June 7, 2025 • 13 mins
The latest 1News/Verian poll has both National and Labour down, while an RNZ-Reid poll done in the same week has National down and Labour up, and a Taxpayers' Union-Curia poll had both up.
You see the same with overlapping polls by different agencies showing significantly different results in preferred Prime Minister.
So why are polls done this far out from an election?
LISTEN ABOVE
See omnystudio.com/listener for privacy information.
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:05):
You're listening to the Weekend collective podcast from News talks'd.
Speaker 2 (00:09):
Be right into the politics solo. Last week, as I
referenced briefly, the latest one News variant pole had national
and labor down, while an r NZ Read poll done
the same week had national down, Labor up, and one
of them had one side forming a government, the other
had another side, and the Taxpayers Union Courier poll had
both up. You'll see the same with overlapping polls by
(00:30):
different agencies showing significantly different results. So why are we
doing polls this far out from an election? What are
we looking to learn from them? And so I thought
we'd have a chat with our Taxpayers Union executive director
Jordan Williams, who joins me now, Jordan, good afternoon.
Speaker 3 (00:44):
Get Tim.
Speaker 2 (00:45):
What's the point of political polls at this time of year.
Speaker 3 (00:49):
Well, it gives you a snapshot on the government. Is
that relative to the other parties of course, or you
always say, you know, don't look at any particular poll.
It's the trend that matters. And what all the polls
are telling us is that it's actually very close each
In the TVNZ and the r n Z Read research.
You know, TVNZ had a majority of the center right,
(01:10):
I think of two seats. R INZ had a majority
of the cent to the center left of two seats.
Both of those tell you it's very close indeed.
Speaker 2 (01:19):
Okay, so just explain to me. I've got a bit
of a beef, which you can probably allay with the
margin of area of three point one percent. So one
News spend a lot of time talking about all such
and such as down one or down two. It's all
within the margin of era being three point one percent,
which I understand means if somebody scores, say ten percent,
we're confident that they're in reality between seven and thirteen percent.
(01:43):
So can you put that in perspective for me, because
it's almost never any not even lip service paid to it.
Speaker 3 (01:51):
That's not quite right, of course, So that when they
talk about the the margin of era, what they mean
is sorry to put a stats head on, but go
for it. It's an era, it's a confidence interval, and
so you you know, statistically it's it's either ninety or
ninety five percent confidence it's within that bound. So there'll
(02:12):
always be rogue poles, you know, I talk about with
a taxais union pole, you know, one and twenty in
theory is rogue, which is why I say you never
put a particular emphasis on one pole. It is the trend.
And that's really why the tax Boers Union started polling
with our monthly polls is because there's the media have
pulled right back. What you get is, I mean ran
zet I think are polling every three or six months,
(02:35):
and you sort of that the media like to dial
it up to eleven, that there's some huge significance in
this pole. The good thing about monthly polling is that
if you've got sort of three consistent poles that start
to show a trend, then that's when you would put
put meaning into it. But just to pull you up.
The confidence interval or the margin of vera is done
(02:56):
at the fiftieth percentile, and if you go back to
your fifth form stats, you'll know that as the number
goes down, the confidence into or the margin of era
goes down. It really infuriates me when I read in
a court judgment some years ago this stupid High court
judges not understanding basic stats. They talk about you know,
(03:17):
that's parties in the margin of era, meaning it could
it could have it could have none, or it could
be double that's nonsense on Steltz because of course the
margin of ERA is about half once you get to
blow about ten percent.
Speaker 2 (03:29):
Okay, So I guess the margin of ERAa is more
relevant to the parties who are polling thirty forty percent.
So if somebody's polling forty percent, it could be you know,
there's a ninety five percent confidence. Basically it could be
within close to three points other side.
Speaker 3 (03:43):
Yeah, And of course the secret source for the different
polsters is that there are various different ratings. I mean
I can, of course I'm not a polster. I simply
contract of polling companies to do the texpos. You can
poll each month, but our supplier, we go back through
the elections, has tended to get that right than say
(04:05):
the closest the least wrong out of all the poles.
And in some of that secret source or the difference
between perhaps Radium New Zealand Television New Zealand, is each
polster waits on slightly different things. So for example, some
polls I think that the are in ZED won the
Australian company that did the worst of the Australian election
(04:27):
are variant. You know, I think they are online only
and you get some that have a particular waiting for
phone calls, some which include cell phones now some just
online panels which can be which you know, attract a
different demographic. And then of course you've got I think
television New Zealand weight for all of age, ethnicity and
(04:49):
education status, where I don't think are in z weight
for the education status. So there's lots of sort of
that modeling behind the scenes is kind of what you're
paying your expert pole stuff for their expertise than were
Sorry they won't let me say this, but from a
third party suspective, it looks a little bit of an
art rather than a strict science. Yeah.
Speaker 2 (05:11):
Actually, by the way, I think I'm pretty sure read
research is online only and I think very into a
mixture for my memory around. Yeah, but I guess the
question is, I mean, how how much does methodology matter?
Because I just asked chat GPT and it gave me
some interesting answers around the propensity of online polling to
(05:35):
maybe favor people who are a little bit more technologically
literate and also anyway, but I mean, what do you
think about the method of online versus phone versus I
don't know, in the middle.
Speaker 3 (05:46):
Of the studio you get every three years, you get
a sort of you know that they have to pony
up and you get the actual results in the leeching,
and of course you can compare against all the different
polling companies. The other reason we use polling and we've
found it so useful though, is the polling of issues.
So do you think back years ago to when we
were doing our three Waters campaign. As part of the polling,
(06:10):
we were not necessarily releasing it every month, but that
for us was quite a useful benchmark to know how
we were going in terms of that campaign. And it
was quite quite incredible the dramatic change within a month,
and it was the moment we won. It was actually
when the mayors of Auckland and christ Church came out
together against three Waters. Now, of course it required twelve
(06:32):
months of campaigning beforehand to get to it, but we
could tell just immediately a dramatic jump. And so from
a lobbying campaign perspective, and I'm sure the parties would
they might not be quite as open, but that's the
sort of thing they're looking for. Similarly, around issues, you
(06:53):
will know that we often release favorable unfavorable results on
politicians I can tell you that I get politicians calling
me saying, why don't you put my name in the
preferred PM. Well, it's actually unprompted, I won't name, and
shame that those particular politicians come on the other one
(07:17):
that we don't release it all. We certainly do release
it every month to our members and supporters, but that
is country right and wrong direction? Do you think the
country is going in the right or wrong direction? And
that tends to be a precursor to the party vote.
I've sort of noticed that if that dramatically shifts, it's
(07:37):
usually three or four months later you start to see
a shift in the party vote between the government and
the opposition.
Speaker 2 (07:44):
I guess i'd probably answer the question do you think
the government's going in the right or wrong direction? Because
the country could be going in the wrong direction for
a whole bunch of reasons, I guess. But that's the
art of the poll, isn't it. Are you happy with
the way the media report polls?
Speaker 3 (07:57):
Well? I find it interesting that and I think it's
it's sort of pasted the parcel of having particularly television
US out and not doing as many poles as they
used to. That they dial the results of every pole
up to eleven. I mean, you think about it. You know,
we're a humble taxpayers union. We're a pressure group. You know,
(08:18):
we have we wear our mission on our sleeves. And
yet I'm pretty confident that a lot of our poles
we just play down the middle for what they are,
as opposed to dial up to eleven as some sort
of disaster melt down the beehive because of one single pole.
As I say, poles will bounce around. It is the
trend over time that matters. And that's why I think
(08:39):
from our perspective, monthly polling is so good because once
you're starting to see sort of three poles in a
row showing a significant difference from say six months ago,
that NAT is telling you something much more than a
single pole or a single snapshot.
Speaker 2 (08:53):
What about internal what about parties political parties internal polling?
What sort of caution notes of caution should we have
if any on those ones, Well, they'll.
Speaker 3 (09:03):
Only lead them of cocause because we can see what's
going on. I find it delightful the way that the
parties only lead their internal poles when it's helpful to
them or to trip up the other side. They're probably
more interested in the reaction to political scandal or events
of the day than a group like the Texpaths Union.
(09:24):
But the reports that we send to our we called
our Texpower Caucus, our most generous supporters, is in effect
the reports that the leader of the opposition and the
Prime Minister is getting also every few weeks or every month,
depending on the party, with that real granular data. So
an example would be how was the center right or
(09:45):
how is a particular party doing in rural communities or
with a particular gender for example. Now, of course, as
you get more and more specific, your sample size gets
lower and lower, and your margin of error gets larger.
But again, you put a number of polls together, you
start to get really meaningful data. We did that with
(10:08):
the mayors last year. We asked over a series of
about eight poles to rank your performance or rate the
performance of your local mayor. And then for those for
those cities where we had enough I think it was
sort of over three hundred responses, we ranked the performance
of every mayor. Now we got some angry emails, but
(10:29):
why didn't you do our mair? Well? The short point is,
and for some small councils, we just didn't have enough
responses even over a long time frame.
Speaker 2 (10:38):
Statistically meaningful data should have a fun inbox to look
at just a couple of quick questions. What's the red
flag tip for you that makes you go no, I'm
not buying that result. Is it just a statistical anomaly
where it stands out from all the others? I mean,
would there be a time when you look at your
own pole and go, oh, hell.
Speaker 3 (10:54):
Well, of course, I'm One of the great things about
doing polling is that any political leader will take your
phone call every month to give them the heads up,
as it really is a courtesy all the leaders officers
or the leaders of each of the political parties and
give them a heads up three or four hours before
public release. And the line I have found I use
(11:16):
when it is quite a shift in the polls, as
I say to them, you know, well, you know one
in twenties a rogue and this could be a rogue.
This hasn't setted recent polls, but then again there's been
a few of our polls where subsequent Television New Zealand
staff Roy Morgan have actually confirmed our results. So you
don't really know at the time, but if it's if
(11:41):
you can't explain a dramatic shift. The one I find
really interesting is that January. We usually first in the
field in sort of mid January, and there's been a
few summers where there's been quite a shift in the
polls that sort of we haven't been able to unable
to explain subsequently confirmed by subsequent polls. But I will
admit that those poles I have thought this be a
(12:05):
roague because as I say, one in twenty has those
two large parties outside the confidence in the marketing of error.
Speaker 2 (12:13):
Hey, I really appreciate your time this Aftener and Jordan, Hey,
just one quick question. Actually, have you ever done a
poll on how we feel about our nuclear free status?
Because I reckon that there's a lot of assumptions made
about that. This is just one a completely different issue.
But can I nominate that for a question? Do you
think we should cling on to our nuclear free status?
And I reckon you do that you'll be in for
an interesting result.
Speaker 3 (12:32):
I actally think we should. We have quite a long
list of things that we'd love to poll on, but
we tend to stick to things that are in the mission,
in the in our wheelhouse of taxes, government waste government accountability.
But I should add that the only thing is is
that if we didn't release it and it leaked, I
would worry that people would say there was some sort
(12:52):
of conspiracy. I love polling on sort of interesting issues,
but I am duty bound to keep, you know, keep
it within the mission.
Speaker 2 (12:59):
Okay, Well, stick around because we're going to ask callers
what they want, what question they'd like you to to
poll on next and I'm going to throw that one
over the great Okay, geers, Jordan, thanks very much for
your time. All rights Jordan Williams, TAXPAS Union executive director.
Speaker 1 (13:15):
For more from the Weekend Collective, listen live to News
talk edb weekends from three pm, or follow the podcast
on iHeartRadio
You're listening to the Weekend collective podcast from News talks'd.
Speaker 2 (00:09):
Be right into the politics solo. Last week, as I
referenced briefly, the latest one News variant pole had national
and labor down, while an r NZ Read poll done
the same week had national down, Labor up, and one
of them had one side forming a government, the other
had another side, and the Taxpayers Union Courier poll had
both up. You'll see the same with overlapping polls by
(00:30):
different agencies showing significantly different results. So why are we
doing polls this far out from an election? What are
we looking to learn from them? And so I thought
we'd have a chat with our Taxpayers Union executive director
Jordan Williams, who joins me now, Jordan, good afternoon.
Speaker 3 (00:44):
Get Tim.
Speaker 2 (00:45):
What's the point of political polls at this time of year.
Speaker 3 (00:49):
Well, it gives you a snapshot on the government. Is
that relative to the other parties of course, or you
always say, you know, don't look at any particular poll.
It's the trend that matters. And what all the polls
are telling us is that it's actually very close each
In the TVNZ and the r n Z Read research.
You know, TVNZ had a majority of the center right,
(01:10):
I think of two seats. R INZ had a majority
of the cent to the center left of two seats.
Both of those tell you it's very close indeed.
Speaker 2 (01:19):
Okay, so just explain to me. I've got a bit
of a beef, which you can probably allay with the
margin of area of three point one percent. So one
News spend a lot of time talking about all such
and such as down one or down two. It's all
within the margin of era being three point one percent,
which I understand means if somebody scores, say ten percent,
we're confident that they're in reality between seven and thirteen percent.
(01:43):
So can you put that in perspective for me, because
it's almost never any not even lip service paid to it.
Speaker 3 (01:51):
That's not quite right, of course, So that when they
talk about the the margin of era, what they mean
is sorry to put a stats head on, but go
for it. It's an era, it's a confidence interval, and
so you you know, statistically it's it's either ninety or
ninety five percent confidence it's within that bound. So there'll
(02:12):
always be rogue poles, you know, I talk about with
a taxais union pole, you know, one and twenty in
theory is rogue, which is why I say you never
put a particular emphasis on one pole. It is the trend.
And that's really why the tax Boers Union started polling
with our monthly polls is because there's the media have
pulled right back. What you get is, I mean ran
zet I think are polling every three or six months,
(02:35):
and you sort of that the media like to dial
it up to eleven, that there's some huge significance in
this pole. The good thing about monthly polling is that
if you've got sort of three consistent poles that start
to show a trend, then that's when you would put
put meaning into it. But just to pull you up.
The confidence interval or the margin of vera is done
(02:56):
at the fiftieth percentile, and if you go back to
your fifth form stats, you'll know that as the number
goes down, the confidence into or the margin of era
goes down. It really infuriates me when I read in
a court judgment some years ago this stupid High court
judges not understanding basic stats. They talk about you know,
(03:17):
that's parties in the margin of era, meaning it could
it could have it could have none, or it could
be double that's nonsense on Steltz because of course the
margin of ERA is about half once you get to
blow about ten percent.
Speaker 2 (03:29):
Okay, So I guess the margin of ERAa is more
relevant to the parties who are polling thirty forty percent.
So if somebody's polling forty percent, it could be you know,
there's a ninety five percent confidence. Basically it could be
within close to three points other side.
Speaker 3 (03:43):
Yeah, And of course the secret source for the different
polsters is that there are various different ratings. I mean
I can, of course I'm not a polster. I simply
contract of polling companies to do the texpos. You can
poll each month, but our supplier, we go back through
the elections, has tended to get that right than say
(04:05):
the closest the least wrong out of all the poles.
And in some of that secret source or the difference
between perhaps Radium New Zealand Television New Zealand, is each
polster waits on slightly different things. So for example, some
polls I think that the are in ZED won the
Australian company that did the worst of the Australian election
(04:27):
are variant. You know, I think they are online only
and you get some that have a particular waiting for
phone calls, some which include cell phones now some just
online panels which can be which you know, attract a
different demographic. And then of course you've got I think
television New Zealand weight for all of age, ethnicity and
(04:49):
education status, where I don't think are in z weight
for the education status. So there's lots of sort of
that modeling behind the scenes is kind of what you're
paying your expert pole stuff for their expertise than were
Sorry they won't let me say this, but from a
third party suspective, it looks a little bit of an
art rather than a strict science. Yeah.
Speaker 2 (05:11):
Actually, by the way, I think I'm pretty sure read
research is online only and I think very into a
mixture for my memory around. Yeah, but I guess the
question is, I mean, how how much does methodology matter?
Because I just asked chat GPT and it gave me
some interesting answers around the propensity of online polling to
(05:35):
maybe favor people who are a little bit more technologically
literate and also anyway, but I mean, what do you
think about the method of online versus phone versus I
don't know, in the middle.
Speaker 3 (05:46):
Of the studio you get every three years, you get
a sort of you know that they have to pony
up and you get the actual results in the leeching,
and of course you can compare against all the different
polling companies. The other reason we use polling and we've
found it so useful though, is the polling of issues.
So do you think back years ago to when we
were doing our three Waters campaign. As part of the polling,
(06:10):
we were not necessarily releasing it every month, but that
for us was quite a useful benchmark to know how
we were going in terms of that campaign. And it
was quite quite incredible the dramatic change within a month,
and it was the moment we won. It was actually
when the mayors of Auckland and christ Church came out
together against three Waters. Now, of course it required twelve
(06:32):
months of campaigning beforehand to get to it, but we
could tell just immediately a dramatic jump. And so from
a lobbying campaign perspective, and I'm sure the parties would
they might not be quite as open, but that's the
sort of thing they're looking for. Similarly, around issues, you
(06:53):
will know that we often release favorable unfavorable results on
politicians I can tell you that I get politicians calling
me saying, why don't you put my name in the
preferred PM. Well, it's actually unprompted, I won't name, and
shame that those particular politicians come on the other one
(07:17):
that we don't release it all. We certainly do release
it every month to our members and supporters, but that
is country right and wrong direction? Do you think the
country is going in the right or wrong direction? And
that tends to be a precursor to the party vote.
I've sort of noticed that if that dramatically shifts, it's
(07:37):
usually three or four months later you start to see
a shift in the party vote between the government and
the opposition.
Speaker 2 (07:44):
I guess i'd probably answer the question do you think
the government's going in the right or wrong direction? Because
the country could be going in the wrong direction for
a whole bunch of reasons, I guess. But that's the
art of the poll, isn't it. Are you happy with
the way the media report polls?
Speaker 3 (07:57):
Well? I find it interesting that and I think it's
it's sort of pasted the parcel of having particularly television
US out and not doing as many poles as they
used to. That they dial the results of every pole
up to eleven. I mean, you think about it. You know,
we're a humble taxpayers union. We're a pressure group. You know,
(08:18):
we have we wear our mission on our sleeves. And
yet I'm pretty confident that a lot of our poles
we just play down the middle for what they are,
as opposed to dial up to eleven as some sort
of disaster melt down the beehive because of one single pole.
As I say, poles will bounce around. It is the
trend over time that matters. And that's why I think
(08:39):
from our perspective, monthly polling is so good because once
you're starting to see sort of three poles in a
row showing a significant difference from say six months ago,
that NAT is telling you something much more than a
single pole or a single snapshot.
Speaker 2 (08:53):
What about internal what about parties political parties internal polling?
What sort of caution notes of caution should we have
if any on those ones, Well, they'll.
Speaker 3 (09:03):
Only lead them of cocause because we can see what's
going on. I find it delightful the way that the
parties only lead their internal poles when it's helpful to
them or to trip up the other side. They're probably
more interested in the reaction to political scandal or events
of the day than a group like the Texpaths Union.
(09:24):
But the reports that we send to our we called
our Texpower Caucus, our most generous supporters, is in effect
the reports that the leader of the opposition and the
Prime Minister is getting also every few weeks or every month,
depending on the party, with that real granular data. So
an example would be how was the center right or
(09:45):
how is a particular party doing in rural communities or
with a particular gender for example. Now, of course, as
you get more and more specific, your sample size gets
lower and lower, and your margin of error gets larger.
But again, you put a number of polls together, you
start to get really meaningful data. We did that with
(10:08):
the mayors last year. We asked over a series of
about eight poles to rank your performance or rate the
performance of your local mayor. And then for those for
those cities where we had enough I think it was
sort of over three hundred responses, we ranked the performance
of every mayor. Now we got some angry emails, but
(10:29):
why didn't you do our mair? Well? The short point is,
and for some small councils, we just didn't have enough
responses even over a long time frame.
Speaker 2 (10:38):
Statistically meaningful data should have a fun inbox to look
at just a couple of quick questions. What's the red
flag tip for you that makes you go no, I'm
not buying that result. Is it just a statistical anomaly
where it stands out from all the others? I mean,
would there be a time when you look at your
own pole and go, oh, hell.
Speaker 3 (10:54):
Well, of course, I'm One of the great things about
doing polling is that any political leader will take your
phone call every month to give them the heads up,
as it really is a courtesy all the leaders officers
or the leaders of each of the political parties and
give them a heads up three or four hours before
public release. And the line I have found I use
(11:16):
when it is quite a shift in the polls, as
I say to them, you know, well, you know one
in twenties a rogue and this could be a rogue.
This hasn't setted recent polls, but then again there's been
a few of our polls where subsequent Television New Zealand
staff Roy Morgan have actually confirmed our results. So you
don't really know at the time, but if it's if
(11:41):
you can't explain a dramatic shift. The one I find
really interesting is that January. We usually first in the
field in sort of mid January, and there's been a
few summers where there's been quite a shift in the
polls that sort of we haven't been able to unable
to explain subsequently confirmed by subsequent polls. But I will
admit that those poles I have thought this be a
(12:05):
roague because as I say, one in twenty has those
two large parties outside the confidence in the marketing of error.
Speaker 2 (12:13):
Hey, I really appreciate your time this Aftener and Jordan, Hey,
just one quick question. Actually, have you ever done a
poll on how we feel about our nuclear free status?
Because I reckon that there's a lot of assumptions made
about that. This is just one a completely different issue.
But can I nominate that for a question? Do you
think we should cling on to our nuclear free status?
And I reckon you do that you'll be in for
an interesting result.
Speaker 3 (12:32):
I actally think we should. We have quite a long
list of things that we'd love to poll on, but
we tend to stick to things that are in the mission,
in the in our wheelhouse of taxes, government waste government accountability.
But I should add that the only thing is is
that if we didn't release it and it leaked, I
would worry that people would say there was some sort
(12:52):
of conspiracy. I love polling on sort of interesting issues,
but I am duty bound to keep, you know, keep
it within the mission.
Speaker 2 (12:59):
Okay, Well, stick around because we're going to ask callers
what they want, what question they'd like you to to
poll on next and I'm going to throw that one
over the great Okay, geers, Jordan, thanks very much for
your time. All rights Jordan Williams, TAXPAS Union executive director.
Speaker 1 (13:15):
For more from the Weekend Collective, listen live to News
talk edb weekends from three pm, or follow the podcast
on iHeartRadio
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.