I think Arnold Kling is off the mark here : TCS: Tech Central Station - An Open Letter to Paul Krugman
Ad hominem attacks, including criticism of an opponent's motives, are, I agree, irrelevant when you want to make an argument like this : I say A -> B (eg. tax cuts imply economic prosperity). If you disagree, the reason why I said it, has no bearing on whether it's true or not.
However, if you want to argue a different point, that I am untrustworthy, then evidence of my malice, stupidity or duplicity is completely relevant.
When Krugman attacks supply-side economics he doesn't do it on the grounds that supply siders are motivated by right-wing politics, he does it on his interpretation of the data. With a perfectly valid "Consequences" argument.
To put this formally, let's say C = "pure motives"
Kling suggests Krugman is arguing
C -> (A->B),
Which clearly has no logical validity.
But I'd say he's arguing
A, (Raegan's tax cuts)
NOT B (No sign of supply side effect in the data)
NOT (A -> B),
NOT C (though this is an independent issue argued for with separate data.)
Krugman does add some logically unnecessary evidence :
NOT A (Clinton's tax increase),
B (economic growth in the 90s)
from which we can conclude nothing about A -> B
But really, all Krugman is guilty of, is adding a second, separate claim, about power and politics, which attempts to explain why supply-side theory is so heavily promoted. And he has evidence for this too.
Kling is right that the second point has no logical bearing on the first. But that's not why it's there. It has a different relevance to the political point being made in the article. It's a warning ... "these guys are not motivated by improving the economy according to supply-side theory, they're motivated by wanting to shrink government, so evidence against supply-side economic theory isn't going to stop them."
Now Krugman may be wrong in his interpretation of the data, and supply-side may not have been falsified. In fact, supply-side economic theory may well be true. But with the assumption that he's right about this, the rest of the logic of the article is fine.
ps : There is a positive side to Kling's attack. It introduced me (and probably many others) to a great Krugman essay I'd have otherwise missed.