How many more elections until we stop paying attention to polls

Opinion research in the internet eraI have to won­der why phone polls are still news and still given so much cre­dence both by the media, media con­sumers and the polit­i­cal class gen­er­ally. Almost the most inter­est­ing thing about the Ontario elec­tion (apart from the dis­in­te­gra­tion of Tim Hudak's writ­ten-for-buz­zfeed cam­paign plat­form) has been how com­pletely whacked opin­ion poll num­bers have been.

Through­out my ever-length­en­ing life polls - regard­less of the firm con­duct­ing them - have always marched in lock­step toward elec­tion day, to the point where the pun­ditry in the 1990s was mus­ing about just can­celling elec­tions and declar­ing a poll-o-cracy.

The ubiq­uity of the land­line, near com­plete cov­er­age of the phone book, the lack of call dis­play, made the polls almost scary con­sis­tent and a much more accu­rate pre­dic­tion of how votes were cast.

But now you've got Bell exec­u­tives telling stu­dent media land­line use is down 50 per cent, and Sta­tis­tics Canada says that only two thirds of Cana­dian homes have land­lines.

That is huge. Poll­sters are required by law to announce their mar­gin of error which they usu­ally put in the three to five per cent cat­e­gory. But now that they can­not get at one third - that would be 33 per cent - of the peo­ple whose opin­ions they claim to mea­sure, surely the mar­gin of error is now much much higher.

The roller coaster ride of polling num­bers through­out this cam­paign is surely evi­dence of that. The "against all opin­ion research evi­dence" vic­tory of the BC lib­er­als over the NDP is surely more.

Add to the fact that so many peo­ple have call dis­play and only pick up the phone for num­bers they know - or at the very least avoid num­bers that look like mar­keters or sur­vey­ors - and I have to won­der how long are we going to put any stock in them at all?

Surely the opin­ion research indus­try is feel­ing the sands shift­ing beneath their feet. Are they com­pen­sat­ing. Are they try­ing to fig­ure out what sort of peo­ple drop their land­lines and how they're likely to vote com­pared to peo­ple who keep their land­lines and answer the phone no mat­ter who's call­ing.

Or are they hum­ming along pre­tend­ing none of this is hap­pen­ing and hop­ing no one notices that - to bor­row a Rick Mer­cer line - they're basi­cally throw­ing darts?

What­ever the case, I'm going to vote today. I hope you are too.