Straight after the Greek referendum was announced actual polling evidence seemed quite light, but there has now been the expected rush in polling. Polls from a handful of different companies are all painting a consistent picture of YES and NO being neck and neck. In fieldwork conducted on Monday and Tuesday there was still a small lead for NO, but across all the polls conducted in the last couple of days the position has been almost a dead heat.

The most recent polls are below:

Metron/Parapolitik (Thurs-Fri) – YES 46%, NO 47% (No ahead by 1%)
GPO/Mega TV (Wed-Fri) – YES 44.1%, NO 43.7% (Yes ahead by 0.4%)
Alco/Proto Thema (Wed-Fri) – YES 41.7%, NO 41.1% (Yes ahead by 0.6%)
Ipsos (Tues-Fri) – YES 44%, NO 43% (Yes ahead by 1%)
Uni of Macedonia/Bloomberg (Thurs) – YES 42.5%, NO 43% (No ahead by 0.5%)

In the week we also had the monthly ComRes/Daily Mail poll. Latest voting intention figures are CON 41%, LAB 29%, LDEM 8%, UKIP 10%, GRN 5%. Tabs are here.

UPDATE: And the actual Greek result (with just over a third of the votes counted) looks like a solid victory for NO, absolutely miles away from what the Greek polls were showing. Ouch! I don’t know enough about Greek politics or Greek polling to hazard any guesses as to what they got wrong, but I imagine a country in economic turmoil is not the easiest to poll correctly in terms of contacting people, or to getting any firm demographic figures to weight or sample by – and that’s before you get to whether people feel able to answer the question honestly. As it happens most of the Greek polls were pretty good at their general election earlier this year, but clearly not this time.


Grexit polls

On Sunday there is a referendum in Greece on whether to accept the deal that was put to the Greek government before negotiations broke down (or at least, there was as I write, who knows what the position will be by the time you read this). What can the polling tell is about the likely result? There have not been any polls since the referendum announcement yet – though I don’t think there is anything preventing any (Greece previously had a ban on polls in the last fortnight of election campaigns, but this was repealed before the election earlier this year. I don’t know about referendums or any subsequent legal changes.)

There were, however, two Greek polls conducted in the three days before the referendum announcement that have been widely reported. A Kapa Research poll conducted between Wednesday and Friday actually asked how people would vote in a then hypothetical referendum, with 47% saying they would vote yes, 33% that they would vote no. Of course the poll was conducted prior to the referendum announcement so may not reflect current Greek opinion at all – people taking it as a sign Greece is about to vote yes should probably hold on a sec. Respondents may have been imagining a referendum on a deal that had the support of the Greek government, rather than a referendum where the government are opposed and backing a No vote.

The rest of the Kapa poll found 72% of Greeks wanted the country to remain within the EU and 68% wanted them to keep the Euro. There was a pretty even split over the government’s strategy – 49% had a positive opinion, 50% a negative opinion.

A second poll by Alco found negative opinions about the proposals on the table, but continuing goodwill towards Syriza. People didn’t think the proposals met their pre-election promises, but by 53% to 34% thought this was because Syriza hadn’t realised how difficult it would be rather than an attempt to mislead the people. By 61% to 33% respondents rejected the idea that the last Greek government would have done any better. Syriza continue to hold a robust lead in voting intention. Again, this is sometimes being reported as showing Greeks will vote Yes, but I’d be wary. It found people would, in principle, prefer a deal to default… but that’s not the same as saying they will vote YES in a referendum on a specific offer that the Greek government doesn’t support.

Turning to the attitude in other countries in Europe, YouGov polled the countries it has panels in a week ago and in most countries the public expected Greece to leave the Euro, and would prefer it if they did. In Britain people would prefer Greece to leave by 35% to 26%, Denmark by 44% to 24%, Sweden 35% to 26%, Finland 47% to 26%. France was the only country polled where people would prefer Greece to stay within the Euro, though only by 36% to 33%. In Germany 53% of the public thought Greece should leave the Eurozone, only 29% would prefer Greece to remain. Note, of course, that the countries YouGov operate in are largely Northern Europe… the public in Southern and Eastern European countries may have different views.

UPDATE: We finally have a poll on the referendum conducted after it was announced. Prorata for Efsyn found 51% of Greeks intending to vote no in the referendum. The fieldwork appears to have straddled the announcement to close the banks – before the announcement NO led by 57% to 30%, after the announcement NO led by only 46% to 37%. On the face of it that looks like No leading, but in a very fluid situation, but I don’t know what the sample size was before and after the bank closure (and indeed, whether the early and late respondents to the poll were comparable) so cannot tell if that apparently shrinking lead is meaningful.


-->

Sooner or later, a pollster gets something wrong. It happens to everyone if they are in the game for long enough. There are two responses to that, there is to deny there is any problem and blame it all on a late swing, or you can go away, work out what went wrong and put it right. The good pollsters do the second one – so when all the companies got it wrong in 1992 there was an industry inquiry, and ICM in particular came up with new innovations that addressed the problem and led to many of the methods companies use today. In 2008 when MORI got the London election wrong they went away, looked at what had happened, and made changes to put it right. A pollster that gets things wrong, admits it, and then puts it right isn’t a bad thing.

Anyway, in the US election last year the most venerable of all polling companies, Gallup, managed to get things wrong, showing a small lead for Mitt Romney rather than the eventual victory for Barack Obama. They put their hands up, invited in some academics to help and went away and looked at their methods – the result is here. Gallup tested about twenty different hypotheses of things that could have gone wrong, and found that in the majority things were working okay and there was no issue to address. They ended up with four issues where they think things went wrong and caused the overestimate of Romney’s support.

Most or all of the actual problems Gallup identified aren’t directly relevant to British political polls – different system, different challenges, different methods, different solutions – but it’s still an interesting look at what can go wrong with a poll and how a company should dig through its methods if something has gone wrong.

Likelihood to Vote – In Britain pollsters have a relatively simple way of approaching likelihood to vote: they ask people how likely they are to vote, and then weight and/or filter people’s responses based upon that, either giving people’s answers more weight based on how likely they say they are to vote or excluding people below a certain threshold. The only exception to this is ICM, who also include whether people voted in the 2010 election in their likelihood to vote model. American pollsters tend to use much more complicated methods, they ask people how likely they are to vote, but also whether they voted last time, how interested they are in politics, whether they know where the polling station is and so on – there are seven questions in all, which they use to work out a likelihood to vote score and then include only those most likely to vote. Other American pollsters do much the same, but Gallup’s method put more weight on whether people voted in the past, and their adjustment ended up being more pro-Romney than some other companies. Gallup are going to go away and do more work on turnout, including whether the sort of people who take part in polls are more likely to vote anyway (something that I would certainly expect to be true).

Sampling – Most telephone pollsters in the USA get their phone numbers in a similar way to British pollsters, by using random digit dialling. This ensures that people who are ex-directory are not excluded from samples, but at the cost of getting lots of dead telephone numbers, faxes, modems, business numbers and so on. In 2011 Gallup started doing something different. Like most companies they do a fair amount of their interviews on mobile phones, and noted that the majority of ex-directory people did have mobile phones, so theorised that it was safe to randomly generate their landline sample from telephone directories, while bumping up the mobile phone sample to catch those ex-directory people on their mobiles (mobile people who reported being ex-directory were weighted up to account for the tiny percentage of ex-directory people without mobiles). In theory it should of worked. In practice, it probably didn’t – before weighting the RDD sample was more democratic, younger and more pro-Obama than the listed one, so Gallup are going back to using the more expensive RDD method.

Time zone skews – this is an interesting one. As you might expect, Gallup sample within and weight by the regions in the USA. But within some of those regions there are different time zones, and because Gallup started polling at 5pm local time, it meant that in regions that covered more than one time zone they ended up doing more interviews in the eastern part of the region. Correcting this problem would have increased Obama’s support in Gallup’s final poll by 1%. Of course, in Britain we don’t have different time zones to worry about, but it illustrates a problem that can effect any methodology design – skews within the categories you weight by. A pollster can have, for example, the correct proportion of people in the DE social class or people over the age of 55, but what if within those categories you control for people are skewed towards more affluent DEs, or people only just over the age of 55?

Race – the final problem was a rather specific one on how Gallup asked about race – instead of giving people a list of race categories and asking which applied to the respondent, they asked them one at a time and got people to say yes or no, which produced some rather odd effects like overstating the proportion of Native Americans and mixed race people.

The full Gallup review is here and if you’re interested I’d also recommend reading the verdict of Mark Blumenthal (who spotted some of the problems before Gallup did) here and who has obviously followed it infinitely more closely than me.


Today part of the country (or at least, a British dependency) actually has an election – the Falklands have a referendum on whether they wish to remain British. The result is a formality of course, people there will overwhelmingly vote to remain British, but what do people in Britain itself and Argentina think?

YouGov and Ibarometro have carried out parallel polling on the issue this week in Britain and Argentina – full results are here. There is a broad perception that the issue matters rather more to Argentina than Britain, only 1% of British people pick the Falklands as one of the most important international issues facing the country, 24% of Argentinians do. Asked directly about Falklands 54% of British respondents think it is an important issue to Britain, 67% of Argentinians think it is an important issue to Argentina.

Unsurprisingly 62% of Argentinians think that the fairest solution to the issue would be for the islands to become Argentinian, 20% would support joint-sovereignty. Only 4% think they should remain British. For British respondents 40% want the islands to remain British, 28% think they should become independent (presumably respondents who don’t realise just how few people live there!), 13% would support shared sovereignty, only 4% think the islands should become Argentinian. There is very little crossover there.

Asked what they think actually will happen there is less contrast. 61% of British respondents think the islands will remain British, 37% of Argentinian respondents think they will remain British. Despite the fuss made over the Islands by the Argentine government (actions that are supported by most Argentinians – 53% think their government is doing a good job on the issue), only 25% of Argentinians think the islands actually will end up becoming Argentinian.

Turning to the referendum this week, British respondents overwhelmingly think that the Falkland Islanders should have a say on their future (88% think they should, 4% do not). Argentinian respondents do not – only 15% think the Falkland Islanders should have a say, 59% do not. Asked who should have the final say on the Islands’ future, 74% of British respondents think it should be down to the islanders themselves. Argentinian respondents were more divided, but the most popular option was for a international organisation to make the decision (36%).


Today is the second Greek general election of the year, following the May election that produced a Parliament unable to agree on a coalition government. Needless to say, the election has great importance beyond Greece, in terms of whether a New Democracy government that will continue with the current bailout agreement emerges or a Syriza government that will reject the bailout agreement.

Greece has a law banning opinion polls from being conducted in the final couple of weeks before an election, so the final polls were all conducted at the tail end of May, two weeks ago. Since then there have been rumours of secret polls showing ND ahead, which illustrates one of the arguments against such bans – the void created by banning proper polls is just filled by rumour and leaks. That aside, the final Greek polls are listed below.

Date ND Syriza Pasok Anel KKE XA DIMAR
Metron 31/05/12 27 26 13 7 5 5 8
Marc/Alpha 31/05/12 29 27 14 7 6 5 6
Kapa* 31/05/12 30 27 12 6 7 6 5
Rass* 30/05/12 30 27 14 7 6 4 6
MRB 30/05/12 28 26 15 7 5 5 7
DataRC 30/05/12 28 26 14 7 6 5 6
Global Link* 30/05/12 27 24 13 8 7 7 7
Alco* 30/05/12 28 25 14 7 6 5 6
Public Issue 30/05/12 26 32 14 6 6 5 8
Pulse RC 29/05/12 27 27 15 8 6 6 6
VPRC 29/05/12 27 30 13 8 6 5 8
LAST GENERAL ELECTION 2012 19 17 13 11 9 7 6

*Greek pollsters differ on whether or not they re-percentage their figures to exclude don’t knows and won’t says. The polls marked with asterisks were not originally re-percentaged, but I have done it manually to make them comparable.

As you can see, two weeks ago the polls were tending to show a small lead for New Democracy, a reverse from the period straight after the May election when Syriza surged ahead for a while. The two polls that show a Syriza lead, VPRC and PublicIssue, apparently have methodological differences involving using time series analysis rather than political weighting – I won’t pretend to understand them given that the technical papers are, literally, all Greek to me.

A final consideration is the Greek electoral system awards an extra 50 seats to the largest party, so while ND and Syriza are very close in the polls, one will emerge with at least 50 seats more than the other.