Explore the vast range of titles available.

Citizens’ exit polls are performed by local voters to verify the official reported election results. Five citizens’ exit polls were run in southeast Kansas during the Nov 8th 2016 election. These exit polls were designed specifically to verify computer generated vote counts and run solely by volunteer labor, all local citizens who were willing to put in the necessary hours on Election Day to conduct the poll and later, to count the results by hand. These exit polls were able to obtain high participation rates resulting in the ability to detect small yet statistically significant differences. All five polling stations surveyed show evidence of multiple statistical anomalies in both the pattern and size of the errors between the official results and exit poll results although biases were not uniformly oriented across sites. The small discrepancies found in the studied races were insufficient to alter the outcomes. Non-response bias and unintentional errors were evaluated as potential causes; those explanations were plausible in some but not all cases. These results show a pattern of discrepancies between the exit polls and computer counted results displaying consistent bias within sites. This would be an expected outcome of a deliberate manipulation of the computer results. While this data doesn’t conclusively prove election interference and manipulation of votes counts, it should be taken seriously as a sign of such interference. Doubts about the accuracy of the reported results are appropriate unless other plausible explanations for the discrepancies can be found.

Despite their numerous advantages, exit polls are not a common tool in the study of Canadian electoral behaviour. In this methodological note, we use data from two pilot projects to test small-scale exit polls’ accuracy when estimating party support. We mobilize exit-polling data collected in the 2018 Quebec provincial election (four voting locations) and the 2019 federal election in Quebec (two voting locations). We focus on chance error and bias error in small samples. Results obtained using parametric linear models suggest that small sample exit polls achieve relatively precise estimations. We do find, however, that right-of-the-centre parties’ vote share tends to be underestimated. These findings shed light on the strengths and shortcomings of small-scale exit polls in Canada.

This article uses data from the American National Election Studies and national exit polls to test Fiorina's assertion that ideological polarization in the American public is a myth. Fiorina argues that twenty-first-century Americans, like the midtwentieth-century Americans described by Converse, “are not very well-informed about politics, do not hold many of their views very strongly, and are not ideological” (2006, 19). However, our evidence indicates that since the 1970s, ideological polarization has increased dramatically among the mass public in the United States as well as among political elites. There are now large differences in outlook between Democrats and Republicans, between red state voters and blue state voters, and between religious voters and secular voters. These divisions are not confined to a small minority of activists—they involve a large segment of the public and the deepest divisions are found among the most interested, informed, and active citizens. Moreover, contrary to Fiorina's suggestion that polarization turns off voters and depresses turnout, our evidence indicates that polarization energizes the electorate and stimulates political participation.

With the growth of Latino and Asian American populations, candidates frequently must appeal to diverse electorates. Strategies for doing so include emphasizing candidates' racial/ethnic identity and securing endorsements from racial/ethnic groups. While many scholars focus on candidates' racial/ethnic attributes, ethnic group endorsements are understudied. Whether such endorsements induce voters to choose ideologically similar candidates (spatial voting), or choose based on race/ethnicity (racial voting) is unclear. We address this question by examining elections in multiethnic local settings. Using original surveys and exit polls, we create comparable measures of candidate and voter ideology, and examine how race/ethnicity and ideology affect voters' choices. We also embed experiments that manipulate ethnic group endorsements. We find that ideology influences voters' choices, but that ethnic group endorsements weaken spatial voting. The latter effect among whites is driven by racial/ethnic stereotypes. These reactions explain why some candidates seek such endorsements and why others might prefer to avoid them.

On November 8, 2016 Donald Trump, a man with no office-holding experience, won the Electoral College, defeating the first woman to receive the presidential nomination from a major party. This paper offers the first observational test of how sexism affects presidential vote choice in the general election, adding to the rich literature on gender and candidate success for lower-level offices. We argue that the 2016 election implicated gender through Hillary Clinton’s candidacy and Donald Trump’s sexist rhetoric, and activated gender attitudes such that sexism is associated with vote choice. Using an Election Day exit poll survey of over 1300 voters conducted at 12 precincts in a mid-size city and a national survey of over 10,000 White and Black Americans, we find that a politically defined measure of sexism—the belief that men are better suited emotionally for politics than women—predicts support for Trump both in terms of vote choice and favorability. We find the effect is strongest and most consistent among White voters. However, a domestically defined measure of sexism—whether men should be in control of their wives—offers little explanatory power over the vote. In total, our results demonstrate the importance of gender in the 2016 election, beyond mere demographic differences in vote choice: beliefs about gender and fitness for office shape both White men and women’s preferences.

A pivotal question raised in the 1980s debate over the durability of the United States-led world order still haunts us today. Does the postwar liberal international order, which the US was central in shaping, serve US interests? President Trump’s answer is a resounding no, promising an ambitious redistributive programme to rebalance global wealth and power. Similar to political platforms in other advanced countries, Trump argues for a fundamental revision of the order, proposing a renewal of international principles to foster a global system tailored to US interests. The major theatres where this battle is being fought are the politics of security, trade and money. Prominent academics align on some aspects of this agenda. Instead, I contend that as the primary beneficiary of the liberal international order, the US will be its first casualty. While President Trump is right to identify the potential role redistribution could play in strengthening America’s global leadership, he misidentifies the nature of the distribution problem. Internationally, ‘America First’ is premised on zero-sum logic and poses a risk to the liberal international order, to US security and prosperity. Domestically, ‘white America First’ promises to restore lost greatness to white Americans, aggravating economic and political inequality in the US. Drawing on presidential exit polls, other survey data, real income and income growth nationally and regionally, I explain the 2016 election outcome as a function of education and ethnicity, contextualized by income concerns and racism. Redistributive domestic policies, particularly expanding higher education, are necessary for US support of the liberal international order to endure.

Political researchers and scientists have long sought to forecast the results of upcoming elections in a democratic setup. Prediction approaches have been combined to improve prediction accuracy. In this work, we focus on India, utilizing exit polls spanning the last 8 years to develop machine learning-based prediction models. Exit polls provided by major media agencies for both provincial as well as the union legislative assemblies have been used as inputs to improve prediction accuracy. The training set comprises of the seats projected for the United Progressive Alliance (UPA) and the National Democratic Alliance (NDA) in state and central government elections from 2014 through 2020, providing ample data for training across a wide range of short and long-term trends. The testing set includes seats projected for the UPA and the NDA in 2021, across states including Assam, Kerala, Pondicherry, Tamil Nadu and West Bengal, ensuring validation of the prediction models using undisclosed ‘out of sample’ data. Trained models have been used to provide projected seats for the 2021 legislative assembly elections in the states of Assam, Kerala, Pondicherry, Tamil Nadu and West Bengal. Efficient machine learning approaches viz. support vector regression (SVR), extreme gradient boosting (XGBoost) and deep CNN-LSTM networks have been explored as predictors. The performance of the predictive models is evaluated by utilizing the metrics such as relative error (RE%), normalised mean square error (NRMSE), fractional bias (FB), coefficient of determination (R 2 ) and normalised absolute error (NAE). XGBoost exhibits the lowest R 2 value of 0.97, indicating superior fit to the observed data compared to other methods. Additionally, it demonstrates the lowest FB value of −0.03, suggesting minimal prediction bias. Notably, XGBoost also yields the lowest NRMSE, NAE, RE% values (4.11, 2.73 and 0.97% respectively), highlighting its superior accuracy in predicting election outcomes. XGBoost is followed by the deep CNN-LSTM model, while the SVR model exhibits the lowest prediction accuracy among the proposed forecasting methods.

What happens after primary elections? Strategies of loyalty or defection in general elections have been addressed by US literature mainly by means of aggregate data. However, we lack similar studies in non-US contexts. This article investigates the strategies followed after primary elections by taking the case of the Italian Partito Democratico as an illustration. We addressed the individual drivers of loyalty or defection strategies by considering three different dimensions: (1) the outcome of the primary election, having backed a winning or losing candidate; (2) the strength of partisanship, meant as ideological congruence with the party and partisan involvement; and (3) the leader effect. We relied on four surveys (exit polls) administered during party leadership selections held in 2009, 2013, 2017 and 2019. The results suggest that all three dimensions have an influence on post-primary strategies, but what counts the most is partisan involvement.

Research shows people share common political facial stereotypes: They associate faces with political ideologies. Moreover, given that many voters rely on party affiliation, political ideology, and appearances to select political candidates, we might expect that political facial stereotypes would sway voting preferences and, by extension, the share of votes going to each candidate in an election. And yet few studies have examined whether having a stereotypically conservative-looking (or liberal-looking) face predicts a candidate's vote shares. Using data from US. election exit polls, we show that the Republican voters within each state are more likely to vote for a candidate (even a Democrat) the more that person has a stereotypically Republican-looking face. By contrast, the voting choices of the Democratic voters within each state are unrelated to political facial stereotypes. Moreover, we show that the relationship between political facial stereotypes and voting does not depend on state-level ideology: Republican voters in both right-leaning (\"red\") and left-leaning (\"blue\") states are more likely to vote for candidates with conservative-looking faces. These results have several important practical and theoretical implications concerning the nature and impact of political facial stereotypes, which we discuss.

Exit poll surveys during elections are conducted to predict the outcome of actual elections. However, such polls historically have been controversial, particularly for multi-phase elections, because they can influence the behavior of voters in the later rounds of voting. If subsequent voters are more likely to vote for the predicted frontrunner, the effect is known as the bandwagon voting phenomenon, whereas if they vote for the predicted trailing candidate the phenomenon is known as underdog voting. To avoid such issues, in 2009 the election administration in the world’s largest democracy (India) introduced a blanket ban on publishing exit polls in the media until all rounds of an election are completed. Exploiting the potentially exogenous timing of this reform, and using administrative data to compare states that went to elections before and after the ban, we find that in response to the policy, vote shares increased for the frontrunner and declined for others. The result implies that in the counterfactual, without the ban, fewer people would have voted for the frontrunner. The evidence is suggestive of underdog voting.