Fact check: Are ABC employees or journalists five times more likely to vote for the Greens than the general population? – Fact Check


Updated

July 13, 2018 10:31:44

The claim

Calls to privatise the ABC have become louder in recent months, with Liberal Party members voting to privatise the national broadcaster at the party’s annual federal council, though the Coalition Government’s policy remains to keep the ABC in public hands.

Sinclair Davidson, an RMIT academic and adjunct fellow at think tank the Institute of Public Affairs, has co-authored a book with fellow IPA colleague Chris Berg, entitled Against Public Broadcasting: Why and how we should privatise the ABC.

In support of the book’s launch, Professor Davidson told Sky News’ Outsiders program that countries with public broadcasters tend to propagandise to their citizens, and “it just so happens that our public broadcaster propagandises from a particular viewpoint of the world”.

“The ABC employees are five times more likely to be Greens voters than the general population,” he said.

As well as ABC employees, Professor Davidson mentioned journalists:

“It kind of suggests that they are incredibly out of touch with ordinary Australians. The whole idea that we have is that well, journos tend to be kind of centre-left anyway so what do you expect? We’re not talking centre-left here, we’re actually talking very hard left wing views here.”

Are ABC employees or journalists five times more likely to vote Greens than other Australians? RMIT ABC Fact Check investigates.

The verdict

Professor Davidson’s claim is flimsy.

In making the claim, Professor Davidson referred to a study published in 2013 which surveyed 605 journalists from a variety of organisations on their voting intentions.

Fifty-nine of these journalists were from the ABC, and only 34 of them answered the question on voting intention, with 25 either undecided or electing not to answer.

Of the 34 who did answer, 41.2 per cent, or 14, said they would vote for the Greens.

But experts told Fact Check that the ABC sub-sample was too small and the rate of undecided and non-response too high to be able to draw accurate conclusions from the survey on ABC journalist voting intention, let alone voting intention of all ABC employees.

Upon releasing the findings in 2013, the author of the study himself, Folker Hanusch, inserted numerous caveats about using sub-samples of the survey, including that the margins of error would be larger than those for the total sample. Professor Davidson neglected to include any of these important caveats in making his claim.

Experts contacted by Fact Check, including Professor Hanusch, also took issue with comparing the results of the survey with larger, more stable studies of the voting intention of the general population, such as Newspoll.

Whilst the survey found that Australian journalists in general tend to skew left, it showed no evidence that ABC journalists were five times more likely to vote for the Greens than the general public, and experts contacted by Fact Check said they did not know of any other recent studies which canvass the voting intentions of ABC journalists.

University of New South Wales statistician Jake Olivier compared the survey results with those of Newspoll over the same survey period, and found ABC journalists were 2.4 times more likely to vote for the Greens.

But he cautioned that the sample size of ABC journalists was too small to make strong conclusions about this result.

Fact Check could find no research on the voting intentions of all ABC employees including those outside of the news division.

Context for the claim

Though the Liberal Party’s federal council voted to privatise the ABC, the decision is not binding on the parliamentary party, and the Government’s position remains to keep the national broadcaster in public hands.

“The ABC will never be sold. That is my commitment. It is a public broadcaster,” Prime Minister Malcolm Turnbull said on June 18.

Party elder and former prime minister John Howard agrees with this policy, telling the Centre for Independent Studies on July 10 that “I think the ABC needs greater balance, but I don’t think the solution is to abolish or to privatise the ABC”.

Nonetheless, the vote shows there is an appetite to privatise the national broadcaster among some sections of the community.

In his interview with Sky News, Professor Davidson said that the ABC should have its government funding removed and be handed over to its employees (both journalists and non-journalists) to “let them realise the value” of the organisation.

A day before, Professor Davidson and Mr Berg co-authored an article for The Spectator, entitled Not our ABC.

In the article, the authors outline their case for gifting the ABC to its employees for free, and taxing them when they sell their shares in the organisation.

Professor Davidson and Mr Berg criticise the ABC for “political bias in its reporting and news coverage”, stating:

“A 2013 survey revealed that ABC journalists are almost five times more likely to be Greens voters than the average voter and twice more likely to vote Greens than the average journalist.”

As Professor Davidson mentioned “journos” in his interview and made this claim in reference to journalists a day earlier, as well as advocating handing over the ABC to all employees, and not just those with some sort of editorial responsibility, Fact Check will assess his claim on the basis of both employees and journalists.

Studying journalism

Fact Check asked Professor Davidson for the source of his claim.

Professor Davidson supplied a study published in 2013 by Folker Hanusch, who was then a senior lecturer in journalism at the University of the Sunshine Coast.

Entitled Journalists in times of change: Evidence from a new survey of Australia’s journalistic workforce, the study surveyed a sample of 605 journalists between May 2012 and March 2013, from a variety of news organisations across print, broadcast and online mediums on a variety of topics, including gender, salary, religious affiliation, and of course, political beliefs.

Of the total sample 9.8 per cent were from the ABC.

Fact Check contacted Professor Hanusch, now at the University of Vienna, Austria and the Queensland University of Technology, to clarify certain points of the study. Professor Hanusch confirmed that the number of ABC journalists surveyed was 59, and that only 34 of that number answered the question about voting intention.

The remaining 25 ABC journalists surveyed either did not yet know who they would vote for, or declined to answer.

The study said: “Among the surveyed ABC journalists who declared a voting intention, 41.2 per cent would vote for the Greens, 32.4 per cent for Labor, and 14.7 per cent for the Coalition.”

Fact Check was unable to find any other recent study which canvassed the voting intentions of ABC journalists; polling and media experts spoken to in researching this fact check said they were unaware of any other recent studies.

The same experts were unaware of any studies which canvassed the voting intentions of ABC employees including those outside the news division.

How much of the general population votes for the Greens?

In a post on his blog on June 18, Professor Davidson indicated that the 2013 election results were used as a basis for the general population in making his claim. In that election the Greens attracted 8.65 per cent of the vote.

However, the election was held on September 7, whilst the survey of journalists he compares it to began gathering data in May 2012 and finished in March 2013 — five months before.

Fact Check asked psephologist Kevin Bonham for help in determining how much of the “general population” votes for the Greens in the context of the media industry survey.

Dr Bonham suggested using an average of the Greens primary vote taken from the Newspolls conducted in the period in which Professor Hanusch’s survey took place.

“Newspoll is a very suitable benchmark for measuring the Green vote for your purpose given that it was accurate to the nearest whole number in its final federal poll in both 2013 (9 per cent versus actual 8.65) and 2016 (10 per cent versus actual 10.23) and I think averaging all Newspoll data over the survey period is an adequate method in the absence of any more refined detail about when exactly surveys were conducted,” he told Fact Check via email.

A flawed comparison?

Jake Olivier, an associate professor in the School of Mathematics and Statistics at the University of New South Wales cautioned that certain differences between the two surveys and certain features of the Hanusch survey may make comparing the two problematic.

“I think the quality of the data makes it difficult to draw strong conclusions here. This is compounded now by comparing results from very different surveys (Hanusch’s versus Newspoll). It is possible to tease out those who are uncommitted or undecided from both. These ‘unknowns’ are a much larger proportion among journalists than the general public irrespective of media outlet.”

Dr Bonham agreed:

“… an issue with using the study to define the voting beliefs of ABC journalists is that not all those sampled answered the voting intention question. Journalists with certain political beliefs may have been more likely to answer this question than others. That doesn’t interfere with the overall findings re: journalists tending to skew to the left but it does make it difficult to sustain claims about specific ratios to the general population.”

Sample size matters

James Baglin, a senior lecturer of statistics in the School of Science at RMIT University told Fact Check that the Hanusch study had gone to a “decent effort” to get a representative sample of Australian journalists, but that the aim was not a representative sample of ABC journalists.

“Their aim was to get a representative sample of Australian journalists. So, it’s possible that the ABC sample was too small to estimate a statistic with a reasonable level of confidence,” Dr Baglin said.

“Given only 34 ABC journalists from the ABC responded, the statistical estimates derived from this survey are expected to be associated with a high degree of uncertainty.”

According to the ABC’s 2012-13 annual report, the national broadcaster that year employed 5,446 people across all states and territories, equivalent to 4,664 full-time employees.

The report said that the news division employed 21.64 per cent of ABC employees at that time, which Fact Check has calculated is roughly 1,178 people, though not all of these people would meet the definition of journalist from Professor Hanusch’s study:

“Only professional journalists were surveyed, ie. those who earn at least 50 per cent of their income from paid work for news media and who are involved in producing and editing journalistic content, editorial supervision or coordination. Hence, news photographers were included by definition, but camera operators, for example, were only included when they independently made editorial decisions.”

Associate Professor Olivier said that given the magnitude of the number of employees and journalists who worked for the ABC at that time, the sample size of 59, with only 34 responding is “not very reassuring in terms of data quality”.

He told Fact Check that there was no “magic response rate”, but the sample in the study was “clearly too small to accurately estimate the proportion among all ABC journalists”.

“Another issue is it is well-known that people who feel strongly about a survey are more likely to respond. With regards to this survey, the results could mean that Greens members are more passionate about their party affiliation than others and thus more likely to respond,” he said.

What does the study’s author say?

In a piece for The Conversation when the study was about to be published, Professor Hanusch said:

“The margin of error for the entire study sample is 4 per cent. Sub-samples of journalists’ responses to some questions — such as voting intentions — are likely to have a higher margin error, however, appropriate statistical methods were used in testing for differences between sub-samples to take account of the smaller sample sizes.”

The study itself echoes this point:

If one were to compare only those journalists from the big three news organisations who would vote for one of the three major parties, we can see a statistically significant result χ2(4)=10.309, p<.05. However, Cramer’s V=.164 indicates a very small effect size, and it must be noted that these sub-samples are smaller in size with a large sample error.

When contacted by Fact Check, Professor Hanusch took issue with the use of the numbers in his survey without important caveats.

In a statement emailed to Fact Check, Professor Hanusch expressed dismay with the media’s portrayal of his study’s findings, and cautioned against comparing the results of his study to studies of the voting intention of the general Australian public, which he said were larger and more robust.

Fact Check has reproduced the statement in full below:

I would like to stress that the partial results of my study of Australian journalists, conducted in 2012-13, need to be interpreted in context.

As I pointed out in my publications on The Conversation, in various interviews and in the academic article I published, the results in relation to journalists’ voting intentions are based on smaller sub-samples than the overall representative sample of the study.

This means, there is a much higher margin of error associated with those results.

While there were some statistically significant differences between the groups of journalists who provided their voting intentions along organizational lines (i.e. between ABC, News Corp and Fairfax Media), I would caution extrapolating from these results to all journalists at these organizations and to generalize, given the small sample sizes involved, as well as the sizeable number of journalists who did not provide their voting intention, or who did not yet know who they might vote for.

In particular, I would caution against comparisons to the results of this study with much larger and far more robust studies of the voting intentions of the general Australian public, in the way that appears to have been done here.

Comparing Aunty with the general population

Despite the problems raised by experts, Fact Check asked Associate Professor Olivier for help in calculating the average Newspoll result and comparing it with Professor Hanusch’s survey.

He said that because of the high rate of undecided and non-response in the journalism survey, these results should be included in any comparison.

The number of journalists who indicated they would vote Greens out of the total sample of 59 was 14, or 23.73 per cent.

The percentage of the population who indicated they would vote Greens during Professor Hanusch’s survey period was 9.82 per cent, or 2,262 of the total 23,030 people surveyed in that period.

“If I compare the ABC and Newspoll results for Greens preference in a simple analysis, ABC journalists are 2.4 times as likely to prefer Greens than the general public,” Associate Professor Olivier said.

Associate Professor Olivier gave his calculation a 95 per cent confidence interval of between 1.5 and 3.8.

In their Statistics Glossary, Valerie J. Easton and John H. McColl define a confidence interval as: “… an estimated range of values which is likely to include an unknown population parameter, the estimated range being calculated from a given set of sample data.”

This means that it can be said with 95 per cent confidence that the real figure falls in between these two values.

“You’ll notice that [the number five] is not in the confidence interval,” Associate Professor Olivier said.

Associate Professor Olivier also calculated an odds ratio, using a logistic regression model, which includes all outlets (ABC, Fairfax and News Corp) as well as Newspoll.

Using this model, he calculated that ABC journalists have a 2.9-fold increase in the odds of preferring the Greens over Newspoll respondents, with a 95 per cent confidence interval between 1.6 and 5.2.

But he noted that Professor Davidson’s statement aligns better with relative risk than an odds ratio, noting that he stated that ABC employees are “five times more likely” rather than that ABC employees have a “five-fold increase in the odds of”.

“So, the estimated 2.4-fold increase seems the more relevant here,” he said.

Once again, Associate Professor Olivier cautioned against making strong conclusions with these calculations:

“There are important limitations to this and any other analysis. As mentioned, the reliability of the sampling frame is unclear. There are perhaps certain types of journalists or media outlets that are harder to identify than others. The analysis is also limited by the number of participants who did not respond to this question. If non-response is related to the question itself, it could invalidate any analysis.

“With those limitations in mind, I think the results should be tempered to reflect the uncertainty in the data and the analysis. For example, it would be reasonable to state ‘Among Australian journalists who participated in the survey, there appears to be…’ or something similar. This in some way acknowledges those who did not respond may have answered differently.”

Principal researcher: Matt Martino

Additional research by Lauren O’Keefe

factcheck@rmit.edu.au

Editor’s note (13/07/2018): A previous version of this fact check contained a line which incorrectly said that the University of the Sunshine Coast had been rebranded as the Queensland University of Technology. The two are separate institutions which both still operate today.

Sources

Topics:

abc,

television-broadcasting,

work,

greens,

political-parties,

australia

First posted

July 13, 2018 07:18:36





Source link

Leave a Reply

Your email address will not be published. Required fields are marked *