Perhaps the silliest debate in environmental economics is whether we should tax greenhouse gas emissions or invest money in clean energy research. The obvious answer is that we should tax greenhouse gas emissions to fund clean energy research, which is more or less what Mark Muro and Jonathan Rothwell propose in a new Brookings paper. Specifically, they want a carbon levy of $20 per ton of CO2 emissions that rises at 4 percent per year in real terms which should "raise on average $150 billion a year over a 10-year period while reducing carbon dioxide emissions 14 percent below 2006 levels by 2020 and 20 percent below 2006 levels by 2050."
That would be helpfully environmentally speaking, but not nearly good enough on its own. Hence "the first $30 billion of carbon tax revenue each year should be deposited into an independently managed fund for supporting top-quality energy-system RD&D activity" along with "technology deployment" and other 'push' measures to ensure that decarbonization proceeds rapidly and adequately." Then the rest can be used for deficit reduction, tax offsets, etc.
My quibble on this is that the initial levy should probably be higher.
You could do a lot to boost the short-term labor market without upsetting the deficit scolds by doing short-term carbon taxation and offsetting it with lower payroll taxes. The only real constraint on this ought to be political feasibility. It's true that a $30 per ton initial price is politically infeasible, but it's also true that a $20 per ton initial price is infeasible. As long as we're just writing white papers we may as well aim for the stars.