About this Issue

Mobile computing has enabled a new way to exchange goods and services. But how should governments respond? Regulatory structures now in place were crafted for a set of concerns that may or may not be relevant now, and we increasingly face the prospect of a two-tiered regulatory system.

Yet companies like Uber and AirBnB are poised to do more than just sidestep traditional regulations. They also raise entirely new questions - including concerns about safety, working conditions, and trust. Are new regulations needed just for them? If so, what might they look like? 

Joining this month to discuss are Matthew Feeney, a Research Fellow at the Cato Institute; Dean Baker, co-founder of the Center for Economic and Policy Research; Christopher Koopman, a Research Fellow at the Mercatus Center; and Avi Asher-Schapiro, a freelance journalist who has done notable work on the subject.

Lead Essay

Level the Playing Field - by Deregulating

Lawmakers and regulators have been struggling in the last few years to adapt to the rise of the so-called “sharing economy,” a relatively new and increasingly popular peer-to-peer economic model. Familiar experiences such as catching a ride, paying to sleep in a bedroom, making some money doing odd-jobs, and cooking a meal in exchange for money are all made easier by sharing economy companies such as Lyft, Airbnb, TaskRabbit, and Feastly. Yet despite the fact that many sharing economy services are demonstrably popular they have encountered opposition across the world from market incumbents as well as regulators, who see the sharing economy as a dangerous and arrogant menace rather than a welcome competitive disruption.

Opponents of the sharing economy often argue that sharing economy companies enjoy an unfair lack of regulatory constraints that puts consumers at risk. Supporters argue that while there are a number of safety issues related to the sharing economy it is unclear that they must be dealt with by new legislation or regulation. Sharing economy companies have major financial incentives to protect providers and consumers and have demonstrated that they are capable of implementing safety improvements and features in the absence of regulation. In my view, policy relating to the sharing economy must be as hands-off as possible, not least because attempts to regulate companies such as Lyft and Airbnb in ways analogous to taxis and hotels could limit innovation and be used to engage in regulatory capture. Rather than pursue legislation that seeks to regulate sharing economy companies like their traditional competitors, lawmakers ought to deregulate these competitors in order to make the regulatory playing field as level as possible while allowing for innovative disruptors to enter markets.


Solving an Information Problem

Everyone reading this essay owns something that they value very little but which might be valued very highly by someone else. While someone might consider their collection of old books, baseball cards, or CDs as a waste of space, someone else might think very differently. Yard sales illustrate this very well. Of course, rather than put household junk on the front lawn homeowners can now advertise goods on eBay, which is much better than a yard sale at allowing buyers to find information about sellers.

The sharing economy helps individuals solve similar informational problems. Most people have spare resources, skills, or time that are not used to make money because of a lack of information. Before home sharing companies such as Airbnb a full-time software engineer who owned a property with a spare bedroom would find it difficult to find a visitor looking for a room in her city. Likewise, before companies like Uber and Lyft arrived on the scene that individual would also have found it difficult to locate passengers to ferry around town in exchange for money in her spare time. Before the sharing economy the price of finding the relevant information to offer spare bedrooms or rides was oftentimes prohibitive. Thanks to the sharing economy people can more easily make money from spare assets and skills that they do not use in their full-time jobs.

Unsurprisingly, this has proven popular. Uber, one of the best known sharing economy companies, is valued at $40 billion and operates in 54 countries, while Airbnb has been valued at $13 billion and available to property overs in over 190 countries.[1] Yet despite their popularity, sharing economy companies have faced regulatory barriers.


Regulatory Concerns

Many of the regulatory issues related to the sharing economy are due to the fact that companies like Uber and Airbnb fit awkwardly into existing regulatory regimes.

Uber, Lyft, Sidecar, and other rideshare companies allow for private car owners who are not commercially licensed to use a smartphone app to find passengers willing to part with some money in exchange for a ride. There is no doubt that rideshare drivers are competing with taxis, but there does seem to be a clear conceptual difference between someone using their own car to provide rides (oftentimes part-time) and a full-time commercially licensed and insured taxi driver.

Likewise, people listing properties on Airbnb, HomeAway, and FlipKey are competing with hotels without themselves being licensed hotel proprietors, and homes used by Feastly and EatWith providers are not restaurants.

Before the sharing economy, the regulatory distinctions between a private car and a taxi, a spare residential bedroom and a hotel room, and a restaurant and a home dining room were much clearer. Now, sharing economy companies are blurring these distinctions. Regulators and lawmakers across the world have dealt with this conceptual blurring in different ways.

For instance, in Las Vegas officers with ski masks have impounded rideshare vehicles.[2] In contrast, both Colorado and California have passed legislation that legalizes ridesharing and provides insurance requirements. In some parts of Europe, Uber’s ridesharing service has been banned.[3] Airbnb hosts have been issued cease and desist letters as well as fines.[4]

Many of the crackdowns on sharing economy companies occur because they are not operating within the regulatory framework for taxis, hotels, and other industries competing with the sharing economy. Lawmakers and regulators dealing with the sharing economy must decide whether companies operating in this new economic model ought to be regulated like their traditional competitors. I believe that they should not. Indeed, lawmakers and regulators ought to view the rise of the sharing economy as an opportunity to deregulate the sharing economy’s competitors.

It is understandable that companies like Lyft and Airbnb upset people working in the taxi and hotel industry. After all, Lyft and Airbnb are competing with taxis and hotels for business without having to overcome the regulatory barriers taxis and hotels face. Taxi drivers must have commercial licenses and insurance, and in many jurisdictions taxi companies must have medallions in order to operate. These medallions artificially limit the number of taxi drivers and as a result are very expensive. In addition, many jurisdictions restrict not only how many taxis are permitted in a given area but also how far a taxi can travel and how much a ride can cost. Hotel owners must pay occupancy taxes and adhere to regulations relating to disabled access, fire escapes, and so on. Home sharing hosts are not required to comply with these same regulations, although Airbnb is voluntarily collecting local tourism taxes in San Francisco, Portland, San Jose, Chicago, and Washington D.C. Airbnb also recommends that hosts have smoke and carbon monoxide detectors in their homes.[5]

While some regulators are tempted to regulate sharing economy companies like their competitors, to do so would be a mistake that betrays a misunderstanding of how the sharing economy works. It would also threaten to stifle innovation. It might seem that because Uber competes with taxis and Airbnb competes with hotels that Uber and Airbnb should be regulated like taxis and hotels. This reasoning is misguided.

Fundamentally, sharing economy companies are, unlike their traditional competitors, providing information via their technology to individual providers and consumers. Airbnb, Uber, and Feastly do not own their users’ homes, cars, or food. Rather, they own a technology or platform that allows drivers to find passengers, homeowners to find temporary guests, and cooks to find mouths to feed. To fit the sharing economy companies into a regulatory regime that includes regulation that predates smartphones and assumes a very different sort of business model would be inappropriate.

Yet, it is the case that rideshare companies like Uber and Lyft have praised the passage of legislation that allows them to operate legally. Lawmakers and regulators in both Colorado and California have legalized ridesharing, but while it might be tempting to welcome such changes, it is worth remembering that these new regulations could hamper innovation and encourage regulatory capture.

For instance, Californian and Coloradan regulations recognize ridesharing companies like Uber and Lyft as so-called “Transportation Network Companies (TNCs).” The regulations in place for TNCs do allow Uber and Lyft to operate in California and Colorado. But the new TNC designation only adds to the number of regulations governing transportation in California and Colorado and could be abused by established companies such as Uber in order to limit competition.

New regulatory designations for the sharing economy not only offer sharing economy companies the opportunity to influence policymakers and engage in regulatory capture. They are also limiting. At the moment Uber is focused solely on transportation, so the TNC designation seems initially appropriate. However, if in the coming decades Uber expands into the logistics industry, competing not only with taxis but also with companies like FedEx and UPS, the TNC designation will be unhelpful as well as inappropriate.

It is conceivable that companies like Airbnb and Feastly could also be subjected to a regulatory regime that introduces new carve-outs rather than implementing deregulation. Given the regulatory grey area discussed above, this is especially worrying for sharing economy providers using their residential property to make some extra money. We ought to be very wary of potential carve-outs that classify someone’s house as anything other than a private residential property.

Adding to the already extensive pages of regulations faced by businesses in the U.S. amid the rise of the sharing economy may seem like an appropriate response now, but in the long term such moves could end up being abused and potentially used to infringe on property rights. Deregulation will not only allow sharing economy companies to develop and innovate, it will also allow for their traditional market incumbents to do the same.


Safety Concerns

Many of the criticisms leveled against the sharing economy take the form of safety concerns. There have been reports of Uber drivers assaulting passengers as well as Airbnb guests damaging property. Critics argue that rideshare companies and home share companies put customers at risk when they flout the regulations that govern taxis and hotels.

Yet in the case of ridesharing, companies already have stricter background check requirements than most taxi companies and include two-way rating systems that encourage good behavior on the part of the passenger as well as the driver.[6] Likewise, Airbnb guests and hosts are rated. Badly behaved providers and consumers in the sharing economy do not last long, and the lack of anonymity means that anyone who does commit a crime is unlikely to escape justice.

Perhaps one of the sharing economy’s most interesting potential benefits is the emphasis that sharing economy companies put on reputations and good behavior. Uber and Lyft drivers face being dropped if they do not maintain high ratings, and Airbnb bans hosts who do not conform to Airbnb’s ethos and culture or who behave badly. EatWith, which also allows for ratings and reviews, is very strict when considering hosts and boasts that only 4 percent of applicants who apply are accepted.[7] As the sharing economy grows we shouldn’t be surprised if something like a “reputation score” becomes a metric considered along with credit scores. There are already websites such as TrustCloud which rate sharing economy users based on participants’ trustworthiness. Unsurprisingly, sharing economy companies are working on providing safety and good service in the absence of burdensome regulation. Without regulation it is still in the best interests of sharing economy companies that their providers and consumers are not harmed and enjoy their experiences.



Lawmakers who wish to introduce new regulations often cite public safety and market failure. A case cannot be made that the sharing economy is awash with market failure. Nor can a case be made that the sharing economy poses a greater risk to consumers than market incumbents. Indeed, innovations such as two-way rating systems and verified accounts remove provider and consumer anonymity from the sharing economy, ensuring that sharing economy participants have a major incentive to behave well.

Sharing economy companies have highlighted the over burdensome nature of many of the regulations that govern transportation, accommodation, and other industries. It is understandable that people working in these industries are frustrated by sharing economy competitors who have not paid the same regulatory costs that they have. However, regulators should resist the urge to impose old regulations on new innovative companies or to write new regulations that could encourage cronyism and stifle growth. Sharing economy competitors deserve the chance to compete, and regulators can help them do this by deregulating the industries the sharing economy is disrupting. The forces of a free  market, not regulatory bodies, should decide which providers survive and which ones fail.



[1]Serena Saitto, “Uber Raises $1.6 Billion in Convertible Debt to Expand,” Bloomberg, January 21, 2015, http://www.bloomberg.com/news/2015-01-21/uber-said-to-raise-1-6-billion-in-convertible-debt-to-expand.html and Tim Bradshaw, “Airbnb valued at $13B ahead of staff stock sale,” CNBC, October 23, 2014, http://www.cnbc.com/id/102117120#

[2] John Kartch, “Vegas Uber Drivers Hounded by Ski-Masked Agents,” Forbes, November 3, 2014 http://www.forbes.com/sites/johnkartch/2014/11/03/vegas-uber-drivers-hounded-by-ski-masked-agents/

[3] Peter Teffer, “Bad week in Europe for Uber,” euobserver, December 15, 2014 https://euobserver.com/regions/126900

[4] Charlie Brennan, “Boulder issues cease-and-desist orders to 20 property owners using Airbnb or VRBO,” Daily Camera, January 6, 2015 http://www.dailycamera.com/news/boulder/ci_27270016/boulder-issues-cease-and-desist-orders-20-property and Rob Lieber “A $2,400 Fine for an Airbnb Host,” The New York Times, May 21, 2013 http://bucks.blogs.nytimes.com/2013/05/21/a-2400-fine-for-an-airbnb-host/

[6] For more information on rideshare safety see my Cato Institute Policy Analysis “Is Ridesharing Safe?” http://www.cato.org/publications/policy-analysis/ridesharing-safe#jo76Bb:0VY

Response Essays

The Sharing Economy Must Share a Level Playing Field

Last year, millions of mom and pop retailers across the country collected sales tax for state and local governments, as required under the law. Amazon, one of the largest retailers in the world, did not collect the same taxes in many states. The trick is that Amazon sells over the Internet and does not have a physical presence in many states. It is therefore able to claim an exemption from this legal obligation. The amount of the tax in most states is comparable to the profit margin in the retail sector.

This massive implicit subsidy for Amazon should infuriate supporters of the free market everywhere, even if it might have many conservatives applauding. After all, it is difficult to find a rationale for government subsidies to one of the largest corporations in the world to the detriment of local family-owned businesses.

This is the situation that we face with Uber and other “sharing economy” companies. Uber is using the Internet to disrupt the taxi market. In most cities this market has been dominated for decades by a small number of well-entrenched incumbents who used their control over the regulatory apparatus to limit competition. The result was higher prices and often bad service.  

Uber and other new entrants to this market have already led to a substantial improvement in the quality of taxi service in many cities. In addition to the increased supply of taxis, many incumbents have responded to the challenge by improving service quality in the form of newer cars and the adoption of Internet apps. This is good news.

However, the incumbent cab companies are still largely subject to a set of regulations that Uber is trying to evade. Many of these regulations serve legitimate public purposes, even if they may not be the best way to achieve the goal intended.

For example, taxis are required to undergo regular safety and brake checks. This is both expensive and time-consuming. (In many cities the checks require going to understaffed city facilities.) It is certainly arguable that the safety requirements could be relaxed in many cases (e.g. do brand new cars need to be inspected every year?), but it is reasonable to have regulations that ensure an out-of-town traveler who gets a cab at the airport at 2:00 in the morning will find it has working brakes. Of course, this traveler could research on the web, and seek out a cab company that has a record of high standards, but many people may not want to go through this effort after a long day of traveling.

Similarly, drivers for incumbent cab companies typically have to get special chauffer licenses and also go through a criminal background check. The former ensures driver quality and the latter guarantees the safety of a passenger. Again, these requirements both involve time and money. Uber has largely avoided such requirements, although it is now doing some screening of drivers in some cities.

Incumbent taxi services also typically have to carry substantial insurance policies to protect passengers who are involved in accidents. Many Uber drivers are effectively uninsured when they are working because standard policies do not cover commercial driving. Here too, Uber is moving to provide insurance to passengers and drivers.

In addition to this set of regulations, incumbent taxi cab companies generally face some requirement to serve the handicapped. This usually means having a minimum number of handicap accessible vehicles. There is also an issue of serving cash-paying customers. Uber is designed as a credit card only service. Those without credit cards would not be able to use Uber. This would effectively exclude a substantial portion of the population from using taxis, if Uber came to dominate the industry. Since lower income people disproportionately do not have credit cards, this could leave many poor people unable to make trips to the doctor or other necessary travel.

There are also labor market regulations that often apply to incumbent taxi services. They are required to pay for workers’ compensation if a driver is injured on the cab. They also have to pay towards unemployment insurance if a worker loses their job. And they have to meet minimum wage and other labor standards. (This is not always the case with incumbent taxi services, since many treat drivers as independent contractors.)

In all of these cases, regulations serve a public purpose. It may not be necessary to maintain the same scope of regulation in all cases, and it certainly is not necessary to maintain the regulations in their current form, but it does not make sense to have one set of rules that apply to incumbent taxi services and a whole different set that applies to Uber. The appropriate policy going forward should be to modernize the regulatory structure and establish rules that apply equally to Uber and incumbent taxi services.

This should not in principle be a difficult task, but it requires some good faith from the parties involved. Amazon argued for years that it should not be required to collect sales tax because its programmers were not smart enough to keep track of the tax rates in different states. We saw similar silliness from Uber when it shared its data with Alan Krueger, a Princeton economist and former head of President Obama’s Council of Economic Advisers.

Krueger used the data to analyze the gross pay of Uber drivers and compare it to the government data on the net pay of drivers for traditional cab companies. However, he was not able to go beyond this apples to oranges comparison because Uber chose not to share its data on miles driven.  Without some data on miles driven it is not possible to produce estimates of drivers’ costs, and thereby convert gross revenue into actual earnings.

Presumably Uber didn’t make this data available to Professor Krueger because it knew that it would lead to earnings numbers that looked bad compared to those of drivers of traditional cabs. We can only speculate on this question, since Uber is sitting on the data.

But if we scrap the “sharing” nonsense, the basic story is relatively straightforward. We need rules that ensure that cabs are basically safe, that drivers are competent, and without a recent history as dangerous felons. We need to make sure that people with physical handicaps can count on getting taxi service in a timely manner and that we have a system in place so that people without credit cards have access to cab service.

We also need to make sure that both the drivers and passengers are insured in the event of an accident. This includes worker compensation for drivers. In addition, drivers should be able to bargain collectively if they choose, if not as a union covered by the National Labor Relations Act, then as some other entity that would serve the same purpose.

And minimum wage and overtime laws should apply to taxi services. This means that drivers should be assured of earning at least the minimum wage net of expenses and one and a half times the minimum if they work more than a 40-hour week. If it is too complicated for Uber to make such calculations then they will be replaced by firms with more numerate management.

Finally, it is reasonable to have an overall limit on the number of cabs in a city. When congestion and pollution are factored in, the optimal wait time for a cab is not zero. It is likely the case that many cities have set these limits too low in the past, but that does not mean that no limits are the best policy.

Uber deserves credit for ending the dominance of the taxi cartels and modernizing the industry. That’s worth a positive newspaper column. The idea that Uber should be able to operate by its own rules may be the dream of its top executives and holders of Uber stock, but it is not sound public policy.

The Sharing Economy Is Propaganda

Taking an Uber is not sharing, just like buying a footlong meatball sandwich at Subway is not “eating fresh.”

Sharing economies do exist: couch surfing, lending your neighbor a power drill, giving your cousin a lift to the airport. But to pay per-mile for a ride, per-night for a room, or per-hour for a house cleaner is a no different than paying per-loaf at the bakery.  That a smartphone or website mediates the exchange does not make it sharing. And it does not change fundamentally the relationship between consumers, laborers, and management.

The managers in this triad— Uber, HomeJoy, and Taskrabbit—prefer that politicians and commentators call their business “sharing.” At the outset, it’s important to dispense with the Orwellian term, because it promotes misunderstanding and bad policy. It leads many to conclude, as Matthew Feeney does, that the primary function of these businesses is to provide “information via their technology to individual providers and consumers.”

We do not live in an imaginary world of dormant assets, aimless drivers, and idle house cleaners begging for a Yellow Pages 2.0. A sophisticated public policy response to the “sharing economy” must start with an understanding of the markets and human relations that companies like Uber and Airbnb promote—and the social benefit or cost they might entail.

In reporting on Uber, I found many drivers in Los Angeles who drive full time to pay rent and feed their families. Executives called these drivers “partners” who can choose their own hours; “entrepreneurs,” who essentially start their own “business.” The drivers I spoke to found themselves in an entirely different sort of relationship. Uber would cut prices, change work rules, and deactivate drivers at will—leaving many feeling disposable and exploited.

My reporting touched a nerve because coverage of the “sharing economy” up to that point largely ignored the economic realities of these workers, which number in the hundreds of thousands. It makes people uncomfortable to imagine that their happy Uber driver is actually struggling to make ends meet, trading a smile for a 5 star rating.

Feeney doesn’t consider the implications of the sharing economy on workers at all. Public policy should, however, put these workers and their economic predicament at the center. Likewise, an honest appraisal of “asset-sharing” platforms like Airbnb must take into account the effect on affordable housing stock, rental rates, and community cohesion (Rachel Monroe has documented the danger Airbnb poses to affordable housing here).

It’s not a coincidence that the so-called sharing economy exploded in the aftermath of the financial crisis. In the first years of the “economic recovery,” the top 1% slurped up 121% of the income gains. Now six years later, the stock market booms and corporate profits smash records. Yet working people’s wages remain unchanged, and nearly 7 million are stuck in temporary jobs against their will.

Enter Uber.

Evangelists for the sharing economy always talk about how important it is to “disrupt” markets. When I spoke to Uber PR, they told me that introducing GPS hailing technology to the taxi industry solved inefficiencies, benefiting consumers and drivers alike. Many of the practices of the taxi industry are indefensible, and introducing GPS to car services is a welcome, if minor, disruption.

In the long run, however, if the “sharing economy” is disruptive of anything, it is disruptive of hard-fought labor protections. Uber brings on 40,000 new drivers a month. But to avoid minimum wage laws and liability claims, Uber will not admit that it actually employs these drivers—though around 20% work full time.

Uber then extracts millions of dollars from its drivers’ labor and invests in ad campaigns and lobbying efforts to spread to other cities. Right now in New York, if you pull up a map of public transportation on Google, the app offers Uber as an alternative to trains and buses, as if a private car amounts to “public transportation.”

In the short term, Uber is encouraging drivers to take out predatory subprime car loans with the Spanish bank Santander.  But ultimately, Uber wants to replace drivers with robots, as soon as the technology becomes viable.

Matthew Feeney undoubtedly considers these drivers consenting adults in an economic relationship. And that’s true. If you believe in the inherent benevolence of market forces—that desperate workers deserve whatever job appears at the intersection of the labor-supply and labor-demand curve—then the sharing economy should be allowed to run its course.

Thankfully we live in a society that rejects that logic outright. Paid medical leave, workers’ compensation, the 40 hour work week—these are all rights won by working people in the face of strong opposition from corporate interests. The sharing economy is just the latest broadside against these hard-fought labor protections.

Other companies, now smelling the potency of the “sharing” brand, sell themselves as the “Uber for laundry,” or the “Uber for childcare.”  All of these companies prey on a pool of desperate workers left out of the economic recovery.

A world in which everything is available on-demand for the wealthy, courtesy of low-wage workers, is neither desirable nor fair.  It is actually quite similar to the on-demand economies of cities like Cairo or New Delhi. Anyone who spends time in the developing world knows how easy it is to summon a delivery boy, carpenter, or house cleaner at all hours.  This so-called-efficiency is the product of a deplorable wealth gap— it’s no wonder that San Francisco, the birthplace of Uber, now has income inequality that rivals Rwanda.

In the long run, companies like Uber do the opposite of what they promise. Instead of promoting well-paid flexible work, they bring the informal sector under the centralized control of big tech companies. They pit workers against each other, discouraging unionization and collective bargaining—necessary components of a fair labor-management equilibrium. Lured by dishonest claims of high-wages, many drivers sink money into cars they can’t afford and keep driving for Uber even when wages fall.

Policymakers should not be ensorcelled by “sharing-economy” propaganda. While Uber is a great deal for Wall Street—drivers front all the capital and take on all the risk, while investors scrape off profit—it’s a bad deal for workers.

Companies that profit from the labor of hundreds of thousands of people should be forced to behave like responsible employers. Thankfully, U.S. District Judge Edward Chen appears poised to do just that.  Last month, in a hearing for a class-action suit brought against Uber and Lyft, Chen rejected the companies’ claim that drivers are not employees, saying: “If all you were doing is selling an app you could sell it at an app store, but Uber does a little more than that, doesn’t it?”

Technology innovators should, of course, be rewarded for their inventive coding.  But just imagine if Adobe took 20% of the profits from every music producer who used Adobe Premiere to edit sound. That arrangement would never be tolerated by workers in the knowledge economy, and it should not be tolerated by drivers.

If “sharing-economy” corporations don’t want to become employers, they can easily sell their technology to the workers who use it—like everyone else. Government can play a constructive role here by requiring that transportation services be owned by workers’ collectives, or at least employ unionized labor.  To leave the sharing economy space entirely unregulated, as Feeney suggests, would usher in a dystopian future, where the precariously employed hover over their smart phones waiting to be summoned by someone lucky enough to have a full-time job.

Today’s Solutions, Tomorrow’s Problems

In his lead essay, Matthew Feeney provides a full-throated defense against the rush to regulate the new, innovative companies making up the “sharing economy.” Many will find his arguments both familiar and welcome, while others will find it unfair to competitors and potentially unsafe for consumers. This divide is ultimately the central feature of most policy debates surrounding the rapid rise of the sharing economy, and the growth of firms like Uber, Lyft, and Airbnb.

Though many aspects of this debate merit greater discussion, I want to limit my comments to ridesharing and the designation of “Transportation Network Companies.”

Some states – including Colorado, California, and (most recently) Virginia – seem to have struck a balance between the two sides of the debate. These states legalized ridesharing services under the designation of “Transportation Network Companies” (TNCs), while also creating specific regulatory requirements in the name of protecting consumers. Feeney is warranted in his skepticism of these attempts, as well as in his belief that these new laws will create opportunities for rent-seeking and regulatory capture in the future.

However, I would take his apprehension one step further and argue that Feeney’s fears are being realized in real time, and it is the consumer who is left worse because of it. As Freeney notes, the TNC designation is a win for incumbent ridesharing companies – not only because they can now operate legally, but because these new regulations limit the ability of smaller competitors to enter the market. While current ridesharing companies seem to welcome the regulatory protection today, it may prove to be the greatest impediment to continued growth and development of ridesharing in the future.

In terms of limiting competition, the TNC designation is an instant win for incumbents like Uber and Lyft. Organizing and running a successful startup is expensive and complicated enough, but the TNC regulations simply add additional barriers that make it costly for smaller companies to enter the market. Under Virginia’s pending TNC laws, for example, any company wishing to offer ridesharing services must pay the state a $100,000 licensing fee and $60,000 every year thereafter. These costs are relatively insignificant when viewed in the context of companies valued in the billions of dollars; however, they may be insurmountable for most startups looking to begin their own ridesharing venture to compete with those established firms.

Many argue that limiting those who can enter the market in this manner serves some public benefit. After all, limiting entry to protect consumers was the driving justification behind the taxicab regulations that persist throughout the United States. In fact the opposite tends to be true.

As I have noted in my research (with Adam Thierer and Matthew Mitchell), when competition is limited, consumer welfare suffers. Limiting the number of entrants and the ways in which they can compete, all in the name of “consumer protection,” actually undermines both the competitive rivalry within the industry as well as the incentive for firms to distinguish themselves in the level of customer service they provide.

As a result, consumers are often left with higher prices, fewer choices, and lower quality service. Barriers to entry mean that incumbent firms have little need to focus on satisfying consumer desires; instead, their success depends upon their ability to court regulators and retain regulatory protections. Over time, barriers to entry ensure that firms become sluggish, lazy, and less alert to the sorts of entrepreneurial innovations that drive consumers to purchase products or use their services.

The general lack of competition caused by taxi regulations explains the absence of customer care among taxicabs that created the opportunity for ridesharing to become as popular as it is today. Conversely, it is the current state of vibrant competition among ridesharing services that has reduced prices and improved conditions for drivers. It has also brought us on-demand ice cream, puppies, and toilet paper.

Moreover, while firms such as Uber and Lyft may be more than happy to accept the TNC regulations as a way to limit competition today, it may ultimately prove fatal to their own continued growth and development tomorrow. While Feeney believes this classification may limit Uber’s ability to grow outside of ridesharing, UberRUSH, UberMovers, and UberESSENTIALS all demonstrate that the company has moved far beyond simply driving people from one point to another.

Instead, I believe that the TNC designation will ultimately limit the ability of firms to grow within ridesharing. Specifically, the TNC classification is built on two basic assumptions about the relationship between the transportation network company and the cars being used to transport passengers: First, it assumes the transportation network company does not own the car. Second, it assumes that a person drives the car. For those states that have adopted the TNC designation, these two assumptions permeate their regulations.

While this may seem innocuous or benign, these regulations may have unintentionally foreclosed the opportunity for firms like Uber or Google to continue to push the margins of how ridesharing is provided. In particular, the issue of who owns or drives the cars takes on an entirely different character in light of Uber’s recent partnership with Carnegie Mellon University to research autonomous technology, or rumors that Google may offer its own ride-hailing service using autonomous cars.

Bringing these services to consumers would necessitate redefining the term “transportation network company.” Regulators would need to rethink how these companies may interact with drivers, or whether drivers are required at all. Regardless of how those questions are resolved, these issues would need to be re-debated each time ridesharing outgrows its current classification.

In addition, given the standard approach among states that otherwise yet-to-be-regulated ridesharing is illegal, those able to employ on-demand autonomous vehicles would first need permission before doing so. This would create a constant cycle of starts and stops, as regulators continually work to catch up to the sharing economy. Of course, this is all dependent on any given regulator’s willingness to adapt and evolve with a changing economy. If history is any guide, this may be the biggest hurdle to overcome.

Consider the destructive effects that regulation of this nature will have on a rapidly evolving space like the sharing economy. The market, as a process, is driven by entrepreneurs discovering new ways to meet consumer needs. The sharing economy, and its use of technology, has accelerated this process compared to many other industries. Regulation by its very nature imposes a type of stasis on this dynamic process. The juxtaposition between the market process and regulation is magnified by the speed with which the sharing economy is continually antiquating the attempts to regulate it.

Does this mean that doing nothing will leave the sharing economy somehow unregulated? Certainly not.

There are already a plethora of laws in existence – including both civil and criminal – that cover liability, theft, fraud, and other potential harms that many of these new regulations seek to address. Policymakers should first apply the existing legal apparatus before attempting to create a new one. And when the existing framework is unworkable they should, as Adam Thierer explains, strive for simple legal principles rather than complex “technology-specific, micromanaged regulatory regimes.” The watchword going forward should be “permissionless innovation.”

In sum, Feeney is correct in his assertion that regulators should resist the urge to impose old taxicab regulations on – or write new regulations for – the sharing economy. However, regulators should not do so because the industry deserves the chance to compete, but because it will make competition healthier and consumers better off. The sharing economy of today must be allowed to compete with and challenge the business models of the past. And – just as important – tomorrow’s innovators must be able to challenge today’s upstarts, who will soon enough be the incumbents.

The Conversation

Some Clarifying Questions

Before I begin my response I’d like to thank Dean Baker, Avi Asher-Schapiro, and Christopher Koopman for participating in this edition of Cato Unbound, I enjoyed writing the first essay and reading the responses. I hope that Christopher will forgive me for engaging with the comments made by Dean and Avi, who both made some points in their essays that I would like to directly address.

Avi begins his essay with a complaint regarding the use of the term “sharing economy.” I share this complaint. The term is clumsy and potentially misleading. It is true that the so-called “sharing economy” does not rely on the sharing of assets but rather offering use of assets (such as a ride in a car or a spare bedroom) in exchange for money.

What is perhaps most interesting about Avi’s response is that it highlights a disagreement we have regarding the role of regulation and public policy. In his essay Avi states the following:

Public policy should, however, put these workers and their economic predicament at the center. Likewise, an honest appraisal of “asset-sharing” platforms like Airbnb must take into account the effect on affordable housing stock, rental rates, and community cohesion

Unfortunately, Avi doesn’t offer an argument for why policy makers ought to concern themselves with economic predicaments or the effect a company like Airbnb has on housing stock or community cohesion. Who decides what an economic predicament is or what counts as community cohesion? Presumably Avi does have limiting principles that describe what ought to be considered economic predicaments and community cohesion, yet he does not explain what these are. My own belief is that regulations are justified only in some instances of market failure. Avi may not like that some Uber drivers are struggling to make ends meet, but that is not a market failure. In addition, it is far from clear to me that the regulatory response to some people making less money than they would like is increased intervention in a market.

Later in his essay Avi points out that he and I agree that drivers who use Uber are “consenting adults in an economic relationship.” He then goes on to mention a regulation also noted by Dean which highlights what I think is a conceptual error.

Both Avi and Dean discuss the 40-hour work week. Avi claims that the 40-hour work week is an example of society’s rejection of the “logic” of consenting adults engaging in an economic relationship, while Dean states that Uber ought to ensure that drivers who work more than 40 hours a week are paid one and half times the minimum wage.

Yet Uber’s business model relies on its rideshare drivers operating whenever they like. An Uber rideshare driver can drive for two hours a week or fifty. Uber cannot control when a driver using its platform picks up passengers, although surge pricing can provide additional incentives for drivers to get on the road. It seems to me that the sharing economy model offers a good degree of choice for drivers. They can work whenever they want, and if they’re not satisfied with their income they may stop using the service.

Uber also relies on an employment model that is different from the model used by many sharing economy competitors. This is because Uber is not a taxi service and its ridesharing drivers are using their own vehicles. Uber is providing a technology that allows private car owners to find passengers willing to part with some money in exchange for a ride. Unfortunately, Avi does not directly address what it is about the relationship between Uber and the drivers and passengers that makes the “employer” designation appropriate. Is it also the case that Airbnb or Feastly hosts should be considered Airbnb and Feastly employees?

In his essay Dean argues for a set of policies that “apply equally to Uber and incumbent taxi services” and goes on to discuss cash payments, handicap access, insurance, background checks, and congestion. However, Dean does not highlight what it is about current rideshare insurance and background check requirements that needs reform. In fact, although Dean mentions that Uber “is now doing some screening of drivers in some cities,” it is actually the case that at least in the United States every Uber driver undergoes a background check.

Dean and I had a brief discussion related to the cash question at a recent Cato Institute event on ridesharing, which can be seen here at around the 1 hour mark. As I understand his position Dean is not arguing that Uber and similar companies should be forced to accept cash, although at the Cato event he did discuss a fee that could subsidize rideshare cash transactions. I would be interested to hear more about the fee, and I hope that Dean will elaborate in his next response.

I found Dean’s suggestion relating to congestion and vehicle caps particularly interesting. Currently, rideshare passengers hail drivers who turn on the app whenever they want. Perhaps Dean’s suggestion that the number of taxi cabs and rideshare vehicles be capped is mostly aimed at traditional cabs, but when it comes to rideshare vehicles I think the proposed vehicle limit is a solution in search of a problem.

I’ll conclude this response with some questions for Avi and Dean.

For Avi:

  1. If the proper role of public policy is to ensure that economic predicaments are avoided and social cohesion is preserved, how should these be measured or defined?
  2. If Uber were to be classified as an employer, should the same designation be applied to Airbnb and other sharing economy companies?

For Dean:

  1. Are the insurance and background check requirements already put in place by rideshare companies sufficient, or should these requirements be the same as those in place for taxis?
  2. How should the number of rideshare drivers be capped? Why not keep the current rideshare system in place in which passengers hail drivers and the rideshare platforms change prices in order to provide incentives for drivers to search for passengers at times of high demand?


Uber’s Problems Are Real, and the Regulatory Solutions Are Simple

I will make a few quick points to clarify some arguments I made in my original piece and also to respond to some comments at the Cato forum.

First, the point on the optimal number of cabs should not be controversial. Cabs clearly impose externalities in the form of increasing congestion and pollution. We should not want Uber cars waiting outside every house and business just in case someone wants to get somewhere. Are  too many cabs likely to be a problem? Certainly it won’t be if Uber drivers are getting paid reasonable wages, a point to which I will return below, but it is certainly possible that Uber and other services will substantially increase congestion in some cities.

The extent to which this could be a problem depends on what an Uber ride is replacing. If it’s simply a question of Uber replacing a traditional cab, then there is no net change in congestion. If an Uber ride is an alternative to the person driving herself then it would increase congestion since the driver must drive to pick-up the person and then drive from the drop-off to another destination. If the Uber ride is in place of public transit, bicycling, or walking or alternatively causes an additional trip to be made that would not otherwise be made, then it is increasing congestion.

This would be a non-issue if cities had in place some sort of congestion pricing, since the Uber car would then be assessed the appropriate fees. But at least in the United States we don’t have congestion pricing in cities, which means that a proliferation of Uber-type services could be a problem. That doesn’t require a medallion system. We could have a system of flexible fees by time of day and location of service, but it is reasonable to assess fees on such services. It certainly does not make sense to assess fees in the form of medallion requirement for traditional cabs and then exempt Uber because customers use a web-based app.

I probably should have fleshed out the issue of cash-paying customers a bit more in my earlier reply. A range of people use traditional taxis for a variety of reasons. Some are relatively affluent and take cabs for convenience. However there are also older and lower income people who may be dependent on cabs for things like major grocery shopping or going for doctor visits. We expect cabs to serve these people. There is undoubtedly some element of cross-subsidy with relatively generous tips from more affluent people subsidizing the trips of lower income customers who may tip their drivers little, if anything.

If Uber and related services skim off the affluent customers, the traditional taxi services may no longer be viable. This could leave this older and poorer segment of the population without necessary transportation service.

This is an appropriate concern for public policy. I suggested that it would be possible to make cash cards available to people who did not have credit cards so they could use Uber-type services. (There is also an issue if the only taxis available have to be summoned with a smart phone and many people don’t have smart phones.) There are many ways in which this problem can be solved, but we should recognize the potential loss of an implicit cross-subsidy and be prepared to develop policies that will address any resulting gap in meeting transportation needs.

The last point I wanted to address was the treatment of Uber drivers. I argued that they need employee-type protections. Uber seems to be operating with the idea that Uber drivers are pseudo-employees when they have a paying customer in their car, but the moment the customer steps out, the driver no longer has a connection to Uber.

This is not a serious standard. It would be like an employer saying that a retail clerk is on the payroll when they are attending to a customer, but they are off the clock the moment the conversation or exchange ends. The exact nature of employer-employee regulation can be argued, but it doesn’t make sense to say that an employer can end it any time the worker is not directly bringing in money for the employer. In the case of Uber, this should mean that when the driver is going to pick up a rider and when they are returning from dropping off a rider, they are still Uber’s responsibility. This would mean being covered by Uber’s insurance and should also mean that these hours should be counted when assessing pay for minimum wage and overtime purposes. (Again, what the minimum wage should be and whether we should have one are arguable points. It is not reasonable to say that it shouldn’t apply to Uber because people order it over the web.)

The fact that Uber is web-based, and they have records of where people are picked up and dropped off, means that it should not be difficult for Uber to know when someone is actually doing Uber-related driving as opposed to taking their family for a picnic. This should make it a relatively simple matter to monitor Uber drivers’ time and ensure that wage and hour standards are being met. (The issue would be average pay, not that a driver will necessarily get the minimum wage for every hour they do Uber-related driving.)

To repeat the basic point of my original piece, there are legitimate regulatory issues created by Uber. These are not difficult to address if they are dealt with honestly. 

Make Safety Standards Universal

On Matthew Feeney’s first questionI don’t know the specifics of Uber’s background and safety checks, but I would want rules written into law. They should not be “Uber only” rules. The rules they apply may in fact be good ones, in which case they should be ones that are applied to the incumbent taxi industry as well as start-ups that may come to compete with Uber in future years. As I’ve said, I recognize the traditional barriers were often put in place in large part to protect incumbents, but the goal should be uniform rules that meet a public purpose. If Uber has largely designed those rules, that’s great, we can appropriate them from the company and apply them to everyone.

I think I’ve answered his second question, on how to limit the number of rideshare drivers, in my earlier reply.

What Is the Sharing Economy?

While the bulk of this conversation has focused on regulating the sharing economy, little time has been spent actually defining what the sharing economy is and is not. The lack of a shared definition is why Matthew Feeney can call it “a relatively new and increasingly popular peer-to-peer economic model,” and why Avi Asher-Schapiro can call it “propaganda,” and how they can both be correct. To perhaps help clarify some of these issues, I’d like to propose a simple definition for the sharing economy.

I agree with Avi’s point that taking an Uber might not be sharing. And I would argue that there may be little sharing actually occurring in the sharing economy.  But that really isn’t the point.  Instead, it is helpful to think of the sharing economy as my colleagues and I have defined it before: any marketplace that uses the Internet to connect distributed networks of individuals to share or exchange otherwise underutilized assets.

When people talk about the sharing economy, they are very rarely focused on remuneration. The term is simply being used as shorthand to describe firms that offer a platform to connect individuals who have something with those who need it.

A cash-strapped homeowner may not have seen her spare bedroom as capital until the Airbnb platform provided a way for her to rent it out to vacationers. A college student with an extra hour between classes may not have viewed his time as a profit opportunity until Instacart and TaskRabbit allowed him to put that time to use for others. A young couple may not have been able to use their couch to connect with other travelers from around the world, but can now do so through Couchsurfing. A retiree with a workbench full of power equipment may not have viewed his tools as a way to supplement his income until 1000 Tools connected him with people in his area wanting to borrow tools.  This is the sharing economy.

While some may choose to call this “the peer economy,” ”peer production,” “the collaborative economy,” or “collaborative consumption,” each of these are simply different attempts to describe the shifts taking place in the way individuals are choosing to transact and interact with one another.

Regardless of what you are calling it, it has very real benefits.  In recent research, Adam Thierer, Matthew Mitchell, and I highlight five distinct ways that the sharing economy is creating real value for both consumers and producers:

  1. By giving people an opportunity to use others’ cars, kitchens, apartments, and other property, it allows underutilized assets or “dead capital” to be put to more productive use.
  2. By bringing together multiple buyers and sellers, it makes both the supply and demand sides of its markets more competitive and allows greater specialization.
  3. By lowering the cost of finding willing partners, haggling over terms, and monitoring performance, it cuts transaction costs and expands the scope of trade.
  4. By aggregating the reviews of past consumers and producers and putting them at the fingertips of new market participants, it can significantly diminish the problem of asymmetric information between producers and consumers.
  5. By offering an “end-run” around regulators who have been captured by existing producers, it allows suppliers to create value for customers long underserved by incumbents that have become inefficient and unresponsive.

Finally, it is also important to remember that many of the policy problems that Avi Asher-Schapiro and Dean Baker presented are often failings of particular firms and business models within the sharing economy and not problems with the entire industry. Viewed in this light, each and every one of these failures represents a profit opportunity for a new or rival firm interested in improving the customer experience. Preemptively regulating an entire class of firms based on these anecdotes can be dangerous. As Jim Dwyer of the New York Times recently warned, “Be careful around anecdotes; they are the black ice of reality.” The sharing economy is too diverse and too rapidly evolving, and these sorts of pixel-sized stories should not be mistaken for larger, universal truths.

The real issues should not be lost in the noise.  Are people sharing? Not always. But, then again, that really isn’t what the sharing economy is about. Instead, they are benefitting from mutually beneficial interactions that would not be possible without the sharing economy’s platforms.

Sharing Economy Realities

I appreciate Matthew Feeney’s thoughtful reply and his important questions.

It seems obvious that our disagreement is largely ideological. He trusts unregulated markets are both efficient and fair, while I suspect those markets aren’t value neutral and rather tilt in favor of business interests. This isn’t a dispute we can resolve here in the pages of Cato Unbound.

But I do think my reporting on Uber can clarify some of our disagreement, and as I wrote before, it may help inject the actual lived experiences of Uber drivers into this discussion.

I met S.J. while working on a piece about Uber’s military recruitment initiative, Uber Military. It’s an ambitious program to bring 50,000 veterans on as drivers. Backed by former Secretary of Defense Robert Gates and General Stanley McChrystal, the initiative has been widely successful, enlisting 10,000 drivers in the first six months.

While serving in the marine corps, S.J. threw out his back carrying a machine gun that weighed as much as he did. He was honorably discharged, moved back home to Los Angeles, and started looking for work. At the time, Uber was were reaching out to veterans, and S.J.  heard CEO Travis Kalanick claim drivers make $100,000 a year. So he signed up.

When we first, spoke S.J. described himself as a “totally free market guy.”  He admired Kalanick’s entrepreneurial machismo. He saw Uber much the same way Feeney does—a good deal for riders, a free market innovator, and an opportunity for enterprising hardworking types to excel.  When S.J.  first started driving for Uber in May, 2014 he cleared $1,200 a week, not exactly $100,000 a year, but a good living.

In preparing this response, I caught up with S.J. He quit Uber in January after his pay dipped to $400 a week. All summer and into the fall, Uber cut its rates in L.A. to edge out competition from SideCar and Lyft (unlike AirBnB hosts who can set their own rates, Uber drivers have no say in the matter). S.J. was able to quit because he hadn’t taken out a loan to finance his car. But many of his co-drivers aren’t so lucky—Uber partners with predatory subprime lenders like Santander to link up its drivers with easy loans, locking drivers into a cycle of debt.

I asked S.J. if he still thought Uber was an example of free market efficiency. It’s fascinating how out of touch Uber is with its drivers, he told me. “Uber used unethical practices to hook drivers onto their platform… and not everyone is working with the same set of information.” He says the job is fine for some side cash, but it’s totally untenable as a full-time job, even though the company continues to market it as such. Uber, he says, encourages unstable employment and preys on those who’ve fallen on hard times and are desperate for any paid work.  “I feel like a lot of people are stuck,” he said.

Above all, S.J. worries about the many recent immigrants who worked with him at Uber—many trusted Uber’s wage projections, took out loans, and are still riding it out, hoping rates will rebound. Without the protection afforded to regular employees in our society (and admittedly these protections could be improved markedly) the economic fate of these drivers is in the hands of billionaire entrepreneurs in Silicon Valley.

Driving for Uber actually shifted S.J.’s political outlook: “I still support the free market, but something has to be done to account for exploitation,” he said. “As long as the company can keep duping drivers into signing up and taking out loans, it has no incentive to improve working conditions… that’s where regulation comes in.”

I agree with S.J.

As for the specifics of such an intervention, I think Dean Baker is on the right track: Uber shouldn’t be allowed to set its own safety standards, termination policies, and wages, outside of the current regulatory framework just because it uses an app to drum up business. Minimum wage laws, insurance mandates, and overtime pay should all apply to drivers who work on the Uber platform. As I suggested in my first essay, if Uber doesn’t want to act like a responsible employer, it can charge a licensing fee for use of its product and walk away.

In his thoughtful response to my initial essay, Feeney asked me to provide a standard of economic “predicament” that would warrant government intervention. Here’s a good starting point: work should pay a living wage, include healthcare, pension benefits, and sick leave—in short all the minimum requirements to pursue a dignified life.

Feeney did bring up “market failure” as his own benchmark for intervention. The U.S. economy now contains over 7 million people forced to work part-time jobs, despite being willing and able to do full-time work. A multi-billion dollar tech company increasingly misleads these workers and corrals them into subprime car loans. If this doesn’t count as “market failure,” I’m not sure the term retains any use.

To specifically answer Feeney’s’ question about Airbnb: since the company does not require its “hosts” to adhere to any pricing structure, it wouldn’t be fair to classify them as employees. Rather, Airbnb is just a platform that allows unlicensed (and untaxed) hotels. As you’d expect, this process is increasingly dominated by capital-rich landlord and speculators—not individuals with a spare room. An investigation by the New York Attorney General found that in New York, just 6% of hosts were responsible for 36% of bookings—one operator alone accounted for 3,000 reservations and $6.8 million in revenue.  That trend will only accelerate, and Airbnb will increasingly be dominated by larger proprietors.  New York should levy taxes on these businesses. In the long term, it may also be prudent for regulators to further discourage speculation—which drives up rental prices—by restricting Airbnb to the actual residents of a home.