The ideal of course is regulators are proactive in regulating, rather than waiting for the horse to bolt before deciding the door should be kept closed.
This is a good idea in the case of low probability, high impact events. For example, counterparty contagion in the financial system or oil spills.
But in the case of low impact events, it's pointless. Take a few obvious precautions (e.g., track all the drivers) but otherwise just wait for problems to arise before fixing them.
The risk of an unregulated system is very low - a few consumers get scammed out of $12. The risk of regulation is very high - many consumers may be deprived of a considerable consumer surplus (i.e., they may not get a ride at all, or may be forced to overpay for a regulated cab).
Well, ignoring that you've pointed to an example of a licensed cab driver, as that's "not the point being made here", there really isn't a point to be made.
What you've pointed to is a risk of allowing people to interact, and really has very little (if any) to do with cab regulation. Being murdered is a risk anytime you're around other people; its just so remote that usually it is appropriately ignored.
Further, if there were an epidemic of cab-murders, which of course a single anecdote doesn't show, such that it made sense to worry about the risk of homicide, then the people calling cabs would be aware of this, and there would be market demand for a cab company that does full background checks, offers life-insurance policies, sends a second person to guard against malfeasance, etc. Of course, murder is illegal, so police would be looking for the perpetrators as well.
Even if for some reason a that didn't happen (say, people enjoy being murdered during their cab ride) and we wanted to prevent it anyway, the type of regulations that'd make sense would be things like background checks before issuing a license (which would need renewing routinely), which would apply to all cab operators, both new and renewing.
Keep in mind that the regulations that people actually object to are mostly of the form "we allow X cabs total in the city".
I am more worried about public safety than consumer protection personally. I know in my city there has been a lot of crime associated with our heavily policed taxi industry, the opportunity for these threats to both passengers and drivers exists with a business like Ubercabs as well.
You don't think a cabbie using Uber knows that there is an electronic record of them picking up a passenger? If anything, Uber is safer than yellow cabs.
But lets suppose this really were an issue. There is a simple solution: charge each taxi a flat $10k/year license fee (with no cap on the number of licenses issued). Use the proceeds to hire full time undercover taxi inspectors.
That gets you about 1 inspector for every 10 cabs. Assuming each inspector rides 10x/day (i.e. 5 hours/day riding, 30 min/ride), each taxi will be inspected every single day on average.
This is a good idea in the case of low probability, high impact events. For example, counterparty contagion in the financial system or oil spills.
But in the case of low impact events, it's pointless. Take a few obvious precautions (e.g., track all the drivers) but otherwise just wait for problems to arise before fixing them.
The risk of an unregulated system is very low - a few consumers get scammed out of $12. The risk of regulation is very high - many consumers may be deprived of a considerable consumer surplus (i.e., they may not get a ride at all, or may be forced to overpay for a regulated cab).