Hacker News Reader

Comment by dlss | Hacker News Reader

I hope this happens soon.

The risk/reward trade-offs don't seem to be understood by most AirBnB landlords (99.99% chance of earning $100, 0.01% chance of losing $50k+). Even though this is strongly positive expected value, for a family that can't afford to lose $50k this is a game of russian roulette.

My heart goes out to the author of this; I hope the person who did this to her is put in jail, and the system (AirBnB) that enabled that person prevents landlords who can't afford the financial & emotional risk from listing properties going forward.

This is one those places where laws make sense to me; the average person needs to be protected from unknowingly gambling with their family's future.

View on HN

> This is one those places where laws make sense to me; the average person needs to be protected from unknowingly gambling with their family's future.

What kind of laws do you suggest?

At the very least requiring theft/fire/vandalism insurance for people renting out their home.

Also interesting: limiting vacation rentals to accredited investors.

(edit: and requiring places like AirBnB to verify those requirements before you can list with them)

Why don't you start a business that offers competitive insurance packages to people who want to use airBnB (and similar products) instead of demanding a law that obviously would have a lot of unwanted side effects on people who _do_ know about the risks they are taking by entering this business?

Protective laws have very high hidden costs. These costs are hard to quantify and routinely are neglected when discussing pros and cons of said laws.

Are you against requiring car insurance to drive? (another situation with a small chance of a large loss that most people can't afford)

If yes, please explain why. If not, please tell me what's different here. I'm genuinely confused why you said what you said.

Usually the only insurance you're required to have is for damage you do to others, not your own car. You're allowed screw yourself over.

But if the driver who hits my car can't afford insurance or the repairs for my car, then the other driver has screwed me over.

Car insurance is different, because it also covers damages you inflict on others and other people's property.

I am speaking out against mandatory insurance of your own property. Yes, there will be cases where people are underinsured. But mandatory insurance would make some people overinsured. There is a long argument to be made for balancing this tradeoff, but in short: Owners are in a better position to judge whether they should insure their property than lawmakers and thus it should be up to them (again: as long as their property doesn't have a high probability of damaging other people; in such cases mandatory insurance might be reasonable).

It's very easy for younger smart people to be anarcho-capitalists (which I take it you are), but IMHO becomes progressively harder as you age.

The basic problem is that a "people know what's best for themselves" policy screws over dumb people for the simple reason that they don't.

Car insurance is a very interesting thing to legislate, because although the expected value of buying insurance is slightly negative, when you factor in the economies of scale on dealing with someone else's car getting repaired (figuring out if a bill is over market, handling multiple bills, tracking any medical expenses, etc), purchasing car insurance is strongly positive EV.

I think landlord insurance is similar.

>> Car insurance is different, because it also covers damages you inflict on others and other people's property.

Landlord insurance of course also covers this (if your property catches fire while you're renting it out, you may very well be legally responsible, if a previous renter breaks in and steals the new renters belongings you are probably legally responsible, etc).

Regardless, it seems based on what you said that you might be in favor of requiring renters to buy insurance. Is this correct?

No. I wasn't even making an argument for mandatory car insurance. I was merely pointing out the difference to mandatory property insurance and said that this difference _might_ be the basis for a convincing argument in favor of the former.

Rental property insurance is not a brand new idea.

Probably not worth to be codified as a law, after all, some may choose to bypass it, but other than that there are plenty of companies selling it (vacation properties and rentals existed long before AirBnB), and AirBnB could benefit from linking to those vendors.

Of course. I didn't mean to say that she/he should actually start an insurance company. I wanted to point out that there are market-incentives that might motivate someone to offer such a service without the need for a law.

I think you are dramatically overstating the risk at 0.01%. Suppose past performance predicts future results. AirBnB has had, what, 10 million rental deals? And this has only happened, as far as we can tell, once. That's not an 0.01% risk. It's more like 0.000001%.

Overstating the risk by a factor of ten thousand is a very substantial exaggeration.

It's hard for human beings to understand very large and very small numbers, so to put that in context, http://en.wikipedia.org/wiki/Road_traffic_safety says about ten people are killed per billion vehicle kilometers in most developed countries, so perhaps five people per billion passenger kilometers. So you have a one in ten million chance of getting killed by riding about twenty kilometers in a car.

Surely you would not advocate making laws to "protect" "the average person" "from unknowingly gambling with their family's future" by traveling tens of kilometers by car, thus possibly orphaning their children?

> I think you are dramatically overstating the risk at 0.01%.

Perhaps, but the chance of losing >$100 one way or another (negligence, accidents, outright theft) is going to be a lot greater than 0.01% anyway.

Update: Airbnb's new post http://techcrunch.com/2011/07/27/on-safety-a-word-from-airbn... says "undermined what had been – for 2 million nights – a case study demonstrating that people are fundamentally good." If I take that to mean that 2 million nights had been booked via Airbnb to date, then I underestimated the (frequentist) risk by a factor of around 350: 2 million nights are only a quarter of a million weeks, so this is a one-in-a-quarter-million chance, 0.00035% rather than 0.000001%. Also, my original figure, "It's more like 0.000001%," was wrong; that would have been one in a hundred million, not one in ten million. And the original "0.01%" was an overestimate by only about a factor of 300, not ten thousand.

So the risk of this happening to you is not like the risk of dying by traveling 20 kilometers by car, but more like the risk of dying by traveling 700 kilometers by car.

(Of course that's assuming that past performance is some kind of a guide to future results. It could be that all kinds of tweakers are going to get on Airbnb now and trash things that one of their friends has had a taste, or it could be that future people thinking of doing this kind of thing will be deterred by the rapid arrest of the apparent culprit in this case.)

That split seems wrong. There seems to be no deterrent to the type of behavior the author experienced, so it would seem that everyone with an inkling will try to steal the renter's identity. I'd estimate that population to be 1-3% of people, which implies the expected value is strongly negative

Reply on news.ycombinator.com