Lumpy Work

For close to ten years, the gig economy has dazzled with its seeming powers of disaggregation. Want a picture mounted on your wall but no other decorating, framing, or assembly help? There’s an app for that (TaskRabbit). Want a place to stay during your holiday but without the frills of turn-down service, room service, and concierge service? There’s an app for that (Airbnb). And of course, if you want a simple means of getting from Point A to Point B with only one intended destination (yours) but without the obligation to call someone, direct someone, or (until quite recently) tip someone, there’s an app for that too (Uber, Lyft). In many ways, gig work has quite radically configured the way we consume services by disaggregating the component parts of many jobs and allowing customers to select from a suite of options that, while not infinite, is greater and very often more flexible than the options that were available before.

Nevertheless, as Lee Fennell points out across a range of examples in Slices and Lumps, disaggregation and reconfiguration are far from self-evidently good. Sometimes, productively slicing up an exchange for one party (say, the customer) doesn’t produce a similarly productive disaggregation for the other party (the service provider). Let’s call this the “parity problem.” At other times, slicing up an exchange highlights or exacerbates negative spillover effects that the relationship has on seemingly unrelated areas. Borrowing Fennell’s own language, we’ll call this the “interaction problem.” Labor and employment law scholars have been exploring these types of issues with respect to gig work for several years now by discussing the relationship between gig labor and existing laws regulating work. However, Slices and Lumps suggests that challenges like the parity problem and the interaction problem exist well beyond the gig economy or even the realm of work regulation: they (and several related challenges) are a feature of modern life writ large, where configuration, according to Fennell, “is a power that has become increasingly pressing to understand and harness”(p2).

This Essay teases out some of the implications of Fennell’s wide-ranging discussion, extends some of her analysis, and underscores some ways in which her work and existing scholarship complement one another. I will focus on Fennell’s observations regarding reconfiguration and work—especially gig work—which means that, for the most part, I will draw on Chapters Seven through Nine of Slices and Lumps. Part I of this Essay offers a brief explanation of what “gig work” encompasses, while Parts II.A–C take up different aspects of Slices and Lumps as they relate to gig work and labor and employment more broadly.

I. Honeycombs, Circles, and Triangles: Defining the Boundaries of the Gig Economy

Early attempts to define and name this new space of technology-mediated exchange frequently relied on visual analogies. Jeremiah Owyang used a honeycomb to refer to the way in which these new exchanges were “structures that efficiently enable many individuals to access, share, and grow resources among a common group.” Seth Harris and Alan Krueger used triangles to describe the way that service providers relate to consumers on the one hand and the companies themselves on the other hand. I have used overlapping circles to argue that only some exchanges actually carried implications for labor and employment regulation.

As a result of these and other definitional efforts, labor and employment scholars today largely focus on the same set of companies: those that connect individual providers with individual consumers in order to provide a service in a way, at a time, or for a price that is not readily available on the market. The terminology still varies: many commentators have now switched from the earlier “sharing economy” to “gig economy” on the grounds that (as Fennell points out, p 141) these models are market-based, not voluntaristic. For ease of reference, I will refer to the broader set of exchanges as the “gig economy” and the companies that facilitate those exchanges via new technology as “platforms.”

Although Chapter Seven largely focuses on the exchange of services and Chapters Eight and Nine emphasize the exchange of goods or resources, Fennell does not artificially assign specific platforms to one category or another—and for good reason. As I’ve previously argued, conversations about the gig economy’s implications for work regulation do not benefit from overly rigid distinctions between “labor” platforms (like Uber or TaskRabbit) and “capital” platforms (like Airbnb). Labor platforms still depend heavily on capital, whether it is material or intellectual—you can’t drive for Uber without a car, or succeed on TaskRabbit without monetizable skills. Additionally, capital platforms can involve significant emotional and physical labor, like conveying attentiveness and responsiveness to guests or scrubbing bathrooms and restocking coffee supplies. In a very real sense, all of these are traditional service relationships that have been sliced up or otherwise reconfigured, and as such they raise concerns with which labor and employment lawyers have been grappling for years.

II. Lumpy Work, All Sliced Up

Fennell explores a range of benefits and complications brought on by the gig economy’s reconfiguration of service exchanges. This Section will consider just a few of those consequences. In some instances (like the “interaction” problem and the “parity” problem), Slices and Lumps offers a new vocabulary with which to think through widely discussed challenges affecting gig workers. In other instances, the new perspective offered by Slices and Lumps—its focus on configuration rather than labor and employment—highlights and reframes specific phenomena that have lurked in the background of earlier, more work-focused discussions.

A. The “Interaction” Problem

The last of the ten brief “lessons” offered by Slices and Lumps—and the place where most labor and employment scholars begin their discussions of gig work—is to Identify Interactions between changing work configurations and how they can “impact configuration issues in other arenas or create spillovers for seemingly unrelated decisions” (p 232). Most importantly, as Fennell notes, “unbundling the job into granular work units may also unbundle the job from the legal protections and benefits that typically accompany employment” (p 120). In fact, this kind of unbundling is something that platforms count on.

The United States funnels an extraordinary range of benefits and protections through work and most of these goods only flow to workers who are classified as employees. Anti-discrimination protections, certain leave and retirement protections, wage and hour guarantees—all of these and more are legal protections that workers miss out on if they are classified as independent contractors. From an employer’s perspective, reclassifying workers as independent contractors instead of as employees makes them less expensive and less administratively cumbersome, and it also facilitates more interesting slicing and dicing of the work relationship. Whatever the case may be in other areas like housing or poverty relief, as far as work regulation is concerned, the “interaction” effect that new, slicing-oriented business models have on labor and employment protections is an intended feature, not an unforeseen bug.

Three things are worth noting as far as the interaction problem is concerned. First, the problem itself is by no means limited to gig work: losing work protections by virtue of losing employee status is an old problem and one that has plagued many of the industry verticals now led by prominent platforms (like transportation-for-hire and delivery services). Taxi companies, for example, reclassified their drivers as independent contractors in the 1970s. Similarly, FedEx has been involved in a long-running battle across state lines over its classification of delivery drivers as independent contractors. And as these examples may suggest, jobs can be “unbundled” from the standard suite of labor and employment protections without any corollary disaggregation of the jobs themselves.

Second, Slices and Lumps is agnostic as to whether the unbundling of work protections from work tasks is an inescapable effect of gig economy–style slicing. This makes sense, since labor and employment commentators themselves disagree about whether slicing up work tasks inevitably leads to slicing up or forfeiting work protections. Some protections that are funneled through work, like minimum wage guarantees, clearly need not be lost despite platforms’ arguments that their business models simply do not allow them to fulfill established standards. Other protections—most significantly, health insurance coverage—have in recent years become somewhat detached from the employee-employer relationship in ways that are not specific to the gig economy. Even labor organizing, which in the United States is not only grounded in employee status but is strongly associated with the physical worksite, is beginning to disaggregate and reconfigure itself. Now, movements like Fight for $15 or various Uber driver strikes are organized across far-flung locations, despite the lack of worksites, and target society and government more than the actual employer. Finally, some work law scholars argue that it is possible and desirable to totally unbundle labor and employment protections from the singular employer-employee relationship: instead of one employer responsible for providing all benefits we might have many employers, each responsible for providing some protections.

Third, a point Fennell makes in Chapter Eight with respect to lumpy housing standards—that there may be good cause to give one party a take-it-or-leave-it (TIOLI) choice between supplying an acceptable option or none at all—is equally valid in the employment context. In fact, Fennell explicitly references minimum employment standards (p 163) but the reference invites some elaboration. Areas of law often carry their own skeletons-in-the-closet, cases or statutes responsible for enshrining principles that are now (and were also frequently then) abhorrent: Dred Scott for citizenship, Buck v. Bell for reproductive rights, and so on. A solid contender for that dishonor in labor and employment law (though it is formally a contract case) is Lochner v. New York, which effectively held that employers could impose TIOLIs on workers but that the state could not impose TIOLIs on employers, at least when it came to the maximum hours of work. Fennell’s cautionary note echoes the concern of many labor advocates and work law scholars that slicing, even when it is possible and appealing, may carry disturbing interaction effects because of the way in which it dismantles some protective TIOLIs.

B. The “Parity” Problem

Lesson number one in Slices and Lumps—Mind the Lump (p 227)—encourages the reader to second-guess her likely intuition that inputs and outputs exist in a linear relationship to one another. This intuition has special significance in the labor and employment context, in which employers often argue that the disaggregation of lumpy jobs into finely sliced and therefore flexible tasks warrants a comparable shift in worker classification status from lumpy “employee” to sliceable “independent contractor.” However, workers and workers’ advocates often argue that the new job format is no more liberating than the one it replaced—“X widgets produced per day” is no more worker-friendly than “Y hours per day spent producing widgets.” In other words, there is often little parity between the job-slicing that an employer (or platform) engages in and the job-slicing that an employee (or gig worker) experiences, although employer rhetoric often suggests otherwise.

Fennell herself points to this phenomenon using a Dutch study in which workers were found to have engaged in significant overtime labor that was unpaid and unwanted even though their formal hours of work had been reduced pursuant to new regulations. The reason, according to the study, was “a new form of lumpiness . . . work itself comes in ‘lumps’ of tasks that require sustained attention and timely completion . . . [and so] workers’ own choices can sustain a gap between preferred hours and actual hours” (p 122). A rough analogy in the gig economy is the acceptance rate requirement articulated by many transportation network companies (TNCs) like Uber and Lyft: no driver is obliged to drive in a given hour, but any driver who chooses to drive must accept a certain percentage of available passengers within that hour or risk deactivation from the platform. Since passenger destinations are unknown at the time drivers accept ride requests, this means that drivers may find themselves driving at times, or for lengths of time that they would rather avoid—all despite the fact that they are not formally required to drive a certain number or set of hours.

Phrasing the problem this way provides a different perspective on a foundational doctrinal problem in labor and employment law: workers are mostly classified as employees or independent contractors based on the amount of employer control they are subjected to, but control is notoriously difficult to measure. What this means is that the “control test” often proves to be a poor sorting mechanism: courts cannot easily determine how a given worker ought to be classified and workers who are classified as independent contractors frequently behave, feel, and are treated like employees despite being ineligible for any of the benefits or protections associated with employee status. Moreover, although platforms regularly cast the inadequacy of the control test as a new problem triggered by the slicing capabilities of disruptive technologies, criticism of the control test’s sorting abilities has been longstanding and plentiful: a 1935 article in this very law review began by stating that “[t]he ‘legal literature’ about the independent contractor has been occupied, almost exclusively, with the question of how best to identify him.”

By drawing attention to reconfiguration’s non-linear effects, Slices and Lumps casts new light on something that labor and employment scholars have been saying for a while and that courts—if recent decisions on gig worker classification are any indication—may just be coming to realize: disaggregation is not synonymous with flexibility because there are different ways to experience both control and its opposite, freedom. The Uber driver who must select and complete an unknown ride in order to remain on the platform or the Dutch worker who must work overtime in order to complete a given task are, in one sense, choosing to perform their respective jobs at a time and in a manner of their own preference. In another sense, however, their choices are very much constrained by the output requirements and performance standards of the platform/employer. Unfortunately, doctrinal arguments, intensively fact-based analysis, and appeals to fairness do not appear to have moved courts and legislators significantly away from control-based analysis. Perhaps an approach that emphasizes aggregation and division and their inconsistent effects in the employment context will fare better.

C. On the Benefits and Changing Prestige of “Slack”

In 2017, the New York Times used the lives and trajectories of two janitors at top companies in the present and in the early 1980s to sketch a broader argument about rising inequality in America. The article noted that the two women, Marta Ramos and Gail Evans, “have each cleaned offices for one of the most innovative, profitable and all-around successful companies in the United States”—Apple (Ramos) and Kodak (Evans)—but that the spread of a new management theory explained why Ramos lacked many benefits that Evans had enjoyed. The management theory in question is that companies should “[f]ocus on core competence and outsource the rest.” Ramos, properly speaking, does not even work for Apple; she works for a contractor hired by Apple to perform the “non-core” but necessary task of maintaining its facilities. Evans, by contrast, was part of a company and a generation when large, cutting-edge employers had in-house employees who performed essential functions that were nonetheless unrelated to core competencies.

Lesson number seven in Slices and Lumps—Cut Yourself Some Slack (p 230)—speaks to this phenomenon in two ways. First, Fennell notes that while much of the excitement surrounding new technologies is based on their ability to maximize resource potential by cutting out wastage or “slack,” “the heretofore hidden virtues of maintaining some slack capacity must also be taken into account” (p 230). Consider one of the more ambitious ways in which platform-enabled labor is being used to cut slack from existing exchanges: the growth of on-demand publicly subsidized transportation. While public transit systems still overwhelmingly operate on a fixed-route basis using buses, subways, and light rail, a few municipal governments are beginning to experiment with TNCs as service providers. These arrangements appeal to public officials for the same reason they appeal to private consumers: they have the potential to expand access to services, to allow services to be offered on-demand, and to cut costs. They hope to achieve all this by eliminating or decreasing reliance on fixed route systems that have high startup and maintenance costs, and instead moving to a system where every vehicle that moves using subsidized dollars is moving in the performance of a requested ride. In other words, they aim to cut out the slack.

Although it’s too early to definitively weigh in on the success of these ventures—many are still in beta phases, some have shuttered, and others have yet to be implemented—one thing is already becoming amply clear: there were unanticipated benefits to having some slack in the system. Perhaps the biggest benefit is that the earlier fixed-route systems allowed for types of information gathering that are crucial for urban planning, like user and usage patterns, service shortcomings, and infrastructure needs. Having these benefits is not necessarily tied to having slack capacity but their absence in the new platform-mediated systems is directly tied to the elimination of slack. TNC platforms are able to finely match usage to need, among other things, because of their extensive and proprietary information collection and they are unsurprisingly averse to sharing that information. By choosing the leaner, sliced up service model municipalities have largely foregone a valuable side-effect of lumpy service.

Second, as the Kodak/Apple example highlights and as Chapter Eight of Slices and Lumps makes implicitly clear, “slack” has gone from being a marker of productivity and prestige to being inversely correlated with both. Modern consumers exhibit privilege by being able to buy small portions of “craft beef” (instead of whole cows or supermarket meat) or by hiring task-specific personal assistants (instead of, most likely, sorting their own mail or assembling their own furniture). Likewise, modern governments exhibit savvy by reconfiguring or paring down services so as to offer what is demonstrably wanted and no more. Nor is the prestige associated with eliminating slack limited to consumer behavior: it applies just as well to the supply side of the gig economy. The abiding appeal of platform-based labor lies in the chance to earn an income without sitting at a desk or engaging in the meetings and paperwork and office politics that may come with one’s preferred occupation—to have the beef, as it were, without having the cow, and to thereby live a fuller life or game an economic system that requires most workers to accept lumpy circumstances.

And yet, once again, slack sometimes has its advantages. For workers in particular, built-in slack “ease[s] the stresses of everyday life and help[s] people absorb various shocks”: the slow day at work that comes as you are preoccupied with a struggling child, or the option to take a sick day to care for an ailing parent (p 230). When there is no slack in the system there is no room to maneuver in the face of ordinary challenges. This is how gig workers (or Yahoo employees) end up feeling like they can never take a holiday. Slack can also be a necessary component of sequences—valuable, intertemporal lumps that “introduce variety, build on each other, or improve over time” (p. 134). Slices and Lumps discusses the most obvious and potentially slack-producing sequence in working life (wage growth), but the nonmaterial benefits of sequences, and consequently of slack, are also profound: social stature, personal growth, camaraderie. These too are potentially lost to fine slicing.

Conclusion

At a time when less—less wastage, less waiting, less red tape—is more, Slices and Lumps asks us to carefully consider the true effects of slicing, lumping, and generally reconfiguring aspects of everyday life. Many of the tensions it describes with respect to the gig economy and work regulation are familiar to labor and employment scholars but are often articulated in the dense or technical language of law. By reframing these considerations in the language of “slices and lumps” and by showing how the challenges of reconfiguring work apply equally to the reconfiguration of housing, social benefits, and consumption (among other things), Slices and Lumps offers a chance to approach the problem of lumpy work, and its transformation, with fresh eyes.