16 September 2009

ABC TV votes Conservative. Who knew?

Jonathan Holmes, of the ABC's Media Watch programme, was, as usual, fulfilling his institutionally self-assigned duty of holding the Opposition to account the other evening. He was gratuitously, but unsurprisingly, openly sticking the boot into a carefully chosen target, the former Coalition Government's Treasurer, because he had had the temerity to be critical of a public broadcaster for using its taxpayer funded media resources to favour a political party. In the process Holmes drew our attention to a piece of research recently published by some Australian National University academics.

Holmes crowed jubilantly, at the Coalition's expense, from his taxpayer funded ABC TV soapbox, and without a hint of irony, that, contrary to Mr Costello's suggestion, research showed that ABC TV did show a political bias in the research period: towards the Coalition. So stick that in your pipe and smoke it you Coalition whingers.

Just why would ABC TV choose to broadcast this? As Holmes himself seems to concede, perhaps it's because it is news. "News" that is, in the sense that such a claim is indeed completely "new". No-one has ever seriously suggested this before. And who is there out there, without an axe to grind like Holmes, that actually believes this claim?

Sure many of us continue to watch ABC TV's current affairs programmes notwithstanding its seemingly open pro-Labor slant, but we don't pretend for a moment that it is not left-wing biased. We make allowances for this bias because we quite are fond of Aunty from habit. It occasionally can be quite exasperating, but we can turn it off or switch channels. To a limited extent some of this bias is forgivable, if only because it is an almost inevitable function of the fact that those with the job security of a taxpayer funded job at the ABC, will naturally tilt their political allegiances (even if unconsciously) towards entities that they perceive are more likely to protect their own state funded economic security. As a consequence there is only one known employee of the ABC who is voluntarily out of the closet as a Liberal supporter, and everyone in the ABC seems to know who he is (some radio jock in Perth), because he is so unusual there.

But for ABC TV to openly disseminate that it has a pro Coalition bias, even whilst it is simultaneously bashing the Coalition, does seem a bit much, even for such brazen leftist provocateurs as Media Watch hosts. What new levels of political distortion does this suggest ABC TV might now be prepared to go to ? For such a manifestly ludicrous assertion to be given currency at the taxpayers' expense, both by a university and on a public broadcaster, undermines one's faith in such institutions. Could this be an elaborate hoax? It is doubly bizarre that it was revealed whilst the broadcaster was in the very act of doing what we all see it doing every day, sticking it to the Coalition in Opposition with its customary glee. It's also very amusing (if a bit disturbing) that the ABC can be quite so self unaware.

Are we now, as a consequence, going to see even more ABC TV political hacks exploit the well trodden ABC path to political power in the ALP, following the likes of ABC TV's former stars, Maxine McHugh and Allan Carpenter? This is only to be done of course in order to offset the blatant ABC TV pro Coalition bias. Are we going to see Tony Jones pitch even more gentle softball questions to Government Ministers like Penny Wong about the economic effects of her carbon trading scheme? Is he now going to treat Tony Abbott with even greater disrespect and with even more hostility than at present? Will Kerry O'Brien actually feel he has to get even nastier on air towards Malcolm Turnbull than he is already? Are they all now going to give Commissar Julia, the Minister for Everything, even more servile deference than they have displayed to date? Will Kerry perhaps now feel even more emboldened to admit electoral swings towards the ABC on election night? All this of course to provide the holy statutory "balance" necessary to offset the brazen pro-Coalition bias that ABC TV apparently carries.

Let's have a quick look at the methodology in play here. Not of ABC TV, we do that nearly every night. I mean of the academic "research". Here's the website of Andrew Leigh, the guy who, with Joshua Gans, recently published the paper: "How Partisan is the Press? Multiple Measures of Media Slant".

How did Messrs Leigh and Gans arrive at their conclusion? Here's their criteria:

"A good measure of media slant ought to reflect the ideological affinity between a particular outlet and one side of politics. In effect, such a measure plots media outlets onto a political spectrum, allowing us to answer questions like: "if this newspaper were a politician, how would it vote ? " (page 2)

That seems pretty clear. But how do they do this?

"To estimate the political position of each media outlet, we simply estimate a weighted OLS regression, in which the dependent variable P is the share of Coalition mentions * by a given public intellectual i in media outlet j in time period t, and the independent variable is a vector of indicator variables of each media outlet:

Pijt = ZIjt
... " (page 9)

Mmm. I'm sure glad their estimate of the weighted OLS regression is simple.

And what conclusion did they arrive at using this sophisticated academic model:

"Only one of the 27 outlets we study (the ABC Channel 2 television station**) is significantly distinguishable from the centre position. These results are robust to various specifications. We also find that there has been no systemic evolution in slant over time. ... " (page 4)

I think this means that their finding of a pro-Coalition slant at ABC TV is not a just temporary state, but has been pretty much a permanent condition.

This sure brings to mind that all too frequently recalled saying about statistics and damned statistics.

You also have to ask yourself how academics can continue to expect to be ceded credibility by the public when they publicise such findings outside academia. The only thing it really seems to prove is that an academic with a grant can demonstrate almost anything. Either that, or that the finding itself demonstrates conclusively that they stuffed up their methodology. What next an ANU computer model showing conclusively that membership of a trade union correlates with a vote for the National Party***?

The ABC would be as well to be little more coy about promoting this finding. It could generate even more derision at the ABC's expense than it receives already and in a lot more places than just the Liberal Party HQ.

* Mentions ? Could this be the explanation for their counter-intuitive finding. They count ABC TV mentions of the Coalition as being pro Coalition.? ABC TV has no compuction about bagging the Coalition out all the time. So Mr Holmes' reference to Mr Costello on Media Watch would thereby count towards ABC TV's pro Coalition slant. Are these academics this stupid or just partisanly dishonest leftists themselves?
** You'd ordinarily reckon academics doing research into the media in Australia would at least know that this is not how the ABC describes this component of its operations, but then you look at what these blokes say they found, and you know they live in a different universe to the rest of us.
*** By counting membership of a commodities co-operative as if it were a trade union.

No comments: