[This is largely a reprint from a blog post originally posted on the Public Knowledge blog.]
The last week or so has highlighted the complete inadequacy of our political advertising rules in an era when even the President of the United States has no hesitation in blasting the world with unproven conspiracy theories about political rivals using both traditional broadcast media and social media. We cannot ignore the urgency of this for maintaining fair and legitimate elections, even if we realistically cannot hope for Congress to address this in a meaningful way any time soon.
To recap for those who have not followed closely, President Trump has run an advertisement repeating a debunked conspiracy theory about former Vice President Joe Biden (a current frontrunner in the Democatic presidential primary). Some cable programming networks such as CNN and those owned by NBCU have refused to run the advertisement. The largest social media platforms — Facebook, Google, and Twitter — have run the advertisement, as have local broadcast stations, despite requests from the Biden campaign to remove the ads as violating the platform policy against running advertising known to contain false or misleading information. The social media platforms refused to drop the ads. Facebook provided further information that it does not submit direct statements by politicians to fact checkers because they consider that “direct speech.”
Elizabeth Warren responded first with harsh criticism for Facebook, then with an advertisement of her own falsely stating that Zuckerberg had endorsed President Trump. Facebook responded that the Trump advertisement has run “on broadcast stations nearly 1,000 times as required by law,” and that Facebook agreed with the Federal Communications Commission that “it’s better to let voters — not companies — decide.” Elizabeth Warren responded with her own tweet that Facebook was “proving her point” that it was Facebook’s choice “whether [to] take money to promote lies. You can be in the disinformation-for-profit business or hold yourself to some standards.”
Quite a week, with quite a lot to unpack here. To summarize briefly, the Communications Act (not just the FCC) does indeed require broadcast stations that accept advertising from political candidates to run the advertisement “without censorship.” (47 U.S.C. §315(a).) While the law does not apply to social media (or to programming networks like NBCU or CNN), there is an underlying principle behind the law that we want to balance the ability of platforms to control their content with preventing platforms from selectively siding with one political candidate over another while at the same time allowing candidates to take their case directly to the people. But, at least in theory, broadcasters also have other restrictions that social media platforms don’t have (such as a limit on the size of their audience reach), which makes social media platforms more like content networks with greater freedom to apply editorial standards. But actual broadcast licensees — the local station that serves the viewing or listening area — effectively become “common carriers” for all “qualified candidates for public office,” and must sell to all candidates the opportunity to speak directly to the audience and charge all candidates the same rate.
All of this begs the real question, applicable to both traditional media and social media: How do we balance the power of these platforms to shape public opinion, the desire to let candidates make their case directly to the people, and the need to safeguard our ability to govern ourselves? Broadcast media remain powerful shapers of public opinion, but they clearly work in a very different way from social media. We need to honor the fundamental values at stake across all media, while tailoring the specific regulations to the specific media.
Until Congress gets off its butt and actually passes some laws we end up with two choices. Either we are totally cool with giant corporation making the decision about which political candidates get heard and whether what they have to say is sufficiently supported and mainstream and inoffensive to get access to the public via social media, or we are totally cool with letting candidates turn social media into giant disinformation machines pushing propaganda and outright lies to the most susceptible audiences targeted by the most sophisticated placement algorithms available. It would be nice to imagine that there is some magic way out of this which doesn’t involve doing the hard work of reaching a consensus via our elected representatives on how to balance competing concerns, but there isn’t. There is no magic third option by which platforms acting “responsibly” somehow substitutes for an actual law. Either we make the choice via our democratic process, or we abdicate the choice to a handful of giant platforms run by a handful of super-rich individuals. So perhaps we could spend less time shaming big companies and more time shaming our members of Congress into actually doing their freaking jobs!!
(OK, spend more time doing both. Just stop thinking that yelling at Facebook is gonna magically solve anything.)
I unpack this below . . .
I addressed a number of these issues in Chapter 5 and Chapter 6 of my book “The Case for the Digital Platform Act” (available for free download at digitalplatformact.com as well as on Amazon). But I want to expand on these issues and stress the urgency of Congress addressing these issues below.
First Step: What Is the Actual Law?
The law governing political advertising goes back to the Federal Radio Act of 1927. Broadcast radio had gone in a short time from a curiosity to a powerful shaper of public opinion. Like modern social media platforms, broadcasters also based their dominant business model on advertising. Congress rejected a nationalized broadcast service such as the BBC in England, and rejected converting broadcasters into common carriers. At the same time, Congress recognized that broadcasters could easily influence an election by refusing to take advertisements from one candidate, or charging one candidate a more favorable rate.
The result was that political advertising became the most regulated content under the Communications Act. Section 315 and Section 312(a)(7) (47 U.S.C. §312(a)(7)) require that commercial broadcasters take advertising from “qualified federal candidates.” (While Section 315 allows broadcasters to refuse all advertising from “qualified candidates,” Section 312(a)(7) requires commercial broadcasters to take advertising from all qualified federal candidates whether the licensee wants to take the advertisement or not. (Since President Trump and Senator Warren are both “qualified federal candidates,” both Section 312(a)(7)’s “reasonable access” requirement and the requirements of Section 315 apply). Additionally, Section 315 requires broadcasters to charge political candidates no more than “the lowest unit rate” for advertising, and to maintain a public file on political advertising. Finally, Section 315 empowers the FCC to enact rules “to carry out the provisions of this section.”
So yes, local radio and television broadcasters need to take advertising from all qualified candidates. Further, the law states broadcasters “shall have no power of censorship over the material broadcast.” Subsequent case law and FCC precedent interpret the “reasonable access” clause of Section 312(7)(a) and the “no power of censorship” clause of Section 315(a) to give political candidates the power to require broadcasters to broadcast the advertisement at specific times of the candidates choosing, and to air the ads unedited even if the advertisement contains material that the broadcast station finds offensive or fears the public will find offensive. Additionally, the FCC by rule through its “ancillary authority” requires local cable systems to follow the rule for any local advertising. The FCC also requires satellite providers (i.e., Dish Network or DirecTV) offering local into local advertising to observe the same rules.
Importantly, this doesn’t apply to programming networks. So NBC the programming network or CNN the cable network can turn down any candidate advertising they want. The rule applies to the local licensee. So even though NBC owns lots of stations, it does not have to run an advertisement on its network programming, but every local station it owns that is approached by a “qualified federal candidate” must comply with Section 312(a)(7) and Section 315 and run the desired advertisement in its viewing area without any editorial discretion.
Needless to say, the law does not apply to social media companies directly. So social media companies, as a matter of law, do not have to accept or deny any specific candidate’s advertising. The choice is theirs.
The Law Doesn’t Apply, but the Concerns Remain.
So is Facebook lying when it invokes the rules that apply to broadcasters? No, but that doesn’t mean they bear no responsibility for their choices. Facebook, and presumably the other social media providers that accepted the Trump advertisement and found it did not violate their policies, does not claim that the law requires the company to carry the Trump advertisement. Rather, Facebook argues that the same fundamental values that prompted Congress (and the FCC) to create the rules for broadcast licensees should guide the company’s own political advertising policies. Therefore, when a political candidate speaks through an advertisement, the social media provider lets the candidate speak directly (as if it were through their own Facebook page or Twitter account) to the public.
So does that make Facebook right — at least in principle? Unfortunately, it’s more complicated than that. It’s not just, as Elizabeth Warren retorts, that Facebook (and other social media providers) are choosing to act like broadcast licensees rather than like broadcast networks and therefore choosing to spread disinformation for money. The problem lies in the fact that mechanically applying broadcast regulations to social media companies doesn’t really strike the balance intended by Congress when it adopted Section 315.
Why the Old Rules Just Don’t Strike the Same Balance for Social Media.
Let’s start with the obvious problem that we don’t use social media, or experience social media advertising, in the same way we experience broadcast advertising. I’m not talking about the question of how influential the medium is. Radio advertising in the 1930s was arguably as powerful in shaping public perception as social media advertising is today. But local broadcasters reached a very limited listening/viewing area. Additionally, two candidates could expect to have their competing advertising heard by approximately the same audience if they bought advertising for the same time. If Biden or Warren want to buy broadcast advertising to counter Trump’s advertising, they can request advertising in the same general time period, and at the same rate, so that the same audience will be able to see the advertisements and compare candidates.
But social media advertising does not work that way. Even with social media companies trying to make adjustments in how they do political advertising, as far as I know they still allow political candidates to use targeted advertising and have no obligation to charge candidates “the lowest unit rate.” Even if social media services did their absolute best to try to replicate the broadcast advertising experience, they simply could not do it. How we use social media, and how it is structured, is simply too different. It is not comparable to programming networks, let alone broadcast licensees.
Furthermore, every social media network is going to have a different way it interprets the policies. Again, assuming the best faith on the part of everyone, and assuming a genuine commitment by social media companies to honor the balance struck by Congress in Section 315, it just doesn’t work. Nor do frustrated candidates have the option to go to a neutral third party like the FCC or the courts to resolve conflicting views of how to interpret the law or to complain that the policy is being applied unfairly.
Once again, mechanical application of the old rules to new media just makes the situation worse.
The Problem Goes Beyond Social Media. Old Media Still Remain Important, And the FCC Has Repealed Most of the Other Protections That Complemented The Law.
While Facebook has — once again — proven the lightning rod for controversy, we shouldn’t ignore the fact that broadcasters still have to take advertisements that are false or utterly unsupported by any facts. We should not ignore that whether programming networks run or don’t run such advertisements remains a matter of corporate discretion. Radio, broadcast television, and cable programming remain powerful influences on the voting public, and a majority of Americans still prefer to get their news through television and radio.
The balance struck by Congress in the 1920s and 1930s was never designed for today’s world of massive, consolidated media, hyper-partisan talk radio, and 24/7 cable news networks breathlessly reporting and debating conspiracy theories as news. Worse, the FCC has repealed many of the protections it previously put in place to discourage politicians from using Section 315 to spread disinformation. For example, the FCC used to have something called the “Personal Attack Rule.” Under this rule, an attack on the “honesty, character, integrity or like personal qualities” of an identified individual required the broadcaster to provide free air time for a personal reply. This created a disincentive for candidates to use Section 315 to run advertisements of false claims of corruption, since these advertisements would give the opposing candidate a bonanza of free reply time. But the personal attack rule was part of the Fairness Doctrine, and the FCC’s repeal of the Fairness Doctrine ultimately meant repeal of the Personal Attack Rule (albeit after nearly 15 years of litigation).
Certainly problems with other existing media should not justify giving social media a free pass to ignore the consequences of their decisions on advertising policy. But we equally should not focus exclusively on social media and imagine that we can adequately address the existential threat to democracy by shaming a few Silicon Valley billionaires or by breaking up social media companies.
Congress Needs to Re-Examine How to Apply Our Fundamental Values in Today’s Media Environment.
The existential threat to democracy is perhaps the worst it has ever been since Congress initially passed the Communications Act of 1934. Then, as now, liberal democracies were crumbling around the globe to be replaced by Fascist strongmen and Communist dictators leveraging the power of mass media to spread propaganda and target political enemies. The fundamental values Congress embodied in the law — respect for the independence of news providers on the one hand, with acknowledgement of the power of electronic platforms to threaten our access to diverse sources of news and perspectives on the other hand — remain the same and as urgent as they were then. But the actual mechanics of how to balance these conflicting concerns no longer work.
It’s up to Congress to protect our democracy by striking the right balance with political ads. We should not outsource a question of how to balance fundamental values and existential threats to democracy to private companies. Not only does this outsource the fundamental role of government to a handful of giant profit-maximizing companies, it simply won’t work. Companies cannot make these judgments in a consistent manner that’s accountable to the public. Nor should they.
The 1920s and 1930s were no less partisan an era than our own. But members of both political parties managed to come together to find a workable balance designed to protect democracy. Despite the temptation, neither the Republicans who dominated Congress and the White House in the 1920s, nor the Democrats who dominated Congress in the 1930s, designed a system of advertising rules to give themselves a partisan advantage. It is not unrealistic for the American people to demand that today’s Congress do the same.
But What Should Facebook (and Other Social Media Platforms) Do Until Congress Acts?
An advantage of rules is that they actually tell companies what to do and how to stay out of trouble in situations like this that require balancing and with no clear right answer. But sadly, no one expects Congress to step up and resolve the issue any time soon. This sticks social media platforms in a classic Kobayashi Maru, aka the “no win” scenario. At the same time, social media platforms can’t avoid the problem. They either take advertisements, or they don’t. They either make judgments about whether these are “political advertisements,” or they don’t. They either allow political candidates to speak directly through advertisements without any fact-checking or filter, or they don’t. And, of course, deciding what to do about direct candidate advertising is only one small piece of the overall problem of addressing disinformation.
But “sufficient unto the day is the evil thereof.” Until Congress can actually do its job, here are some recommendations for ways social media platforms can address the issues raised by direct candidate advertising.
Don’t Change Anything. This is, frankly, the rational thing for companies to do in the absence of any direction from Congress — as long as the companies apply this policy consistently. Facebook has applied things consistently by passing through Elizabeth Warren’s “Zuckerberg endorses Trump” ad high-lightng Facebook’s policy. The fact is there is no good choice here. Whether you like Facebook’s (or Google’s or Twitter’s) decision to let candidates speak directly (through both their accounts and advertising) or not, the law places the ball in their court and the decision here rests on an important principle of letting candidates speak directly to the voting public. Whether or not you like the results, it’s not a crazy decision. It’s one possible choice out of a list of bad possible choices. To quote Gilbert & Sullivan from the Mikado: “it’s an unjust world, and virtue is only triumphant in theatrical performances.”
True, as I said above, I don’t think this is good policy overall. But that is why Congress needs to step up here. Bluntly, the more pressure for Congress to step up, the better. In fact, I would hope that companies would actively tell Congress “hey guys, we’re happy to follow the rules, but you need to actually figure out the rules. No one elected social media companies to balance the dangers of gatekeepers picking what candidate advertisements are unacceptable v. the danger of candidates spreading lies and crazy conspiracy theories. Our job is to provide a platform and take ads. We do our job, when are you in Congress going to do yours?” I don’t expect the companies to be so blunt, mind you. On the other hand, I will point out that we voters can — and should — say this directly and bluntly to our elected officials.
Don’t take advertisements that don’t meet fact-checking standards. Alternatively, companies can stick to what the law requires and refuse to distinguish between direct advertisements from candidates and any other kind of advertisement. Mind you, this will promptly set off a screaming fit from anyone who finds their advertisement rejected. But again, there is no good choice here. The principle of allowing candidates to speak directly to the public is not absolute, and needs to be balanced in an environment where some candidates deliberately trying to manipulate the truth, undermine confidence in the electoral process, and generally ‘break’ the system — or act with reckless disregard for the truth of their allegations. Given the state-of-play today, one can easily argue that the balance tilts away from allowing candidates to speak directly and more toward refusing to take advertising that relies on false allegations or acts in reckless disregard of the truth.
To the extent a candidate’s supporters don’t like this approach, let them call their member of Congress and push them to pass legislation requiring social media to pass through advertising from “qualified candidates . . . without the power of censorship,” like Congress did for broadcasters almost 90 years ago. If there is one group of people in the United States who can (and should!) actually tell companies like Facebook what to do, it’s Congress. Also, if tech companies want to get Congress to stop trying to browbeat them into submission, then they should make it clear Congress needs to pass a law if they want to see anything change.
Social media platforms can still allow candidates to speak directly through unpaid posts on their personal or even campaign accounts. Yes, there are also questions about what is appropriate on candidate personal accounts. But the argument for allowing a candidate to speak directly in the context where individuals encounter the statements because they chose to follow the candidate (either because they support the candidate or wish to rebut the candidate) is somewhat different than using targeted advertising to find individuals particularly receptive to the false or deceptive message the candidate wants to send.
Block political advertising (or at least direct statements by candidates) for some set period of days before an election. Political pressure being what it is, companies may want a third option to show critics they have heard the criticism and take it seriously. Some countries use “election silence” periods before an election for certain types of content. Many attribute France’s election silence rules for minimizing the ability of hackers to undermine the French national elections in 2017. Social media platforms could announce that they will observe “election silence” periods and refuse direct candidate advertising exempt from fact checking for some period (such as a week) before the primary or general election day.
The advantage of this approach is it would cut off efforts to manipulate facts and press coverage in the last few days before an election, when opponents do not have the time to mount an effective counter-message and where the advantage of disseminating outrageous content that can grab press and public attention is highest. Even if the public generally disbelieves the specific content, an outrageous and untrue advertisement can shape press coverage and discussion in the last few days before the election, undermining the integrity of the process.
Run the advertisement with a disclaimer. Platforms can accept advertisements but make it clear if the advertisement has not been submitted to fact-checking. Alternatively, platforms can submit the advertising to fact-checking and run candidate ads whether they pass the fact check or not — but with a disclaimer saying whether the advertisement would violate the platform’s policy if it were not a direct candidate advertisement. Platforms could also run an intermediate approach. Platforms could allow candidates to voluntarily submit their direct advertisements to fact-checking, or carry a prominent disclaimer that the candidate chose to bypass fact-checking and that the content may therefore be false or deliberately misleading.
All of these approaches have the same problem. All these solutions invariably require platforms to make judgments that are hard to make consistently even by a single individual — let alone by a massive corporation with dozens of different review teams trying to apply a set of generally applicable guidelines. We can also expect that those seeking to undermine the political process will actively try to hack whatever rules platforms put in place, and that some percentage of these hacks will prove successful. At the same time, we will have numerous cases of good faith advertisements blocked, and high-profile cases where reasonable people (to the extent any reasonable people are still out there on social media) might disagree on how to apply whatever standards the platform adopts.
Until Congress steps up and does its job, however, platforms will continue to make consequential choices that potentially impact our ability to govern ourselves — in the certainty that they cannot possibly avoid controversy. But hey, no pressure.