Gerhard Schröder managed to recover around 10pp during two successive election campaigns, one in 2002 and then again in 2005. Angela Merkel’s CDU/CSU polled at 50% during the 2005 campaign at one point, and ended up at 35%. Martin Schulz’s SPD peaked at 33% in the 2017 campaign, and ended up at 20.5%. This German election is only just starting, and we have seen swings of similar magnitude already.

What happened in the last month is that a Green bubble inflated and burst. People were intrigued by the Greens' choice of Annalena Baerbock as the chancellor candidate. She is a sure-footed performer on the political stage, but was not as well known as Armin Laschet or Olaf Scholz.

She has made a number of unforced errors since her nomination. Her CV is full of annoying but relatively minor inaccuracies. She did not declare payments she received from her party. This is particularly embarrassing since the Greens have been campaigning for transparency of politicians' income. Neither the irrational optimism in April, nor the current disappointment, will ultimately matter in this election. What matters is whether the Greens can persuade Germans to support what would constitute a historic change of direction, on a scale similar to the shift that took place in 1969, when Willy Brandt became chancellor, enacted social reforms, and changed foreign policy.

In April, it looked possible that the Greens might succeed. Today, less so. I believe the Greens will eventually succeed, but I am not sure that they will go all the way this time. There are many polls that have registered the ups and downs of the Green party. But the most important one is an Infratest dimap poll on public attitudes towards Green issues. This one tells us what the Greens are up against.

People were asked: Should the state outlaw behaviour that is particularly damaging to the climate? 53% say No. Are you in favour of higher petrol prices? 75% say No. Should the government encourage a shift from fuel-driven to electrical cars? 57% say No.

Baerbock and her co-leader Robert Habeck both know what they are up against. At the Green’s party conference over the weekend, Habeck gave a clever speech in which he defined Green politics as being about freedom. This was a clear attempt to reconnect to voters who don't agree with higher petrol prices, but who are open to parts of the Green agenda otherwise. This will be the battleground for the upcoming elections.

The Greens are a divided lot. Some delegates spoke in favour of a higher CO2 price than the party’s officially agreed €60 per tonne. Habeck told them they would lose the elections. The Green leadership prevailed. The so-called fundis, or purists, still matter, but they are not as powerful as they used to be. They want a prohibitions of short-haul flight; a mandatory ban on fuel-driven cars in 2025; and much stricter climate targets. All this is mingled with the policies of the old left, including the nationalisation of industries.

The other section, realos known in Green-speak, are focused more on incentives and investment, with the role of the state as a green regulator. Baerbock is talking a lot about the digital economy, and about her plans for the digitalisation of the civil service. That’s the part of the Green agenda that is highly attractive for those Germans who are in favour of Green investments, but who oppose any form of prohibition. The swing in the Green support is not only to do with Baerbock. It is also do with what they choose to talk about: petrol prices, or investment?

I would, however, not characterise the debate among the Greens as one between the left and the centre - because the realos' agenda does not fit this category. The German centre ground as exemplified by the grand coalitions and stands for industrial corporatism; for coal and cars; for reliance on Russian gas and oil; and for a mercantilist foreign policy. The Baerbock/Habeck wing of the Green Party deviates from the old centre on a different axis: they want a departure from fossil fuels; a new foreign policy in which climate change and human rights play a much bigger role. This is not left, right or centre. It is a category shift. This is what makes the promise of a Green government so enticing.

The question is whether a sufficient minority of Germans is now ready to support that categorical leap. A base support of around 20% is a substantial number in a Bundestag of seven political parties - if you count CDU and CSU separately. The polls show the Greens as the second largest party - indeed the largest if you count CDU and CSU separately. This is a big leap from the outgoing Bundestag, in which the Greens are the second smallest party, larger only than the CSU. Even after the Baerbock bubble has burst, support for the Greens is now massively higher than it was 2017. But at current levels, it may not be big enough for the category shift to happen.

I see only two coalitions where such a shift has a chance of occurring. In coalition between Greens and the CDU/CSU, where both parties are of roughly equal strength. Whether the chancellor will be Baerbock or Laschet is not the main issue. The other construction would be a coalition between Greens, SPD and FDP, the so-called traffic-light coalition.

But even then, it is far from clear that the Greens can pull this off. In an ideal situation, a CDU/CSU/Green coalition could end up with a Green investment programme, a Green fiscal policy and a Green foreign policy, with the CDU/CSU focusing on defence, internal security, and the stuff that keeps rural Germany happy. But there is no guarantee that it will end up like this.

I am confident that an agenda with a focus on green tax incentives and investment will eventually emerge. Germany is a wealthy country way past its prime. It is at a stage where people realise that the something is not quite right, but they are not yet ready to draw strong conclusions from that observation.

A version of the Green Revolution may happen, but it won’t be what it appeared to be six weeks ago, when the Greens briefly looked like they would sweep the floor.