[Paper Review] Devam vs. Tamam: 2018 Turkish Elections

Mücahid Kutlu, Kareem Darwish+1 more|arXiv (Cornell University)|Jul 17, 2018

Social Media and Politics1 references5 citations

TL;DR

This study analyzes 108 million tweets from the two months before the 2018 Turkish presidential election to examine political polarization between pro-Erdogan ("Devam") and anti-Erdogan ("Tamam") supporters. Using semi-automatic labeling and label propagation, it identifies distinct media consumption, ideological preferences, and entertainment habits—revealing deep societal polarization beyond party lines.

ABSTRACT

On June 24, 2018, Turkey held a historical election, transforming its parliamentary system to a presidential one. One of the main questions for Turkish voters was whether to start this new political era with reelecting its long-time political leader Recep Tayyip Erdogan or not. In this paper, we analyzed 108M tweets posted in the two months leading to the election to understand the groups that supported or opposed Erdogan's reelection. We examined the most distinguishing hashtags and retweeted accounts for both groups. Our findings indicate strong polarization between both groups as they differ in terms of ideology, news sources they follow, and preferred TV entertainment.

Motivation & Objective

• To understand the ideological and behavioral divides between pro-Erdogan and anti-Erdogan supporters during the 2018 Turkish elections.
• To identify the most distinguishing hashtags and retweeted accounts used by each political group.
• To examine differences in news sources, political demands, and preferred entertainment (TV series) between the two groups.
• To assess the extent of societal polarization beyond party affiliation, including media consumption and cultural preferences.
• To provide empirical evidence of how digital discourse reflects broader societal cleavages in Turkey’s political landscape.

Proposed method

• Collected 108 million tweets from April 29 to June 23, 2018, using keywords related to the election, candidates, parties, and hashtags.
• Applied a semi-automatic labeling approach followed by label propagation to classify approximately 652,700 Twitter users as pro-Erdogan (279.2K) or anti-Erdogan (373.5K).
• Analyzed the most frequently used hashtags and retweeted accounts for each group to identify key discourse markers and information sources.
• Categorized media consumption by distinguishing between news media, journalists, political parties, and regular users.
• Mapped TV series preferences by analyzing mentions of specific shows and their alignment with each group’s preferences.
• Used quantitative frequency analysis of hashtags and retweets to compare ideological, cultural, and media preferences between groups.

Experimental results

Research questions

• RQ1 What are the most distinguishing hashtags used by pro-Erdogan and anti-Erdogan Twitter users during the 2018 Turkish election?
• RQ2 Which news sources and political accounts are most frequently retweeted by pro- and anti-Erdogan supporters?
• RQ3 How do the two groups differ in their consumption of political and cultural content, such as TV series?
• RQ4 To what extent does political polarization in Turkey extend beyond party lines into media consumption and entertainment preferences?
• RQ5 What role do religious and nationalist symbols (e.g., Jerusalem, Quds) play in shaping the discourse of each group?

Key findings

• The pro-Erdogan group predominantly used hashtags related to national unity, such as #Devam and #Kudüs, while the anti-Erdogan group favored #Tamam and pro-Palestine slogans like #WeStandForQuds.
• Anti-Erdogan users were 99.3% likely to mention the drama series 'Vatanım Sensin,' compared to only 0.7% for pro-Erdogan users, indicating strong cultural polarization.
• Pro-Erdogan users were 79.7% likely to mention 'Diriliş Ertuğrul,' a historically themed series, while anti-Erdogan users were 99.5% likely to mention 'Erkenci Kuş,' suggesting divergent cultural identities.
• The anti-Erdogan group was 51.8% more likely to retweet accounts from opposition parties (CHP, HDP, IYI, etc.) compared to pro-Erdogan users, who favored AKP and MHP-affiliated accounts.
• Pro-Erdogan users were 33.5% more likely to retweet accounts from the government or pro-government media, while anti-Erdogan users were 17.4% more likely to retweet independent media and journalists.
• The pro-Erdogan group showed a 26.7% higher retweet rate for AKP-related accounts, while the anti-Erdogan group had a 28.8% higher retweet rate for CHP-affiliated accounts, indicating strong party-based media engagement.