How conservative are young German voters?

A lesson in how we narrativize election results

Last Sunday’s European Election results in Germany were disturbing enough to be noticed even in corners of the internet that do not usually take a ton of interest in election results in Germany. That’s because they seemed bad in a way that speaks a very clear language. 

I shared a version of the following graph – election results in the state of Saxony, with the two posts listed below – and I stand by my overall point.

Still, a lot of people, especially from the area, replied to my Tweet, agreeing with me in principle, but pointing out that the numbers were preliminary and likely to change quite a bit. The change in the end wasn’t significant enough to make me want to take down the posts entirely—but their intervention got me interested in the way the numbers emerging from the European election seemed to predetermine how the broader political and media conversation about the results would be conducted, where it would focus and what it would pointedly avoid. 

There were two major takeaways: truly staggering numbers for the far-right AfD in the East German Länder (Brandenburg, Mecklenburg-Vorpommern, Saxony, Saxony-Anhalt and Thuringia). And the Green Party facing massive desertions among the youngest voting block (16-24 year-olds, 16-18 year-olds getting to vote for the first time in a European election in Germany ever), and the AfD gaining significantly in that group.

Probably because the first takeaway was just so staggeringly depressing, people tended to fasten on to the second one. To be very clear: an 11 percentage point gain for the AfD among young people is fucking horrible. This post is not intended to debunk that away. It is intended to think through what was done with these numbers in public debate in the immediate aftermath of the election — and how the framing in German media might give us some clues about the dynamics that brought about this result in the first place.

From the very beginning I was noticing two things: First, an astonishing number of interpretations quickly got attached to these numbers. And second, seasoned reporters in Germany didn’t seem to believe them. It was a classic case of watching a media narrative taking shape, while people who follow election results for a living kept trying to point out that it was probably to early to draw specific conclusions beyond “the Far Right" gained a lot of votes”. In general, what seems to have happened is that the large cities in East Germany — Leipzig, Dresden, etc. — reported results slightly later, meaning that the AfD outperformed in the early count. Something that US election-viewers will be only too familiar with. The youth vote figure too found the AfD’s success significantly overstated, though more of the vote coming in didn’t really help mitigate the Greens’ losses.

It’s certainly a phenomenon that begs for an explanation – and it has occasioned a ton of attempts at one since the election. After all, these are the very kids that were supposed to have been radicalized into climate action, politicized by Fridays for Future, the school strikes and Greta Thunberg. You can imagine why there was a  lot riding on the answer to what happened. Not just for the Greens, obviously, but also for media types trying to make sense of this election.

Unfortunately, the way the election results entered into the media discourse — i.e. the fact that the AfD seemed to overperform at first, when in fact it ended up underperforming its recent polling —, has led to some pretty egregious motivated reasoning. For this piece I spoke with Mark Schieritz who covers media for Die Zeit, and who was one of the first people on Twitter on election night telling people to pump the brakes. In the event, no one pumped the breaks. How could they? You have to cover the election somehow, and writing an analysis, inviting experts, etc. takes time. You have to make a judgment one way or the other. Still, the snap judgments themselves are kind of interesting, because they tell us a lot about the priors specific media actors bring to an evolving story: what their first instinct is, what they look for as the numbers come in, where their mind goes when the numbers start telling a certain kind of story.

Here are the numbers for the voting cohort 16-24:

Mark Schieritz and Dana Hajek wrote a very enlightening explainer why the emerging consensus was wrong, and they did a very good job suggesting what that consensus was. The narrative was, in their formula, that “Generation Greta” had turned into “Generation Gauland” (after the founder of the AfD). The idea was, in other words, that somehow the voters the AfD gained were the ones the Greens lost. And this turns out not to be true: Quite literally we know that a fraction of the voters the Greens lost went to the AfD, though perhaps observers didn’t mean it quite that literally. But even the generational shift is a little hard to pin down: 16-24 year olds were not less likely to vote for the AfD, they voted for them in exactly the proportion as the overall population. But the AfD’s strength lay in different, older age cohorts — and their biggest weakness, by the way, was among the very old.

I also want to emphasize that this is the first result of an election in which 16-18 year olds could vote, that is to say people who are, in Germany, overwhelmingly likely to (i) live with their parents, (ii) to still live wherever they grew up and (iii) to be ensconced in a fairly homogeneous community. Since these age brackets haven’t yet been broken down further (I believe this will happen down the line), it’s hard to say, but there’s of course a chance that the votes of 16-18 year olds still reflect their parents’ opinions in some measure alongside their own, evolving opinions. Two other things to note is that 16-18 year olds in Germany are probably more equitably distributed geographically than demographics that have to move for training/work; and that the data aren’t broken down by gender, which would be fascinating to see. My suspicion in both cases is that the age breakdown basically shows what the other vectors already show: that if you live in East Germany, you are far more likely to vote for the AfD.

Finally, it’s important to note that, when it comes to the European Elections, Germany does not have its usual “5 percent threshold” meant to keep small parties out of parliament. Meaning: EU election voters in Germany know that they can vote for tiny parties and still possibly send a few folks to Brussels. And that’s what they do, and the young do it at disproportionate numbers.

Whether it’s the “Animal Rights Party” or the satirical “The Party”, or whether it’s the new offshoot from Die Linke headed by charismatic anti-woke leftist Sahra Wagenknecht (called “Bündnis Sahra Wagenknecht”, just so you don’t have to memorize too many names), the young went for a bunch of parties that are basically a wasted vote in national elections. As Hajek and Schieritz point out, it’s especially the pro-EU party Volt (the only party to my knowledge that runs an EU-wide campaign as a unified European party) that attracted the youth vote — a full 9% of the vote, which is as much as the SPD, Chancellor Scholz’s party! 

Hajek and Schieritz point out that Volt “is a pro-European mini-party which calls for more renewable energy, slashing carbon emissions from industry, and wants a federalized EU.” In other words, “it’s a program that is in some senses greener than the Grens”; what is more, Volt is also younger: Volt’s members “have an average age of 35 years, compared to 48 years for the Greens.” I’ll have more to say about the questions people are and aren’t asking about Volt in analyzing these results, but on the face of it, Hajek’s and Schieritz’s point seems intuitively right. Why vote for a party that, as a member of a not-particularly-popular government, has been busy watering down their profile on the issues that drew voters to them in the first place, when there’s something on offer that promises the original article?

So the first thing to note is that young people don’t appear to have swung that far to the right. They seem to have splintered, but not in a way that is electorally ineffective. A vote for Volt is not like a vote for Jill Stein or RFK Jr.! This party will have seats in the European Parliament, and it will have a say in how Europe is governed in the next few years. But it’s noticeable that the generation that was busy analyzing those results has a very different attitude to party affiliation than the young do. Meaning, they were unlikely to see this as a perfectly reasonable strategic shift, and instead saw it as part of a shift in ideological alignment. Maybe it is that, maybe the Greens cannot recapture the people who voted for Volt this time around — but these numbers alone don’t really suggest that.

The second instinct was: TikTok! “All across Europe, the far right made gains,” the TV-magazine Kulturzeit (a co-production between a German, Austrian and Swiss channel) began its show on June 10. “How much does generation TikTok have to do with that?” The framing here is super interesting because while Kulturzeit emphasized that the Far Right had made gains “across Europe”, the show mostly focused on France, Italy and Germany. The fact that in Spain, Portugal and the Scandinavian countries, as well as Poland, the results looked very different, was acknowledged, but really didn’t become part of the framing. And part of that strange disconnect seemed to have to do with the youth vote: “The EU did not manage to convince especially young voters of the importance of a united Europe in these times.” That seems odd: anti-EU parties did not score “especially” well with “young voters” — it is true that young voters didn’t act as a failsafe when their elders rushed to the right. But unlike in France, where the Rassemblement National (RN) scored about 30 percent of the youth vote, the AfD’s strength lay elsewhere, the young just didn’t make up for it.

“What was decisive was [the AfD's] presence in social media networks like TikTok. Their deliberate play on the keyboard of political emotions especially appeals to you voters.”

“Decisive” and “especially” here have the distinction of being both absolutely accurate and kind of misleading. Insofar as we’re talking about recruitment of voters via TikTok, it’s quite likely the app had more of an impact on 17 year olds than, say, on 80-somethings. But there is a narrowing of the narrative focus here that risks distorting what actually happened. And pretending we know how social media influence people’s voting — but only for very specific populations.

There had been pieces in German media for weeks about how the AfD’s ground game on TikTok was way better than that of other parties. The interviews in Kulturzeit point to their use of memes, their effective use of short clips, their relentless information warfare. Der Spiegel did a massive cover story on TikTok the week before the election — these stories come together well ahead of time, so this wasn’t necessarily about the EU election. But the framing of course could not help become infected by the framing that — as Der Spiegel was putting the finishing touches on the new issue — was already taking shape.

It seems right that the Far Right is pushing a lot of content on TikTok, though, as longtime readers of this newsletter know, I tend to be weary of these sorts of assessments. Der Spiegel’s title story in some way exemplifies the problem: they basically profile and interview users who get a lot of these videos and report seeing horrible comments on them. Then the authors say — accurately — that since TikTok is extremely stingy with disclosure/API/etc., there’s no way to say for sure how widespread this phenomenon is and what effect it has.

I would never deny the effectiveness of some of these strategies, but at the same time I worry about overemphasizing them. It all reminds me fatally on the US-fixation on the “alt right” in the wake of the Trump election — in fact memes, use of media, and disinformation are basically the three things Angela Nagle posited as the central strategies of an ascendant right in Kill All Normies. I would never deny that these sorts of voters exist, and there’s clearly a kind of right-coded libertarian that is particularly easily targeted on social media platforms that all promote some kind of grindset-ideology. But Nagle’s argument made sense as an explanation for a very specific set of political actors; it didn’t really hold up that well for an explanation for large-scale shifts in voting pattern (I haven’t read the book in years, but my recollection is that that wasn’t an argument she was making — but it was one that others kept projecting onto the book.) So yes, it’s easy to imagine a kind of person who this kind of TikTok campaign really works for, and it’s at least possible to imagine that this person is persuadable, and wasn’t gonna vote for the AfD to begin with. But is TikTok the most likely explanation for voting changes of the magnitude we were talking about above? I worry that this focus tends to exoticize what are inescapably shifts that happen within the existing parties and electorates: focusing on outside influence, outside technologies and new voters.

I should explain why I am always a little bit skeptical of this take. TikTok is even more of a black box than your average social media giant, at least when I’m on it it seems rife with coordinated inauthentic interactions. If something really blows up in your Facebook feed or gets ton of engagement on Twitter, that is fairly likely to be indicative of something. But on TikTok, it’s just so much harder to say what that something is. So we might be able to say: this is a well-produced video and it seems to get a ton of engagement. But how much, of what kind, and what bearing this engagement has on things that happen offline, is of course a harder thing to measure.

Moreover, there is a tendency when we talk about disinformation to assume that we can infer from someone’s media diet what they draw from that diet. But that’s of course not true, in both directions. People can get Twitter-brain who don’t actually spend that much time on Twitter. The number of people in the US who reflect what we might describe as a Fox News worldview far outstrips the people who actually watch Fox News. And Fox News viewers put what they get on the network to a variety of purposes. That doesn’t mean that we shouldn’t criticize disinformation when we see it, and annoy our relatives about it. But it is interesting, for instance, that we look at 17 year-olds voting in a way that seems incomprehensible to us and we think “TikTok!” — but we see their 35 year-old elders vote for the same party in greater numbers (20 percent versus 16 percent), and no one goes “Telegram!” Media reports tend to credit the role of disinformation a lot more the further you move away from whatever media is doing the reporting. After all, this media is the one informing you about the disinformation — others might be dupes, but you are well-informed and getting your news the way you’re supposed to.

Finally, as readers of this newsletter, you will not be shocked to hear that some brainiacs decided that the reason for the decline in the Green vote was — drumroll, please — wokeness.

These are both fascinating Tweets. The first one, by frequent talk-show guest Ahmad Mansour, says:

“Could it be that the horrifying dimensions of antisemitic excesses on Europe’s streets and universities, the partnership between postcolonial identity politics and Islamists, and the open sympathy for islamic terror groups, scared off, annoyed and shocked enough people that they doubted the ability of left wing parties to grapple with those problems adequately?”

It’s a rich text, as they say. There’s the fact that this rehashes the basic dynamic of right wing moral panics the world over — taking “postcolonial identity politics” and campus controversies as givens and pretending that it is them, and not their relentless discussion in media, that is shaping public opinion. But there’s a bigger problem here: the tweet invites us to identify the Greens and the SPD (the parties that lost the most significant support on the left) with a positions that they not only do not hold, but that they’ve spent the last eight months calling out. If anything, the most visible politicians among the Greens and the SPD have been busy joining the chorus calling out “the left” for its “postcolonial identity politics” etc. Sure, they may not be using the word “woke” to do so, but they’ve fully joined in the language game. Mansour is actually half correct here in that he seems to sense what the leadership of the Greens does not seem to have: that even if they joined this moral panic, it was designed to damage them, and it would do exactly that. But he’s wrong on the merits: if you’re mad about “postcolonial identity politics” (that scourge of our time), then you’re just fine voting for the Greens, the SPD, etc. It’s this interesting dynamic where “woke” has no constituency, but then gets identified with vast swaths of the political spectrum. It can shrink or stretch to fit however far you want to spray your resentments.

In some way it’s inevitable that anti-woke grifters would look at any election result and say “hey, you know what this reminds me of?”

But the very roteness of the “wokeness”-argument highlights an omission: From a US vantage point it’s noticeable just how resoundingly pundits refuse to entertain the opposite possibility. I won’t claim that this is true, but I will say that I’ve barely seen this raised in German public discourse — something Mark Schieritz confirmed for me. Here’s what this theory of the case would look like: The other thing that’s true of Germans 16-24 is that they’re way more diverse than any other age cohort in Germany that can vote. Many of them are of Arabic descent, a fact of which Germany takes pains to remind them whenever possible. And one of the generational divides may well have to do with Israel/Palestine. There’s shockingly little data on the issue, but here’s what we can say: The Greens have remained staunchly supportive of the Israeli war effort in Gaza; they have as a member of the governing coalition ensured material and financial support of the effort; and they have not been shy decrying criticisms of Israel’s war as antisemitic. 

All of which, we should note, Volt did not, or not as publicly. Not that they were necessarily critical of Israel (their party program for the European election doesn’t mention the conflict, and why should it), but their main (and most visible) foreign policy demand is a meta-demand: creating a robust EU foreign ministry that would centralize European efforts to “stand up for those in need of protection and shape a future in which the [EU] stands for positive change.” This includes “climate-positive projects”, “ensur[ing] climate justice”, “promot[ing] the economies of non-European countries” in order to combat the causes of displacement, “promot[ing] a strong awareness of the legacies of colonialism and imperialism” among member states, and establishing a “feminist foreign policy.” 

All of which, just as a side note, sounds really good! The point is that this is essentially a foreign policy that has everything a Green voter would like – minus the baggage that comes with taking a very strong stand in favor Israel’s actions in the Gaza war, which was something the Greens, holding the Foreign Ministry, basically had to do. Or felt like they had to do.

The reason I mention this is that there’s some evidence that the Greens and their own voters don’t exactly see eye-to-eye on this. There’s only one poll asking voters to identify their party preference and then answer the question “Should Germany put more pressure on Israel in order to put an end to the war in Gaza?” Among Greens, the number answering “yes” was 69%, compared to 67% for Die Linke, 60% for the liberal FDP, 59% for the SPD, 59% for the new Bündnis Sahra Wagenknecht, 53 % for the CDU/CSU, and 37% of AfD supporters. This is the only poll of this kind I found, and it of course doesn’t break things down by age cohort. And I admit, it would require further explanation why this kind of divide should be important for the Greens’ electoral outcome and not for the SPD (69% is higher than 59%, but it’s not exactly a massive difference).

But I think these separate pieces of evidence, when taken together, are quite suggestive: A disproportionately diverse German voting population (with many voters from Middle Eastern countries); a Green voter profile that, on average, wants a tougher stance on Israel’s conduct of the war compared to what the Greens have demanded; and Green politicians – from Vice Chancellor Robert Habeck via Foreign Secretary Annalena Baerbock – who are very much identified with the pro-Israel Staatsräson. 

So perhaps it’s not that the “wokeness” put them off (though it clearly puts off newspaper writers who write about “wokeness” — who are, however, usually long past 16-24 years old). Perhaps it’s the fact that the Greens in fact went pretty hard on identification with Israel in the Gaza War. I’ve presented what evidence we have to support that idea — and it ain’t that much, to be quite honest. But it’s a bit more than we have for a bunch of the narratives that did take hold — from the “they abandoned the left for the AfD”, via “they’re no longer Generation Greta”, to “it’s the wokeness done did it” and “ban Tiktok”. Which I think suggests that media framing both helps interpret election results — and to some extent helps bring them about. And that this is particularly true for young voters and activists, whose access to media will frequently run through people of very, very different age cohorts.