From the Jan. 27 New Republic article "The Great Savings Scare," by John B. Judis (for the full text, click here):

... [S]avings hawks ... assume that the amount of savings determines the amount of investment, rather than vice versa. But the question of which determines which is a contested one. Current policy debate pivots on it, and so does the difference between classical and Keynesian economics. ...

... Investment decisions are not based simply on interest rates and the volume of savings, but also on the expected rate of return. ... In calculating this rate of return, businessmen consider demand for their products and the general state of the national and world economy. If businessmen's uncertainty about future profits outweighs the lure of lower interest rates, then the savings that were initially created will disappear.

Take an economy in which workers save an extra 10 percent of their income and put the money into savings accounts that create potential funds for lending. That results in lower interest rates, but also in lower demand for consumer goods and, by extension, capital goods--the goods like steel and machine tools that are used to produce other goods. If businessmen, worried about future demand, don't invest in spite of lower rates, then they will eventually have to lay off workers because of reduced demand. As workers are laid off, total income will decline, so that even if workers continue to save the extra 10 percent, their total amount of savings will decline until it is equal to the amount of investment that business is now willing to undertake. Lower investment will cause lower savings, rather than vice versa.