Fair enough.
Here is the original article[1] and a rebuttal[2], both on 538. Doesn't read like wild-eyed denialism to me.
[1]: http://fivethirtyeight.com/features/disasters-cost-more-than...
[2]: http://fivethirtyeight.com/features/mit-climate-scientist-re...
It is hard to mount one's best arguments when one is operating under tight resource constraints and time pressures. HN threads get stale very quickly.
> You could cherry-pick tons of evidence to support either position, if that's your thing.
Heh, this whole disagreement may actually be a mistake on my part about the meaning of the word "scant." I thought "scant" meant "rare or non-existent" but it doesn't, it means "barely sufficient or adequate." So I was about to say that if there's tons of evidence to be cherry-picked then it cannot be scant. But I suppose it could be.
In any case, I did finally manage to find some drought data:
http://www.cgd.ucar.edu/cas/adai/papers/Dai-drought_WIRES201...
That's a review with 137 references, so the evidence doesn't look "scant" to me. But I suppose a reasonable person might disagree.