities of law. Liability outcomes may be all-or-nothing, but conditions may be placed on the legal consequences. This can lead the legal consequences to be more granular in spite of the indivisibility of liability. This is generally the role of the law of remedies. For example, when a defendant is found to be liable in tort, the damages the plaintiff is awarded may be nominal, compensatory, or punitive. If only nominal damages are awarded, even though the holding was contrary to the defendant’s wishes, the defendant is relatively closer to “winning” than she would be if she had to pay more in damages. Similarly, the sentencing phase of a defendant’s trial can be as significant for a defendant as the judgement itself. Sentencing impacts whether a guilty defendant spends time in jail or, on the more severe end, faces the death penalty. This provides gradation in outcome even though the liability result is binary.
Creating a legal consequences cliff does not merely involve the question of whether an indivisible liability choice exists. To create a legal consequences cliff, the law surrounding the binary rule must reinforce the liability cliff. As the previous two examples show, remedies and sentencing can be used, in civil and criminal law respectively, to create gradation in legal consequences even while liability outcomes are indivisible.
In summary, indivisibilities of fact exist where there are indivisibilities with respect to facts about the world (that is, a ball cannot be divided), and indivisibilities of law exist where there are indivisibilities with respect to legal results (that is, a defendant cannot be both guilty and innocent of the same crime).
What is the relationship between indivisibilities of law and indivisibilities of fact, if any? Fennell suggests that where indivisibilities exist in fact, legal results that are also indivisible may do a good job in regulating them. She states, “Binary outcomes mesh well with the indivisible entitlements that are frequently at issue in legal disputes” (p 191). One wonders if the inverse may be true as well, though Fennell does not expressly contend this. That is, granular facts may do best with granular assessments of liability. This points towards a potential method of evaluating law through assessing the correspondence and interplay between the indivisibilities of fact and indivisibilities of law. It may be an analytical ground for choosing one rule over another.
The next two parts will illustrate the concepts I have discussed through examples from law and technology.
II. Indivisible Privacy
The technosocial environment may render information privacy an indivisible resource. I will use this as an example of the indivisibility of fact framework.
Many commentators have argued that there is a suboptimal level of consumer privacy in society. Consumers routinely report in surveys that they feel that they do not have sufficient privacy and report that they take active steps to protect their privacy. As a result of this preference, companies compete to create the impression that they are privacy friendly in their advertising. The frequent Federal Trade Commission deceptive practices actions against companies for misrepresenting their privacy practices as more extensive than they are show a darker side of the character of competition: it is based on perception, not substance. This is why the Federal Trade Commission, an agency in part dedicated to promoting free markets, has taken a prominent role in regulating privacy. Few regulations limit data processing and sale in the United States and those that exist are sector-specific, regulating data processing for specific industries, such as banking, or particular people, such as children. By contrast, there is extensive regulation of data processing and use in the European Union, led most recently by the General Data Protection Regulation.
I do not intend here to defend the thesis that more privacy protection is desirable. Instead, my goal is to consider whether the overall level of information privacy in society can be regarded as indivisible. This, in turn, limits the regulatory approaches that would be effective in the event that state or federal government pursues regulation to protect consumer privacy and cybersecurity. A threshold issue may be whether each individual can determine her own level of privacy protection. If each individual can express her privacy preferences in the market, an argument can be made that regulatory intervention is unnecessary, and that there is no need to interrogate the overall level of the resource provided. This relates to the thorny question in the literature regarding whether privacy policies are contracts. If privacy policies are regarded as contracts, there is at least the formal argument that each person who uses a device or application has chosen her individual level of privacy. However, due to the complexity and mutability of data processing, most consumers cannot distinguish between products on this basis, and this leads to a “market for lemons” situation with respect to the market for privacy.
Indivisibility can be due to complementariness or expense and complexity. I will briefly apply each concept to this case.
A low level of information privacy and the presence of digital economy services are probably not complementary. Complementariness is when an item has less value when separated from its constituent components. Data privacy loss is associated with information economy resources. As more devices play more prominent roles in the lives of consumers, so too are there increased opportunities for data collection and processing. But it does not seem that data insecurity is a necessary part of why these services are valuable. Email services, internet search platforms, and social media websites have arisen with stronger privacy and cybersecurity protections. These applications have not been able to compete with more widely used services, but they prove the concept that weak privacy and cybersecurity protections are not fundamental to providing information age services. Lacking privacy or cybersecurity provisions does not make a service inherently better. In fact, one of the primary reasons for the market’s failure is because of the invisibility of the features that make an application more privacy protective or data secure. Saving costs on privacy or cybersecurity may allow for more resources for research and development or perhaps a product reaching market more quickly. That is not the same as complementariness, which has it that that Thing A (lack of privacy protective policies) itself adds to the value of Thing B (provision of information economy resources), and vice versa, and that the two are worth more together than apart. It is not apparent, nor even contended by skeptics of regulation, that this is the case.
It is probably expensive and complex for individual companies to provide a greater level of information privacy while also providing digital economy services in the current economic and regulatory environment. This is a possible source of indivisibility in the social allocation of information privacy. While current conventional industry wisdom states that privacy and cybersecurity measures should be considered throughout the development of a new product, growth and market share imperatives in the technology industry mean that businesses seeking to disrupt or create a market for a service tend not to prioritize privacy and security. What is more, modern applications and devices often are reliant on several third-party services to function. Even if a company prioritizes privacy and security in its own services, it is difficult to guarantee that all parties they work with follow the same principles. Given that the majority of consumers do not have the technical knowledge to tell the difference between privacy and security-protective goods and services, providers have every incentive to make available services that are insecure and non-privacy protective. Given that any provider would be competing with such unscrupulous firms, it makes it expensive not to do the same, because the privacy-protective provider would be taking on costs and processes that the other would not without receiving compensation for it on the market. As Paul Ohm put it, “Privacy seems to be a market for lemons where promises are easy to make and quality is difficult to inspect. As with all such markets, there seems to be little incentive to compete for privacy.”
The significance of a finding that the overall information privacy level is indivisible due to expense and complexity would be to limit the types of regulatory options by those seeking to raise the level of privacy. In order to raise the societal level of information privacy, either the underlying market failure would need to be corrected or mandatory privacy and security policies would need to be imposed across the board. These are both major moves that are not sector-specific and would be difficult to do on a state-level basis. The indivisibility argument is one way that privacy advocates could argue that major reform, rather than incremental reform, is necessary. In a similar vein, Fennell argues, “Thinking about poverty relief from the perspective of functional adequacy—the lumpy amount needed to get from a bad situation to a sustainable one—can alter the discourse around welfare policy” (p 118). This would also counsel against seeking episodic, small victories along the way because, if the level of information privacy is indivisible, such steps would waste political capital while not raising the overall privacy level. It would also support the argument, which has been made in the literature on at least one occasion, that privacy is a public good.
On the other hand, this framework also offers a path for privacy regulation skeptics to contest this type of reform. If they can show that incremental improvements in privacy protections are not prohibitively expensive for individual market actors, there is no need for wide-ranging reform.
III. Indivisible Consent
Consent may be an indivisible concept at law. The factors leading to finding an indivisibility of law differ from the factors determining an indivisibility of fact. The rules of thumb of complementariness and expense, which are key factors for finding indivisibilities of fact, do not apply when it comes to finding indivisibilities of law. The process of finding an indivisibility of law that Fennell follows is more direct—when a legal rule takes a binary form, we say that the law is lumpy, or indivisible. For instance, she writes, “Legal rules and litigation outcomes may also exhibit lumpiness, operating in an all-or-nothing fashion, or producing results only when some threshold of compliance or deterrence is reached” (p 8). There are several known rules of this type in American law. For example, as Part I discussed, a criminal defendant party is either guilty or innocent of a given crime, and civil defendant is either liable or not for a given claim. For historical and professional cultural reasons, it is unlikely for courts and lawmakers to deviate from the binary structures of these rules.
The point I wish to make here is that the concept of consent may be as indivisible, or binary, a concept at law as guilt. As a result, consent may be resistant to division into sub-categories such as “informed consent” or “meaningful consent” versus some lesser form of consent. This in not to say this it is impossible to introduce sub-categories. Informed consent is an important part of health law, for example. Yet courts have largely resisted a large chorus of scholarly commentary suggesting that meaningful consent, rather than mechanical, non-thoughtful consent, should be the standard for enforcing consumer form contract terms. For reference, Nancy Kim outlines the insensitivity of current contract law doctrine to concerns about meaningful consent in Wrap Contracts: Foundations and Ramifications (pp 93–126) and Brian Bix discusses theoretical approaches to contract formation that require meaningful assent in Contract Law: Rules, Theory, and Context (pp 137–38).
Whether or not a modifier like “meaningful” is added, consent takes a binary form. There is either a finding of consent or not, for a given term. The fact that consent takes this binary form may complicate attempts to render it more granular by introducing modifiers create different levels of liability.
If consent is indivisible, its indivisibility limits the regulatory choices available for those who believe that consumers need protection from exploitative terms in form contracts. As my analysis in Part I showed, indivisible liability findings do not necessarily lead to indivisible legal outcomes. Consent is not a finding of liability, but as modern courts use it, where consent is found, a properly formed contract will likely be found as well.
One option, already available at law, is limiting remedies available for breach of contract in this context. Another option is to increase the prominence of the unconscionability defense to contract enforcement and thereby finding that, notwithstanding consumer consent, a defense prevents the enforcement of a contract when the terms are substantively or procedurally unfair. Finally, as I have argued elsewhere, the law could imply fiduciary duties to firms in consumer contracts with consumers. What all three approaches have in common is that they do not contest or modify the indivisibility of consent.
While the goals of imposing a blanket standard of meaningful consent in consumer transactions are laudable, understanding that consent is an indivisible concept at law explains why this approach has not succeeded in the courts, and why other options to limit the impact of the indivisible consent on legal outcomes may be more successful.
Conclusion
Attention to indivisibilities of fact and indivisibilities of law can have salutary effects on legal analysis. This Essay has shown the illuminating effect this perspective can have on addressing information-age problems. It adds an additional analytical method to the lawmaker’s toolbox for choosing a legal rule. Indivisibilities should influence the selection of regulatory possibilities by observing that indivisibilities of fact and law may make proposed rules less effective, for reasons unrelated to their theoretical desirability.