This is very entertaining, because it takes real skill to muck up reporting this PPP poll about Scott Walker’s chances in a hypothetical recall election; fortunately, Joe Conason is up to the challenge. Let’s look at what Conason wrote (bolding mine):
Asked whether they would support or oppose [Scott Walker’s] removal from office in a recall election, 50 percent said yes and only 47 percent said no.
The same poll found that Wisconsin voters are also apparently sorry that they replaced progressive Democratic Sen. Russ Feingold last fall with a tea party extremist named Ron Johnson. Today, they would re-elect Feingold with a comfortable margin over any Republican…
I’ve already gotten a screenshot of this, by the way. Just in case Truthdig decides to memory-hole the entire thing. 
Anyway, again: Conason is reporting that 50% of PPP’s respondents favor a recall of Walker. So far, so good for the Democrats. PPP also reports that in that hypothetical election Feingold would win. OK, piece of data to consider. But what Conason did here – and probably deliberately, seeing as he didn’t link to the original poll – was falsely claim that this means that Feingold would win re-election against Senator Ron Johnson, despite the fact that PPP did not poll that hypothetical match-up. You see, Feingold was never governor of Wisconsin, so he cannot be re-elected to that position. The poll is strictly about the Wisconsin recall situation*; not about Ron Johnson. Perhaps PPP will poll a hypothetical rematch between the two, although why anybody would bother is beyond me completely (it’s not going to matter before 2016 anyway); but until then, it’s dishonest to use polling results in this manner.
Yes, ‘dishonest.’ Remember, we know that Conason meant this poll, because the numbers that he did specifically quote (but not source) are the same; and we know that Conason meant Feingold’s re-election as Senator, because he did specifically use the word “re-elect.” If Joe Conason tries to claim that this was all an innocent mistake – which he undoubtedly will – then said claim should be seen as the calculated insult to his readers’ intelligence that it is.
Moe Lane (crosspost)
*As to the poll itself: yup, pretty harsh. Guess we’ll see in July how those legislature recall numbers hold up, huh? And the obvious problem that the Democrats face in exploiting their hypothetical advantage is left as an exercise for the reader: I don’t give free hints to the Left unless it suits me.