Ad transparency transformed digital politics. After four years, what’s next?
Big tech’s 2018 move to publicly catalog political ads was a win for everyone. They should build on it.
Four years ago this month, Facebook (now Meta) launched a searchable database of all political ads running on its sites. The move was quickly followed by other digital ad platforms (including Google, Snapchat, Twitter, Reddit, and Roku) providing some type of political ad transparency in the Iead-up to the 2020 election. Since then, these valuable data sources provided researchers, journalists, and the public with a comprehensive look at paid political influence campaigns online - and set a new standard for transparency that should be expanded in the future.
We’re not exaggerating when we say we’ve probably spent more time scrolling through these various transparency databases than anyone else over the past 3+ years. Here’s what we’ve learned:
Meta and Google have led the way
Meta’s Ad Library has been a powerful example of what platforms are capable of building when they invest the time and resources to do so. The company’s ad database is the most comprehensive and user-friendly in the industry and provides users with multiple options for browsing the $3.2 billion of ad dollars spent on Facebook and Instagram since May 2018.
Despite some debate over whether the company should ban political ads altogether (they shouldn’t), Meta continues to release meaningful updates to its Ad Library tool, showing an ongoing commitment to improvement ahead of the 2022 and 2024 election cycles.
Similarly, Google’s Transparency Report has provided valuable insight into the enormous political ad spending on YouTube, where campaigns and PACs often blanket battleground states with video ads ahead of Election Day. The company continues to build upon their report in a way that improves the public’s ability to navigate the site.
While Snapchat’s Political Ads Library provides some of the most granular ad targeting information of the three, it lacks a clean user interface and is really just a downloadable spreadsheet.
What’s considered a “political ad” is open to interpretation
Companies like Meta, Google, and Twitter define political advertising in different ways, and that has far-reaching impacts on their transparency work. Meta, for example, chooses to define “political and issue advertising” pretty broadly, often resulting in objectively non-political ads about health, nutrition, or clean water being archived in its Ad Library with frequency.
Google, on the other hand, fails to provide transparency into many political advocacy and issue campaigns, opting to focus its transparency work on ads mentioning candidates for federal and state offices or ballot measures. Although Twitter bans “political ads,” the company allows advertising about certain political issues and advocacy, as long as the ad copy is vague enough. Twitter does not archive those ads.
Ad transparency has helped hold bad actors accountable (sometimes).
The transparency provided by these companies has produced some solid results. In 2019, our newsletter revealed that the far-right news outlet Epoch Times was quietly spending millions of dollars on deceptive pro-Trump Facebook advertising, which ultimately led Facebook to ban Epoch’s ads altogether. Over the past four years, we and many other groups have been able to keep some of the most insidious advertisers in check thanks to the insight provided by Meta’s library
It’s tough to estimate if or how many bad actors were dissuaded from launching political advertising on platforms like Google, Facebook, or Snapchat due to increased advertiser disclosure and verification requirements. However, we know that political campaigns and their digital consultants are now well aware that every ad they run will be archived for journalists and the public to find.
Ad libraries have also revealed various shortcomings in platform self-regulation and policy enforcement. For instance, despite Donald Trump being “deplatformed” from Facebook in January 2021, we broke the news last summer that his team was still allowed to run political ads on the platform. Other journalists have kept an eye on the Google Transparency report, calling attention to inconsistencies when they find them.
Platforms should expand, not retreat on transparency investments
Today, Twitter and Reddit no longer archive political advertising on their platforms (Twitter outright banned electoral ads), and Roku’s ad archive is limited to the point of uselessness. As political advertisers continue to diversify their media mix online, those companies should reverse course and re-commit resources to following Meta, Google, and Snap’s lead on transparency.
Most importantly, transparency should no longer apply to political ads alone. The majority of “misinformation” or political mal-intent spreads organically in places like YouTube, Facebook’s Feed or via TikTok’s viral algorithm. Tools like Meta’s CrowdTangle have shown the public that the platforms are capable of providing transparency into the virality of organic content, but are so far reluctant to do so. The next major frontier in political platform transparency should be the creation of more CrowdTangle-like databases for all major social platforms - systems that would allow the public to understand the most viral political messages and the people behind them.
What it means for 2022, 2024
Each election cycle, digital advertising continues to be a bigger piece of political campaigns’ paid media budgets. Ahead of this fall’s midterms, big tech companies like Facebook, Google, Snapchat, and others should build on their transparency efforts, and others like TikTok, Twitter, and Reddit should re-think their approaches to political content transparency for both paid and organic content.
When it comes to stemming the spread of misinformation and misleading political content online, these past four years have shown that transparency alone won’t save us - but it’s an important first step.