It’s only one poll

The latest Roy Morgan poll continues the positive trend for National:

. . . Today’s New Zealand Roy Morgan Poll shows a strong gain in support for National (52.5%, up 7%) now at their highest since before the last New Zealand Election and well ahead of a potential Labour/Greens alliance (38%, down 6%) – almost matching their performance at the 2011 New Zealand Election at which the two parties polled a combined 38.5%.

Support for Key’s Coalition partners has also improved with the Maori Party 1.5% (up 0.5%), ACT NZ (1%, up 0.5%) and United Future 0% (unchanged).

Support has fallen significantly for all Opposition parties with the Labour Party down 1.5% to 29%, the Greens down 4.5% to 9% (the lowest support for the Greens since September 2011), New Zealand First 4.5% (down 1.5%) and Mana Party 0.5% (down 0.5%). Support for the Conservative Party of NZ is 1% (unchanged) and the Internet Party is 0.5% (unchanged).

If a National Election were held now the latest New Zealand Roy Morgan Poll shows that the result would be a landslide victory for the National Party and a third term for Prime Minister John Key. . .

But wait, there’s more good news:

The latest NZ Roy Morgan Government Confidence Rating has also improved considerably – up 8.5pts to 140.5pts with 64.5% (up 4.5%) of New Zealanders saying New Zealand is ‘heading in the right direction’ compared to 24% (down 4%) that say New Zealand is ‘heading in the wrong direction’.

Gary Morgan says:

“Today’s New Zealand Roy Morgan Poll shows a strong positive response to the predicted Budget Surplus of $372 million handed down by Finance Minister Bill English with National surging to 52.5% (up 7%) – it’s highest since the last New Zealand Election. National has surged to a huge lead over a potential Labour/ Greens alliance (38%, down 6%).

“The closer the election, it appears the less support there is for the main opposition parties with support for Labour (29%, down 1.5%) now stuck below the level that prompted the resignation of previous leader David Shearer for most of 2014. The initial surge provided by David Cunliffe has well and truly worn off. In addition the Greens (9%, down 4.5%) have slumped to their lowest level of support since before the last New Zealand election after announcing last weekend a proposal to introduce a Carbon Tax in New Zealand in place of the current Emissions Trading Scheme.

“Last week’s merger announcement of the Internet Party (0.5%) and Mana Party (0.5%) to contest this year’s election offers both parties a better chance of attaining the 5% threshold required to elect a slate of Party List MPs. However, the combined support for the two parties has never exceeded 2%, and it would appear unlikely the merged party can bridge this gap in the next few months.” . . .

Polls can be too good, of course.

This level of support for National could make supporters complacent.

Some might think they can afford to vote for another party, others might not bother to vote at all.

However, while it continues the positive trend for National of other recent polls, it is only one poll and the one which is usually regarded as the least reliable.

But is it?

Thomas Lumley at Stats Chat, says it’s not:

. . . In fact, there’s not much difference between the major polling companies in the variability of their estimates. . .

There really is not much to see here. So why do people feel that Roy Morgan comes out with strange results more often? Probably because Roy Morgan comes out with results more often.

For example, the proportion of poll-to-poll changes over 3 percentage points is 0.22 for One News/Colmar Brunton, 0.18 for Roy Morgan, and 0.23 for 3 News/Reid Research, all about the same, but the number of changes over 3 percentage points in this time frame is 5 for One News/Colmar Brunton, 14 for Roy Morgan, and 5 for 3 News/Reid Research. . .

What that shows is voter preference is volatile and that more frequent polls reflect that volatility.

That’s why it doesn’t pay to get too excited about a single poll, or even several with the same trend.

The volatility of support merely reinforces the oft repeated phrase, there’s only one poll that counts.



3 Responses to It’s only one poll

  1. Neil says:

    The trend is what is important. National has been ahead for the whole of the three years. Only briefly have we seen the left “rabble” lead in a coalition possibility. There appears to be a cementing of this lead for National.
    Compare to Europe where populist parties have made ground because of the struggling economic situation.Europe is still a basket case.
    Not so in New Zealand.People have more confidence in NZ. Witness Winston First who is not getting traction among the “angry” brigade. Good news started back last year and people close to politics had a fair idea this current situation would occur. Things are moving ahead.
    Even the purveyors of the inequality clause must be disappointed to see crime down, reduced welfare rolls and employment moving ahead. Still a lot of room to break that 20% poverty trap in education. That will take years and changes in attitudes..
    The Banks fiasco will come and go however I don’t see it as a major problem in the election. People want competent government, this counts out the Greens and the current Labour Party.
    However, to right wing parties get out there and campaign !!! Let’s whip the left wingers.


  2. TraceyS says:

    I’m sure that voter preference is pretty variable over time but to what degree? It is difficult to know how much of the variation in the Roy Morgan poll is due to variations in voter intentions or variation in the measurement process itself. Some variation could be due to differences in, for example;

    – procedures for collecting the data
    – level of training researchers receive
    – the interview script
    – public perception of the particular poll
    – how much time is given for respondents to answer
    – qualities that are sought when recruiting interviewers
    – overall quality control

    These kind of differences could definitely affect the amount of variation in results between, and within, different research organisations. The margin of error doesn’t explain this as that is representing random sampling error, not systematic errors as described above.

    The way to understand how much of the variance is truly attributable to variation in voter intentions would be for other polls to run more frequently and see if they prove to be more volatile over shorter periods or not. If not, it would tend to suggest that some research organisations have more consistent, accurate and tightly controlled measurement process than others, which is entirely likely.


  3. badtom:
    One party wants to legalise a medical procedure & stop climate change. Another says electoral fraud is ok. Guess which one’s called “crazy”?


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: