EMC, Essential Report, peter lewis, political reporting, political research, Politics, Polling, polls, The Drum
First Published on The Drum 21/12/2010
If our politicians are feeling bloody and bruised after a torrid year of spills and thrills, spare a thought for the one group on the national stage that had a harder time of it in 2010 – the humble pollster.
This was the year when hand wringing about the malign impact of political research became a national sport; if there was a problem with politics, political research was the cause. We were to blame for the uninspiring election campaign; for the dumbing down of political reporting; for robbing the ALP of its soul.
So as a purveyor of the Dark Art I want to end the year with a plea for understanding for the minority group I am part of, stealing the catchcry of the Shooters Lobby in the 1990s to proudly declare ‘polls don’t kill people, people kill people’.
Before doing so, I will readily admit that political research can be put to poor use – running focus groups to get ‘permission’ from the electorate to pursue particular policies is to misunderstand the point of such research.
Comments »
Coalition, Election, Essential, Galaxy, Greens, Independents, Labor, Morgan, Newspoll, Nielsen, Polling
All pollsters performed well in estimating the 2PP vote – all were within 0.3% to 1.3% of the current result. The Essential Report and Morgan Research were closest with 51/49. Newspoll’s 50.2/49.8 was next closest with Nielsen and Galaxy 1.3% off at 52/48.
However, a better way to compare the polls is to look at their first preferences for the major parties. Because the 2PP is based on an assumed distribution of preferences – not on the actual measurement of voting intentions.
Although all polls were within the margin of error, based on estimates for the 3 major party groupings, Essential Report was clearly the closest – their average difference being just 0.5%. Most polls over-estimated the Greens vote and Newspoll underestimated the Labor vote.
This is the first Federal election where public online polling has been used extensively, and the performance of the Essential Report poll is significant in that it has shown this methodology can provide reliable and valid measures of public opinion.
| Actual current | Newspoll | Nielsen | Galaxy | Morgan | Essential | |
| Labor | 38.5% | 36.2% | 39% | 38% | 38% | 38% |
| Coalition | 43.5% | 43.2% | 41.5% | 41% | 42% | 43% |
| Greens | 11.4% | 13.9% | 13% | 14% | 13% | 12% |
| Others | 6.6% | 6.5% | 6.5% | 7% | 7% | 7% |
| Average difference (Labor, Coalition & Greens) | 1.7 | 1.4 | 1.9 | 1.2 | 0.5 |
| Actual current | Newspoll | Nielsen | Galaxy | Morgan | Essential | |
| Labor 2PP | 50.7% | 50.2% | 52% | 52 | 51% | 51% |
| Difference | 0.5 | 1.3 | 1.3 | 0.3 | 0.3 |
2PP, Election, Gillard, Labor, Liberal, Polling, Prime Minister, Voting intention
Q. If there was a Federal election held today, to which party would you probably give your first preference?
Q. If you ‘don’t know’ on the above question, which party are you currently leaning to?
1,797 sample size
| First preference/leaning to | 6 months ago | 4 weeks ago | 2 weeks ago | Last week | This week
|
| Liberal | 34% | 38% | 37% | 36% | 36% |
| National | 3% | 3% | 3% | 3% | 3% |
| Total Lib/Nat | 37% | 40% | 40% | 39% | 39% |
| Labor | 47% | 37% | 38% | 42% | 42% |
| Greens | 8% | 12% | 11% | 9% | 11% |
| Family First | 2% | 3% | 3% | 3% | 2% |
| Other/Independent | 6% | 7% | 8% | 7% | 6% |
| 2PP | 6 months ago | 4 weeks ago | 2 weeks ago | Last week | This week
|
| Total Lib/Nat | 43% | 48% | 48% | 46% | 46% |
| Labor | 57% | 52% | 52% | 54% | 54% |
NB. The data in the above tables comprise 2-week averages derived the first preference/leaning to voting questions. Respondents who select ‘don’t know’ are not included in the results.
* Sample is the aggregation of two weeks’ polling data. Comments »
2PP, Abbott, Election, essential report, Greens, Labor, Liberal, Polling, Prime Minister, rudd, Voting intention
Q. If there was a Federal election held today, to which party would you probably give your first preference?
Q. If you ‘don’t know’ on the above question, which party are you currently leaning to?
1,803sample size
| First preference/leaning to | 6 months ago | 4 weeks ago | 2 weeks ago | Last week | This week
|
| Liberal | 34% | 39% | 38% | 37% | 36% |
| National | 3% | 2% | 3% | 3% | 3% |
| Total Lib/Nat | 37% | 41% | 41% | 40% | 39% |
| Labor | 47% | 39% | 35% | 38% | 42% |
| Greens | 8% | 9% | 14% | 11% | 9% |
| Family First | 2% | 2% | 2% | 3% | 3% |
| Other/Independent | 6% | 8% | 8% | 8% | 7% |
| 2PP | 6 months ago | 4 weeks ago | 2 weeks ago | Last week | This week
|
| Total Lib/Nat | 43% | 49% | 49% | 48% | 46% |
| Labor | 57% | 51% | 51% | 52% | 54% |
NB. The data in the above tables comprise 2-week averages derived the first preference/leaning to voting questions. Respondents who select ‘don’t know’ are not included in the results.
* Sample is the aggregation of two weeks’ polling data. Comments »
Election, Gillard, Labor, Polling
Q. Does having Julia Gillard as Prime Minister make you more or less likely to vote for the Labor Party?
| Total | Vote Labor | Vote Lib/Nat | Vote Greens | |
| Total more likely | 26% | 44% | 11% | 31% |
| Total less likely | 24% | 11% | 38% | 22% |
| Much more likely | 10% | 24% | 1% | 5% |
| A bit more likely | 16% | 20% | 10% | 26% |
| A bit less likely | 8% | 8% | 6% | 12% |
| Much less likely | 16% | 3% | 32% | 10% |
| Makes no difference | 41% | 40% | 47% | 40% |
| Don’t know | 8% | 5% | 3% | 7% |
Sample = 818
In aggregate, the results indicate the change will have little impact on voting intentions. 41% say it makes no difference to their vote, 26% say they are more likely to vote Labor and 24% less likely. Among Labor voters 44% say they are more likely to vote Labor and 11% less likely.
Greens voters were overall a little more positive about voting Labor – 31% more likely and 22% less likely. Comments »
andrew bunn, Campaigning, Essential Report, Polling
First Published on Crikey Thursday 24th June
Essential Reports polls taken over the last 2 years show how Kevin Rudd’s approval ratings have declined since he almost unprecedented figures he achieved following his election through to early last year. Up to March last year approval hovered around the high 60% level into the low 70%. Is net approval (i.e. approve minus disapprove) was around the mid 40% level.
Throughout 2009 and early 2010 his approval went into steady decline but still remained in positive territory. At the end of March he recorded 53% approval and 36% disapproval. However, the most dramatic shift occurred in April and May when his approval first entered negative territory with 41% approve and 47% disapprove. The decline appeared to be accelerating.
Some of the reasons for this sudden decline can be found in how the personal perceptions of Kevin Rudd changed from 2009 to May this year. His attribute ratings showed significant falls in the percentage of people who considered him to be a capable leader (72% to 55%), good in a crisis (60%-44%) and trustworthy (51%-41%). Meanwhile his ratings for “out of touch with ordinary people” increased from 41% to 55%. It indicates a significant loss of confidence in his performance as Prime Minister.
Last week’s Essential Report showed only weak support for Kevin Rudd to lead the Labor Party to the next election – 40% thought the Labor Party would have better chance of winning the election if they changed leader and 37% thought Kevin Rudd was the best person to lead the labor Party. Even among Labor voters only 66% supported Kevin Rudd as leader and 23% though they should change. Comments »
EMC, Essential Report, peter lewis, Polling, The Punch
First published in The Punch 21st June 2010
This is the week when another round of disastrous opinion polls was meant to spark a mass uprising within the Labor Caucus, as members convinced they were facing one-term oblivion hitched their wagon to Julia Gillard.
Everything was in place, a bunch of unsourced comment pieces predicting a move on the PM, an early Newspoll published in Monday’s Australian. And then? Well apart from a slight narrowing in preferred PM, no real movement in the polls.
Entering the spirit of leadership speculation, Essential Research asked our own series of leadership questions. What emerges is a completely different story – the failure of Tony Abbott to convince voters he is the man to lead the Coalition to the election. Now before you all start flaming me (again) for being a Left-ist agent of the evil ALP, let’s have a look at the questions we asked. Comments »
2PP, Election, Polling, Voting intention
Q. If there was a Federal election held today, to which party would you probably give your first preference?
Q. If you ‘don’t know’ on the above question, which party are you currently leaning to?
1,842 sample size
| First preference/leaning to | 6 months ago | 4 weeks ago | 2 weeks ago | Last week | This week
|
| Liberal | 34% | 39% | 38% | 38% | 37% |
| National | 3% | 2% | 3% | 3% | 3% |
| Total Lib/Nat | 37% | 41% | 40% | 41% | 40% |
| Labor | 47% | 40% | 37% | 35% | 38% |
| Greens | 8% | 10% | 12% | 14% | 11% |
| Family First | 2% | 2% | 3% | 2% | 3% |
| Other/Independent | 6% | 7% | 7% | 8% | 8% |
| 2PP | 6 months ago | 4 weeks ago | 2 weeks ago | Last week | This week
|
| Total Lib/Nat | 43% | 48% | 48% | 49% | 48% |
| Labor | 57% | 52% | 52% | 51% | 52% |
NB. The data in the above tables comprise 2-week averages derived the first preference/leaning to voting questions. Respondents who select ‘don’t know’ are not included in the results.
* Sample is the aggregation of two weeks’ polling data. Comments »