“As researchers, we were given a live election with two candidates that in many ways were almost caricatures of these two academic concepts of how you gain influence,” said Niro Sivanathan, an associate professor at London Business School who co-authored the paper with a colleague.
Sivanathan said in an interview that although past researchers have looked at the preference people have for strong leadership when they feel they have lost control, his was one of the first to link his studies to this theoretical framework, which describes leaders as trying to gain power through displays of dominance or through “prestige” — gaining the admiration of followers through experience and success in the associated field.
Also unlike previous studies, Sivanathan said, his research linked economic data at the Zip code level to a sense of economic anxiety and a preference for strong leaders, rather than just experiments that manipulate feelings of uncertainty in a lab.
“The main takeaway is that whenever individuals feel a lack of control in their environment or their lives,” he said, they believe that “having a dominant leader is likely to help them regain that control.”
The research paper includes several studies, the first of which looked at voters' preference for Clinton or Trump and several questions about who they saw as a more dominant or prestigious leader just before the third debate. It compared that with the economic conditions in their individual Zip codes, controlling for issues such as political orientation, gender, age and income. “If people were living in a Zip code where the poverty rate, unemployment rate and housing vacancy rate were high, they showed a greater preference for voting for Trump,” Sivanathan said.
Again and again, Trump displayed a particularly dominant — even autocratic or authoritarian — concept of leadership while on the campaign trail. He said he would return the use of waterboarding and advocated for the use of torture. He commanded rally attendees to “get them out” when a protester was mobbed by a crowd. He said he thinks he has the best temperament, or one of the best of “anybody that’s ever run for the office of president” because “I know how to win.” At the Republican National Convention, he vowed that “I alone can fix it.”
To broaden the study beyond Trump and Clinton, Sivanathan and his colleagues also asked respondents about a hypothetical local election, asking voters in various Zip codes whether they preferred a candidate described as more dominant or one who was most admired and less forceful. Again, Zip codes with more economic uncertainty preferred the more dominant leader. A third study used global data maintained by the World Bank that finds a similar effect at the global level.
The research, Sivanathan said, is a reminder that yes, candidates matter — the style and approach of the individual people who run, not just the political party, are critical — but so does context.
“Regardless of the party, if people are feeling they don't have control or things are not certain, these are the types of leaders [voters] are going to prefer,” he said.
“The conditions were one where they felt very upset — and then along came the candidate who had all the rhetoric and all the behavior of someone who, psychologically, they believed would be the one to get them out of this state.”