Keynesian and tax multipliers
Topics: Stimulus
03 Dec 2010
From: Ervan Darnell
The most significant debate before the lame duck Congress is extending
the "Bush tax cuts" (I object to calling a tax increase extending a cut,
but that's a different matter). Obama having argued for spending
stimulus is now arguing that tax cut stimulus is too expensive (can no
longer find the reference where he said any tax cut was a deficit buster).
To first approximation, borrowing to finance government spending and
borrowing to finance a tax cut, most of which will go into spending, are
both the same thing: borrowing against the future to encourage
consumption now. With even Mankiw finding QE2 tolerable [1], I'm in a
small minority but this still makes no sense to me. Consumption needs
to shrink (since we were spending money we didn't have by borrowing
against equity we didn't have). Stimulating the economy back to
unsustainable levels only makes the future worse. The Keynesian
argument has to be that the market overreacts and drops productivity
below where it should be, while the government somehow knows the right
amount.
Even accepting that for the moment, which is better for stimulus,
spending or tax cuts? The argument is usually phrased in terms of the
multipliers. The Keynesian multiplier is how much GDP you get for $1 in
extra government spending. Estimates range from 1.0 to 1.6 [2]
(ignoring the future pain of paying back the debt of course). The tax
multiplier is the other way, how much new spending do you get from
cutting taxes. Estimates range from 1.0 to 3.0 (none less than
Christine Romer found the 3.0) [3], and ignoring future debt again. I
don't know that one of these estimates is better or worse than another,
and they do vary widely.
My thought for the day is how this measures the wrong thing though.
These numbers look only at the GDP. They don't say anything about how
desirable that new GDP is. When the government buys $1 extra with
deficit spending, the actual value of what it buys might be a lot less.
If it's make work jobs, doing things we don't really want done, then
actual value is $0. If it's health care spending at market rates, then
value is nearly $1. If it's keeping a state education bureaucrat
employed (as the Obama stimulus did), then it's actual value is more
like -$1 for the extra harm that person does by performing their job.
On the other hand, cutting taxes means that extra spending is actually
spent on things we want and thus it has full value. So, if we are to
borrow in order to stimulate the economy, then tax cuts are likely to
produce more real value. The downside is that Congress loses vote
buying opportunities with that money. The other downside is that tax
cuts first help people who pay taxes, and indirectly help everyone else
with the higher demand creating new jobs. Spending could theoretically
be targeted at the unemployed (though it hasn't been outside of
unemployment "insurance").
-------
[1] http://gregmankiw.blogspot.com/2010/11/qe2.html
[2] http://gregmankiw.blogspot.com/2008/12/spending-and-tax-multipliers.html
[3] http://www.voxeu.org/index.php?q=node/3962 (not sure of the
credibility of this site, but these numbers are common).
Home