How *should* we be able to trigger an early election?

[andrewducker]

The UK has no mechanism for triggering a general election early that isn't dependent on the existing representatives choosing put themselves at risk.

If you were to institute a mechanism for triggering a general election that didn't involve representatives choosing to hold one, what would it be?

Mine would be to specify that if reputable polling institutions (either certified by the British Polling Council or the Electoral Council) found there was ¾ support for the question "There should be a general election" for a continuous period of one month then one should be held.

75% because it should only be only in very unusual circumstances where the government had gone very wrong, and continuous for a month to make sure that a couple of statistical anomalies didn't throw things off.

You?

Comments

[haggis]

I don't trust polling enough to genuinely reflect the opinions of the population. I trust them *even less* if the fate of the government depends on them,because the pressure to manipulate the poll would be incredible.

[andrewducker]

How would you socially pressure people to privately answer a poll about whether they wanted an election?

[haggis]

I wouldn't pressure the respondents. But I would definitely ensure that I had party loyalists involved in all the steps of the process - exact wording of the question, decisions about where and how to recruit people for the poll, and especially data analysis, eg how the data is processed to reflect the differences between the sample group and the national population.

[andrewducker]

I specified the precise question in my suggestion.

I agree that recruitment and data analysis are open to hackery. Would replacing a polling company with the ONS make you any happier?

[haggis]

No, public and private organisations would be stacked in different ways but still emd up stacked.

For a decision of this magnitude, you either need an election (to actually determine what the public wants) or give the appointed representatives the powers to do it.

[andrewducker]

Then what do you do to trigger that election when the representatives turn out to have lied and cheated their way into power?

[haggis]

The Truman vs Dewey US election gives a pretty good example of how sampling bias destroys a poll.

https://en.m.wikipedia.org/wiki/Dewey_Defeats_Truman

[andrewducker]

That was also part of the reason why I went for 75% - a sampling error of 25% would take a *lot* of effort.

[snippy]

Oh dear. People lie to pollsters all the time. And they'd be motivated to do so if they didn't like the status quo but also didn't want an election for other reasons.

[andrewducker]

I don't understand - they'd be motivated to say they wanted an election when they didn't? How would that work?

[snippy]

Some people just want to be spoilers. They might not even consider how it would work, so long as they were preventing others from getting what they want.