I’ve been trying, and failing, to find any formal documentation — preferably official policies or similar from gov.uk — that state or set forth the principle that in something like a referendum or a public consultation, the views of respondents should be taken as statistically representative of those who did not respond/vote/express a view.

To explain further:

For example: if 65% of voters vote 'yes' on an issue, then that percentage is extrapolated proportionally and it is assumed, for the purposes of calculating a result, that 65% of all of those eligible to vote on the issue would vote 'yes' (or, to equal effect, that the views of those who did not vote can be discounted when evaluating the result).

That’s a wordy way of describing what we intuitively know about elections. To put it another way, if 65% of voters vote 'Yes' but turnout was only 50%, then it is generally not considered legitimate to assume that all non-voters would have voted 'No' and to thereby reach a conclusion that the final tally was Yes: 32.5%, No: 67.5%.

My question:

Does anyone have a link to a document that puts this seemingly-obvious principle into writing, preferably as UK government official policy or guidance documents?

  • 3
    I don't think they do consider a referendum to be a statistical sample or consider the views of the voters to be statistically representative of the non-voters. Opinion pollsters might have more data, but what matters in a referendum is who wins, not who would have won if the non-voters had voted. So what you descrive as "seemingly obvious" I think is "obviously false".
    – James K
    Oct 17 at 16:09
  • Perhaps I explained myself badly. The point I’m trying to find evidence of is that assumptions cannot be made about the views of non-voters, and that only the views of those who expressed an opinion should be considered legitimate in calculating the outcome. Whether that means extrapolating the vote spread of received responses to apply to non voters, or whether it means ignoring everyone who didn’t vote, results in the same result, right? (In practice, at least). Oct 17 at 17:32
  • Incidentally, maybe me using the ‘referendum’ word was a mistake as it might imply that this is a Brexit-related question. It’s not. Oct 17 at 17:34
  • "only the views of those who expressed an opinion should be considered legitimate in calculating the outcome": yes; how else would you do it? Unless you're suggesting a system based on statistical modelling - in which case, see Isaac Asimov's satirical short story Franchise. Oct 19 at 10:21
  • @SteveMelnikoff yes, quite. For context, this question was prompted by an official response made by a public official who was confronted with the results of a public consultation in which 88% of respondents were opposed to the plans. When asked what level of objection would be required in order for the local authority to change its plans, he replied that “88% of respondents is not the same as 88% of residents.” Which is true — but also, I believe, democratically irrelevant. I was looking for some tangible published policy enshrining this principle. Oct 19 at 10:33

In the UK, elections, and referendums are decided by those who vote. Those who don't vote are ignored. There is no claim made that the voters form a statistically representative sample of the electorate. Merely that the winner of the election is decided only by the votes that are cast.

From the AV election legislation:

The Minister must make an order bringing into force section 9, Schedule 10 and Part 1 of Schedule 12 (“the alternative vote provisions”) if—

(a)more votes are cast in the referendum in favour of the answer “Yes” than in favour of the answer “No”...

And that really is that.

The result of the referendum is decided by the votes cast.

Other systems have different rules, and in some there is a quorum. The election is considered invalid if too few people vote. Or a majority of the electorate is needed, and not a majority of the voters. These rules have to be written into the legislation that establishes the vote.

These rules can have paradoxical effect (eg if more people were to vote against the proposal, then quorum is reached and the proposal can pass, but if those people choose not to vote, then quorum fails and the vote doesn't pass, making it tactically a bad idea to vote against a proposal)

  • 1
    I will add that it’s not uncommon for a politician to verbally blur these lines and assume that there’s a direct relationship between percentage of actual voters and percentage of the whole electorate, but that’s in the same category as claiming that 51% of the vote means “most” people support them.
    – Bobson
    Oct 17 at 19:37
  • @Bobson isnt that kind of why we also track total number of votes for a party in addition to seats or positions won in an election, as it allows parties to not only claim a victory but also the "popular vote" and, depending on which version you subscribe to, whether the winner has a "mandate" or not. Its common in UK politics for the government to win with a "popular vote" of much less than 50%, so in the UK the "mandate" is usually based on whether they have a simple majority or not - but elsewhere it can be based on popular vote percentages.
    – user16741
    Oct 17 at 20:32

Not the answer you're looking for? Browse other questions tagged .