It would seem fairly straightforward for Facebook to collect information about whether an advertisement pertains to something that has anti-discriminatory regulations tied to it (housing, employment).
I generally agree that FB should not be allowed to permit users to segment these types of advertisements by the dimensions claimed. However, what piques my curiosity here is whether “indirect” discrimination in choice of advertising medium constitutes discriminatory practice.
For instance, what if some of these advertisers, instead of segmenting directly by race to target whites only, targeted fans of pages that were obviously but not specifically comprised of whites (e.g., “Irish American Heritage” or “Blonde Hair Tips”). Would there be a case to be made for discrimination? And against whom?
It's easy until you realize how global Facebook's audience is. What happens if California makes this type of advertising illegal for hotels and Sri Lanka makes it illegal for movie theaters? Now repeat for every possible advertising law.
The telecom industry is already 40 years down this path, the level of regulations, state, county and city E911 fees and operating regulations is impressive. If scrappy, insolvent players like Sprint can write code to comply with onerous, numerous and varying regulations across 7000 different LATAs, Facebook can handle writing code to comply with national laws in 200 odd countries.
Facebook (and others) can choose to show ads in any jurisdiction it wants, but if it does it has to follow the rules there. If they want to show ads in day Sri Lanka then they have to read up on and follow the laws there. In some cases it will not be worth it.
IMNAL and I dont have a source, but I believe credit agencies were once sued for this type of indirect discrimination. If I remember correctly, the argument was that credit scores unfairly treated some protected classes.
It seems to me that even though the intention of that is discriminatory in some way (and not necessarily against protected classes - the Whole Foods ads might be simply to target higher-income tenants), it's not in fact exclusionary. There's nothing preventing anyone who's not the normal target of these ads from going to Whole Foods or buying those newspapers. In FB's case the excluded audience has no chance of ever seeing those ads.
Sure. But income or wealth aren't protected classes. And anyone can walk into a Whole Foods just to see the ads, without spending any money. There is targeting happening when you advertise in Whole Foods but it's not discriminatory against a protected class.
There's actually something called "disparate impact" in Fair Housing that covers this. So if someone does something like that, especially intentionally, and it impacts a protected class, they can be sued and fined.
You don't need an insane amount of data. It's easy to figure out if any given criteria will have a disparate impact -- all criteria have a disparate impact.
Is your position that it should be illegal to use a criterion in ad targeting if the use of this criterion results in an ad being seen by one arm of a protected category more than another arm?
That's a consistent position, but it amounts to banning of ad targeting. Is that the world you want?
The law explicitly says you can’t discriminate housing ads against certain groups. Courts have held that Things that aren’t illegal on their face can be illegal if they have highly discriminatory effect in practice.
So I don’t see why any of that should be allowed here.
And it would only apply to things that have a heavy discriminatory effect. I don’t see why it would matter if you chose not to advertise your non-pet friendly apartment complex to people who have pets. Unless someone can show that’s a pretty direct proxy to a protected class it seems fine.
Again, this only applies to kinds of ads covered by this (or similar) laws. Housing and employment are the only two kinds I know of. I see no reason why you should be restricted in who you choose to advertise your new T-shirt to.
> I don’t see why it would matter if you chose not to advertise your non-pet friendly apartment complex to people who have pets. Unless someone can show that’s a pretty direct proxy to a protected class it seems fine.
As a matter of fact, pet ownership is a strong proxy for a protected class [1].
Is your position that an apartment building owner should have to advertise to pet owners and non pet owners equally even if the apartment complex doesn't allow pets?
Come to think of it, isn't having a no-pets policy itself discriminatory?
I didn’t know that. Well, if it meets whatever HUD/the courts’ threshold is for something that’s too discriminatory then maybe that’s off the list.
I know that pet friendly and non-pet friendly housing exists, so there must be some legal basis.
But if the current legal standard meant that it would be illegal to advertise to (or away from) pet owners then yes, I would expect that FB would have to remove those options for housing ads.
Guess that wasn’t the clear example I was hoping for.
Is a "no pets" policy illegal housing discrimination? Why or why not?
A "no pets" policy, on its face, would have a disparate impact on different arms of a protected group, and so, on its face, should be illegal. I know that "no pets" policies tolerated at this point.
What I want to know is how you or anyone else can justify a "no pets" policy considering the protected group issue I raised above. Is the "no pets" policy just an unprincipled exception [1]?
I see no explanation for allowing "no pets" policies other than "yeah, 'no pets' amounts to illegal discrimination, but everyone does it, so it's okay". That's not a good basis on which to organize a society. Why or why not shouldn't people make another unprincipled exception for ad targeting?
Again, I don’t know the actual reason, just that it’s clearly legal.
Many people are very allergic to pets or scared of them and won’t live in a building/complex if there are animals there. If you’re a landlord and you allow pets you also have to deal with any possible damage they may do to your property, even though you can make the renter reimburse you for that.
It may be that because of those factors courts have determined that it’s perfectly reasonable for landlords to choose not to allow pets no matter what effect that may have on the number of people of different races who apply to their properties.
I generally agree that FB should not be allowed to permit users to segment these types of advertisements by the dimensions claimed. However, what piques my curiosity here is whether “indirect” discrimination in choice of advertising medium constitutes discriminatory practice.
For instance, what if some of these advertisers, instead of segmenting directly by race to target whites only, targeted fans of pages that were obviously but not specifically comprised of whites (e.g., “Irish American Heritage” or “Blonde Hair Tips”). Would there be a case to be made for discrimination? And against whom?