Slant and wind


When Andrew Leigh and I set out to replicate the US studies of media slant for Australia, we knew that — apart from our own curiosity about the answer — it would be interesting to observe the reaction of others. Our study found — using several different measures — that the Australian media is decidedly unslanted. Only a couple of outlets were significantly different from the mean but, as a whole, despite what anyone might have thought, the Australian media is pretty darn neutral. If you are a consumer of more extreme right or left wing packaged news, this is a market where the mainstream isn’t serving you. (This is in contrast to the US where there is significant slanting — mostly to the left — going on; we speculate that this might be a result of a lack of competition in Australian media).

[DDET Read more]

I’ll come to the mainstream reaction in a moment. But let’s look at the blogosphere reaction. Both Andrew Norton and Sinclair Davidson noted that our public intellectuals were not necessarily biased in the way you might have thought. I was surprised too to find that Glyn Davis was mentioned (positively) more in parliament by the Coalition than the Government. But that is not a classification but a fact. In any case, there were 21 public intellectuals who were significantly mentioned more by one side of politics than the other and it is they that were more ‘important’ in constructing our slant measure than those, like Glyn, who were more evenly mentioned. (By the way, yours truly has been mentioned 3 times in Parliament the last time I looked — 2 Coalition and 1 Democrat. Work that one out.). So it was Les Murray versus Helen Hughes that was important here.

Now Andrew Norton builds on this to argue that our public intellectuals are not partisan in their public work (i.e., they are more discipline based). Funny he should say that when he was expecting to find certain public intellectuals cited more favourably by one side of politics than the other. But that observation isn’t a problem with the study, it is a feature. The point is to pick people who stand out based on public reputation per se and then use revealed behaviour to classify and match. Again, the issue is that there are some public intellectuals identified as significantly partisan and we want to see how media outlets pay attention to them. In any case, we used multiple measures and they came up with the same basic finding.

Sinclair Davidson notes that our various tests point the same way but finds it unconvincing anyway. He writes:

To the extent that the robustness tests confirm a non-convincing result I suspect there may be a deeper problem with the study.

Yes, that sounds as biased as it is. But Sinclair is introspective:

Alternativily, of course, it may be that I am slanted and don’t appreciate the fine job the ABC have done in promoting the Coalition all these years.

And that is the point and it is one that Peter Costello is going to have to think about before his anti-ABC rants (that said, we didn’t measure intra-Coalitional factional bias!). What we are suggesting is that no one appreciated that and someone will have to come up with an alternative test to counter it. We would welcome any such exercise.

But Sinclair also accuses us of ‘survivorship bias.’ This arises when academics study firm profitability and its drivers but fail to taken into account that firms with bad strategies may not survive to be part of our investigation. In this case, Sinclair says we ignore those public intellectuals who didn’t survive (either physically or in the public) until 2005. True, we only look at ‘survivors’ but we have to also think about what bias this might give rise to. Were Labor public intellectuals so distraught with 10 years of Coalition that they carked it? Or was it the other way? Were they energised by having a Government to critique? There is no theory of bias related to survivorship that I can think of. So that is a weak attack.

Finally, on the media reaction. Let’s look at the headlines. For the Fairfax media:

And for News Ltd

Draw your own conclusions.

[Update: This post appeared in Crikey, 3rd September 2009.]


4 Responses to "Slant and wind"
  1. Joshua – I said that most public intellectuals are not party partisan. I think Sinc is right that your methodology isn’t very good at picking up the political stances of public intellectuals. On a quick look at Phillip Adams mentions in Hansard, for example, your positive or neutral test really just shows that Coalition MPs can mention Adams’ long history in the arts and media or some story he has reported without a digression on his left-of-centre political views, not that he is pro-Coalition. Marie Bashir seems to be mentioned because she is Governor of NSW, not because MPs have any particular opinions on her ideology or party preference.

    If you want to use public intellectuals surely there are far better tests of their political position: for example their own stated views, or using the coding methodology, where the coders have been expressly asked to put their mind to the issue of slant/bias, rather than using mentions by MPs who may have been talking about something else at the time.

    I think public intellectual mentions could pick up ideological or issue bias by newspapers, but party slant is more difficult.

    Most of your specific findings aren’t highly counter-intuitive, but on the other hand the methodology isn’t really strong enough to replace readers or listeners own impresssions if formed over an extended period of time, not just one election campaign.

  2. In your study you write, ‘We do not regard differences in the volume of coverage in itself as being a form of media slant.’
    But to me, this is one of the most important forms of media slant (or ‘bias’ – which I understand is a word you deliberately avoided). It is true that incumbents typically get more coverage than opposition – but public policy debates are more mixed. I bet you would find that the ABC gave a disproportionate amount of time to opponents of league tables for instance, as well as giving a disproportionate amount of time to opponents of the PC’s report on the parallel importation of books – two issues where the left/right divide is clear but the Coalition/ALP divide is less clear.

%d bloggers like this: