ARRA and growth in 2010

Menzie Chinn and Paul Krugman comment on a Deutsche Bank study showing that while ARRA (the fiscal stimulus package passed last February) should increase the level of GDP in 2010 relative to what it would have been without ARRA, it subtracts from economic growth in the 2nd half of the year.

This is not evidence that ARRA is doing more harm than good. The reason ARRA subtracts from growth in the second half of 2010 is that spending under the program starts declining at that point from its peak in 2010Q2. To use Krugman's numbers, if stimulus spending is $125 billion in 2010Q2 and $120 billion in 2010Q3, and if government spending is 15 percent of GDP, then the $5 billion decline in spending would contribute (5/125)x.15 = .006 (6/10 of 1 percent) in that quarter.

The lesson is not that we would be better off in 2010H2 without the stimulus: as the graph above shows, GDP is still 2% higher than it would be otherwise. Repealing ARRA would be tantamount to accelerating the phaseout, which would increase the negative contribution to growth.

Krugman wishes we had built in more stimulus in 2010H2 - the phaseout should occur in 2011 or later. This makes sense if you believe that growth apart from the stimulus will be sluggish in 2010H2. If on the other hand you believe that the economy is poised for a strong rebound in 2010, then phasing out the stimulus in the second half, at a time when the rest of the economy has a full head of steam, is perfectly appropriate.