There is a threat of an election earthquake in the east that could shake all of Germany, headlined Bild Online. The headline sounds current, but it is five years old, published on June 6, 2021, the day of the last state election in Saxony-Anhalt. It was said that the AfD could become the strongest party. That would have been a sensation back then. It didn't fit at all with the poll by the Elections Research Group, which saw the CDU seven points ahead of the AfD shortly before the election.
Around the same time, Bild and Spiegel suddenly had completely different numbers: The online institute Civey (CDU 29, AfD 28 percent) and the Bild partner INSA (CDU 27, AfD 26 percent) each saw the CDU's lead at only a single point and thus led to the expectation of a head-to-head race. The Republic was shocked.
That subsided on election day. Many people no longer understood the world: the CDU won with 37.1 percent, the AfD fell back to 20.8 percent. Pre-election polls and the results have never been so far apart.
Can this also happen in 2026?
Even today, Saxony-Anhalt moves the entire republic. The AfD is likely to become the strongest party in the state elections next Sunday. Depending on the institute, it is currently given at 40 to 43 percent and is therefore well ahead of the CDU. If the AfD gains a little more, it could win an absolute majority of seats for the first time. But what is important is the subjunctive, the possible form. In eleven different surveys since the beginning of the year, not a single one shows such a mandate majority.
Can that still change?
Basically yes. Even if it doesn't seem likely according to the polls. But surveys are not predictions and are not meant to be that way. They measure the political mood at the time of the survey. In addition, the numbers have a statistical fluctuation range, which some institutes also report. Infratest dimap puts it at two to three percentage points for the latest Saxony-Anhalt survey. This means: If a party is shown with 40 percent, it is very likely between 37 and 43 percent at the time of the survey. That's a wide window.
In addition, a surprising number of voters only decide on election day whether or who they will vote for. Differences between surveys and election results are therefore completely normal.
Where did the big differences come from in 2021?
In 2021 in Saxony-Anhalt the gap was still greater than in all other elections before and after. In the end, the CDU was more than 16 points behind the AfD. A trend for the CDU emerged in the days before the election, but not to this extent. At the Elections Research Group, the lead had grown to seven points by Thursday. Infratest dimap, commissioned by ARD, came within four points of the gap ten days before the election. ARD generally refrains from publishing any further surveys in the week of the election. One reason for this is that results cannot be predicted and the measurements often fluctuate.
Stefan Merz, Senior Director of Elections at infratest dimap, remembers: "Our internal numbers in the last few days are not weighted, but the trend for the CDU and against the AfD was visible shortly before the election. But by no means as strong as it was reflected in the result."
How the type of survey can play a role
The two surveys by Civey and INSA, which saw the CDU and AfD close to each other shortly before the 2021 election Sunday, do not match other measurements or the result. There must have been major methodological problems. Civey only collected data via an online panel; the online share is not known at INSA. Purely online surveys are not considered representative because part of the electorate is overrepresented and another part is excluded from participation.
The public impact was nevertheless immense, and the attention was greater than for the other classic, representative surveys, which did not reveal any sensation. There is some evidence to suggest that this attention also had an impact on some voting decisions. And the election result would have looked different without these surprising numbers.
What influence the “AfD election victory” scenario had
The motivation to prevent the AfD from winning the election was not only among CDU supporters. Many who originally leaned towards the SPD, FDP and Greens may have voted for the CDU, even if the party was not their first choice. This was facilitated by Prime Minister Reiner Haseloff, who was well rated at the time. Long-time incumbents have already gained a lot of votes in the final stretch in some federal states. This year, however, the official bonus of the current CDU Prime Minister Sven Schulze is likely to be lower: he only took over from Haseloff in January. He doesn't come close to his approval ratings from previous years.
The influence that the special features have in Saxony-Anhalt
What could also play a role in 2026 are the special features of the federal state of Saxony-Anhalt. Stefan Merz from infratest dimap describes the country as a “special case of election research”. Party loyalty is generally lower in the East and willingness to change is particularly high. However, Saxony-Anhalt is the state with the lowest average voter turnout in state and federal elections since 1990. And the biggest changes between the last surveys and the result. They were extreme in 2021, but also clear in previous elections.
Imponderables in protest parties
In Saxony-Anhalt, right-wing protest parties often played a major role. In 1998, the DVU surprisingly achieved 12.9 percent with a last-minute election campaign financed by the publisher Gerhard Frey. Four years later, the Schill party narrowly failed with 4.5 percent. In 2016, the AfD got 24.3 percent from a standing start, its highest state result to date.
“Protest parties are the hardest for us to determine,” says Merz. "If they occur new and suddenly, the empirical data is missing." But even if they have been on the market for longer, like the AfD, it doesn't necessarily get any easier. The AfD has grown so much in the polls since the last election in the country that its supporters are no more predictable than they were in the founding phase.
If the result on Sunday in Saxony-Anhalt were very close to the pre-election polls, that would be really unusual for Saxony-Anhalt.
