Suchergebnisse

AbstractThis report examines accuracy and bias in national and state‐level preelection polls conducted during the 2020 U.S. general election cycle. Overall, national and statewide polls in 2020 were considerably less accurate than in 2016. In fact, there is evidence the final, national presidential preelection polls were less accurate than in any cycle since 1996. Polls across the board underestimated Republican support in the presidential, U.S. Senate, and gubernatorial races. The results show the pro‐Democratic biases reflected in the final 2020 preelection polls were systematic.

Individuals behave differently in the presence of others. Some scholars argue the mere presence of others has the capacity to improve task performance, while other researchers demonstrate individuals become "social loafers," exerting less effort in group tasks unless they perceive potential for evaluation. I investigate these claims in the context of evaluating poll‐worker performance using two randomized field experiments conducted in New York City during a special election that took place in February 2009 and during the presidential primary conducted in April 2016. The results suggest that manipulating the presence of observers, or even poll workers' perceptions of the potential for evaluation, likely enhances their performance overall, resulting, especially, in greater efficiency and potentially strengthening voters' confidence that their ballots were accurately counted and diminishing perceptions of electoral fraud.

Scholars and practitioners alike are increasingly concerned about growing and differential rates of nonresponse in exit polls. In this paper, I examine how exit poll sponsorship affects response intentions. I leverage an experiment imbedded in a nationally-representative probability sample telephone sample that probed respondents about exit poll participation intentions. The survey, fielded by Opinion Research Corporation for CNN in November 2006, randomly varied whether respondents were informed the exit poll would be conducted by the "television networks," specifically, or generically, by "some organizations." The results reveal that respondents are significantly more likely to report they would be likely to participate in exit polls when television networks are explicitly identified as sponsors. These effects are robust across a series of specifications and do not appear to be moderated by key observable attributes including age, gender, race, partisanship or educational attainment.

The current study presents evidence that partisan campaign strategies in US presidential elections have shifted in recent years to reflect a growing emphasis on base mobilization compared to persuading independent, undecided or swing voters. Pursuing reliable supporters appears to be increasingly desirable for political campaigns, especially since 2000, relative to the risky and uncertain prospects of targeting less reliable supporters. I speculate this shift is attributable, at least in part, to changes in the context in which contemporary campaigns operate, coinciding with the 2000 presidential election; advances in microtargeting, particularly through new e-campaigning and e-mobilization technologies, and behavioral science related to voter mobilization represent two of the most potent developments. By contrast, persuasion is a far more difficult and risky approach. I conclude that growing emphasis on base mobilization in elections has likely contributed to intensifying partisan polarization in America.

Implicit social pressure, applied via exposure to eyespots in nonpartisan, direct‐mail blandishments to vote, has been shown using randomized field experiments to raise turnout in elections. Similar eyespot effects have been observed across a wide range of prosocial behaviors. A series of recent replications conducted by Matland and Murray (2015) have failed to consistently produce statistically significant eyespot effects on voter turnout, however, leading the authors to conclude the effects observed in previous research were likely illusory. In this article, I rebut this claim, arguing that an alternative, more circumspect interpretation of the authors' key results points to a different conclusion that supports the notion that eyespots likely stimulate voting, especially when taken together with previous findings.

Explicit social pressure has been shown to be a powerful motivator of prosocial behavior-like voting in elections. In this study, I report the findings of a randomized field experiment designed to study the impact of more subtle, implicit social-pressure treatments. The results of the experiment, conducted in the October 2011 municipal elections in Key West, Florida, demonstrate that even subtle, implicit observability cues can effectively mobilize citizens to vote, perhaps as much as explicit surveillance cues. The findings speak more broadly to our understanding of human decision making, and even evolution, and provide fodder for the claim that humans are evolutionarily programmed to respond to certain stimuli. I interpret the evidence to support the notion that evolutionarily charged impulses, like exposure to images that implicitly signal the potential for surveillance and observability, are sufficient to overcome powerful collective action incentives to free ride. Adapted from the source document.