The Technology 202: New Duke paper calls Washington to increase transparency around online political ads

Researchers trying to answer these questions instead encountered a major black box. The report estimates that in the final months of the election, campaigns directed 94 percent of their advertising spending through consulting firms. 

Duke researchers want Washington to change rules so that political organizations have to report how advertising agencies are spending on their behalf. Currently political campaigns have to disclose that they paid an agency or consulting firm, and researchers can’t see how those firms are spending the money.

This makes it challenging for researchers to tell whether the ad bans were effective – and how they impacted political discourse. “There’s so much PR and regulatory risk that platforms are closing off political advertisers in ways that potentially have a negative effect on political speech,” said Matt Perault, a co-author of the report and former director of public policy at Facebook. 

Perault said the goal of the report is to “shed some light on a set of solutions that the government and platforms could consider that would address some of the challenges but facilitate the good of political advertising.” 

Companies such as Twitter, LinkedIn, Pinterest and TikTok don’t allow political ads, while Facebook and Google instituted temporary political ad blackouts. Google resumed selling such ads last week, while they continue to remain blocked on Facebook.

The report also calls on the Federal Election Commission to ensure that there’s greater standardization and specificity in how campaigns report their online ad spending, and for the agency to release these disclosures in a more timely fashion so they can be studied more effectively by researchers. 

There has been growing pressure to pass legislation to address problems with online political advertising since the aftermath of the 2016 election, when Russian trolls were able to purchase ads on social networks including Facebook to sow chaos and division. More recently, companies have faced backlash for not strictly policing ads from domestic politicians, including former president Donald Trump, who were spreading falsehoods. 

In the absence of action from Washington, tech companies have taken matters into their own hands in the face of intense public and political pressure. Facebook, Google and Snap have launched their own transparency tools for researchers and journalists to study how ads are being leveraged. But they vary greatly from platform to platform, making it difficult for researchers to answer pressing questions about social media that could influence the regulatory debate. Other companies, most notably Twitter, have pulled out of the political ad business altogether to avoid the increased risk in an area that makes up only a fraction of their total revenue. 

J. Scott Babwah Brennen, a co-author of the report, told me in an interview that he initially set out to explore whether these bans were effective at stopping the spread of misinformation, or whether they had implications for certain candidates. But he soon realized that there is not enough transparency around the ads to effectively study that, and intervention from Washington is needed. 

“It got us more interested in looking into some of the access and data challenges,” he told me.

In addition to changing regulations at the Federal Election Commission, Brennen and Perault call for Congress to pass an updated version of the Honest Ads Act, bipartisan legislation that was initially introduced in 2017 to create greater transparency around political ads. They say that the Congress should consider modifying the bill to require companies “to maintain standardized, publicly accessible, and searchable ad databases.”

The researchers also say in the meantime, there are clear steps the tech companies can make to make it easier for people to study political ads and the impact of their policies. Right now, each company has designed their transparency tools on an ad hoc basis. But the Duke paper calls for the companies to “standardize data types and variable names, improve search functionality, and increase user access.”

Brennen and Perault tell me that this will likely just be the first in a series of papers studying the impact of online political ads. They said in the report they continue to have three remaining questions:

1. How did political campaigns and organizations shift their spending in the wake of advertising bans or blackouts from social media companies?
2. The bans were intended to stop the spread of misinformation. Did the content of the ads on social media change at all following these measures?
3. How did bans impact individual campaigns? Are there meaningful differences in how they impacted challengers versus incumbents? Was one political party impacted more than another?

Charlotte Newman, a manager in Amazon’s cloud-computing division, said that Amazon executives used racial stereotypes and held back Black employees, Jay Greene reports. The lawsuit is just the latest sign of employee unrest at the tech giant, as Amazon fends off claims of discrimination by workers and opposes a unionization drive at a warehouse in Bessemer, Ala., where many workers are Black. (Amazon founder Jeff Bezos owns The Washington Post.)

Newman’s lawsuit, filed in a D.C. court, accuses former co-workers of describing her as “too direct,” “just scary,” and saying she looked “like a gorilla.” The suit also alleges that another former Amazon employee grabbed Newman’s thigh at a work dinner and propositioned her for sex. 

The suit names Amazon, two Amazon employees and another person who has left the company.

Amazon does not tolerate discrimination or harassment, spokesperson Kate Brinks said. “We are currently investigating the new allegations included in this lawsuit,” Brinks said in an emailed statement.

The company says it will add labels and a five-strike system to root out misleading posts and the users who push them. It follows pressure from the Biden administration and Democrats on social media giants to update their policies around false claims about coronavirus vaccinations.

Twitter said that the company has removed 8,400 tweets and challenged 11.5 million accounts since December, when it announced a new policy to crack down on claims about vaccines. 

The Justice Department and some state attorneys general are asking the search giant for data on search results and related ads from 2015 and 2020, Bloomberg News’s Nico Grant reports. The request indicates that the Biden administration is pressing ahead with the antitrust lawsuit brought against Google under former president Donald Trump, and it signals they’re looking into how the company’s position in search evolved over time. 

The government has gone after a variety of records in the case, but Google has fought back on the grounds that the requests are too broad. The landmark antitrust trial is set to begin in 2023.

Google has rejected the claims in the government’s antitrust lawsuit, saying that consumers are able to go to its competitors.

Twitter’s new policy for misleading claims about coronavirus vaccinations left some people wanting more. Author Dan Sinker:

Digital Democracy Project Director Alex Howard:

Others noted the policy appeared clear, in sharp contrast to some tech companies’ misinformation rules. Renee DiResta, technical research manager at the Stanford Internet Observatory:

Ian Conner, the director of the Trump administration Federal Trade Commission’s competition bureau, has joined Latham & Watkins as a partner.

• Former acting assistant attorney general for national security Mary McCord speaks at a Carnegie Endowment for International Peace event on fighting right-wing extremism online today at 10 a.m.
• The Senate Banking Committee holds a hearing on President Biden’s nominees to lead the Securities and Exchange Commission and Consumer Financial Protection Bureau, Gary Gensler and Rohit Chopra, today at 10 a.m.
• The Center for Strategic and International Studies hosts an event on quantum computing on Wednesday at 1 p.m.
• YouTube CEO Susan Wojcicki speaks at an Atlantic Council event on Thursday at 1 p.m.
• The Brookings Institution hosts an event on the government’s role in reducing bias in algorithms on March 12 at 9 a.m.