Fluctuation and standard errors in election polling British politics is very interesting lately; particularly so since Corbyn was elected Labour leader in 2015, and later with Brexit happening and so on.
Since then, multiple events have had dramatic effect on opinion polls, which can be seen here. 

Compare that to also quite interesting situation in Austria, where a right wing government fell because of corruption (link). 

This stark contrast strikes me as odd. In Austria, the political events of the past two months not only had absolutely no effect on polling numbers and only a very small effect in March-April (which is surprising from a political point of view), but also itra-poller agreement is unbelievably high (which is surprising from a statistical point of view). For example, consider the 26 most recent polls, conducted between 8 July and 23 September, only 2 report the FPÖ outside the 20±1% range (one of them at 22%, the other is denoted "raw data").
In the UK, however, we observe several instances of polls being conducted at the very same day, yet they often differ substantially. For example, Labour is at 21% in a 17-18 Sep poll, at 27% in a 18-19 Sep poll, and again at 22% in a 19-20 Sep poll.
I understand that it is very much possible that the public mood can swing back and forth considerably within hours, but I wonder if that sufficiently explains the stark and persisting contrast in variation between Austrian and UK polling numbers. 
Note also that Austrian polls, which are much more stable, are usually based on a smaller sample size. 
What is going on here? 
 A: For change over time -- well, as you note, British politics has had a lot of big events lately. I'd expect change over time. I don't know much about Austrian politics -- but the very fact that I don't know anything is a tiny indicator that nothing too crazy has happened lately.
For inter-poller analysis:
One possibility is that the weighting methods used by British pollsters vary a lot, while those used by Austrian pollsters vary relatively little.
Good pollsters don't use the raw data they collect, because they know that it is not a random sample of any particular population (e.g. voters or adults). They weight the data to match the population they want. This is very tricky when you are trying to match the voting population because we don't know who exactly voted in the last election and we don't know if that will change in the next election.  Different pollsters have different ideas about who will vote.  Some are less accurate than others; sometimes they are all wrong.
To use one recent and fairly famous example from USA politics, Alexandria Ocasio-Cortez was predicted to be way behind in the primary in her district. She won fairly easily. Her explanation (almost certainly correct) is that the pollsters used a weighting scheme that was based on other elections and, so, figured that few young people would vote. This turned out to be incorrect. 
