Search

This study examines the effect of telephone survey dialing patterns on lab productivity and survey responses. Using an original data set of paradata from 2010 to 2017 and a machine learning technique for variable selection, we find that early and late afternoon shifts are as productive as late evening shifts for both landline and cellphone Random Digit Dialing (RDD) samples. Also, early weekdays are more productive than the weekend for the cellphone RDD samples. Most importantly, time of the interview affects survey responses; therefore, survey practitioners and scholars should be cognizant of this effect when scheduling calls.

How did the pandemic impact turnout of young voters living in university communities? Leveraging the mandatory vacating of Florida college students living on campuses and drawing on administrative data from Florida's voter file, we argue that on-campus registered young voters who had to leave their university housing in the days prior to Florida's 2020 Presidential Preference Primary (PPP) were less likely to turn out compared to adjacent off-campus young voters because they lost the opportunity to cast early in-person and Election Day ballots. Using a difference-in-differences (DiD) design, we find that on-campus students, in part because they had early and Election Day voting available to them on campus in the 2020 general election, were more likely than comparable off-campus student-aged registered voters to cast ballots in the November election. Our study has important implications for academic debates concerning the turnout effects of convenience voting reforms and the ability of voters to cast ballots prior to Election Day. [ABSTRACT FROM AUTHOR]

Recent partisan claims about the illegitimacy of the 2020 election highlight a need to better understand the determinants underlying Americans' trust in the electoral process. In this study, we focus on African Americans and conservatives, two groups that stand out both historically and contemporaneously for high levels of distrust in elections. Utilizing nationally representative survey data, we analyze the degree to which election distrust is associated with respondents' attitudes towards policies addressing voter access (photo ID requirements, vote by-mail, and felon voting), perceptions of disenfranchisement, and their personal experience while voting. We find evidence that distrust is not tied to one's personal voting experience, but rather to one's policy attitudes towards electoral access. Importantly, for conservative and Black voters, the policy remedies that would lead to increased trust for one group would only further exacerbate the concerns of the other, suggesting that distrust towards American elections will be difficult to attenuate. [ABSTRACT FROM AUTHOR]

Are more politically knowledgeable registered voters more likely to cast their ballots prior to Election Day when given an option to do so? We argue that individuals with high political knowledge are more likely to take advantage of convenience voting opportunities because they have command over static-general facts, enabling them to make informed choices when voting. Drawing on five Cooperative Congressional Election Study (CCES) pre- and post-election national surveys and an original survey of registered voters in Florida conducted in 2017, we test if political knowledge affects the method of voting. Estimating a series of multinomial logistic regression models with fixed-effects for voters registered in American states that permit convenience voting, and controlling for socio-economic, political, and campaign effects, we find that voters with greater static-general political knowledge are more likely to vote early, and those with low political knowledge are more likely to wait until Election Day to cast their ballot. [ABSTRACT FROM AUTHOR]

Is there an optimum method to elicit public records from election officials? Using a field experiment that randomly assigned the wording and email domains used to solicit public records, we test how county election offices respond to requests under given conditions. We find that the response rates of Florida's 67 Supervisors of Elections (SOE) to a request for election administration data and election protocols after the 2018 General Election varies considerably, but that the formality of the language and the email domain randomly assigned only marginally affects the compliance of election officials to abide by the state's public records statutes. The variance in responses and lack of compliance seems to not be related to who asks for data nor how it is solicited, but rather is idiosyncratic. Despite the largely null findings, our study raises concerns about local election administration and transparency. [ABSTRACT FROM AUTHOR]

Because of the COVID-19 threat to in-person voting in the November 2020 election, state and local election officials have pivoted to mail-in voting as a potential solution. This method of voting—while safe from a public health standpoint—comes with its own set of problems, as increased use of mail voting risks amplifying existing discrepancies in rejected mail ballots. While some mail ballot rejections are to be expected, a lack of uniformity in whose ballots get rejected among subgroups of voters—whether for mistakes on a ballot return envelope (BRE) or lateness—raise concerns about equal representation. We draw on official statewide voter file and mail-in ballot data from the 2018 midterm election in Georgia, a state that until the pandemic did not have widespread use of mail voting, to test whether some voters are more likely to cast a mail ballot that does not count. Most importantly, we distinguish between ballots rejected for lateness and those rejected for a mistake on the return envelope. We find that newly registered, young, and minority voters have higher rejection rates compared with their counterparts. [ABSTRACT FROM AUTHOR]

With the increasing usage of dual-mode data collection, researchers of public opinion have shown considerable interest in understanding response differences across different interview modes. Are mode effects an outcome of representation or measurement differences across modes? We conducted a dual-mode survey (web and telephone) using Florida's voter file as the sampling frame, randomly assigning registered voters into one mode versus the other. Having a priori information about the respondents allows us to gauge whether and how sample composition differences may be driven by mode effects, and whether mode affects estimated models of political behavior. Survey mode effects are still significant for issue voting even when sampling design is similar for both modes. [ABSTRACT FROM AUTHOR]

Declining response rates and increasing costs of collecting public opinion survey data have led to an increasing usage of dual-mode surveys. Revisiting foundational theories of political knowledge, issue constraint, and issue voting, we gauge the theoretical implications this methodological change has in our understanding of the public opinion and voter behavior. We see different patterns in the 2012 and 2016 American National Election Studies and observe clear mode effects in an original dual-mode survey experiment of Florida registered voters. Overall, we find that web respondents appear to be more politically knowledgeable, ideologically constrained, and have greater correspondence between issue and vote preferences compared to face-to-face or phone respondents. Survey mode matters, even with a common sampling frame. [ABSTRACT FROM AUTHOR]