You wonder how the CDU/CSU playing with the "Brandmauer" to AfD last week has influenced German voters? @mafaldapratas.bsky.social @jmfernandes86.bsky.social & I have you covered. We surveyed sample of 2,000 German voters. 🧵with first results 👇. Spoiler: It did not go down as well as CDU/CSU hoped. - ThreadSky

chriswratil.bsky.social • 20 days ago

You wonder how the CDU/CSU playing with the "Brandmauer" to AfD last week has influenced German voters? @mafaldapratas.bsky.social @jmfernandes86.bsky.social & I have you covered. We surveyed sample of 2,000 German voters. 🧵with first results 👇. Spoiler: It did not go down as well as CDU/CSU hoped.

Comments

grinsgefal.bsky.social•19 days ago

Ein Blick auf die jüngste Geschichte vieler Länder zeigt, dass wir uns zwischen zwei Übeln entscheiden müssen. Entweder man lässt die Brandmauer intakt und die Unzufriedenheit wächst, oder man arbeitet mit ihnen zusammen und setzt eine Politik durch, die den Rechtsstaat beeinträchtigt.

thomas.escritt.eu•20 days ago

Fascinating! Is there any reason why that doesn't seem to be reflected in the polls yet?

johanneszeller.bsky.social•20 days ago

It doesn’t? All the changes in voter intentions that are demonstrated here are very minor. I would say that matches exactly what we see in the polls: no noticeable movement in either direction.

thomas.escritt.eu•20 days ago

Fair enough - I guess I was reading this data as suggesting there'd be a slight headline shift away from the CDUCSU in polls, but the evidence on that is very contradictory.

klauspforr.bsky.social•20 days ago

Das hypothetische kontrafaktische Design erscheint zweifelhaft

chriswratil.bsky.social•20 days ago

Wie hätten Sie es denn gemacht? Uns ist klar, dass es hier keine gute Lösung gibt. Wir orientieren uns an Graham and Coppock (2021) in POQ.

klauspforr.bsky.social•20 days ago

die @gles.bsky.social rolling cross section studie bietet m.e. ein besseres design. @axburg.bsky.social und @joross.bsky.social wissen hier besser bescheid https://www.gesis.org/gles/news-details/article/fragebogenveroeffentlichung-frageprogramm-der-gles-rolling-cross-section-2025-za101001-version-02-ist-verfuegbar

chriswratil.bsky.social•20 days ago

Dann bin ich gespannt auf die Ergebnisse. Wir wollten halt ganz bestimmte Fragen stellen (wir haben auch noch offene zu dem Thema) und auch auf intra-individual correlates of change schauen.

klauspforr.bsky.social•20 days ago

Wobei man bei diesem Problem ja mit rolling Cross-section Design ziemlich weit kommt, bzw Probleme der selektiven Panel Attrition vermeidet

chriswratil.bsky.social•20 days ago

We tell respondents about the events last week in a neutral news-based way & ask them whether they had heard about them previously (94% say "yes"). Among other things, we then measure propensity to vote (PTV, 1-7 scale) for all major parties & a counterfactual PTV had the events *not* happened. 1/10

chriswratil.bsky.social•20 days ago

All results are from a subset of 1785 respondents who declare themselves willing to report their counterfactual vote choice (other refuse or pay no attention). The histograms show distribution of change in PTV between reality & counterfactual for CDU/CSU & AfD. Both distributions peak on "-1". 2/10

chriswratil.bsky.social•20 days ago

Korrektur hierzu: Wir hatten "no change" (0) nicht entfernt & ich Grafik falsch interpretiert. Nun nochmals die Histogramme ohne Personen, deren PTV für AfD oder Union unverändert belibt. Das spiegelt auch klarer Mittelwertunterscheide in folgenden Grafiken wider. 🙏 Danke @danbischof.bsky.social

klarapernsteiner.bsky.social•20 days ago

Additionally, these graphs allow a more detailed look into how the PTV-score of CDU/CSU changed specifically for voters of SPD and Greens. 41% of Green voters and 38% of SPD voters report that their PTV for CDU/CSU would be higher, had the accomodation not happend @thomasmmeyer.bsky.social

chriswratil.bsky.social•20 days ago

Next, we consider overall PTV differences for all parties. First graph shows that only CDU/CSU have a small downward movement on average. But this is not big, it can be something like some people moving from 7 to 6 on PTV scale. The other two figures subset by vote in EP elections last June. 3/10

chriswratil.bsky.social•20 days ago

It shows that CDU/CSU loses among their voters last year, but gains slightly among AfD voters in the EP elections. Note however, that most AfD voters are very set on AfD & movement is small. The overall picture is below: CDU loses among own, wins among AfD & loses with voters of other parties. 4/10

samuelbrielmaier.bsky.social•20 days ago

thanks for sharing these results. The differences aren’t significant, right? I‘d be curious how @tabouchadi.bsky.social looks at these - do they show a hint of contradiction to the „voters prefer the original“ argument?

chriswratil.bsky.social•20 days ago

We have not tested for statistical significance yet, but note that the significance of the *difference* between the means is not indicated by the non-overlap of these confidence intervals for the separate means.

pluggedchris.bsky.social•20 days ago

Thanks, very interesting and highly relevant! Could be good news for democracy if CDU wins from AfD and loses to the left with such a strategy?

chriswratil.bsky.social•20 days ago

Real interesting are subsets by other factors. CDU/CSU ONLY loses in Western Germany, East is not affected on average. It gains among people who think that restricting immigration is one of two most important issues (MIP), while it loses among all others with other priorities. Finally, ... 5/10

francesconicoli.bsky.social•20 days ago

One point though. Merz might have willingly lost some votes to increase his post election leverage. By keeping a 2-party coalition with AfD on the table, he enormously increases his negotiating leverage with spd and greens after the election.

juliusk.bsky.social•20 days ago

I knew that the outcome here is PTV, but it would be super interesting to explore whether, compared to a counterfactual scenario, East German CDU voters have become more willing to support a coalition with the AfD as a result of shifts in party elites’ behavior.

chriswratil.bsky.social•20 days ago

But this is potentially endogenous, of course, we ask people whether they think 🇩🇪 has a special responsibility in dealing with extreme right parties, given its history. People who see this possibility move away from CDU/CSU, those that do not, move towards it! We also checked how many people... 6/10

bijanmoini.de•20 days ago

CDU/CSU‘s logic probably is that they would have lost voters to the AfD had they not acted the way they did (push even more extreme demands, effectively cooperate with the AfD instead of toning down/rationalize the issue). Is there any way to confirm or disprove this?

chriswratil.bsky.social•20 days ago

No, not with this data, at least.

Comments

Posting Rules

Reply