Milton Friedman get credit for the title, but the current Congress and the Obama Administration get credit for the content. When the stimulus bill was being debated, a number of economists explained how "stimulative" results would be less than promised because the composition of stimulus spending was tilted toward industries (like health care) that were already doing well. Now the AP reports that the same pattern is found regionally: road construction is going to regions that are already doing well.
That is the likely trade off of government spending: in order to even come close to stimulating, the spending must be on items where private spending is low -- that is, items that the private sector current finds to have little value. One example is home building. Another example is expansion in depressed areas (eg., a road project in an area that is losing population).
Conversely, if the government spending is to be considered valuable (or at least better than stupid), it will likely have to occur in areas where the private sector is also spending, such as health care and construction in places that are gaining population. But, when it comes to areas where the private sector is already spending, you have to wonder why the government needs to add to it.