• Facebook
  • Facebook
  • Facebook
  • Facebook

Search This Blog

Visit our new website.

Friday, June 20, 2008

Did the Commission lie about its own poll result?

Looking at the official results of the Eurobarometer 'snap poll' of Irish voters following the No vote (released today), there's a striking contrast between the figures in it and the figures that were leaked out to journalists before the summit.

The Times story on 18 June ran with the headline “Irish voters failed to understand the Lisbon Treaty”, citing a figure from the EU Commission stating that, amongst No voters, “40 per cent blamed the fact that they did not understand the treaty.”

Below is the extract from the briefing note pushed around by the Commission – the basis for the story in the Times, and other papers:


However, when we compare this with the official Eurobarometer poll, published today (20 June) – after the Summit had concluded – we see a very different picture:


Just 22% of No voters said their vote was informed by a lack of knowledge/ familiarity with the Treaty.

The EU Commission’s unofficial briefing seems to have arbitrarily doubled the proportion of voters falling into this category – presumably if this had not been done, pre-summit headlines focussing on Irish voters “failing to understand the treaty” would not have occurred.

Are we missing something here?

If not, this certainly looks like a very clumsy piece of politically-driven propaganda on the part of the Commission designed to make the Irish voters look ignorant.

Unless we are missing something, they have lied to journalists, and this should be a big story.


Anonymous said...

Noticed 2 things that might explain the discrepancy, in case the "leak" is authentic:
1. "Why did you vote no" (ONLY ONE OPTION) -- cannot be true, must have been multiple response, as it is over 100% if we add up the responses.
2. the footnote to the just released and quoted table is referring to "% of ALL ANSWERS"

It is obvious, that what might happened is that if you are looking at something in the % of ALL answers, and one respondent can say more then one reason, why he/ she voted no, the percentages will be much smaller numbers, compared to a situation, where you are looking at the numbers as if it were only one answer per respondent, and you calculate the percentages on a different base. (but again, it cannot be that there was only one answer option, otherwise the percentages could not be more than 100% when you add it up.
In other words, the caution might have been correct, the numbers in the "leaked" document were not carefully doublechecked.

Anonymous said...

"Just 22% of No voters said their vote was informed by a lack of knowledge/ familiarity with the Treaty."
If the basis of %s (as noted in the footnote in the table) is NOT the respondents, but the answers; AND, if on the average somebody gave 2 answer to the "WHY" question -- then both can be true. IF the first leaked results are referring to respondents, and the just released figures are referring to the "answers" as the base of percentages then something like that should be the result.

Polymath said...

The first set of percentages adds to 105%.

Open Europe blog team said...

Its hard to see that any different method of presenting the same data can be responsble for the difference.

No simple transformation can account for the radical difference in the answer for "don't trust politicians" which is below or level pegging with neutrality in the final answers, but three times as big in the origninal briefing. (Another point that europhiles want to push, conveniently).

Here are the ratios comparing the top six answers in the original note and the final version.

-22 -40 -0.55
-12 -20 -0.60
-6 -17 -0.35
-6 -10 -0.60
-6 -10 -0.60
-6 -8 -0.75

Anonymous: If it is just a matter of presenting the data as a share of no voters who responded versus a share of all no voters the ratio should be the same because all you are doing is taking out DK which would push up all the results proportionately...

Anonymous 2 - The question was obviously asked as a single choice (and only for no voters) and in the final form the result sums to 99% (presumably only not 100% because of rounding error). It says single choice in the original briefing too.

I don't think it can just be a matter of "not checking" the data.

The key question here is: even if we assume that this was a convenient accident, which just "happened" to give the Commission the headline it wanted, should the Commission be pushing around innacurate and unverified data?

Any more thoughts much appreciated...

Anonymous said...

The only rational conclusion is that - shock horror - Eurocrats lie.

The whole of the current UK debate about whether or not we should have a referendum is actually based on the fact that Eurocrats lie as a matter of course - if they didn't lie they would admit that the Lisbon Treaty and the Constitution are, in all material respects, the same.

I'm delighted that they have (at last) been caught out over such a blatant issue as the Irish poll - perhaps now we might see the MSM and especially the BBC look more closely as the propaganda pumped out at vast expense from the EU and its many, many client 'independant' bodies.

P O'Neill said...

Some Irish bloggers have spent the last week trying to figure out the provenance and methods of this poll. One issue that is the pollsters appear to have done unstated calculations with the non-voting population to arrive at their numbers e.g. adding those who said they voted No through not understanding to those who didn't vote because they said they didn't understand. But that's a different decision.

Anonymous said...

Actually, not understanding the thing is an excellent and intelligent reason to vote against it. The US Constitution was written so as to be understandable to the man in the street of 1787. It is short enough, including the amendments, to be memorized in a couple of days. I had to do exactly that when I took a class on Constitutional Law in university. It deals with the essentials of government and leaves the policy details up to the future political leadership to flesh out in accordance with the needs of some future time.

The EU Constitution, now masquerading as the Lisbon Traty, by constrast, is an abomination that not even its drafters can understand, as Valery Giscard-d'Estaing himself admitted. It tries to freeze in ice the current status quo, and empowers a government completely unaccountable to the people. It is a recipe for disaster, and the Irish were quite intelligent to recognize this and vote it down.

The EU Constitution mongers, if they want to succeed and not establish something that will be ripped apart by 2020, should go back a write another document to replace the proposed "Treaty". Make it ten pages at most, written so most people can understand it. Then submit it to the various electorates. Such a constitution, including only the minimum necessary provisions, will have a chance of producing a state which might survive the next crisis, and will have a much better chance of being accepted by the people. All government rests on the consent, tacit or explicit, of the governed, and to evade such consent in setting up the EU is asking for trouble.

Eveleigh Moore-Dutton said...

It would be very interesting to know what proportion of the Irish voters who voted for the Lisbon Treaty understood what they were voting for. I suspect that very few did and in that case it makes far more sense to vote to retain the status quo. Therefore I conclude that the 'No' voters showed greater sense than those who voted for a leap in the dark.

Anonymous said...

Michael Lonie and Eveleigh Moore-Dutton have clearly got this right. If anyone were to be offered a deal which would affect their legal or democratic rights and obligations they would surely refuse if they couldn't understand it.

We know from their own mouths that several, perhaps all ministers argued in favour of a Treaty they had not themselves read. Clearly they could not have understood something they had not read.

Anonymous said...

"Are we missing something here?"

Yes: "only one option".