BC Municipal Madness: Making sense of the polls

VintageBallot

There’s something happening here.  What it is ain’t exactly clear.

     -Buffalo Springfield

For what it’s worth, this is not your typical British Columbia municipal election cycle.  There is no man with a gun, but we are already seeing biblical-sized change:

Amid the chaos and emerging public opinion research: 5 things to consider looking at public political polls:

  1. Start with the fundamentals. Especially in the early days of a campaign,  the relative brand strength of the parties and candidate name recognition can be better predictors than a ballot question that no one has really considered yet.
  2. Ease into the horserace.  Asking mayoral vote preference is like asking about entré choice for a lunch on October 20.  Both answers largely depend on the menu options.  Right now, few have even opened their menus.  Be careful about reading too much into pre-Thanksgiving ballot numbers.
  3. Geography matters.  In Vancouver municipal politics, there is a North / South divide (sorry Eastside / Westside traditionalists.)  Surrey has a large population spread over enormous geography.  This results in some distinct neighbourhoods with great divides between them—Newton is not Crescent Beach.  My point that is a good pollster will consider regional differences when sampling and weighting.  Find out how.
  4. Elections are decided by those that show up.  In BC civic elections, most don’t. Public polls typically report the intent of eligible voters.  This is why experienced It is a good idea told consider voter turnout in their analysis.  Turn out tends to increase with age and can vary dramatically between demographic groups.
  5. Get the crosstabs.  These are the tables that show questions tabulated against each other.  Most pollsters will provide tables of a publicly released poll’s “marquee questions” cross-tabulated against key demographic and geographic segments.  If you are looking for insight, the crosstabs are where you’ll find it.  If these aren’t available, ask why not.

NEXT BLOG: The 5 questions I would ask pollsters if I were a journalist