Labor law and related regulations were created long before the current growth of the on-demand economy and the associated innovation in provision of goods and services. The current labor law structure contains only two primary categories of employment classifications—workers are either employees or contractors. Within this binary structure, the status of workers providing services in the on-demand economy is unclear and subject to ongoing litigation, creating an uncertain operating environment for on-demand companies. In this paper, I argue that these jobs in the on-demand economy represent a form of labor relationship not well captured by the current employee/contractor model. Forcing these on-demand workers to become clearly employees or clearly contractors would, in either case, be replacing the current services with distinct and quite likely inferior services from those offered in the on-demand economy today. I conclude that the preferred resolution of the current uncertain legal and regulatory environment surrounding employment in the on-demand economy would be through modification of the classification scheme to reflect this new alternative type of labor relationship, rather than by forgoing the value created by the on-demand economy.


I.     Introduction

The rapid growth of the on-demand economy in the United States is creating innovative goods and services while simultaneously challenging existing labor laws and regulations. Specifically, companies such as Uber, Lyft, Instacart, Postmates, and Handy that rely on a contractor-based business model are experiencing waves of labor lawsuits that may threaten the fundamental business structure of the on-demand economy space. There is some debate about which companies fall into the on-demand economy. Nonetheless, one of the primary explanations given is that it involves an economic activity created by digital marketplaces that match consumer wants with providers to immediately deliver those goods and services.1

For example, Uber and other ride-sharing companies (i.e. Lyft, Juno, Gett) provide a platform that immediately matches individuals who need a ride (the “riders”) with nearby individuals who can give them a ride to their locations (the “drivers”). Uber does this through a software app that allows riders to submit a trip request, which is automatically sent to nearby drivers who are signed up on the Uber platform. The rider and the driver then “match,” and the Uber app sends the rider location information to the driver. Within minutes, the driver is at the rider’s front door. Riders have their payment information saved on the app so that when the ride is complete, the driver is automatically paid through the app. Uber launched in 2011, and since then has created an “Uberification” effect that has led to many other companies providing a similar on-demand service.2 For example, Instacart is the “Uber for groceries,” matching shoppers with grocery deliverers; Handy is the “Uber for household cleaners”; Luxe is the “Uber for valet parking.” And now there is even an “Uber for blood”—Iggbo—providing on-demand phlebotomists.

In addition to the on-demand feature, these Uber-like companies also have a common business model: the workers who are supplying these various on-demand goods and services look economically different than a standard company’s workers. Take Uber for example: the drivers sign up to be on the Uber platform and have no set hours or requirements for how many hours or days they must work. A driver can turn on his app when he wants to drive someone and he can turn off his app at any time of the day, week, or month. Additionally, Uber has workers who look more like a “typical” worker. For example, it employs software developers who work hours similar to a 9-to-5 job, who report to work in the same physical location every day, and who share other job features with more standard company employees. For the company, these workers are economically different from drivers who have no set hours and no real “loyalty” to the company.

Under the law, these workers are different as well. Software developers and marketing associates are considered “employees” and drivers are considered “independent contractors” (hereafter: contractors) or “self-employed” individuals. Another name for traditional employees is “W-2 workers,” and another name for contractors is “1099 workers;” these terms come from an Internal Revenue Service (IRS) distinction.3 This is because employees file a W-2 form and contractors a 1099 form, and they are treated differently for tax filing purposes.4

For purposes of this paper, I do not discuss in detail the tax differences between the employees and contractors; instead, I focus on the regulatory distinctions between these two groups. From a regulatory perspective, companies have to meet a set of labor requirements for employees. Some of these labor laws include the Fair Labor Standards Act (FLSA),5 which requires employers to meet minimum wage and overtime requirements for their employees;6 the Employment Retirement Income Security Act (ERISA),7 which regulates the standards employers must meet for aspects of employees’ benefit plans, typically in the context of retirement;8 and the Family Medical Leave Act (FMLA),9 which demands employers provide eligible employees with up to twelve weeks of unpaid leave per year when those employees face vital life circumstances.10 Employers do not have to meet any of these same requirements for workers who are contractors with the company.11

The problems this legal distinction presents today are both the uncertainty surrounding whether on-demand workers should be classified as contractors or employees and, as I argue below, the consequences of attempting to fit this new type of worker into either of the two worker classifications. The uncertainty has led to hundreds of class-action lawsuits across the country on worker misclassification issues. For example, Uber has at least fifteen pending misclassification lawsuits in federal and California state courts alone.12 Uber is also being sued in, at least, Florida and Illinois.13 This list does not include Uber’s most high-profile lawsuit, O’Connor v. Uber Technologies, Inc.,14 which, after being litigated for three years, reached initial settlement in April 2016 that was later rejected by a judge.15 The settlement is still pending. Similarly, Lyft is facing numerous misclassification lawsuits and settled a high-profile one in February 2016 with its California drivers.16 HomeJoy, a company that provided similar on-demand household services like Handy, was sued multiple times after its launch. The former CEO of HomeJoy cited the ongoing legal battles as one of the main reasons they had to shut down the company.17 Handy is now also being sued in California and Massachusetts for the same misclassification issue.18 Instacart recently reclassified part of their workforce into part-time employees rather than contractors amid pressure of these misclassification battles.19 These are a handful of examples illustrating the surge of misclassification lawsuits facing on-demand companies, and in Appendix 1, I provide a table of some of these major class-action lawsuits.20

The companies argue that their non-traditional workers (e.g., the drivers, the cleaners, the phlebotomists) are contractors rather than employees because they have independence, flexibility, can work any hours they choose, and can work for competing companies, among a host of other factors that make them different from the companies’ standard employees. Plaintiffs argue that companies are treating contractors like employees by exerting control and imposing requirements on many aspects of their job.21 For example, in the lawsuit against Handy Technologies, the plaintiff argues that Handy’s requirements include fixing service prices, telling workers how the cleaning must be performed, and instructing them as to what cleaning products they must use.22 Handy argues these are just the guidelines for joining their platform because they want to ensure good customer service.23

There is clearly some uncertainty regarding whether this new type of worker (the “on-demand worker”) should be classified as employees or contractors under the law. In the remainder of this paper, I investigate these tensions and confusions between the worker classification laws and the on-demand economy jobs. In doing so, I argue that the uncertainty indicates that the old bifurcation of the law is not capturing a new type of worker that has emerged in the on-demand economy. I show this by analyzing what the service would be like it if it were clear that the new workers were definitely employees or definitely contractors. The result in both cases is an inferior and different type of service than the one that exists in the market today. I then conclude that there is some unique value created from the on-demand workers occupying this “other space” between contractors and employees, and I explore the idea of a third way solution. This does not have to be a radical change, as there are already “third way” types of workers in labor law (e.g. wait staff, agricultural workers, etc.).

The paper will proceed as follows: In Section II of this paper I look into the factors for determining an employee versus contractor relationship and illustrate the ambiguity created by attempting to place the on-demand suppliers into either category. In Section III, I provide an analysis of what the service would look like if the on-demand suppliers were transformed to fit clearly into either the employee or the contractor definitions. In Section IV, I conclude and discuss the implications.

II.    Contractors or Employees?

There are three tests to determine a worker’s classification type: the common-law test, the economic realities test, and a hybrid test that includes factors from both the common law and the economic realities test.24 The classification of a worker as an employee or an independent contractor can vary because different tests have been applied to different federal statutes.25 For example, the common-law test is used by the IRS, and the courts have applied this test to the Federal Unemployment Tax Act, the income withholding tax, and ERISA, among others.26 On the other hand, the economic realities test is used by courts for statutes such as the FLSA and the Americans with Disabilities Act (ADA).27 Thus, a worker can be classified as an employee under one test, but as an independent contractor under another test.28

The class-action misclassification lawsuits discussed in this paper deal entirely with the FLSA or a state’s own labor laws on things such as minimum wage and overtime regulations. As such, I will focus on the standards of determining whether a worker is an employee or contractor as covered under the FLSA. Before the passage of the FLSA, the common law “control test” was the most used test for determining whether a worker was an employee or contractor.29 When drafting the FLSA, Congress specifically rejected this common law “control test” for the “economic realities test.”30 The economic realities test looks to see whether the worker is economically dependent on an employer, or whether the worker is indeed in business for himself or herself. Thus, Congress preferred this definition because the broader definition covered more workers as employees.31 In a recent issue to clarify the distinctions, David Weil, head of the U.S. Department of Labor’s Wage and Hour Division, explained that, “A worker who is economically dependent on an employer is suffered or permitted to work by the employer.”32 The factors enumerated in the economic realities test help determine whether the worker is economically dependent on the employer (making him or her an “employee”) or whether the worker is really in business for himself or herself (thus making him or her an independent contractor).33

Furthermore, each state has its own version of the economic realities test and may place different weight on the factors or use stricter versions of the test, but they substantively follow the FLSA factors outlined below. For example, California and several other states passed the “presumptive employee law,” which means that in cases where employment status is an issue, the Division of Labor Standards Enforcement starts with the presumption that the worker is an employee.34 More-over, in applying the economic realities test, the most significant factors for California are whether the employer exerts control over the worker and the manner and means in which the work is performed.35 For the purposes of this paper, I will use the Department of Labor’s baseline (as opposed to going state by state) for how to determine whether the worker is an employee or contractor. The factors of the economic realities test include: (1) Integrality, (2) Risk, (3) Investment in Facilities, (4) Skill, (5) Continuing Relationship, and (6) Right to Control.36

Below I explain these factors and provide an application to Uber. I use Uber for simplicity: because the nature of these on-demand jobs are similar, this can be applied to Lyft, Handy, or many of the other on-demand economy jobs.

(1)   Integrality: “[T]he extent to which the work performed is an integral part of the employer’s business.”37

The idea behind the first factor is that “[i]f the work performed by a worker is integral to the employer’s business, it is more likely that the worker is economically dependent on the employer.”38 Work can be integral to a business even if the work is just one component of the business and/or is performed by hundreds or thousands of other workers.39

Application: It can certainly be argued that Uber drivers are not involved in an occupation that is separate and distinct from Uber’s core business. Uber’s main business is in connecting riders and drivers, and as such, drivers are seen as an essential component of Uber’s regular business operations. However, it can also be argued that Uber’s main business is the development and maintenance of the Uber app—dealing with all the matching problems, surge pricing, ensuring enough supply-side uptake to make it a useful application, vetting for quality drivers and riders, and a host of other technological functions.40

       (2)   Risk: “[T]he worker’s opportunity for profit or loss depending on his or her managerial skill.”41

If a worker does not exercise managerial skill that affects his or her profit, the worker is more likely considered an employee.42 This does not include whether the worker chooses to work more hours in order to increase his/her profits.43 Examples of using managerial skill to affect profits could be things like workers engaging in advertising activities, purchasing materials and equipment, deciding which jobs to perform, managing timetables, or other managerial-type actions that can lead to greater profits or losses for the worker.44 The basic point here is that if a worker is exercising managerial skill that affects his/her opportunity for profit or loss, this can be an indication that this worker is more like an independent contractor.45

Application: Under this factor, it seems as if the Uber drivers could be exercising managerial skills to affect their profits. Uber drivers can decide to perform jobs during times when there is surge pricing and can go to certain areas during certain times when they believe there will be surge pricing or go into areas where there will be increased demand for their services. Uber drivers can also decide to risk and purchase a more expensive “black car” to be on the UberBlack platform, which charges a higher price to customers, but is less frequently used than UberX.46 Drivers can also add extra “perks” to the experience (having bottled water for customers or playing their music) in order to obtain a greater ranking or tip from the customer. On the other hand, once a driver accepts to be connected to a specific customer, he cannot easily “turn down” the ride. The Uber platform does not show the customer’s end location until the Uber driver has indicated that he has “picked up” the passenger.47 This means that once an Uber driver accepts a ride request, he must pick up his passenger, even if he no longer finds it worthwhile to drive to that end location once it is revealed to him on the smartphone app. If he chooses to cancel the ride, there are repercussions.48 This feature of the Uber platform can perhaps be seen as either a limit to his managerial skills to affect profits or an indication that the drivers are not really exercising managerial skills in this respect.

       (3)   Investment in Facilities: “[T]he extent of the relative investments of the employer and the worker.”49

If workers are contractors, they should make some investments that involve at least some risk for a loss.50 However, these investments need to be considered relative to the employer’s investments.51 For example, if a worker makes some investments (buying equipment or tools), but these investments are miniscule compared to the investments an employer is making for the employee—things like purchasing vehicles or advertising strategies or providing insurance—then the worker’s relative investments are not sufficient enough for this factor.52 Workers who are more like contractors should be making a majority of the investments required to do their job and workers who are more like employees will have greater investments made by their companies.53

Application: Uber does not provide the most essential tool of this job—the car. Drivers who want to sign up with the Uber platform must have their own cars.54 Uber can provide a used iPhone for a weekly fee if drivers do not have one.55 A recent development regarding this factor, however, is that Uber has partnered with rental car companies and dealerships to get special rates for its drivers.56 Under this factor, and especially because it is a relative investment point, it seems as though drivers are making their own significant investments to do this job. And in fact, this was one of the arguments used by Judge Chen in O’Connor v. Uber Technologies, Inc.57 when he discussed factors that made it seem as though the drivers were independent contractors: “all drivers invest considerably more in tools and equipment by obtaining their own vehicle than Uber does by arguably providing certain of its drivers with a smartphone.”58

      (4)   Skill: “whether the work performed requires special skills and initiative.”59

This factor states that individuals who are using specialized skills or initiative to perform a job are more likely to be seen as contractors.60 However, the U.S. Department of Labor emphasizes that the mere fact that workers are skilled is not itself indicative of independent contractor status.61 A highly skilled carpenter who does not make any independent judgments at the job site beyond the work that he is doing for that job is just a skilled employee.62 By contrast, an independent contractor would need to demonstrate skill and initiative such as investing in marketing, determining the number of jobs he would like to complete, or ordering his own materials.63

Application: This factor is not very informative for the Uber example, and has not been cited much. Some have discussed that driving is not a special skill, so it is a point against the argument that Uber drivers are contractors.

(5)   Continuing Relationship: “[T]he permanency of the relationship.”64

If a worker lacks a permanent or indefinite relationship with an employer, that can be indicative of an independent contractor status.65 The lack of permanence with one company and not relying on that one company as his or her primary source of income would make a worker more like an employer’s contractor rather than its employee.66

Application: Under this category, it seems as though drivers are not in any permanent or indefinite relationship with Uber. Uber openly tells its drivers they can be involved in another occupation67 and does not restrict them from working for multiple other ride-sharing companies, even for its main competitors like Lyft.68 Especially in large cities, many times a worker will be driving with two or three different companies and will have multiple platforms open all at once on his dashboard. Most drivers working for Uber (eighty-five percent) also work less than thirty-five hours a week for Uber.69 All of these facets and the general “at-will” nature of the job highlight the impermanence of the relationship between the worker and the employer in Uber’s case.

(6)   Right to Control: “[T]he degree of control exercised or retained by the employer.”70

If the worker controls meaningful aspects of his or her work, this can be an indication that the worker is an independent contractor.71 However, things like flexibility in worker schedules (i.e., a worker saying that he only wants to work until noon everyday) or the ability to work from home are not signs of “lack of control.”72 Instead, a business’s lack of direct control would mean that the worker can work as many or as few hours as he pleases and the worker can negotiate his own price with his customers, among several other features.73

Application: It is unclear which interpretation this factor supports. It seems as though Uber does exert at least some significant control over the drivers. Uber conducts background and vehicle checks, monitors the approval ratings of its drivers, and deactivates the accounts of drivers who fall below a rating of 4.6 stars and who have a high cancellation rate.74 Uber also exerts control by setting the prices for the customers—meaning that drivers cannot negotiate their own prices with riders. On the other hand, drivers can set their own hours and work whenever they please and for however many hours they wish to work. Drivers are also not supervised on the job and can accept or reject any customer. All of these different facets of the work make it difficult to determine whether the employer does have significant control over the driver. It suggests that there is strong control in some areas and weak control in other areas, and that makes it hard to determine which way this factor sides.

Uber as a company is not unique to this analysis. Take, for example, Uber’s main competitor, Lyft. Almost all of the above discussions would also apply to the relationship between Lyft and its drivers. As is the case for Uber, Lyft drivers are not restricted from driving with other ride-sharing companies. In fact, in a random survey of ten thousand California Lyft drivers, it was found that over half of the drivers have driven with another ride-sharing company.75 And furthermore eighty-three percent of drivers indicated that they drove with Lyft and another ride-sharing company in the same week, and of those that drove with another ridesharing company in the same week, seventy-five percent indicated that they drove with another ride-sharing company besides Lyft in the same hour.76 Data from the recent Lyft California lawsuit also indicate that less than one percent of California Lyft class drivers (755 out of 150,000) worked thirty or more hours in at least half of the weeks they drove with Lyft.77 And, over one hundred thousand people (more than two-thirds of the class members) “have driven less than sixty hours in total for Lyft.”78 In discussing the economic realities test for Uber or Lyft, the point is not to argue whether drivers should be classified as either employees or contractors—rather the goal is to demonstrate that the relationship is a lot like both employment and contracting, which suggests that it is difficult to regulate in a binary employment regulatory scheme. In 2015, when Uber and Lyft were facing two separate suits in U.S. district court in San Francisco, U.S. District Judge Vince Chhabria in Cotter v. Lyft, Inc.79 noted that it was unclear whether either the employee or contractor label applies to the drivers, and turned the decision over to a jury. He wrote:

At first glance, Lyft drivers don’t seem much like employees. . . . But Lyft drivers don’t seem much like independent contractors either. . . . [T]he jury in this case will be handed a square peg and asked to choose between two round holes. The test the California courts have developed over the 20th Century for classifying workers isn’t very helpful in addressing this 21st Century problem.80

He goes on to suggest a solution: “perhaps Lyft drivers should be considered a new category of worker altogether, requiring a different set of protections.”81 Judge Chen, in O’Connor, similarly noted: “The application of the traditional test of employment—a test which evolved under an economic model very different from the new ‘sharing economy’—to Uber’s business model creates significant challenges. Arguably, many of the factors in that test appear outmoded in this context.” 82 Like Judge Chhabria, he suggests a legislative or appellate solution to “refine or revise that test in the context of the new economy. It is conceivable that the legislature would enact rules particular to the new so-called ‘sharing economy.’” 83

Thus, the problem with following this regulation in classifying these sets of on-demand suppliers is that it is not clear whether the Uber or Lyft drivers should be considered employees or contractors. It is difficult to say “they are definitely employees” and it is also difficult to say “they are definitely contractors.” These on-demand workers seem to be occupying a position somewhere in between these two classifications.

III.  A New Type of Job

The wave of uncertainty regarding whether these on-demand supplier jobs fit neatly into either the “contractors” or “employees” classification seems to indicate that perhaps the binary system fails to sufficiently capture the new type of worker that has emerged at the onset of our new economy. To confirm this suspicion, it is helpful to examine a hypothetical on-demand worker at the extremes of the binary system’s two ends. In this section, I will analyze what the on-demand worker job and its respective good/service would look like if it were clear that the worker was an employee. I will then provide a similar analysis if it were clear that the on-demand suppliers were contractors. My argument is that if Uber’s business model changed in a way that made it more certain that drivers are employees, the outcome would be weakening the “on-demand” feature of the service and eliminating the beneficial aspects of an on-demand supplier. And if Uber’s business model changed in ways that make it more certain that drivers are contractors, the result would be a different type of service where Uber would look more like a Craigslist for rides and would lead to a decrease in the quality of the service provided to the customer.

It is important to note here that I am not making a specific prediction about how Uber would respond to a ruling that required drivers to be classified as employees. Uber could respond by shutting down completely (due to the radical increase in its cost structure and the disappearance of efficiency gains from a flexible labor model), or it could change its business model in a variety of creative ways that are outside the purview of this paper, with implications ranging from higher consumer prices to laying off their new employees to restructuring their entire product and image. Instead, my aim is to illustrate that it is inconsistent with the core of Uber’s business when its drivers look clearly like employees or when its drivers look clearly like contractors. In other words, Uber ceases to be Uber when its drivers fit neatly into either category.

A.    Drivers as Employees

In O’Connor, Judge Chen argued that it was unclear whether Uber drivers were employees because “Uber does not control any of its drivers’ schedules—all Uber drivers are free to work as much or as little as they like,” and later discussed how all class members “drove passengers without supervision.”84 He also considered how Uber “exercises no control” over the location of the driver or what routes the driver chooses to follow.85

Judge Chhabria in Cotter similarly argued that Lyft drivers do not seem like standard employees because “Lyft drivers enjoy great flexibility in when and how often to work—far more flexibility than the typical employee” and that drivers had “minimal contact with Lyft management,” which did not actively supervise them.86 He also argued that a driver might treat Lyft as a “side activity, to be fit into his schedule when time permits and when he needs a little extra income.”87

Following this discussion in utilizing the factors of the economic realities test and following the overarching notion that an employee is someone who is economically dependent on the employer, what adjustments can be made to make it certain that the drivers on the Uber platform are employees? What do Uber drivers who are clearly employees under the economic realities factors look like?

Under factor (2) Risk, Uber could weaken the opportunity for drivers to exercise managerial skills to affect their profits. Taking away the uncertainty of the second factor’s influence would make it easier to classify the drivers as employees. Currently, drivers have greater freedom to choose which jobs to accept and how they are going to respond to the surge pricing in a nearby neighborhood. Uber could eliminate this opportunity and automatically assign jobs to the drivers. The drivers would indicate that they would like to work for a certain period of time, and then Uber could eliminate the mechanism that allows drivers to accept or reject certain jobs. By doing so, drivers would have less of an opportunity to impact their profits.

Furthermore, under factor (5) Continuing Relationship, Uber could set up certain rules for its drivers to make the relationship between Uber and the drivers more permanent. For example, Uber could forbid drivers to work for other companies, and it could force drivers to work at least thirty-five hours a week. By doing so, Uber would transform the relationship between the driver and the employer to reflect greater permanence and indefiniteness.

Under factor (6) Right to Control, Uber could exert greater control over the drivers to make it clearer that they are employees. Uber could set the drivers’ schedules and exert greater supervision over their workdays. As mentioned above, Uber could require that drivers accept all rides during their hours of work. Uber drivers would have less independence regarding when they can work or when they want time off.

From the product side, what would Uber’s service look like in the hypothetical case where drivers looked more like standard employees? While it is certainly not impossible for Uber’s service to continue operating even if drivers were clearly employees, the reliability and on-demand aspects of Uber’s service would diminish. It would look similar to what taxi cab drivers do now. For many cities, taxi drivers rent out their cabs from a taxi medallion owner or fleet. They have the vehicle rented out in time-shifts—in New York City, taxicabs are rented in twelve-hour shifts by drivers, and it costs them about $95–139 per shift.88 To recoup the costs, cabs drive around hunting for passengers in that twelve-hour shift. If drivers were effectively employees, Uber’s model would be similar, with the exception that riders do not have to hail the car or call a dispatcher. Riders could still use their smartphones to call the drivers. Drivers for Uber would functionally resemble a taxi cab driver with a smart phone, which is what is happening in New York City already with the launch of Curb.89 Curb is an app that taxi drivers use to connect with riders, which is similar to the other ride-sharing apps. Nothing else about the taxicab model has changed—they are now just cab drivers with a smartphone.

But the “cab drivers with a smartphone” is a fundamentally different service than Uber because the Uber business model utilizes an algorithm that matches supply and demand. Within the algorithm, there is a built-in incentive structure that increases the pay to drivers (i.e., surge pricing) to encourage drivers to provide rides when that service is needed most, and this instantaneous supply-demand matching requires a flexible labor model (along with a nonstandard wage and compensation structure). That is the core of Uber’s business model, and it is functionally different than having workers drive around during a given period of time in a given location and waiting for an app or dispatcher to connect them. In fact, this difference is precisely why economists Judd Cramer and Alan Krueger found that the capacity utilization rate—measured by “the fraction of time a driver has a fare-paying passenger in the car while he or she is working, and by the share of total miles that drivers log in which a passenger is in their car”90 —is significantly higher for UberX drivers than for taxi drivers.91 That is, UberX drivers spend a significantly higher fraction of their time, and drive a substantially higher share of miles, with a passenger in their car than do taxi drivers.92 Their paper concludes that part of the reason Uber drivers are able to do so is because Uber’s flexible labor model and surge pricing more closely match supply with demand throughout the day.93

Thus, this hypothetical transformation would weaken the “on-demand” feature of the service. It is true that Uber could theoretically use its data to figure out the most popular times that people need rides at each location. Then it could staff drivers to be in that particular area during those particular times. On average days, it may get it right, though there would likely be idle drivers as there are with taxi drivers. But this new Uber would not be able to predict the circumstances like rainy days, local concerts, or subway malfunctions and delays. The beauty of Uber’s business model is that they do not need to plan or predict anything in advance to meet the immediate needs of the riders. The on-demand aspect of these on-demand services comes from users requesting more Uber rides on rainy days when Manhattan’s cabs are nowhere to be found. The surge pricing mechanism sends a signal to drivers to get out on the road. In a paper studying how drivers respond to surge pricing, M. Keith Chen and Michael Sheldon conclude that, “Uber partners both drive at times with higher demand for rides, and dynamically extend their sessions when surge pricing raises earnings.”94 Uber drivers are able to respond to the surge pricing because their smartphone interface “allows them to know current prices and session statistics like cumulative earnings, time, and trips.”95

Other on-demand companies would see a similar change in their services if the workers were transformed into employees. For Handy, the company would bear a strong resemblance to a local management company that staffs cleaners. The main difference between those two models would then be that the customer either requests the maid by tapping the app or requests the maid by calling the company. Handy would then also have to predict the most requested times and staff employees accordingly to meet customer demands. Like Uber, Handy has a “peak pricing” algorithm that incentivizes more Handy cleaners to supply their services at popular times.96

Thus, changes to the above factors would help eliminate the uncertainty about whether Uber drivers are in fact employees while also significantly altering the services and products provided by the on-demand economy. With the proposed changes, there would be very little disagreement that drivers on the Uber platform are economically dependent on an employer and therefore should be considered “employees” under the law.

Next, what would be the impact of this on the nature of the on-demand supplier job? Uber’s business model (and those of other companies who have followed suit) attracts many individuals who would like supplemental income opportunities and flexible working conditions that may not be available in the traditional labor market. It may be the case that some workers do utilize Uber as their primary source of income. But this is only true for a fraction of them. According to a Benenson Strategy Group survey, sixty-nine percent of drivers have other full-time or part-time work and half of Uber drivers use the platform for less than ten hours a week.97 Seventy-three percent of the drivers said they would rather have a job where you choose your own schedule and where you are your own boss, as opposed to the traditional model of a steady 9-to-5 job with some benefits and a set salary.98 Furthermore, sixty-three percent of Uber drivers said they specifically use Uber to have more flexibility so they can balance work and family.99 Drivers for Lyft look similar. A survey of 3,100 Lyft drivers found that eighty-two percent of Lyft drivers agreed or strongly agreed with the statement “I like being an independent contractor” and ninety-nine percent of Lyft drivers agreed with the statement “I like to choose when I work.”100

The most general characteristic across these on-demand suppliers is that they have the freedom to select the jobs they want and they can work the hours they want, whenever they want to work them. They can take time off for a long period. These characteristics are not standard of a typical job. For people who have personal obligations (i.e., caring for children or elderly parents, going to school), it is difficult to be in a standard salary or hourly-set job. Thus, there is at least some value generated in the labor market from having these “freelance” drivers, or household cleaners, or phlebotomists. If Uber’s business model changed in way that made it more certain that drivers are employees, the outcome would be elimination of the beneficial aspects of an on-demand supplier.101 The fact that there is some value generated on both sides—on-demand supply and the on-demand product—indicates that there is some value in having this driver not be a “real” employee, as defined by the law.

B.    Drivers as Contractors

What adjustments would Uber need to make in order to transform drivers from “neither-employees-nor-contractors” into just contractors?

Under factor (2) Risk, Uber could give greater freedom to drivers in canceling the rides once the driver and the rider have been matched. Or, Uber could just allow drivers to see the end location of the rider and make a decision ex ante about whether they wanted to accept or reject the ride given the end the location.

Furthermore, Uber could make significant changes to factor (6) Right to Control, which would help eliminate this uncertainty about the classification of drivers. Recall that the major problem argued in these lawsuits is that Uber exerts a lot of control over its drivers. Uber does this by monitoring the drivers and removing them from the platform if they have below a 4.6 ranking.102 It also performs driver background and DMV checks.103 Another point of contention is that Uber controls the price and does not give drivers the freedom to negotiate their own prices. Furthermore, Uber has a whole set of requirements for the drivers’ cars—for one, drivers must have “new” vehicles (no older than 2001) to be listed on the platform.104 In order to transform the Uber driver into an obvious contractor, Uber could give drivers more freedom by allowing them to negotiate their own prices with passengers. Uber could also stop prohibiting drivers from transacting with riders just because they have a quality rating below 4.6 stars. This is actually one of the main complaints by Uber drivers and it was also one of the main complaints filed by the Lyft drivers in the California misclassification lawsuit.105 Moreover, Uber could forgo conducting background and DMV checks on their drivers. All of these facets of control have been cited by the plaintiffs and judges. For example, in Berwick v. Uber Technologies, Inc.,106 California determined that the driver (Berwick) was an employee rather than an independent contactor because the work performed by the driver was an integral part of Uber’s business model and because of the control Uber maintained over drivers.107 In a summary, the Labor Commission concluded:

Defendants are involved in every aspect of the operation. Defendants vet prospective drivers, who must provide to Defendants their personal banking and residence information, as well as their Social Security Number. Drivers cannot use Defendants’ application unless they pass Defendants’ background and DMV checks. Defendants control the tools the drivers use; for example, drivers must register their cars with Defendants, and none of their cars can be more than ten years old. . . . Defendants monitor the Transportation Drivers’ approval ratings and terminate their access to the application if the rating falls below a specific level (4.6 stars). 108

What would the outcome be if Uber made those changes to fit their drivers into the legal definition of a contractor? The result would be a new type of service where Uber would be more like a Craigslist for rides or a job-listing site (e.g., Yellow Pages). It would be a platform where drivers and riders connected and then negotiated their own terms of pickups, price, and the like. Uber’s role would be a more explicit middleman that charges a percentage from drivers for helping them to connect to riders. Thus, if Uber relaxed most of their control over the drivers, Uber drivers would almost certainly look like a typical contractor, as defined by the law, and the service would look like Craigslist or Yellow Pages where you are paying for access to some information/a pool of potential clients, rather than engaging in a valuable matching service. To reinforce this point, consider the benefits of Uber setting the price. Not needing to negotiate a price bilaterally is a lot of value, and without price setting power, Uber would lose surge pricing entirely, as well as almost all of the informational value of pricing.

Another problem is that the more Uber has to relinquish “control” over aspects of their platform in order to fit into the original definition of a contractor, the quality of the service provided to customers will decrease. Uber’s true business model is not Craigslist for drivers. Uber is in the business of vetting for quality because it wants to be seen as a reliable ride. Uber’s idea of the service is to have all of these features in place such that it can increase the quality of the ride-sharing experience and lower the persistence of unfavorable rides; in other words, it does not want high risk and “bad” drivers on its platform. If a driver has a low ranking, Uber has an incentive to deactivate the driver because of the type of service it is providing—a reliable way to get around.

Thus, if drivers were definitely contractors, Uber would be precluded from vetting from various types of quality on its platform. But there is a value to having a quality-controlled service, which is more than just listing or delisting services. While no formal survey data is available yet, there are mechanisms in place that force Uber drivers to be safe. As mentioned, precisely because Uber drivers will be deactivated if they fall below a 4.6 rating, drivers have an incentive to drive more safely. Uber also immediately deactivates any drivers when there are complaints of sexual assault or harassment by the riders.109 Compare this to standard taxi systems of handling misconduct: first the passenger needs to obtain either the license number or plate number of the cab; then the passenger files a complaint (via phone or online) through a meticulous procedure of answering “what, where, when, who,” etc.; then when the complaint is processed by the city, the passenger will be contacted by a prosecutor to further investigate the complaint.110 Finally, the driver will be notified of the complaint, at which point he/she has the opportunity to plead guilty, which then opens up a whole new set of legal procedures.111 This, of course, becomes an even more difficult task if the passenger is a tourist of a foreign country. This is not to say that Uber’s process is always the desirable one, but rather that there is a value to consumers for using a ride-sharing system that contains stronger quality mechanisms.

There are, of course, many instances of Uber drivers speeding, driving carelessly, or being involved in sexual assault or harassment issues with passengers. The point is not that the quality of Uber rides is perfect, but rather that the quality would get worse if Uber drivers were transformed to fit into the legal definition of a contractor. The mechanisms in place that help riders get home safely are also ones that are cited legally as Uber “exerting” control over their drivers. There is a value to getting into a stranger’s car that has been vetted for quality: the background and DMV checks, the “new” car restrictions, and the 4.6 star ratings are all meant to provide a service of a better-quality experience.

This is not to say that a “Craigslist” for drivers would not be valuable service either. It would just be a different type of service than what Uber is providing right now. As discussed above, one of the main aspects of Uber’s service is the reliability of the ride. And thus, in order to provide this reliable ride, Uber wants to help ensure that their workers have reliable cars and they are reliable drivers who also will not cancel on you when they have accepted your ride.

The fact that drivers do not fit perfectly into the contractor definition and that the service would look different if they did fit perfectly into that definition means there is some value being provided by Uber by occupying this middle space where it is not clear whether the drivers are employees or contractors. The outcome of Uber drivers being contractors would be a different (perhaps inferior) good, so from the customer’s perspective, it is not necessarily a “negative” that Uber monitors rankings and exerts control over ride cancellations, background checks, and car checks. That is, Uber is providing some valuable service by operating in the different space that is not captured by the bifurcation of the law.

IV.  Conclusion

While it’s true that simple categorizations cannot possibly be a perfect fit for each situation, they also should not create a host of ongoing problems and uncertainties. And in fact, when they do, it may indicate that the problem rests with the categorization itself. This is the case for the employee-independent contractor bifurcation in current labor law. The legal uncertainty on this issue is creating inefficiencies and problems in the economy for all parties—including workers and the companies. Emerging on-demand startups might also be unsure about the legality of their business models. The ongoing litigation battles are costly and continue to leave us uncertain, and even judges have called for a legislative solution.

I do not argue whether the Uber driver should legally be an employee or contractor. Rather, I argue that the uncertainty over whether the Uber driver is an employee or contractor indicates that the bifurcation of the law is not capturing a new type of worker that has emerged in the on-demand economy. When we attempt to fit the worker into what is clearly an employee or what is clearly a contractor, the outcome is a different (perhaps less desirable) type of service than the one we see on the market. If it were clear that drivers were employees, the valuable aspects of independence and labor flexibility would worsen for the workers, and the reliability and on-demand feature of these services would fall. Similarly, if it were clear that the drivers were contractors, the outcome would be a Craigslist for riders and drivers rather than a quality-controlled ride. In other words, the less control that Uber exerts on its drivers, the more they look like the legal definition of independent drivers, but then the less of a reliable and “quality” ride they become for users.

Rather than restrict a new type of job and the on-demand services and suppliers that are emerging as we enter into an information economy, the law should change to reflect these modifications in the economy on both the consumer and the supplier sides. This may include having “middle-path workers” that are under some labor regulations but not others. And it may include having a “third-way” tax classification. This is not a radical idea. There are “third way” categories for all types of workers already in labor law. Wait staff, for example, have a different compensation structure and model, and as a category of workers they have their own sets of rules and regulations that are separate from standard employees.112 Agricultural workers are also their own category of workers under the law and receive certain exemptions because the nature and compensation structure of their job do not fit neatly into boxed categories of labor law.113 It is not certain exactly what this new type of worker should look like under the law, but it is important that the law should be updated when it has become outdated and inapplicable to the realities of the modern world. Former Deputy U.S. Secretary of Labor Seth Harris and economist Alan Krueger have proposed a few guiding principles for one idea of a third-way category, which they call the “independent worker.” They provide an analysis of which labor law exemptions would make sense based on the nature of this type of job. For example, one of their guiding principles is the “immeasurability of work hours,” and it refers to the challenges of gig economy workers in determining “work” and “nonwork” (and even “for whom work”); as a result, the authors argue that it makes little sense to apply hour-based rules such as the minimum wage and overtime regulations. They conclude their analysis by saying that, “Legal rules defining ‘independent workers’ can and should more closely reflect the actual experience of workers in that category than the current definitions of ‘employee’ and ‘independent contractor.’”114 Furthermore, countries like Canada have created two types of contractor categories, “dependent” and “independent,” to better capture the different nature of the jobs.115

The problem is that the employee-contractor bifurcation laws in the United States were developed more than seventy years ago at a time when it may have been easier to classify workers into either contractors or employees. An “on-demand” service and thereby “on-demand” suppliers were non-existent at the time of the law, and as such the unique features of this new type of job were not captured by the law. The United States is headed toward an information economy and the laws that were written in the 1930s may be too outdated and inapplicable for the new type of jobs that are emerging as a result of the growth in our economy.


  • 1The On-Demand Economy, (last visited Jan. 30, 2017) (The definition used by the coalition of on-demand companies is as follows: “The On-Demand Economy is defined as the economic activity created by digital marketplaces that fulfill consumer demand via immediate access to and convenient provisioning of goods and services.”). “Gig” and “sharing” economy have also been used to describe this platform.
  • 2David Adeleke, How Uberification Is Changing the World, Techcabal (Oct. 16, 2015),,
  • 3See IRS, Forms and Associated Taxes for Independent Contractors (Dec. 2, 2016),,; IRS, Depositing and Reporting Emp-loyment Taxes (Dec. 14, 2016),,
  • 4Employees (the W-2 workers) have payroll taxes automatically deducted from their paychecks, and they split paying with their companies the tax that covers Social Security and Medicare taxes. On the other hand, companies do not withhold income taxes from contractors, and contractors pay 15.3% of their income directly to the government for Social Security and Medicare taxes. As a result, contractors pay the full 15.3% of the tax while employees only pay 7.65% (their companies pay the other 7.65% for them). However, contractors are able to write-off their direct expenses for tax-purposes whereas employees are not able to write-off their work expenses for tax purposes. See IRS, Employer’s Tax Guide 1–2 (2017),; IRS, Self-Employment Tax (Social Security and Medicare Taxes),,; IRS, Travel, Entertain-ment, Gift, and Car Expenses 33 (2016),,
  • 5Fair Labor Standards Act, 29 U.S.C. § 201 (2012).
  • 6Id.
  • 7Employment Retirement Income Security Act, 29 U.S.C. § 1001 (2012).
  • 8Id.
  • 9Family Medical Leave Act, 29 U.S.C. § 2601 (2012).
  • 10Id.
  • 11Charles J. Muhl, What Is an Employee? The Answer Depends on the Federal Law, Monthly Lab. Rev., Jan. 2002, at 4.
  • 12See, e.g., Price v. Uber Techs., Inc., No. BC554512 (Cal. Super. Ct. Aug. 12, 2014); Del Rio v. Uber Techs., Inc., No. 3:15-cv-03667 (N.D. Cal. Aug. 11, 2015); Berger v. Uber Techs., Inc., No. 3:16-cv-00041 (N.D. Cal. Jan. 5, 2016); In re Uber FCRA Litig., No. 3:14-cv-05200 (N.D. Cal. Nov. 24, 2014); Ghazi v. Uber Techs., Inc., No. CGC15545532 (Cal. Super. Ct. Apr. 28, 2015); Richardson v. Uber Techs., Inc., No. RG15775562 (Cal. Super. Ct. June 24, 2015); Zine v. Uber Techs., Inc., No. BC591351 (Cal. Super. Ct. Aug. 14, 2015); Narsis v. Uber Techs., Inc., No. BC599027 (Cal. Super. Ct. Oct. 27, 2015); Tabola v. Uber Techs., Inc., No. CGC16550992 (Cal. Super. Ct. Mar. 16, 2016); Barajas v. Uber Techs., Inc., No. CGC16550198 (Cal. Super. Ct. Feb. 1, 2016); Aquino v. Uber Techs., Inc., No. BC608873 (Cal. Super. Ct. Feb. 3, 2016); Adzhemyan v. Uber Techs., Inc., No. BC608874 (Cal. Super. Ct. Feb. 3, 2016); Gollnick v. Uber Techs., Inc., No. CGC15547878 (Cal. Super. Ct. Sept. 11, 2015); Mokeddas v. Uber Techs., Inc., No. RG16807483 (Cal. Super. Ct. Mar. 14, 2016); Uber Techs., Inc. v. Berwick, No. CGC15546378 (Cal. Super. Ct. June 16, 2015) (appealing Labor Commissioner award).
  • 13For Illinois, see Trosper v. Uber Techs., Inc., No. 1:2016-cv-04842 (N.D. Ill. May 1, 2016); for Florida, see Rojas v. Uber Techs., Inc., No. 1:16-cv-23670 (S.D. Fla. Aug. 26, 2016).
  • 14201 F. Supp. 3d 1110 (N.D. Cal. 2016).
  • 15Id.; Joel Rosenblatt, Uber’s $100 Million Driver Pay Settlement Rejected by Judge, Bloomberg (Aug. 18, 2016),,
  • 16Cotter v. Lyft, Inc., 176 F. Supp. 3d 930 (N.D. Cal. 2016).
  • 17Carmel Deamicis, Homejoy Shuts Down After Battling Worker Classification Lawsuits, Recode (Jul. 17, 2015),, (“The company had already been facing growth and revenue challenges, but CEO Adora Cheung said the ‘deciding factor’ was the four lawsuits it was fighting over whether its workers should be classified as employees or contractors.”).
  • 18Emmanuel v. Handy Technologies, Inc., No. 1:15-cv-12914 (D. Mass. Jul. 7, 2015).
  • 19Brad Stone, Instacart Reclassifies Part of Its Workforce Amid Regulatory Pressure on Uber, Bloomberg (June 22, 2015),,
  • 20The table provides additional information on whether the company is being sued under state or federal labor law. There are also administrative lawsuits on this issue. See Order, Decision, or Award of the Labor Commissioner, Berwick v. Uber Techs., Inc., No. 11-46739 EK (June 3, 2015). This was an important misclassification lawsuit (but not class action) that was filed with the California Labor Commissioner in September 2014, and a decision for the Plaintiff was reached in June 2015—the Commissioner found the Uber driver to be an “employee” and awarded her $4,000 as reimbursement for business expenses and interest. Uber has filed an appeal to the California Superior Court, where it is now pending. See Berwick, No. CGC15546378.
  • 21It is important to note that many of these companies require independent contractors to sign private arbitration agreements, in addition to class waivers. As a result, large aspects of these cases involve the companies asking judges to compel arbitration.
  • 22Class Action Complaint for Damages at 8, Zenelaj v. Handybook, Inc., No. 3:14-cv-05449-TEH (N.D. Cal. Dec. 15, 2014).
  • 23See Alison Griswold, Dirty Work, Slate (July 24, 2015),,
  • 24Muhl, supra note 11, at 5.
  • 25Id.
  • 26Id. at 6 exhibit 1.
  • 27Id.
  • 28See id. at 7 (“Accordingly, a worker could be classified as an employee for the purposes of dealing with one Federal law, such as the Fair Labor Standards Act, but as an independent contractor under another, like FICA.”).
  • 29U.S. Dep’t of Labor, Wage & Hour Div., Adm’r’s Interpretation No. 2015-1, The Application of the Fair Labor Standards Act’s “Suffer or Permit” Standard in the Identification of Employees Who Are Misclassified as Independent Contractors 1 (July 15, 2015) (withdrawn as of June 7, 2017).
  • 30Id. (citing Walling v. Portland Terminal Co., 330 U.S. 148, 150–51 (1947)).
  • 31Id. at 2.
  • 32Id.
  • 33Muhl, supra note 11, at 7, 9.
  • 34State of Cal. Dep’t of Indus. Relations, Independent Contractor Versus Employee,,; see also AFL-CIO Dep’t for Prof ’l. Emps., Misclassification of Employees as Independent Contractors Fact Sheet 2016 at 7 & n.46,, (citing Catherine K. Ruckelshaus and Sarah Leberstein, NELP Summary of Independent Contractor Reforms: New State and Federal Activity at 6, Nat’l Emp. L. Project (Nov. 2011)).
  • 35State of Cal. Dep’t of Indus. Relations, supra note 34 (California’s primary factor is the “right to control,” and its secondary factors include all of the FLSA factors.); see also Cotter v. Lyft, Inc., 60 F. Supp. 3d 1067, 1076 (N.D. Cal. 2015) (quoting S.G. Borello & Sons, Inc. v. Dep’t of Indus. Relations, 769 P.2d 399, 402 (Cal. 1989)) (denying cross-motions for summary judgment and noting that the court “recognized, as ‘logically pertinent to the inherently difficult determination’ of employee status, the six-factor test employed by other jurisdictions”).
  • 36Adm’r’s Interpretation No. 2015-1, supra note 29, at 4 (clarifying the factors).
  • 37Id.
  • 38Id. at 6.
  • 39Id.
  • 40See Uber Help, What Is Surge?,, (last visited Sept. 30, 2017) (for background information on surge pricing and supply issues).
  • 41 2015-1, supra note 29, at 4.
  • 42Id. at 7.
  • 43Id.
  • 44Id.
  • 45Id. at 7 n.7 (“This factor should not focus, however, just on whether there is opportunity for profit or loss, but rather on whether the worker has the ability to make decisions and use his or her managerial skill and initiative to affect opportunity for profit or loss.”).
  • 46See UberBLACK: Professionally Driven Black Cars, Uber,
    black/, (last visited Sept. 30, 2017).
  • 47See Requesting and Taking an Uber Ride, Uber Help,, (last visited Sept. 30, 2017).
  • 48See Uber Community Guidelines, Uber,, (last visited Sept. 30, 2017) (“There are several ways we measure driver quality, with the most important being Star Ratings and Cancellation Rate. . . . Your cancellation rate is based on the number of trips canceled out of the total number of trips you accept. . . . Each city has a maximum cancellation rate, based on the average cancellation rate of drivers in that area. . . . If your cancellation rate continues to exceed the maximum limit, you may lose access to your account.”).
  • 49 2015-1, supra note 29, at 9.
  • 50Id.
  • 51Id.
  • 52Id. (citing Dole v. Snell, 875 F.3d 802, 810–811 (10th Cir. 1989); Sec’y of Labor v. Lauritzen, 835 F.2d 1529, 1537 (7th Cir. 1987); Baker v. Flint Eng’g & Constr. Co., 137 F.3d 1436, 1442 (10th Cir. 1998); Hopkins v. Cornerstone Am., 545 F.3d 338, 344 (5th Cir. 2008); Real v. Driscoll Strawberry Assocs., Inc., 603 F.2d 748, 755 (9th Cir. 1979)).
  • 53Id. at 10.
  • 54See Driving Jobs vs Driving with Uber, Uber,, (last visited Sept. 30, 2017).
  • 55Taylor Soper, Uber Now Charging Drivers $520 per Year to Lease Company iPhone, GeekWire (July 25, 2014),,
  • 56Nicole Arata, 5 Ways to Get a Car You Need to Drive for Uber or Lyft, USA Today (Jan. 8, 2017), (discussing these new developments).
  • 57No. C-13-3826 EMC, 2015 WL 5138097 (N.D. Cal. Sept. 1, 2015).
  • 58Id. at *25.
  • 59 2015-1, supra note 29, at 4.
  • 60Id. at 10.
  • 61Id. (citing Brock v. Superior Care, Inc., 840 F.2d 1054,1060 (2d Cir. 1988)).
  • 62Id.
  • 63Id. at 11.
  • 64Id. at 4.
  • 65Adm’r’s Interpretation No. 2015-1, supra note 29, at 11–12.
  • 66Id. at 12.
  • 67See, for example, Uber directly advertising itself as a potential part-time job: “It’s simple and perfect for those looking for seasonal, work from home, entry level, temporary, or any type of part time job opportunity.” Uber Driving Partner - A Great Part Time Opportunity, Uber,, (last visited Sept. 30, 2017).
  • 68See O’Connor v. Uber Techs., Inc., No. C-13-3826 EMC, 2015 WL 5138097 (N.D. Cal. Sept. 1, 2015), at *18 (Uber attorneys are cited indicating that “Uber does not limit drivers’ ability to seek and obtain employment with third-party employers” and specifically that Uber “never restricts [drivers] from engaging in another occupation or business, and never restricts them from simultaneous use of other apps like Lyft and Sidecar.” (emphasis in original)).
  • 69Jonathan V. Hall & Alan B. Krueger, An Analysis of the Labor Market for Uber’s Driver-Partners in the United States 18, (Princeton U. Working Papers, Jan. 2015),
  • 70 2015-1, supra note 29,at 4.
  • 71Id. at 13.
  • 72Id.
  • 73Id. at 14–15.
  • 74Order, Decision, or Award of the Labor Commissioner at 9, Berwick v. Uber Techs., Inc., No. 11-46739 EK (June 3, 2015); see also Uber Community Guidelines, supra note 48. Previously, Uber also deactivated drivers who had been inactive for more than 180 days. This policy is no longer listed on their website, but Uber attorneys had alluded to this in O’Connor: “Uber further claims that the right to control element is not met because drivers can work as much or as little as they like, as long as they give at least one ride every 180 days (if on the uberX platform) or every 30 days (if on the UberBlack platform).” See O’Connor v. Uber Techs., Inc., 82 F. Supp. 3d 1133, 1139 (N.D. Cal. 2015).
  • 75Declaration of Simona A. Agnolucci in Support of Lyft Inc.’s Brief Regarding Preliminary Approval of Class Action Settlement at 2, Cotter v. Lyft, Inc., No. 3:13-cv-04065 (N.D. Cal. Sept. 3, 2013).
  • 76Id. (emphasis added).
  • 77Order Denying Motion for Preliminary Approval of Class Action Settlement at 9, Cotter v. Lyft, Inc., No. 3:13-cv-04065 (N.D. Cal. Sept. 3, 2013).
  • 78Id.
  • 7960 F. Supp. 3d 1067 (N.D. Cal. 2015).
  • 80Id. at 1069, 1081.
  • 81Id. at 1082.
  • 82O’Connor v. Uber Techs., Inc., 82 F. Supp. 3d 1133, 1153 (N.D. Cal. 2015).
  • 83Id.
  • 84O’Connor v. Uber Techs., Inc., 2015 WL 5138097 at *17, *24.
  • 85Id. at *17.
  • 86Cotter v. Lyft, Inc., 60 F. Supp. 3d 1067, 1081 (N.D. Cal. 2015).
  • 87Id. at 1069.
  • 88New York City Taxi & Limousine Comm’n, Notice of Promulgation of Rules,, (last visited Oct. 2, 2017).
  • 89See Curb,, (last visited Oct. 2, 2017).
  • 90Judd Cramer & Alan B. Krueger, Disruptive Change in the Taxi Business: The Case of Uber, 106 Am. Econ. Rev. 177, 177 (2016).
  • 91Id. at 179.
  • 92Id. at 177.
  • 93Id.
  • 94M. Keith Chen & Michael Sheldon, Dynamic Pricing in a Labor Market: Surge Pricing and Flexible Work on the Uber Platform 13 (Working Paper, Dec. 11, 2015),,
  • 95Id. at 15.
  • 96Handy, What Is Peak Pricing?,, (last visited Oct. 2, 2017).
  • 97Uber: The Driver Roadmap 2.0, Benenson Strategy Group, (last visited Oct. 2, 2017).
  • 98Id. (“If both were available to you, at this point in your life, would you rather have a steady 9-to-5 job with some benefits and a set salary or a job where you choose your own schedule and be your own boss?”).
  • 99Id.
  • 100Declaration of Simona A. Agnolucci in Support of Lyft Inc.’s Brief Regarding Preliminary Approval of Class Action Settlement, Cotter v. Lyft, Inc., No. 3:13-cv-04065-VC (N.D. Cal. Mar. 17, 2016).
  • 101Furthermore, one of the features emphasized in Uber’s advertisements to attract more drivers is a slogan of being “your own boss.” Be Your Own Boss: Drive on the Uber Platform, Uber,, (last visited Oct. 2, 2017). Recall that most suppliers want to sign up on the Uber (or Lyft) platform because they value the flexibility and independence they have, but Uber may not be the same “reliable” and on-demand service it advertises itself to be if it cannot attract suppliers to join its platform.
  • 102Order, Decision, or Award of the Labor Commissioner at 9, Berwick v. Uber Techs., Inc., No. 11-46739 EK (June 3, 2015).
  • 103Driver Requirements: How to Drive with Uber, Uber,, (last visited Oct. 2, 2017).
  • 104See, e.g., Vehicle Requirements Chicago, Uber,, (last visited Oct. 2, 2017).
  • 105Notice of Motion and Motion for Preliminary Approval and Memorandum of Points and Authorities in Support Thereof at 1, O’Connor v. Uber Techs., Inc., No. 3:13-cv-03826, 2016 WL 2607229 (N.D. Cal. Apr. 21, 2016) (“The forward-looking non-monetary component of the settlement is significant and includes numerous changes to Uber’s business practices. . . . Specifically, Uber will only be able to deactivate drivers from the Uber platform for sufficient cause, and drivers will be provided with at least two warnings prior to many types of deactivations, a written explanation of the reasons for any deactivation, and an appeals process overseen by fellow drivers for certain types of deactivations.”); Cotter v. Lyft, Inc., 193 F. Supp. 3d 1030, 1033 (N.D. Cal. 2016) (“The settlement did not reclassify the drivers as employees, but it limited the circumstances in which Lyft could terminate drivers, and it created a process by which drivers could challenge certain termination decisions.”).
  • 106No. 11-46739 EK (June 3, 2015).
  • 107Order, Decision, or Award of the Labor Commissioner at 9–10, Berwick, No. 11-46739 EK.
  • 108Id. at 9.
  • 109Uber Community Guidelines, supra note 48.
  • 110Filing a Complaint with the TLC, N.Y.C. Taxi & Limousine Comm’n,, (last visited Oct. 2, 2017); N.Y.C. Taxi & Limousine Comm’n,, form) (last visited Oct 2. 2017).
  • 111Filing a Complaint with the TLC, supra note 110.
  • 112Fact Sheet #15: Tipped Employees Under the Fair Labor Standards Act (FLSA), U.S. Dep’t of Labor,, (last visited Oct. 2, 2017).
  • 113Fact Sheet #12: Agricultural Employers Under the Fair Labor Standards Act (FLSA), U.S. Dep’t of Labor,, (last visited Oct. 2, 2017).
  • 114Seth Harris & Alan Krueger, A Proposal for Modernizing Labor Laws for Twenty-First-Century Work: The “Independent Worker” 25 (2015).
  • 115A. Irvin Schein, Canada: When Is a Worker a Dependent Contractor?, Mondaq, (Aug. 26, 2014),