Well every commentator and journalist tells us that this is the most difficult election to call in history, and in the same breath tell us what will happen due to opinion polls.
Since 1950 the lab + con voteshare has been in steady historic decline from 98.5% down to 65% in 2010.
Now almost every single poll, whether commissioned by a tory paper or a labour one, seems to buck the trend, generally claiming voteshares of 67-70%, which frankly just doesn't sit well with the mood of the country or the reaction the leaders get in public. Add in the massive loss of Labour votes in Scotland and it becomes difficult to accept the findings of any of the polls.
Election forecast strangely predict a 71% chance of this vote share being lower than any previous election, whilst still putting the lab + con average on much greater than 65%.
Manipulating polls is clearly in the interests of the 'narrative'. Individual seats are rarely ( about a dozen cases) won by a constituency vote of less than 33%, hence if a party is polling under that nationally it makes the notion that they are going to win hundreds of seats statistically unlikely.
Also the polls are selective, generally using newspaper readership as the criterion. Hence x many Guardian readers plus Mirror etc are used to represent likely labour voters.
Trouble is paper rounds are thin on the ground as newspaper readership has been in freefall for at least 15 years and tallies rather well with the overall lab + con voteshare ( most households took more than one in the 50s). Nowadays less than a third of households regularly buy a paper and the respect accorded to journalists is not that different from the public perception of a paedophile or politician.
This is flawed to begin with, though the polls will then weight the results by past election performance. So the lib dems, despite regularly being beaten by Elvis impersonators and "save the Church Road bus shelter" candidates since 2010, are given credit for their 20+% 2010 result. The Greens on the other hand ( Gawd knows which newspaper they think Greens read) are weighted by their 1%.
When it comes to UKIP it appears to be outright bias rather than statistical anomalies in sampling. Indeed the concept of a UKIP surge seems to be more to explain the deliberately low weight given to them during a campaign than an actually event on the ground. Reckless and Carswell won big despite polls showing they were losing.
It is interesting to note that of the 27% and 36% who tribally voted Labour and Tory in 2010 around a seventh ( 9% overall) are likely dead. Also the turnout ( 65% in 2010) is historically on the rise which isn't explained by young voters and likely isn't people who regularly buy papers.
Please or to access all these features
Please
or
to access all these features
News
Opinion polls... Can't even be close to right.
11 replies
Spinflight · 04/05/2015 17:35
OP posts:
Please create an account
To comment on this thread you need to create a Mumsnet account.