Loading presentation…
Opening in Keynote for Mac...
Your browser isn’t fully supported.
For the best Keynote for iCloud experience, use a supported browser.
Learn More
Cancel
Continue
TO:
Interested parties
FROM:
Brad DeLong
SUBJECT:
FISCAL DRAG FROM MORE DEBT-FINANCED SPENDING
TAKEAWAY:
RIGHT NOW THERE IS NONE
DATE:
Mo 2020-11-17
You asked me to think about long-run downside via fi scal drag and higher required tax rates and
revenues in the future, after the economy has returned to full employment, from additional debt-
fi nanced COVID depression-fi ghting stimulus expenditures. You asked me to think in the context
of Larry Summers’s and my “Fiscal Policy in a Depressed Economy” of a decade ago.
My conclusion:
RIGHT NOW THERE IS NO PROSPECT OF ANY FUTURE FISCAL
DRAG FROM ADDITIONAL DEBT-FINANCED FISCAL STIMULUS.
Let
u
s call the short-run fi scal
-expansion
multiplier
μ
, let's call the marginal tax rate
τ
, let's call
the proportional
‘
hysteresis shadow
’
cast on future potential output by depressed production
today
η
, let's call the marginal real long-term interest rate on Treasury debt
r
, let's call the
average growth rate of production per worker
g
, and let's call the growth rate of the labor force
n
.
Then debt-fi nanced expansionary fi scal policy now imposes no fi scal drag but instead provides a
fi scal boost to the economy in the future as long as:
The left-hand side is the permanent boost to revenue from higher long-run production, as a result
of the smaller shadow cast by a reduced current depression, via stimulus funded by a dollar of
net debt. The right
-
hand side is the cost of fi nancing an extra dollar of debt along a path that
stabilizes the debt-to-national-income ratio.
Right now, in per-year units:
•
μ
is, with the economy depressed but with expansion in some important labor-intensive
industries impossible, perhaps 1.5
•
τ
is roughly 1/3.
That means we can simplify our equation to say that there is no long-run fi scal cost to the
government from marginal debt-fi nanced stimulus as long as:
η >
r
–
g
–
n
Right now, in per-year terms:
•
r
from the 30-year TIPS rate is -0.2
%
•
n
is, with reasonable immigration policy, roughly 1
%
•
g
is unknown, but unlikely to be less than 1
%
This tells us that our hurdle is:
η >
–
2.2
%
For debt-fi nanced stimulus now to
generate
zero
future
fi scal
drag
, a dollar of debt-fi nanced
2020-11-17: DeLong Debt Memo
2020-11-17
<
https://www.icloud.com/keynote/0UJ2pg4UkvegNtUljgTL-Qhdg
> <
https://github.com/braddelong/public-fi les/blob/master/bernstein-delong-debt-memo-2020-11-17.pptx
>
stimulus now does not need to reduc
e
the long-run shadow cast by a depression
and so
increase
future
full-employment
production
at all
. It only needs to not
reduce
future production when the
economy gets back to full employment by more than
–
2.2 cents.
If the 'hysteresis-shadow' parameter η is in fact our decade-ago base case
of
0.2, then our
condition for
zero
fi scal
burden is
for the 30-year TIPS rate
to remain at
:
r
< 22.2
%
This is so far away from
today’s –0.2
%
30-year TIPS rate
that it
is senseless to
worry now about
the consequences of fi scal drag from measures that increase the national debt.
Now there are debt fi nancing issues: We surely do not want the Treasury to fi nance anything
additional by selling more 30-year TIPS exclusively. We equally surely do not want the Treasury
to sell only more 3-month T-Bills. Debt management, and the interaction of debt management
with safety-and-soundness regulation—i.e., how many Treasury securities and reserve deposits at
the Fed we require leveraged fi nancial institutions to hold as reserves rather than letting them get
away with dodgy things that someone has rated AAA—
are
important and complex
issues
. And
the
‘
r
’
in the equations above needs to be grossed up
,
by taking account of how additional debt
may raise the rate at which existing debt is refi nanced when it is rolled over.
The high-interest rate era of the 1980s and 1990s in which we worried about fi scal drag from
expanded national debt is now long gone. Right now we are in what Larry calls “secular
stagnation”, in which Treasury rates are very low, markets expect Treasury rates to remain very
low for decades, and the term structure gives the U.S. government power to lock in very low
interest rates on its debt for the long term.
In some ways the right way to think about government debt is not as a cost but as a profi t center
for the government—like the Medici Bank of the Late Middle Ages, savers are willing to pay the
U.S. government to accept their money because they believe the U.S. government can keep it
safe. And it is in that context that we should think about debt fi nance.
====
<
https://github.com/braddelong/private-fi les/blob/master/bernstein-delong-debt-memo-2020-11-17.pdf
>
<
https://
www.icloud.com/pages/0cyGMZiTxl67Uw4VlubYkHr4Q
>
<
https://www.icloud.com/keynote/0UJ2pg4UkvegNtUljgTL-Qhdg
>
<
https://github.com/braddelong/public-fi les/blob/
master/bernstein-delong-debt-memo-2020-11-17.pptx
>
The Not-
Quite-
So-Simple Arithmetic of Fiscal Policy in a Depressed Economy
<
https://www.bradford-delong.com/
2012/03/lunch-march-14-2012-the-not-so-simple-arithmetic-of-fi scal-policy-in-a-depressed-economy.html?
asset_id=6a00e551f080038834016763ce31e3970b
> 2012-03-14
Expansionary Fiscal Policy in a Depressed Economy
<
https://delong.typepad.com/fi les/20120417-econ-191-fi scal-
policy-in-a-depressed-economy.ppt.pdf
> 2012-04-17
Fiscal Policy in a Depressed Economy
<
https://www.bradford-delong.com/2012/04/econ-191-april-17-2011-fi scal-
policy-in-a-depressed-economy.html
> 2012-04-17
Fiscal Policy in a Depressed Economy: Further Thoughts
<
https://www.bradford-delong.com/2013/04/fi scal-policy-
in-a-depressed-economy-further-thoughts.html
> <
https://delong.typepad.com/20130408-delong-pres--fi scal-policy-in-a-
depressed-economy-further-thoughts.pdf
> <
https://delong.typepad.com/20130408-delong-pres--fi scal-policy-in-a-
depressed-economy-further-thoughts.
ppt
> 2013-04-07
Full recovery (well, to 2007, not 2000) in prime-age
employment; ZERO recovery in production or income
relative to the 2000-2007 trend…
