Facebook Accused by HUD of Housing Bias Over Use of Targeted Ads | Hacker News Reader

Facebook Accused by HUD of Housing Bias Over Use of Targeted Ads

bloomberg.com

111 pointsokket142 comments

I agree that facebook's advertising practices allowed for illegal discrimination. A good case of "move fast and break things." Some journalists go farther though, and suggest that even the concept of a credit score is racist.

https://www.theguardian.com/commentisfree/2015/oct/13/your-c...

Consider the 2 possible worlds: 1. One in which disadvantaged minorities have poor credit because of their poor socioeconomic status. Once statistical controls are used to control for socioeconomic status, all ethnic groups are equivalent. 2. One in which, even after controlling for socioeconomic status, disadvantaged minorities still have worse credit than the general population.

In both of these worlds, race can be used as a weak predictor of credit quality. Many of the moral claims made by The Guardian are absurd, and overly protectionist.

>The insidious notion that our credit history speaks to our reliability as human beings is largely taken for granted.

If you, as a business owner, had to choose to hire someone with a good credit score or a bad one, all else being equal, what would you choose? Conscientiousness and intelligence both correlate with socioeconomic status, both being good traits for a worker.

As an owner, I don't know why the finances of my employee should be any of my business. Even in assessing credit worthiness, there are plenty of flaws with the current system.

National security reasons if you're a government or contractor, and likewise financial security reasons if you're a corporation.

The former is valid, the latter I disagree with strongly. Specifically...

  Too bad.

There are costs and risks in employment and employees are exposed to lots of risks that can result in health issues, sudden termination (layoff) and a myriad of other problems, the slight risk an employer accepts when hiring new employees and investing in training absolutely can be offset by insurance (employers can get targeted insurance against key personnel departures and against sudden wave departures). So... too bad. Violating this layer of privacy yields nearly no business benefit and costs sanity.

Is it typically free to invoke that insurance? Typically my driving insurance goes up if I make a claim. (Assuming someone stealing from you is not your fault)

I can't answer this with confidence, I suspect it would go up if you made a claim similar to other insurances. Though car insurance is a weird racket so business interruption insurances may not be able to get away with it.

> Though car insurance is a weird racket

Ha, I don't think I agree with you on the topic at hand, but this is something I'll get behind.

To the contrary, commercial insurance is often written to explicitly factor in the amount of claims that occur during the policy term, e.g. using retrospective rating[0]. Additionally, commercial underwriters typically have much more legal latitude in what they can consider when deciding whether—and for how much—to offer a policy contract: all states in the US have very strict rules about which sorts of claims can be considered when determining home and auto insurance rates for individuals (e.g. not-at-fault accidents and certain classes of weather damage are typically excluded)[1].

That being said, the brunt of the risk associated with a hiring decision lies with the employee in most situations (certainly in all the ones I’ve encountered personally working in retail and as a white-collar professional) in the US, and I would have to see some pretty startling numbers for an implied increase in insurance premiums to get anywhere near tipping that balance.

[0] https://en.m.wikipedia.org/wiki/Retrospectively_rated_insura... [1] NB I’m only familiar with US insurance regulation, so perhaps things work differently in other countries.

Well, I'm all for saying it should be none of your business (in the general case) after considering that there are reasons that contribute to the case that it might be, and dismissing them. But insurance is a weak argument. Ultimately that means people fit for employment pay for others' incompetence and malfeasance.

I mean, this is a thing that happens. The malfeasance is pretty indefensible, but if you start talking about the employ-ability (or lack thereof) of people judged "incompetent" then we're going to start walking down a really dark path. It'd be nice if everyone was the perfect employee but some people have short tempers and some people are a bit less efficient than others - should these people (if they enjoy and are trained in software development) be unable to find employment because they fall short of a standard?

If so, how should they support themselves? Is it right and just to relegate everyone you judge incompetent (due to training or innate ability) as ineligible to work? There are obviously extremes here that we, as a society, accept - but if we raise that bar and make it more precise with big data some really scary social effects will come out of the woodwork.

Personally, I'm in the camp that we need to shift the morality and realism of working so that everyone has the safety net of a UBI to fallback on in desperate times because I am reading the tea leaves and fewer and fewer people have more and more control of the economy.

It is right that people judged incompetent are free to work for anybody that wants to pay them or judge them differently. They might just want to get their gambling debts in order before being CFO or holding the keys to Gmail.

Edit: Also, the more you know about candidates, the less risky it is to hire them. This generally makes it easier for people to get jobs.

I've really never understood the point of using credit scores when hiring.

I completely understand it when making decisions on actually loaning money, leasing agreement, etc...

If you're hiring someone who would have access to valuable information that could be sold, who would have information before the market, or who could potentially be paid to sabotage the company, it's important to ensure the candidate has good credit so they're less likely to do those things, since they at least wouldn't feel a financial pressure to do so and would only do it out of greed.

Is there direct evidence this is true or is it just a "common sense" thing that makes sense but might very well be false?

Before I or others go looking for such, what would you consider an acceptable standard of direct evidence?

I honestly don't have an opinion on it. It just seems something that's repeated and accepted often and I'm being skeptical is all.

I will probably look into it myself too! So whatever you can add to the discussion is great :)

Do you mean evidence as to whether people with lower credit are likely to do such things against the company's interest? If so, I'm not sure. I just know it's something that is considered for govt purposes re: security clearances and that logic seems to carry over to private sector.

I feel like "compensate them appropriately" should be the actual solution to this problem.....

There are real enforceable legal salves you can take for these purposes included targeted NDAs (as opposed to the less enforceable "Don't talk about our stuff") that makes it really easy to prove wrong doings - additionally there are forms of business insurance out there to cover the short term sudden costs of corporate espionage/sabotage.

Basically, trusting in the person to be trustworthy because equifax is giving you a big thumbs up with a stupid grin on their face is a stupid idea. Just set up protections for your company, properly utilize the practice of minimum access and call it a day.

Credit scores aren't an indicator for morals. Do you really have so much faith in the current credit system that you'd be willing to turn someone from a job they're otherwise qualified. The argument could be totally flipped on it's head. It's important to hire someone with a bad credit score (using it as a measure of wealth apparently) because they will value their job more and not do anything to risk losing it.

I didn't say credit scores are an indicator for morals or wealth, or that they're a system I support. But bad debt is a possible incentive to act against your employer's interest that can be controlled against by using credit score as proxy for knowing their exact financials. I'm not saying that's fair, or good, just that that's the thought process behind it.

Me neither. I've never understood why it isn't illegal.

I'm not so sure that saying the concept of a credit score is racist is quite the same as saying credit scores perpetuate and reflect racial injustice/inequalities.

Moreover, the article you cited isn't simply making moral claims, but specific examples. In addition, there is a well documented history of racial discrimination within the financial sector (see the Fair Housing Act).

ALL else being equal? I'd flip a coin. Because if all else is equal, the credit score (that indicates past interactions) isn't predictive of future outcomes.

If we're assuming conscientiousness and intelligence aren't equal across these workers, that's a different story. In which case, it depends on particulars; I might take the one with the lower credit score because they've demonstrated willingness to take risks.

But really, the issue is that credit score is a lousy predictor of work performance. I have no reason to believe it's actually correlated.

A credit score is not racist because racism implies intent, however credit scores do disproportionately discriminate based on race and thus should not be used for hiring.

Consider the fact that to access any form of credit, and thus a credit score requires a bank account. When you consider the racial disparities between those who are un-banked or under banked it's clear that a credit score disproportionately affects certain communities.

Not only that, but credit ratings in general can be incredibly flawed. Many institutions had their bonds downgraded by the large credit rating agencies after the financial crisis simply due to the fact that balanced sheets were scrutinized more closely due to all the public outrage.

>>The insidious notion that our credit history speaks to our reliability as human beings is largely taken for granted.

That's literally what a credit score is. A record of (financial) reliability.

What an article

I am not sure I understand the way Fair Housing Act is enforced. Facebook allows ads targeting. This is how it earns money. Apparently some people decided that they want to advertise houses to white Christians population, which is illegal in the USA - shouldn't be HUD after those who configured such criteria, not Facebook?

Facebook provided only a tool. Nobody is after hardware store because a hammer bought there was used to kill someone.

Using your analogy, this is probably more akin to somebody going after a hardware store specializing in hammers to kill people. Sure, the government could spend the time trying to track down each individual patron and bring individual charges against each of them, but it's probably easier to go after the supplier and cut everyone off.

For that, the shop must be in violation of a law. I am not a lawyer, but there are plenty of cases where the crime is punishable, but goods and services involved are not, even though these are clearly designed to facilitate.

I guess you don't remember Backpage getting taken down?

Facebook creates the market for these types of targeted advertising to exist, and they should have at least a basic understanding of the laws and not permit companies to advertise to certain demographics when it would be in clear violation of the law.

That's like saying, oh we were too oblivious to understand how the platform we created could be used while we still profited significantly from our mistakes in the meantime.

Yeah and Ill argue that in todays current environment thats the “correct” business decision because thats how all the forces line up to push you.

These platforms can be overrun by a majority of discriminatory ads. I remember using Craigslist about a decade ago to rent housing in Virginia. Over half the ads easily contained some kind of gender, age, family or religious discrimination (interestingly enough I never saw an ad that explicitly listed race, I suspect that is because explicit racial discrimination is so taboo people used more implicit means of discrimination).

>interestingly enough I never saw an ad that explicitly listed race...

???

That must've just been Virginia? Where I lived in Houston, they listed unacceptable races all the time. (I suppose they didn't have to though, it was always the same two races listed.)

Sorry, but I'm curious. I'm guessing white and ???.

I'm guessing white was not one of the UNacceptable races. Being that it was in Texas I think it's fairly obvious which two races they were referring to, but it's not something I'd like to type out.

> Where I lived in Houston, they listed unacceptable races all the time.

Definitely going to need a citation for this. I grew up in Houston and never saw an add specifying "unacceptable races" -- perhaps 50+ years ago maybe. My parents were in the rental business and we spent a lot of time looking at (and placing) ads -- never was race listed on an ad that I've ever seen for as long as I can remember (since the mid 1980s.) That doesn't mean it didn't happen, but there'll need to be some proof that the practice was widespread or common if you're to come on a forum and essentially libel a city as racist.

Woah.

Calm down guy.

The original comment was saying that, in Virginia, ...CRAIGSLIST... would not openly mention the desired or unacceptable races for rentals. My only point was that, again, CRAIGSLIST ads certainly did have desired or unacceptable races listed in Houston.

The fact is it's happened in Houston Craigslist, just as it's happened in Craigslist in lots of other places in the nation. There was a class action, as well as several individual lawsuits, filed. My memory is hazy, but I believe Craigslist prevailed in all of this on the fact that they were acting only as a provider of publishing services and not a publisher themselves??? (Not really sure about Craigslist winning, but I seem to remember that they did. In any case, it's easily Google-able if you're really that curious as to whether or not all of this really happened.)

In other words, the comment is about Craigslist. Houston and Virginia are incidental to the story of what was happening, all across the nation, on Craigslist at the time.

Also, come on man, I lived in Houston. For a long time. A time period that, coincidentally, saw me doing a lot of renting. (And using Craigslist to do so, because that's what you did in those days.) Houston Craigslist was certainly not magically immune to this issue. I was there. I saw it. Further, the fact that there were lawsuits all around the nation makes me feel confident in saying that this was very likely not a problem limited to Houston.

I'll take at face value your assertion that YOUR parents didn't post discriminatory ads. That said, the point of fact that discriminatory ads were placed on Craigslist on a systemic and nationwide scale was not in question in any of these lawsuits. The only question was, "Was Craigslist responsible?" Put another way, even the government, courts, and lawyers all across the US agree that racist ads were placed in Craigslist all across the nation. All parties accepted that as fact based on the evidence. They simply disagree on who was responsible.

So I'm not really sure why you're trying to say that racist ads never appeared in Houston like they did everyplace else? My sense is that you believe what you're saying, so my head is telling me that you must not have lived in the social strata of people who had to check Craigslist for cheap apartments or new roommates.

You can't be the "???", italics, and caps guy if you're gonna be the "Woah. Calm down guy." guy.

Either you're the one reacting strongly or you're the one chiding people for doing so. Doing both is poor form, to an extent that it stops mattering what else you type.

Facebook doesn't have to allow you to target demographics. If you create an ad, mark it as type "housing", then things like race and wealth shouldn't be an option to target. It's illegal.

Using your analogy, when you buy a hammer you don't tell the cashier it's for murder.

You're targeting a demographic with every type of ad. Sure, the radio station doesn't have checkboxes of who hears your ad, you just pick the radio station you want your target to be listening too.

Same goes for billboards. Suburbs? Subway? The highway 2 towns over?

Print media, TV time...the list goes on.

Well, they're free to litigate that question if they think it will help them. That's the beauty of our legal system.

The sticking point is that those advertisements are made to the public and everyone has relatively the same equal access to them. Targeted advertisements through social media completely remove any semblance of equal access.

You are looking at it from the wrong perspective. Laws don't have to be perfect. But they have to be enforceable.

As far as I can see neither Federal fair housing rules nor any state fair housing rules would cover wealth. Federal covers race, color, religion, national origin, sex, family status, and disability. Many states add more categories to those protected classes.

13 states (California, Connecticut, Delaware, Maine, Massachusetts, Minnesota, New Jersey, North Dakota, Oklahoma, Oregon, Utah, Vermont, and Wisconsin) and DC consider "source of income" a basis for a protected class, and so you probably couldn't target ads in those states based on that, but I don't see anything that would cover wealth level rather than wealth source.

(Neat site with a nice way to look up which state protects what [1])

Since you can have a wealth requirement for renting (e.g., you can require that the person be able to afford the rent), there wouldn't be much point in prohibiting targeting based on that.

[1] http://lawatlas.org/datasets/state-fair-housing-protections-...

Facebook provided only a tool. Nobody is after hardware store because a hammer bought there was used to kill someone.

Your analogy is faulty.

People, governments, and law enforcement agencies go after gun stores that sell guns that are used to kill people.

They're actually not allowed to do that unless the sale was illegal.

It depends on how you define "allowed." Gun stores are pressured out of business and zoned into extinction all the time. There's a reason people from Chicago go to Indiana to buy guns.

Also, note the "kill people" phrase in the message you're responding to, and the context of this conversation: committing crimes.

They're specifically not allowed to do that because the gun manufacturing and sales industry's lobby realized it would essentially kill gun sales.

Yes, that would be the consequence of allowing frivolous lawsuits to be brought against companies simply because you don't like the product.

42 U.S. Code § 3604.

...it shall be unlawful—

(c) To make, print, or publish, or cause to be made, printed, or published any notice, statement, or advertisement, with respect to the sale or rental of a dwelling that indicates any preference, limitation, or discrimination based on race, color, religion, sex, handicap,familial status, or national origin, or an intention to make any such preference, limitation, or discrimination.

Yep.

>or cause to be made...

Key words here.

Sure, the government should be going after individual advertisers too, but selective prosecution or not, Facebook is clearly in violation.

>Sure, the government should be going after individual advertisers too

They probably will.

But in this instance it’s much more appropriate the end FBs illegal business practices, rather than continuing to go after ad creators in perpetuity. Turn off the water and fix the leak instead of catching the drops, similar theory to local police flipping low level drug dealers to go after the suppliers.

Exactly, and the net effect of a single action against Facebook is greater than hundreds of actions against individual discriminatory landlords.Facebook can clean up it's act and prevent (to the best of its ability) legally discriminatory ads, which would deal a huge blow to those individuals who would have committed future discrimination.

How did Facebook cause this to be made? The wording implies that it's meant to address the administrators that demanded such an ad be made. The ones who decided to make it. Who caused it to be made.

Facebook is just the hardware store in this sense. They provide tools. And you make the billboard. They did not cause this to be made.

edit: They still may be responsible for publishing it, which is illegal. I would think its a much better argument to go after them for that then the 'cause' to be made' argument.

The point is that it doesn't matter whether Facebook was involved in the creation of the ads, whether they support the ads, or anything else. They have a duty not to allow their platform to be used in this way, and until recently, they were failing to do that.

Maybe. Or maybe not. Can you find/cite the law that requires that?

edit: ah. tunnel vision. yes the other parts of the statute would cover FB as partially culpable here. particularly the 'publish' part.

It's right above in this subthread.

"To make, print, or publish, or cause to be made, printed, or published"

Facebook likely restricts and prevents other kinds of illegal ads. Why shouldn't they be responsible for restricting and preventing this kind of illegal ad?

It's unreasonable to require a private company to do the governments job of policing and enforcing laws.

Nobody would consider forcing hammer manufactures to police their customers not to hurt people with said hammers either

This is a nonsensical comparison. Facebook is the party that is serving these ads, using their own proprietary methodology for targeting. It's not like these consumers are printing fliers and distributing them on their own. Everything that is happening is happening 'within Facebook's house' so-to-speak.

It's unreasonable to not require a private company to adhere to the law. If you sell hammers, they must meet the laws about hammers, tools, and general goods. If the hammer is unfit for purpose -- say, because the heads are not firmly attached and a user might unwittingly hurt themselves -- that's the company's fault.

If the ad selection tools are unfit for purpose -- say because it allows a user to target/untarget recipients in a blatantly illegal manner -- that's the company's fault.

Fine, it's not their job to police the law. However, every time they break the law by publishing one of these ads, HUD fines them if they're caught. It's illegal to publish the ad, not just create it. They can't host illegal ads on their site.

Can we agree to this compromise?

Facebook isn't manufacturing hammers. They're offering their hammering services. If you say "sure, we'll hammer some people for the right price," you're offering illegal hitman services.

There have always been specific laws governing what publishers could publish when it came to real estate and employment.

It's really an argument of platform vs publisher. Google, Facebook, and others have been treated as platforms (it's really what the DMCA protected). But as they have been grown larger, governments want to treat them more like publishers and hold them accountable for the content that they put on their site (treating them more like publishers).

Probably some truth to that.

Also likely the reason you'll start to see a shift over time in Facebook and Google's business models. They won't just allow any old thing on their servers any more.

It seems to me that Facebook, Google, Twitter, etc all want to be treated like a publisher -- they keep making publisher like decisions on who can and who cannot use their platform in whatever way that they decide.

Sort of like CBS or Fox can and do decide what kind of ads they would or they would not air.

I think Facebook/Google/Twitter/etc are going to find out that by being a publisher one has to actually act as a publisher all the time and not just when it is convenient.

> They provide tools. And you make the billboard.

Facebook caused the advertisements to not show for protected groups. The "tools" ran on their servers, not yours, so they are responsible for the outcome.

Providing the tools to specifically do terrible things makes you liable.

This comes up rather often but the "disruptor economy" style business that try and replace existing things that are burdened by regulations and other boring things with shiny new things with Vue, rounded corners, blinking lights and a lack of those regulations are, in fact, totally responsible for their failure to adhere to those regulations.

I feel like D&D and friends are ultimately partially to blame here, and their long (and wonderful, I'm not knocking table top RPGs) history of encouraging pedantic rules lawyering. The US legal code isn't like a tabletop RPG where if nothing specifically says your Thri-Kreen can't quadra-wield bastard swords with monkey grip then go ahead - when laws fail to specifically address situations then the courts are present to try and sanely interpret the intent of the original law and apply it to the new situation.

That's sort of like how murder is illegal, instead of it being illegal to hit someone on the head with a garbage can, hit someone on the head with a tire iron, hit someone on the head with a crow bar... "Murder" isn't an action you can take without a lot of abstract assumptions.

So perhaps Facebook is like a hardware store that sells the...

  Burgler Bundle - a complete set of a ski-mask, crow bar, knife and lockpick set
  (do note "Burgler Bundles" are not intended for burglery, theft, larceny
  or other crimes, by purchasing the markers of the "Burgler Bundle"
  are hereby absolved of responsibility for any law breaking committed)

Trying to manufacture or sell that in the store would probably get you arrested (unless done in a valid parody form I guess?) similarly Facebook set up tools here to make it absolutely trivial to break some laws "Oh but we didn't realize we were breaking any laws officer" - welp, that's why newspapers and such have legal departments that review ads being placed in them, skipping that step to save money doesn't absolve you from guilt.

None of which is actually alleged. The allegation is that FB allowed ad targeting, not that they allowed ads that "indicated a preference... or an intention to make any such preference". It's classic regulatory overreach over the course of 50 years, from disallowing ads that explicitly forbid blacks, to forbidding advertising in venues that haven't paid their diversity tax.

Targeting an ad specifically to avoid certain races indicates your intention to make such a preference. You're just telling Facebook you don't want those races rather than putting it in the actual ad text.

I forgot, legally you're supposed to take the upsell on your ad package so you can target based on a love of kayaking and golden retrievers. (This is what I mean by "diversity tax".)

> shouldn't be HUD after those who configured such criteria, not Facebook?

Indeed they should. This is basically like going after the phone company because somebody else was using number prefixes as a proxy for race. Even though this is a fundamental property of any system that doesn't explicitly preclude such targeting, Facebook already took steps to limit this type of discrimination in ads for housing or credit. That was good enough for the ACLU, Fair Housing Alliance, and others who have a sincere interest in the matter. It should have been good enough for HUD too, which brings us to the real crux of the matter.

Sufficiency of the new limits on targeting was not the sticking point between Facebook and HUD. The sticking point was HUD's demand for all user data, without any safeguards. It's a fishing expedition. What would HUD do with that data? It's not a stretch to believe they'd use it to further this administration's own discriminatory agenda. HUD is not fighting against discrimination at all. They're on the other side of that battle. Rejecting their demand was IMO the right thing to do.

>Nobody is after hardware store because a hammer bought there was used to kill someone.

s/hammer/gun/ and you'll find some who are. Not agreeing or disagreeing, only raising the point that some people do seem to judge different tools by different standards based on some criteria that is not immediately visible/available to me.

Very clever, but alas, Facebook stopped being a hammer when they started censoring and manipulating content to support their political agenda. If the suit is successful, Facebook will be the official owner of all content on their platform under U.S. Law and be required to exercise due diligence in upholding the law or face the consequences. This will open the door legally to monetizing the value of Facebook's algorithm to the political campaigns it supports which is good news for conservatives in 2020 as Facebook is heavily biased toward the Democrat Party. The new attorney general, Bill Barr, served on Verizon's board of directors and said, during his confirmation hearing, that he intended to address the issue of political bias on social media platforms. Asked, would he use antitrust law? His answer, no, probably something else. Over the past few weeks Congressman Devin Nunez filed suit against Twitter for libel, and now HUD is filing against Facebook for discrimination in housing, also content manipulation. Finally US law is beginning to address the oversized impact of social media on politics, culture, finance, et.al.

We have regulations on the books that are there as a result of 100 years of painful human experience.

Wave a computer screen in someone's face and all of a sudden it's a different plane of existence and some people feel like these laws don't apply. I'm glad to see there is finally being some push back.

> some people feel like these laws don't apply

That's catchy, but doesn't address the crucial question of which people should be held accountable for using a tool in a discriminatory way. You could argue that the publisher bears some or all of that responsibility, and I'd be fine with that, but shouldn't the advertisers bear some too? Why is HUD only going after Facebook and not them? Hint: the advertisers don't have all of that juicy user data.

I disagree, if the tool/service is by its very nature both discriminatory and a 'black box', the company who made/provides it is responsible for what it does. You can't just make millions selling an ethically questionable service and shift the blame to your customers who you kept in the dark.

> if the tool/service is by its very nature both discriminatory

Is it, in this case? Ad targeting has a zillion legitimate use cases, including some that are outright humanitarian. Yes, it can also be used in a discriminatory way, but Facebook already tried to address that in the cases HUD only nominally cares about. Phone numbers and ZIP codes can also be (and often are) used as proxies for race. Should HUD go after the phone companies and the postal service too? You could probably apply the same argument to Uber, Airbnb, and every other "gig economy" service as well. What you're suggesting is tantamount to making all targeted advertising illegal. If that's the argument you want to make, make it.

I Agree certain ads can legitimately target on the basis of age, sex, race and religions.

However, as a multi billion dollar ad company FB knows these options are illegal in other instances. I’m not even sure those other companies you mention are ad companies or give advertisers the option to discriminate on the basis of age, sex, race and religion like Facebook.

Last I used FB ads, the process included a submission of my ad to FB for review. Therefore, FB knows if an ad targets based on age, sex, race, and religion and I do think the onus shifts to them to throughly review these types of potentially illegal ads before publishing the same.

Certainly FB prohibits advertisers from publishing other types of illegal ads to avoid potential liability (try publishing a cigarette or alcohol ad targeting children) so it’s hard for them to bury their head in the sand on these matters and say they don’t bear any liability for the acts of their users.

> I’m not even sure those other companies you mention are ad companies

Why limit it to ad companies now? Two posts ago you mentioned any service that is "by its very nature both discriminatory..." without such limitation.

> the onus shifts to them

All of it? One hundred percent? Even though they are now taking measures to weed out these illegal ads? Again why is HUD only going after Facebook and not any other company or the advertisers themselves?

> it’s hard for them to bury their head in the sand

They haven't. They've already taken action on this exact matter, meeting the approval of ACLU and the Fair Housing Alliance. Since when does HUD really have a higher standard than those?

> Why limit it to ad companies now? Two posts ago you mentioned any service that is "by its very nature both discriminatory..." without such limitation.

I think you are misattributing someone’s else’s quote to me, in my comment you replied to I specifically state these filters can be used in some instance lawfully, but Again FB is a multi billion dollar ad company and is aware of the unlawful potential of these tools.

Discrimination certainly isn’t limited to ad companies, but again I dont think those other companies you reference in the gig economy offer options to only accept riders/guests of a certain age, sex, race or religion. Those types of filters are limited to online ads as far as I am aware. But please shed some light is those gig platforms allow drivers/hosts to filter who they do business with and discriminate against protected classes.

> All of it? One hundred percent? Even though they are now taking measures to weed out these illegal ads? Again why is HUD only going after Facebook and not any other company or the advertisers themselves?

Certainly both the advertiser and FB could be liable for discrimination. But yes I think FB dragged themselves in by #1 giving these options to users (knowing these filters are also protected classes of people and in certain industries will violate the law); and #2 Reviewing/approving these illegal ads before publication.

I certainly never claimed the individual advertisers could not be liable, “onus shifting” here essentially means FB was aware not only the filters used for these ads but they approved the ads themselves.

Answering further, it’s obvious tax payer dollars are more efficiently spent going after FB and ending the unlawful practices on the platform entirely than an indefinite game of cat and mouse with individual advertisers.

> it’s obvious tax payer dollars are more efficiently spent going after FB

I'd love to see you try that reasoning with gun companies. I'm pro-gun-control myself, but the reaction you'd get would help to illuminate the difference between being efficient and being right. There are all sorts of things we could make illegal. There are all sorts of companies we could forbid from operating. It would be efficient as hell, but that's always the lure dangled by police-state proponents.

The Federal Government does go after illegal arms dealers, so I’m not sure your point.

Otherwise arms dealers have a federal law that protects them in gun/ammo sales from the acts of their buyers.

That said notwithstanding federal protections arms dealers can be liable for the acts of their buyers if the seller knew or should have known the buyer was making the purchase to commit a crime.

Here FB knew or should have known they were being paid to publish an unlawful ad (at least factually FB knew the filters were applied, reviewed these ads and approved them for publication in violation of the law).

I’m not sure why you think preventing FB from knowingly publishing illegal ads is on par with anything else you are suggesting (ie “going after gun sellers”). Gun sales and ads are both legal...until they are not.

Why do you think it’s right to go after the ad creator only and allow FB to continue to profit on and facilitate the illegal acts themselves?

> Why do you think it’s right to go after the ad creator only

Never said or suggested any such thing. In fact I've explicitly said that assigning some responsibility to the publisher seems reasonable. What I find questionable is the singular targeting of one publisher but neither the advertisers nor other publishers. That's not how one fights for equality or justice, so there must be some other goal.

The Gov can and maybe even will go after the individual advertisers.

>That's not how one fights for equality or justice, so there must be some other goal.

It happens all the time, for example, low level drug dealers or even arms dealers are given immunity just for cooperation and just the chance to go after the suppliers. It’s easy to say well, there is some other goal, but generally the goal is to combat the illegal activity.

Another example, which I have been using for years is Uber. Many Uber drivers have been arrested and/or civilly fined for illegally operating rides for hire without a permit. Yet despite all the arrests no law enforcement ever went after the company but only the drivers. I think this is one of the biggest injustices of our time, you have a multi billion dollar tech company hiring “contractors” to knowing break the law (in those jurisdictions) and then only punishing the individuals who are basically being exploited (i say this because Uber knew the acitivity was illegal in those jurisdictions, but they kept hiring drivers and scheduling rides in violation of the law, whereas the drivers didn’t necessarily know they were being hired to provide illegal rides). In some cases Uber even paid bonuses to drivers to leave jurisdictions where operations were legal to give rides in jurisdictions where it was illegal.

Is that really equality and justice to you? If law enforcement went after uber instead of the driver is that really indiciative by itself of another goal rather than stopping the illegal activity?

> If law enforcement went after uber

If they only went after Uber, giving Lyft and traditional cab companies and the drivers for all of them a pass, then yeah, I'd suspect there was something besides fair enforcement of the law going on. That would go double if they were stretching the definition of the law to include Uber's actions when its drafters had no such intent, or if they were simultaneously going after Uber for doing and not doing the same thing (parallel to the flak Facebook gets both for censoring and not censoring content).

> Phone numbers and ZIP codes can also be (and often are) used as proxies for race. Should HUD go after the phone companies and the postal service too?

I love the arguments that amount to: "if we can't stop <bad thing> everywhere, we shouldn't bother stopping it here either". I'm not sure how one expects any kind of social progress to happen with that attitude.

> Is it, in this case? Ad targeting has a zillion legitimate use cases, including some that are outright humanitarian.

Let's be real. The kind of ad targeting that happens on the internet is something quite different than demographic targeting by sending flyers to certain zip codes or putting ads in certain magazines. Because the process is largely a black box and the ads are only visible to those being targeted it's nearly impossible to audit by watchdog groups or go after the parties who placed the ads. In the realm of commerce this leads to a lot of borderline scams targeting the ignorant. In politics, this leads to radicalization and weird conspiracy theories like QAnon.

At best, modern targeted ads are a way to convince people they need something they do not. Something that "classic" advertising did as well. At worst, it's a channel for misinformation, scams, and sometimes malware. Completely invisible to you and me because we're not the "target demographic".

“ZIP codes can also be (and often are) used as proxies for race. ”

It is called ‘redlining’.

https://en.m.wikipedia.org/wiki/Redlining

FB is a great big platform. Of course they are going to be targeted first. Especially since they have the ability to stop the problem, just as the classified editor at a newspaper can decide, "No, I won't run this ad that says "No Blacks". As that would be a bad idea for both myself and the newspaper given the laws on the books that everyone knows about."

Given a finite amount of law enforcement manpower, it makes sense to target the pimp first. Especially if they are national in scale and flaunting it.

Why do people have this notion that Facebook is some unbiased public space, like an undisturbed forest or a beach where passersby can pedal their wares or something? Facebook is like a dystopian shopping mall combined with Clockwerk Orange reeducation techniques.

Due to it's algorithmic nature, Facebook ultimately decides every single post, ad, like, and person you see, and in what order. Creating a housing ad and targeting white people is not like putting an ad on the smooth jazz station, it is like having a storefront in the rich-whites-only section of the Facebook mall. In fact it's even worse because other people don't even know the rich-whites-only section exists, since they can never see it, even if they are your friend! I actually have a hard time imagining how an online system could reinforce existing class and race structures any more than Facebook currently does, short of disallowing users to become friends if it detects that this would be financially disadvantageous to allow.

copy/past from another comment:

42 U.S. Code § 3604. ...it shall be unlawful—

The law is clear. The only difference is that the publisher of the advertisement is Facebook instead of the New York Times web or newspaper.

> The law is clear.

As usual with law, not so much. Can you say with any certainty that "indicates any preference" doesn't refer to the content of the ad? Allowing an ad to be targeted in a discriminatory way might be wrong morally, but it's not at all clear that the law as written and (so far) interpreted by courts supports your conclusion.

I can totally see how this wasn't anything malicious and just done out of pure ignorance of the law. That said, I would expect this out of a small startup; not a massive corporation. This is actually where I can see the argument for increases in diversity as I can see how a homogeneous group of people being completely ignorant of the ramifications of this.

> I can totally see how this wasn't anything malicious and just done out of pure ignorance of the law.

I agree that's possible, but that seems to be Facebook's goto excuse, and you're right, that's not something a megacorp should be allowed to reach to.

A large corporation should be able to afford lawyers that specialize in the legality of advertisements...

This is the weird thing. They don't have a legal team? Or are they just not listening to their lawyers? That's a bad idea for anybody.

The historical precedent for Facebook ad targeting is allocating one's print ad budget according to the readership of a given periodical, or one's TV ad budget according to Nielsen's analysis of a show's audience.

Or more personally, advertising the room for rent in your house at your white-collar job but not at the welfare office.

Does the Fair Housing Act address this? Doesn't look like it.

How comparable is the historical precedent? Are facebook audiences comparable to different periodicals, or does the level of microtargetting available now make it more like putting your ad in copies of Time delivered to one part of town but not the addresses outside the redline?

“We’re surprised by HUD’s decision, as we’ve been working with them to address their concerns and have taken significant steps to prevent” ad discrimination, the company said in a statement. It added that it had been negotiating with the housing agency over the issue, but that the talks had broken down because federal officials were seeking access to too much user information “without adequate safeguards.”

Interesting. I also wonder if they are going after the landlords who knowingly discriminated using Facebook as a platform.

Could definitely see this. They could also be trying to prevent setting a data-sharing precedent that other Federal agencies will look to utilize.

Yes - I just submitted a related article (https://news.ycombinator.com/item?id=19516173) that has additional details on the data access request and reports that the investigation has spread to Twitter and Google.

Weird that they go after the ad platform instead of the advertisers. Pretty sure magazines also publish stats about their members enabling advertisers to specifically target a certain demographic. Actually the NYTimes does this as well: https://nytmediakit.com/index.php?p=sunday-magazine

Nothing prevents racial minorities purchasing a magazine, unlike FB ads.

They were accused before.

I thought this was old news. Weren't they being accused of showing certain results to a certain demographic because the advertisers got to call the shots on who saw the ads?

This is them getting charged. I think there is a distinction. Now lawyers are involved.

That's fine for some things. E.g. showing an ad for specific hair care products to black people.

Not OK for housing/rental ads, credit/mortgage, etc. Facebook should have known better from the start and never enabled that kind of targeting for those types of ads.

The thing its not a new law. Its been around for 50 something years. I remember they stressed this in American history and it was on the AP test too.

Still gets violated all the time too. When my parents bought a house outside New Orleans the realtor flat out told them which neighborhoods would be acceptable for them to look in.

Oh I really agree. Play stupid games, win stupid prizes.

I was just saying the headline sounds like old news when it’s not.

Why is it okay sometimes but not all the time? If there are predatory ads for mortgages, there will be predatory ads for snacks and video games as well.

The goal of most advertising is to change your consumer behaviors regardless of what best for the consumer. In my opinion it's a huge net negative on society.

This has nothing to do with predatory ads. It’s because these ad policies can be used to exclude certain races from certain housing which is illegal.

Facebook should have known better from the start and never enabled that kind of targeting for those types of ads.

If only Facebook has some money it wasn't using, it could hire a bunch of lawyers to look at its products and processes to make sure they are legal and ethical and lawful. Some kind of a team of lawyers. They could even made it a department of some sort.

"Move fast and break laws."

Big tech is going to have to learn how to balance automation with social history. Race based housing discrimination has a long history in America, it's hard to believe that no one in a position of power to question these methods brought up how these ads could be used to discriminate.

The biggest benefit that diversity will have on big tech is alternative views that prevent these embarrassments from occuring. Seems every few months another company comes to the realization that it's models simply don't have the social context to be effective.

There's sometimes a lot of chatter here on HN about a liberal arts education being worthless.

I have to wonder if more tech employees had taken classes in ethics, philosophy, and other "soft" subjects, if people in the industry would have a better capacity to make decisions that get their companies in trouble less often. Or at least be able to explain to their superiors why a particular business decision is bad.

>> There's sometimes a lot of chatter here on HN about a liberal arts education being worthless.

Plenty of people here will argue that you shouldn't take out $80k in student loans to get a liberal arts degree. Worthelss in a ROI sense

But there are a large amount of people here with pretty liberal views. So they certain don't disagree with the viewpoint.

liberal arts != liberal viewpoints

Full disclosure: I have a liberal arts degree.

When I was looking at colleges, I had absolutely no idea what I wanted to do with the rest of my life. In high school, I really hated science and math classes. Years later I realized that I actually didn't but the single math teacher I had for four years of high school (it was a small school, I just had bad luck with what teachers were assigned to the classes I took). If I had gone to college with the goal of a career that I thought I would have been suited for, I probably would be doing something much different.

I picked a school that offered me the opportunity to study a broad range of topics and had a curriculum based in seminars & discussion. Rather than learning technical subject or a trade, I learned how to accomplish tasks independently. What I feel I got were the skills to teach myself anything. Sure I'd need to learn harder career based skills, but I knew how to best teach myself something IMO, that's no different than a career paths that requires additional schooling to get a job. Looking back, I think it's ridiculous that we ask 16-18 year olds to make decisions about their careers.

>Looking back, I think it's ridiculous that we ask 16-18 year olds to make decisions about their careers.

And just how far do you think extended adolescence should last? People have to start transitioning into adulthood and preparing for career and life at some point, even if they aren't fully mature and ready for it.

It's interesting to me that there's no single legal definition of adulthood. Voting, drinking, signing contracts, joining military, count as a dependent on insurance, buy a gun, rent a car... We don't seem to be able to decide, as a culture, when a person is responsible for themselves.

You’re drawing a false dichotomy between career training and liberal arts.

I studied pure math which is as nerdy STEM as it gets, and absolutely zero of it was job training.

Yeah, you're right. In my head I was thinking of pre-law, pre-med, dental & pharm programs as the opposite of liberal arts.

When I was looking at colleges, I had absolutely no idea what I wanted to do with the rest of my life.

You're not supposed to. That's what a liberal arts degree is about: Exposing you to a lot of possibilities so you can choose.

I went to college for Computer Science, but half way through changed my major because I found other fields more interesting. A friend of mine went in for psychology and came out journalism. It happens all the time. I know very few people who came out of college with the degree they went in for. That's how it's supposed to work.

> That's what a liberal arts degree is about: Exposing you to a lot of possibilities so you can choose.

Why can't this be what high school is about, like it is in 99% of developed societies?

High school is about giving you basic education and social skills.

like it is in 99% of developed societies

This is a fatuous argument. Neither you nor I are education experts, and throwing out a fake statistic just makes you look foolish. You have not done a study of 100% of developed societies to draw your conclusion. You're just making assumptions to justify your pre-conceived notions.

I seriously doubt you can find examples of countries outside the US where studying a non-specific subject with a lot of broad distribution requirements is normal. I know for a fact that it is not common in the UK, France, Germany, Switzerland, Serbia, or Russia. I will concede that I'm not too sure what the situation is like outside of Europe.

Hate to break it to you, buddy, but the liberal arts kids grew up to work in online advertising too. I know, I worked on a targeted ads team with loads of them.

Facebook ads in the residential real estate space absolutely allow violation of the 1968 Fair Housing Act on the part of real estate agents, sellers, marketers and brokers. The mechanism of FB ad targeting inherently grants permission to exclude groups from ad targeting by way of discriminating on the basis of demographic detail. The FHA was passed to redress the widespread and normalized practices of racial and socioeconomic exclusion from home ownership, and as such FB's practice is a major step backwards in fairness and transparency for persons participating in housing markets.

Does anybody know if FB's practice has been attacked in court sooner than this?

They're probably going to settle. This is not new and there is already precedent[1] about these sort of postings. Even a halfwit counsel could have told them this wasn't allowed.

I've been seeing this go on for a while and I'm not surprised at all. I'm more surprised it took this long. As soon as they added rentals and properties to their marketplace they should have immediately tweaked their algorithms to not fall outside the law

[1] https://www.npr.org/templates/story/story.php?storyId=554032...

Hot on the heels of their announcement yesterday. The irony is sweet

That's likely WHY they announced they were getting rid of extremists. They would have had knowledge of these charges today, as well as knowledge of other things the government is looking into vis-a-vis their business. So I'd imagine their lawyers are preempting a lot of things because they seem to be entering an age where there is potential to be charged even if you are secondarily associated with criminal activity. (Potential not just for Facebook, but for everyone else too.)

It's a crazy time right now, because you never know what some person you provided a service for might do.

"...you never know what some person you provided a service for might do."

Well, you are in the business of knowing everything inferrable about all people, everywhere.

> knowing everything inferrable about all people, everywhere.

Which makes me wonder if the FAANGs and others will go this route - basically robo-ban people from their services who meet certain criteria (or maybe no criteria at all - maybe some kind of neural network or something decides who should be allowed to use the service and who shouldn't).

Make the ban based on various obvious and non-obvious metrics. Like credit score, comments or whatnot made using the service, how they treated other people or products in their comments - as well as things like how they typed, or moved their mouse around, or other info from their browser (or even side-channel info if that can be done from a browser - which it probably can).

So once banned - even if they tried to go back with a completely different computer, home address, IP, etc - they'd still eventually get flagged as being banned in the past.

Then have a service that all of them use to make a centralized "blacklist".

It wouldn't surprise me if they don't already have this in testing in some manner.

It's all PR. If a company knows some bad press is about to resurface, they'll try to dominate the news with their own good press. People think it's them being virtuous, but it's all a game.

Why are they not just subpoening Facebook to go after the people that actually did the ad targeting? It is probably happening but it would make more sense, and Facebook would understand the stance of the enforcement agency. Since HUD or a private citizen already is doing this, Facebook probably offered non-answers or something cocky and then got slapped with this.

They might be. It's got to be easier from an enforcement point of view to go after publishers than all the individuals who place ads. They have limited resources, just like the rest of us.

What we need is a law to make all advertisements and their targeting criteria publicly listed. It will make finding out about such problems much easier.

This is kind of ironic in many aspects but the most is one of members under HUD worked for Cambridge Analytica.

It's important to keep in mind that FaceBook is an enterprise founded by an individual who appears to have many of the traits of sociopathy, and that an overwhelming amount of shady, anti-privacy, socially-destructive, and downright exploitative behavior has been credibly attributed to them over the years. The second sentence of TFA claims that the HUD's opinion of FaceBook is that the company "...enables and encourages discrimination based on things like race and religion, as well as sex, by restricting who can see housing-related ads on its platforms and across the Internet." This should come as no surprise from a company with the kind of track record that FaceBook has; the company's founding purpose was, after all, to judge people based on superficial traits.

Rot of this kind, when located in the head of an organization or creature, does not go away easily. It has been obvious for some time now that the culture in the leadership echelons of FaceBook is deeply amoral and disinterested in societal health. Although it's a very touchy subject (especially around tech sites), I think there's a strong justification to start really questioning whether or not a person who is interested in the health of society can continue working for FaceBook.

I find it interesting that you are getting downvotes, but no replies. Come on people, if you disagree, why not say why you disagree? Just downvoting isn't helping the discussion.

I agree. Facebook is no doubt a great place to learn about building software at massive scale while earning a ton of money. But at some point you have to step back and look at the impact of what you are doing.

I remember reading an Obama administration report about this in 2016, but nothing ever came out of it. Nice to see Dr. Carson take up this issue.

Reply on news.ycombinator.com