Greg Mankiw's column in today's business section of the New York Times compares economic research on the stimulative impact of tax cuts relative to government spending. He cites research papers that find a larger impact for tax cuts, including work done by the President's advisor, Christina Romer. I know space is limited in a column, but I wish Mankiw had emphasized more the difference between temporary changes and permanent changes. As anyone who has had any microeconomics knows (or at least learned at one time) all elasticities are greater the longer the time period.
The stimulus offered while George W. Bush was still president was a tax cut, but a temporary tax cut. The effect was that most people used the tax cut to pay down debt or increase savings. I think these are good things, but they do not stimulate the economy. Similarly, increased spending that is know to be temporary in nature will not have as much effect as an increase in spending that should last a long time. If a firm wants to take advantage of a temporary increase in spending, it will not make long-term investments as part of the process. If the firm believed the spending might be available for many years, it would respond in a different manner. One of the reasons tax cuts provide more stimulus when the cuts are expected to be permanent, or as permanent as anything can be that involves the government, then there is both a spending effect and an incentive effect. The latter is not in place for temporary tax cuts or rebates. Probably the worst think Keynes ever said was, "In the long run, we are all dead." But time passes and constant focus on the short run leads to ad hoc measures that lead to more measures later on.