These modern rideservices can expel drivers (from the dispatch-and-payment system) far more effectively than traditional regulations could suppress gypsy cabs.
Policies should be driven by real results and demonstrated harm... not by assuming that the risks/regulatory-tools of decades ago still apply.
Ban their DL. Ban their device. Ban their car tags. Include a photo of their approved drivers.
It's then at least as hard and probably harder to masquerade to Uber as someone completely different than it is to masquerade to traditional regulatory authorities.
Entering an Uber car, my own trusted device can supply me with car and driver info, including photograph. In traditional systems, I'd have to trust unauthenticated documents (including car markings) provided by the car/driver. Such forged traditional documentation can even fool local law enforcement, in some gypsy cab situations.
And even if someone can occasionally slip into the role/reputation of another driver in good standing, with that other driver's cooperation, there is still a chain-of-control/motivation incenting good behavior and allowing punishment of transgressions (through the accomplice who loans out their valuable identity).
We would need to empirically study whether this system offers better results, overall, than the traditional system. I suspect it does: even if you can contrive a few exploits, they could still be less prevalent/practical than the weaknesses of the traditional system.
Straining one's imagination is a bad way to inform policy. Observe, don't imagine.
There won't be 10 Ubers: the reputation angle and network effects strongly favor a few large recognizable operators. And maybe just one!
But even if there were 10: accident, citation, and criminal records are easy to check. Consumers will prefer the branded/authenticated services which self-police most effectively.
Traditionally, this may have been a problem needing regulation: at the moment of flagging a streetcab, the fare was at the mercy of the provider, with minimal authentication/reputation information available.
Now with the modern mobile-dispatched services, this problem is obsolete... like the problem of needing to buy a paper street map when arriving in a new city.
You have to strain your imagination to see 10 competing companies? This is another bias routinely evident on HN: we think of public policy for new business models solely in terms of the best-known operator of that model; it's not about regulating livery, it's about regulating Uber. No.
You don't have to change the subject to be about my capacity for imagination, or some bias you perceive in other HN conversations.
There aren't 10 such companies now. There might never be. Nor is there yet any evidence of consumer harm.
Imagining the worst when it might never materialize is a bad basis for policy. Especially with all the new factors which provide stronger, more-immediate checks on bad-actors: many of which apply to one operator or 100 equally well.
Why not give them a chance to work before constraining new operators to an ancient system which (a) disappointed customers with scarcity; and (b) got captured by incumbents?
This thread began with me describing the general case, "these modern rideservices".
I happen to think there will be few of these modern rideservices, rather than many, because the matching/reputation market has natural monopoly characteristics. But that's just an aside.
The tech/market/reputation checks I've been describing don't depend on Uber or any particular company. They are inherent to the category. I am making a case for the modern mobile-dispatched-and-billed rideservices category, not Uber specifically.
(My guess would also be that the average customer is more likely to someday face economic 'harm' from an eventual dispatcher monopoly, than from the sort of dangerous/abusive/unaccountable car/driver issues that you've been mentioning. But those sorts of antitrust concerns also need to be handled in retrospect, after observing how they develop, and not based on anticipatory paranoia about what might someday happen.)
There is no way that banning someone from a single website is more effective than putting them in jail or prison (depending on the state). For that to work, all of the psuedocab/ride-sharing services would have to work together to blacklist drivers who violate their rules, but this is illegal in many states, especially California, in which several of these companies are based.
A driver doing something so bad the penalty is incarceration can be incarcerated whether they do it in a traditional cab or a new carservice. Also, accident/citation records are available to insurance companies and others, even for non-commercial drivers.
We can make the penalties for any real problems that arise as strong as needed... when they arise. Expulsion from Uber/etc might be enough. If it isn't, it doesn't require traditional licensing/regulation to increase the penalties for specific actions.
Comments
These modern rideservices can expel drivers (from the dispatch-and-payment system) far more effectively than traditional regulations could suppress gypsy cabs.
Policies should be driven by real results and demonstrated harm... not by assuming that the risks/regulatory-tools of decades ago still apply.
Exactly how does Uber reliably "expel" a driver?
Ban their DL. Ban their device. Ban their car tags. Include a photo of their approved drivers.
It's then at least as hard and probably harder to masquerade to Uber as someone completely different than it is to masquerade to traditional regulatory authorities.
Entering an Uber car, my own trusted device can supply me with car and driver info, including photograph. In traditional systems, I'd have to trust unauthenticated documents (including car markings) provided by the car/driver. Such forged traditional documentation can even fool local law enforcement, in some gypsy cab situations.
And even if someone can occasionally slip into the role/reputation of another driver in good standing, with that other driver's cooperation, there is still a chain-of-control/motivation incenting good behavior and allowing punishment of transgressions (through the accomplice who loans out their valuable identity).
We would need to empirically study whether this system offers better results, overall, than the traditional system. I suspect it does: even if you can contrive a few exploits, they could still be less prevalent/practical than the weaknesses of the traditional system.
Ok, now imagine there are 10 competing Ubers.
Straining one's imagination is a bad way to inform policy. Observe, don't imagine.
There won't be 10 Ubers: the reputation angle and network effects strongly favor a few large recognizable operators. And maybe just one!
But even if there were 10: accident, citation, and criminal records are easy to check. Consumers will prefer the branded/authenticated services which self-police most effectively.
Traditionally, this may have been a problem needing regulation: at the moment of flagging a streetcab, the fare was at the mercy of the provider, with minimal authentication/reputation information available.
Now with the modern mobile-dispatched services, this problem is obsolete... like the problem of needing to buy a paper street map when arriving in a new city.
You have to strain your imagination to see 10 competing companies? This is another bias routinely evident on HN: we think of public policy for new business models solely in terms of the best-known operator of that model; it's not about regulating livery, it's about regulating Uber. No.
You don't have to change the subject to be about my capacity for imagination, or some bias you perceive in other HN conversations.
There aren't 10 such companies now. There might never be. Nor is there yet any evidence of consumer harm.
Imagining the worst when it might never materialize is a bad basis for policy. Especially with all the new factors which provide stronger, more-immediate checks on bad-actors: many of which apply to one operator or 100 equally well.
Why not give them a chance to work before constraining new operators to an ancient system which (a) disappointed customers with scarcity; and (b) got captured by incumbents?
You don't design public policy based on the interests and controls of one single private company.
This thread began with me describing the general case, "these modern rideservices".
I happen to think there will be few of these modern rideservices, rather than many, because the matching/reputation market has natural monopoly characteristics. But that's just an aside.
The tech/market/reputation checks I've been describing don't depend on Uber or any particular company. They are inherent to the category. I am making a case for the modern mobile-dispatched-and-billed rideservices category, not Uber specifically.
(My guess would also be that the average customer is more likely to someday face economic 'harm' from an eventual dispatcher monopoly, than from the sort of dangerous/abusive/unaccountable car/driver issues that you've been mentioning. But those sorts of antitrust concerns also need to be handled in retrospect, after observing how they develop, and not based on anticipatory paranoia about what might someday happen.)
Don't tell the driver where the customers are.
There is no way that banning someone from a single website is more effective than putting them in jail or prison (depending on the state). For that to work, all of the psuedocab/ride-sharing services would have to work together to blacklist drivers who violate their rules, but this is illegal in many states, especially California, in which several of these companies are based.
A driver doing something so bad the penalty is incarceration can be incarcerated whether they do it in a traditional cab or a new carservice. Also, accident/citation records are available to insurance companies and others, even for non-commercial drivers.
You missed his point. By regulating cabs, we can easily make the penalties for breaking the cab rules strong enough to deter bad actors. Uber can't.
We can make the penalties for any real problems that arise as strong as needed... when they arise. Expulsion from Uber/etc might be enough. If it isn't, it doesn't require traditional licensing/regulation to increase the penalties for specific actions.