Today, the Women in the Media project tackles the sources journalists turn to for quotes and expert opinion. Whose voices get heard? The results show men feature overwhelmingly.

Analysis was compiled by Wendy Bacon, Elise Dalley, Julie Posetti, Lauren Frost, Joanne Griffiths and Rochelle Widdowsen. Data analysis and visualisation by Elise Dalley.

So far in our Women in the Media investigation, we have found that:

Today our project shifts to what voices were heard in Australian media on 4 March, 2013. Are women fairly represented by the nine mastheads chosen for this study? The answer on this day is a resounding "no". To appreciate the findings, remind yourself that women are slightly more than 50 per cent of the Australian population, and that the purpose of the media in a democratic society is meant to be to represent or give voice to all communities.

Key findings

  • Male sources overwhelmingly dominate the coverage.
  • The biggest round of sport is overwhelming male.
  • Even when sports round is removed, male sources were three times more likely to be quoted than female ones.
  • Business people quoted in stories were more than 90 per cent male. The voices of men dominated in politics, business and international news. Female sources were slightly more numerous than male sources in the education round. Female journalists were more likely to use female sources than male journalists.

A crucial aspect of reporting is journalists' selection of their sources. Through this process, reporters and editors exercise power over who will get a voice and who will remain silent. Many articles have just one source, often effectively giving that source a "free kick" at promoting their views. Where there is more than one source, the first source tends to define the terms of discussion within the article. Sources further down the story are more likely to be in response to the first source.

There were 760 stories across nine publications. The mastheads were The Australian, five metropolitan newspapers, The Age, Courier Mail, Daily Telegraph, Herald Sun, Sydney Morning Herald and three regional newspapers, Cairns Post, Geelong Advertiser and the Illawarra Mercury.

Of the stories we investigated, 235 were sports stories. As reported on Tuesday, these are overwhelmingly about male sports and sports people, by male reporters, quoting men, with images of men. With few exceptions, the sports pages are by men and about men. For the rest of the remainder of this analysis we have excluded sports stories. If sports stories were included, the proportion of male voices would have been even greater.

We coded all stories according to the field of reporting or "round". Where it could be identified, we coded sources male or female. We also coded sources according to type of source — for example, business sources, professional sources, NGO sources and other sources from the community.

To understand the role of the sources more clearly, we further divided the sources between first, second and third sources quoted in stories. (There were very few stories with more sources and we did not code further sources.)

On 4 March, 2013, there were 646 sources that could clearly be identified as male or female. Of these, 486 (75 per cent) were men, and 160 (25 per cent) were women.

There were 371 stories which had one source. Of these, 281 (76 per cent) were male and 90 (24 per cent) were female sources. Second sources were even more likely to be male, while third sources were still twice as likely to be male than female.

There are two different ways of looking at sources. One way is to look at the type or role of the person quoted. For example, is the source a business person or a professional person? The other way is to examine whether sources of all types in different rounds of reporting are more likely to be male or female.

Types of sources

When journalists selected a first source with a business background to tell a story, they chose a male in 92 per cent of cases. When they chose a source from police or corrective services, women didn't even get a look with; 100 per cent of first sources with police or corrective services background were men.

Professionals (eg. doctors, lawyers and teachers) were far more likely to be men and even the category including "all other workers" was strongly male dominated.

Even in community sources, which included "people in the street" and other sources not identified with a particular role, men were more strongly represented. Only arts and entertainment sources, and those representing non-government or civil society organisations, had slightly more women than men.

(A deeper analysis would compare these figures with ABS statistical data on gender breakdowns of occupation).

Proportion of male and female sources across rounds

For this second type of analysis, we examined the five biggest rounds to identify the gender of sources

The top five rounds were economy (including business and finance), politics, international, crime/law and media. Male voices dominated in all these rounds.

Male voices accounted for 136 of the sources (85 per cent) in the 78 economy stories we analysed. Politics was the second biggest round in which women were only quoted on 34 per cent of sources. The proportion of female sources dropped dramatically for the third biggest round, international. Of the 78 sources referred to in this round, male sources accounted for 92 per cent.

Interestingly, Julia Gillard accounted for 11 of 33 female sources in the politics round (33 per cent). It goes without saying that if we didn't have a female prime minister, women would be far less visible in politics.

Of all rounds,reporters in the education round gave women the strongest voice with 53 per cent of the 19 stories. The gender balance was also close to even in the health and entertainment rounds, with 45 per cent and 39 per cent of sources respectively.

Are female journalists more likely to use female sources?

In short, yes. To answer this question, we divided stories into three categories: those written exclusively by males, those written exclusively by females and those with joint male and female bylines. The results are as follows:

A total of 282 gendered sources featured in the 152 stories with a male-only byline. An overwhelming 233 (83 per cent) of these sources were male.

From the 119 stories with a female-only byline, there was a total of 212 gendered sources. Of these, 147 (69 per cent) of sources were male and 65 (31 per cent) were female, almost double the number of female sources present in stories written exclusively by males.

Twenty-six stories had a shared male and female byline. In these stories there were 48 gendered sources. This category had the highest percentage of female sources (35 per cent). Although it was a much smaller sample size than the other those categories, the results raise the possibility that, when male and female journalists share the reporting role, they are more likely to have a greater gender balance with sources.

Part of the differences between sources selected by male and female journalists might be accounted for by considering the fact that women reporters tend to be concentrated in fields where women are more likely to be quoted. These findings should lead to further research as they have important implications for the visibility of women in public life.

Masthead comparisons

We analysed data in this category to determine the gender gap across masthead for females in terms of voices quoted in stories. We included all stories with at least one gendered source in this count. As can be seen from the graph, the gender gap was greater in the Queensland regional publication Cairns Post and best in the Victorian regional publication the Geelong Advertiser. The gender gap was greater in the big News Ltd publications The Herald Sun, The Daily Telegraph and The Australian than Fairfax Media's Sydney Morning Herald and The Age.

Actors in stories - male and female

Casting a wider net, we also counted the number of individuals or "actors" in a story, whether or not they were quoted.

In our sample of data, there were 1101 male actors and 462 female actors. The 70:30 ratio continues. The finding does suggest that women were slightly more likely to be characters or subjects in stories than be quoted on this day.

Our methodology of selecting a single day for parts three and four of our project is similar to that used by the Global Media Monitoring Project (GMMP) which since 1995 has surveyed women in the media at five year intervals. In her introduction to the international GMMP 2010 report, Margaret Gallagher wrote:

"If through an unequal distribution of narrative resources, the materials from which some people build their account of themselves are not theirs to adapt and control, then this represents a deep denial of voice, a deep form of oppression."

In 1995 the global snapshot was startling in the evenness of its findings of gender gaps across newspapers, radio and television in 71 countries. Womens' visibility was uniformly low. While the picture has not been static, the project reported in 2010 that "voices remain overwhelmingly those of men" across 108 countries.

In 2010, QUT journalism academic Angela Romano reported for the GMMP that on a single day in November 2009, women accounted for 24 per cent of the 1012 news sources who were heard, read about or seen across a sample of media. Neglect of female sources was particularly noticeable in sports news. Women made up only 1 per cent of the 142 sources who were talked about or quoted in sports stories in Australian media.

Our findings suggest that the situation does not appear to have improved over more than three years.

Asked to respond to New Matilda's findings, Romano said that our findings confirm other academic studies of women in the media. She wrote:

"This snapshot of newspaper activity confirms and strengthens the findings by other researchers, who find that women are underrepresented as bylined reporters in the Australian news media. Several researchers have conducted snapshots of Australia's news media, and they offer the same picture. Women reporters are a minority when it comes to bylined stories. There are occasionally exceptions in reporting so-called "women's topics", such as lifestyle stories. The clearest disparity is in sport news, where female reporters are seldom seen.

"The research also shows that women are also minority figures as sources who are quoted or discussed in news media stories. While a snapshot will always provide a limited view of what is happening, it is important to note that regardless of when these snapshots have been taken or which cities have been studied, they always tell the same story. Women are substantially outnumbered in the Australian news media as both reporters and sources of news."

How to change women's lack of visibility and voice is the question.

Analysis was compiled by Wendy Bacon, Elise Dalley, Julie Posetti, Lauren Frost, Joanne Griffiths and Rochelle Widdowsen. Data analysis and visualisation by Elise Dalley.