The Advent of Internet Surveys for Political Research: A Comparison of Telephone and Internet Samples

The Internet offers a number of advantages as a survey mode:low marginal cost per completed response, capabilities for providingrespondents with large quantities of information, speed, andelimination of interviewer bias. Those seeking these advantagesconfront the problem of representativeness both in terms ofcoverage of the population and capabilities for drawing randomsamples. Two major strategies have been pursued commerciallyto develop the Internet as a survey mode. One strategy, usedby Harris Interactive, involves assembling a large panel ofwilling respondents who can be sampled. Another strategy, usedby Knowledge Networks, involves using random digit dialing (RDD)telephone methods to recruit households to a panel of Web-TVenabled respondents. Do these panels adequately deal with theproblem of representativeness to be useful in political scienceresearch? The authors address this question with results fromparallel surveys on global climate change and the Kyoto Protocoladministered by telephone to a national probability sample andby Internet to samples of the Harris Interactive and KnowledgeNetworks panels. Knowledge and opinion questions generally showstatistically significant but substantively modest differenceacross the modes. With inclusion of standard demographic controls,typical relational models of interest to political scientistsproduce similar estimates of parameters across modes. It thusappears that, with appropriate weighting, samples from thesepanels are sufficiently representative of the U.S. populationto be reasonable alternatives in many applications to samplesgathered through RDD telephone surveys.     

Voter Learning in State Primary Elections

When voters learn about candidates' issue positions during election campaigns, does it affect how they vote? This basic question about voters remains unanswered in part because of a methodological obstacle: learning candidates' issue positions may influence not only voters' vote choice but also their issue positions. To surmount this obstacle, we attempt to answer this question by examining statewide primary elections, which are arguably less vulnerable to this reverse causation problem because they lack partisan cues and are of much lower salience than presidential elections. Using both existing polling data and our own panel Internet surveys, we find that voters learn about the ideologies of candidates during statewide primary campaigns and that this learning affects their voting decisions in senate and gubernatorial primaries. We fail to find similar results for down‐ballot primaries, raising questions about voters' ability to make informed judgments for these types of elections.     

Individual level dynamics of PTV change across the electoral cycle

Propensity-to-vote (PTV) scores are ever more commonly used in electoral research as a measure of electoral utilities. Yet a growing literature employs them as dependent variable in the voting equation in place of the lower information granted by vote recall questions. However, this choice can be seen as problematic because of the very structure of election survey research. To the extent that voters' PTVs are measured in post-election surveys (as it is often the case) these are likely to result endogenously produced by actual voting behavior in the past election – thus partly undermining the validity of the PTV question which, ideally, should not be related to any specific election. In this paper, we try to disentangle the relationship between short-term political attitudes (leader evaluations, issue proximity, economic assessments) and voters' changing patterns of propensities to vote in both an electoral and a non-electoral context. The latter scenario serves as a means to rule out the potentially contaminating effect of voting choices on voters' PTVs. The data comes from two panel surveys of Italian voters conducted by ITANES in occasion of the 2006 general election, and in 2011 (that is, in a non-electoral year) respectively.     

Voter overrepresentation,vote misreporting,and turnout bias in postelection surveys

Figures from postelection surveys often grossly overestimate election turnout. Two distinct phenomena are responsible for this gap: overrepresentation of actual voters and vote misreporting by actual nonvoters among survey respondents. Previous accounts of turnout bias are inconclusive in that they either focus on a single component, or fail to separate between the two. In this paper, we formally decompose turnout bias in election surveys into its constituent parts, assess their empirical prevalence and heterogeneity using an extensive collection of 49 vote validation studies from six countries, and employ Bayesian meta regression techniques to account for cross-study differences. Our results indicate that both election and survey characteristics such as actual voter turnout and survey response rates differentially affect the components of turnout bias. We conclude with a discussion of the threats and potentials of our findings for survey-based comparative electoral research.     

The costs of electoral fraud: establishing the link between electoral integrity,winning an election,and satisfaction with democracy

Previous research has shown that voters’ perception of electoral fairness has an impact on their attitudes and behaviors. However, less research has attempted to link objective measurements of electoral integrity on voters’ attitudes about the democratic process. Drawing on data from the Comparative Study of Electoral Systems and the Quality of Elections Data, we investigate whether cross-national differences in electoral integrity have significant influences on citizens’ level of satisfaction with democracy. We hypothesize that higher levels of observed electoral fraud will have a negative impact on evaluations of the democratic process, and that this effect will be mediated by a respondent’s status as a winner or loser of an election. The article’s main finding is that high levels of electoral fraud are indeed linked to less satisfaction with democracy. However, we show that winning only matters in elections that are conducted in an impartial way. The moment elections start to display the telltale signs of manipulation and malpractice, winning and losing no longer have different effects on voter’s levels of satisfaction with democracy.     

A measure of survey mode differences

We evaluate the effects of different survey modes on respondents’ patterns of answers using an entropy measure of variability. While measures of centrality show little differences between face-to-face and Internet surveys, we find strong patterns of distributional differences between these modes where Internet responses tend towards more diffuse positions due to lack of personal contact during the process and the social forces provided by that format. We introduce an entropy measure of dispersion for survey responses and illustrate its utility with election data from 2012. Our results provide clear evidence that mode matters in modern survey research, and we make recommendations for interpreting results from different modes.     

Support for electoral system reform among voters and politicians: Studying information effects through survey experiments

Does information about the consequences of proposals to change the Norwegian parliamentary electoral system influence voters' and politicians' attitudes towards the system? Is the willingness to accept change greater among voters/politicians who “lose” under the present electoral system? These questions are illuminated using empirical data from two identical survey experiments, with responses from both voters and politicians about 1) increased proportionality between parties (more seats for smaller parties) and 2) increased geographical proportionality (stronger representation for the more populous counties). The results show that being informed about the consequences of the proposals has a major effect on voters' and politicians' attitudes. This applies especially to the question of increased proportionality between parties, where feedback was particularly negative from respondents who were told that the proposal might weaken the larger parties’ representation and make it more difficult to establish viable governments. The responses to the question about increased proportionality between parties were also influenced by partisanship; politicians who belonged to or voters who voted for one of the smaller parties favour increased proportionality. We also find that there is limited support for the proposal to distribute parliamentary seats according to the number of inhabitants in the counties, and this support is further reduced when the respondents are informed that the measure will increase representation from the more populous parts of the country.     

Parties are always right: the effects of party cues and policy information on attitudes towards EU issues

Abstract

This article analyses the influence that political parties exert upon citizens’ opinions about European Union issues. By measuring at the same time the content and source effects on political attitudes, the article considers the possibility that voters pay less attention to the arguments used in a political message than to its source. Results from an online survey experiment in Spain show that partisan voters use a heuristic model of processing when taking positions on an unfamiliar EU issue, even though the prevalence of the source effect is moderated by the respondent’s political sophistication and party attachment. The results also indicate that some respondents tend to pay less attention to a message’s content when the message comes from their preferred party. Such findings raise concerns about the possibility for EU issue voting to guarantee the accountability of political elites and party–voter linkages.     

From voter ID to party ID: How political parties affect perceptions of election fraud in the U.S.

This paper uses a survey experiment to assess what individuals understand about election fraud and under what circumstances they see it as a problem. I argue that political parties are central to answering both these questions. Results from the 2011 CCES survey suggest respondents are able to differentiate between the relative incentives of Democrats and Republicans where fraud tactics are concerned, but whether voters see these tactics as problematic is heavily influenced by partisan bias. The results show little support for the notion that partisan ideology drives fraud assessments, and suggest support for the idea that individual concerns for fraud are shaped a desire for their preferred candidate to win. These results offer insights that might be applied more broadly to questions of perceptions of electoral integrity and procedural fairness in democracies.     

Different Surveys,Different Results? A Comparison of Two Surveys on the 2009 European Parliamentary Election

The European Election Survey (EES) is carried out in all member states at the time of each European Parliament election. The mode of data collection (mainly telephone interviews) and the sampling procedure (achieving 1,000 interviews in each country) raise doubts about the data quality. Until now knowledge has been lacking about the extent to which the mode of data collection and the sampling procedure bias the results. In one European country an independently designed election survey is carried out: The Swedish National European Parliament Election Study (SNES). The survey consists of face-to-face interviews of a random net sample of 2,682 individuals (response rate 67 per cent compared to 11.2 in the Swedish EES survey). In addition, SNES includes a large number of variables from official register data (including validated voting) that facilitates analyses without any non-responses generating missing data. This quasi-experimental methodological set up is used to compare the data from the two surveys for voter turnout, left–right placement and party choice. Results show that EES overestimates turnout levels more than SNES. EES also has a large overrepresentation of highly educated respondents, and thus underestimates differences in turnout between highly and less educated citizens. As for left–right placement, respondents in EES place both themselves and the parties on more extreme positions. Regarding party choice, the main difference between the surveys is that the EES largely underestimates the share of Social Democratic voters.     

An Experiment Testing the Relative Effectiveness of Encouraging Voter Participation by Inducing Feelings of Pride or Shame

Prior experimental research has demonstrated that voter turnout rises substantially when people receive mailings that indicate whether they voted in previous elections. This effect suggests that voters are sensitive to whether their compliance with the norm of voting is being monitored. The present study extends this line of research by investigating whether disclosure of past participation has a stronger effect on turnout when it calls attention to a past abstention or a past vote. A sample of 369,211 registered voters who voted in just one of two recent elections were randomly assigned to receive no mail, mail that encouraged them to vote, and mail that both encouraged them to vote and indicated their turnout in one previous election. The latter type of mailing randomly reported either the election in which they voted or the one in which they abstained. Results suggest that mailings disclosing past voting behavior had strong effects on voter turnout and that these effects were significantly enhanced when it disclosed an abstention in a recent election.     

The effect of socioeconomic policy and competence messages on populist radical right support: Evidence from a pre-election survey experiment

Does highlighting socioeconomic policy considerations or mainstream parties’ government competence reduce support for populist radical right (PRR) parties? Such “defuse” messages may attract PRR voters without alienating mainstream parties’ core electorate and thus, have advantages over an accommodative strategy. This study tests four “defuse” messages in an original survey experiment on a sample of 1,786 likely PRR voters in the context of the 2017 German federal election. The findings show that potential PRR voters are hardly swayed by these messages. This result is in notable contrast to findings from prior experimental studies about the malleability of PRR support. Exploratory analyses suggest that some of these null findings may mask heterogeneities. Both respondents who were surveyed during the first days of fieldwork and those with less political knowledge responded to some treatments in the expected way. Overall, these findings point to a limited responsiveness of PRR voters to “defuse” messages.     

Artificial Inflation or Deflation? Assessing the Item Count Technique in Comparative Surveys

While the popularity of using the item count technique (ICT) or list experiment to obtain estimates of attitudes and behaviors subject to social desirability bias has increased in recent years among political scientists, many of the empirical properties of the technique remain untested. In this paper, we explore whether estimates are biased due to the different list lengths provided to control and treatment groups rather than due to the substance of the treatment items. By using face-to-face survey data from national probability samples of households in Uruguay and Honduras, we assess how effective the ICT is in the context of face-to-face surveys—where social desirability bias should be strongest—and in developing contexts—where literacy rates raise questions about the capability of respondents to engage in cognitively taxing process required by ICT. We find little evidence that the ICT overestimates the incidence of behaviors and instead find that the ICT provides extremely conservative estimates of high incidence behaviors. Thus, the ICT may be more useful for detecting low prevalence attitudes and behaviors and may overstate social desirability bias when the technique is used for higher frequency socially desirable attitudes and behaviors. However, we do not find strong evidence of variance in deflationary effects across common demographic subgroups, suggesting that multivariate estimates using the ICT may not be biased.     

Why don't people vote in U.S. primary elections? Assessing theoretical explanations for reduced participation

Primary election participation in the United States is consistently lower than general election turnout. Despite this well-documented voting gap, our knowledge is limited as to the individual-level factors that explain why some general election voters do not show up for primary contests. We provide important insights into this question, using a novel new survey to examine three theoretical perspectives on participation never before empirically applied to primary races. Compared to general elections, we find that for U.S. House primary elections sizable segments of the electorate consider the stakes lower and the costs of voting greater, feel less social pressure to turn out and hold exclusionary beliefs about who should participate, and are more willing to defer to those who know and care more about the contests. Multivariate analysis reveals that these attitudes explain validated primary election participation. These findings point to new directions for future research.     

Sampling People or People in Places? The BES as an Election Study

Electoral surveys conducted as part of the regular series of British Election Studies have used a stratified, clustered sampling design. This is constructed to ensure a nationally representative sample of voters (after weighting) but does not necessarily ensure a representative sample of the different social areas within the country. Much recent work has indicated the important role of local social and electoral milieux in the structuring of electoral behaviour but these are not reflected in the sampling design. An analysis of the geography of the face-to-face pre-election component of the 2005 BES shows not only that it was unrepresentative of certain types of area (defined using bespoke neighbourhood data) but also that it was less representative of such areas than the two (larger) internet samples (pre- and post-election) also undertaken as part of the 2005 BES. This suggests the need to reconsider the nature of the sample designs (even the entire survey methodology) to be deployed at future election studies in Great Britain.     

Assessing (and fixing?) Election Day lines: Evidence from a survey of local election officials

Despite recent studies that find few people face significant wait times when attempting to vote in U.S. elections, the 2012 election produced numerous anecdotal and journalistic accounts claiming otherwise. This study relies on a national survey of local election officials to systematically ascertain their views about the challenges and successes they had in administering the 2012 general election. Consistent with surveys of voters, most officials report that wait times and lines were minimal. Furthermore, the relative amount of money available to a jurisdiction for election administration was unrelated to the occurrence of these problems, while the presence of more poll workers—especially first-timers—may actually exacerbate them.     

Voter‐weighted environmental preferences

This article examines the political economy of preferences with respect to the environment using a new stated preference survey that presents the first benefit values for national water quality levels. The mean valuation greatly exceeds the median value, as the distribution of valuations is highly skewed. The study couples the survey valuations with unique and extensive information on respondent voting patterns. Preferences of registered voters are similar to the preferences of the population at large, but median voters value water quality more than nonvoters. The strongest contrast related to voter‐weighted preferences is among voters for different candidates, as those who voted for Gore in the 2000 presidential election have the highest environmental values. © 2009 by the Association for Public Policy Analysis and Management.     

Socioeconomic status,unemployment experience,and political participation: A disentangling of main and interaction effects

A clarification of the effects of unemployment on political participation attitudes and behaviors is developed by contrasting the effect of unemployment experience across categories of socioeconomic status. Data on employed and unemployed heads of household are drawn from the 1976 University of Michigan national presidential election survey. The results indicate both main and interaction effects. Regardless of employment status, lower socioeconomic status respondents are less committed to voting, feel less efficacious, are less interested in politics, and are less politically active than persons of higher status. However, participation attitudes and behaviors are more adversely affected by unemployment experience among those of lower than higher status. Among higher status respondents, attitudes toward self (i.e., feelings of efficacy) and political interest—but not political activity or attitude toward the importance of participation—are altered by unemployment.An earlier version of this paper was presented at the Southwestern Sociological meetings in Ft. Worth, Texas, March 1979.     

Same Considerations,Different Decisions: Motivations for Split‐ticket Voting among Swedish Feminist Initiative Supporters

This article explores why supporters of small, non‐established parties choose to vote for different parties in the elections to the European Parliament (EP) and elections to the national parliament. It uses individual‐level data with open‐ended questions from an online survey on supporters of Feminist Initiative (Fi) – a comparatively small and new Swedish feminist party – to map voters’ own motivations for split‐ticket voting in the 2014 elections. Contrary to expectations based on second‐order election theory, it is found that voters ticket‐split in both directions: there are those voting for Fi in the EP election but not in the national election, and those voting for Fi in the national election but not in the EP election. These voters take the same types of considerations into account but nevertheless end up making opposite voting decisions. Voters clearly distinguish between the two levels – for example, by prioritizing different issues.