Not long ago Paul Krugman wrote:
To a first approximation, in other words, the effect of current fiscal policy — whether stimulus or austerity — an [on?] the actions of future governments is zero.
He makes further points at the link, although there is not a citation to the literature. I thought we should look at the evidence a little more closely. Some of it contradicts Krugman as read literally, though it is not all bad news for his larger point.
Here is an abstract from Brian Goff:
In spite of Peacock and Wiseman’s 1961 NBER study demonstrating the “displacement effect”, simplistic theoretical and empirical distinctions between temporary and permanent spending are common. In this paper, impulse response functions from ARMA models as well as Cochrane’s non-parametric method support Peacock and Wiseman’s conclusion by showing 1) government spending in the aggregate displays strong persistence to temporary shocks, 2) simple decomposition methods intended to yield a “temporary” spending series have a weak statistical foundation, and 3) persistence in spending has increased during this century. Also, as a basic “fact” of government spending behavior, the displacement effect lends support to interest group and bureaucracy models of government spending growth.
There is persistence to spending, although this study does not create a category for stimulus spending per se, however that concept might be defined. The work of Robert Higgs also provides a clear look at ratchet effects on government spending, control, and regulation, although Higgs focuses on war rather than spending. State governments also seem to exhibit a ratchet effect, whereby good times bring about permanently higher budgetary demands, if only through endowment effects, lock-in, and status quo bias.
That said, the federal debt/gdp ratio seems to show mean reversion, as does the measure of primary surplus. That would mean that fiscally troubled situations are followed by improvements, though not necessarily from spending decreases. In fact there has been considerable reliance on a “growth dividend.” And here is Henning Bohn from the QJE:
How do governments react to the accumulation of debt? Do they take corrective measures, or do they let the debt grow? Whereas standard time series tests cannot reject a unit root in the U. S. debt-GDP ratio, this paper provides evidence of corrective action: the U. S. primary surplus is an increasing function of the debt-GDP ratio. The debt-GDP ratio displays mean-reversion if one controls for war-time spending and for cyclical fluctuations. The positive response of the primary surplus to changes in debt also shows that U. S. fiscal policy is satisfying an intertemporal budget constraint.
In other words, we make up for first-temporary-then-permanent spending boosts by a mix of growth and higher taxes. Krugman might well be happy with that scenario, but the data do show intertemporal interdependence for budgetary decisions, with a mix of persistence on one variable (spending) and mean-reversion on another (debt-gdp ratio). And if you think a lot of government spending is inefficient, you should still be troubled by apparently “temporary” spending bursts.
As with much of macroeconomics, I would apply a good dose of agnosticism to these results (noting that agnosticism is not the same as assuming zero effect), but still the correlations are consistent with my intuitions more generally.