I recently started reading Paul Krugman's column in the New York Times and was particularly interested in his mathematical analysis of Obama's $775b stimulus plan.
Based on his back-of-napkin calculations, I created a simple (and ugly) tool to simulate a stimulus's effect on unemployment. The tool is pre-populated for Obama's proposed stimulus.
The bottom line? Obama's proposal, if Okun's law holds, would result in a 6.75% increase in GDP and a 3.4% decrease in the expected unemployment.
Strangely, Krugman only shows the net result after one year (1.7% decrease in expected unemployment), not the net result by the end of the stimulus in 2 years. It's this first-year result that he is afraid is not enough to appear effective to conservatives. In fact, it almost sounds like he accidentally thought that his first-year calculations were the end-result:
"Suppose that we’re looking at an economy that, absent stimulus, would have an average unemployment rate of 9 percent over the next two years; this plan would cut that to 7.3 percent, which would be a help but could easily be spun by critics as a failure."
He may just be saying that the failure will be spun after just one year of the plan, but I'm still surprised he made no mention of the overall effect.
If you're interested, here are the mathy details:
"Okun's Law" relates GDP growth to unemployment. Assuming "Okun's coefficient" is 2, it says:
You need a 2% increase in GDP to reduce unemployment by 1%So if we can increase GDP with, say, a stimulus, we can decrease unemployment. Here are some rules of thumb for how a stimulus increases GDP (
Krugman's source):
For every $1 the government spends, they make $1.50For every $1 cut in payroll taxes, they make $1.30For every $1 cut in business taxes, they make $1So a simplistic equation for decreasing unemployment by stimulus is:
Decrease in Unemployment = 100 *
[spending*1.5 + (payroll tax cuts)*1.29 + (business tax cuts)*1] / (2 * GDP)
UPDATE: Krugman also stated that the latest forecast is a high of 8.4% even with the stimulus going into 2010. Obviously this means there's some other variables at work here wonking with the equation.
.jpg)