Effects of Terms of Trade Gains and Tariff Changes on the Measurement of U.S. Productivity Growth
description
Transcript of Effects of Terms of Trade Gains and Tariff Changes on the Measurement of U.S. Productivity Growth
Effects of Terms of Trade Gains and Tariff Changes on the Measurement of U.S. Productivity Growth
Rob Feenstra, UC DavisMarshall Reinsdorf, BEA Matt Slaughter, Dartmouth
2008 World Congress on National Accounts and Economic Performance Measures for Nations
Rosslyn, Virginia
May 13, 2008
www.bea.gov
Motivation for Paper▪ US productivity speedup of 1 percent per
year starting in 1995, with ITC as key driver.▪ Trading gains can look like productivity
gains.▪ Growth of imports may have price driver. ▪ Multilateral elimination of tariffs on ITC
goods over 1997-1999 under the WTO Information Technology Agreement.
▪ Improvement in non-petroleum terms of trade coincided with the productivity speedup.
2
www.bea.gov
Globalization of ITC IndustryTable 1: International Trade in ICT Industries ($million)
Industry 1992 1996 2000
Computers (End-use 21300) Exports 8,277 10,422 10,263 Imports 5,042 6,927 14,284 Trade Balance 3,235 3,495 -4,022
Computer accessories Exports 16,730 27,550 34,686 (End-use 21301) Imports 26,659 54,590 75,514 Trade Balance -9,929 -27,040 -40,828
Semiconductors Exports 11,527 24,135 45,118 (End-use 21320) Imports 15,477 36,713 48,341 Trade Balance -3,950 -12,579 -3,223
Telecommunications Exports 10,520 19,137 28,987 Equipment (End-use 214) Imports 10,773 14,505 38,203 Trade Balance -253 4,633 -9,216 Total (End-use 213+214) Exports 47,054 81,244 119,054 Imports 57,952 112,735 176,343
Trade Balance -10,898 -31,491 -57,289
Share in Overall Trade (percent) Exports 10.7 13.3 15.4 Imports 12.0 15.4 16.0 Trade Balance 24.1 26.6 17.2
www.bea.gov
US Nonpetroleum Terms of Trade, MFP and Detrended MFP (1995=100)
90
95
100
105
110
115
120
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
Nonpetroleum terms of trade
MFP of Private Business
Deviation of MFP from its post-1987 average
www.bea.gov 5
Trade in GDP
▪ In current dollars, GDP = D + X – M,
where D = C+I+G, “gross domestic purchases” or “domestic absorption.”
▪ Diewert and Morrison’s (1986) GDP function treats M as intermediate inputs and D and X as outputs.
▪ Can also think of GDP as consolidated Value Added for the whole economy in which the domestic intermediate inputs cancel out (but note that M is measured at tariff-free prices in calculating GDP.)
www.bea.gov 6
Terms of Trade Effects can Resemble Productivity Effects
Terms of Trade defined as PX/PM, where
PGDP sDPD + sXPx –
sMPM.
▪ Real income depends on production (GDP) and on gains from trade, which grow when terms of trade improve.
▪ Improved terms of trade raise nominal GDP and the real D attainable for a given current account balance, ceteris paribus.
www.bea.gov 7
Change in Terms of Trade from PP to PP Reduces Real Consumption from D to D but Shift in Production from A to A has no Effect on Real GDP
www.bea.gov
Theory of Productivity Measurement
with International TradeThe revenue function for the economy gives its gross value added at basic prices:
Rt(Pt, t, vt)
)v(Sy|m)1(pxpqp ttt
N
1i
C
1j
tij
tij
tmij
M
1i
N
1i
ti
txi
ti
ti0y
maxt
where:
Domestic final goods i = 1,…,M have quantities tiq > 0 and prices t
ip > 0
Exports i = 1,…,N have quantities tix > 0 and international prices t
xip > 0.
Imported intermediate inputs i = 1,…,N, in j = 1,…,C varieties indexed by country
Import quantities tijm > 0, international prices t
mijp > 0, domestic prices )1(p tij
tmij .
www.bea.gov
Tariff Revenue in GDP
GDP function differs from Rtby the amount of tariff revenue.
Gt(Pt, t, vt) = Rt(Pt, t, vt)
N
1i
C
1j
tij
tij
tmij mp ,
Proposition 1
Holding fixed Pt and vt, the value of GDP is maximized at t = 0.
www.bea.gov 10
A
P'1
P1
G* R G
U1
y2 (import)
0
C
Figure 4: Tariff on an Imported Final Consumption Good
y1 (export)
A
www.bea.gov
Theoretical Measures of Productivity
We define “true” productivity, as in Diewert and Morrison (1986):
)v,,P(R
)v,,P(RA
1t1t1t1t
1t1t1tt1t
, or )v,,P(R
)v,,P(RA
ttt1t
ttttt
.
These concepts of productivity change are not measurable because both the
numerator of At-1 and the denominator of At are unobservable. Yet their geometric mean
can be measured, once we assume a specific form for the revenue function.
www.bea.gov
Measure of True Productivity 2/1t1t AA =
1t
t
R
R / [PT( 1tP
~ , tP~ ,yt-1,yt) QT(vt-1, vt, wt-1, wt)] ,
where PT( 1tP~ , tP~ ,yt-1,yt) is a Törnqvist price index over final goods, exports and imports,
and QT(vt-1, vt, wt-1, wt)is a Törnqvist quantity index over primary factors.
The Törnqvist price index is defined as:
ln PT( 1tP~ , tP~ ,yt-1,yt)
M
1i1t
i
ti
t
ti
ti
1t
1ti
1ti
p
pln
R
qp
R
qp
2
1
N
1i1t
xi
txi
t
ti
txi
1t
1ti
1txi
p
pln
R
xp
R
xp
2
1
N
1i1t
mi
tmi
t
tij
tij
tmij
1t
1tij
1tij
1tmij
p~p~
lnR
m)1(p
R
m)1(p
2
1 C1j
C1j
.
www.bea.gov
Adjustment for Net Entry of Varieties from New Supplying
CountriesOur price indexes incorporate the effect of new (and also disappearing) countries that sell
good i using Feenstra (1994), which shows that the ratio of CES price aggregates is:
1t
mi
tmi
p~p~
= )1/(1
1tmi
tmit,1t
miG
i
P
,
where the terms tmi < 1 equal one minus the share of period t expenditure of “new”
suppling countries. As new supplying countries sell more, tmi falls.
We construct )/( 1tmh
tmh
for each HS product h, raising to the power 1/(h – 1);
then take the geometric mean across products iHh within an End-use industry i:
)1/(w
Hh
1tmh
tmh
tmi
htmhi
i
/
.
www.bea.gov
Conventional TFP – True Productivity
ln TFPt – 2/1t1t AA ln =
t
t
R
Gln
1t
1t
R
Gln
+
N
1i
t,1txiLt
ti
txi
1t
1ti
1txi
1txi
txi
t
ti
txi
1t
1ti
1txi Pln
G
xp
G
xp
2
1
p
pln
R
xp
R
xp
2
1
N
1i1t
mi
tmi
t
tij
tij
tmij
1t
1tij
1tij
1tmij
p~p~
lnR
m)1(p
R
m)1(p
2
1 C1j
C1j
t,1tmiLt
tij
tmij
1t
1tij
1tmij
N
1i
PlnG
mp
G
mp
2
1 C1j
C1j
.
Gap between conventional and true TFP comprises:
Difference between exact (Törnqvist) and Laspeyres formula for exports
Difference between exact and Laspeyres formula for imports
Correction for tariffs in the “true” import prices and weights
Correction for net growth in import variety in the “true” import prices
www.bea.gov
Estimation of Influence of Unmeasured
Gains from Trade on GDP/Productivity ▪ BEA aggregates Laspeyres indexes from BLS
to find deflators for exports/imports in GDP. ▪ We calculate alternative versions of the
indexes used by BEA to construct deflators.▪ To aggregate our indexes, we use weights
from the NIPAs and Fisher index formula. ▪ Removing nonmarket sectors of GDP leaves
Value Added of Private Business, the output part of aggregate productivity calculation.
15
www.bea.gov
Three Kinds of Corrections Tested
16
True detailed index = (linked index)(tt-1)1/( 1)
t = 1 – share of new supplying countries in period t
▪ Törnqvist rather than Laspeyres formula for deflators of trade components of GDP.
▪ Tariff-inclusive weights and index for M.
▪ Adjust PM for net entry of new countries supplying imports using CES formula.
www.bea.gov
If the improvement in the terms of trade is understated, then productivity is overstated.
Sources of mismeasurement: see Figure 2.
Figure 2: Alternative Terms of Trade Indexes
95
100
105
110
115
120
1989
01
1989
07
1990
01
1990
07
1991
01
1991
07
1992
01
1992
07
1993
01
1993
07
1994
01
1994
07
1995
01
1995
07
1996
01
1996
07
1997
01
1997
07
1998
01
1998
07
1999
01
1999
07
BLS Laspeyres Tornqvist w/ Tariff w/ Variety
Nonpetroleum terms of trade before taking account of BEA’s use
of Fisher
www.bea.gov
Figure 9: Import Prices, Semiconductors (21320)
40
50
60
70
80
90
100
BLS Laspeyres Tornqvist w/ Tariff w/ Variety
Example of Detailed Indexes: Semiconductor Imports
www.bea.gov
Figure 10: Export Prices, Semiconductors (21320)
40
50
60
70
80
90
100
BLS Laspeyres Tornqvist
Example of Detailed Indexes: Semiconductor Exports
www.bea.gov
Telecom Equipment Imports
Figure 11: Import Prices, Telecommunications (21400)
75
80
85
90
95
100
BLS Laspeyres Tornqvist w/ Tariff w/ Variety
www.bea.gov
Telecom Equipment Exports
Figure 12: Export Prices, Telecommunications (21400)
80
85
90
95
100
105
BLS Laspeyres Tornqvist
www.bea.gov
Effects on Price Indexes (Averages for 1996-1999)
Corrections Revision to
Exports Index Revision to
Imports Index Revision to Terms of Trade
Törnqvist formula 0.4 %/year 0.6 %/year 0.2 %/year
Add: Tariff-inclusive prices 0.7 %/year 0.3 %/year
Add: New import suppliers 1.5 %/year 1.1 %/year
www.bea.gov 23
Table 6: Effect of Lower-level Index Formula on Aggregate Fisher Price Indexes for Exports and Imports
1994 1995 1996 1997 1998 1999 Ave
96-99
I. Export Indexes Formula for lower-level aggregates Laspeyres 100 104.1 104.1 102.5 99.4 98.2 101.1 Törnqvist 100 103.5 103.4 101.5 97.7 96.1 99.7 Growth rate difference of Törnqvist from Laspeyres -0.58 -0.16 -0.33 -0.66 -0.43 -0.40
Contribution of ICT goods to above difference -0.10 -0.18 -0.23 -0.41 -0.27 -0.27 II. Import Indexes Formula for lower-level aggregates Laspeyres 100 103.2 101.6 99.0 95.1 93.5 97.3 Törnqvist 100 102.8 100.9 97.6 93.2 91.0 95.7 Törnqvist, tariffs included in prices 100 102.6 100.5 97.3 92.8 90.4 95.3 Törnqvist, adjusted for tariffs and varieties 100 102.6 99.8 95.7 90.3 87.5 93.3 Growth rate difference from Laspeyres Törnqvist -0.35 -0.37 -0.62 -0.65 -0.72 -0.59 Törnqvist, tariffs included in prices -0.55 -0.48 -0.60 -0.82 -0.87 -0.69 Törnqvist, adjusted for tariffs and varieties -0.60 -1.14 -1.54 -1.77 -1.41 -1.47 Contribution of ICT goods to above difference -0.01 -0.33 -0.78 -0.59 -0.56 -0.57
www.bea.gov
Effects on Measures of GDP and Productivity
▪ Reduction in real GDP growth is about 0.1 percent per year after 1995 (allowing for lack of effect of capital goods & tariffs.)
▪ Growth rate of price index for private business value added increased by 0.2%/year.
▪ Growth of productivity reduced by almost 0.2 %/year, a fifth of the productivity speedup.
www.bea.gov 25
Price and Volume Indexes for GDP and Value Added of Private
Business 1994 1995 1996 1997 1998 1999
Price index for GDP Official (rebased to 1994=100) 100 102.0 104.0 105.7 106.9 108.4 After adjustment for Törnqvist formula, tariffs & variety 100 102.1 104.1 106.0 107.3 108.9 Growth rate difference 0.01 0.10 0.13 0.13 0.12 Price Index for Private Business Value Added Official Laspeyres 100 101.8 103.4 104.9 105.5 106.5 Törnqvist formula without tariffs 100 101.7 103.4 104.9 105.6 106.7 Törnqvist formula with tariffs in prices and weights 100 101.8 103.4 105.0 105.7 106.8 Törnqvist formula, with tariffs and variety adjustment 100 101.8 103.5 105.3 106.1 107.3 Difference from Laspeyres in annual growth rate Törnqvist -0.01 0.03 0.05 0.02 0.06 Törnqvist, tariffs included in prices 0.01 0.07 0.06 0.06 0.09 Törnqvist formula, with tariffs and variety adjustment 0.02 0.16 0.19 0.19 0.16 Implied difference in growth rate of real output 0.01 -0.13 -0.20 -0.18 -0.15
www.bea.gov
End