Anti-Obama viral email: Impact tested

by Mark Silva

This election campaign has seen no shortage of scurrilous and "viral'' chain email - you know, like the ones that suggest that Barack Obama is a radical Muslim or that he is unpatriotic, eschewing the flag pin and even allegiance to his nation.

The impact of all this remains unknown.

But, with the help of some students of political science, professors at Elmhurst College set out early this election season to gauge the potential impact of the email-campaign on voters.

They ran an experiment: They randomly assigned 363 students to double-blind tests. One group would receive a piece of "hate-mail.'' Another would get "second-hand exposure,'' such as reading about a hateful e-mail in a newspaper account.

Their preliminary findings: Voter-choice was not altered by exposure to the email, either firsthand or secondhand. Yet, still Obama's "feeling thermometer scores'' dropped in the condition where subjects read the email, and the temperature for him rose in conditions where they read a rebuttal from Obama's Web-site.

"Our results suggest that these types of personal candidate attacks can have an affect on citizen preferences,'' Professors Phillip Hardy and Mary Walsh write in a draft of their report graciously shared with the Tribune. "The preliminary findings from our study point to (1) decreased feeling thermometer scores toward the target candidate; (2) a diminished sense of empathy and morality associated with him; and (3) fear of Obama's religious background.

They've also concluded that "this type of communication (i.e., chain emails) is likely to continue in future campaigns, and it does not appear to be benign or meaningless,'' in the words of Hardy, who will present the study with Walsh at the Illinois association of political scientists after the election.

"In the political science literature,'' Hardy writes to us, "there has been extensive work on evaluations of presidential candidates, political misinformation, negative campaigning, and character attacks, but no one to our knowledge has looked at the effects of these types of chain emails on citizen preferences and/or voting....So we focused on this relatively new medium, the Net, and the specific phenomenon of chain emails.''

The study found that those who read the Obama smear-mail voiced concerns about Obama's religion and a stronger desire to have "Judeo-Christian candidates" run for high office, Hardy notes.

"We think these preliminary results point to (1) citizen vulnerability in some cases to these types of rumors and messages, (2) the strong need for campaigns to aggressively respond to rumors - early and often, as has been the case with Obama's campaign, and (3) the need for further study of this type of political communication.''

Some limitations to this study, perhaps?

"The most glaring is that our sample consisted of a majority of Democrats, so the effects could have been more pronounced if we had more Republicans in the sample.'' Hardy allows. "However, we think it's somewhat impressive that we've
observed these findings among a sample of mostly Democrats.''

In the paper that they are preparing, Hardy and Walsh write:

"Scholars are only beginning to understand the transformative potential of the Internet in the politics of the United States and throughout the world. Writing in the 1990s, the political scientist W. Russell Neuman anticipated a new mass media technology that would "blur the distinction between mass and interpersonal communications"

The Internet is not only ripe for manipulation, they note, but also a potent political vehicle.

"The Internet gives citizens virtually unlimited access to information about candidates, policies, government spending, current events, and more. What is more, it provides opportunities for political discourse, including blogging and social networking, as well as communication via email messaging.

"In the U.S. one of the first tangible signs of the power of the Web was manifest in Howard Dean's bid to win the Democratic nomination for the presidency in 2003. Dean and his chief strategist, Joe Trippi, tapped into a groundswell of anti-war activists and critics of the Bush administration who set up shop online and began hosting online chats with candidate Dean and his supporters on the political left. Dean's campaign used these forums to gather an unprecedented amount of money - nearly $50 million dollars in campaign contributions, many as small donations - which allowed Dean to capture the attention of the mainstream media and the electorate.''

Yet its potential for mischief is huge.

"A widely circulated chain email titled "Who Is Barack Hussein Obama?" begins and ends with a plea for the recipient to forward the message to "all you know" and purports to convey 'facts' about Barack Obama's upbringing, religion, and patriotism,'' Hardy and Walsh write. " The forwarded email alleges that Obama "takes great care to conceal the fact that he is a Muslim."

"The unidentified authors go on to describe Obama's (1) familial ties to radical Islam; (2) education in a Wahabi school in Jakarta, Indonesia; (3) refusal to recite the Pledge of Allegiance; and (4) disregard for the American flag. The first of these chain emails (there have been several variants, all with the same theme) was distributed as early as 2006, but these types of messages gained momentum and started bouncing around the Net as "pass alongs" during the 2007 Democratic primaries.

"Both the mainstream press and new media sources (e.g., online fact checking groups, liberal and conservative blogs) began investigating the claims presented in the emails. Barack Obama's campaign also aggressively denounced the emails and constructed a "facts" page on Obama's website in order to refute lingering rumors. Despite his campaign's attempt at "rapid response" to these personal attacks, Barack Obama's candidacy continues to be plagued by smears and innuendo. For example, at a town hall meeting for the Republican nominee for president, John McCain, held in late October 2008 a supporter declared "I can't trust Obama. I have read about him and he's not, he's not...he's an Arab"

"The 'Obama is Muslim' cyber-rumors communicate the sense that Barack Obama is anti-Christian and un-American, and as such is a threat to American values and the American way of life....

"Barack Obama has faced similar "innuendos, shruggings and hunchings" about his faith and patriotism, but the mode of communication is far more sophisticated than the forms utilized by foes of Smith or McCain. The electronic transmission of the Obama rumors offers the potential for endless forwarding and replication, but also provides anonymity for the source or sources of the misinformation. These "cybercascades" are particularly elusive and difficult to stop.

"As Obama explains, "We have no way of tracing where these emails come from, but what we know is they come in waves, and they somehow appear magically wherever the next primary or caucus is." And, as the case of Obama indicates, the rumors persist despite concerted efforts to correct the misperceptions by the Obama campaign and others.

"A Pew Research Center study in June 2008 found that 12% of registered voters believed Obama was Muslim, up from 10% in March 2008, and that 25% of respondents stated they didn't know about Obama's religion. This poll also indicated that the impact of the rumors extended beyond conservative circles, as concerns about Obama's religion were observed among both Democrat and Republican likely voters

These are the questions they pursued:

Research Question 1 (RQ1): Will reading a negative chain email about a presidential candidate affect citizens' (1) attitudes about and preferences for the candidate; (2) vote choice; or (3) tolerance toward the group or object identified as dangerous in the message?

Research Question 2 (RQ2): Will firsthand exposure to the message be more persuasive than a secondhand (e.g., reading a newspaper article that describes the email) encounter with information about the message?

Research Question 3 (RQ3): Are campaign counterattacks or rebuttals effective at diminishing the persuasiveness of fear appeals disseminated via forwarded emails?

Accordingly, we have proposed several testable hypotheses for the present investigation:

Hypothesis 1 (H1): Fear-inducing email messages that personally attack a presidential candidate will persuade the recipient of the message to vote for a different presidential candidate.

Hypothesis 2 (H2): Recipients of emails smearing a presidential candidate will have colder feelings for that presidential candidate and warmer feelings for their competitor(s).

Hypothesis 3 (H3): Recipients of emails smearing a presidential candidate will give lower evaluations of the candidate regarding a number of presidential traits (i.e., morality, caring, knowledgeable, leadership, honesty, and intelligent).

Hypothesis 4 (H4): Recipients of emails smearing a presidential candidate will be less tolerant of the objects or groups identified as dangerous in the message.

Hypothesis 5 (H5): Firsthand exposure (i.e., reading the email) will be more persuasive than will secondhand exposure (e.g., reading about the email in an article) to information about the message.

Hypothesis 6 (H6): Campaign counterattacks and rebuttals will effectively cancel out the negative affects of firsthand or secondhand exposure to the fear appeal (i.e., the chain email).

The political scientist put 363 subjects to work on a study between late February and mid-April 2008 -- 197 randomly assigned to Experiment 1 (4 conditions) and 166 randomly assigned to Experiment 2 (3 conditions). All of the subjects were undergraduates at a small liberal arts college in Chicago's western suburbs. about political learning and opinion formation.

They found:

"The most pronounced findings have to do with subjects' preferences for Senator Obama. In experiment one, we observed a statistically significant difference across conditions two (reading the email alone) and four (reading the campaign's rebuttal alone...

"Reading the email resulted in a drop in feeling thermometer scores for Senator Obama, while reading the rebuttal alone from the campaign (i.e., a rebuttal to the rumor) caused an increase in feeling thermometer scores for Obama. Subjects who read the campaign's response to the Internet rumors appears to have caused individuals to express warmer feelings about Obama than those people who read the chain email....''

In the second experiment, they looked at "secondhand" exposure to the Obama email.
"Interestingly, more than 250 of 363 subjects indicated they had heard something about the Obama rumors, just 55 of 363 subjects stated they had personally read the email (this question was asked on the final page of the posttest survey). Thus, it seems that many people have gotten bits and pieces from the mainstream press about the Obama rumors, but only a small percentage, about 15%, personally received this viral email or something like it.

"Looking specifically at effects on presidential candidate preferences or vote choice, as stated above we do not find any evidence that secondhand exposure to the rumor impacted vote choice. Additionally, there were no statistically significant findings on feeling thermometer scores, presidential traits (e.g., honesty, morality, caring, knowledgeable), or tolerance statements.

"There is one qualification, however - reading about the rumors regarding Barack Obama's patriotism and religious background in the context of a three candidate profile article did have an effect on high knowledge subjects (an eight point political knowledge battery was administered in the pretest), and the relationship approached statistical significance. High knowledge subjects in condition five (profiles of the three candidates, with a paragraph length insert emphasizing the rumors inserted into the Obama profile) had lower feeling thermometer scores than did low knowledge subjects in the same conditions or the remaining subjects in conditions six and seven. We discuss the possible meaning of this finding in the following section of this paper.''

In summary: "Our results suggest that these types of personal candidate attacks can have an affect on citizen preferences. The preliminary findings from our study point to (1) decreased feeling thermometer scores toward the target candidate; (2) a diminished sense of empathy and morality associated with him; and (3) fear of Obama's religious background.

"These findings are restricted to conditions associated with firsthand exposure to the attack email, and there are qualifications to be considered... However, we believe they provide evidence, albeit limited, of the power of this type of communication in the political process. ''

And they suggest:

"Campaigns would be well-served to combat early and often these types of Internet-based smear campaigns, from politically insensitive and hateful cartoons, to "urgent" messages passed about via electronic messaging, and to look out for early signs of these types of personal slurs.
"In this vein, one of the findings from our first experiment, namely that thermometer scores for Obama increased among subjects who read the "facts" page created by his campaign, and that subjects in this same treatment felt agreement with statements that Obama "cares" and is "moral," indicates that this type of campaign correction may actually shore up support for the candidate. ''