Millions turn to the online encyclopedia for information about the presidential candidates, and page views may even have predictive power.

Share story

For a website with no paid writing staff that is still overcoming an out-of-date reputation for inaccuracy, Wikipedia punches above its weight. As a primer for just about any topic, it is especially powerful in an election season: On the day of the 2012 election, Barack Obama’s and Mitt Romney’s entries alone were read 1.6 million times.

The online encyclopedia famously allows the public to edit it, but it also publishes reams of data about itself: about what articles used to say, who added or deleted passages and how many people read the articles. And if you know where to look, that data can provide a window into what the public is saying about the 2016 presidential candidates.

In the Wikipedia primary, it is not much of a contest. On any day, Donald Trump’s entry usually attracts more views than those of his Republican rivals — and on some days, more than all of them combined. On the Democratic side, Bernie Sanders gets more attention than Hillary Clinton. Even Martin O’Malley outpaces Clinton on debate nights.

Page views are not votes, of course, but they could have some predictive power. In each of the first five Republican contests in 2012, the candidate whose Wikipedia page received the most views on voting day (but before winners were announced) got the most votes. The pattern was not as strong in 2008, when two candidates drew outsize attention, garnering the most page views even on days when they got fewer votes: one was Ron Paul, the other, Obama.

While most readers will not notice every edit, each article is a palimpsest of added or deleted nuggets, often several in a single day, which are archived under the “view history” tab. Some are crude or juvenile smears that are quickly removed by hawk-eyed contributors on the lookout for pranks.

On Sept. 19, for example, the name of Ted Cruz’s spouse was changed to “Edward McMahon,” then changed back to “Heidi Nelson” six minutes later.

But other edits reflect substantive disputes that are often debated on Wikipedia’s discussion pages (the “talk” tabs) before they make it to newspapers, mainstream blogs and TV shows.

Cruz’s birthplace — Calgary, Alberta — became a major campaign issue last month after Trump questioned whether his rival would be eligible for the presidency. But mentions of his Canadian birth have been added to, modified or deleted from the page at least 636 times dating to 2009, before Cruz had even run for the Senate. Eventually, the fact led the second paragraph of Cruz’s entry.

Until recently, the struggles with drug addiction of Jeb Bush’s daughter, too, were not widely covered in the media out of deference to the family’s private struggles. But on Wikipedia, details of those struggles had been added or deleted at least 90 times over many years.

One Wikipedia editor who removed a passage about Bush’s daughter, Noelle, wrote that “these are little more than attacks.” But another complained that the removals were “a rather pathetic attempt at propaganda, the sort of thing the Stalinists used to do, trying to whitewash something out of existence. In this digital age, it’s absurd.”

The latter opinion prevailed, once Bush himself brought it up, and it has remained on the page since.

Wikipedia’s rules generally forbid anyone with a “conflict of interest” from editing pages, but this rule is difficult to enforce because editors are almost never identified by their real names. Instead, editors are identified by user names, which are generally untraceable, or, if they log on anonymously, by their numerical Internet protocol, or IP, address.

In 2015, a British member of Parliament was accused of sprucing up his own page while adding derogatory information to the articles about his rivals. Also last year, someone using the New York Police Department’s network apparently scrubbed the pages about the department’s darker moments, including the deaths of Eric Garner, Sean Bell and Amadou Diallo. In 2006, several congressional staff members were found to have edited their bosses’ pages. Now, a Twitter account automatically keeps track of any page edited from congressional buildings.

It probably does not flatter Bush that you can read about his 1980s-era work with Miguel G. Recarey Jr. Recarey, then a medical entrepreneur in Florida, is now a fugitive after being indicted on charges relating to defrauding Medicare.

A passage about the episode was deleted from Bush’s Wikipedia page three times in 2012. Whoever deleted the Recarey passage also deleted a portion claiming that Bush “interceded” with his father, who was president at the time, to release Orlando Bosch, an anti-Castro militant who had been accused of terrorism in Cuba and was in danger of being deported. (The younger Bush, among others, did lobby his father to free Bosch, a hero in Miami’s Cuban-exile community, and eventually the Justice Department released him from detention.)

In 2012 and 2013, a contributor from the same IP address also deleted passages about Bush’s daughter’s drug addiction and rewrote the section on Bush’s education policy. Who was making those edits?

The Internet connection for the account demonstrated that whoever made the edits was from the Foundation for Florida’s Future, a nonprofit group that promotes certain education policies championed by Bush. Bush founded the organization and, until 2014, served as its chairman and president.

Asked about the edits, Jaryn Emhof, a spokeswoman for the foundation, said only, “The page was reviewed for accuracy.”