Pedro Magalhães

Margens de Erro

Marktest, 15-19 Nov., N=804, Tel.

Posted November 30th, 2011 at 3:22 pm4 Comments

Voting intention:

PSD: 45.4% (+3.8)
PS: 19.7% (=)
CDU: 7.9% (-2.6)
CDS-PP: 5.0% (-0.3)
BE: 4.1% (-0.1)
Others+Blank: 17.9% (-0.8)

PM approval: 45.5% (up more than 9 pts from October). All results here.

by Pedro Magalhães

Spain: results, polls, and forecast

Posted November 21st, 2011 at 2:52 pm4 Comments

PP obtained 44.6% of the vote and 186 (53%) MP's, up 4.7 points in votes and 9 points in MP's in relation to 2008. The results basically match (slightly surpassing) those of 2000, when PP had obtained its first absolute majority, but are nonetheless the best results ever for the party. Still, expectations were for an even (slightly) better result. A simple and rough (not taking sample sizes into account) average of the very last polls (those whose fieldwork took place after November 7th) gave PP a 45.9% average. Fernández-i-Marín's more sophisticated approach generated the same average, with an interval between 45% and 46.8%.

PSOE obtained 28.7% of the vote and 110 (31.4% of) MP's, down more than 15 percentage points in votes and 17 points in MP's. It's the party's worst result ever. Although the debacle was predicted, somewhat better was expected. A simple average of the very last polls was 30.6%, Fernández-i-Marín estimated 31.1%, and our forecast based on May 2011 data estimated 34.5%. Of the 86 polls published since January 2011, only 4 estimated PSOE at 28.7% or less, and only one of them is recent (a study by the Ortega y Gasset Foundation).

IU obtained 6.9% of the vote and 3.1% of MP's, a definitive improvement over 2008 and the party's best result since 1996. Extremely close to what the polls suggested. CiU got 4.2% of the vote and 4.6% of MP's (better than expected on the basis of polls) and UPyD got 4.7% of the vote and 5 MP's (also better than what the last polls suggested). Overall, then, PP did slightly worse, PSOE rather worse, and smaller parties better than expected.

In 2008, PP and PSOE captured 83% of the vote and and more then 90% of MP's. Yesterday, they got 73% of the vote and 85% of MP's. Turnout decreased, from 74% to 72%.

by Pedro Magalhães

Eurosondagem, 10-15 Nov., N=1025, Tel.

Posted November 18th, 2011 at 3:25 pm4 Comments

PSD: 36.3% (-0.6)
PS: 29.6% (+0.4)
CDS-PP: 12.1% (-0.4)
CDU: 9.0% (+0.2)
BE: 6.1% (-0.2)

Here. Changes in evaluations of party leaders and institutions are equally small. Interestingly, leaders are mostly evaluated positively, while institutions are evaluated negatively (in the piece, there's a typo in the parliament's net evaluation result).


by Pedro Magalhães

Spain: the last polls

Posted November 14th, 2011 at 7:40 pm4 Comments

The Spanish election is less than a week away. In the last few days, several polls coming out from several different sources. Here's the updated graph, for the three major parties, since Rubalcaba became PSOE's candidate (smoother 25% bandwidth):




















The smoother seems to capture a slight decline for PP and a rise for IU, but that could be a function of the particular mix of companies that have done polls recently and the kind of "house effects" attached to their polls. A better way to ascertain this is to regress scores on dummies for companies (leaving CIS as the reference category) and dummies for time periods (months, in this case, considering the small number of polls). The results suggest that:

1. After their peak in April, PSOE lost about 3 points until today and has remained mostly stable since September.
2. Estimates clean of house effects for PP show a very small decline from October to November (.6%) but that only leaves them at the point they already were in September. Effect of the debate? An hypothesis, but impossible to say just on the basis of these data, really.
3. IU is the only party experiencing significant changes recently, up more than 1 p.p. since September.

Starting tomorrow, publication of polls in the media is not allowed in Spain.

P.S.- For estimates resulting from a Bayesian state-space model for pooling poll data first developed by Simon Jackman, see here.

by Pedro Magalhães

Polls on the Spanish debate

Posted November 8th, 2011 at 4:46 pm4 Comments

12 million people watched the debate, for a 54% share, down 5 points and 1 million from the last Zapatero/Rajoy debate (which was, however, the most watched debate in Spanish democratic history). Many polls, apparently, as reported here. But pay attention, that some of these "polls" seem to be online voluntary votes. The ones below are those I could establish actually used randomly selected samples. So who won?

1. El Pais/Metroscopia, N=501, Phone. Rajoy: 46%; Rubalcaba: 41%.
2. Antena 3 and Onda Cero/TNS, N=?, Phone. Rajoy: 43.9%; Rubalcaba: 33.1%.
3. El Mundo/Sigma Dos, N=?, Phone. Rajoy: 51.4%; Rubalcaba: 44.2%.
4. La Sexta/Invymark, N=1100, Phone. Rajoy: 48.6%; Rubalcaba: 39.9%.

Little doubts here, although one might argue that Rubalcaba did less badly than one would expect from current voting intentions. Don't forget, however, that these are self-selected samples too in a way, in this case, of those voters who watched the debate in whole or in part. And these "flash" polls take place very quickly after the debate, which makes fieldwork difficult and is certainly associated with low contact and response rates, with what that implies for sample quality. And that the effects of debates are thought to be relatively small. Etc.

BTW, Publico.es has a nice Twittometer.

by Pedro Magalhães

Spanish polls update

Posted November 4th, 2011 at 2:14 pm4 Comments

New polls:

1. By Sigma Dos, with fieldwork ending October 31st, showing almost no change in relation to previous poll finished 3 days before.

2. The new CIS study is out. The last one, from July, had PP at 43% and PSOE at 36%. Now, in congruence with what other polls have shown since then, that lead has expanded: 47% to PP and 30% to PSOE,

by Pedro Magalhães

Posted November 3rd, 2011 at 5:37 pm4 Comments

Referendum: a call for democracy and citizen participation a ploy to force ND to accept the deal.

by Pedro Magalhães

More (and probably last) in the Greek referendum

Posted November 3rd, 2011 at 5:35 pm4 Comments

Not going to happen. Had I mentioned this already?

by Pedro Magalhães

What poll results to publish?

Posted November 2nd, 2011 at 6:07 pm4 Comments

SIC Notícias made this piece about the last Eurosondagem poll. Expresso published this piece summarizing the results concerning mass opinion about the budget measures. Maybe more about this poll has been published and I just missed it. What I did not miss, however, are the actual marginals published here. And there are results I haven't seen published anywhere (rounded):


P.20 Acha que o governo cortou o suficiente na despesa?
Sim: 38%
Não: 43%
NS/NR: 20%

P21. Tendo em conta o Orçamento de Estado, vai mudar hábitos de vida no próximo ano?
Sim: 69%
Não: 27%
Ns/NR: 4%

P.27 Se estivesse em idade ou tivesse possibilidades considerava emigrar?
Sim: 61%
Não: 32%
NS/NR:7%

I really don't like P.27, as it deals with too many hypotheticals. P21 was followed by a question for those who answered "Yes", and "Eating Out" was the most selected option for cuts in personal expenditures, and this is vaguely interesting too (although not "political" enough, I assume). But it is P20, and its deemphasis in the news coverage, that I find particularly interesting. Sometimes, journalists are puzzled by apparent contradictions in polls: "If a huge majority of people disagree with the cuts, as responses to other questions in the poll show, how can then a plurality believe the government has not made enough cuts?", I imagine them thinking. But the results are not necessarily contradictory:

1. People may disagree with the concrete cuts and prefer other ways of reducing expenditure. This is especially inviting in this poll because those "other cuts" that might make them "enough" remain unspecified. 
2. "Expenditure" may generically sound bad, in the sense of wasteful. People may react to the question by thinking that it is a positive thing to cut "waste", that more should be done in cutting it, and that the government has not done enough of it. This does not mean they should support the concrete government plans at all.
3. Talking about "enough", without specifying "enough for what", further increases vagueness and allows for these apparent contradictions. Does "enough" mean that there will not be more? Does it mean that it will be "enough to satisfy our creditors"? Does it mean "enough to prevent Portugal from entering the Greek death spiral"? Etc.

In sum, it is a very good thing that ERC allows us access to the poll data unfiltered by whatever journalists think is newsworthy in those data.

P.S.- A few days later, these results were published.

by Pedro Magalhães

More on the referendum

Posted November 2nd, 2011 at 2:52 pm4 Comments

Apparently, the Daily Telegraph reports that a 3/5 majority is needed to pass a resolution permitting a referendum. If so, this would probably make it impossible. But although this may be totally clear to a constitutional lawyer, it is not clear to me at all. Article 44(2) of the Greek constitution provides for two sorts of referendums: on "crucial national issues" and on "serious social issues". In the former, government proposals must be passed by an absolute majority in Parliament. In the latter, 2/5 of parliament proposes and a 3/5 majority is needed. But although we could spend a lot of time thinking about the fascinating distinction between "crucial national issues" and "serious social issues", I don't think we need to. The main distinction seems to be between who proposes the referendum. If government, absolute majority. If parliament, qualified majority. Therefore, if we are talking about the former, as I think we are, the referendum is not as impossible as a 3/5 majority would suggest. Unlikely, but not impossible.

It should also be mentioned that when the possibility of such a referendum was first announced - last June, no less - the government also announced that it would introduce changes to several procedural aspects of referendums, which nonetheless must abide by article 44(2) of the Constitution.




by Pedro Magalhães