Christina Gahn

@christinagahn

Postdoctoral Researcher at University of Vienna. Dr in Political Science from Humboldt University of Berlin Political science, political communication, electoral campaigns, targeting, party competition, public opinion polls

670 Siguiendo1357 Seguidores192 Publicaciones

Bluesky

¿Es tu perfil?Reclámalo y christinagahn.dirty.social es tuyo.Reclamar perfil

Christina Gahn@christinagahn· 14d

We asked voters to donate the real political ads they got on their phones during the 2021 German election and the 2022 US midterms. We find that microtargeting works largely through perceived tailoring. An ad from a party you support feels more relevant, and these are evaluated more positively. 🧵2/3

Diagram of the two-country data collection. On the left, the United States, on the right, Germany, each with its national flag. For each country a two-week timeline leading up to election day: in the US from October 27 to election day on November 8, 2022, and in Germany from September 13 to election day on September 26, 2021. Over these two weeks, participants repeatedly filled in questionnaires on their smartphones and uploaded screenshots of the political ads they had received, up to three per day in the US and up to one per day in Germany.

Christina Gahn@christinagahn· 14d

🥳 New Paper with @jadevrielink.bsky.social and @skruikemeier.bsky.social: on how voters perceive the political ads they receive, and why that tailoring perception matters more than we usually assume. 🧵 1/3 📄 doi.org/10.1080/1369...

Hand-drawn doodle illustration in three steps. On the left, a large smartphone with "VOTE for me" on the screen, surrounded by small icons such as animals and speech bubbles. An arrow points to a smiling person looking thoughtful, with thought bubbles above showing the same motifs (pets, music, people). Another arrow points to the same person, now beaming, holding up the smartphone with the "VOTE for me" ad and giving a thumbs up. The illustration shows how political advertising that feels personally relevant is evaluated more positively.

Christina Gahn@christinagahn· 27d

Hier das ganze Interview, vielen Dank nochmal @irisbonavida.bsky.social! 📰📊 Zum Paper mit @krausewe.bsky.social: doi.org/10.1086/742632

Zeitungsseite der Salzburger Nachrichten vom 21. Juli 2026, Seite 4, Ressort Österreich. Interview mit Christina Gahn unter dem Titel „Die Macht der Umfragen". Großes Foto: eine Hand wirft einen gefalteten Stimmzettel in eine Wahlurne. Rechts unten ein Porträtfoto von Christina Gahn mit dem hervorgehobenen Zitat: „Ein Problem ist nur, wenn taktisches Wählen auf falschen Annahmen beruht."

Christina Gahn@christinagahn· Jun 17

In terms of issue distribution, while also preliminary (both data + classification), see the descriptive plot attached. It shows both variation in terms of composition and breath of issues, and adaption to the local context (eg housing).

Christina Gahn@christinagahn· Jun 17

Our first results show that parties use posters for mobilization, near precincts where they were strong in the last election. But we have many more plans with this data! 🗺️ Explore the posters in this interactive map: christinagahn.github.io/posters-berl... & let me know if you want to meet up! ☕️

Academic paper title page and abstract. Title: "The Geography of Electioneering in Multi-Party Systems: Evidence from Campaign Poster Locations in Central Berlin." Authors: Christina Gahn (University of Vienna) and Tristan Klingelhöfer (The Hebrew University of Jerusalem). Version: June 11, 2026. The abstract describes a study of geolocated campaign poster locations in Friedrichshain-Kreuzberg, Berlin, collected during the 2025 German federal election. Key findings: consistent evidence for mobilization and intra-bloc persuasion strategies in poster placement.

Christina Gahn@christinagahn· Jun 17

Looking forward to presenting at #EPSS2026 in Belfast this week! @klingelt.bsky.social and I hand-collected 10,000+ campaign posters in Berlin and use automated image analysis to study placement and content strategies across parties THU, 5:10–6:50pm, Brookfield (Hilton) @epssnet.bsky.social

Map showing the spatial distribution of campaign posters by party across electoral precincts in Friedrichshain-Kreuzberg, Berlin. Nine panels display poster locations (dots) overlaid on precinct-level 2021 vote shares (shading) for: Die Linke (n=1,129), Die Grünen (n=1,459), SPD+JUSOS (n=693), CDU (n=639), AfD (n=107), FDP (n=348), Volt (n=780), and BSW (n=248). Each panel uses the party's color. Based on automated classification of 5,854 detected posters (conference version, ~50% of full corpus).

Christina Gahn@christinagahn· May 15

🔍 Key finding: parties rarely combine targeting & tailoring. Targeting boosts ad impressions (but with diminishing returns at high degrees), but tailoring or combining them doesn't. Campaigns are more selective and pragmatic than often assumed, and we propose tools to study them properly. 4/4

The figure shows three heatmap charts that summarize political ad counts and mean spending in euro by targeting and tailoring. Each chart has 3 rows labeled mis-targeted, targeted, and not targeted, and 2 columns labeled not tailored and tailored. In the Region chart, not tailored and not targeted contains n=1934 and 268 euro. Tailored and not targeted contains n=26 and 304 euro. Not tailored and targeted shows n=2552 and 546 euro. Tailored and targeted shows n=188 and 961 euro. Mis-targeted and tailored shows n=55 and 217 euro. In the Women chart, not tailored and not targeted contains n=4234 and 452 euro. Tailored and not targeted contains n=20 and 722 euro. Not tailored and targeted shows n=242 and 470 euro. Tailored and targeted shows n=11 and 141 euro. Mis-targeted and tailored shows n=5 and 50 euro. In the Men chart, not tailored and not targeted contains n=4254 and 453 euro. Not tailored and targeted shows n=253 and 278 euro. All tailored cells for Men show n=0 and 0 euro. A horizontal color scale below all charts is labeled Mean spending on ads and ranges from 0 to 3000 euro, with darker shading indicating higher values.

Christina Gahn@christinagahn· May 15

🇩🇪 We apply our framework to all 4,760 Facebook & Instagram ads by German parties in the final 4 weeks of the 2021 federal election - combining manual content analysis with Meta's Ad Targeting Dataset to measure both strategies independently. 3/4

The figure shows a grouped bar chart comparing how many political ads use different numbers of criteria for tailoring, targeting, and matching. The horizontal axis is labeled Amount of criteria and has 5 tick marks at 0, 1, 2, 3, and 4 criteria. The vertical axis is labeled Number of ads and runs from 0 to 4500 with tick marks every 500. For 0 criteria, tailored ads total 889, targeted ads total 1322, and matched ads total 4276. For 1 criterion, tailored ads total 3605, targeted ads total 2696, and matched ads total 465. For 2 criteria, tailored ads total 265, targeted ads total 642, and matched ads total 19. For 3 criteria, tailored ads total 1, targeted ads total 94, and matched ads total 0. For 4 criteria, tailored ads total 0, targeted ads total 6, and matched ads total 0. A legend below the chart identifies tailored, targeted, and matched as the three bar groups. Overall, most ads cluster at 0 (for matching) or 1 (for targeting and tailoring) criterion, and very few ads use 3 or 4 criteria.

Christina Gahn@christinagahn· May 15

📢 Despite often being used interchangeably, targeting (deciding who sees an ad) and tailoring (adapting the ad content to the audience) are distinct strategies. Ads can be targeted-but-not-tailored, tailored-but-not-targeted, or both - and research needs to account for that. 2/4

Christina Gahn@christinagahn· Apr 22

Most research on voting advice applications (VAAs) focuses only on the top match. @belad.bsky.social looks at effects across the whole party system and finds that VAA use does shift party preferences, but not in ways that consistently increase or decrease political polarization.

Christina Gahn@christinagahn· Apr 22

Does social media spread lone-actor terrorism? Fabian Feilmeier's analysis of 16 democracies (2013–2020) finds the opposite: higher social media penetration is actually associated with slightly lower attack rates. Individual radicalization pathways don't simply scale up to the national level.

Christina Gahn@christinagahn· Apr 22

Why are young women and young men increasingly voting differently? Benedikt Unkel digs into the 2025 German federal election and finds that feeling "left behind" and differing attitudes toward gender issues both help explain the growing gender gap in support for the populist radical right.

Christina Gahn@christinagahn· Apr 22

Not all civil war endings are equal. @benjaminhofer.bsky.social analyzes 414 terminated intrastate conflict episodes worldwide and shows that conflicts fizzling out through "low activity" (rather than peace agreements or military victory) much more likely to recur within five years.

Christina Gahn@christinagahn· Apr 22

Social isolation and populism are both on the rise, but are they connected? Using ESS data from EU member states, Veronika Wimmer finds that higher social isolation predicts voting for right-wing populist parties, but not for left-wing ones.

Christina Gahn@christinagahn· Apr 22

Do people who feel politically powerless see immigration as more threatening? Anna Treutlein, using ESS data from Germany and creating an immigration threat index, finds that lower internal political efficacy is linked to higher immigration threat perceptions, but political attitudes matter too.

Christina Gahn@christinagahn· Nov 20

Using geolocated data from 10,000+ posters in Friedrichshain-Kreuzberg during the 2025 German federal election, we explore where parties place posters - and why. Preliminary results suggest that parties mostly use them for mobilization. Looking forward to all feedback! Thanks for inviting me!

Christina Gahn@christinagahn· Nov 20

I‘m excited to take part in the @zweitstimme.bsky.social Workshop on Elections, Public Opinion & Elite Behavior at the @hertieschool.bsky.social this week! I'll be presenting first results from a project with @klingelt.bsky.social on the Geography of Electioneering in Multi-Party Systems.

Christina Gahn@christinagahn· Nov 19

Im Oktober erschien in den Salzburger Nachrichten (@sn.at) ein Artikel von @irisbonavida.bsky.social zu personalisierter Onlinewerbung mit Kommentaren von mir. Darin verweise ich auf aktuelle Forschung von Kolleg*innen sowie auf Ergebnisse meines eigenen Experiments: doi.org/10.1177/1940...

Christina Gahn@christinagahn· Sep 10

… and on Saturday at 12pm PDT I’ll present joint work with @meinungsfuehrer.bsky.social on targeting issue messages in online campaigns 🎯 It’s part of a panel on Political Parties’ Online Communication Strategies in Election Campaigns organized by @zeynsom.bsky.social 👇👀 tinyurl.com/ypujwc5l

Christina Gahn@christinagahn· Aug 13

Die Regierungsbildung: 🤝 🕒 Nach 155 Tagen (der längsten in der öst. Geschichte) bildeten ÖVP, SPÖ und NEOS die erste 3-Parteien-Koalition seit 1947. ❌ Die FPÖ, obwohl stärkste Partei, wurde von der Regierung ausgeschlossen. ⚖️ Die neue Koalition fand hohe Zustimmung bei ihren Unterstützer:innen.

Christina Gahn@christinagahn· Aug 13

Die Ergebnisse: 📊 🇦🇹 FPÖ erstmals stärkste Partei (Erholung von Verlusten 2019), ÖVP und Grüne mit starken Verlusten. 💉 Besseres FPÖ-Ergebnis in Regionen mit niedrigen Impfquoten, aber Pandemie spielte nur eine begrenzte Rolle. 🌆 Städtisch-ländliche und demografische Spaltung vertieft.

Christina Gahn@christinagahn· Aug 13

Der Wahlkampf: 📢 😴 Trotz turbulenter Vorgeschichte verlief der Wahlkampf 2024 überraschend unspektakulär. ♻️ Wichtige Themen für Wähler:innen waren Inflation, Migration, Sozialpolitik und Klima. 🌊 Eine schwere Überschwemmung lenkte den Fokus auf Krisenmanagement und verstärkte die Klimadebatte.

Christina Gahn@christinagahn· Aug 13

Wichtigste Ergebnisse: 🧐 💡 Migration wurde wieder zu einem zentralen Thema für Wähler:innen. 💡 Die FPÖ wurde erstmals stärkste Partei, blieb aber außerhalb der Regierung. 💡 Nach rekordlangen Verhandlungen bildeten ÖVP, SPÖ & NEOS die erste 3-Parteien-Koalition seit 1947. Mehr Details im Thread 👇🧵

Christina Gahn@christinagahn· Jun 27

Packed panel kicked off by @lenamariahuber.bsky.social on group identities in political competition. 🙌 #EPSA2025 @epsanet.bsky.social

Christina Gahn@christinagahn· Jun 24

Preliminary findings: 1️⃣ Solo-male authorship is declining, with mixed-gender teams now the norm. 2️⃣ All-female authorship plateaued after the mid-2000s. 3️⃣ Gender disparities persist, especially in specific sub-fields. More details (and plenty of graphs!) in my presentation—swing by! 📊✨

Proportion of Papers by sub-fields and Team Type

Christina Gahn@christinagahn· Jun 24

Excited to kick off #EPSA2025 on Day 1, Panel 1! 🎉 I’ll present my co-authored research with @michaelimre.bsky.social on 50 years of gender representation in political science publications. We analyze nearly 100,000 articles across 72 journals to uncover long-term trends. 🔍 @epsanet.bsky.social

Proportion of Papers by Year and Team Type

Christina Gahn@christinagahn· Feb 12

There’s a strong assumption in academia that working overtime is the norm. But I don’t think that has to be true. While I had some intense months, I also took vacations and treated most weekends seriously.

Christina Gahn@christinagahn· Feb 12

Since defending my thesis two weeks ago, I have reflected on my time spent in the last five years (tracked by a time-tracking tool): ⏱️ Main Results on Average:   - 35% Dissertation + Training - 10% Other Research  - 25% Teaching  - 18% Admin + Service   - 10% Feedback/Conferences 🧵

Christina Gahn@christinagahn· Feb 6

Heute auf der Titelseite der @szde.bsky.social: die neue (Vorab-)Studie von @krausewe.bsky.social (Universität Potsdam) und mir (@stawi-univie.bsky.social @univie.ac.at) zu dem Einfluss von Umfragen auf den Erfolg kleiner Parteien. 📊

Christina Gahn@christinagahn· Jan 29

I can’t believe it yet but I‘m #PhDone! A special thanks to my committee @heikekluever.bsky.social @thomasmmeyer.bsky.social @anselmhager.bsky.social @markuskollberg.bsky.social & @tiondelisa.bsky.social! I learned a lot and I am truly grateful for the experience @humboldtuni.bsky.social 🎓🥳