He said the goal of the Sunlight Society was "to detect, to study and to expose new technologies that threaten human freedom."
Three of the 12 academics publicly supporting the Sunlight Society, when contacted by The Washington Post, said they were uncertain about the sources of bias but favored creating a group for studying the issue because search engines had become dominant sources of online information. Results often vary from person to person based on location and other factors, often including browser and search history.
"One of the questions we are still grappling with is, where is the bias coming from?" said one of Epstein's supporters, Natali Helberger, a professor of information law at the University of Amsterdam. "Part of this bias is probably in ourselves."
Google dismissed Epstein's research as "nothing more than a poorly constructed conspiracy theory."
"We have never re-ranked search results on any topic (including elections) to manipulate political sentiment," the company said in the statement to The Washington Post. "Moreover, we do not make any ranking tweaks that are specific to elections or political candidates, period. We always strive to provide our users with the most accurate, relevant answers to their queries."
Microsoft, which operates the Bing search engine, and Yahoo both declined to comment.
The importance of search rankings - and especially their order as they appear on the screen of a computer or other device - is well established when it comes to how consumers select online shopping offers, websites or news stories. Epstein summarized research into the impact of search results on political behavior in the Proceedings of the National Academy of Sciences in 2015.
University of Maryland law professor Frank Pasquale, who in 2007 publicly raised the possibility of a "Federal Search Commission" to monitor the increasingly powerful effects of search engines, said a monitoring group is a valuable step. "It's part of an ecosystem of accountability," said Pasquale, who is a public supporter of Epstein's project.
Dennis Allison, a lecturer at Stanford's Computer Systems Laboratory and a supporter of Epstein's project, also called the study of bias in new technologies important. "The source of that bias has not been established as far as I can see but likely is present," he said.
In Epstein's study on the 2016 election, a summary of which is scheduled to be presented at an international psychological conference in Vienna this month, bias was most prominent in results displayed to people who had already made up their minds about how to vote. Results in Democratic-leaning states were also more biased, Epstein said, with the lowest measures of bias in swing states.
Panagiotis Metaxas, a Wellesley College computer science professor who has studied the workings of search engines but has no affiliation with Epstein's project, said the biggest risk of ranking bias comes from spammers or political activists who seek to manipulate results - not from the search companies.
He questioned whether Epstein, based on his writings, will focus the Sunlight Society on this target, as opposed to scrutinizing Google and other tech companies.
"The battle is elsewhere," Metaxas said. "But overall I think it does not hurt to have such an organization."