Election Survey Research Articles

Ensuring Appropriate Representation in Artificial Intelligence-Generated Medical Imagery: Protocol for a Methodological Approach to Address Skin Tone Bias.

In medical education, particularly in anatomy and dermatology, generative artificial intelligence (AI) can be used to create customized illustrations. However, the underrepresentation of darker skin tones in medical textbooks and elsewhere, which serve as training data for AI, poses a significant challenge in ensuring diverse and inclusive educational materials. This study aims to evaluate the extent of skin tone diversity in AI-generated medical images and to test whether the representation of skin tones can be improved by modifying AI prompts to better reflect the demographic makeup of the US population. In total, 2 standard AI models (Dall-E [OpenAI] and Midjourney [Midjourney Inc]) each generated 100 images of people with psoriasis. In addition, a custom model was developed that incorporated a prompt injection aimed at "forcing" the AI (Dall-E 3) to reflect the skin tone distribution of the US population according to the 2012 American National Election Survey. This custom model generated another set of 100 images. The skin tones in these images were assessed by 3 researchers using the New Immigrant Survey skin tone scale, with the median value representing each image. A chi-square goodness of fit analysis compared the skin tone distributions from each set of images to that of the US population. The standard AI models (Dalle-3 and Midjourney) demonstrated a significant difference between the expected skin tones of the US population and the observed tones in the generated images (P<.001). Both standard AI models overrepresented lighter skin. Conversely, the custom model with the modified prompt yielded a distribution of skin tones that closely matched the expected demographic representation, showing no significant difference (P=.04). This study reveals a notable bias in AI-generated medical images, predominantly underrepresenting darker skin tones. This bias can be effectively addressed by modifying AI prompts to incorporate real-life demographic distributions. The findings emphasize the need for conscious efforts in AI development to ensure diverse and representative outputs, particularly in educational and medical contexts. Users of generative AI tools should be aware that these biases exist, and that similar tendencies may also exist in other types of generative AI (eg, large language models) and in other characteristics (eg, sex, gender, culture, and ethnicity). Injecting demographic data into AI prompts may effectively counteract these biases, ensuring a more accurate representation of the general population.

Representation of diverse skin tones in Nelson's Textbook of Pediatrics.

Physicians are trained to visually recognize disease using images. Many pediatric dermatologic conditions are initially identified and treated by pediatricians, who need to diagnose on varied skin tones. The objective was to evaluate if figures depicting cutaneous disease in the preeminent pediatrics textbook reflect diverse skin tones. Figures depicting dermatologic findings in Nelson's Textbook of Pediatrics were assessed using the Fitzpatrick, Massey-Martin, and Color Bar scales. The distribution was compared to the US population using American National Election Survey 2012 data. The three scales were compared for concordance. Statistical analysis included chi square with P < 0.05 significant. Of 515 figures, 484 were classifiable. Light skin tones were depicted in 453 (93.6%) by Fitzpatrick, 364 (75.2%) by Massey-Martin, and 406 (83.9%) by Color Bar, moderate tones in 92 (19.0%) by Massey-Martin and 53 (11.0%) by Color Bar, and dark tones in 31 (6.4%) by Fitzpatrick, 28 (5.8%) by Massey-Martin, and 25 (5.2%) by Color Bar. The textbook skin tone distribution did not reflect the US population: light 75.2% vs. 63.3%, moderate 19.0% vs. 25.8%, dark 5.8% vs. 11.0%, respectively (P < 0.00001). The three scales yielded consistent proportions for light/moderate vs. dark tones (P = 0.71). Certain conditions were mostly depicted on dark (burns, leprosy, urticaria pigmentosa) or light skin (psoriasis, acne, hemangioma, molluscum, herpes, keloid). Figures demonstrating dermatologic manifestations are predominantly depicted on light skin tones, and are not representative of the US population. Certain conditions were more commonly shown on dark or light skin tones, unrelated to epidemiology.

Partisan (A) Symmetries in Hardball: Mass Level Support of Hardball Equivalent Across Party and Race

Constitutional hardball is when partisan actors engage in legal tactics that violate democratic norms for political advantage. The Republican Party seems to have engaged in hardball more than the Democratic Party in recent years, which is surprising as both parties should be similarly incentivized to use hardball. This asymmetry in elite behavior could reflect differences in what each parties’ coalitions tolerate. For example, non-white voters might oppose hardball tactics more than white voters for reasons such as hardball historically being used to enforce racial hierarchies. If true, that could constrain on Democratic elites’ actions more Republican elites due to Democrats’ more racially diverse base. I draw on a long line of research on partisanship and group interest to argue instead that both white and non-white voters of both parties should be similarly disposed towards hardball. This article then marshals data from the American National Election Survey (ANES) and original surveys to test that argument. In support of my hypotheses, I find white and non-white members of both parties have similar levels of support and similar motivations for hardball. Political elites are likely free to engage in undemocratic acts such as hardball with little fear of backlash from their supporters.

Voices from the margins: How national stories are linked with support for populist radical right parties.

How do national stories shape voter behavior? Do they affect all voters equally, or are some groups more influenced by these narratives? This article examines the impact of "boundary national stories," which highlight clear distinctions between "us" and "them" in national identity, on voting patterns for populist radical right parties (PRRPs). Using original representative election surveys conducted in four Western democracies, we find that voters who embrace a Boundary national story are more likely to vote for Populist Radical Right Parties (PRRPs) than those who do not hold such stories, and that the electoral effect of such stories is more salient for marginalized groups in society. Our findings demonstrate that, while national stories can foster cohesion, they can also drive us apart and polarize our politics. We conclude by discussing the broader implications of these findings for the study of populism in political science.

Reactive Transnationalism and the Ascent of Donald Trump: Evidence From the Latino Immigrant National Election Survey

ABSTRACTBetween 2017 and 2020, the Donald Trump Presidency made for a particularly hostile context of reception for immigrants in the United States, but did it have any measurable influence on their adaptation? Hostility in the form of interpersonal discrimination is known to increase transnational practices—a pattern dubbed reactive transnationalism. But can a hostile government context produce the same result? The current study considers this by examining Latino/a immigrants’ transnational practices at three time points—the 2016 presidential campaign, between Trump's election and inauguration and during his early presidency. The latter should present the most hostile context. Pooled cross‐sectional regression models using data from the Latino Immigrant National Election Study confirm significant increases in transnational remittances and contacts after Trump took office. However, Trump's rise also coincides with a significant decrease in transnational political interest. Overall, the findings provide greater insight into reactive transnationalism and present several directions for future research.

Electoral change through generational replacement: An age-period-cohort analysis of vote choice across 21 countries between 1948 and 2021

People’s generation may have become one of the core predictors of their vote choice. This study examines this hypothesis across 21 Western established democracies between 1948 and 2021. An age-period-cohort analysis on 258 national election surveys (N = 462.084) reveals that the most recent generations are much less likely to vote for the major right-wing party in two-party systems. In multi-party systems, the gradual decline of Christian democratic parties has been largely driven by the generational replacement of pre-WW2 cohorts. Social democratic and conservative parties may face a challenge in future decades because their support is particularly low among the most recent generations whereas liberal, socialist, and particularly green parties stand to gain from generational replacement. Far-right parties have been least popular among voters who came of age during the 1930s and 1940s. A small life-cycle effect points out that people over the age of 65 vote slightly more conservative.

Direct and Indirect Prejudice in Scotland

This paper estimates levels of prejudice in Scotland with data from the 2021 Scottish Election Survey. It examines both direct and indirect indicators of prejudice to discover that somewhere between 15 and 20 percent of respondents indicate prejudicial feelings towards immigrants. Comparisons with results from other surveys in the UK and abroad suggest lower rates of prejudice in Scotland, and declining rates of prejudice over time. Empirical comparisons reveal direct and indirect prejudice to be closely related in Scotland, and regression analysis discovers common determinants of directly and indirectly prejudicial feelings. Political positions, such as ideological conservativism, identifying as a Remainer in the Brexit debate, and support for Scottish independence are important determinants across indicators of prejudice, and the strength of these political indicators demonstrates how Scottish political culture has been linked to minority politics in such a way as to re-draw social boundaries around expressions of prejudice.

Analyzing Public Trust in Presidential Election Surveys: A Study Using SVM and Logistic Regression on Social Media Comments

In the context of democracy in Indonesia, elections play a crucial role, and survey agencies often publish their results on social media. User responses, especially from voters, often express dissatisfaction, including distrust, insults, and negative comments, if the candidate they support receives low survey results. Therefore, this study aims to examine the level of public trust in the survey results of Presidential candidates in 2024 using the Support Vector Machine (SVM) and Logistic Regression algorithms. The study utilized data from 1778 Instagram comments and 985 Twitter tweets. The process involved problem identification, data collection, and system implementation, such as preprocessing, labeling, SMOTE, TF-IDF, data splitting, model classification, and evaluation. The results show that SVM with an 80% training data and 20% test data scenario provides high accuracy, namely 93.19% from Instagram and 91.19% from Twitter. Logistic Regression, with the highest accuracy of 89.79% from Instagram and 88.01% from Twitter in the same scenario. Sentiment analysis using SVM scenario one resulted in 195 positive comments and 216 negative comments. Logistic Regression scenario one shows 180 positive sentiments and 216 negative sentiments. From the classification results, it can be concluded that the level of public trust tends to be negative towards the survey results of the 2024 Presidential candidates, both using SVM and Logistic Regression.

Does perceived inequality promote abstention and protest voting? A study of the 2022 Italian general election

ABSTRACT We focus on the role of perceived economic inequality as a possible antecedent of abstention and protest voting in the 2022 Italian general election. In particular, we hypothesize that citizens’ perceived economic inequality (and the perception that it will grow), reinforced by the negative emotions the perception trigger, may undermine citizens’ trust in institutions. We also investigate whether the erosion of trust is associated with abstention (as an exit option) and protest voting, operationalized as voting for Fratelli d’Italia (FdI), i.e. the only party that did not support the Draghi government (as a voting voice option). We analysed data from the two waves of an online election survey conducted by ITANES (the Italian National Election Study) on a quota sample of the Italian adult population and tested a moderated – mediated model aimed at predicting the association between perceived inequality when eliciting negative emotions, and the probability of voting for FdI or not casting a valid vote. The mediating variable was institutional distrust and the moderating variable was (lack of) political interest. The analyses confirmed that perceived economic inequality does play a role in promoting abstention and protest voting.

BLENDING DOMINANCE, REGIONALISM AND COALITIONS: A NEW READING OF INDIA’S FOURTH PARTY SYSTEM

Since 2014, India’s Bharatiya Janata Party (BJP) has won more elections at the national and state levels than it has lost, returning with an even larger majority in the 2019 General Election. However, the question of whether there has been any break from the past in the nature of the Indian party system remains open. What is the nature of the BJP’s dominance in this period? What is the status of the opposition and the vitality of regional players in this new era? Have coalitions really been relegated to the past? This article investigates the current trends of the party system and compares them with tendencies exhibited historically. It applies mixed methods, largely quantitative analysis of secondary data based on ECI and Lokniti – CSDS election surveys. It argues that what has been called India’s “Fourth Party System” is not completely separate from the trends of the three preceding epochs but rather blends them in unique ways. A re-examination of the election results would reveal how dominance, regionalism and coalitions coexist in an enigmatic symbiosis offering a new reading of the structure of India’s current party system. The 2024 General Elections which witnessed a humbling but historic third consecutive BJP victory albeit reliant on coalition partners only reinforce how the three dimensions of dominance, regionalism and coalitions constantly play out in shaping the character of India’s Party System where changes seem more of a degree rather than of kind.

Political mobilisation and socioeconomic inequality in policy congruence

AbstractIn recent years, scholars have observed that political parties’ policy positions frequently fit the preferences of well‐to‐do voters better than those of the less well‐to‐do; a phenomenon known as policy congruence inequality. While the existence of inequality in policy congruence is well‐established, we currently only have a modest understanding of the causes of it. We develop an argument proposing that the political mobilisation of citizens with low socioeconomic status (SES) both in the parliamentary channel, in the form of high turnout, and in the extra‐parliamentary channel, in the form of high union density, is pivotal. Both high turnout and union density force parties to pay more attention to the preferences of the disadvantaged, thereby creating lower policy congruence inequality. To test the argument, we have collected and harmonised election surveys and party manifestos covering 90 elections in Australia, Denmark, Germany, Sweden, Switzerland, the United Kingdom and the United States, covering several decades until today, yielding more than 120,000 voter–party dyads. Employing this new dataset, our results confirm that the political mobilisation of citizens with low SES is a strong predictor of policy congruence inequality. This finding nuances the conclusion of extant research by showing that low‐SES citizens are not always on the losing side politically. It also implies the important role of maintaining or maybe even increasing turnout and union membership among the disadvantaged in society. Places where either turnout or union density is slipping in these years are likely to witness further increases in policy congruence inequality in the years to come.

Opinion incongruence and public support for direct decision‐making

AbstractPolitical representation does not function well for citizens whose positions on political issues differ from those of elected representatives. In this paper, we argue that opinion incongruence leads citizens to want to bypass elected representatives and place more decision‐making power in the hands of the public. We theorise that this is because incongruent citizens are highly dissatisfied with the existing political system and/or think they will benefit from direct decision‐making in terms of improved policy responsiveness. Using data from the 2019 Belgian Election Survey (n = 3413) and Party Leadership Survey, we find that greater incongruence between citizens’ positions and those of their elected representatives is related to higher support for direct decision‐making. This holds for opinion incongruence with the party voted for and incongruence with Parliament as a whole. This paper contributes novel insights into the consequences of the quality of political representation as well as the drivers of citizens’ support for direct decision‐making processes.

Political Ideology and Attitudinal Ambivalence: Investigating the Role of Ideological Extremity

Are individual differences in political ideology associated with inclinations to hold more or less ambivalent attitudes? Extant research on the linear association of political ideology with attitudinal ambivalence yielded inconsistent findings. The present research tested the hypotheses (a) that the association of political ideology with attitudinal ambivalence is curvilinear with lower levels of ambivalence at both extremes of the ideological spectrum and (b) that higher political interest is associated with lower levels of attitudinal ambivalence. It used data from large and demographically diverse electoral surveys in a set of three studies (Study 1: N = 13,808; Study 2: N = 6,528; Study 3: N = 4,789) that focused on attitudes toward political candidates (Studies 1 and 2) as well as political parties (Study 3) in Germany. Overall, the results support both hypotheses even when general attitudes toward the politicians and parties are controlled in the analyses.

Let the people’s will prevail: Self-uncertainty and authoritarianism predict support for populism

Recent years have witnessed a widespread rise of right-wing populism—an ideology maintaining that the will of the people is supreme, but is subverted by antagonistic elites. Drawing on uncertainty-identity theory, three studies (total N = 5,882) tested the hypothesis that uncertainty would only strengthen support for populism among low authoritarian respondents. Studies 1 and 2 were secondary analyses of American National Election Survey (ANES) 2012 and 2016 data. They supported our key hypotheses in explaining support for the populist American Tea Party (Study 1, N = 1,917), and support for right-wing populist ideology and voting preference for Donald Trump rather than Hillary Clinton (Study 2, N = 2,520). Study 3 ( N = 1,445) experimentally manipulated self-uncertainty to allow a causal interpretation, and focused on right-wing populist ideology. The results are discussed in terms of their contribution to uncertainty-identity theory, and their societal relevance in an environment of growing populism.

Economic Hardship Increases Justification of Political Violence

ABSTRACT Acts of political violence have reached a new level of intensity with the storming of the United States Capitol in 2021. In this paper, I investigate whether and how economic hardship affects political violence justification. Using an original two-wave panel survey around the 2020 US Presidential election, two studies from the American National Election Survey (ANES), and an original study from Germany, the results show substantial support for political violence justification among respondents. In both countries, political violence justification is strongly and consistently associated with economic hardship. The study thus shows that political violence justification is widespread, not just in polarised two-party systems, but also in other, less polarised multiparty systems.

The artificial intelligence shock and socio-political polarization

Are artificial intelligence's complimenting and displacing labor market effects corresponding with similar polarization in socio-political beliefs? This study answers this question by analyzing survey data of 26,311 Americans, collected from the American National Election Survey. Data is deployed alongside 22-category Manyika et al. (2017) ‘automation potential’ estimates (proxying for ‘displacement due to AI’) and Michael Webb (2019) ‘AI-exposure’ estimates (proxying for labor complimented by AI). The study summarizes the demographic characteristics and socio-political views of the highly AI-exposed and automation-susceptible groups. It deploys a year and region fixed effects OLS model, with standard error clustered at the occupation level. This study then finds that automation-susceptible ‘losers’ of AI are more likely to be culturally conservative and economically left-leaning. Those complimented by AI are more likely to hold socially liberal and fiscally conservative views. The results suggest AI's labor market polarization may accompany radicalization and socio-political divergence, with implications for vote capture and representation of working-class interests in government.

Media Exposure and Vote Choice in India, 1996–2019

This article examines the relationship between media consumption and voting choices in India in the context of increasing exposure to media, both traditional and new. Using media exposure-related data from the Centre for the Study of Developing Societies’ (CSDS) national election surveys since 1996, it makes a few key assertions. Historically, there has been a positive relationship between traditional news media exposure and voting for the BJP, i.e., the greater the voters’ TV and newspaper consumption, the higher their support for the BJP. With respect to the Congress party, the tendency has generally been the opposite. This pattern now extends to social media exposure as well. However, this impact of media exposure on voting preferences that’s visible at the broad level weakens considerably when slicing the data further by socio-demographic factors. It has been found that the trend of the BJP’s vote share rising with increasing exposure to media is not consistent across the categories of age, education, and caste during many elections. If there is one election, however, that bucked this trend, it was the 2014 Lok Sabha election. The article argues that the BJP was successful in recognizing this and capitalized on it through an extensive media and communications campaign in 2014.

The “Us vs. Them” Mentality: The Role of Affective Polarization in Deepening the Partisan Divide in Media Bias Perception

This study examines how affective polarization impacts the partisan divide in perceptions of cable news networks’ political biases. Results from a 2016 U.S. presidential election survey show that affective polarization deepens this divide, even after controlling for election involvement, partisanship strength, and cable news usage. Partisanship strength is more closely associated with in-party love than out-party hate, but out-party hate has a stronger association with media bias perception, indicating that out-group hate is more influential in shaping partisans’ media bias perception.

Politicized science and rural–urban divides: Exploring how rurality in the place of residence and red and blue media use affected attitudes toward scientists before and during the COVID‐19 pandemic

AbstractBased on the concept of place‐based identity, this study disentangles the effects of rurality and partisanship and considers their interactions with red/blue media use in rural–urban divides in attitudes toward scientists. By employing the 2016 American National Election Survey and our 2021 survey data, we investigated how partisanship, rurality, and red/blue media use related to attitudes toward scientists. Analyses revealed distinct roles for partisanship and rurality in predicting attitudes, even after controlling for demographic variables, including family income and education. Compared to those who resided in large metro areas, residents of more rural areas tended to harbor more negative attitudes toward scientists. This tendency was reinforced by red media use in 2021 but not in 2016. Given the dissimilar results in 2016 and 2021 in terms of the interaction of red media use and rurality, we should pay more attention to rural–urban divides in attitudes toward scientists.

‘It's the party that counts’? The Rise of Labour and the Image of the Woman Politician at English Elections, c.1929–1950

AbstractThis article uses election addresses to consider how the early women parliamentary candidates sought to make their case to English voters. It then explores the insights that Mass Observation's election surveys offer into public attitudes to women politicians, and gender and political leadership more broadly, from the late 1930s to the early 1950s. While the pioneer female candidates argued that they should have more representatives at Westminster to better uphold the ‘woman's point of view’ this approach was gradually undermined from the 1920s onwards with the growth of programmatic politics led by Labour. Mass Observation found that voters claimed to focus more on which party had the best programme rather than the personalities of candidates. However, their findings also indicate that women candidates continued to face many additional prejudices which their male opponents did not.