Riedl (Heritage) shoots down the usual Keynsian fallacies. A couple of them:
...among proponents of government spending there is a strong focus on whether people are spending or saving, with the implication that spending circulates through the economy while savings effectively drop out.
But savings do not drop out of the economy. Nearly all people put their savings in: (1) banks, which quickly lend the money to others to spend; (2) investments in stocks and bonds; or (3) personal debt reduction. In each of these situations, the financial system transfers one person's savings to someone else who can spend it. So all money is quickly spent regardless of whether it was initially consumed or saved. The only savings that drop out of the economy are those hoarded in mattresses and safes.
OK. Next?...government spending can put under-utilized factories and individuals to work--but only by idling other resources in whatever part of the economy supplied the funds. If adding $1 billion would create 40,000 jobs in one depressed part of the economy, then losing $1 billion will cost roughly the same number of jobs in whatever part of the economy supplied Washington with the funds
And this one:
Declining consumption means that either: (A) more income is diverted into investment or government spending (which is zero-sum in the short run); or (B) less income is created overall, which typically leads to less spending across all categories. For the latter situation, the solution is to create incentives for productivity that create more wealth and income for people to spend across all categories.
There's also the question of "bubbles," which are a result of mal-investment. Were dotcoms really worth all that money? Houses? Tulips?