Here's an interesting nugget from the Washington Post's write-up of its new poll:

Public opinion is split down the middle on the question of whether the government should spend more money to stimulate the economy in a way that leads to job creation. Among those who support such new spending, 18 percent change their minds when asked what they think if such outlays could sharply increase the budget deficit. In that scenario, 57 percent opposed another round of spending.

No, the interesting part isn't that 18 percent changed their minds and opposed stimulus spending when they were told it could “sharply increase the budget deficit.” There's nothing surprising about that.

The interesting thing is that the Post tested what happens when you give people negative information about stimulus spending -- but didn't test what happens when you give them positive information about such spending, or what happens when you give them negative information about the consequences of not spending money on stimulus.

So the Post asked people if they supported stimulus spending. Then they asked those who support spending if they would still support it if it “sharply” (a loaded word) increased the deficit. Then they took everyone who said “no” to the first question and added them to the people who said “no” to the second question, and reported that number as indicative of opposition to stimulus spending. In other words, they completely stacked the deck.

The lack of balance in the Post's stimulus questions not only means that the results the paper touted are of questionable value, it also raises the question of why the Washington Post needs polling to tell it that if you give people negative information about only one side of a public policy question, you're going to drive them to the other side. That's pretty obvious.

UPDATE: Last year, I pointed out a similarly skewed Post poll that stacked the deck against the public option.