Dan Hamermesh has a nice post up at Freakonomics discussing how Obama's $3000 per job tax credit for companies who expand employment above current levels over the next 2 years is a great example of a well targeted economic policy:
Unlike inefficient subsidies that provide funds for an activity that would have been undertaken anyway, this kind of marginal tax credit only subsidizes new activity. The benefit per dollar of credit is greater with this approach; it is more target-effective.
Indeed, a number of studies evaluating the old N.J.T.C. suggested it had substantial effects in stimulating employment.
Moreover, the jobs created especially benefited low-wage workers: not surprisingly, since the cap on the credit per worker made it a more attractive percentage subsidy for hiring lower-skilled, lower-wage workers.
Theoretical work suggests it is especially likely to be successful in an economy that is sliding further away from full employment, as we now are.