Stalling the State: How Digital Platforms Contribute to and Profit From Delays in the Enforcement and Adoption of Regulations

Abstract

While it is often claimed that the pace of digital transformation is such that its own, often glacial changes do not allow the state to catch up, we argue that technological companies, with the help of some state actors, have been slowing the state down. To capture this phenomenon, we introduce the notion of stalling strategies. We argue that stalling strategies have allowed digital platforms to create time that they have spent generating revenue and accumulating platform power, which later protected them from state actions. Drawing on a case study of Uber in Poland and a number of shadow cases, we distinguish five stalling strategies: reinventing classifications, dragging out court proceedings, stealing the time of street-level bureaucrats, delaying new regulations, and taking time to (not) comply. By analyzing stalling strategies, this article contributes to discussions about the politics of platform capitalism, the temporality of digitalization, and institutional drift.

Introduction

Why have digital platforms been able to grow so quickly, even in the institutionally hostile environments of heavily regulated markets? The expansion of digital platforms into social life has been taking place at an astounding speed. This is clear from the growing economic value of digital platforms such as Uber, Airbnb, Facebook, Google, and Tinder, but also from the growing number of social interactions that digital platforms now mediate (Kenney et al., 2021).

The speed of digitalization is perhaps best illustrated by the rise of Uber (Adler, 2021; Chan & Kwok, 2021; Calo & Rosenblat, 2017; Collier et al., 2018; Rahman & Thelen, 2019; Seidl, 2021; Serafin, 2019). In the course of less than 10 years, Uber has been able to successfully enter many heavily regulated taxi markets and turn from a small startup located in Silicon Valley into a global company with 14 billion dollars in revenue. Its business strategy has been based on fast and effective expansion into new markets, which is characterized by—among other things—its constant eagerness to highlight the number of cities in which it operates. But the strong emphasis on the pace of expansion is not unique to Uber. It is a characteristic of the business models of other digital platforms as well, as they aim to become “too big to ban” (Pollman & Barry, 2017, pp. 400–403). In order to be successful, digital platforms need to achieve a scale that ensures the effectiveness of linking various groups of clients (Evans & Schmalensee, 2010; Rahman & Thelen, 2019; Culpepper & Thelen, 2020).

The rapid growth of digital platforms should not be seen as self-evident. Like other technological innovations (Bijker, 2007), the growth of platforms has been a contested process (Schüßler et al., 2021). It is easy to forget now, but Uber was not always the powerful incumbent it is today, with its dominant position in most taxi markets around the world. It started as a small challenger that, although armed with the deep pockets of venture capital, was entering heavily regulated markets with powerful incumbent taxi corporations and entrenched rent-seeking interests of medallion owners (Collier et al., 2018, p. 921).

One explanation for the rapid growth of digital platforms can be summarized by Facebook’s motto “move fast and break things” or Uber’s “it’s better to beg forgiveness than ask permission” (Tusk, 2018, p. 109; Kirchner & Schüßler, 2020). The speed of digitalization, the argument goes, is related to platforms disrupting markets, acting too quickly for the state to catch up with their innovations. In contrast, in this article, we show how digital platforms—with the helping hand of certain state actors—have been able to grow so quickly by actively slowing down the pace of various actions taken by other state actors aimed at curtailing their expansion. Drawing on an exploratory case study of Uber in Poland, as well as a few shadow cases, we examine what we propose to call stalling strategies that have been used by digital platforms to slow political and legal processes down. We argue that, by stalling the state, digital platforms have been making time that they have been able to use to accumulate platform power (Culpepper & Thelen, 2020); that is, to establish a relationship with a large number of consumers and become an architecture for their everyday lives, thus making it difficult for the state to curtail them.

The paper starts with a theory section in which we discuss the primitive accumulation of platform power and introduce the notion of stalling strategies. We then turn to our case selection and our data. We move on to the empirical analysis and discuss various stalling strategies. The paper concludes by highlighting the role of stalling strategies beyond the case of Uber in Poland.

Theory: Primitive Accumulation of Platform Power and Stalling Strategies

In a recent paper, Culpepper and Thelen (2020) argued that digital platforms are difficult to regulate because they have accumulated a new form of political power, which they call “platform power.” According to them, digital platforms such as Uber and Airbnb do not obtain their political power from being large employers, as was the case with previous large companies whose threat of exit would prevent states from regulating them. Rather, digital platforms have power because regulators fear the political fallout of going against a service that their constituents see as essential to their everyday lives. If regulators nevertheless decide to act against digital platforms, platforms are able to quickly mobilize a large number of devoted consumers to oppose the regulation. This was the case with Uber’s struggles in New York in 2015, where Mayor Bill de Blasio was defeated in his attempt to regulate the company (Seidl, 2021), as well as in London, where more than 600,000 Londoners signed a petition to save Uber (Doward, 2017). In other words, digital platforms profit not only from an economies of scale, which reduces their costs per unit the more users they have, but also from a politics of scale, which increases the political cost of acting against their interests the more users they have.

While Culpepper and Thelen convincingly theorize the power of platforms once they have already achieved a certain scale, the theoretical question their paper leaves unanswered is, how do digital platforms operate when they have not yet formed an alliance with a large number of consumers? Put differently, how does the primitive accumulation of platform power take place?

One theory explaining how digital platforms have been able to accumulate platform power is the two-stage model of expansion (Collier et al., 2018). According to this model, in the first stage a platform disrupts regulation and in the second it regulates disruption. What we call the primitive accumulation of platform power happens between the two stages. Ruth Berins Collier, V. B. Dubal, and Christopher L. Carter refer to this period as one of “regulatory inaction” (2018, p. 921). Yet, as the authors of this special issue point out, “state inaction is characterized by its opposite: a series of actions that seek to alter, modify, falsify, block, delay or accelerate actions or processes related to the enforcement of the law.” The notion of regulatory inaction obscures all the actions taken by digital platforms, as well as by various state actors, which either prevent the state from enforcing existing regulation, help it to enforce the law selectively, or force it to engage in some other form of action. The notion of regulatory inaction also hides all the actions taken by various actors that slow down the process of amending existing regulations or creating new ones, during a period of rapidly changing market conditions. In this article, we refer to those actions as stalling strategies and argue that such strategies have been central to the primitive accumulation of platform power by Uber and other digital platforms.

The primitive accumulation of platform power is a process of building strong relations with a large number of consumers. While there is an ongoing debate on whether the relationship between a digital platform and its users is based on purely market transactions (Schneider, 2017), dispossession (Zuboff, 2019; Cohen, 2019), or gift (Elder-Vass, 2016; Mikołajewska-Zając, 2018; Fourcade & Kluttz, 2020), it is clear that establishing this relationship takes both capital and time. In the case of Uber, attracting customers away from conventional taxis, from public or private transport, gaining their trust and affection, to the point of becoming an infrastructure for their everyday lives takes large amounts of capital: major investment in user-friendly technology with innovative features; large subsidies of fares that attract both drivers and customers; and different forms of advertisement to popularize the application.

But building a large base of loyal customers also takes time. In other words, platform power is not bought, which would suggest an instantaneous transaction, it is accumulated. Paul Pierson pointed out that “the most fundamental point is that power is something that develops over time and simultaneously becomes less visible as it does so” (2016, p. 134). This has been the case with platform power as well. Stalling strategies extend the time a platform has to accumulate platform power by preventing state institutions from enforcing existing regulations or creating new ones. The notion is thus closely related to institutional drift, that is, “the failure of relevant decision makers to update formal rules when shifting circumstances change the social effects of those rules in ways that are recognized by at least some political actors” (Hacker et al., 2015, p. 184). Following this special issue’s actor-centered perspective on law enforcement, we argue that stalling strategies force institutional drift by preventing the state from enforcing existing rules or adapting existing rules to the new situation created by the arrival of digital platforms.

The Temporality of Digitalization and Stalling Strategies

There is a growing body of literature that shows how digital platforms contribute to and profit from accelerating social life. Digital capital has effective new technologies to speed up social life with the use of quantification, surveillance (Zuboff, 2019), and choice architecture (Fourcade, 2017). In relation to workers, digital platforms compress time and create a 24-hour economy with an “on-demand” just-in-time workforce (Wajcman, 2015; Griesbach et al., 2019; Gray & Suri, 2019). Amazon earns more money by increasing quotas for its workers and surveilling them to work faster so that they reach these quotas (Evans, 2019). Similarly, digital platforms also speed up social life for consumers. If, as Shoshana Zuboff has pointed out (2019, pp. 293–328), digital platforms have an interest in making people dance, implicit in Zuboff’s metaphor is that the faster they dance the better for the platform. The more and the quicker people interact using platforms the more “behavioral surplus” is created, allowing platforms to better predict and shape what people will do next, earning more money in the process.

In contrast to the literature on how digital platforms accelerate social life, this article shows that digital platforms also contribute to and profit from decelerating political and legal processes. Various authors have already pointed out how digital platforms have been able to slow down the process of entering the stock market as they rely on “patient” capital to fuel their growth (Rahman & Thelen, 2019, pp. 179–180; Rothstein, 2021, pp. 16–17). Taking time to enter the stock market allows platforms to achieve long-term market domination, without having to provide short-term profits or disclose detailed information about the company.

Digital platforms also profit from actively slowing down legal and political processes. They benifit from stalling the state. Stalling helps a platform to entrench itself in society through technological lock-in and institutional deepening (Starr, 2019). Stalling creates time for technological lock-in to set in, which increases the economic costs of switching to a different platform or going off platform altogether. It also creates time for institutional deepening, enabling the platform to establish legitimacy (Serafin, 2019; Adler, 2021), permeate social relations, and become an infrastructure to everyday practices and beliefs. A digital platform profits not only because it is able to use the time earned through stalling to generate revenue but also because it can use this time to grow, learn about consumers, improve its services, lobby, delegitimize existing regulation, and accumulate platform power. For a digital platform, growing and accumulating platform power facilitates future profit-making by improving its structural position vis a vis any potential competitor (Muennich, 2019). It leads to the “institutionalization of advantage” (Pierson, 2016, p. 131). But growing and accumulating platform power also makes the platform more difficult to regulate due to policymakers’ growing fear of opposing a company that has become a basic infrastructure in the lives of many of their constituents. Moreover, as we shall show in the case of Facebook later, once a platform has accumulated platform power it continues to use stalling strategies as a defensive strategy during scandals, which as Culpepper and Thelen point out (2020), threaten the platform–consumer alliance at the heart of platform power.

When stalling the state, digital platforms can sometimes rely on certain state actors and state institutions who are more than happy for the state to be stalled. ¹ For digital platforms, some “hands of the state” (Morgan & Orloff, 2017) are usually more helpful than others. Some state actors turn a blind eye to a lack of law enforcement or might even actively contribute to stalling by delegitimizing existing laws or blocking their enforcement or prolonging the process of creating new regulations. The contribution of state actors to stalling cannot always be explained by state weakness. State actors might stall not necessarily because of personal interests (direct corruption, campaign contributions, and hope of a future job) but because stalling allows them to achieve political goals that would otherwise require changing existing regulations and taking on powerful stakeholders. For policymakers, helping platforms stall can be a form of industrial policy, which allows them to “circumvent structural impediments in public policymaking” (Dewey & Di Carlo, 2021, p. 16). In the case of Uber, by either passively allowing the state to be stalled or actively contributing to stalling, policymakers are able to provide their constituents with cheaper rides, create additional jobs, lowering unemployment figures, as well as push forward what they and some of their constituents perceive as the “fourth industrial revolution” without having to pass new laws that would deregulate the taxi market and go against the entrenched interests of the traditional taxi industry.

To recapitulate, like accumulation by gift (Fourcade & Kluttz, 2020) and accumulation by dispossession (Zuboff, 2019, p. 99), accumulation by stalling has contributed to the emergence and entrenchment of platform capitalism. Digital platforms are able to use the time created by stalling to accumulate platform power, which once accumulated makes it difficult for the state to regulate them.

Law’s Delay and Stalling Strategies

Building on the recent literature on the constitutive role of law in contemporary capitalism (Cohen, 2019; Deakin et al., 2017; Pistor, 2019; Starr, 2019), we focus in particular on stalling strategies related to law. More specifically we focus on how, to borrow a phrase from Hamlet, platforms have contributed to and profited from “the law’s delay.” ²

In recent years, various researchers have noted the discrepancy between the ongoing social acceleration in modern societies and the slowness of law (Francot, 2018). The relationship between technology and the pace of regulation has been discussed, for example, in the context of measures considered desirable to allow law to keep pace with technology (Fenwick et al., 2017; Slating & Kesan, 2011; Marchant et al., 2011). Others have argued that, due to its slowness, law provides time for discussing and analyzing new technologies and may restrain their potential negative effects (Drechsler & Kostakis, 2014), or that the slowness of regulation corresponds to the fact that the effects of the usage of new technologies become apparent only after some time (Mandel, 2007, p. 574). Slowness, this “rather notorious trademark” (Francot & Mommers, 2017, p. 282) of law, was shown to be both a guarantee of stability and a hindrance to adjustment to changing social and economic conditions, leading to the conclusion that “law seems to be either too late or too soon, but never ‘in time’” (van Klink, 2018, p. 33).

We know that platforms sometimes push for quick application of the existing legal principles or doctrines in a way which enables them to conduct their business. One example of this, as Julie Cohen notes, are data brokers and platforms making efforts to be recognized as owners of personal data by showing how their use of personal data results in innovation (Cohen, 2019, p. 72). Shaping the discourse according to which platforms—as innovators creating new services or products—should be recognized as the owners of personal data, happens at the expense of data subjects whose data are appropriated. This has enabled the development of a business model built on exploitation of personal data.

While some constitutive decisions regarding digital capitalism were made quickly without much debate or public awareness of their consequences, we show how in other situations digital platforms have been slowing down legal processes. We trace the strategies that can be observed in terms of digital platforms’ efforts to slow down the process of adoption and enforcement of laws. Thus, we focus on how “legal tempo” (Luskin & Luskin, 1986) is influenced by what Lyana Francot evokes as “strategic delays by parties” in the context of legal proceedings (2018, p. 98). We distinguish five stalling strategies that can be observed in relation to postponing the adoption of legislative measures, slowing down court proceedings, and enforcement of regulations.

Case Selection and Data

We develop our argument with an exploratory case study of Uber in Poland. To avoid the pitfalls of generalizing from a single case, however, we situate it in the context of a number of shadow cases that confirm our findings for other digital platforms and in other institutional settings.

Our case study is exploratory with the aim of theory construction (Mahoney, 2015, p. 212): it seeks to identify one of the causes of the rapid growth of digital platforms even in hostile institutional environments. We justify choosing Uber because it is a typical case (Beach & Pedersen, 2018): other digital platforms have used Uber’s strategy for dealing with regulations as a blueprint (Burfield & Harrison, 2018). We use Poland as the case study because Uber has been able to grow rapidly there and successfully enter 10 cities, even though there were laws regulating taxi markets that should have made this difficult. Yet these laws were not successfully enforced, nor were they updated to meet the regulatory challenges created by Uber.

This article investigates how stalling strategies gave the company time to grow and accumulate platform power. Unfortunately, as Culpepper and Thelen point out, platform power is difficult to measure. Traditional indicators such as market share or size understate dominance (Rahman & Thelen, 2019, p. 179; Culpepper & Thelen, 2020, p. 294). Moreover, even data on size is hard to come by and usually comes from platforms themselves, which are known for being rather secretive about their activities. To make our case that Uber was accumulating platform power, we thus had to rely on imperfect measures such as the number of markets in which it operates, the limited data provided by the company to the general public, its prevalence in the media, but also our qualitative judgments of Uber’s position in Poland.

Studying the contribution of stalling strategies to the accumulation of platform power required the use of methods that focus not only on whether or not the law has been enforced or changed—that is, on an outcome—but rather on law enforcement and law-making as a process. To develop our theory, we relied on process tracing (Mahoney, 2015), which focuses on the unfolding of events over time and draws on within-case observations, causal-process observations, and clues (Barrenechea & Mahoney, 2019, p. 454). We drew on the case law of administrative and regular courts, press articles and online media, Facebook posts of organizations of taxi drivers and Uber drivers, discussions in the Polish Parliament and various committee meetings, comments of legal practitioners and officials concerning Uber, and interpretations of law provided by state officials and institutions.

We also drew on ethnographic observations and interviews conducted during court cases of Uber drivers, which took place in Cracow between November 2016 and June 2017. ³ Ethnographic observations in court have been crucial as the subtleties of stalling strategies are not necessarily envisioned by case law: judgments usually present the final outcome, not necessarily the path that leads to it. During ethnographic observations multiple interviews were conducted with a municipality official and the lawyer representing Uber drivers, as well as with taxi drivers who were observing the trials. Our findings from ethnographic observation were validated with the analysis of the case law and media coverage.

Stalling Strategies

Having discussed the theory and method, we now turn to a detailed analysis of how stalling strategies have helped Uber to accumulate platform power in Poland. We distinguish between five such strategies: reinventing classifications; stealing time from street-level bureaucrats; dragging out court proceedings; delaying the introduction of new regulation; and taking time to (not) comply with new regulations.

Making Time by Reinventing Classifications

One way in which Uber has been able to stall the state is by creating a complex legal structure that both avoids and challenges existing state classifications. The power of the state rests, to a large degree, on its ability to create and enforce classifications (Morgan & Orloff, 2017, p. 11). Uber challenges this power by claiming not to operate in the transport market at all, thus reinventing classifications.

In Poland, since the beginning of its operations, Uber has also claimed that it is not a transport services provider and is not subject to the rules governing passenger transport. In the National Court Register, Uber Poland’s core business was initially classified as “Intermediary in sales for advertising purposes in electronic media (Internet)” and after a couple of weeks changed to “Other consulting in the field of business economic and management.” ⁴ Also, in the media the company presented itself solely as an intermediary between drivers and passengers.

Uber’s strategy clashed with state institutions’ uncoordinated attempts to somehow clarify the state’s attitude toward the company, for example, under which legal category it should be classified and what its obligations are. The passenger transport market in Poland is regulated by the 2001 Law on Road Transport (2001). Before the 2019 amendments, familiarly known as “lex Uber,” regulations targeted mainly drivers. According to the Law on Road Transport, passenger transportation services could be provided either by taxis or by so-called occasional passenger transportation (equivalent to London’s private hire cabs). Both types of services required that drivers fulfill several obligations. While focused on regulating drivers, the Law on Road Transport did not provide any categories describing intermediaries. Such services were usually provided by corporations. Taxi drivers, usually self-employed micro-entrepreneurs, used the services of corporations to obtain access to fares in exchange for a monthly fee. Uber’s entry into Poland disrupted the existing order as the company claimed that it was neither a taxi corporation nor a transport company.

A number of state institutions, at both the central and local levels, took issue with the way Uber was reinventing classifications. The Ministry of Infrastructure, the Road Transport Inspectorate, and local regulators in Cracow contested the characterization of Uber’s services as something different from passenger transportation. But in their attempt to challenge existing classifications Uber obtained support from two other state institutions. In a statement released in April 2016, the Polish Office of Competition and Consumer Protection (UOKiK) praised Uber for its positive impact on competition and consumers. The agency “found no need for the state to intervene with regard to protecting consumers and competition” (UOKiK, 2016). Citing this passage from the statement, the Ministry of Finance went further, arguing that the existing regulatory framework should be liberalized because Uber “provides an interesting solution for people wishing to perform an additional job” and solves some of the existing “pathologies” of taxi markets, such as taxi drivers not turning on the taximeter or overcharging their customers (Banaś, 2017). The ministry of finance was helping Uber delegitimize existing regulations.

Stealing Time From Street-Level Bureaucrats

While the strategy of reinventing classifications protected Uber, it did not protect Uber drivers, who could still be fined for providing the service of a taxi driver without a license. As drivers were necessary for expanding Uber’s market share, other stalling strategies were needed in order to hamper the state’s enforcement of regulations. Uber has been stealing time from street-level bureaucrats: it made law enforcement time-consuming by facilitating what, following the authors of the introduction to this special issue, can be called jurisdictional, domain and organizational complexity.

Enforcing the law in the digital age has been a challenge. As the head of the Labor Inspectorate, Wiesław Łyszczek, pointed out during a committee meeting in the Polish Parliament:

“The growing complexity of economic and social relations, together with the availability of information and communication technologies are changing the labor market. The complexity of the problems (…) is clearly visible in work performed with the use of platforms, such as Uber and Bolt. This outcome has the multi-sided and interorganizational character of control, on one hand, and the time pressure facing labor inspectors, on the other.” (Łyszczek, 2020)

The labor inspectorate faced a number of practical problems that made law enforcement more time-consuming. For example, in Poland, Uber relies on brokers, called Uber partners, to help recruit, handle and pay drivers. When labor inspectors tried to inspect these companies to see whether they were breaking the law—whether, for example, they were unlawfully employing foreigners or paying below the minimum wage—it would turn out that their headquarters were not where they were supposed to be or, on the day of the inspection, that the given Uber partner no longer employs anybody and was therefore no longer subject to labor inspection (Chief Labor Inspectorate, 2020).

Similarly, institutions at the local level faced problems trying to enforce the law. This was the case in Cracow, which in 2016 began to forcefully enforce the law against Uber drivers. Local officials in Cracow organized sting operations and began to fine Uber drivers for breaking the law. Here too, state officials faced a number of problems that stalled their ability to enforce the law as Uber began to deactivate their apps and restrict their credit cards, forcing the local authorities to order new ones (Łazarczyk, 2016). As one state official told us, this created a number of practical problems for a small local state agency to be able to acquire a larger number of credit cards and postponed the whole process of law enforcement.

Another method that Uber used to hamper state institutions’ ability to control drivers was to request access to public information to hamper the use of sting operations by showing that the law is being enforced without following all the necessary legal procedures. The first requests contained questions concerning, among other things: the name and surname of the clerk who used the rides obtained via the app at the given time mentioned in the request, whether the institution is the owner of the phones used to order the rides, the personal data of the City officials who received credit cards to be used for payment of transportation services ordered via the app and the numbers on the cards. Demanding information on the exact details of checks was a way of impeding what were fairly innovative actions aimed at enforcing regulations.

Dragging Out Court Proceedings

When state officials tried to enforce the law, Uber also stalled by dragging law enforcement through courtrooms and slowing trials down. For Uber, entering a new market is often followed by a legal battle concerning its activities. These activities cause concern in areas such as competition, labor, social policy, taxation, and consumer safety (Crespo, 2016; Thelen, 2018; Cherry, 2019; Kirchner & Schüßler, 2020; Marin et al., 2020). In their legal battles, the company not only tried to affect the outcome but also their pace.

As Uber was claiming not to be a taxi corporation and thus not to be subject to existing regulations in Poland, a number of state institutions, especially at municipality level, were nevertheless treating Uber drivers as if they were taxi drivers and fining them for breaking the law. In response, Uber introduced a policy of advising their drivers to reject fines, forcing the state to go to court, thereby prolonging the process of law enforcement further.

Courts faced the necessity of assessing the character of the work performed by Uber drivers. As a company, Uber was not officially a party in any of these proceedings. Yet, as reported by the media—and as we were able to verify through our ethnographic observations of court cases—Uber did provide drivers with lawyers and coordinated their defense. Uber was under no legal obligation to do this. The contract that drivers signed with Uber meant that the responsibility for following the law lay with the drivers. In contracts, the company claims that drivers are “customers” of the company. Thus, for Uber, to provide drivers with lawyers was analogous to a radar-detection company helping its customers fight speeding tickets. ⁵

Providing drivers with legal counsel enabled Uber to represent its interests in court and hinder the progress of the trial. It seems that delaying court cases was not an unintended consequence for Uber but rather a deliberate strategy. It was reported in the Polish media that Uber would inform people who were considering becoming Uber drivers that, if they were fined by the state authorities for driving without a license, Uber lawyers would take care of the situation and, if necessary, drag out the court case (Szczepaniak & Szczygieł, 2016; Ułan, 2016). We were able to observe the use of this strategy in court as drivers’ lawyers used various motions to stall proceedings. For example, the lawyer would show up in court without the driver and ask the court to postpone the proceedings until the driver was available.

Another strategy we observed was for the driver’s lawyer, who was being paid by Uber, to motion the court to ask Uber Poland whether or not the accused driver did in fact receive money for the fare during the sting operation and whether it was possible that somebody other than the defendant was the driver. ⁶ The lawyer was thus asking the court to get Uber to provide information that the lawyer, hired by Uber, could bring into evidence. The court would grant the motion, adjourn the trial to wait for the company’s response, which would take time, and then answer that Uber Poland is just an advertising company. The lawyer was able to use the privileged position of representing the interests of an actor—Uber—that was not formally on trial, as well as the jurisdictional complexity of the fact that Uber drivers sign contracts not with Uber in Poland but with Uber in the Netherlands. Just with this single motion the lawyer was able to prolong the court case a number of weeks. ⁷ And when a driver would be found guilty of breaking the law, the case was appealed, thereby dragging out the whole process even further.

Delaying New Regulations

Another stalling strategy is related not to enforcement of existing regulation, but to delaying new regulation. In September 2017, the Ministry of Infrastructure proposed regulatory changes to the Law on Road Transport. The proposal had a number of provisions that were very problematic for Uber. First, it included a broad definition of intermediaries, which would make it impossible for Uber to avoid fulfilling the obligations foreseen by law. It introduced a number of new penalties for not following the law, including the option to block the platform altogether. Equally problematic was that the proposal also retained the possibility for cities to demand that drivers pass an exam in knowledge of the law and the city. Until that point Uber had been able to grow quickly by relying on drivers, in many cases recent immigrants, who would have problems passing such an exam, as they did not know the city well enough and relied on GPS navigation for their work.

Rather than dismiss the law altogether or push for complete deregulation, in their response to the Ministry of Infrastructure Uber agreed with the need for new regulation. It criticized various parts of the proposal, however, making the case for further dialogue in the construction of “pro-consumer regulation.” The company also made the point that the state should not hurry in creating new regulations because “technological change is happening so fast that restrictive and rigid rules could very quickly again become obsolete.”

By this point, Uber was not only using stalling to continue to grow but was able to draw on its popularity to undercut the legitimacy of existing regulation and argue against a need for new regulation. Between 2014 and 2017, Uber successfully entered 10 different cities and 11.2% of inhabitants of large cities—which are the markets in which Uber operates—were using the service (Statistics Poland, 2017, p. 137). In their letter to the ministry, Uber did not hesitate to evoke its growth and popularity as an argument against the proposal. Uber claimed that research shows that “already 70% of young people in Poland cannot imagine life without mobile apps,” and that there are over one million users of Uber in Poland. ⁸ An analysis of Uber’s public statements shows that, while prior to 2017, Uber emphasized the qualities of their users (that they are well educated, professionals, more likely to vote, do not own a car), and how because of those qualities (e.g., lack of car) Uber will bring about positive changes related to the environment and city congestion, by 2017 it was emphasizing the quantity of people who use the app. It was making the case that it has become an essential service and an infrastructure to everyday life.

Nevertheless, the Ministry of Infrastructure, headed by Andrzej Adamczyk who was the member of the major coalition partner, Law and Justice (PiS), continued to pursue their proposed project against the objections of Uber. At the time, it seemed possible that the law might pass, as a similar law had just passed in Hungary (Makó et al., 2021), a country that PiS was following in many other policy decisions.

In their attempt to block or delay new regulations created by the Ministry of Infrastructure, Uber was aided by Mateusz Morawiecki, at the time Deputy Prime Minister and Minister of Finance and a member of PiS, and by Jarosław Gowin, Deputy Prime Minister and Minister of Higher Education and leader of PiS’s small coalition partner. Morawiecki argued that, by enabling car sharing, Uber was increasing labor productivity, which was important because Poland was lacking labor resources and that “the sharing economy” was a “trend of the future.” Gowin publicly compared the Ministry of Infrastructure to Luddites destroying machinery with hammers during the Industrial Revolution (Jarosław, 2017), promoting digital platforms as agents of the “fourth industrial revolution” (Gowin, 2016).

Gowin was not only blocking the new law being proposed by the Ministry of Infrastructure but was also not supporting smaller measures proposed by the National Labor Inspectorate, and supported by the Ombudsman, that would give the agency better tools to enforce existing law. In other words, he was engaging both in organizational sabotage and legal sabotage, pursuing policy through regulatory forbearance (Dewey & Di Carlo, 2021). It is worth noting that in 2012, before the arrival of Uber, the same Jarosław Gowin, then the Minister of Law, had tried to fully deregulate taxi markets but failed after taxi drivers protested. Five years later, by simply stalling the introduction of new regulations, Gowin was able achieve the same goal of opening up the taxi market to drivers without a license without having to repeat the fight with traditional taxi corporations and licensed taxi drivers that he had previously been unable to win. By blocking the Ministry of Infrastructure’s proposal and other measures that would have made law enforcement easier, Gowin was contributing to institutional drift benefiting Uber.

In blocking new regulations, Uber was also able to rely on the helping hand of the United States. On October 17, 2018, the Ministry of Infrastructure proposed new regulations. One week later the US ambassador in Poland, Georgette Mosbacher, sent a letter to the Minister of Infrastructure, which was subsequently leaked to the press. The letter voiced strong opposition, arguing that the law would have negative consequences for investments, as well as for Polish–US relations. She added a handwritten note “please don’t make such a far reaching mistake” (Gruca, 2019).

Six months later, in April 2019 the Polish parliament passed a weakened amendment to the Law on Road Transport, which the media called “lex Uber.” The amendment was much more favorable to Uber than the one proposed by the Ministry of Infrastructure in 2017. It confirmed the obligation for Uber drivers to possess the relevant license. The option of requiring drivers to pass an exam in the bigger cities was, however, removed, legalizing the already existing situation of Uber using drivers who had not passed the exam. The new law allowed drivers to use the mobile app instead of taximeters, enabling the executive branch to issue the relevant detailed regulation.

But Uber was able to additionally delay the introduction of new regulations by asking for a prolongation of vacatio legis, the transitional period between the promulgation of a law and the time at which the law takes effect. Together with several other organizations, Uber voiced its concerns regarding the 30-day period proposed and asked for it to be extended (Prime Minister, 2019, p. 66). This was granted and a 90-day vacatio legis was written into the law.

Hacker, Pierson and Thelen have pointed out that “doing nothing in the legislature often means doing something quite big in the world” (2015, p. 193). Uber entered Poland in 2014, but nothing was done in terms of regulation that would impact it until 2020. But nothing was done because of various actions taken by Uber and various state actors, who stalled the introduction of new regulation, contributing to institutional drift. By the time the regulation was passed, not only was it different from the one initially proposed but so was the situation. In the late 2016, when the law was beginning to be discussed, Uber was claiming 350,000 users who had used the app at least once (Bereszczyński, 2016). In 2021, when the law came into effect, the figure was allegedly 4.5 million, with 60% of people living in the city in which Uber was operating having used the app at least once (ISB Tech, 2021). The company had built an extensive network of partners and a large fleet of drivers, and expanded its operations to Uber Eats and electric scooters. It had changed from being a small market challenger to an incumbent, established legitimacy among policymakers and the wider public, and delegitimized existing regulation. It had accumulated platform power.

Taking Time to (not) Comply

The last strategy refers to the time Uber takes to adjust to the new regulatory framework. There has been a tension between Uber’s declared readiness to support drivers in the process of fulfilling new regulatory obligations and its efforts once again to escape the legal classification created by the amended law.

The amended law introduced a category of “intermediation in transportation of passengers,” obliged to work solely with licensed drivers under threat of financial penalty (Art. 92a); correspondingly, drivers were obliged to work only for licensed intermediaries. Thus, technically, the amended law clarified the role played by Uber in the transportation market and the obligations the company was supposed to fulfill. But Uber again implemented a strategy that prolonged its existence outside existing regulations.

With the prolonged vacatio legis, the new law required intermediaries to adapt by April 2020. By that time, the company was supposed to register as an official intermediary. Uber did not comply with the new law in time. At the beginning of April, the period was subsequently prolonged as part of an anti-Covid law until 30 September 2020, when Uber finally complied with the new law (Chief Inspector of Road Transport, 2020).

Even though some elements of the law came into effect in January 2020, Uber presented drivers with new contracts that contained a provision which contradicted the new law. It contained the following disclaimer: “The customer acknowledges and agrees that Uber is a technology services provider that does not provide transportation services, function as a transportation carrier, nor operate as an agent for the transportation of passengers,” ⁹ which showed that Uber sought to bypass the introduction of the legal category of intermediary.

On the drivers’ side, it was not clear whether drivers needed a taximeter or could use the app to calculate the distance. Uber’s explanation of the new policy on the possession of a taximeter and other specific provisions included in lex Uber seemed to be written in order to promote the interpretation most suitable for the company’s needs. In an interview from January 9, 2020, Uber’s CEE Cities Lead from Uber Poland claimed that the executive acts merely accompanied the Law on Road Transport and it was possible to act solely on their basis to justify the use of apps instead of taximeters. He evoked the policies of certain cities that allow this (Konowrocki, 2020). He did not explicitly claim that there is no obligation to have a taximeter, but gave the impression that the Law on Road Transport should be prioritized over the executive acts, which were not prepared yet. However, neither he nor the company addressed the fact that in some cities drivers were being fined for not complying with the new law and not possessing a taximeter.

Beyond the Particular Case: Stalling to Accumulate Platform Power, Stalling to Protect It

While the use of stalling was not a sufficient condition for Uber to accumulate platform power in Poland, it was a contributing condition and, perhaps, even an “‘important’ necessary condition” (Mahoney, 2015, p. 214). Stalling prolonged Uber’s unregulated operations, creating time that it could use to build a fleet of unlicensed drivers and a large customer base, which became accustomed to low prices and the manner in which the company operates. It could use this customer base to show its popularity to policymakers and state officials trying to curtail it.

But Uber was stalling not only in Poland. Across the world, the company was stealing the time of street-level bureaucrats with software called Greyball, which made it more time-consuming to organize successful sting operations (Portland Bureau of Transportation, 2017). The software made it impossible for officials to see drivers in the app, limiting their ability to enforce the law. Uber used another application, Ripley, to lock down computers in its foreign offices and protect them from police raids, making it more time-consuming to obtain evidence of law breaking (Zalewski & Newcomer, 2018).

Uber was also using various stalling strategies in California. In 2019, California passed Assembly Bill 5 (AB-5), which was supposed to provide a test which would classify Uber drivers as employees (Win, 2020). The company was reinventing classification and taking time to (not) comply as its Chief Legal Officer claimed that: “We continue to believe drivers are properly classified as independent (…) And because we’ll continue to be responsive to what the vast majority of drivers tell us they want most – flexibility – drivers will not be automatically reclassified as employees” (O’Brien, 2019). California took Uber to court to force compliance (O’Brien, 2020). As it was stalling the enforcement of new law, Uber was pushing for the adoption of Proposition 22, which would exempt drivers from the classification as workers foreseen in AB-5. The company spent 57 million dollars on a campaign supporting Proposition 22 and “weaponized their popular apps” (Ongweso, 2020b) in order to encourage both customers and drivers to vote “yes” in a referendum on this issue (Midgley, 2021). During this time, Uber was also taking time to (not) comply and dragging out court proceedings related to an investigation into sexual assaults (Ongweso, 2020a). In December 2019, an Administrative Law Judge in California ordered the company to turn over data related to sexual assaults that took place in an Uber but the company did not comply, forcing the state to go to court. Uber ended up settling and agreeing to provide the data in July 2021 after Proposition 22 was passed in November 2020. Providing data to state authorities before the vote could have given fuel to the campaign against Uber. Forcing the state to go to court to gain access to the data and settling after the measure was passed prevented this from happening. ¹⁰

Other digital platforms too have engaged in stalling. In the United States, Airbnb stalled new regulation with lobbying and made law enforcement more time-consuming for street-level bureaucrats by, for example, making it appear as if listings are located in slightly different locations (Martineau, 2019; Jia & Wagman, 2020). In Europe, a recent audit report for the European Commission highlighted the slow speed of law enforcement, which was especially slow in the digital sector. In the case of antitrust law, it took on average of 4 years, and in some cases, as long as 8 years, from the beginning of the preliminary investigation into a company to the termination of the investigation with a formal decision. The slow speed was due not only to the growing complexity of cases or the result of choices made by the Commission, but because “companies under investigation systematically requested prolongations of deadlines and replied to requests for information only with substantial delays of between four to eight months and up to 1.5 years, or prolonged proceedings by submitting ineffective proposals” (European Court of Auditors, 2020, p. 27). Companies would further delay law enforcement by appealing the Commission’s decisions before European Union (EU) courts.

To give an example, in 2010, the EU Commission launched an investigation into Google Search (Google Shopping) for antitrust violations but finished the investigation only in 2017, fining the company €2.42 billion, a decision that Google appealed in 2020. While many focused on the record fine, one competitor working for Yelp complained that “The commission was sending an ambulance to a funeral” (Van Dorpe & Nylen, 2020). By the time Google’s conduct was declared illegal and it was fined, the platform had entrenched its dominance among shopping comparison services. Google had benefited from the law’s delay.

As we saw in the case of Uber in Poland, even 4 years can be enough for a platform to entrench itself. The EU commissioner for competition, Margrethe Vestager, highlighted the problem with delayed law enforcement when discussing her ongoing inquiry into Apple possibly abusing the position of its platform, the App Store: “We have seen in other cases how damaging it can be if things take a lot of time – then the market moves on. So obviously, we never compromise on the quality of our case work and on due process, but we need speed, because a digital marketplace and a digital world is a world where things are moving fast” (Swisher, 2021). We thus see a struggle over speed playing itself out within the EU, with regulators trying to speed up law enforcement as platforms have been able to slow it down.

We also saw various stalling strategies at work during the Cambridge Analytica scandal. After the scandal broke in March 2018, Facebook’s lobbyists unsuccessfully tried to prevent Mark Zuckerberg from being called to testify before the US Congress (Frenkel & Kang, 2021, p. 161). In April 2018, Zuckerberg appeared for the first time to face questions from policymakers. During his testimony Zuckerberg was reinventing classification as he claimed that “I do not consider ourselves to be a financial institution although you are right that we do provide tools for people to send money.” Carefully prepped by a team of litigators (Frenkel & Kang, 2021, p. 161), Zuckerberg was also stealing time from state officials by asking them to reprise their questions, by providing long answers to yes-or-no questions until they ran out of time or avoiding answering potentially very damaging questions altogether by saying “we’ll get back to you.” It would be easy to dismiss this as an insignificant figure of speech. But with the use of this simple phrase, and similar expressions, Zuckerberg was avoiding having to provide public answers at a moment when Facebook was under intense public scrutiny. Facebook did get back to Congressmen with answers but only 2 months later (Marketplace, 2018). During those 2 months, the media together with public opinion had moved on from the scandal to other topics and so Facebook’s answers did not receive the public scrutiny that the public hearing had got. ¹¹

While Zuckerberg’s stalling in Congress was very public, other Facebook stalling strategies related to the scandal were less visible. Behind the scenes, Facebook was taking time to (not) comply. In 2019, 1 year after launching an investigation into the Cambridge Analytica scandal, California’s Attorney General had to petition the San Francisco Superior Court to force Facebook to comply with the investigation (Palmer, 2019). And after more than 2 years it was still an open question whether the Attorney General would be able even to collect the evidence to build a strong case against Facebook (Swift, 2020).

As the example of Facebook during the Cambridge Analytica scandal shows, stalling not only helps platforms to accumulate platform power but also helps them to protect this power, once accumulated. Put differently, stalling is not only a proactive strategy that is used at an early period of growth and quest for domination, as was the case of Uber in Poland, but can also be a defensive strategy of an already dominant platform protecting their position. When the Cambridge Analytica scandal broke out in March 2018 Facebook’s shares slid 6.77% in response to uncertainty concerning what the scandal would mean for the company. In contrast, when in July 2019, as part of a larger settlement, Facebook was fined five billion dollars for the Cambridge Analytica breach, its shares hit highest price in nearly a year. Like financial institutions (Woll, 2019), large digital platforms are resilient even in the face of hefty fines for misconduct (Pistor, 2020, p. 10). Facebook does not face budgetary constraints or pressure from competition. Moreover, its corporate structure gives Zuckerberg almost complete control over the company, preventing the possibility of a hostile takeover. For large digital platforms such as Uber or Facebook, even a hefty fine is therefore not a big threat.

But an escalating scandal is a big threat. It can become a critical juncture that changes discourse and coalitions (Laurer & Seidl, 2021, p. 15), and takes away consumer support. This not only decreases revenue but creates a window of opportunity for lawmakers and regulators (Rossi, 2018; Culpepper & Thelen, 2020). During a scandal—Cambridge Analytica or #DeleteUber—the coalition between a digital platform and its users at the heart of platform power is at risk and it thus becomes easier to enforce existing regulations or create new ones. A scandal can embolden lawmakers and regulators who want to introduce new regulations or more strictly enforce existing ones but who had previously feared a public backlash. It can also force lawmakers, who previously might have been quite happy to be stalled—either because of personal interests (corruption, financial contributions, and hope of a future job) or public goals (cheap services for constituents, lower unemployment rates)—to take a more critical stance and, perhaps, even switch coalitions. Stalling can help a platform to wait out a scandal until the window of opportunity for lawmakers and regulators closes, thereby allowing the platform to conserve the power it was able to accumulate, in part, because of stalling.

Conclusion: Platform Power as a Fait Accompli

While Silicon Valley cultivates a discourse about how it disrupts and challenges the “political status quo” (Tusk, 2018, p. 10), we have shown that platforms sometimes both contribute to and thrive in the status quo. What members of the industry called “regulatory hacking” (Burfield & Harrison, 2018), in practice often means regulatory hampering. We have argued that the period of time between a platform disrupting regulations and the regulation of disruption, characterized by the literature as a moment of regulatory inaction, is filled with various stalling strategies used both by the platform and by certain state actors. We argued that such strategies have helped platforms grow and accumulate platform power and later protect them during scandals.

Digital platforms are not the first companies to stall the state. The strategies described in this article, with the possible exception of reinventing classifications, have been used by other corporations long before the emergence of the digital economy. They have been part of the toolbox of lawyers and lobbyists and an element of business power. Stalling strategies, however, are especially significant in the digital economy. This is because “a particular characteristic of the digital age is that companies compete for a market instead of in a market” (European Court of Auditors, 2020, p. 29). The victor in a winner-takes-all market is often the first to pass a size threshold, which triggers positive feedback for further growth. Stalling can help to get past that threshold first, which will more than make up for any potential legal and financial consequences of using this strategy.

The growth of platforms is hardly limited by anything but the number of their users (Petit, 2020, pp. 81–87), as the number of users, next to their technologies, is the most important production factor for platforms. Because platforms grow quickly due to economies of scale and network effects, being able to stall for even a few years can go a long way. It can create a space of regulatory inaction long enough for the platform to get so big that consumers demand them and states feel like they have to cave in to their demands. What makes stalling of digital platforms different from that of analogue companies is not only the speed of their growth but also its effects, which go beyond the economy and transform social relations. As a platform quickly grows and permeates social relations, not only can it more easily afford a major fine from regulators but policymakers will find it more difficult to act against the platform’s interest. Moreover, during the period it is stalling, the platform will be able to institutionalize its advantage, creating barriers of entry for competitors and high switching costs for users. By the time the state catches up with a stalling platform, its regulatory actions no longer have the legitimacy or the efficacy they previously had. Since many platforms benefit from the law’s delay, their slogan seems to be not only “move fast and break things” but also “run down the clock until opposition is futile.” ¹²

Stalling the state is expensive. This is not because the strategy itself leads to legal fines because it rarely does. This is because it requires building an army made up of lawyers who create complex legal structures across various jurisdictions and drag law enforcement through courtrooms; programmers who make law enforcement more time-consuming for street-level bureaucrats; lobbyists who protract the process of creating new regulations; and consultants who help wait out a scandal. Even though building such an army requires a lot of capital, it ends up being a good investment as it helps to win the war for the market.

While our exploratory research made a case for the contribution of stalling to the entrenchment of platform capitalism, further comparative research on stalling strategies is clearly required. Such research should map out the population of potential cases (Beach & Pedersen, 2018), including the cases that, like Uber in Poland, involve both stalling and the accumulation of platform power, and cases where either accumulation of platform power was achieved without stalling or there was stalling without accumulation of platform power. Looking at those cases where stalling was unsuccessful would help us see potential omitted contextual conditions that are required for stalling to work (Beach & Pedersen, 2018, p. 864). Uber lost their regulatory battles in certain countries (Chan & Kwok, 2021; Makó et al., 2021; Pelzer et al., 2019), so it seems likely that there are political and legal systems that are less prone to stalling than others. It would be particularly interesting to study the extent to which successful stalling requires the helping hand of at least some state officials. Such a helping hand was clearly present in the case of Uber in Poland, as well as Uber in Boston (Adler, 2021). Yet, as we saw with the example of California struggling to enforce its laws against Uber, and the EU struggling to enforce competition law quickly enough, it may not be necessary. Finally, stalling is clearly not always successful as a defensive strategy. Platforms have not been able to simply stall their way out of every scandal, as is illustrated by the fact that the creator and CEO of Uber, Travis Kalanick, was forced to resign after a number of scandals. Further comparative research would help us to better theorize the contribution of stalling to the accumulation of platform power and the entrenchment of digital capitalism.

To conclude, we already know much about how digital platforms contribute to and profit from speeding up economic processes. In this article, we have argued that digital platforms also contribute to and profit from slowing down the enforcement and adaptation of regulation. Digital platforms often claim that they are solely intermediaries and that their main economic activity is marketing. There is a grain of truth in this. When entering a market, a platform wants to capture as big a share of the market as possible, and lock-in customers by becoming part of the infrastructure of their everyday lives. As customers grow accustomed to the standards and the price of the, often free, service, it becomes more difficult for the state to curtail the platform. Unlike financial institutions, which are perceived by policymakers as “too big to fail,” digital platforms want to be perceived as too popular to be curtailed. They stall the state until the accumulation of platform power becomes a fait accompli.