In his quest to disqualify John Cook's ERL study, showing a 97% consensus on human induced global warming in the scientific literature, Richard Tol reanalyzed consensus estimates from other studies:
That's a rather strange claim, since most other studies find high to very high levels of consensus among climate scientists. Even stranger when Tol mentions the various sources of his datapoints:
The green datapoints apparently come from Oreskes, Bray, Anderegg, Doran, and Verheggen. There's even a datapoint on Tol's graph with a level of scientific consensus less than 10%! Something's up here...
That's when I knew he misrepresented our work:
It's such a bizarre argument that I can't help but becoming a little sarcastic:
Richard being Richard, he plays the ball back and we start a ping-pong match:
I'm flabbergasted that he actually digs his heels in, as the data in his graph don't at all mean what he implies they mean.
ATTP chimes in with a blogpost outlining the most egregious ways in which Tol's reanalysis goes wrong: