(Hopefully this will convince you that the answer is we don’t know).
In Part 2 I mentioned the many different models that have tried to predict whether the stimulus worked. One often hears that there is a consensus that the government stimulus worked. On the contrary, there isn’t even a consensus about the correct assumptions used to build these models.
One key portion of these estimates that is still contentious is what the so-called “multiplier” should be. The multiplier is the effect that a one dollar increase in government spending has on total output. One theory is that the multiplier is large, meaning that if the government spends a dollar, on say, a road, the construction worker who earned that dollar will take some of the money and buy shoes. The shoe salesmen will in turn spend some of the money at the grocery store and so on. In this way the dollar increase in government spending ripples (or “multiplies”) through the economy. Alternatively, because the government is using real resources by spending, the increase might actually hurt the economy by “crowding out,” as economists say, the private sector.
“If the multiplier is larger than 1, it means that government purchases multiply or stimulate private sector economic activity. If it is between 0 and 1, it means that purchases displace or crowd out private sector economic activity. And if it is less than 0, it means that government purchases crowd out enough private sector economic activity to offset any increase in public sector activity.”
He also links to a chart he presented during a February congressional testimony that clearly shows the disagrement in the field on what the true value of this multiplier is in the current US economy (and by implication the disagreement over what effect any type of government spending increase will have):
Mitchell summarizes the results:
“Note that there is a wide range in the estimates both across and within studies. If the optimistic scenarios are correct, an additional $1.00 in deficit-financed government spending spurs $2.70 in new private sector economic activity. But if the less-optimistic scenarios are correct, then an additional $1.00 in spending destroys $3.80 in private sector activity.”
And because probably every statement about the success of the government stimulus relies on models that have to assume a value somewhere along this spectrum, it’s clear that the results are in some sense arbitrary. Not arbitrary in the sense that they are unreasonable, simply arbitrary because you have to pick some value (or range of values) to have a functioning model. As far as I know, most models assume a multiplier greater than 1, implying that there is some effect of increased spending even if it is less than the full dollar value spent. This seems reasonable in some respect since “only” 5 out of the 21 studies show a possibility of a negative multiplier (but these studies themselves rely on still more assumptions).
For even more evidence that there is a lack of agreement on the effects of fiscal stimulus see the preliminary drafts of the National Bureau of Economic Research’s forthcoming book on entitled, “Fiscal Policy After the Financial Crisis.”