I think Paul is disingenuous (there I go again, more hysterics) when he says that the blog-post in question
makes it clear he is not in favour of acting on the short-run argument
for protection. You be the judge. In my view, he says it is all very
complicated. He kind of supports free trade in an ideal world--that
would be an Econ 101 world--but there are circumstances, which happen
to be very like today's circumstances, in which protection could make
sense. This "needs to be taken seriously", and the case gets stronger
if optimal macro co-ordination is not forthcoming (which it won't be).
Ask any Democratic congressman what Paul's advice on "Buy American" is.
The answer will not be, "Don't do it." It will be, "The most brilliant
economist I know of thinks there's a case."
My exchange with Robert Barro was via email. With his permission, here it is:
I hope that all is well. It seems from your writing that I at least
had the beneficial effect of neutralizing Krugman. I am not sure what
you mean by "consensus." You don't think it's pretty wild to have a
multiplier of 1.5, so that government spending is not only free, it has
negative costs? The only informative empirical work I know of (from
Valerie Ramey and my own work) finds multipliers associated with
defense spending that are positive but less than one. In Valerie's
work, this shows up as a negative effect from added government military
purchases on private consumer spending. Multipliers for other forms of
government spending are imprecisely determined but are not
significantly different from zero. Actually, I hope that my current
long-term empirical project will shed more light, including macro
effects from changes in marginal tax rates.
You want me to ignore good economic theory and empirical analysis
and pretend that a voodoo multiplier above one should be respected as a
Thank you for writing. It is good of you to take the trouble.
If we are to have Paul doing what he does in the NYT, it might be
better to have you writing in a similarly gladiatorial way than not at
all. But I do think it would be best of all if you, Paul, and other
scholars of such distinction gave a little more thought to the harm
that this unscholarly politicised jousting can do.
Rather than asking them to think, Paul gives comfort to the massed
ranks of liberal economic illiterates--and you do the same for their
conservative equivalents. In the middle, where there are open minds
willing to be enlightened, the typical response is to switch off. I
think that this is a shame. Needless to say, I am not arguing that you
(or Paul) should be untrue to your own views when writing for a wider
audience, only that you should not deny or dismiss the professional
consensus, such as it is, out of hand, or advance your own views with
deliberately exaggerated certainty.
Yes, I do think there is a broad consensus on fiscal policy--one
that you, I don't need to say, have helped shape. It is that fiscal
policy can deliver a stimulus under certain circumstances, albeit not
as powerfully or reliably as was widely believed pre-Barro.
The consensus has absorbed your contribution by advocating
counter-cyclical fiscal stimulus more cautiously than before, and with
much closer attention to the implications for public debt and
financing. Of course you would go further--and everybody who cares
about this subject will pay close attention to your ongoing research.
You may very well push the consensus further yet. But, for the moment,
you can press for greater caution in attempting fiscal stimulus, or
even argue against it altogether, without saying that advocates of
stimulus under present circumstances--including economists like Paul or
Larry or Christina Romer, whom I know you respect--are dealing in
By the way, is it helpful to say that a multiplier of 1 implies that
government spending is free? The implication would be that even a
multiplier of 0.5 (which you seem willing to contemplate) would buy you
real resources at half-price. Obviously public spending eventually has
to be paid for, in full, one way or another--which is true regardless
of the multiplier. It is quite another thing to say that a multiplier
of more than zero commits you to the view that we can have a free
lunch. That is a debating point, something to please one's allies and
annoy one's opponents--not an aid to understanding.
Perhaps all I am saying is that economists of the stature of you and
Paul and Larry and so on should address each other in your pop writings
with something like the same courtesy and intellectual seriousness you
would use in an academic setting, rather than pushing custard pies into
each other's faces. That just looks bad--and I find it hard to believe
that you or Paul value the approbation of the idiots on either side who
are delighted by it.
Sorry, but I do not believe you are getting the economic substance
right. A multiplier of 1 means that, from a social perspective, the
added government purchases really are free--they are provided by the
utilization of idle resources. This point holds even if the spending
is deficit financed so that the public debt rises, although there can
be deadweight losses in the future from financing the larger debt. In
the same sense, if the multiplier is greater than 1, say 1.5, the free
public goods really do come with the bonus of free private consumption
or investment. I think this view is the one actually held by Summers,
C. Romer, etc., and I really do think it's voodoo macroeconomics. I
think you do not regard it as voodoo because you find it familiar and
If, as you say, the multiplier is positive but less than 1--say
0.5--the public goods cost only 50 cents on the dollar. This is
correct, but I think, in practice, the only evidence for positive
multipliers comes from responses to (temporary) military outlays during
wars (such as in the U.S., where the wars have not involved a lot of
physical destruction or loss of life). In these cases, the added outlay
did require a lot of added work beyond reduced unemployment, notably by
women, so this extra effort has to be included in the social cost. For
non-defense outlays, the best empirical estimate I have at this point
is a multiplier of 0 (though not precisely determined). This is the
standard cost-benefit case where the benefit of an extra unit of public
spending has to justify the added cost one-to-one. This multiplier of
0 should not be regarded as a pole, but rather as a central point. In
the medium run, the multiplier is likely to be negative, reflecting the
adverse effect of larger government on economic growth.
By the way, although WWII raised U.S. real GDP a lot, this response
is not typical for the OECD. If fact, WWII is the biggest economic
disaster of the 20th century, out-stripping the Great Depression. This
is, of course, because many countries suffered greatly from physical
destruction and loss of life (the U.K. not nearly as much as many
countries on the European continent).
As to Krugman, my response would likely have been more moderate if
he had not referred to my ideas as bone-headed. But I promise to
behave better in the future.
Yes, I take your point, of course. From a social welfare point of
view--ignoring distributional issues as they affect taxpayers v
bondholders, this generation v later generations, residents v
foreigners--then the cost of a fiscal stimulus that puts idle resources
to work is confined to the deadweight losses of financing the debt. It
feels strange for me to be reminding Robert Barro
that those losses and distributional issues might be a big deal, and
that one should be cautious on those grounds (among others) about
advocating immoderate fiscal activism--but I just did, so there you are.
I would never presume to quarrel with you (or Paul) on the
economics. That would be absurd. I am quarrelling with you about your
journalism. Your view that counter-cyclical fiscal policy cannot--even
in present circumstances--put idle resources to work may for all I know
be true. I resolve to keep an open mind about it. But I am sure that
for the moment the larger part of the economics profession disagrees
with you. And I am sure that in challenging that consensus by calling
it voodoo economics you are doing the discipline no favours.
Did Paul call you bone-headed? That was a bone-headed thing for him
to say. But you see my point. What are non-economists to make of this
exchange of insults and exaggerated certainties? Can you blame them for
hearing only the political message they want to hear, and tuning out
Oh yes, the bone-headed remark is what started things.
One thing I think you need to be clear on is the distinction between
Ricardian equivalence and Keynesian multipliers. The first bears, for
example, on how a deficit-finance tax cut affects aggregate demand.
The Ricardian view is no effect, and the "standard" view is that the
effect is positive but less than one. The multiplier has to do with
how a change in aggregate demand affects output. It is possible to
have a large multiplier even with Ricardian equivalence, and it is
possible to have a small multiplier even without Ricardian equivalence.
Thanks, Robert, especially for that last clarification. I have to
admit that I haven't been giving much thought to the scenario in which a
deficit-financed stimulus has no effect (under Ricardian equivalence)
on aggregate demand and yet still has a big multiplier. I will have to
think about that one...