Introducing the Politicalbetting Polling Averages

With the declaration of the May 5th General election, only, perhaps, forty-eight hours away a big decision for gamblers could be whether or not they believe the telephone pollsters or are they going to put their money on those firms that use other methods to assess public opinion.

For a big gap has emerged between the firms that rely on the phone and those that don’t and during the election campaign we will be producing the Politicalbetting Polling Averages which distinguish between them so that punters can see the difference. Based on their latest published surveys these are the figures:-

Telephone interview pollsters: LAB 40.25%: CON 32.25%: LD 19.5%
Non-telephone pollsters: LAB 36%: CON 35.5%: LD 21%

For this the phone pollsters are ICM, NOP, Populus and Communicate Research. The non-phone pollsters are Mori, which uses quota sampling and interviews people in their own homes, and YouGov which uses the internet to survey opinion from a specially recruited panel. Anthony Wells’s excellent UK Polling Report site has good descriptions of the way each of the firms operates.

The methodology distinction coud be significant because on a uniform national swing using Martin Baxter’s calculator the telephone interview pollsters’ average produces a projected Labour majority of 148 while the non-telephone pollsters’ average comes out with Labour majority of 50.

On the spread betting markets the balance of money is currently nearer to the non-phone pollsters. The latest IG Index projected Labour majority spread is 56-62 seats.

Mike Smithson