Transcript The U. S. Productivity .Growth "Explosion"

The U. S. Productivity
Growth “Explosion”:
Dimensions, Causes,
Consequences, Aftermath
Robert J. Gordon,
Northwestern University and NBER,
NABE 48th Annual Meeting,
NBER Session,
Copley Place Marriott, Boston, September 11,
2006
Brief Survey of NBER
Research on Macro
Productivity Issues




Research Papers on my web site, google
“Robert J. Gordon”
Topic #1 on quarterly productivity behavior,
today’s unpublished update of BPEA 2003
Topic #2 on inflation, unpublished update of
BPEA 2005 with added dire implications for
Bernanke
Topic #3 on Europe, new paper on web site
(unfortunately no time to talk about this)
Broad-ranging
Interpretation of U. S. 200204 Productivity “Explosion”




#1. Causes, Were they Temporary,
Dimensions of Future Slowdown
#2. Effects of Productivity Growth on Core
Inflation and the Fed’s Dilemma
For this NABE Audience the quarterly
analysis of U. S. data will be emphasized
Lots of charts, new analysis here done since
early August BEA and BLS data releases
Topic #1: Behavior of
Productivity Growth in
Quarterly Data




Important to understand the dynamics
They have nothing to do with the
NBER business cycle chronology
The behavior of productivity is driven
by the lag of hours behind output
This was a topic of the early 1960s,
Okun’s Law and Walter Oi on labor as
a “quasi-fixed factor”
2006:02
2005:01
2003:04
2002:03
2001:02
2000:01
1998:04
1997:03
1996:02
1995:01
1993:04
1992:03
1991:02
1990:01
1988:04
1987:03
1986:02
1985:01
1983:04
1982:03
1981:02
1980:01
1978:04
1977:03
1976:02
1975:01
1973:04
1972:03
1971:02
1970:01
1968:04
1967:03
1966:02
1965:01
1963:04
1962:03
1961:02
1960:01
1958:04
1957:03
1956:02
1955:01
8-quarter Change in NFPB
Output and Hours, 19552006
10
8
Output NFPB
6
4
2
0
-2
-4
Hours
Key Implications of Lag
in Hours Behind Output




Productivity Growth is not
Synchronized with the utilization of
resources
Because hours lags, productivity leads
Productivity Growth is fastest at the
beginning of the recovery
The “early recovery productivity
bubble”
2006:02
2005:01
2003:04
2002:03
2001:02
2000:01
1998:04
1997:03
1996:02
1995:01
1993:04
1992:03
1991:02
1990:01
1988:04
1987:03
1986:02
1985:01
1983:04
1982:03
1981:02
1980:01
1978:04
1977:03
1976:02
1975:01
1973:04
1972:03
1971:02
1970:01
1968:04
1967:03
1966:02
1965:01
0
1963:04
1962:03
1961:02
1960:01
1958:04
1957:03
1956:02
1955:01
Notice the
“Early Recovery Bubble”,
8-qtr changes 1955-2006
10
8
Output NFPB
6
4
2
Output per hour
-2
-4
Trend Methodology


Two Leading Methods of Detrending
Hodrick-Prescott Filter
– Normal Paramater of 1600 bends too
much and allows too much of the “cycle”
to get into the trend.
– Everything here uses HP 6400

Kalman Filter. Allows correcting for
changes in the business cycle
Alternative U. S. NFPB
Productivity Trends in
Quarterly Data, 1954-2006
3.5
H-P 6,400
3
Kalman filter with gap term
2.5
2
Kalman filter without gap term
1.5
1
0.5
0
1954:01 1956:04 1959:03 1962:02 1965:01 1967:04 1970:03 1973:02 1976:01 1978:04 1981:03 1984:02 1987:01 1989:04 1992:03 1995:02 1998:01 2000:04 2003:03
8-quarter Actual LP
vs. the Average Trend
Percent
6
5
Actual LP
4
3
2
1
Trend LP
0
-1
-2
1955:01 1959:03 1964:01 1968:03 1973:01 1977:03 1982:01 1986:03 1991:01 1995:03 2000:01 2004:03
The Early Recovery Bubble,
How Much “Payback” is
Left?

2001:3-2004:2, 11 quarter AAGR
– Actual 3.87
– Trend 2.92
– Difference 0.95, or cumulatively 2.62

2004:2-2006:2, 8 quarter average
– Actual 2.00
– Trend 2.68
– Difference -0.68, or cumulatively -1.37
What Implications for
2006:Q2-2008:Q2?




Start by Assuming that the trend slows
a bit further from 2.68 to 2.50
Remaining “payback” of 2.62 (01-04)
minus 1.37 (04-06) equals 1.25
Distributing that over next two years
implies actual AAGR = 1.88
Anything below that would imply the
trend is lower than 2.50
The “Output Identity”
Organizational Tool for
Trends, Cycles, and
Residuals

In its Simplest Form Makes Output (Q) Equal to the product
of:
–
–
–
–
–

Productivity (Q/A)
Hours per Employee (A/E)
Employment Rate (E/L), that’s just (1 – U/L)
Labor-force Participation Rate (L/N)
Working-age Population (N)
Hiding Inside the Output Identity are Numerous Useful Trend
and Cyclical Relationships, including OKUN’s Law.
Five-term Output Identity Cannot
be Used for Empirical Analysis


Productivity data for the NFPB sector
Expand the identity to identify NFPB variables and links to total
economy:
Q B AB E L
Q/ EP EP
Q  B  B .  N B B 
A E L N
Q /E E


Mix effect – ratio of output per employee: total/NFPB sector
Employment ratio of payroll to household
The Novelty here
is to Display the Seven
Components



We’ll look through each of them,
plotting actuals (8-qtr MAs) vs. trends
We’ll pay special attention to what has
happened to each over the past six
years
Then we’ll multiply them together to
see what has happened to potential
real GDP growth
Actual vs. Trend Growth
for Hours per Employee
Percent
1.5
1
Actual AE
0.5
0
-0.5
Trend AE
-1
-1.5
-2
1955:01 1959:03 1964:01 1968:03 1973:01 1977:03 1982:01 1986:03 1991:01 1995:03 2000:01 2004:03
Actual vs. Trend Growth
for Labor Force Participation
Percent
2
1.5
Actual LN
1
0.5
0
Trend LN
-0.5
-1
-1.5
1955:01 1959:03 1964:01 1968:03 1973:01 1977:03 1982:01 1986:03 1991:01 1995:03 2000:01 2004:03
Actual vs. Trend Growth
for the Employment rate
2
1.5
Actual E/L
1
0.5
Trend E/L
0
-0.5
-1
-1.5
-2
-2.5
1955:01 1957:03 1960:01 1962:03 1965:01 1967:03 1970:01 1972:03 1975:01 1977:03 1980:01 1982:03 1985:01 1987:03 1990:01 1992:03 1995:01 1997:03 2000:01 2002:03 2005:01
Actual vs. Trend Growth
for Wkg-Age Population
Percent
2.5
Actual N
2
1.5
Trend N
1
0.5
0
1955:01 1959:03 1964:01 1968:03 1973:01 1977:03 1982:01 1986:03 1991:01 1995:03 2000:01 2004:03
Actual vs. Trend Growth
for the “Mix Effect”
Percent
1.5
Actual Mix
1
0.5
0
Trend Mix
-0.5
-1
-1.5
-2
1955:01 1959:03 1964:01 1968:03 1973:01 1977:03 1982:01 1986:03 1991:01 1995:03 2000:01 2004:03
Actual vs. Trend Growth
for Payroll vs. Household
Employment
Percent
2.5
2
Actual EPE
1.5
1
0.5
0
Trend EPE
-0.5
-1
-1.5
1955:01 1959:03 1964:01 1968:03 1973:01 1977:03 1982:01 1986:03 1991:01 1995:03 2000:01 2004:03
Two Measures of Trend
Potential GDP Growth
8
Actual Output
7
6
5
4
3
Trend Output
2
1
Alternative Trend Output
0
-1
-2
1955:01 1957:03 1960:01 1962:03 1965:01 1967:03 1970:01 1972:03 1975:01 1977:03 1980:01 1982:03 1985:01 1987:03 1990:01 1992:03 1995:01 1997:03 2000:01 2002:03 2005:01
Potential GDP vs.
Productivity: the Trend
Story in Table 2

Potential GDP growth (Δq*) ranged from:
– 4.03 in 1963-72 to 2.69 in 1987-94
– Differences accounted for by



Productivity (peak 1954-63)
Population growth (peak 1972-78)
LFPR (peak 1972-78)
– Offset by hours/employee (peak 1963-72)

Currently growth rate is 2.9 percent by one
measure and 3.0 percent by the other
Log Ratio of Actual to
Potential Real GDP
Percent
8
6
4
2
0
-2
-4
-6
-8
-10
1955:01 1957:02 1959:03 1961:04 1964:01 1966:02 1968:03 1970:04 1973:01 1975:02 1977:03 1979:04 1982:01 1984:02 1986:03 1988:04 1991:01 1993:02 1995:03 1997:04 2000:01 2002:02 2004:03
Okun’s Law Updated: What
Happens with Changes in
Ratio of Actual to Potential?






When the ratio of actual to potential real
GDP rises by 1 percent, the following
happens (after allowing for lags)
The unemployment rate falls by 0.50
Productivity growth rises by 0.16
Hours per employee rise by 0.10
LFPR rises by 0.10
Residual distributed across other factors
Why Did Productivity
Growth Accelerate Further,
2002-04?

Hypotheses in 2003 BPEA paper
– Savage corporate cost cutting (profits hardly fell
in 1990-91 but fell by half in 2000-02)
– Intangible capital

Implications of Industry Decomposition
– Jorgenson and Stiroh: IT no role after 2000
– Sichel’s new numbers show IT no more
important in 2000-04 than in pre-1995

Added Element from Jorgenson-Stiroh,
forthcoming slowdown in “labor quality”
Implications for
Interpretation

What does it mean that no special role of IT
use in 00-04 or 01-05 acceleration?
– Stiroh’s interpretation: a broad cross-the-board
upsurge in TFP growth unrelated to IT
investment
– My 2003 BPEA interpretation, unusual pressure
for corporate cost-cutting due to late 1990s
bubble, overshooting, accounting scandals



Stiroh: It’s permanent, but we don’t know
why
Me: Big but temporary adjustment in
corporate organization and cost structure.
Trend is headed from 2.6 now to 2.25.
Potential GDP headed from 2.9 to 2.6
Effects of Productivity
Growth on Inflation



Details of the inflation model and its
treatment of productivity are in BPEA
2005, no. 2 (and on my web site)
“Where Did the Productivity Growth
Go? Inflation Dynamics and the
Distribution of Income”
The 2005 BPEA paper was co-authored
with Ian Dew-Becker
“Triangle” Model
of Inflation



Developed in late 1970s, intact since
1980
Two sides of the triangle are demand
and supply
The base of the triangle is inertia
How it Works: Explains
Headline PCE Deflator


Demand enters through the
unemployment gap (TV-NAIRU)
Supply shocks (changes relative to
zero)
– Food-energy effect
– Relative price of imports
– Change in trend productivity growth

Inertia: allow 24 quarters to enter
How the Model has
Changed since 1980




Before 1995, assumed the NAIRU was fixed
at 6.0 percent
Actuals fell below predictions in 1994-95
Adopted Stock-Watson technique of
estimating a NAIRU that varied over time
The technique simultaneously estimates the
inflation equation coefficients and the TVNAIRU
The TV-NAIRU with and
without Supply Shocks
12
10
TV-NAIRU w ithout Supply Shocks
Actual
8
6
4
TV-NAIRU w ith Supply Shocks
2
0
1962 1965 1967 1970 1972 1975 1977 1980 1982 1985 1987 1990 1992 1995 1997 2000 2002 2005
The Model Also Produces


Post-Sample Dynamic Simulations
Have Coefficients Changed?
– Instead of 1962-2006, estimate only for
1962-1996
– Lagged inertia effect for 1996-2006 then
generated endogenously
– This is the key technique to reveal
changes in coefficients, or “drift”
Here are the Three Supply
Shocks

Food and Energy Effect
– This is simply headline PCE inflation
minus core PCE inflation



Change in relative price of imports
Change in productivity trend growth,
from the research summarized earlier
A Consistent Theme:
Greenspan’s Gifts!
2016
2014
2012
2010
2008
2006
2004
2002
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
1980
1978
1976
1974
1972
1970
1968
1966
1964
1962
1960
Food-Energy Effect,
1960-2016
4
3
2
1
0
-1
FAE
-2
-3
Change in Relative Import
Price, 1960-2006
20
15
10
5
0
-5
IMP
-10
-15
1
9
17
25
33
41
49
57
65
73
81
89
97 105 113 121 129 137 145 153 161 169 177 185 193 201 209 217 225
2016
2014
2012
2010
2008
2006
2004
2002
2000
1998
1996
1994
1992
1990
1988
1986
1984
1982
1980
1978
1976
1974
1972
1970
1968
1966
1964
1962
1960
Change in the Productivity
Growth Trend, 1960-2016
0.6
0.5
0.4
HPDEV
0.3
0.2
0.1
0
-0.1
-0.2
-0.3
-0.4
-0.5
2016:02
2015:03
2014:04
2014:01
2013:02
2012:03
2011:04
2011:01
2010:02
2009:03
2008:04
2008:01
2007:02
2006:03
2005:04
2005:01
2004:02
2003:03
2002:04
2002:01
2001:02
2000:03
1999:04
1999:01
1998:02
1997:03
1996:04
1996:01
The Effect of Future
Changes in Prod’y Growth
4
Productivity dow n to 2.2
3.5
3
Predicted
2.5
Productivity up to 3.0
2
1.5
1
0.5
Actual
0
20
00
20
00
.8
20
01
.5
20
02
.3
20
03
20
03
.8
20
04
.5
20
05
.3
20
06
20
06
.8
20
07
.5
20
08
.3
20
09
20
09
.8
20
10
.5
20
11
.3
20
12
20
12
.8
20
13
.5
20
14
.3
20
15
20
15
.8
The Dilemma that
Bernanke has Inherited
3.5
3
2.5
2
Inflation Lag
1.5
Unemployment Deviation
1
FAE
0.5
HPDEV
0
-0.5
IMP
-1
With a “Neutral” Set of SS,
What is Need to
Maintain 2.0% “Comfort
Zone”?
6.6
UB
UC
UD
U
E
6.4
6.2
6
5.8
5.6
UA
5.4
5.2
5
40
38
36
34
32
30
28
26
24
22
20
18
16
14
12
10
8
6
4
2
0
-2
-4
-6
-8
-1
0
-1
2
-1
4
-1
6
-1
8
-2
0
4.8
3.5
UA
3
UB
2.5
UC
UD
2
UE
1.5
-20
-17
-14
-11
-8
-5
-2
1
4
7
10
13
16
19
22
25
28
31
34
37
40
Conclusion about
Bernanke’s Dilemma

Due to long inertia lags, nobody has
noticed
– Oil prices really do get into core inflation
– So do rising relative import prices if/when
dollar falls
– So does the productivity turnaround

Bernanke is now stuck with ~3.0 not
2.0 core inflation
There are only
Two Choices



The first choice is to create a recession with
several years of unemployment above 6
percent
The second choice is to abandon any
pretense of a 2.0 (or even 2.5) core inflation
target
NYT Jackson Hole coverage suggested the
second option has already been chosen but
they won’t say so (of course)
No Time for Topic #3,
Europe’s Turnaround




I’ll just leave you with two enticing graphs
The details are on my web site, look for the
paper about “Tigers and Tortoises” (also coauthored with Ian Dew-Becker).
Post-1995 reversal in EU-US productivity
growth
Post-1995 reversal in EU-US hours growth
Here’s Productivity
Growth, 1981-2004
3.5
3.0
U.S. Output per Hour
E.U. Output Per Hour
Percent
2.5
2.0
1.5
1.0
0.5
0.0
1981
1986
1991
1996
2001
The Stunning Turnaround
in Hours per Capita
4
US Output per Capita
3
EU-15 Output per Capita
2
1
US Hours per Capita
0
-1
EU-15 Hours per Capita
-2
-3
1981
1986
1991
1996
2001