Mark Thoma points to an analysis by Ted Gayer and Emily Parker.
The $2.85 billion program provided a short-term boost in vehicle sales, but the small increase in employment came at a far higher implied cost per job created ($1.4 million) than other fiscal stimulus programs, such as increasing unemployment aid, reducing employers’ and employees’ payroll taxes, or allowing the expensing of investment costs.
Pointer from Mark Thoma.
Although this analysis supports a view that this program was not a very effective stimulus, I think this sort of analysis has to be somewhat tenuous. Any government spending involves a diversion of funds from some other use. Any government spending redistributes income. As far as I can tell, Gayer and Parker assume that the subsidies accrued to car buyers. But maybe the subsidies accrued to auto companies or auto workers, in which case the multiplier effects would show up rather indirectly. The concept of what is seen and what is not seen casts suspicion on any calculations of the sort attempted here.
I am not trying to defend the Cash for Clunkers program, of course. I am just trying to point out how difficult it is to draw firm conclusions about the effect of macroeconomic policy.
It’s interesting that he lists extended unemployment insurance as a stimulus program. The data seems very clear to me that extended UI prolongs unemployment duration. The idea that it is a short term stimulus is crazy, but I’m sure there is some model somewhere that proves it created some thousands of jobs.
You say:
“Any government spending redistributes income. ”
Well, that might be true if we are to include income that may be earned or produced in future generations.
All government spending in excess of revenues collected, does not redistribute “income” in the strict understanding of that term. Spending funds borrowed from sources outside the economy should not be qualified as redistributing income. A case can probably be argued that funds borrowed internally, whether they come from capital accumulated through prior incomes, or come from current incomes, is a redistribution of those sources of incomes.
Beyond that the statement is too broad.
R Richard Schweitzer:
“Spending funds borrowed from sources outside the economy should not be qualified as redistributing income.”
Why not?