The Role of Government & Regulatory Systems (Tim O’Reilly Interview 3/3)
This is part three of three of Tim O’Reilly’s interview with Peter Leyden as part of Reinvent’s Future of Sharing series. A lightly edited transcript of the above segment follows. The full interview is available here.
If you’re interested in learning more about the future of work and the sharing economy, check out O’Reilly Media’s Next:Economy summit in San Francisco, October 10–11.
Okay, so focus a little bit on the tech and business crew. What would you say about government, folks in government, particularly people in city government, which is a lot of what we’re talking about. What’s your analysis essentially of the receptivity to change, and what is the kind of sense of what advice you might give in terms of, particularly around the sharing economy, but in general this whole transition?
Well, the first thing I would say is, when we’re thinking about regulation and public policy, the first thing to think about should be what are the legitimate social goals of any regulation. What are you trying to do? Let’s be really transparent about that. For example, I recently landed at Newark Airport. I took a Lyft into New York City and I asked the driver, “When you do this, do you pick up in New York?” He said, “Oh no, I can’t do that. I have to go back to New Jersey.” Now you go, “What?” They’ve effectively recreated some of the same territoriality that they had for taxi companies. What’s the public policy goal there? It’s protectionism. That should be super clear. Let’s lay out there and say, “Okay. We as a city have decided that we want to protect our industry. Does everybody agree that this is the best public policy outcome? What are some of the other alternatives?” So, for example, you would look at, reducing congestion.
We want to protect consumers. We want to get the best prices for consumers. We want to improve the income of drivers. If you think about that particular example, none of those other objectives, which are far more legitimate, are being met. So, that’s the first thing. Be clear about what your objectives are.
There’s actually a really wonderful new tool I’ve been investigating called pol.is. Basically it’s a public policy debate tool put together by Colin Megill, Chris Small and Mike Bjorkegren. Actually in Taiwan, they ran a process where everybody from the cities, through Uber, all kind of participated in a public debate about what the legitimate goals of public policy were and they came to a set of solutions that I think everybody was quite happy with. I think there are some new tools for having open, informed, debates about public policy. I think on the one hand you have the cities who have to be really clear about what are they trying to accomplish, and also we need to realize that one size doesn’t fit all. There may be some cities, for example, where they say, “Okay, we could use a lot more housing, you know, in certain neighborhoods, for people coming in because it will improve our tourist industry.” It’s a great example for Airbnb.
When I think about the benefits of Airbnb, I don’t just want to stay in a hotel, I may want to stay in a different part of the city. I go up to the Yuba River and I can stay in an Airbnb in Nevada City. There’s a cheap motel, a couple of old school B&Bs, but Airbnb gives these wide range of wonderful new opportunities. So, in a city like that you go, yeah, we want some more Airbnb because it will bring more people in.
Clear public policy objective. You want to support this. There are other places where you go, well, we don’t have enough housing for our residents. People being gentrified out and people buying second homes and turning them over for Airbnb rentals is actually contrary to that public policy objective. So, if you’re clear about what the objectives are, then I think you’re going to have a much more informed discussion. Then the second piece is how do we use data on both sides more effectively?
Tell me about that.
Some years back I wrote a piece called “Open Data and Algorithmic Regulation.” In it I made the point that regulation has become a bugaboo. Everybody says, “ugh, regulation.” But that’s because regulation is interpreted as a set of rules. But in fact, we live with all kinds of regulatory systems that we love. Just try driving a car without the regulatory system of your carburetor or your fuel injection system. Think about the regulatory system that we call the “autopilot” that flies us around. I was on a flight recently where I was talking with the pilot, and he’s like yeah, in a busy airspace like San Francisco, we can’t actually fly manually in and out of the airport. You’ve got to do it on autopilot. You know what I mean? Crazy! You think, oh well, don’t you do at least the takeoffs and landings on autopilot? No, actually, that’s sometimes the place where they most use the autopilot now. But the point is, here’s a regulatory system. It works, we accepted it because it works. Credit card fraud detection. Right? We’re being monitored and we accept it because it’s protecting us.
Anti-spam, a regulatory system. Google Search Quality, a regulatory system. So what do all of these regulatory systems have in common? They actually have a clear objective, kind of what I was talking about earlier, that everybody’s agreed on. And it can be imposed unilaterally. Google says “Here’s our regulatory system. We want to have the best search results. We want people to click on the top result and go away happy.” Right?
Yeah.
At least that’s the theory of it. Uber has a regulatory system. They say they want people to be picked up in three minutes. Right? The question is, are there other factors that we could take into account? My sense is that if we could get, first of all, the clear statement of here are the outcomes we’re going for, then you start measuring are we achieving those outcomes. I love Google Search Quality as a model, because it’s something they’ve written a lot about, they’ve talked about. Things like short clicks and long clicks. This idea that if somebody clicks and goes away, and doesn’t come back, it’s a long click. If they click on the result and they come right back, then click on the next one. That’s a short click. A short click means that they weren’t happy with the results. So it gets demoted.
So they built this wonderful feedback loop and that kind of feedback loop dynamic, which we’ve really mastered in Silicon Valley… It’s what drives Google, what drives Facebook, what drives any of these modern apps. We have the data that’s feeding back to making the service get better and better as people use it. It is so largely absent from the way that government does regulation. What I would love to see is, I’d love to see a partnership between cities and the services they want to regulate, where there is real shared data, shared visibility, clear objectives.
Oh, and by the way, credit card fraud detection is another really good model of a regulatory system in which government actually doesn’t micromanage. It simply said, “Hey, guess what guys, you only get to charge the consumers for 50 bucks of bad debt, and all of a sudden, it becomes the credit card company’s duty to regulate. So, the question I have, and I don’t know what the right answers are, but if you were thinking about how would I regulate Airbnb, how would I regulate Uber and Lyft, how would I regulate any of these companies, I would be asking myself what are those interventions that will squarely put the shoe on their foot? If cities wanted to come down on the side of labor and say, “Okay, we’re not going to tell you how much you can charge, but we’re going to tell you that you can’t go below this amount.”
This came to me one day, reading an article about Uber’s rates in Detroit. Travis does a lot of hand-waving about the perpetual ride and how great it is. At one point, they were offering drivers $0.30 a mile in Detroit. Just a simple thought experiment to me said, “Okay, well, let’s assume that with the perpetual ride, I get $0.30 a mile, maybe I’m going 30 miles an hour, that’s $9.00 an hour before expenses. The city could say, your rates have got to add up for the perpetual ride at the very least or whatever; for some percentage utilization, to this amount. You can’t go lower than that.
We have ideas like minimum wage, and we go, well we don’t know how to apply these in a world where it’s on demand and people work for multiple people. Sure you can. You say we want to know what your typical utilization rate is. You could do some math and come up with some rules. Then you go, “Okay, we’re going to periodically audit those rules and if you’re not paying enough, you’re going to have to pay back the drivers.” There area bunch of ways to intervene much more creatively than, “Oh you can’t pick up in New York City if you’re from New Jersey.” That’s just stupid.
Yeah. Really what we’re talking about is the difference between pushing these square pegs into these round, old holes, from the old system; or getting creative and really thinking about how to innovate.
Yeah. What is it that we’re trying to accomplish? Also, really trying to get to a meeting of minds. I do think that there are real challenges. One of the things I think really muddies the water is the lack of financial literacy on the part of so many people, where they confuse market value with profit and loss. How many times have you seen Uber’s worth $60 billion dollars? They’re so rich, you don’t know what their actual profitability is. They’re losing billions of dollars a month in some markets. So, the question is, what’s the actual profit and loss there? How would we kind of back off a little bit of the notion of Uber’s screwing its drivers because they’re worth so much.
That’s a totally different system of money. It has nothing to do with whether they’re succeeding as a business. At some point, if the stock market stops giving them money, they could be out of business. We don’t know that though. It’s harder right now in a world where so many of these companies are private so late. That’s why again, you go back to Walmart, where it becomes very clear that Walmart is shoving cost off onto society and capturing it for its shareholders. Would somebody say, “I wouldn’t want to own Walmart if they were only making 15 billion dollars a year instead of 20?” That’s silly. If you think, what it would do, it would depress the stock price and who does that actually affect? All the shareholders will go, well how much of that is pension funds. Well it turns out, a lot of the pension funds… They’re making all their money churning stuff anyway. It’s like if they were in Vanguard or some other Index Fund, they’d be doing a lot better for people.
As we’re winding down here, you spent a lot of time with people in government. You ran these whole government 2.0 conferences.
When you look at the next couple years here around city government, state government, even federal government, do you see some kind of potential big shifts happening in terms of regulation or categorization of workers? Are we in a window here, a kind of fluid window that some stuff could happen quickly, or are we in the early days of experimenting and it could be five years more?
Boy, you know… “Government” and “quickly” are terms that don’t usually go together. Which is a shame, because government can play such an enormous role; and I do think that while it’s very easy and popular to rag on government, it does an enormous amount of good in our society. Even the most staunch libertarian who thinks that government should be out of there and everything, they’re quite happy to use the roads, they’re quite happy to have regulated airways. We have an education system. You can go to a country without government in it and it doesn’t look very good. I would just say government is us.
It’s a mechanism for collective action. So the question is, how do we make it more responsive? How do we make it more effective? That’s a lot of what we work on at Code for America. It’s like okay, you have policies, do they work? How do you make them work? What we all saw on display was the failure of healthcare.gov; it was that you can have the best policy in the world, but if you can’t implement it, it doesn’t do you any good. The fact that the government couldn’t actually build a website, even though they threw a billion dollars at it, is criminal. I think what we really have to do is we have to get better at government. We don’t need less government.
We just need more effective government. Taxes as a share of GDP is pretty constant over decades. This notion that government has been growing, growing, growing, is wrong. What is the case is that government collects way more money than it should, and doesn’t do as much with it as it could. One of the things the private sector is good at is being efficient with things, because there’s creative destruction, whereas government doesn’t really have that. So there needs to be a discipline of government getting better at what it does. Technology has taught us some amazing lessons about how to build services at scale that get better as people use them. I think we need to take some of those lessons from technology, about how you instrument services, how you measure, how you constantly improve; and take those into government. That’s again what we’re doing at Code for America.
Okay, so I see what you’re saying. I totally get the long-term on that, but some people we’ve been talking to in this series, they’ve been talking about a couple-year window. A lot of stuff could go wrong in the next few years that would set this back. On the other hand, if we did it right, it could actually propel positive things forward. Is that your — ?
Absolutely. I think it ranges from movements like portable benefits, which are a different approach to the safety net. I was just talking with a guy from the UK. He was also talking about how we build better marketplaces for connecting people to work that’s effectively occasional work. Apparently, they’ve built a system in the UK. The U.S. has this great workforce training system they spend a huge amount of money on; but it’s really only for “jobs.” Again, that sort of pivot we need is from jobs to work. What work needs doing. It’s one of those crucial next economy pivots that I see. How do we stop talking about jobs, start talking about work? How do we stop talking about getting rid of people, and start talking about how to augment people?
How do we start talking about the legitimate aims of public policy, rather than just sort of having people do these things, which are really not in the interest of society as a whole? How do we have a robust discussion about what those goals ought to be? I think there’s a certain role for that kind of antagonism. We have a society that values the butting of heads to get to truth. Whether it’s through the legal system, or whether it’s through just public policy lobbying and so on. But we also could do a lot better at creative engagement. I think there are some really wonderful opportunities.
Think about Airbnb. What kind of city do we want to build, and what makes it a great place to live? What can we learn? There should be massive amounts of data sharing. Nobody should be writing rules yet. They should be writing the smallest possible rule set while they are figuring out what’s going on, and be focusing on what’s really happening and how does it relate to what we wish was happening. Think about drones and all that’s possible there. You think about self-driving cars. You think about just employment and the nature of employment. All these things are going through these massive technological transformations, and if we have regulations that are only backward looking, that try to make things look like they used to look, that don’t actually say let’s take a fresh look at what we’re trying to accomplish as a society, we’re really missing a huge opportunity.
You said you’re an optimist in the beginning there. How optimistic are you that this will play out with a backdrop of the conventions, and polarization in the country, and all the different things we know here? Are you optimistic that this is going to move ahead?
There was a wonderful piece in The Atlantic recently called “In Praise of Short-Term Thinking.” You can believe in the long-term, but you still have to worry about disruption in the short-term, and that we have to proactive at looking for that disruption. So I guess I would say I’m incredibly optimistic about the long-term.
Going back to what I said earlier, those people in the Luddite Rebellion couldn’t have imagined what the descendants of those machines would do for our society, in making us all so much richer than even the richest aristocrats of their day. We should be thinking the same way about our future. Oh my God, these new technologies! What will they make possible? What could we do if we took all this technology and said, “Wow, let’s go reinvent everything.” I keep going back to Uber. They had this brilliant insight on how they could use technology to do this thing that was previously impossible.
They built this great company, and then they start talking about, well now we’re just going to get rid of the drivers. Why aren’t they talking about their flu shot experiment? In 2014, they gave flu shots on-demand. These self-driving cars, think about all the other things that could be on demand. Yeah, sure the drivers won’t be there anymore, but there’ll be all kinds of other new kinds of services that could be on-demand. What an amazing future! Let’s get all freaking optimistic about all these things that are possible. Zip line in Rwanda. There’s no roads, there’s no hospitals, but guess what? We can use drones to deliver blood on-demand to women who are dying. Postpartum hemorrhage is one of the leading causes of death; but we can get blood there, anywhere, in the country. Drop it with a parachute from a drone.
Let’s rethink the world using today’s technology. That fundamental optimism of how much possibility there is. What will AIs help us to do? What kinds of creative worlds will we see in virtual reality? Just like we’re in the age of Hollywood. What’s that going to look like when we’ll be able to make this world spanning visions with virtual reality?
Let’s stop optimizing for the short term, let’s start optimizing for the long term. Let’s think about how to make the society that we want. Let’s harness technology to do the impossible. I guess I would say one of the things that history teaches us is that government plays an enormous role in that process. Our current distaste for, and distrust of government is part of the problem. We have to get rid of that. We have to actually understand how to make government work again, because when government works, it really becomes a governor for society.
It helps moderate the excesses. It helps deal with market failure both on the upside and on the downside. We’re going to need everything we’ve got. We’re going to need government; we’re going to need technology. We’re going to need everybody on this planet to pull together to solve some of the challenges we’ve got in the years ahead.
Perfect place to end — on an optimistic note, and I loved your theme of reinvention, which is something that obviously at Reinvent we’re totally into. So, a fantastic interview. You really laid it out, from the big picture to the granular. I didn’t expect anything less from Tim O’Reilly and I really appreciate you coming here.
All right. Thank you. All right. This was fun.
