Here’s the solution to the Uber and Airbnb problems — and no one will like it

It’s been a fascinating week to watch the war between Uber and the De Blasio administration play out.

Not surprisingly, Uber ended up carrying the day using a combination of its dedicated user base and its sophisticated political machine.

This is yet another very early round in what will be a long and hard war — not just between Uber and NYC, or Uber and other cities, but between every high-growth startup innovating in a regulated sector and every regulator and lawmaker overseeing those sectors.

Watching the big battles that have played out so far — in particular around Uber and Airbnb — we’ve seen the same pattern several times over: new startup delivers a creative and delightful new service which breaks the old rules, ignoring those rules until they have critical mass of happy customers; regulators and incumbents respond by trying to shut down the new innovation; startups and their happy users rain hellfire on the regulators; questions arise about the actual impact of the new innovation; a tiny amount of data is shared to settle the dispute.  Rinse and repeat, over and over.

I am not sure there’s a near term alternative to this process — new ways of doing things will never see the light of day if step 1 is always “ask permission”.  The answer will nearly always be no, and new ideas won’t have a chance to prove themselves.

Luckily, though, we have somewhat of a model to follow for a better future.  It’s the way that these new platforms are regulating themselves.  My colleague Brad has long said that web platforms are like governments, and that’s becoming clearer by the day (just look at Reddit for the latest chapter).

The primary innovation that modern web platforms have created is, essentially, how to regulate, adaptively, at scale.  Using tons and tons of real-time data as their primary tool, they’ve inverted the regulatory model.  Rather than seek onerous up-front permission to onboard, users onboard easily, but are then held to strict accountability through the data about their actions:


Contrast this with the traditional regulatory model — the one government uses to regulate the private sector, and it’s the opposite — regulations focus on up-front permission as the primary tool:


The reason for this makes lots of sense: when today’s regulations were designed (largely at the beginning of the progressive era in the early 20th century), we didn’t have access to real-time data.  So the only feasible approach was to build high barriers to entry.

Today, things are different.  We have data, lots of it.   In the case of the relationship between web platforms (companies) and their users, we are leveraging that data to introduce a regulatory regime of data-driven accountability.  Just ask any Uber driver what their chief complaint is, and you’ll likely hear that they can get booted off the platform for poor performance, very quickly.

Now, the question is: how can we transform our public regulations to adopt this kind of model?  Here’s the part that no one will like:

1) Regulators need to accept a new model where they focus less on making it hard for people to get started.  That means things relaxing licensing requirements (for example, all the states working on Bitcoin licensing right now) and increase the freedom to operate.  This is critical for experimentation and innovation.

2) In exchange for that freedom to operate, companies will need to share data with regulators — un-massaged, and in real time, just like their users do with them.  AND, will need to accept that that data may result in forms of accountability.  For example, we should give ourselves the opportunity to enjoy the obvious benefits of the Ubers and Airbnbs of the world, but also recognize that Uber could be making NYC traffic worse, and Airbnb could be making SF housing affordability worse.

In other words, grant companies the freedoms they grant their users, but also bring the same data-driven accountability:


That is going to be a tough pill to swallow, on both sides, so I’m not sure how we get there.  But I believe that if we’re honest with ourselves, we will recognize that the approach to regulation that web platforms have brought to their users is an innovation in its own right, and is one that we should aim to apply to the public layer.

Over at TechCrunch, Kim-Mai Cutler has been exploring this idea in depth. In her article today, she rightly points out that “Those decisions are tough if no one trusts each other” — platforms (rightly) don’t trust regulators not to instinctively clamp down on new innovations, and regulators don’t trust platforms to EITHER play by the existing rules OR provide in-depth data for the sake of accountability.

In the meantime, we’ll get to observe more battles as the war wages on.



  • De Blasio must not be familiar with Uber’s tech. Otherwise, he could have easily parried Uber’s claim that capping new drivers was hurting underserved minority neighborhoods. He could have said Uber can keep adding new drivers if it didn’t let the new ones pick up riders in Manhattan below 96th Street. Uber has the capability to do that easily (the same way they keep Uber NJ drivers from picking up in NY).

  • Very interesting. Maybe there is a role for some intermediary organization. Some type of impartial escrow of data.

    • Yeah I think that might be right

      Also needed is analysis which regulators likely won’t have full capacity for in house

      • Bill Goodwin

        Right, because overlooked in this excellent analysis is another trust issue: I don’t trust the government to have my data. As willing as private corporations are, at times, to use and misuse data, there are market feedback mechanisms. The last thing I want the government to have is information about where I travel and who I stay with, and when. A capacity to audit data provided? Perhaps. But more data centralized in government hands? No thanks.

    • Noel Niles

      Quis custodiet ipsos custodes? Who will watch the watchers? I hope you are joking. There really is no way apply a legal warrant to this kind of data and at the same time ensure that the data in question is: 1. Applicable to the warrant, 2. Valid, and 3. Secure said data from improper access.

      I’m sure nothing bad could happen by trusting the government with personal information such as religious background. What’s the harm?

      • of course a valid question

        the answer could be something to do with aggregated data – for example it would be possible to ascertain the impacts of uber on traffic or airbnb on housing prices without revealing granular personal data

        • I doubt that it’s possible to mask granular personal data. Numerous experiments have demonstrated otherwise.

    • Jannick Malling

      Owned/funded by who?

      • Maybe on the blockchain, w multi sig control

      • A corporation, like.. Verisign or something. Or a non-profit, funded by both corps and gov grants.

  • Interesting to consider how this “use first, self-regulate later” could apply to consumers. For example, we could skip driving tests, and just have drivers rate each other, with the low-rated ones losing their licenses.

    Or make it easier to buy and carry firearms, with users losing the privilege due to low ratings.

    • Maybe!

    • Esp if we used real-Time GPS and accelerometer data

      • Or traffic cameras. But there’s the concern with those that municipalities might use them dishonestly to collect revenue via tickets.

        • Delana Quinn

          The problem that we’re skirting is that government has gotten out of control and no longer represents nor serves the interests of the people. Unless and until the people step up and make their own rules and force the government back in its place, it’ll always be a case of whiny candyass citizens begging for permission to live at all.

          The Great Disruption is at hand and the middle man is being eliminated…and people are taking back their own power to choose and decide what’s best. The B2B/P2P model clearly works. – books, movies, food, music, transportation, energy – it’s all being dismantled by one on one style business models that keep a barrier up from government meddling.

          This needs to continue and when the government figures out how much power its losing and goes full draconian, the people need to step up and not back down.

    • Isn’t the firearm model already in play in the united states? It’s easy to buy firearms (clearly) and users often lose their privilege (jailed) due to low ratings (killed someone.)

      • I suppose, but the data part isn’t wired in very well

        • prison census and murders are well known figures.

      • There are background checks required to buy firearms nationally, and then different states have their own requirements on top of that for a permit to buy one. My state’s process is particularly onerous. Put it this way: if it were as difficult to vote here, Romney would have won the state in 2012. And that’s just for a permit to buy a weapon to keep in your home – there’s a separate and more difficult process to apply for the right to carry one here.

        • I still don’t get it. How exactly is one “rated” (and come to receive high or low ratings) while passively carrying a firearm?

          • It’s a binary “rating” system. Other legal firearm carriers who don’t draw on you are giving you a positive rating; those who do draw on you are giving you a negative rating. According to John Lott (“More Guns, Less Crime”), it seems to work.

            More generally, though, my comment meant to highlight the contradiction between the laissez-faire attitude Nick’s colleague Fred Wilson has toward startup regulation, and his embrace of Bloomberg-style strict regulation of individuals/consumers. Further, this dichotomy brings to mind a recent blog post elsewhere, that quotes a prescient Georgetown history professor about the endgame of liberal and conservative thought:

            Conservatives now are telling us that we must curtail government, cut government spending, cut government powers, reduce government personnel for the sake of making individuals more free.

            “Liberals, on the other hand, are still telling us, as they have for a long, long time, that in order to make individuals free, we must destroy communities. By communities I mean villages. Ghettos and cities. Ethnic groupings. Religious groupings. Anything which is segregated. We must destroy them. So that all individuals would be, if possible, identical. Including boys and girls.

            “But the area of political action … in which you have government, individuals … three others: voluntary associations (which I’ll say no more about), corporations and communities. And if the liberals destroy communities for the sake of the individual, and the conservatives destroy the government for the sake of individuals, you’re going to have an area of political action in which irresponsible, immensely powerful corporations are engaged in opposition to individuals who are socially naked and defenseless.

          • Delana Quinn

            Or, if they show up on HLN or Nancy Grace you can consider their ratings low ;p

  • berkana

    Brilliant insight. Our police and other public servants should be held accountable using real-time data. That would be awesome.

  • mikemikemikemikemike

    This is a naive approach to solving a “problem” that is already being solved by force. Giving the government data in real-time is not going to replace the fact that the government (read: largest gang) is the problem in the first place, and quite frankly, it’s unnecessary, because the government (read: largest gang) is quickly losing this battle.

    I also find it ironic that, at a time when capitalism is quickly replacing government (read: largest gang), hordes of delusional imbeciles are talking about the “end of capitalism”. It’s a bit like watching a giant snowball doubling in size every 100 yards as it rolls down a hill towards you, and saying, “Well Bob, looks like that snowball is near the end of its life; so, we’re good.”.

    Open your eyes, or you lose.

    • Matt Wilkie

      ‘…a sneaky euphemism for “give the government all the data”…’

      I think the point isn’t give “the government” all the data, but give everyone open access to the data. By extension and pressure that would/could eventually include any organization that provides service, including government.

  • Mark

    This is really smart. Nice way to think about it.

  • GNerd

    I think it’ll have the same problem as the patent office, where the government agencies and the civil servants staffing them will be understaffed and underpaid in comparison to those overseen. You’ll also have companies gaming them by purposefully giving them so much data that they are paralyzed by it and unable to extract the information needed to impose accountability.

  • Sarah M

    So then I won’t use any of those services. Neither will many others. It’s not the government’s business where I travel or stay.

    I believe enough people will feel the same that I say that this “solution” is not one.

  • Eddie Wharton

    1) how does any new policy get passed when both parties with resources and inertia categorically oppose it?

    2) it’s quite interesting how new technology enables the changing of regulation by breaking it. Before mobile, no company could have gotten enough users by flouting laws in time to effectively lobby to change them.

    • on #1 I think it will have to do with finding the right incentives or moment. won’t happen on its own, but we may begin to find opportunities where this kind of trade is workable on both sides

      on #2 what about napster? that didn’t necessarily change copyright law, but it certainly teed up a big debate. i’m sure there are other examples. Prohibition?

      • Eddie Wharton

        1) good point. There will probably be a moment when both sides get so tired of the current system, they will become more open to new ideas.

        2) Those are two of my favorite examples of what happens when something is technologically easy to do (or becomes so) and also expected to be available by society. I had a good convo with Joel around the topic when I was in NYC. I actually think are many lessons from prohibition that make for an interesting lens to look at decentralized technologies and their adoption. If you are interested in prohibition, Ken Burns made a great documentary series on the topic.

  • FWIW, I like it

  • Pierre Noizat

    very interesting but why not outline the inevitable: companies will be forced to or will decide to operate on open data, peer-to-peer networks. In five years from now, Uber is likely top be challenged by a decentralized app run by a decentralized organization that will not even be incorporated.

    • so that’s actually a really interesting point

      in that scenario, both Uber and the regulators have an interest in NOT seeing that happen. For uber, to protect their economics, and for the regulators, to be able to retain control. So that could actually provide the perfect opportunity for a compromise

      You are seeing this exact scenario playing out right now in the bitcoin space, where you’ve got both large companies (e.g., coinbase) that work with regulators and comply with laws, and pure software (e.g., open source wallets) which are outside the reach of regulators. What’s happening there, and what would likely happen in the transportation space, would be a specialization / division on service offerings based on interaction w the regulatory scheme or not.

  • Richard Lazar

    Nick, your “Regulation 2.0” writings are fascinating, timely and thought provoking. As an appointed member of Portland’s ridesharing Task Force, I’ve had a front row seat on the interplay between the regulators and regulated. Data is a key component of the regulatory framework developed by the Task Force and can, as you suggest, enable a less onerous regulatory approach. As a public policy advocate for tech companies, I think elements of the model you suggest are essential in the long run. But . . . .

    While the benefits of a data-driven regulatory model are many, the current challenges impeding this type of approach are significant. For example, it is not just about how many cars Uber has on the streets of New York or how many rooms Airbnb rents in San Francisco (objective and quantifiable variables). It is also about the politics stemming from disruption of powerful incumbents, potential loss of tax dollars, data privacy and security concerns, economic and environmental impacts, public safety issues and many other largely political considerations (subjective and unpredictable variables). Unfortunately, data alone lacks the capacity to overcome political obstacles in the current environment.

    Bottom line, politics weighs heavily on the arc and velocity of innovator-inspired change in the regulatory arena. To be successful, innovators in regulated sectors must recognize, address and overcome political and regulatory barriers. This process will sometimes (often?) be messy. Then again, meaningful change and disruption are seldom easy, particularly given the dynamics and parties involved. That said, it’s certainly worth the effort.

    Thanks for your contributions.

  • They are not innovations, it’s just new, mostly illegal ways to do something from your home or car instead of a licensed, insured, inspected, tax paying, car/facility. All kinds of things will be coming down the pike from restaurants to day car, NONE of which is really original, it’s just a different facility for the same old service AND it’s illegal.

  • Travis

    There are two fundamental issues with the premise of this story:

    1) Uber has been tightly complying with all regulations In new york’s for hire transportation industTy. These rules were set up long before uber arrived and we likely comply with these rules better than any other player in the transportation industry in New York.

    2) Uber already provides data for every trip we do in New York, per existing regulations.

    given Uber complies with all existing regulations in New York, and already provides New York City data on each trip per those regulations, it makes a lot more sense to conclude that the proposed cap on vehicles was simple taxi protectionism.

  • I am comfortable with #1. Not sure I am comfortable with #2 unless it’s totally anonymous. Government bureaucrats aren’t angels. They are human. What if my competitor pays a govt bureaucrat for data they shouldn’t have?

    • I had a feeling you might say something like that!

      The trick to getting comfortable with number two is that it would only work ***in exchange*** for number one. And then, it would only work under the right conditions of trust and security

      Food for thought: your competitors likely use google docs, the same as you do. Do you worry about them paying off google for access? (Or Verizon, or Facebook, etc?)

    • I’m OK with #2 if the data go into the public domain. If data are for use by the city, but still treated as proprietary, I would see that as worse than the status quo.

      • Totally agree

        This is important for a number of reasons. More analysis by more third parties. More basis to stand on when making regulatory decisions based on the data.

  • Dan Phillips

    But the data that uber / airbnb collect is not holistic. It’s short term smile ratings between customers and employees. This has nothing to do with underlying and critical social and environmental impacts such as equity, accessibility, pollution, pricing etc. These web platform businesses take the most profitable components of a city’s ecosystem, extract earnings from the local environment, avoid paying for social infrastructure and potentially cripple public service alternatives. They need to be regulated hard and ultimately should not be allowed to build an online monopoly. The web was not invented to support monopolies and it needs to be regulated to ensure that it doesn’t. Investors and innovators who try to own whole markets need to expect reduced returns.

    • I think there’s a lot more data there than you give them credit for, which could give a view into the issues you mention

      Also: I am not arguing for no regulation, just different, better regulation
      Fact is the regulators are trying to make decisions now based on almost no data. My point is that they need to make collecting data a high priority