Moderation problems at the Bolt blog (updated)
In my previous post, I noted that, while Andrew Bolt had correctly calculated the impact of the carbon tax for the year 2020, he hadn’t completed the analysis by evaluating the impact over the relevant policy timeframe. While I was working on this, Bolt produced another post, linking to this piece by John Humphreys, which suggested errors in my original analysis. I submitted comments to both sites. John noted the error in Bolt’s analysis, but advises me that he is not going to publish comments, and hasn’t yet corrected his own post. I assume he’ll get around to this soon.
I submitted the following to Bolt’s blog
John Humphreys has updated his post to note “John Quiggin has pointed out that there is also a significant problem with the Bolt estimate, since it only calculates the benefit from reduced emissions for one year (2020) instead of adding up the cumulative reductions over multiple years. Good point. This means the Bolt methodology just got a while lot more complicated since it now requires an expected future emissions time series and an expected future emissions time series counter-factual. That task is too big for me at the moment, but [b]it’s fair to say that such a number is going to be quite a bit higher than Bolt’s original estimate[/b].”
(emphasis added) I give a corrected estimate here
Sadly, the comment didn’t make it through moderation, presumably due to an error, so I’m publishing it here.
Update: Another go-round on moderation Andrew Bolt has posted again, indicating that the non-publication of my comment was indeed a moderation error, and acknowledging the need to use cumulative effects rather than those for a single year. As he will see when he does this, his sensitivity estimate is consistent with mine.
Unfortunately, Bolt didn’t follow the link I gave, and therefore repeated the already-refuted claim that my estimated was out by a factor of five, relative to that of Roger Jones. As I’d already pointed out here, the error was due to Michael Bachelard, who applied Roger’s sensitivity analysis to an emissions reduction of 5 per cent, when the reduction relative to BAU is 25 per cent. That obviously explains the factor of 5 divergence. I’ve posted a comment to Bolt’s blog pointing this out, but that comment too is awaiting moderation.
fn1. In the meantime, John H. has noted the erroneous estimates by Michael Bachelard, corrected here, and also some estimates by Christopher Monckton, presumably as a reductio ad absurdam