Long Live Bad Models and Government Spending!

[Casino67]

NationalReview.com:

Thanks to my recent discovery of Twitter, I see that Slate‘s Dave Weigel is repeating the line that the $61 billion in spending cuts just approved by the Republican-controlled House will hurt the economy:

Democrats have been arguing for weeks that the GOP’s spending cuts will be bad for the economy, and in the last week they’ve celebrated two new analyses that bear none of their fingerprints but validate just about everything they’ve been saying.

I am not quite sure what there is to “celebrate” in these two studies.

Many of us have said it before, but it’s worth repeating: Both studies are based on the same faulty models that predicted that the stimulus of 2009 would stimulate economic growth. This theory assumes that $1 in government spending triggers well above $1 in economic growth and creates many jobs. The Goldman Sachs study goes even further: Its estimate seems to imply a multiplier effect greater than 3, well above anything in recent literature. A review of the literature about the value of the multiplier reveals that in most cases, a dollar in government spending produces less than a dollar in economic growth — and these findings often ignore the impact of paying for this government dollar by increasing taxes. Harvard University’s Robert Barro finds a multiplier between 0.4 and 0.7; at Stanford, John Taylor and John Cogan find that the stimulus package couldn’t have had a multiplier much greater than zero. Even the multipliers used by Christina Romer and Jared Bernstein in their January 2009 paper “The Job Impact of the American Recovery and Reinvestment Plan” ranged from 1.05 to 1.55 for the output effect of government purchases. More recently, Dartmouth economists James Feyrer and Bruce Sacerdote acknowledged that the stimulus didn’t boost the economy nearly as much as the administration models claimed that it would. These are only some of the many studies on the issues.


----------------

A little down in the weeds, but interesting.