If you haven’t noticed already, the polling about President Obama’s jobs proposal is very confusing. Ask people whether they like the plan as a whole, or how they rate Obama’s handling of the economy, and you’re likely to get a negative, or at best middling, response.

In last week’s CBS/New York Times poll, for example, only 34 percent of respondents said they approved of Obama’s management of the economy while 47 percent said they were confident the jobs bill would help. That was roughly the same percentage who said they weren’t confident it would help.

But ask people whether they like the elements of the plan and you’ll get a positive to very positive response. In that same CBS/Times poll, 52 percent favored giving more money to the states, 56 percent favored a payroll tax cut, and 80 percent (!) favored infrastructure spending.

Nor is that enthusiasm limited to Democrats. In Gallup’s latest poll, for example, 56 percent of Republican respondents said they supported using taxpayer funds to hire public workers and 50 percent said they supported public works, including school repair. (The numbers for non-Republicans, naturally, were higher still.)

That last result led Greg Sargent, who has been following this story poll by poll, to declare “this is just crazy.”

Crazy, yes. But familiar too. Polls on health care reform have generated similar responses since early in the debate: Disapproval of the plan as a whole, but strong approval of its individual elements. Come to think of it, polls on health care reform broke down in the same way back in the 1990s, when it was Bill Clinton trying it.

So what does this tell us about how the public really feels – and what liberal politicians can do to win them over? I haven’t a clue, so I put the question to two people who do.

One was Guy Molyneux, the veteran pollster and writer with Hart Research. Via e-mail, he wrote:

I think the main explanation is that these types of poll questions that break out individual components of the plan imply that the policy will do what it's supposed to in each case. That is, that cops and firefighters really will be hired; that job-creating businesses will see their taxes go down; that public works will be built and workers hired to build them.
However, a lot of voters -- and most Republicans -- don't actually believe that any of those things will happen if "Obama's bill" is passed (just as they don't think the original stimulus did any of these things). They don't believe it in part because of general cynicism about government, and -- if they are a Republican -- because the noise machine tells them this every day. They think the money will instead be spent on corrupt payoffs to ACORN, or god knows what.
That's why, if you follow the Gallup links, you will see that only 19 percent of Republicans support the Obama jobs bill without any explanation of what's in it. And it would be 9 percent if you called it the "Obama stimulus bill.”
That said, I do think voters want action, and some elements of the bill will be pretty popular if the president keeps pushing them. The payroll tax cut in particular will be popular, I believe, once people know they are looking at a 2 percent increase in withholding come January, 2012, if nothing is done. (It’ll help if they are reassured this doesn't reduce Social Security benefits.”

The other person I consulted was John Sides, the political scientist from George Washington University:

1) Questions about the state of the economy and Obama's stewardship of the economy generate answers that are based on economic indicators and news about the economy. These questions are measures of performance, and answers to them will become more favorable when the economy improves.
2) Questions about policy proposals are, if Obama's name is mentioned in the question, basically measures of affection for Obama. If you like him, you like policies attributed to him, etc. You'll see some of these questions in current polls about the jobs bill.
3) Questions about policy proposals that don't mention Obama are more complicated. Respondents who pay a lot of attention to politics will be able to attribute policies to Obama, and they will then be polarized along partisan lines. For the findings about Republicans supporting elements of the bill, I'd be interested to see if politically attentive Republicans are less favorable to the elements of the jobs bill than other Republicans.
But other respondents, who may not know where these policies are coming from, would just be evaluating the policy off the top of their head. In essence, these respondents are just asking themselves, "Does this sound good?" And a lot of the proposals in Obama's jobs bill appear to sound pretty good.
Essentially, people survey questions off the top of their heads. And for many people, the considerations on the top of their heads vary from question to question. Their answers may appear incoherent to someone who spends enough time learning and thinking about politics. Political junkies can "link" their opinions on all of these issues. But other people, who are just answering questions when a pollster phones and interrupts their dinner, do not necessarily do so.

OK, clear now?