October 3, 2022

New Study Analyzing Political Advertising on Facebook, Google, and TikTok

By Orestis Papakyriakopoulos, Christelle Tessono, Arvind Narayanan, Mihir Kshirsagar

With the 2022 midterm elections in the United States fast approaching, political campaigns are poised to spend heavily to influence prospective voters through digital advertising. Online platforms such as Facebook, Google, and TikTok will play an important role in distributing that content. But our new study – How Algorithms Shape the Distribution of Political Advertising: Case Studies of Facebook, Google, and TikTok — that will appear in the Artificial Intelligence, Ethics, and Society conference in August, shows that the platforms’ tools for voluntary disclosures about political ads do not provide the necessary transparency the public needs. More details can also be found on our website: campaigndisclosures.princeton.edu.

Our paper conducts the first large-scale analysis of public data from the 2020 presidential election cycle to critically evaluate how online platforms affect the distribution of political advertisements. We analyzed a dataset containing over 800,000 ads about the 2020 U.S. presidential election that ran in the 2 months prior to the election, which we obtained from the ad libraries of Facebook and Google that were created by the companies to offer more transparency about political ads. We also collected and analyzed 2.5 million TikTok videos from the same time period. These ad libraries were created by the platforms in an attempt to stave off potential regulation such as the Honest Ads Act, which sought to impose greater transparency requirements for platforms carrying political ads. But our study shows that these ad libraries fall woefully short of their own objectives to be more transparent about who pays for the ads and who sees the ads, as well the objectives of bringing greater transparency about the role of online platforms in shaping the distribution of political advertising. 

We developed a three-part evaluative framework to assess the platform disclosures: 

1. Do the disclosures meet the platforms’ self-described objective of making political advertisers accountable?

2. How do the platforms’ disclosures compare against what the law requires for radio and television broadcasters?

3. Do the platforms disclose all that they know about the ad targeting criteria, the audience for the ads, and how their algorithms distribute or moderate content?

Our analysis shows that the ad libraries do not meet any of the objectives. First, the ad libraries only have partial disclosures of audience characteristics and targeting parameters of placed political ads. But these disclosures do not allow us to understand how political advertisers reached prospective voters. For example, we compared ads in the ad libraries that were shown to different audiences with dummy ads that we created on the platforms (Figure 1). In many cases, we measured a significant difference between the calculated cost-per-impression between the two types of ads, which we could not explain with the available data.

  • Figure 1. We plot the generated cost per impression of ads in the ad-libraries that were (1) targeted to all genders & ages on Google, (2) to Females, between 25-34 on YouTube, (3) were seen by all genders & ages in the US on Facebook, and (4) only by females of all ages located in California on Facebook.  For Facebook, lower & upper bounds are provided for the impressions. For Google, lower & upper bounds are provided for cost & impressions, given the extensive “bucketing” of the parameters performed by the ad libraries when reporting them, which are denoted in the figures with boxes. Points represent the median value of the boxes. We compare the generated cost-per impression of ads with the cost-per impression of a set of dummy ads we placed on the platforms with the exact same targeting parameters & audience characteristics. Black lines represent the upper and lower boundaries of an ad’s cost-per-impression as we extracted them from the dummy ads. We label an ad placement as “plausible targeting”, when the ad cost-per-impression overlaps with the one we calculated, denoting that we can assume that the ad library provides all relevant targeting parameters/audience characteristics about an ad.  Similarly, an placement labeled as `”unexplainable targeting’”  represents an ad whose cost-per-impression is outside the upper and lower reach values that we calculated, meaning that potentially platforms do not disclose full information about the distribution of the ad.

Second, broadcasters are required to offer advertising space at the same price to political advertisers as they do to commercial advertisers. But we find that the platforms charged campaigns different prices for distributing ads. For example, on average, the Trump campaign on Facebook paid more per impression (~18 impressions/dollar) compared to the Biden campaign (~27 impressions/dollar). On Google, the Biden campaign paid more per impression compared to the Trump campaign. Unfortunately, while we attempted to control for factors that might account for different prices for different audiences, the data does not allow us to probe the precise reason for the differential pricing. 

Third, the platforms do not disclose the detailed information about the audience characteristics that they make available to advertisers. They also do not explain how the algorithms distribute or moderate the ads. For example, we see that campaigns placed ads on Facebook that were not ostensibly targeted by age, but the ad was not distributed uniformly.  We also find that platforms applied their ad moderation policies inconsistently, with some instances of moderated ads being removed and some others not, and without any explanation for the decision to remove an ad. (Figure 2) 

  • Figure 2. Comparison of different instances of moderated ads across platforms. The light blue bars show how many instances of a single ad were moderated, and maroon bars show how many instances of the same ad were not. Results suggests an inconsistent moderation of content across platforms, with some instances of the same ad being removed and some others not.

Finally, we observed new forms of political advertising that are not captured in the ad libraries. Specifically, campaigns appear to have used influencers to promote their messages without adequate disclosure. For example, on TikTok, we document how political influencers, who were often linked with PACs, generated billions of impressions from their political content. This new type of campaigning still remains unregulated and little is known about the practices and relations between influencers and political campaigns.  

In short, the online platform self-regulatory disclosures are inadequate and we need more comprehensive disclosures from platforms to understand their role in the political process. Our key recommendations include:

– Requiring that each political entity registered with the FEC use a single, universal identifier for campaign spending across platforms to allow the public to track their activity.

– Developing a cross-platform data repository, hosted and maintained by a government or independent entity, that collects political ads, their targeting criteria, and the audience characteristics that received them. 

– Requiring platforms to disclose information that will allow the public to understand how the algorithms distribute content and how platforms price the distribution of political ads. 

– Developing a comprehensive definition of political advertising that includes influencers and other forms of paid promotional activity.

AdNauseam, Google, and the Myth of the “Acceptable Ad”

Earlier this month, we (Helen Nissenbaum, Mushon Zer-Aviv, and I), released a new and improved AdNauseam 3.0. For those not familiar, AdNauseam is the adblocker that clicks every ad in an effort to obfuscate tracking profiles and inject doubt into the lucrative economic system that drives advertising-based surveillance. The 3.0 release contains some new features we’ve been excited to discuss with users and critics, but the discussion was quickly derailed when we learned that Google had banned AdNauseam from its store, where it had been available for the past year. We also learned that Google has disallowed users from manually installing or updating AdNauseam on Chrome, effectively locking them out of their own saved data, all without prior notice or warning.

Whether or not you are a fan of AdNauseam’s strategy, it is disconcerting to know that Google can quietly make one’s extensions and data disappear at any moment, without so much as a warning. Today it is a privacy tool that is disabled, but tomorrow it could be your photo album, chat app, or password manager. You don’t just lose the app, you lose your stored data as well: photos, chat transcripts, passwords, etc. For developers, who, incidentally, must pay a fee to post items in the Chrome store, this should cause one to think twice. Not only can your software be banned and removed without warning, with thousands of users left in the lurch, but all comments, ratings, reviews, and statistics are deleted as well.

When we wrote Google to ask the reason for the removal, they responded that AdNauseam had breached the Web Store’s Terms of Service, stating that “An extension should have a single purpose that is clear to users”[1]. However, the sole purpose of AdNauseam seems readily apparent to us—namely to resist the non-consensual surveillance conducted by advertising networks, of which Google is a prime example. Now we can certainly understand why Google would prefer users not to install AdNauseam, as it opposes their core business model, but the Web Store’s Terms of Service do not (at least thus far) require extensions to endorse Google’s business model. Moreover, this is not the justification cited for the software’s removal.

So we are left to speculate as to the underlying cause for the takedown. Our guess is that Google’s real objection is to our newly added support for the EFF’s Do Not Track mechanism[2]. For anyone unfamiliar, this is not the ill-fated DNT of yore, but a new, machine-verifiable (and potentially legally-binding) assertion on the part of websites that commit to not violating the privacy of users who choose to send the DNT header. A new generation of blockers including the EFF’s Privacy Badger, and now AdNauseam, have support for this mechanism built-in, which means that they don’t (by default) block ads and other resources from DNT sites, and, in the case of AdNauseam, don’t simulate clicks on these ads.

So why is this so threatening to Google? Perhaps because it could represent a real means for users, advertisers, and content-providers to move away from surveillance-based advertising. If enough sites commit to Do Not Track, there will be significant financial incentive for advertisers to place ads on those sites, and these too will be bound by DNT, as the mechanism also applies to a site’s third-party partners. And this could possibly set off a chain reaction of adoption that would leave Google, which has committed to surveillance as its core business model, out in the cold.

But wait, you may be thinking, why did the EFF develop this new DNT mechanism when there is AdBlock Plus’ “Acceptable Ads” programs, which Google and other major ad networks already participate in?

That’s because there are crucial differences between the two. For one, “Acceptable Ads” is pay-to-play; large ad networks pay Eyeo, the company behind Adblock Plus, to whitelist their sites. But the more important reason is that the program is all about aesthetics—so-called “annoying” or “intrusive” ads—which the ad industry would like us to believe is the only problem with the current system. An entity like Google is fine with “Acceptable Ads” because they have more than enough resources to pay for whitelisting[3] . Further, they are quite willing to make their ads more aesthetically acceptable to users (after all, an annoyed user is unlikely to click)[4]. What they refuse to change (though we hope we’re wrong about this) is their commitment to surreptitious tracking on a scale never before seen. And this, of course, is what we, the EFF, and a growing number of users find truly “unacceptable” about the current advertising landscape.

 

[1]  In the one subsequent email we received, a Google representative stated that a single extension should not perform both blocking and hiding. This is difficult to accept at face value as nearly all ad blockers (including uBlock, Adblock Plus, Adblock, Adguard, etc., all of which are allowed in the store) also perform blocking and hiding of ads, trackers, and malware. Update (Feb 17, 2017): it has been a month since we have received any message from Google despite repeated requests for clarification, and despite the fact that they claim, in a recent Consumerist article, to be “in touch with the developer to help them resubmit their extension to get included back in the store.”

[2] This is indeed speculation. However, as mention in [1], the stated reason for Google’s ban of AdNauseam does not hold up to scrutiny.

[3]  In September of this year, Eyeo announced that it would partner with a UK-based ad tech startup called ComboTag to launch the“Acceptable Ads Platform” with which they would act also as an ad exchange, selling placements for “Acceptable Ad” slots.  Google, as might be expected, reacted negatively, stating that it would no longer do business with ComboTag. Some assumed that this might also signal an end to their participation in“Acceptable Ads” as well. However, this does not appear to be the case. Google still comprises a significant portion of the exception list on which “Acceptable Ads” is based and, as one ad industry observer put it, “Google is likely Adblock Plus’ largest, most lucrative customer.”

[4]  Google is also a member of the “Coalition for Better Ads”, an industry-wide effort which, like “Acceptable Ads”, focuses exclusively on issues of aesthetics and user experience, as opposed to surveillance and data profiling.

 

On distracted driving and required phone searches

A recent Arstechnica article discussed several U.S. states that are considering adding a “roadside textalyzer” that operates analogously to roadside Breathalyzer tests. In the same way that alcohol and drugs can impair a driver’s ability to navigate the road, so can paying attention to your phone rather than the world beyond. Many states “require” drivers to consent to Breathalyzer tests, where that “requirement” boils down to serious penalties if the driver declines. Vendors like Cellebrite are pushing for analogous requirements, for which they just happen to sell products.
[Read more…]

Android WebView security and the mobile advertising marketplace

Freedom to Tinker readers are probably aware of the current controversy over Google’s handling of ongoing security vulnerabilities in its Android WebView component. What sounds at first like a routine security problem turns out to have some deep challenges.  Let’s start by filling in some background and build up to the big problem they’re not talking about: Android advertising.
[Read more…]

Google Fights Genericide Claim (and Wins)

Google’s famous trademark in its name has just survived a challenger’s attempt to have it declared generic. In Elliott v. Google, a federal court in Arizona held last week that despite the public’s use of the word “googling” to mean “searching on the Internet,” the “Google” word mark still functions in the minds of consumers primarily to identify Google, the Mountain View-based Internet company, as the source of the search service associated with the “Google” mark. The plaintiff in the case argued that the public’s use of a trademark as a verb necessarily signifies that the mark has become generic. The court disagreed:

Verb use of a trademark is not fundamentally incapable of identifying a producer or denoting source. A mark can be used as a verb in a discriminate sense so as to refer to an activity with a particular product or service, e.g., “I will PHOTOSHOP the image” could mean the act of manipulating an image by using the trademarked Photoshop graphics editing software developed and sold by Adobe Systems. This discriminate mark-as-verb usage clearly performs the statutory source-denoting function of a trademark.

The court went on to explain that a problem arises for a mark owner only if mark-as-verb usage is indiscriminate, and the mark becomes referentially unmoored in the public’s mind from the mark owner’s product or service.

[Read more…]