If you’re going to propose things that can pass Congress and they create jobs, then I don’t think it matters whether or not they’re popular. The job creation will be rewarded. But if you’re going to pass something that can’t pass Congress, then it doesn’t matter at all whether it would hypothetically work, all that matters is that it polls well. And as Chait says, the things that Keynesian analysis suggests would create jobs — much larger budget deficits, higher inflation — are not popular things to campaign on. The smart move, if you’re just going to give a speech for speech’s sake, is to make the speech be full of nonsense bromides that voters like to hear. Except one problem President Obama will face is that for a “nonsense bromides” strategy to be maximally effective, it would be really useful for the entire progressive echo chamber to get really excited about his bromide agenda and start loudly insisting that the bromides would be super-successful in reducing unemployment if implemented. But Paul Krugman, Rachel Maddow, etc. won’t do that. A speech full of bromides will be disparaged as bromidish. These are the wages of the “hack gap,” the fact that the progressive media ecology is less leadable than the Conintern.
This gets right to the heart of why the 11-dimensional bully pulpit doesn’t work for Democratic presidents very well. Pulpits are for spewing bullshit and the left is queasy about cheering bullshit too loudly.