Dodgy Electoral Analysis

[andrewducker]

Chatting about the electoral results, I was wondering how much skew was produced by the country being polled in regions, rather than as a whole.

Party	Percentage	Potential MEPs	Actual MEPs	Unearned MEPs
Conservative	27.7	19.1	25	5.9
UKIP	16.5	11.4	13	1.6
Labour	15.7	10.8	13	2.2
Liberal Democrat	13.7	9.4	11	1.6
Green	8.6	5.9	2	-3.9
BNP	6.2	4.3	2	-2.3
SNP	2.1	1.4	2	0.6
Plaid Cymru	0.8	0.6	1	0.4
English Democrat	1.8	1.2	0	-1.2
Christian	1.6	1.1	0	-1.1
Socialist Labour	1.1	0.8	0	-0.8

What's interesting is that Plaid Cymru (the Welsh nationalist party) only got their seat because Wales was polled as a seperate region, and it's clear that this system does allow small regional parties more clout, but minor national parties (the Greens, BNP, English Democrats, Christian People's Alliance and Socialist Labour) do worse out of it.

Not sure how I feel about that...

Edit to Add:
8.5% of people voted for a party that got no MEPs at all. That's a lot of disgruntles people, I'd imagine.

Comments

[meihua.livejournal.com]

Small clarification, I think you mean Liberal Democrats rather than this lot?

[likeneontubing.livejournal.com]

You raise a very interesting question there I think, to what extent should region and regional representation matter in these things?

Whilst I think giving Wales/Scotland a chance to get some seats in is obviously a very good idea, I think the smaller national parties will always suffer from *something* on account of being small. If we did the vote nationally, would they be much better off? Wouldn't they still be ridiculously small?

[e-halmac.livejournal.com]

Hmmm yeah that is a little odd. And worrying but I'm only saying that because it seems to have enabled the BNP to get seats and I think they are bad news.

Out of interest, how does this system compare to the one in Scotland? I think we vote for local candidates and regional party list (or is it national?). Meaning it should (?!) be more representative of the vote overall. But have never checked in realation to the figures as you have done.

[andrewducker]

All figures from here, by the way:
http://news.bbc.co.uk/1/shared/bsp/hi/elections/euro/09/html/ukregion_999999.stm

[skington.livejournal.com]

Far fewer disgruntled people than if we'd been using first past the post, though.

Here's what the results would look like nationwide according to a d'Hondt simulator. It gives slightly different figures from yours, possibly because d'Hondt isn't simply a matter of "divide the votes by 69 and see what happens".

Also, nationwide you'd probably have some sort of minimum score required (I think many countries use a 5% threshold) to avoid a plethora of tiny parties.

[lpetrazickis.livejournal.com]

Disclaimer: I'm Canadian. We have First Past the Post. Bloc Quebecois invariably gets 1/2 the votes but 2x the seats of the New Democratic Party.

Regarding these results: Nation-wide votes that distribute seats proportionately tend to also have a minimum percent cutoff (no seats for 4% or less). Really small parties have a hard slog in either model.

[red-phil.livejournal.com]

Perhaps the answer is to divide the country onto large regions. Say 6.
Each region gets a number of MEPs proportional to their population.

Then use proportional representation within each region to get the overall national result.

I'm not big on politics, but that seems to me to give a balance between regional representation and overall proportional representation.

This is largely beside your specific point, but...

[0olong.livejournal.com]

It really disturbs me that Labour gets less than twice the votes of the Greens, and more than six times the seats. While the BNP getting less representation is a pleasing side-product of this bizarre way of doing it, the overall result strikes me as spectacularly undemocratic.

It *is* difficult to know exactly how to trade off local representation against overall proportionality though...