In this episode, Chris Willis reviews the settlement between the U.S. Department of Justice and Meta Platforms, Inc. related to online targeted advertising.
Please join Troutman Pepper Partner Chris Willis for this solo episode as he reviews the settlement between the U.S. Department of Justice (DOJ) and Meta Platforms, Inc. related to online targeted advertising.
Chris digs into DOJ's complaint, alleging that Meta allowed its advertisers to use user's characteristics, such as race, color, religion, sex, and more, to determine which Facebook users would be eligible to receive certain housing ads. He also discusses the June 27, 2022 preliminary settlement that resolved allegations against Facebook, as well as the most recent January 9 settlement and its implications for financial services companies advertising on social media and web platforms.
The Consumer Finance Podcast: Settlement Agreement Update Between the DOJ and Meta
Host: Chris Willis
Posted: January 19, 2023
Chris Willis:
Welcome to the Consumer Finance Podcast. I'm Chris Willis, the co-leader of Troutman Pepper's Consumer Financial Services regulatory practice. And I'd like to welcome you to our special edition today, where we're going to be talking about the recent DOJ settlement with Facebook, relating to advertising and targeted advertising in particular. It's an area that I think is of great interest to a lot of our audience and certainly to me. But before I jump into that topic, let me remind you to visit and subscribe to our blog, consumerfinancialserviceslawmonitor.com, where we post every day about interesting events like this one in the consumer financial services world. And while you're at it, check out our other podcast, FCRA Focus, which of course, is all about credit reporting, The Crypto Exchange, about all things crypto, and our privacy and data security podcast called Unauthorized Access. All of which are available on all the popular podcast platforms. And if you like this podcast, let us know. Leave us a review on your podcast platform of choice and let us know how we're doing.
Now, as I said, today I'm going to be talking about the finalization of the settlement between the US Department of Justice and Facebook that arises from a claim that the Department of Justice filed back in June 2022, in which it alleged that the Facebook social media platform showed housing related ads to users of Facebook on a basis that was discriminatory on the basis of race. The parties filed a preliminary settlement agreement and stated essentially that they'd be working together between then, which was June 2022, and the end of the year, in December, to try to arrive at an agreement on what the remedial action should be on a going forward basis in connection with the Department of Justice's allegations.
Well, on January the 9th, 2023, they announced that they reached an agreement, and that the proposed settlement agreement has now become a final settlement agreement. So, I want to talk about what's happened there, but then more importantly, talk about what the implications are for creditors and other users of social media and web platforms for targeted advertising, and what do we know now and what still don't we know. That's what we're going to be talking about in today's podcast, and I'll be doing it all by myself. I don't have a special guest today. It's just me talking about this case.
The preliminary settlement between the Department of Justice and Facebook basically said the parties would work together to determine whether there was an agreement they could reach on numerical measurements for showing housing related ads to users in a more equal way than the DOJ had alleged was happening in its complaint filed back in June.
And so, what the parties announced via a letter from the Department of Justice to the presiding judge in the case dated on January 9th, 2023, is that indeed they did reach an agreement. And essentially what it involves is that the Facebook platform will now have an automated variance reduction system. And the idea there is it will review the audiences who see housing related ads and it will measure the demographic distribution of the audiences that receive those ads, both from the standpoint of gender as well as on race and ethnicity, as determined by a version of BISG, which is the same proxy method that many regulators use to try to determine someone's race or ethnicity when you don't know. What essentially the parties have agreed to is for different thresholds of percentage variances between the audiences that see ads that will become stricter over time.
Now, there are two different matrices in the letter that was sent to the court on January the 9th. One for ads that have more than a thousand people seeing them, and one for ads with at least 300 but less than a thousand people seeing them. I'm going to concentrate on the larger one because that seems like the more material one. And what it basically said is that initially by April 30th 2023, Facebook will ensure that there is a less than or equal to 10% variance on the basis of gender in ad impressions for 82.6% of the housing ads. And at that same time, there will be less than or equal to 5% variance in 73.2% of the ads. These numbers must have been negotiated somehow, but there's no context for them in the latter. As the year goes on, the year 2023 goes on, the standards become tighter so that by the end of 2023, 91.7% of the ads must have less than or equal to 10% of variance between men and women seeing them. And 84.5% of the ads have to have a less than or equal to 5% variance.
Now, there are also numbers in this chart for estimated race and ethnicity. Interestingly, the tolerance for variance is higher with respect to that. So, the final, December 31, 2023, allowable tolerance for that is that 81% of the ads must be viewed with a less than or equal to 10% variance across racial or ethnic groups. And 61% of the ads must achieve less than or equal to 5% of variants between racial or ethnic groups as determined by this modified version of BISG. We've got this matrix here and we've got this automated system that the parties have told the court about that Facebook is going to be implementing in order to ensure not perfectly uniform distribution of ads between gender and race and ethnicity groups, but ones that meet the thresholds that I just mentioned.
If you want to see the chart, we have a blog up about this as well, and we've linked to the letter sent to the court so you can actually see the chart yourself if you didn't happen to take notes on those percentages as I was giving them to you just a moment ago.
But I think, what I want to do now is step back and ask what does this mean to us, the credit world, and particularly creditors who may use Facebook or other social media platforms or web platforms to engage in targeted advertising of financial products or services? Remember, the DOJ case against Facebook was solely about housing under the Fair Housing Act, didn't have anything to do with credit. The first implication that I think there is to this is that a lot of financial institutions were a little bit reluctant to advertise on Facebook with this lawsuit unresolved and hanging out over the platform, because the idea was, "Well, hey, we're on official notice that the US government thinks there's discrimination going on on this ad targeting platform, I would really be using it at my own risk, if I'm a creditor, it's subject to the Equal Credit Opportunity Act, et cetera."
But now that the parties have actually finalized their settlement, and assuming Facebook will implement something like this for credit related ads, which we don't know yet, but perhaps they will, then I think the degree of confidence that supervised and even unsupervised financial institutions can have in advertising on the Facebook platform should be significantly increased. They shouldn't be as worried as they were beforehand because now the issue's been resolved and we now have a system that's acceptable to the Department of Justice in terms of reducing variances in the audience of who sees these various ads. So, I think that's a good thing for the industry and for Facebook as well, actually. Because there's been a lot of industry demand to advertise on Facebook because it's a very powerful platform, but there's been fear from a regulatory standpoint as well. And this may help to end that.
Also, I think we can say, that for other platforms, be they web search or other social media platforms, one would think that there would be some pressure on them to adopt something similar to what we've seen memorialized in this agreement with the Department of Justice in order to either avoid potential liability on the part of those platforms or to satisfy the concerns of advertisers who are sensitive about this, like financial institutions. And so, I foresee there will be a lot of conversations between financial services advertisers and the platforms that they use of, "Hey, are you doing anything like what's described in this settlement? And can you reassure me that you're doing that so that I can use your platform with confidence as well?" Those are two of the implications, but unfortunately, we still have some very significant unanswered questions that come out of this case.
Remember, that the degree of regulatory guidance that we have on the subject of targeted advertising is very limited, and almost all of it, if not all of it, has been in the form of regulatory activity directed at the social media platforms or web search platforms themselves rather than advertisers. And so, if you're sitting in the shoes of being an advertiser, which I think most of the listeners of this podcast would be, the real question is, is there an expectation that an advertiser needs to try to comply with these kinds of ratios and numbers in its own advertising efforts? And if so, how would an advertiser do that? Because a social media platform, of course, has complete information about everybody who sees an ad and why, but an advertiser who merely places ads on the platform probably doesn't have that insight. They don't know who saw the ad and who didn't, in a lot of instances. And so, they may not have the ability to measure the performance of their ad against the yardstick that this consent order and agreement have given us.
And so, there's a real question about does this apply to advertisers at all? Will a regulator like the DOJ or the CFPB, or somebody else, the Federal Trade Commission apply this to an advertiser at all? And if so, how is an advertiser supposed to figure out whether it's in compliance or not? We don't know the answer to that question, because there's nothing in this case that would tell us the answer to that. And the other big unanswered question, and I think it's a really important one, is to what extent would the kind of logic that we see both in the complaint in this case as well as in the resolution of the case apply under the Equal Credit Opportunity Act or under Dodd-Frank UDAAP as the CFPB has restyled it to cover discrimination through its announcement back in March 2022?
There's an obvious basis under the Fair Housing Act to attack advertising practices as discriminatory, because the Fair Housing Act directly covers advertising and prohibits discrimination in advertising, but there's not an analogous provision in the Equal Credit Opportunity Act. The only thing that it has in it that covers advertising, and it's not even in the statute, it's in Reg B, is that it prohibits discouraging someone from submitting an application. It's kind of a stretch to say that just because an ad doesn't come up in your social media feed that you've been discouraged from submitting an application, that seems like a kind of aggressive reading of Reg B and otherwise the Equal Credit Opportunity Act only prohibits discrimination against applicants. Well, an applicant means somebody who submit an application, and when you're sitting there scrolling your social media feed or looking at web search results, you're not an applicant.
And so, there's a very strong basis for distinguishing the Equal Credit Opportunity Act from the Fair Housing Act in this regard. But the question remains, will the CFPB and other regulators observe that distinction and be less aggressive on the credit side than we saw the Department of Justice be in this case under the Fair Housing Act? Or will they ignore the differences between the statutes and say, "Look, this is the standard for everything. We don't know the answer to that." And the final question that we still also don't know the answer to is, will the regulators be looking at targeted advertising through the lens of disparate treatment or disparate impact? In this case, the DOJ case that we're talking about today, also doesn't answer that question. When the HUD charge of discrimination that started this case came out, it read very much like disparate treatment.
It alleged that there was direct use of protected characteristics or extremely close proxies for protected characteristics that were used to select audiences for ads, and that therefore, that was a violation of the Fair Housing Act. Well, okay, that sounds like disparate treatment. You're directly taking into account protected characteristics and deciding who sees the ad and who doesn't. If you read the DOJ complaint that was filed back in June 2022, you see similar allegations that say, essentially, there was direct use of protected characteristics in placing ads. So, okay, still sounds like disparate treatment. However, the way the violations were pled in the DOJ's complaint, they were pled as both disparate treatment and disparate impact violations, which isn't helpful to anybody trying to figure out what the proper legal standard is. And moreover, the agreement that the parties have reached measures in a very impact related way, the final product of who sees what ads in what protected group, and in fact, it mandates that members of protected classes have to see these housing ads at least 80%, 90% of the rate of their non-Hispanic white counterparts or their male counterparts for the female population.
And so, that feels like a very disparate impact kind of measurement. The question remains, at least in my mind, when we think about targeted advertising, should we think about it in terms of treatment? That is, we shouldn't directly use protected characteristics like race or gender or ethnicity in saying who we want to see an ad and who we don't. That's something that's relatively easy for advertisers to do. Just don't do that. Or, on the other hand, do we have to be worried that facially neutral targeting criteria that are designed to find people who are interested in our product and who are likely to qualify for our product if they apply, might have a disparate impact because the population of people who are interested in a product and likely to be approved for a product does not mirror the population of the United States, which it probably doesn't.
And so, to me, it's very unrealistic, and unworkable, and unfair to think about targeted advertising from a disparate impact standpoint, both because I think it's unrealistic, and I also think it's probably impossible to monitor and measure, but we're left not knowing what the regulators are thinking in that regard because we have this mixture in this DOJ case, between these very clear allegations that look like disparate treatment, but yet a remedy and a measurement at the end that looks like it's trying to address disparate impact.
And so, that's a question that we still have to stay tuned and wait to hear the answer for through whatever further regulatory development occurs on this front, which frankly hasn't been very rapid in its development, because the regulators in general have not been saying much about this issue. Even the CFPB didn't say anything about it, when it made its UDAAP announcement in March 2022, it stated it was going to use its UDAAP authority to look at discrimination in targeted advertising, but then said nothing else about how it might be measured or assessed.
So, we still have to make our best guess on that, even though this is a very important data point for us in our understanding of the world of targeted advertising, and it gives us something more than we had before, it doesn't come even close to answering all the questions that I know we all have as members of the Consumer Financial Services marketplace.
As always, we're going to continue watching this issue very closely, and you can stay tuned to this podcast and keep reading our blog for any other updates that may happen with respect to it, and we'll all stay tuned together and see what happens. I want to thank you for joining me to listen to me talk about this important settlement today. Don't forget, as I said, to visit our blog, consumerfinancialserviceslawmonitor.com, and hit that subscribe button, so that you can see all of our daily updates about what's going on in the consumer financial services world. And while you're at it, go visit us over at troutman.com, and add yourself to our Consumer Financial Services email list. That way you'll get copies of our alerts as well as invitations to our industry only webinars. And of course, don't forget to tune in every Thursday afternoon for a great new episode of this podcast. Thank you all for listening.
Copyright, Troutman Pepper Hamilton Sanders LLP. These recorded materials are designed for educational purposes only. This podcast is not legal advice and does not create an attorney-client relationship.The views and opinions expressed in this podcast are solely those of the individual participants. Troutman Pepper does not make any representations or warranties, express or implied, regarding the contents of this podcast.Information on previous case results does not guarantee a similar future result. Users of this podcast may save and use the podcast only for personal or other non-commercial, educational purposes. No other use, including, without limitation, reproduction, retransmission or editing of this podcast may be made without the prior written permission of Troutman Pepper. If you have any questions, please contact us at troutman.com.