Twin Deficits and Twin
Decades
Jeffrey Frankel
Harpel Professor of Capital Formation and Growth
Kennedy School of Government
Harvard University
conference on the Macroeconomics of Fiscal Policy
sponsored by the Federal Reserve Bank of Boston,
at Wequasett, June 14-16, 2004
Three propositions were put to the
test by fiscal expansion of 1980s
Proposition Prediction Outcome in
1980s
Barro’s debt BD would be offset by increased Private & National
neutrality private saving, so National Saving rates fell
Saving would not fall.
Mundell- The fall in national saving rate As predicted.
Fleming: high would crowd out TB, more than
capital crowding out investment The Twin Deficits
mobility & (via rise in real interest rate and $) are born.
floating rates
Feldstein- Fall in NS => fall in investment, Saving-retention
Horioka big deficits • “Starve the Beast” claim rings
(despite claims to long-run hollow national saving
• incl. the Laffer hypothesis: falls. (Figure 2 or 2b)
“US tax rate cuts => growth up • Economists debate if deficits
=> tax receipts up” affect interest rates
• optimistic forecasts soon • Fall in national saving reflected
shattered, although in current account deficit.
Administration still blames (Figure 3) Twin deficits.
deficits on recession and • (Admittedly, cyclical factors
claims they will go away responsible for exaggerating
soon (Figure 5) decline in budget, NS, & I
• failure of budget outcomes to at the beginning of decade, but
follow script => switch to not for overall pattern.)
“Starve the Beast” claim
-7
-6
-5
-4
-3
-2
-1
0
1
2
3
J. Carter 1977
1978
1979
1980
R. Reagan 1981
1982
1983
1984
1985
1986
1987
1988
G.H.W. Bush 1989
1990
1991
1992
W.J. Clinton 1993
1994
by Presidential Term
1995
1996
1997
1998
1999
2000
G.W. Bush 2001
2002
2003est
Fig.1: US Federal Deficit as Share of GDP
Fig. 4: Three years of budget forecasts
that soon proved too optimistic
400
300
200
100
US$ bn
0
-100
-200
-300
-400
-500
Jan.
Aug. Jan.
2001 Aug.
2001 2002 Jan.
2002 Aug.
2003 Jan.
2003
2004
2002 2003 2004
Fig. 5: As of 2004, official budget
forecasts are still too optimistic
Source:
Alice Rivlin &
Isabel
Sawhill,
Brookings,
1/13/2004
18
19
20
21
22
23
24
1977 J. Carter
1978
1979
1980
1981 R. Reagan
1982
1983
1984
1985
1986
1987
1988
1989 G.H.W. Bush
1990
1991
1992
1993 W.J. Clinton
1994
1995
1996
1997
1998
1999
2000
G.W. Bush
Fig. 6: US Federal Spending
2001
as a Share of GDP, by Presidential Term
2002
2003est
0.0%
2.0%
4.0%
6.0%
8.0%
10.0%
12.0%
14.0%
-8.0%
-6.0%
-4.0%
-2.0%
1980-I
R. Regan 1980-III
1981-I
1981-III
1982-I
1982-III
1983-I
1983-III
1984-I
1984-III
1985-I
1985-III
1986-I
1986-III
1987-I
Net Private Saving
1987-III
1988-I
G.H.W. Bush 1988-III
1989-I
1989-III
1990-I
1990-III
1991-I
1991-III
1992-I
W.J. Clinton
1992-III
1993-I
1993-III
1994-I
Net Government Saving
1994-III
1995-I
1995-III
1996-I
1996-III
1997-I
1997-III
1998-I
1998-III
1999-I
1999-III
2000-I
G.W. Bush
2000-III
Net National Saving
2001-I
2001-III
2002-I
Saving Falls (shares of GDP 1980-2002)
Fig. 2b: Private Saving Fails to Rise to
2002-III
Offset Budget Deficits; Rather National
2003-I
2003-III
Fig. 2: Budget Balances, Private Saving, &
National Saving, as Shares of GDP
0.140
0.120
0.100
0.080
0.060
0.040
0.020
0.000
1961
1963
1965
1967
1969
1971
1973
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
-0.020
-0.040
-0.060
Net Private Saving (% of GDP) Net Govt Saving (% of GDP) Net State & Local Saving (% of GDP)
Net Federal Saving (% of GDP) Net Natl Saving (% of GDP)
Fig. 3: National Saving, Investment, %
Current Account, as Shares of GDP
0.140
0.120
0.100
0.080
0.060
0.040
0.020
0.000
1961
1963
1965
1967
1969
1971
1973
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
-0.020
-0.040
-0.060
Net Natl Saving (% of GDP) Net Domestic Investment (% of GDP) CA (% of GDP)
Table 2: Feldstein-Horioka
Regressions for US -- 1964-2003
• Dependent Variable:
Current Account / GDP, Cyclically Adjusted
• Coefficient is about 1/2
• Some evidence of upward trend in capital mobility
National Savings / GDP, .5027 ** .3554 **
Cyclically Adjusted (.0332) (.0803)
(CyclAdjNS2)
Trend in Coefficient on .0018 *
Saving (CyclAdjNS2 * T) (.0009)
Constant .0018
(.0012)
R-sqd. .59 .60
Differences between the two decades
• Differences between early 00s and early 80s:
– Interest rates have been low, 2001-2004
– The dollar began to depreciate in 2003
• Not surprising:
– The Fed has been very accommodating, so far
– Asian central banks, too, have bought US Treasury
securities
– People are apparently not yet fully aware of the likely
magnitude of coming deficits.
• Likely to change: i trend will be up
Interest Rates & Expected Budget Deficits
Source: Menzie Chinn and Jef f rey Frankel, October 15, 2003
OLS regression using annual data, in levels (Newey-West robust standard errors in parentheses).
% variables def ined in decimal f orm. *(**)[***] denotes signif icance at the 10%(5%)[1%] level.
US Germany France Italy Spain UK
Constant -0.001 -0.122*** -0.022 -0.081 -0.043* -0.034
(0.008) (0.038) (0.027) (0.041) (0.023) (0.030)
Inflation 1.00 1.00 1.00 1.00 1.00 1.00
debt ratio 0.060** 0.182*** 0.027 0.109 0.031 0.067
(0.019) (0.047) (0.040) (0.062) (0.051) (0.044)
expected
0.289**
change 0.144** 0.112*** 0.177** 0.324** * 0.066
in debt ratio (0.061) (0.032) (0.073) (0.106) (0.048) (0.110)
output gap 0.388** 0.608** 0.252 0.297 0.218 -0.316
(0.174) (0.219) (0.202) (0.484) (0.223) (0.324)
Foreign
0.923** 1.204**
interest rate 0.096 1.529*** * 0.390 * 0.815**
(0.122) (0.327) (0.241) (0.446) (0.145) (0.348)
N 15 15 15 15 15 15
Adj.R2 0.32 0.51 0.82 0.77 0.82 0.55
DW 2.24 2.50 2.47 1.70 2.47 1.44