Now that the Uberization of everything is breaking into mainstream politics, it's time to talk about what on-demand apps such as Uber or Pager will mean for transport, health care or other aspects of modern life and work.
MIT research professor Andrew McAfee is the co-author of The Second Machine Age, an acclaimed book about how automation, artificial intelligence and digital disruption are changing society. I interviewed McAfee twice in Cambridge this week. Video of the second discussion is below, followed by our extended interview.
You've said politicization of the on-demand economy, gig economy or platform economy -- whatever we're calling the model of connecting the supply for services to demand using software and mobile devices -- was inevitable. Why?
Number one, there is a U.S. presidential election heating up. One of the big issues is absolutely going to be the economic prospects for the American middle class. No one seems to be thrilled about where that is right now. Candidates are trying to stake out what is the right recipe or playbook for improving those prospects. That very quickly gets into the impact of technology, which then very quickly gets into whether technology is a job creator or job killer, and what do we think of these new platform companies.
They're clearly creating labor, but it's different than how we used to think about work. They're growing like a weed, all around the world. They're a fact of life on the ground in many cities. As a disruptive business trend like this gets big and clearly important, it starts to activate the classic things it alway does, in the antibodies of incumbents. For instance the taxi industry, which is very large and has had time to familiarize itself with the political and regulatory landscape, is clearly taking notice of Uber.
Is the growth of Uber contributing to congestion in cities?
The best way to answer is with the best objective evidence possible. I find it hard to believe with trucks and deliveries that Uber is a big contributor, but maybe I'm wrong about that. Let's figure that out.
Will on-demand platforms exacerbate the formation of tiered transport systems for the wealthy and the poor?
Welcome to reality! When has that not been the case? Do you really want to forbid people from spending their money on options that bring them the combination of quality and comfort and convenience that can be delivered?
The idea that cabs provide a gold standard for point-to-point [transportation] is a bit of a joke to anyone who's lived in a city, with the possible exception of London, where cabs are awesome.
Should part of the premium paid for these kinds of services be taxed and applied to public transit for people who cannot afford to pay, say, surge pricing? Should we be concerned about tax avoidance?
Is a nickel of every cab ride devoted to public transit? I don't know the facts on that, but if that's the case, I would have no problem with the equivalent nickel of an Uber ride going to public transit. UberX rides are already cheaper than a taxi ride. If there's an option that's cheaper than that, that's a good thing.
[Editor's note: In New York City, there is a 50-cent-per-ride surcharge for every taxi ride that goes to the Metropolitan Transportation Authority to fund mass transit. Riders in metered yellow cabs pay an additional 30 cents to fund wheelchair-accessible taxis. The editors of the New York Daily News say Uber should contribute.]
If these kinds of systems replace cab companies, should a part of Uber's fleet be wheelchair-accessible? What do you make of stories about denied rides for handicapped patrons?
If we decide that that regulation is there for a good reason, then I have no problem with it applying to all urban transport. Does it apply to black cars? Cab companies? It's not technically difficult to add a handicap button to what Uber is doing.
There are going to be anecdotes. These things are going to happen. How hard would it be to find cab horror stories, though? The idea that Uber treats groups worse or provides inferior service to a cab? I'm not buying it. In general, with a rating system, I have a hard time believing Uber treats customers worse.
If you worried about transparency and disclosure of data around accessibility, there needs to be a level playing field. I'm guessing we don't have one. We'd have a hard time getting good data from cab companies, because my guess is they don't track it. Uber is a mighty engine of data. They did share a huge amount of data with a good economist and gave him carte blanche to publish on it. I'm not aware of cab companies that have done that.
Why should on-demand companies be treated different from industry incumbents? Is there an exploitation of regulatory loopholes going on here?
This will be tested in court cases. The question is whether this company is providing a valuable service. What is our rationale and what is our justification for regulating it? Are we doing that to make sure consumers are looked after?
Uber takes better care of customers than any cab company, let alone the industry. That is why people are flocking to Uber. That is why the cost of a taxi medallion is going down. That is why cab companies are upset. Customers are voting with their feet. I don't have a lot of sympathy for lazy incumbents dug into an industry with what they thought were high barriers to entry.
Here's a thought experiment. Let's say that we thought the wrong people were not getting rich from this, and Uber was not run by jerks. Would we still have the same problem with it? My guess is "not really."
Do they just have a PR problem?
They all do. Uber has a particularly rich share of it. What is it we are upset about? Is it that handicapped people are disadvantaged? Do we know that? I don't think we do. Is it cab companies we don't like? Monopolistic companies? I'm not bothered. That's bad service. So no, clearly not.
Sometimes the value of things like taxi medallions goes down, but it's not the job of governments to protect the value of investments. Protecting different flavors of incumbents is a well-known phenomenon, and it is not one we should get behind.
If we like innovation, new goods and services, we have to dislike this dance between incumbents and regulators. The reason we're all so interested in Uber is that it's a crucible for all these issues all at once.
Should there be flexible benefits for people primarily earning income on these platforms?
Yes, this is a big part of what we need to do. We are not going to legislate full employment in good, old-fashioned industrial-age jobs. Many people are trying, but it is not going to work. You can't get back there by any kind of fiat. The question is what the smart thing is to do if we're worried about opportunities for classic, middle-class Americans.
These new companies have come along and are offering, at scale, new ways for middle-class families to earn a living. Should we demonize or celebrate them?
I don't mean to cheer on Uber, pat them on the back, or give them pass on their obligations, but these companies are producing labor opportunities. They are the wave of the future.
A big part of it is to decouple the social safety net from your employment status. When you're employed in a classic industrial job, it's kind of when you need it the least. For more and more people, labor is more contingent, more flexible and more precarious. Let's not bemoan; let's find a way to help.
Do you see the Affordable Care Act as a good thing, in this context?
I don't think most cab companies are providing industrial-era health care to drivers. Taxi companies are opaque, but I don't think they are a paragon of employment. If you can correct me, that would be super interesting.
The percentage of Americans getting health care as part of their job has been declining for a while. It was declining well before the "gig economy" really kicked in. We chose the ACA as a way to address this decoupling. It needed to happen as a way to get people insured.
The trend toward more contingent employment is older than the last couple of years?
It's been steady for quite a long time. It's easy to look at Easy and TaskRabbit and think it's a surge, but it's just not true. I don't have exact numbers, but the percentage of people getting good old-fashioned W-2 income is stable. The decoupling needs to happen. A smaller portion of the population now has good old-fashioned jobs with pension, retirement and health care.
Instead of mandating that those must exist, rethink how the safety net has to work. I'm a huge fan of expanding the earned income tax credit and having a much larger negative income tax. For me, the basic belief I have is that if someone is willing to get out, and do a full week of work and hustle, I want to reward that effort with an acceptable standard of living. If the so-called market is not getting them there, well, we're a wealthy society. Let's top them off.
Every country redistributes wealth. It's a question of how much and how you go about it. I want to reward people who get out to work and give them a path to even better living standards.
What's next?
This is a big deal -- I read that 1 in 10 people visiting Paris last year stayed in Airbnb -- but we ain't seen nothing yet. The changes will keep coming as innovators do their work. Autonomous cars. Machine learning. Drones. These things will not run out of steam. They will get better quickly. When they do, they will bring changes to the economy, labor force and society. Many of these changes are already becoming salient and un-ignorable.
What's coming sooner than people realize?
The advances we're seeing in artificial intelligence and machine learning will infiltrate the broader economy quickly. We'll see digital customer service representatives before much longer, answering complicated questions, doing trouble-shooting and setting up appointments. A lot of people make a living today by listening to other people, figuring out what they want and giving that to them. We have always needed a person throughout history for that work.
What about self-driving vehicles?
We have them in some mines in Canada or Australia, in huge empty places. If you're contemplating autonomous vehicles, how many states have set up the rules of the game? We don't know who will win the war between driver-assisted versus driver-free, but I hope driver-free wins it.
How much can we trust autonomous vehicles on our roads if they're faced with complex ethical decisions?
We always reach for the trolley problem. What I like to think about now is are people thinking of the greatest good with a deer in the road? No, they swerve to avoid it. The idea that any decently configured piece of technology will do worse than drivers do today makes no sense.
I hear what Nick Carr is saying about automation bias and skill decay, and I've debated him. His argument is about what has been lost with progress, and yes, things get lost. I can't use a slide rule. I'm just young enough to have started doing things in the calculator era. Would it be better if I knew base 10 and logarithms in detail? Yes, but in no sense is it worth going back.
Look what's happening with aircraft automation. The data are super clear: As we've added more automation, air travel has become remarkably safer. Nick says yeah, but are we getting to the point of reversal, where pilots are getting complacent? You have to torture data to get there. We don't know yet.
As tech has advanced, you and I are unbelievably safer. That's not coincidence; it's cause and effect. Should we get rid of the pilot completely? I'm not sure. I'm enough of an optimist to think that within not too many years, a human pilot will only be introducing bias and error. Other people disagree, but that's where i am.
By extension, I do not think people should be driving cars for much longer, unless they want to. We will be killing and injuring many, many fewer people if we turn over driving.
It's one of the most important things we can do. If we value human life, let's not end it as often.
This interview has been edited for clarity and length.