Why the PM re­ally is on an up­ward trend

Scottish Daily Mail - - The Brexmas Election - Philip Cow­ley is Pro­fes­sor of Pol­i­tics at Queen Mary Univer­sity of Lon­don and one of the edi­tors of Sex, Lies And Pol­i­tics: The Se­cret In­flu­ences That Drive Our Po­lit­i­cal Choices, pub­lished by Bite­back Philip Cow­ley’s

OF the 28 na­tion­wide opin­ion polls car­ried out since Par­lia­ment voted for an elec­tion, all agree that the Con­ser­va­tives are ahead – yet they vary hugely on by how much.

One, for ex­am­ple, has sug­gested the lead is 17 per­cent­age points, while an­other puts it at just six. How are you sup­posed to make sense of that?

Well, per­haps the sin­gle most im­por­tant thing to re­mem­ber with opin­ion polling is that re­sults are not as pre­cise as they are some­times pre­sented.

Even a poll run by a rep­utable com­pany that has ag­o­nised for hours over its ques­tion word­ing and tar­geted a rep­re­sen­ta­tive sam­ple of the pop­u­la­tion has the po­ten­tial for er­ror of plus or mi­nus three per­cent­age points on any sin­gle find­ing.

This gap can in­crease when de­ter­min­ing the lead of one party over an­other, which re­lies on us­ing two vary­ing val­ues.

So, as some­one who spends al­most all of his time analysing such polls, my ad­vice to you is this – never get ex­cited about small changes in party sup­port in a sin­gle poll. In­stead, fo­cus on trends over time.

It’s also worth not­ing that poll­sters dif­fer in how they go about their business – how they con­tact peo­ple, how they weight the re­sults and so on.

This pro­duces vari­a­tions, which means there is lit­tle point com­par­ing a poll from one com­pany di­rectly with one from an­other.

Yet for all these pit­falls, some trends have be­come ap­par­ent in re­cent weeks.

The Tories and Labour have, for ex­am­ple, trended up­wards since the elec­tion was called, at the ex­pense of both the Lib Dems and the Brexit Party.

As a re­sult, the com­bined share of the vote for the two main par­ties has av­er­aged 71 per cent in the last nine polls.

That may seem ob­vi­ous now, but don’t for­get there were polls in May and June which had the Brexit Party and/or the Lib Dems in first or sec­ond place. Since then, the squeeze has been re­morse­less.

The Con­ser­va­tives will be es­pe­cially pleased that there is as yet no sign of Labour eat­ing into their poll lead. For, although by this point of the 2017 cam­paign the Con­ser­va­tives had a larger per­cent­age lead, there were then al­ready signs of Labour catch­ing up.

The re­verse is now true, with the Con­ser­va­tives slightly pulling ahead. In­deed, of the five largest Con­ser­va­tive poll leads dur­ing the elec­tion, four have come in polls in the last week.

One should be­ware, how­ever, that the most re­cent polling shifts have been ex­ag­ger­ated by the de­ci­sion of the Brexit Party to stand down in those seats which the Con­ser­va­tives won at the last elec­tion, mean­ing sev­eral of the more re­cent polls have only of­fered sur­vey re­spon­dents the choice of choos­ing the Brexit Party if it’s stand­ing in their con­stituency.

The most re­cent Con­ser­va­tive uptick in sup­port is there­fore to be ex­pected. In fact, if the Con­ser­va­tives were not trend­ing up­wards right now, it would be a sign that some­thing was go­ing se­ri­ously wrong with their cam­paign.

And one should never for­get there is al­ways the pos­si­bil­ity that a large na­tional lead won’t trans­late into a Com­mons ma­jor­ity, if it is mostly caused by pil­ing up ever larger ma­jori­ties in seats they al­ready hold.

So, while look­ing at polling trends may be more use­ful than study­ing snap­shots, it’s still no sub­sti­tute for a crys­tal ball.

Newspapers in English

Newspapers from UK

© PressReader. All rights reserved.