I am a skeptic. I believe the one-year multiplier is about one. Four years forward, insignificant from zero.
The fiscal multiplier is the increase in GDP per marginal dollar of government spending. The time element involved in this is not often mentioned. The idea of how a dollar of goverment spending can increase production by more than a dollar is basically that someone gets the dollar the government spends and spends it again, and this cycle repeats.
When I refer to a one-year multiplier of one, I mean "if the government spends $1 today, GDP will be $1 higher this time next year". The four year multiplier of x similarly means "if the government spends $1 today, GDP will be $x higher in four years". My numbers are lower than those chosen by many people who endorse substantial fiscal policy, hence I am a skeptic.
Here's a canonical blogosphere post on multipliers.