Just a few days before May 7 the UK 2015 Election Day, I was lucky enough to watch the celebrated US poll aggregator Nate Silver in action in Britain on BBC Television. As you very well know, poll aggregators are websites that take the published polls results and combine them statistically to report their own election predictions.
According to Wikipedia, notable poll aggregators are Real Clear Politics; Electoral-vote.com; Princeton Election Consortium; FiveThirtyEight (founded by Nate Silver); Pollster.com or Huffpost Pollster; the political blog Talking Points Memo; Votamatic; Frontloading HQ; Election Projection; and Politics by the Numbers.
Nate Silver's prediction on BBC Television was that though the Conservative Party will beat the Labor Party the share of parliamentary seats will be so close that it will result in a hung-parliament. As someone writes: "... after near 30 minutes of mind numbingly boring footage of basically a caravan driving around the UK, finally has Nate Silver state his forecast conclusion that put the Conservatives on 283 seats, Labour 270, SNP 48, Lib Dems 24 and UKIP on just 1 seat. Then he states the obvious that no major party even with the Lib Dem support could form a majority:"
Living in UK and having seen something like this polling scene in 2010 elections, ordinary folks may very well have a gut-feeling that the Tories will win the most seats but couldn't get a majority by themselves. So it's nothing new. In fact, this latter part of their gut-feeling may as well have been influenced by the predictions of the pollsters.
Meanwhile, one UK polling company, Survation, said they played too safe and threw away the only would be correct prediction. It claimed that its election eve poll had been close to the final result with the Conservatives on 37% and Labour on 31% (the final results were 36.9% to 30.4%). According to its CEO, "The results seemed so 'out of line' with all the polling conducted by ourselves and our peers -- what poll commentators would term an 'outlier' -- that I 'chickened out' of publishing the figures -- something I'm sure I'll always regret".
As it happened, some of the election predictions, according to one pollster were:
Source:
Market Oracle
|
Market Oracle
|
May2015.com
|
Electoralcalculus.co.uk
|
ElectionForecast.co.uk
|
The Guardian
|
28th Feb
|
26th Apr
|
26th Apr
|
27th Apr
|
27th Apr
| |
Conservative
|
296
|
272
|
279
|
286
|
274
|
Labour
|
262
|
271
|
282
|
267
|
270
|
SNP
|
35
|
55
|
47
|
48
|
54
|
Lib Dem
|
30
|
26
|
18
|
24
|
27
|
UKIP
|
5
|
3
|
1
|
1
|
3
|
Others
|
22
|
22
|
22
|
22
|
22
|
Predictions from other sources, for example, as given in Wikipedia, were not much different from these. In terms of the vote share percentages, mostly it was like one or two percentage points over the Labor in favor of the Conservative party.
The exit poll result given at 10 PM at the end of voting on the Election Day gives a more realistic number of seats for the Conservatives, but still not enough to give a majority.
Why couldn't all the best pollsters predict the majority for the Conservatives? And not even by the exit poll?
One of the biggest problems for pollsters in the recent times is that the response rate for pre-election polls was getting really low. So that getting a traditional poll done on a good representative sample of voters is getting well-nigh impossible or getting prohibitively expensive. Some hails internet polling as used by YouGov and others as possible remedy for such flaws of the traditional polls.
But, YouGov’s poll on Thursday night, conducted after votes had been cast, was baffling. It again showed that the parties were neck-and-neck (How 'shy Tories' confounded the polls and gave David Cameron victory, Jessica Eglot, The Guardian, 8 May 2015). Besides, it didn't do well recently in the 2014 Scottish independence referendum. One poll by YouGov created a stir when it published a 2% lead for "Yes" and the actual outcome was an 11% lead for "No" (Election 2015: Hold on, this isn't what you said would happen, BBC newsbeat).
Generally, it would be more difficult to get the polls right in the political system prevailing in UK than in the US. On the FiveThirtyEight website Nate Silver said "When there are only two major candidates, the choice isn’t very complicated. ... UK has become less and less of a two-party system. While the Conservatives and Labour collectively accounted for about 90 percent of the vote through the election of 1970, they’ll be down to somewhere in the neighborhood of 65 percent to 70 percent this year."
He was worried that "The World May Have A Polling Problem":
Consider what are probably the four highest-profile elections of the past year, at least from the standpoint of the U.S. and U.K. media:
- The final polls showed a close result in the Scottish independence referendum, with the “no” side projected to win by just 2 to 3 percentage points. In fact, “no” won by almost 11 percentage points.
- Although polls correctly implied that Republicans were favored to win the Senate in the 2014 U.S. midterms, they nevertheless significantly underestimated the GOP’s performance. Republicans’ margins over Democrats were about 4 points better than the polls in the average Senate race.
- Pre-election polls badly underestimated Likud’s performance in the Israeli legislative elections earlier this year, projecting the party to about 22 seats in the Knesset when it in fact won 30. (Exit polls on election night weren’t very good either.)
Perhaps it’s just been a run of bad luck. But there are lots of reasons to worry about the state of the polling industry. Voters are becoming harder to contact, especially on landline telephones. Online polls have become commonplace, but some eschew probability sampling, historically the bedrock of polling methodology. And in the U.S., some pollsters have been caught withholding results when they differ from other surveys, “herding” toward a false consensus about a race instead of behaving independently. There may be more difficult times ahead for the polling industry.
The British Polling Council (an association of UK polling organizations) declared on 7 May that it will set up an independent inquiry to examine "the possible causes of this apparent bias" in the UK 2015 election polls and make recommendations for future polling.
As reported by CNN, Council president John Curtice -- Professor of Politics at Strathclyde University – said that while polls should be judged on their percentages rather than seats and while it could well be true that many had fallen within their margins of error, an inquiry was still needed. The polls had been accurate on the SNP, Liberal Democrats and Greens -- but they all had an error in the same direction, he said. And that "The reason an inquiry has been set up is that actually the industry collectively clearly underestimated the Conservative lead over Labour".
What would happen if the UK2015 election system were proportional representation than first-past-the-post as it was done? This is what The Electoral Reform Society, a campaign group, arrived at through their calculation using the D'Hondt method of converting votes to seats.
Well, this is interesting, but the real challenge for pollsters in the present electoral system in UK is another kind of vote conversion. It is the need to convert estimates of votes received by parties to estimates of seats in the parliament. So even if you could get the number of votes within the margin of error you planned for, you may still get the error in projected parliamentary seats that hurts.
Meanwhile, International New York Times labeled pollsters as British Election's Other Losers (Dan Bilefsky, 8 May, 2015). It cited Alberto Nardelli, the Guardian's data editor, who said there was no simple explanation to what went wrong with the polling.
“It could be simply that people lied to the pollsters, that they were shy or that they genuinely had a change of heart on polling day,” he said. “Or there could be more complicated underlying challenges within the polling industry, due, for example, to the fact that a diminishing number of people use landlines or that Internet polls are ultimately based on a self-selected sample.”
It also cited Korteweg who thinks that polling disasters are becoming a trend in UK:
Rem Korteweg, a senior research fellow at the Center for European Reform in London, said that British pollsters were going through a particularly bad period. He cited the referendum last year on Scottish independence, when pollsters’ predictions of a neck-and-neck performance for the no and yes camps were upended by results in which 55 percent of Scots voted against becoming independent compared with 45 percent in favor.
“This isn’t the first time in 18 months when the polls got it wrong,” he said. “This is starting to become a trend in this country.”
Mr. Korteweg attributed the pollster’s failings in this latest election to the fact that voters often give socially desired responses during polling, only to behave differently when they vote. “People say who they are voting for with their heart and then vote with their wallets,” he said.
And who even suggested the possibility of having exploited dark fears in politics:
He said that the last-minute surge by the conservatives could also be explained by the fact that the Conservatives had adroitly exploited fears among voters that the Labour Party would be able to govern only in coalition with the Scottish National Party, which wants an independent Scotland.
What really happened in the UK 2015 General Elections was that the Conservatives got a majority since 1992, and the Labor Party suffered their worst defeat since 1987 (United Kingdom general election, 2015, Wikipedia).
Party
|
Leader
|
Votes
|
Votes %
|
Seats
|
Seats %
|
Conservative Party
|
David Cameron
|
11,334,920
|
36.9%
|
331
|
50.9%
|
Labour Party
|
Ed Miliband
|
9,344,328
|
30.4%
|
232
|
35.7%
|
UK Independence Party
|
Nigel Farage
|
3,881,129
|
12.6%
|
1
|
0.2%
|
Liberal Democrats
|
Nick Clegg
|
2,415,888
|
7.9%
|
8
|
1.2%
|
Scottish National Party
|
Nicola Sturgeon
|
1,454,436
|
4.7%
|
56
|
8.6%
|
Green Party
|
Natalie Bennett
|
1,154,562
|
3.8%
|
1
|
0.2%
|
Democratic Unionist Party
|
Peter Robinson
|
184,260
|
0.6%
|
8
|
1.2%
|
Plaid Cymru
|
Leanne Wood
|
181,694
|
0.6%
|
3
|
0.5%
|
Sinn Féin
|
Gerry Adams
|
176,232
|
0.6%
|
4
|
0.6%
|
Ulster Unionist Party
|
Mike Nesbitt
|
114,935
|
0.4%
|
2
|
0.3%
|
Social Democratic & Labour Party
|
Alasdair McDonnell
|
99,809
|
0.3%
|
3
|
0.5%
|
Others
|
N/A
|
349,487
|
1.1%
|
1
|
0.2%
|
No comments:
Post a Comment