[johnr2]

> It matters at least somewhat to 90% of the general public and is very important to almost 50%: https://www.instituteforgovernment.org.uk/sites/default/file...

From the linked web page: "sample size 1006 respondents". Given the UK population of around 68 million that seems a small sample.

[dTal]

The size of the UK population is irrelevant to the statistics. Absolute sample size determines the expected error, and a thousand people is plenty provided they are selected at random.

[rsynnott]

Assuming it was a properly conducted poll(I have no idea if it was or not), you get surprisingly accurate results with that sort of sample size.