The Media Research Center believes Google "does change its algorithm to benefit certain political candidates," and is strictly focused on the Senate after a new study proved its "search bias" didn’t carry over to House races where the GOP is already favored.
"This is more evidence that Google deliberately manipulated data for the Senate races. Google needs to explain itself and stop trying to manipulate elections," MRC founder and president Brent Bozell told Fox News Digital.
Earlier this week, the conservative MRC unveiled its findings that focused on 12 Senate races identified as important to watch. It revealed campaign websites for 10 Republicans among the 12 tight races were found significantly lower on results pages compared to their opponent’s sites among organic search results.
Seven Republican candidate websites were completely hidden from the first page of Google search results, according to the MRC. While Republicans were hidden, eight Democratic candidates had websites that were promoted within the first six results on Google.
Google denied any wrongdoing when reached by Fox News Digital after the original MRC report on Senate candidates.
"This report is designed to mislead, testing uncommon search terms that people rarely use. Anyone who searches for these candidate names on Google can clearly see that their campaign websites rank at the top of results - in fact, all of these candidates currently rank in the top three and often in the first spot in Google Search results," a Google spokesperson told Fox News Digital.
In another statement, a Google spokesman said the claims by the MRC continued to be untrue.
"Earlier this week, this organization said we manipulate search results to favor one party -- and now, just two days later, they're saying we favor the other. Neither claim is true. As we have said, this misleading analysis is testing uncommon search terms that people rarely use. We don’t and would never manipulate search results in any way to promote or disadvantage any particular political ideology, viewpoint or candidate – and third party research consistently shows that our results are not politically-biased," the spokesperson said.
Google cited third-party studies in The Economist and from Stanford University that found its search results aren't biased based on political viewpoints. Stanford's study said Google's search analysis emphasized so-called authoritative sources.
Google also said that, regarding the question of why campaign sites didn't always show up high in searches, its algorithms interpret the intent behind the search term "Senate race" as someone who is looking for the latest news on a certain contest, so news articles frequently appear up top.
The MRC’s Free Speech America said it applied the same methodology from its Senate study to 36 House races, where polling indicates control doesn’t hang in the balance.
"Amazingly, somehow Google showed no bias against Republicans in House races," MRC associate editor Brian Bradley wrote. "If the search terms used in the Senate study were flawed as Google says, why did they work fine with no indication of bias in the House study?"
The MRC believes that Google is "targeting the Senate" because polling indicates the Republicans will take control of the House anyway.
"MRC Free Speech America’s methodology was not only correct, but when comparing Google’s search results with Bing and DuckDuckGo, Google’s search bias becomes even more clear," Bradley wrote.
"Google buried 10 of 12 Senate Republican Party candidates’ campaign websites while highlighting their opponents' campaign sites in organic search results," Bradley continued. "Meanwhile… Google placed Republican campaign websites higher in search results than Democrat campaign websites by a 21-11 margin in House races, while giving four races equal treatment."
The initial races analyzed were the "Top Senate Races" on Oct. 7, 2022, as determined by RealClearPolitics, which includes key contests in Arizona, Colorado, Florida, Georgia, Missouri, Nevada, New Hampshire, North Carolina, Ohio, Pennsylvania, Washington and Wisconsin.
The House study looking at 36 races based on the "Toss Ups" list from RealClearPolitics yielded small, predictable variance between Republicans and Democrats running for House seats across the search engines of Google, Bing and DuckDuckGo, according to the MRC.
The House "Toss Up Races" included the Democratic Party and Republican Party candidates from Alaska, Arizona, California, Connecticut, Georgia, Illinois, Indiana, Maine, Maryland, Michigan, Minnesota, Nevada, New Hampshire, New York, North Carolina, Ohio, Oregon, Pennsylvania, Texas, Virginia and Washington.
"Google apparently does change its algorithm to benefit certain political candidates. But the two studies suggest it’s targeting the Senate," Bradley wrote.