Researchers from the University of Nebraska at Omaha (UNO), the University of Nebraska-Lincoln (UNL) and Rice University said while participation in electoral politics is affected by a host of social and demographic variables, there are also biological factors that may play a role.
Specifically, the paper points to low levels of the stress hormone cortisol as a strong predictor of actual voting behaviour, determined via voting records maintained by the US Secretary of State.
"Politics and political participation is an inherently stressful activity," said the paper's lead author, Jeff French.
"It would logically follow that those individuals with low thresholds for stress might avoid engaging in that activity and our study confirmed that hypothesis.
"It's long been known that cortisol levels are associated with your willingness to interact socially - that's something fairly well established in the research literature. The big contribution here is that nobody really looked at politics and voting behaviours before," said French.
"This research shows that cortisol is related to a willingness to participate in politics," he said.
Researchers collected the saliva of over 100 participants who identified themselves as highly conservative, highly liberal or disinterested in politics altogether and analysed the levels of cortisol found.
Cortisol was measured in saliva collected from the participants before and during activities designed to raise and lower stress.
These data were then compared against the participants' earlier responses regarding involvement in political activities (voting and nonvoting) and religious participation.
"Not only did the study show, expectedly, that high-stress activities led to higher levels of cortisol production, but that political participation was significantly correlated with low baseline levels of cortisol," French said.
"Participation in another group-oriented activity, specifically religious participation, was not as strongly associated with cortisol levels. Involvement in nonvoting political activities, such as volunteering for a campaign, financial political contributions, or correspondence with elected officials, was not predicted by levels of stress hormones," he said.
According to the study, the only other factor that was predictive of voting behaviour was age; older adults were likely to have voted more often than younger adults.
The study was published in the journal Physiology and Behaviour.