So, not entirely happy with the new Australian Politifact. I’ve seen reports of the US version being politicised, including this (rather partisan) post on the general matter of fact checking sites in the last US election. (HT: Kiwiblog). I had a little worry about the fact that Politifact Australia was being set up by a guy who used to run the Sydney Morning Herald, which has a noted left wing editorial line.
I was looking at the Australian site yesterday after a day of claim and counter-claim around the budget. I perhaps expected that there would be some coverage of “the $70B black hole” in the Coalition budget plans, amongst other topics. I’d seen some earlier analysis on the site, and reserved judgement – it was neither great nor dreadful.
Looking at the latest few posts, I’m quite concerned. I think it’s really important that a fact checking site be clearly and obviously non-partisan, and it doesn’t look that way at the moment.
Firstly, let’s deal with the format. In order to make the material more digestible the site has a commentary that goes with each analysis. I think this commentary is a good idea if done well, as it gives the reader some context on the political environment that a claim is made within, and the reasons why a particular fact check is important. However, this analysis needs to be very carefully done – this is an area that can easily introduce bias. Further, the actual fact checks need to be sensible – they need to go to the heart of the claim, not tie themselves up in technicalities.
For the items that I see on the site, I’m disappointed. Let’s look at some of them in turn.
Rating on this one is “pants on fire,” basicially seeing this as a blatant lie. My problem here is that gold-plated has two meanings that I think every Australian would know, refer here to the first dictionary entry I get on google:
verb (used with object), gold-plat·ed, gold-plat·ing.
The coffee machines are noted in the report as having cost up to $15,000 each. In the dictionary definition above, I think this arguably fits exactly into definition 2. Having said that, I have some experience with coffee machines, and I suspect $15,000 for a high volume coffee machine isn’t outrageous. But if that is the point of the politifact article, it should perhaps have assessed what the normal going rate for a coffee machine is, not based the entire article around whether the coffee machine literally had gold plating on it.
The “pants on fire” rating is a quite substantial one. It implies that someone is blatantly lying. I’m OK that it should be given in that situation. But why would you give your first rating of this nature based on a technicality – based on your literal interpretation of a statement that can also clearly be taken figuratively? And why would you do that with only a passing mention that the Liberals have said that they meant gold-plated figuratively, and that it had arisen at a time when the Labour party were talking about gold-plated electricity networks? Should we expect to see an assessment that power lines were not literally gold plated, and therefore Gillard lied?
In summary, not a good start.
This one has many redeeming features – it explains gross and net debt, and it explains a little background as to which measure means what. The context that I think would have been useful is that the debt ceiling is a really big deal in the US, and that the Libs are, I believe, attempting to import a little of that media hype by pointing to it. But that’s a small niggle, and I can see why including that would arguably be attributing motives that perhaps aren’t provable.
My bigger concern is that this is rated mostly false largely based on a careful wording of the allegation. So politifact appear to agree that the gross-debt ceiling will almost certainly be breached, but by defining the allegation as ‘the budget papers reveal” they’re able to say that this is mostly false because they can’t see anywhere in the budget papers that says this, notwithstanding that it’s true from other sources.
Again, I feel like this is being marked false on a technicality. The statement that the $300B debt ceiling will be breached is found true, it’s found false that the budget said that. I don’t think I would say that’s mostly false, I’d probably say that is mostly true. And I’d argue that you could derive the assertion from the budget papers, since they provide enough information to calculate or approximate the gross debt.
This one goes in the other direction (see any pattern here). The finding is that the budget does, in fact, have the largest revenue writedown since the great depression. But the context that goes with this is not provided, and the assumptions are generous. The implication from Wayne Swan is that this is the cause of the budget not being in surplus, and that this is outside his control. But if we look carefully at what is going on, I would have said:
a. The previous budget forecast a 12% growth in govt revenues
b. The treasurer is somewhat responsible for that forecast – Treasury give a range, but the Treasurer picks where in that range he’ll use in his budget
c. The revenues actually grew about 7%, so it’s not like revenue is shrinking
d. The deficit is $20B, the writedown $17B. So even leaving aside that the writedown is arguably the treasurers fault (as opposed to the writedown in the great depression being unforecastable), the writedown isn’t really the cause of not getting to a surplus.
My view is that these analyses are not consistent. The benefit of the interpretation is consistently leaning one way, and the analysis is not getting to the heart of the claim. Basing a finding on a technicality I think is misleading, and if you are doing that I think you should be clear about it. In general I’d say if the underlying claim is basically true, but a technicality means the way that claim was framed is inaccurate or could be misinterpreted, that’s a mostly true. At the moment this site feels partisan, as these calls appear to all be going in one direction. Time will tell as more content is added.