BK Drinkwater has posted a good response to some of the comments on Bryce Edwards’ synopses of chapters from the book Informing Voters? Politics, Media and the New Zealand Election 2008 (edited by Chris Rudd, Janine Hayward and Geoff Craig of the University of Otago Politics department). In comments to BK’s post, Eric Crampton recommended Groseclose & Milyo‘s paper on the topic. Having not read the book, I’ll constrain my comments to the posts, comments and paper which I have read.
[Apologies, this is a long and dry post on a topic very dear to my heart. I also banged it up in a spare couple of hours while I ought to have been sleeping, and haven’t proofed it, so it may be incoherent. I reserve the right to subedit it without notice. The rest is over the break.]
The word ‘bias’ connotes directionality – you are biased toward a point of view and therefore away fromthe opposing point of view. This indicates purposiveness, and certainly the meaning with which people use ‘media bias’ is to talk about how ‘the media’ are ‘biased’ against or toward something as a political position. Purposiveness further entails consistency. So let me begin with my own position: I don’t know, but I believe there is no significant, consistent, purposive ideological bias in the NZ news and current affairs media. Some media outlets and journalists are more partial to certain ideological viewpoints or certain types of event than others, and I’m not arguing that there is no bias whatsoever, but I am arguing that the biases which inevitably exist in a media ecology are not significant, consistent, purposive and ideological.
That said, I have practically no hard evidence to support this belief. There exists very little hard research-based evidence to support or undermine this belief. With few exceptions, when people talk about media bias in NZ they are talking anecdotally. Consequently, I have little time for these sort of claims; and less yet because there always seems to be an ideological or political motive behind the complaint (“they’re going hard on $party!”). While I am sometimes guilty of using this to argue a false mean (“Well, since the KBR say the media are Communists and the Standardistas say they’re Tories, they must be doing a good job”), mostly I’m just suspicious that there’s a lack of critical distance on the part of the complainers. There frequently is. Critically assessing the media is hard work, and it’s frighteningly easy to get caught up in the ideological and political lines which the media run; a long-term study documented in Nick Davies in his book Flat Earth News found that more than half the content of UK daily newspapers is wholly or partially derived from PR agencies or communications staff, aka ‘spin doctors’; this content is designed to suck you in and make you believe it. You’re not immune, and I’m not either.
This also gives rise to my second objection, which is the No True Scotsman fallacy employed by Chris Trotter in his comment to Bryce’s post on Rudd and Hayward’s chapter about newspapers, and a very common refrain about the ‘sphere too. He said “How anyone with even a passing interest in NZ politics could fail to notice The Herald’s consistent and outrageous political bias in favour of the National Party…”, in other words: No reasonable person could fail to see the bias. If you don’t see the bias, you’re clearly (a) one of them; (b) not looking hard enough. This is a pointless statement because it consists solely of and invites ‘does not; does too’ arguments and is therefore untestable. It’s anecdotal. Moreover it argues that anyone who disagrees with the definer is unreasonable – not much help, really.
So, absent proper research my belief, too, is anecdotal, but I’m not trying to argue a positive position without evidence; I’m arguing a null position, which is much more justifiable. And I do have the defence that I’m up to my eyeballs in the media on a daily basis, watching, listening or reading it for a job for between four and 12 hours per day, every day. My work is to conduct media research of just the type which is required to assess bias (although little of my research is assessing bias, per se). So I back my anecdotal belief pretty strongly.
The reason for the absence of hard evidence is twofold: firstly, defining media bias is devilishly hard; and secondly, it’s extremely difficult, time-consuming and methodologically fraught to measure, as you’ll see if you read the Groseclose and Milyo paper. G&M do an arguably good job of the second (developing a methodology and measuring the bias) but I find their definition of media bias (roughly, ‘significantly more liberal or conservative than Joe Lieberman’) is rather useless because of all the reasons politicians are dissimilar from the population, which makes them a poor yardstick. I don’t want to launch into a dedicated methodological critique of their work, but let’s just say I’m referring here to research which indicates that politicians tend to have certain personality traits; and those personality traits can and frequently do indicate ideological differences; therefore they don’t adequately reflect the population. And the sheer number of layers of abstraction their methodology employs puts their research on what I deem to be shaky ground. Also, they’re economists and they’ll do what they need to get numbers to work with, while (working from a discourse/content analysis perspective) I’ll do what I can to work with utterances and their meanings. But nevertheless, this all just goes to show: media research is hard to do, and harder to do meaningfully.
However, to argue that the media are ‘biased’ is different from arguing the media distort reality, which should be obviously true to anyone who’s witnessed something first-hand and then seen it in the media at a later point. They are a form of mediation, and distortions which seem purposive aren’t necessarily so. Journalism theorist Andrew Cline runs over a number of these distorting factors in his list of structual biases (as well as some actual purposive biases as well).
Many of the most common complaints about bias are simply objections to content choices – someone who cares especially about a particular thing (say: the environment) wants to see that sort of angle given prominent play in any article in which it might reasonably appear, and will cry anti-environmental bias whenever it’s given short shrift, even though the point of the article might be something else entirely. Others are simply objections to the ideological backdrop of NZ, such as the frequently-begged question that the government has a right to demand taxes from the populace, or that reasonably free markets are a necessity for a modern economy. These are legitimate complaints, but it’s tough to argue that they’re really bias. People tend to view distortions in ways which make sense to them due to the sorts of cognitive biases which BK discusses and for other reasons. One of my favourites is Michael Shermer’s hope springs eternal (formally, a sort of post-purchase rationalisation), which I think neatly explains the media bias rationale currently being run by certain Labour supporters, viz “The voters were manipulated by the media! Our party is good and competent after all, and we would have won if it weren’t for you dastardly journalists!”. BK doesn’t think this is the case, and that’s very charitable of him.
Davies’ book is all about industry bias. There are others, as well, which go into great detail on how and why the media choose what to cover and how. It’s a strongly disputed field with many views and very little actual data, not helped by the fact that a study in one country doesn’t apply elsewhere (yes, I know I’ve cited Davies’ UK study as if it’s true in NZ) and a study conducted now doesn’t necessarily tell us anything about the media next year. It’s all just a bit too pat to say that the media are biased because journalists tend to be ‘liberals’ (whatever that means), or because Rupert Murdoch is a capitalist. The strongest theoretical justification for my null position (no bias) is the market: if the media all swung a certain way it’d open up an opportunity which would quickly be filled. However, as G&M (and plenty of others) observe this is a nice theory but is only a theory until people do the research to back it with actual real-world observations (and their findings suggest it doesn’t hold). Davies, on the other hand, presents a strong case for the importance of structural biases, and against ideological biases.
As you may have picked up from my constantly repeating it, what is needed is actual qualitative and quantitative research into media bias. Until there is some reasonable amount of useful information and discussion from an evidence-based perspective in NZ, all we have is what we’ve got now: people saying “I reckon” and talking as if ‘data’ is the plural of ‘anecdote’, and alienating each other when they disagree with one another’s articles of media faith. Not very useful. We have some such recent work, especially on the 2008 election campaign – the chapters in Rudd, Hayward and Craig’s book, for instance, and Babak Bahador’s research (which I can’t find on the original site any more, but it’s helpfully still up on Kiwiblog.
However, while interesting in their own right and fit to their purposes, these pieces of research don’t strike me as being rigorous, consistent or methodologically transparent enough to provide any real answers beyond the indicative. To really answer these questions, such research needs to be conducted over the long-term and on a large scale. A methodological approach or approaches need to be developed; G&M’s is useless in NZ with its less-codified political structure and small sample size (both political and media), not to mention the absence of an ADA which helpfully rates legislators. Discourse or content analysis is the way I would do this, but that opens up a whole different kettle of worms about the notional audience, and what baselines are taken as ‘unbiased’ (ultimately the problem in any such system, hence Joe Lieberman). But the real difficulties aren’t methodological: we have the talent in NZ to do this. The difficulties are in getting the time and money to do the work; thousands of hours of it, and hundreds of thousands of dollars’ worth. Media research is expensive, and that really is the nub of why nobody has yet definitively answered the question of whether the media is biased: it’s not worth it.