In his December 3^rd column, Washington Post writer Harold Myerson argues for a second stimulus but bases that argument on information known to be unreliable and on misrepresentation of basic economic principles.

First, Myerson notes that a recent CBO report says “that federal spending has saved or created more than 640,000 jobs, and possibly as many as 1.6 million.” He does not mention the well-known inaccuracies in reported data, such as thousands of jobs being “saved or created” in Congressional districts that do not exist, firms reporting more jobs “saved or created” than their total number of employees, and multiple instances of firms saying they did not understand how to fill out the relevant forms.

More importantly, Myerson neglects to mention the CBO’s own critical caveat regarding the type of economic model used in their calculations of the effects of the stimulus. In particular, the CBO relies on “general-equilibrium” models which are widely believed to be unrealistic representations of actual human economic behavior and which produce “multipliers” much higher than either empirical data or other models suggest. (A multiplier is essentially the total effect on the economy of $1 spent. So if the government spends $1 in a situation where there is a multiplier effect of 2, that $1 will have created $2 of total economic activity.)

The CBO says they explicitly do not use other models despite skepticism about the model they do use. Thus they ignore the Life-Cycle Hypothesis (for which MIT economist Franco Modigliani won the 1985 Nobel Prize in Economics) and Milton Friedman’s Permanent Income Hypothesis, each of which attempts to better model human behavior.

The effect of the choice of model is to give a much higher range of multiplier to government spending and certain transfer payments than is widely believed outside of the few Keynesian economists left – most of whom happen to work for government, giving them a self-interest in increased government spending.

For example, the CBO estimates a minimum multiplier for purchases of goods and services by the government of 1 and a maximum of 2.5. A number of studies, however, conclude that the multiplier more likely has a maximum just above one and is generally less. In a Wall Street Journal opinion piece earlier this year, Harvard economics professor Robert Barro says that “when I attempted to estimate directly the multiplier associated with peacetime government purchases, I got a number insignificantly different from zero.”

These errors in multiplier assumptions make the lower end of the CBO’s estimate of jobs “created or saved” more likely to be the maximum than the minimum.

Myerson also argues that there should be a second stimulus because spending cutbacks and tax increases by state and local governments are partially offsetting federal spending. But, like the CBO’s refusal to use a model that represents how people actually behave, Myerson assumes that government can magically create wealth by spending even though government only gets money by taking it (in the future) as taxes or through debasing the currency, the fear of either being enough to cause entrepreneurs to avoid starting or expanding a business. In typical Keynsian fashion, Myerson assumes that you can take money from your left pocket, put it in your right pocket, and say you’ve gotten richer.

Myerson also notes that much of the “stimulus” spending has not yet occurred, being planned for 2010 despite being passed in February. Instead of concluding that we should enact a second stimulus, Myerson should have reached the more obvious conclusion that the stimulus was not intended – even by people who believe in Kenyesian pump-priming – to stimulate the economy as soon as possible. Instead, it was intended to stimulate the Democrats’ chances of minimizing electoral defeat in the 2010 elections by buying shovel-ready votes with our money.