pn812. Covid and the lastest One TV News Colmar Brunton Poll

Colmar Brunton

The latest Colman Brunton Poll largely confirms the results of the two earlier polls by Curia and Talbot Mills reported in my previous post (pn809). Its results, however, are more detailed and the methodology is spelt out, and therefore possibly more reliable. For example:

"Between 6-10th November 2021 1001 eligible voters were polled by mobile phone (501) and online (500). The maximum sampling error is approximately ±3.1%-points at the 95% confidence level. The data has been weighted to align with Stats NZ population counts for age, gender, region and ethnic identification. The sample for mobile phones is selected by random dialling using probability sampling, and the online sample is collected using an online panel."

I'm confident the results are reasonably accurate, at least for the four larger parties, but I have three small reservations: 

1) Those polled by mobile phone were randomly selected, but the online results were chosen from a previously selected panel. More information is needed on how the panel was selected; 

2) "Percentages do not add to 100 due to rounding."  They certainly don't. Party Support totalled 108, possibly close to 10% of those polled.

And the total percentage for Preferred PM only came to 91 which suggests something wrong or missing. 

Colmar Brunton says the results were rounded up or down to the nearest whole number. In my opinion, it would have been far better to round only after the first decimal point; 

3) Percentages for "Don't know" and "refused to answer", which totalled 9% in the current poll, were not given for the earlier September poll which makes true comparisons difficult.  

Here are the main results in a complete table which is the better way to show such information than the fancy ways used on TV:

The increase in National Party support from September to November seems to have come at the expense of Labour, not ACT. I think the party support question is more important than the preferred PM.  Too much negative attention has been given to Judith Collins, and the National support is fragmented.

The results seem to reflect reactions to how people think Labour has handled the pandemic. No other policy concerns are apparent.  Indeed, the results of the question on how people saw the economy heading also see Labour heading downhill, even in the context of contrary data.

Yet, as Finance Minister Grant Robinson says, 

the economy is in fact doing rather well.  We have low unemployment, very strong economic growth and manageable debt.  

He thinks the poll results reflect our reaction to Covid (and not Labour's policies) which is "affecting the mood of people. That is completely understandable."

So there you have it.  The trends in our choices of preferred —and not so preferred— political party and Prime Minister reflect in large part how we're responding — and how we see them responding— to the pandemic.  

It is not an accurate guide to what we think of their policies on housing, poverty, education, general health,  immigration, the environment and so on which come election time will be more to the forefront of the news. 

Colman Brunton usually indicates the environment or context that could influence poll results but if they did so with this poll, it was unfortunately omitted in the OneNews report. 

Was there nothing other than a Covid context affecting this poll's results?

-- ACW