Who or what rigs elections? Sneaky algorithms, or unskeptical voters?
A Facebook comment has decided to become its own blog post. Maybe it’s trying to boost its own search ranking. (Yeah, I question that logic too.)
Happy Friday. A vocal critic of the ubiquitous search engine Google has published an article and a study that claims the technology firm may be influencing the outcome of elections by changing the order in which search results appear when users search for candidates.
The study imagines a phenomenon called Search Engine Manipulation Effect (SEME) and says, in summary: “In laboratory and online experiments conducted in the United States, we were able to boost the proportion of people who favored any candidate by between 37 and 63 percent after just one search session.”
I have neither the time nor the tools to refute the study’s merits, but let’s just take one step back. Even if it were true that a search tool could (and would) sway public opinion about candidates via search result ranking, the woolly mammoth in the room is that people’s favorability meters are moved at all, about anything, by search result ranking!
The right to vote in free and fair elections is a hard-won, precious thing. It is not to be taken lightly. Citizens who exercise the franchise have the responsibility to examine their choices with the utmost care. We should summon every ounce of critical thinking we have, and use it.
Google’s refutation of the study’s allegations mentions civic tools that they have helped create and provide to the public. I’ve examined and used* these tools, and they are very helpful towards the ideal that an informed public uses unbiased information to choose its government.
But if the public chooses to rely simply on the top three or so results from lazy web searches, then whether or not those search results are (unwittingly or otherwise) manipulated matters not when it comes time to lay blame.
The aforementioned Facebook comment:
Since Google makes available an API with a trove of accurate election data that developers can use to present to users in whatever ways they choose, they’d have to be really cocky to still rely on the approach posited in Epstein’s article, supplanting their own “altruistic” endeavors in the meantime. It’s not impossible, just unlikely. The bigger issue is that internet users are uncritical when choosing from search results, no matter how or why those results are ranked. Sure, that behavior may influence how they view the candidates returned, but what’s the real root of that problem? It’s not the search engine.
*The app I attempted to build in 2014 did not get completed in time for the August elections. I plan to use the API and other tools more in the coming months to build a resource for Tennessee voters in 2016 and beyond.