Muhammet Furkan Karakaya_

This research note proposes and implements an innovative computational approach to measuring party cohesion and conceptualizes as the ideological congruence between party's ideological platform (manifesto) and the discourses of each nominated candidates. Traditional approaches focus on party discipline that rely on roll-call vote, party unity score and survey-based measures based on subjective perception. Moving beyond the traditional approaches, this study conceptualize party cohesion as an outcome of candidate nomination process and introduces a method that uses sentence transformer language models to compute ideological congruence score based on quantifying semantic similarity between party manifestos and candidates' public statements on X (Twitter). The proposed party cohesion measure, ideological congruence score (ICS) of electoral candidates, differentiates itself by capturing both elected and unelected candidates' ideological congruence to their party and provides a scalable continuous measure that allows intraparty, interparty comparison across different party systems. This paper aims to fill the gap in the literature on party cohesion that relies on existing measurement approaches and maps out the details of the computational methodology with empirical illustrations from Turkey and the United States. This approach expands the empirical scope of party cohesion and opens new avenues for comparative party politics.

When and how do political parties implement the broad-appeal strategy? I argue that political parties use the broad-appeal strategy based on constituent composition across electoral districts. Political parties need to balance party cohesion and electoral competition when they implement the broad-appeal strategy. This study aims to fill the gap in the literature by going beyond the dichotomous approach of whether appealing to a broader audience harms party cohesion or not. I explore what explains variation in party cohesion based on the percentage of core supporters across electoral districts. Moreover, I present an innovative measurement for party cohesion to assess the effects of the broad-appeal strategy based on the candidates' ideological congruence to party manifestos. The findings indicate that parties nominate ideologically more congruent candidates in electoral districts where the population percentage of core supporters is high and vice versa. To test the argument, I use an ethnic party from Turkey as a case to examine its broad-appeal nomination strategies for the 2015, 2018, and 2023 general elections.

This article examines the extent to which the Democratic and Republican parties implement broad-appeal strategies in Congressional elections. American political parties are notorious for their internal diversity and lack of cohesion. Among the many factors affecting party cohesion in the U.S. is the single-member-district plurality method of election, which encourages candidates to tailor their electoral appeals to their geographic constituencies and tends to reduce ideological cohesion. At the same time, American parties can also appear to be more ideologically unified than one might expect, particularly in recent years as the nationalization of elections has increased. Here, we draw on a form of semantic text analysis to investigate the extent to which Democratic and Republican Party Congressional candidates are ideologically cohesive across electoral districts, as well as the extent to which party cohesion is shaped by the number of party supporters across electoral districts. We draw on a unique computational approach to estimate the ideological distance between party manifestos and each Congressional candidate's political discourse scraped from their Twitter profiles in the 2016, 2018, 2020, and 2022 Congressional elections, regardless of candidates' electoral victory or loser status. Our general expectation is that Congressional candidates running in electorally safe districts will be ideologically closer to their party than Congressional candidates running in competitive or volatile districts, where candidates are more likely to tailor their electoral appeals to the needs of the district. Our findings generally support this expectation. We do find less party cohesion in competitive, rather than safe, districts. These findings bear important implications for party strategies and ideological representation, suggesting that high numbers of safe districts encourage greater levels of party cohesion.