Airbnb Unveils Changes To Address Racial Discrimination (npr.org)

An anonymous reader quotes a report from NPR: Acknowledging that his company has "been slow on this issue," Airbnb CEO and co-founder Brian Chesky is rolling out changes aimed at addressing discrimination complaints against the home rental service. Among the changes: de-emphasizing the role of user photos in arranging stays. Here are some of the other changes Airbnb announced Thursday: Providing assistance to people who feel they've experienced discrimination; Anti-bias training for all staff; Setting public diversity goals for staff; Partnering with historically black colleges and universities to strengthen their recruitment pipeline. The move comes after longstanding complaints from African-American Airbnb customers who said their booking requests were turned down at a high rate. Black Airbnb users vented their frustration with the phenomenon of being rejected for a booking date -- only to see the same place get listed once again -- spawning the hashtag #AirbnbWhileBlack on Twitter. And those frustrations were borne out in a study that sent 6,400 requests to AirBnb hosts in five large U.S. cities; the requests were identical except for the customer's name. As the Hidden Brain podcast reported, "requests with African-American sounding names were roughly 16 percent less likely to be accepted than their white-sounding counterparts."

Comments

they also found...

[turkeydance]

....that black hosts were also less likely to accept requests from guests with African American-sounding names than with white-sounding ones.

Re: they also found...

[Anonymous Coward]

Weird. Almost as though it isn't racism causing this behavior, but knowledge and experience with a culture that is a higher risk to invite in your home. ... Nah! Let's just blame a boogeyman and commit corporate suicide as a sacrifice to white guilt.

Re: they also found...

[Kjella]

They fear them because of the very real and demonstrably accurate statistical fact of black violence. Blacks kill other Blacks at astronomically high levels. Cops would be suicidal to pretend otherwise.

Roughly speaking, black murderers constitute around 0.03% of the black population. Meanwhile, white murderers constitute around 0.005% of the white population. (...) >99% of Muslims are NOT terrorists.

So say I'm a cop approaching a suspect's car, is a 0.03%/0.005% = 600% difference in risk of getting shot trivial or not? Or if I'm the security service, the ratio of Muslim to non-Muslim terrorists? Most rentals go well, but one horror story can easily wipe out the profit of a hundred ordinary rentals. It could happen with white people, it could happen with black people, it's unlikely to happen with white people and it's unlikely to happen with black people. But if the unlikely is still a lot more likely to happen with black people rather than white people, is that bigotry or risk management?

If I've been out partying and is walking home late at night and see a woman walking the same way I bet she's a lot more worried I'll jump her and rape her in the bushes than I am that she'll do the same to me. Because I'm male and she's female and most rapists are male and rape victims female. I'd say it's pretty daft to call that sexist, even though it's entirely based on our sex with no regards to the actual person and wouldn't happen if it were two men or two women. That 99%+ don't drag women into bushes to rape them doesn't mean that fear is false or misguided.

Of course a selective bias may become a self-fulfilling prophecy, if everyone looks to the more likely suspect they might also be disproportionally often investigated and caught compared to non-likely suspects leading to excessive confirmation of the bias. That is to say if the real numbers are 60/40 and the chance of getting caught is also 60/40 the actual figures will look like 0.6*0.6 = 36% vs 0.4*0.4 = 16% leading to the false conclusion that one group is more than twice as likely to be the perp rather than 50%. But in a world of limited knowledge and resources we tend to apply what we have, even though it's unfair to those who belong to a group but go against the statistics of that group.

Re:Will the renters be COMPELLED to rent?

[mi]

They can rank the discriminatory hosts last so that most potential renters won't see them.

That will mean violating the customer's trust — if I am sorting by rankings, the site will be lying to me if they let (alleged) bigotry weight somebody down beyond the low rating the alleging party has left.

Maybe, they can add a separate criteria for "political correctness" or "adherence to Social Justice principles" — and see, how many will choose their next room based on that...

Perhaps more importantly, if the discrimination really is as widespread as is being alleged, your method simply will not help...

Re:Did they collect risk and damage data?

[HBI]

That's not what the numbers would show, so we'll never see that data.

The truth is something today's authoritarians can't handle.

Re:In other words. . .

[ShanghaiBill]

If the renter can supply a good credit rating it should be safe.

Nobody runs a credit check for a one or two night AirBNB stay. Even if they did, that is not a fix to discrimination, since blacks, on average, have worse credit scores.

Because blacks tend to have worse credit scores, in 2007 the state of Washington banned the use of credit reports in hiring. This of course, made discrimination worse. This is an example of a market for lemons. Since employers could no longer tell "good" black candidates from "bad" black candidates, they played it safe by just hiring fewer blacks overall. Yet another perverse unintended consequence of a regulation.

Re:Flip Side

[ShanghaiBill]

Lets see the stats on homes being trashed, by race.

I host on Airbnb, and the biggest problem is not being "trashed", but getting unjustified bad reviews the destroy your reputation and diminish future bookings. I have rented to black people, and they tend to be easy going and tolerant of minor problems that may crop up. I have never had a bad review from any of them. The worse renters are French people. They whine about everything.

Re:Flip Side

[Anonymous Coward]

I live in France, I am not sure if it counts as whining. I think they just have a different attitude - I don't think they ever gush with praise.

Some guy I (an anonymous coward) met said their education system works as follows. All kids start the year with 20/20 full marks. But for each bit of work they do that is less than perfect they lose credit. And everyone always loses credit. In say the UK, it wouldn't be too hard to get around 90% for most things, but I think anything over 80% in France is really pretty exceptional. I took some courses with clever hard working people and the top score in the class might only be 15/20 or 16/20.

I saw some of a patisserie competition where people were making ridiculously beautiful delicious looking things that in the UK would only receive gushing praise and scores of 8.5/10 or more, but in France the praise was very much tempered with criticism and scores were closer to 6/10.

A friend of mine has a French wife and he thinks that the French in general have a very fatalistic attitude towards things, don't generally expecct things to go well etc.

In general I would give a 5 star review on airbnb and consider it being polite, unless there was a problem, similarly with amazon. But I think really a 3 star review means everything was fine right? 5 stars should be for exceptional? Or at least I think that is how the French might see it.

I don't know what the point I am trying to make is - definitely don't take it personally, it is probably a cultural difference which airbnb and any other future renters (frustratingly) won't really be able to take into account.