The macroeconomic effect of the stimulus

Bruce Bartlett has collected some recent research on the effects of the 2009 fiscal stimulus (ARRA). John Taylor in Congressional testimony makes the following arguments:

1. Fiscal stimulus was not effective

- models suggesting it was have conclusions "baked in" from the beginning

- other models show small effects

- the 2008 tax cut and cash-for-clunkers had very little effect on consumption, consistent with the permanent income hypothesis

- most of the turnaround in GDP growth can be explained by inventory growth, not government purchases

2. The main reason growth is slowing is lack of confidence resulting from a scary fiscal situation

- CBO projects that debt will rise to 947% of GDP by 2084

- there's a lot of uncertainty about financial reform, effects of health reform, future taxes

3. Therefore what we need is fiscal consolidation. Reversing the stimulus will not hurt because it has had such a small effect anyway.

I will agree with Taylor on his very first point, that models suggesting a large impact from the stimulus are not terribly convincing because their conclusions are baked in. What we need is to rerun the world economy with no stimulus and compare it to what we have now. We can't do that literally, but we can do it using theoretical models. But the theoretical models will by necessity assume the conclusions one way or another, so the exercise is of limited utility. I'll also agree with his claim that the 2008 tax cuts (and maybe cash-for-clunkers) had relatively small effects.

The rest of his argument strikes me as very sloppy reasoning. Take the inventory vs. government purchases argument. Inventory investment turned positive because inventories had been drawn down to a very low level and businesses were anticipating an increase in sales. Where did that expectation come from? It is inconceivable that a significant part of that expectation did not arise from the fact that the government had just committed to spending $787 billion over a three year period.

The challenge that stimulus pessimists face is coming up with another theory to explain the effects of fiscal policy. The Keynesian story is simple: in conditions of deep recession, if someone in the economy (consumers, business, government) spends, that creates demand for goods and services, which is met by new production and employment. How could it be otherwise? Taylor and others have dug through pre-Keynesian business cycle theories and hit upon the problem of "confidence". Government spending creates uncertainty, which spooks investors and keeps them from spending. Let me enumerate the problems with this line of argument:

1. There is no evidence, none, that the uncertainty created by government deficits has a greater impact than uncertainty concerning the length and severity of the recession. I don't know of any studies that show that the fear of budget deficits has any effect on investment at all, except through long-term interest rates (which are now at extremely low levels). Proponents of this theory I think are engaged in projection: budget deficits make me uncomfortable, so they must make businesses uncomfortable, which must explain why they're not investing.

2. The idea that anyone is worried enough about the debt-GDP ratio in 2084 under the CBO's alternate fiscal scenario to put off investment spending is ludicrous. As we speak the stock market is lurching to and fro in response to company earnings reports for 2010Q2. Consumers are worried about whether they can pay their bills this month. We've seen how ridiculously short-sighted businesses have been, from the financial sector to BP to GM. They're not investing because of what they fear the fiscal situation will look like when my grandchildren are old and gray? Please.

3. There was going to be uncertainty about the health and financial sectors regardless of whether or not reform bills passed. Arguably the passage of comprehensive reform bills has reduced uncertainty rather than increased it. Furthermore, the health reform legislation was the most serious attempt ever to control costs in Medicare over the long haul; you would think that people like Taylor who are concerned about the deficit (and any businesspeople out there who are hinging their investment decisions on fiscal policy) would give the Democrats some credit for that. But noooooo.

So the truth is that we do not know what the effect of fiscal stimulus was. But the Keynesian logic suggesting a substantial positive effect is far more compelling than any alternative that has been proposed.