RNC loses complaint claiming Gmail spam filter is biased against Republicans

The Federal Election Commission (FEC) has rejected a Republican National Committee (RNC) complaint claiming that Google violated US law by using Gmail’s spam filter on Republican campaign emails. Republicans had claimed Gmail’s spam filtering amounted to “illegal in-kind contributions made by Google to Biden For President and other Democrat candidates.”

But an FEC decision last week, which Google provided to Ars today, said the commission found “no reason to believe” that Google made prohibited in-kind corporate contributions. The FEC, an independent agency of the US government, also found no evidence that the Biden for President campaign committee knowingly accepted illegal in-kind contributions in the form of spam filtering preference.

The FEC told Google in a letter that it has “closed its file in this matter” and that documents related to the case will be placed on the public record within 30 days.

“The Commission’s bipartisan decision to dismiss this complaint reaffirms that Gmail does not filter emails for political purposes,” Google said in a statement. “We’ll continue to invest in our Gmail industry-leading spam filters because, as the FEC notes, they’re important to protecting people’s inboxes from receiving unwanted, unsolicited, or dangerous messages.”

FEC: Study cited by Republicans doesn’t prove bias

Republicans made their case in a joint complaint filed in April by the RNC, National Republican Senatorial Committee (NRSC), and National Republican Congressional Committee (NRCC). “By overwhelmingly and disproportionately suppressing emails from Republican candidates, including President Trump and others, Google used its corporate resources to provide a massive service to their Democrat opponents by denying Republican candidates the same ability to communicate with voters,” the complaint said.

The RNC complaint cited a North Carolina State University study that found “Gmail marks a significantly higher percentage (67.6 percent) of emails from the right as spam compared to the emails from left (just 8.2 percent).”

But the FEC decision said that “the NCSU Study does not make any findings as to the reasons why Google’s spam filter appears to treat Republican and Democratic campaign emails differently.” The FEC also described several limitations in the study:

First, the NCSU Study observed a limited time period, and the authors note that they “cannot make inferences about the behaviors of the SFAs [spam filtering algorithms] in general. Second, a number of variables were outside the NCSU Study’s control, including how many emails the individual campaigns sent; how long the campaigns had been active; and how many users outside of the study group had marked a specific email as spam. Thus, the study authors conclude, they “cannot determine the extent to which these factors influence spam percentages.” Finally, the authors note that their study selected for particular variables, and that other variables may have other effects not shown in the NCSU Study. They state that “we have no reason to believe that there were deliberate attempts from these email services to create these biases to influence the voters.”

The FEC decision noted that Google’s response “contends that the NCSU Study was small, analyzing only 34 recipient email addresses when Gmail has 1.5 billion users, and over ‘a mere 153 days, 27 of which were post-election.'” Google also responded that “only four Gmail accounts were included in each of the Read, Spam, and Inbox Groups,” that “the emails sent by Republican and Democratic campaigns were not identical, and that there are multiple factors that impact the action of Google’s SFA that were outside the authors’ control,” the FEC said.