This articleclaims that $5 spent on food stamps brings $9 of extra economic activity. Which kind of sounds like value for money doesn’t it? But there’s a flaw in that argument, namely that the "extra economic activity" created by a particular type of spending (i.e. the multiplier effect) is actually irrelevant. I set out the reasons hereand here.
And that (to repeat) is not to argue against more money being spent on food stamps. The point is that in determining the merits of spending $X on Y or Z, the costs and benefits of Y and Z are the crucial factors: the multiplier effect is irrelevant.