Should Google's Search Results Be Protected by the First Amendment?

Algorithmically generated editorial judgment is still judgment.

Wikimedia Commons

Is Google using its power as the number-one search engine to promote its own products such as Google Places and Google Maps? That was the question at hand last fall when Google executive chairman Eric Schmidt testified in Congress. "I see you magically coming up third every time," Senator Mike Lee of Utah said. "I don't know whether you call this a separate algorithm or whether you've reverse engineered one algorithm, but either way you've cooked it, so that you're always third."

Schmidt responded, "Senator, may I simply say that I can assure you we've not cooked anything."

Schmidt's sounded definite but it didn't really answer anything -- what precisely would qualify as cooking the results? Google could claim not to "cook" anything, but it's indisputable that it writes the algorithms that underlie its search tool, and that those algorithms include choices -- judgment -- that whittle the Internet's billions of pages down to a ranked list of 10.

That process -- that judgment -- is now at the heart of an argument in defense of Google against the accusations of anti-competitive behavior. Coming from Eugene Volokh, the UCLA law professor and prominent blogger, is a paper that makes the case the Google's search results are an editorially curated product, not different from a curated list of likes like the Drudge Report, multiplied ad infinitum. And, as the Drudge Report would be protected by the First Amendment, so should Google search results. With this argument, Google (who funded Volokh's paper) takes the argument against the company and turns it on its head: You think we are not being "neutral" in our search results? Damn right we are not being neutral. And it's for exactly that reason that you can't do anything about it.

The idea of a "neutral" or "unbiased" search was always a bit specious. If Google can provide the same information -- or better information -- from within its own tool shed, why should it not? Certainly some information is objectively better than others. If you doubt this, try this search engine which does not return results from the top one million websites. You'll find that the results are pretty terrible -- junk from paid content sites, reprints of Wikipedia articles, not-very-funny Tumblrs and so on. The results remind you just how much muck is out there, and just how good Google is at finding the needles in the Internet's haystack time and time again.

But at the high-quality end of the Internet's curve -- how do you sort and rank the very best information? What if the information returned by two sites -- Google Places and Yelp, for example -- is nearly identical? Those decisions are judgment calls, coded into Google's algorithm by humans. Not neutral, not the unbiased calculations of a machine, no matter how it works in a given instance. Volokh's paper rests on this idea (he uses the word judgment 34 times) that in exercising judgment, Google's engineers are essentially acting as editors, curators, or, even, parade organizers -- all of whom the First Amendment protects in their decisions to include or exclude content, even when they themselves are not the creators of that content.

There's a lot of support for Volokh's argument including two lower court decisions (2003 and 2007), and, as First Amendment and technology law expert Marvin Ammori argues, other courts -- and even the Court -- would likely agree. The result would be greater protection for Google and its preference for its own products -- something we may not like. But the First Amendment has never been interested in curating society to our liking -- quite the opposite in fact. The results of a strong First Amendment are often distasteful in varying degrees, with hateful speech at the extreme end. But the converse is much worse -- would we really want the government to have a say in the content of Google's returns? Could you imagine what it would like to do with something like this?

The law is always under revision as new technologies emerge and challenge the old categories we had created. Is Google like a publication -- such as the New York Times -- or a utility like the gas company that merely conveys information "neutrally"? Which set of laws should we apply? These comparisons never work perfectly, and refining their raggedy edges is the work of the courts and the participants in their adversarial process. In the case of Google, its search results do seem more like the handiwork of a newspaper editor or a parade organizer than an electrical utility. But the impact of its choices -- judgment, if you will -- are so much greater, so much more central to our civic life that it can be scary to give it such free reign. But that free reign is at the core of our grand experiment with free speech and a free press, an experiment you just have to hold your breath and hope for, because the alternative is much, much worse.

About the Author

Rebecca J. Rosen is a senior editor at The Atlantic, where she oversees the Business Channel. She was previously an associate editor at The Wilson Quarterly.

A Washington Post report on 2016 election interference raises the question: What could Obama have done differently?

If there is one thing The Washington Post’s story on the Obama administration’s anemic response to Russian meddling in the 2016 election makes clear, it’s that it took two to make the meddling effective.

There is a reason the tactics Russia used on the American elections—which are similar to things they’ve done in former Soviet republics and in Europe—are referred to as “asymmetric warfare”: They embody the art of leverage, of doing a lot with a little. As former Director of National Intelligence James Clapper told Congress in May, the Russians “succeeded beyond their wildest dreams and at minimal cost.” The whole operation, according to Clapper, cost a mere $200 million—a pittance in military spending terms. But the Russians used that money not the way a conventional army would, but the way a band of guerrillas would, feeling around for pressure points, and pressing—or not. Though, as Bloomberg reported this month, the Russians were clearly exploring ways to attack voting infrastructure in parts of the country, it still appears they ultimately decided not to pull the trigger, sticking instead with the hack-and-dump and the manufacturing of fake news. “It was ad hoc,” an Obama administration official told me shortly after the inauguration. “They were kind of throwing spaghetti at the wall and seeing what would stick.”