Transcript Principles of Economic Growth

Ownership
and Growth
Thorvaldur Gylfason
Presentation in Two Parts
1.
2.
General discussion of economic
growth and the analytical
background of the paper
Specific discussion of the topic at
hand: public vs. private
ownership and the difference it
makes for growth
 Joint work with Tryggvi Thor
Herbertsson and Gylfi Zoega
Economic Growth
As old as economics itself
Adam Smith, John Stuart Mill, Alfred
Marshall, Joseph Schumpeter were all
growth theorists
Then Harrod and Domar came along,
summarizing the old literature by
g = sE - 
where E = Y/K.
This means, inter alia, that saving is
good for growth.
National economic output
Economic Growth:
The Short Run vs. the Long Run
Economic growth
in the long run
Potential output
Actual output
Upswing
Downswing
Business cycles
in the short run
Thus, increased saving may
reduce actual output, while
increasing potential output.
Time
Economic Growth:
The Short Run vs. the Long Run
To analyze the movements of actual
output from year to year, viz., in the
short run
Need short-run macroeconomic theory
Keynesian or neoclassical
To analyze the path of potential output
over long periods
Need modern theory of economic growth
Neoclassical or endogenous
Economic Growth:
Solow’s Rebellion
Solow rebelled against Harrod and Domar
by pointing out that economic growth in
the long run must be exogenous.
In the long run, growth depends solely on
population growth and technological
progress: g = n + q
To show this, Solow made E = Y/K
endogenous.
Hence, saving cannot affect long-run growth.
Why not?
Take another look at
g = sE - 
When s increases, E begins to decrease,
according to Solow, and continues to do
so until g is restored to its intial,
exogenously determined value.
This occurs essentially because an
increase in s increases K, so that E =
Y/K goes down.
Economic Growth:
The Solow Model
But this approach distracted attention
from two key questions:
1. How long is the long run?
–
Is it possible that the long run is so long
as to be almost irrelevant from the point
of view of economic policy?
2. What determines output per head?
–
Is is possible that the level of output per
head is endogenous in the long run
even if its rate of growth is not?
Economic Growth:
The Solow Model
In long-run
steady-state
equilibrium at A,
growth is
exogenous:
y = Y/L
A
g=n+q
E
k = K/L
Economic Growth:
The Solow Model
y  Ak
The parameter
A represents
technology and
efficiency.
1 a
n  q 
y
k
s

s
y  A 
 n  q 
1
a



1 a
a

dy
0
dA
The Neoclassical Theory of
Exogenous Economic Growth
Traces the rate of growth of output
per capita to a single source:
Technological progress
Hence, economic growth in the
long run is immune to economic
policy, good or bad
“To change the rate of growth of real
output per head you have to change the
rate of technical progress.”
ROBERT SOLOW
The New Theory of
Endogenous Economic Growth
Traces the rate of growth of output per
capita to three main sources:
Saving
Efficiency
Depreciation
“The proximate causes of economic growth are
the effort to economize, the accumulation of
knowledge, and the accumulation of capital.”
ARTHUR LEWIS
A Simple Model of
Endogenous Growth
Four building blocks:
 S=I
Saving equals investment in equilibrium.

S = sY
Saving is proportional to income.

I = K + K
Investment involves addition to capital stock.

Y = EK
Output depends on quality and quantity of capital.
Endogenous Growth in the
Harrod-Domar Model
This version of the endogenous
growth model is simply a
restatement of the HarrodDomar model
where growth depends on:
A. the saving rate
B. the capital/output ratio
C. the depreciation rate
The Main Point about
Endogenous Growth
If it increases economic efficiency, it
is also good for growth.
This idea seems to go a long way
towards explaining per capita income
differentials of 30 or 60 across countries.
And even if long-run growth is
exogenous, the level of income is
endogenous.
This is why endogenous growth theory is
perhaps best viewed as an extension of
the Solow model of exogenous growth.
Specific Model
Y  AL K
a
Two equations in
two unknowns
1 a
Now, assume learning by doing:
A  E K / L 
a
It follows that
Y  EK
g  sE  
and that
q  g n
A Picture of the Model
g
45°
sE - 
A
q
n
Properties of the Model
dg
0
dE
s
E

n
g
+
+
-
0
q
+
+
-
-
Sources of Endogenous
Growth
Saving
Fits real world experience quite well.
No coincidence that, in East Asia, saving rates of 3040% of GDP went along with rapid economic
growth.
No coincidence either that many African economies
with saving rates around 10% of GDP have been
stagnant.
OECD countries: saving rates of about 20% of GDP
Important implication for economic policy:
Economic stability with low inflation and positive real
interest rates encourages saving, and thus is good
for growth.
Sources of Endogenous
Growth
Income
per capita
400
East Asia
300
200
OECD
100
Africa
1965
1990
Sources of Endogenous
Examples:
Growth
Depreciation
1. Oil shocks
2. Soviet collapse
The effect of depreciation on growth is related
to that of saving on growth.
Unprofitable investment in the past reduces the
quality of capital and thus makes it depreciate
more rapidly, necessitating more replacement
investment to make up for wear and tear.
The more national saving has to be set aside
for replacement investment, the less will be
available for the buildup of new capital.
Sources of Endogenous
Growth
Efficiency
Also fits real world experience quite well
Technical progress good for growth because it allows
us to squeeze more output from given inputs.
But that is exactly what increased efficiency is all
about!
Thus, technology is best viewed as an aspect of
general economic efficiency.
Important implication for economic policy:
Everything that increases economic efficiency, no
matter what, is also good for growth.
Sources of Endogenous
Growth
Five sources of increased efficiency
1. Liberalization of prices and trade increases
efficiency, and thus is good for growth.
2. Stabilization reduces the inefficiency associated
with inflation, and thus is good for growth.
3. Privatization reduces the inefficiency associated
with state-owned enterprises, and thus …
4. Education makes the labor force more efficient.
5. Technological progress also enhances efficiency.
The possibilities are virtually endless!
Sources of Endogenous
Growth
This is good news!
If growth were merely a matter of technology,
we would not be able to do much about it …
… except to follow technology-friendly policies by
supporting R&D and such.
But if growth depends on saving and efficiency,
there are things that we can do, in the private
sector as well as through the public sector, to
foster rapid economic growth.
Because everything that is good for saving and
efficiency is also good for growth.
What to Do to Encourage
Economic Growth
Maintain strong incentives to save
Keep inflation low and real interest rates positive
Maintain financial system in good health
so as to channel saving into high-quality investment
Place strong emphasis on efficiency
1.
2.
3.
4.
Liberal price and trade regimes
Low inflation
Strong private sector
More and better education
The Bottom Line
Whatever increases efficiency also
increases economic growth.
Solow: Output per head increases, and
growth also increases for a while.
Romer: Economic growth increases (and
thereby also output per head).
This idea has triggered a large empirical
literature since the early 1990s.
My Work in this Area:
Three Main Channels
High inflation hurts growth.
Follows directly from g  sE  
Inflation may reduce both saving and
efficiency.
Excessive dependence on natural
resources impedes growth.
Has to do with
The Dutch disease
 Rent seeking
 Education

Public vs. private ownership
Which brings
me, at last, to
the point of
this paper ...
Privatization and
Economic Growth
Privatization means that profit-oriented
owners and able managers are allowed to
direct enterprises.
Profit motive replaces political considerations
as the guiding principle of business
operations.
Profit-maximizing owners generally want to appoint
managers and staff on merit rather than on the
basis of political connections, for example.
Hypothesis: Private enterprise is generally
more efficient than state-owned enterprises.
The Model
We derive r = r(v) by
applying Romer’s model of
expanding product variety.
Start with optimal growth à la Ramsey:
g   (r   )
Then assume
r  r (v )
with
r (v )  0
Why?
 SOEs may be inefficient, and may thus depend on low
interest and little growth.
 SOEs produce low-quality output, so that privatization
increases quality (i.e., efficiency) and growth.
What is the
evidence?
Empirical Results
Correlation analysis
To explore the relationship between
SOEs and relative productivity,
investment, and education
Regression analysis
To attempt to provide a fuller picture
of the interrelations among these
variables
Sample of Countries
Code
DZA
ARG
BEN
BOL
BWA
BRA
BDI
CMR
CHL
COL
COG
CIV
EGY
GMB
GHA
GRD
GIN
IND
IDN
KEN
KOR
Name
Algeria
Argentina
Benin
Bolivia
Botswana
Brazil
Burundi
Cameroon
Chile
Colombia
Congo
Cote d'Ivoire
Egypt, Arab Rep.
Gambia, The
Ghana
Grenada
Guinea
India
Indonesia
Kenya
Korea, Rep.
Code
MDG
MWI
MLI
MUS
MEX
NAM
PER
PHL
SEN
SLE
LKA
TAW
TZA
THA
TGO
TTO
TUN
TUR
ZAR
ZMB
Name
Madagascar
Malawi
Mali
Mauritius
Mexico
Namibia
Peru
Philippines
Senegal
Sierra Leone
Sri Lanka
Taiwan
Tanzania
Thailand
Togo
Trinidad and Tobago
Tunisia
Turkey
Zaire
Zambia
Figure 1. The Share of SOEs in Employ ment
and Relativ e Labor Productiv ity
Relative Labour Productivity 1978-92
1.0
IDN
THA BOLDZA
KOR
BRA CHL
EGY
COL
0.8
PER/TAW ZAR
SLE
IND
PHL
CMR
TUR
TUN
TTO
ARG
MLI
0.6
MDG
KEN
LKA
BWA
TZA
CIV
MWI
COG
0.4
Correlation
= -0.40
TGO
ZMB
GRD
SEN
MUS
BDI
0.2
GMB
GHA
0.0
0.0
0.1
0.2
0.3
Share of SOEs in Employment 1978-92
0.4
Figure 2. The Share of SOEs in Employ ment
and Inv estment
Share of Investment in GDP 1978-92
0.4
Correlation
= -0.50
0.3
KOR
IDN
TAW
CHL TUR DZA
THA
BWA
BRA PER
PHL MEX
COL ARG IND LKA
NAM
TTO KEN
CMR
0.1
MUS
MWI
BOL
ZAR EGY
0.2
MDG
GRD
TUN
CIV
TGO
TZA
BEN
BDI
GMB
MLI COG
SEN
SLE
ZMB
GHA
0.0
0.0
0.1
0.2
0.3
Share of SOEs in Employment 1978-92
0.4
Figure 3. The Share of SOEs in Employ ment
and Secondary Education
Secondary School-enrolment 1978
0.8
Correlation
= -0.58
KOR
0.6 PHL
TTO
ARG
PER
LKA
CHL
MUS
EGY
COL
MEX
0.4
BOL
TUR
BRA DZA
IND
THA
TUN TGO
ZAR
IDN
BWA
0.2
CIV
CMR
KEN
SLE GMB
MWI
MLI SEN
TZA
GHA
ZMB
BEN
BDI
0.0
0.0
0.1
0.2
0.3
Share of SOEs in Employment 1978-92
0.4
Figure 4. The Share of SOEs in Employ ment
and Economic Growth
Rate of growth of GDP per capita 1978-92
8
6
Correlation
= -0.35
KOR
TAW
IDN
4
THA
CHLBWA
2
CIV
MEX MUS
SLE
INDLGY
TUR
PHL
0
CMR LKA
BRA
ARG MWI
NAM
-2
PER
BOL
SEN
TZA
TGO
TUN
BDI
GRD
GHA
GMB
MLI
ZMB
BEN
MDG
-4
COG
-6
0.0
0.1
0.2
0.3
Share of SOEs in Employment 1978-92
0.4
Interpretation
An increase in the SOEs’ employment share
by 10 percentage points is associated with
a decrease in annual economic growth by
almost 1%.
A large SOE sector generally goes hand in
hand with slow growth.
No country with an SOE sector accounting for
more than 10% of employment had
economic growth of 2% per year or more.
 Sole exception: Côte d’Ivoire.
Regression Results
Table 2. Cross-section Results
Economic Growth , 1978-92
(1)
(2a)
(3a)
(4a)
(5)
Initial GDP
Investment
Secondary
education
SOE5
SOE2
SOE11
Constant
-0.003*
(1.19)
0,172
(4.74)
-
-0.008
(2.11)
0.130
(3.60)
0.030
(2.10)
-
-0.011
(1.73)
0.186
(3.07)
0.021*
(0.83)
-0.045*
(1.07)
-
-0.008 -0.009*
(-1.83) (1.55)
0.167
0.186
(3.79) (3.20)
0.019* 0.021*
(1.13) (0.87)
-0.034*
(0.82)
-0.034
(2.13)
-0.083*
(1.62)
0.003* 0.038* 0.056* 0.041* 0.051*
(0.16) (1.47) (1.29) (1.37) (1.23)
Investment
(2b)
(3b)
(4b)
0.024
(1.73)
-
0.024
(1.70)
-
0.053
(7.99)
-
-
-
-
-0.265
(2.56)
-0.277
(2.67)
-
-
-
0.078
(1.73)
-
-0.017* -0.112* -0.272
(0.15) (0.10) (-5.11)
Education
(2c) (3c)
(4c)
0.147 0.147 0.240
(4.47) (4.48) (13.15)
-
-
-0.420 -0.413
(1.72) (1.698)
-
-
0.036*
(0.30)
-
-0.730 -0.733 -1.463
(2.79) (2.80) (10.00)
SE
0.019
0.018
0.019
0.018
0.019
0.056
0.056
0.052
0.125 0.125 0.138
R2
# Countries
0.22
96
0.25
88
0.31
34
0.25
0.67
0.34
34
0.33
39
0.33
39
0.44
74
0.55
35
Note: t -values appear within parentheses below the coefficients. * Not significant at the 5
per cent level in a one-tail test.
0.55 0.72
35 69
Regression Results
Table 2. Cross-section Results
Economic Growth , 1978-92
(1)
(2a)
(3a)
(4a)
(5)
Initial GDP
Investment
Secondary
education
SOE5
SOE2
SOE11
Constant
-0.003*
(1.19)
0,172
(4.74)
-
-0.008
(2.11)
0.130
(3.60)
0.030
(2.10)
-
-0.011
(1.73)
0.186
(3.07)
0.021*
(0.83)
-0.045*
(1.07)
-
-0.008 -0.009*
(-1.83) (1.55)
0.167
0.186
(3.79) (3.20)
0.019* 0.021*
(1.13) (0.87)
-0.034*
(0.82)
-0.034
(2.13)
-0.083*
(1.62)
0.003* 0.038* 0.056* 0.041* 0.051*
(0.16) (1.47) (1.29) (1.37) (1.23)
Investment
(2b)
(3b)
(4b)
0.024
(1.73)
-
0.024
(1.70)
-
0.053
(7.99)
-
-
-
-
-0.265 -0.277
(2.56) (2.67)
-
-
0.078
(1.73)
-
-0.017* -0.112* -0.272
(0.15) (0.10) (-5.11)
Education
(2c) (3c)
(4c)
0.147 0.147 0.240
(4.47) (4.48) (13.15)
-
-
-0.420 -0.413
(1.72) (1.698)
-
-
0.036*
(0.30)
-
-0.730 -0.733 -1.463
(2.79) (2.80) (10.00)
SE
0.019
0.018
0.019
0.018
0.019
0.056
0.056
0.052
0.125 0.125 0.138
R2
# Countries
0.22
96
0.25
88
0.31
34
0.25
0.67
0.34
34
0.33
39
0.33
39
0.44
74
0.55
35
Note: t -values appear within parentheses below the coefficients. * Not significant at the 5
per cent level in a one-tail test.
0.55 0.72
35 69
Regression Results
Table 2. Cross-section Results
Economic Growth , 1978-92
(1)
(2a)
(3a)
(4a)
(5)
Initial GDP
Investment
Secondary
education
SOE5
SOE2
SOE11
Constant
-0.003*
(1.19)
0,172
(4.74)
-
-0.008
(2.11)
0.130
(3.60)
0.030
(2.10)
-
-0.011
(1.73)
0.186
(3.07)
0.021*
(0.83)
-0.045*
(1.07)
-
-0.008 -0.009*
(-1.83) (1.55)
0.167
0.186
(3.79) (3.20)
0.019* 0.021*
(1.13) (0.87)
-0.034*
(0.82)
-0.034
(2.13)
-0.083*
(1.62)
0.003* 0.038* 0.056* 0.041* 0.051*
(0.16) (1.47) (1.29) (1.37) (1.23)
Investment
(2b)
(3b)
(4b)
0.024
(1.73)
-
0.024
(1.70)
-
0.053
(7.99)
-
-
-
-
-0.265 -0.277
(2.56) (2.67)
-
-
0.078
(1.73)
-
-0.017* -0.112* -0.272
(0.15) (0.10) (-5.11)
Education
(2c) (3c)
(4c)
0.147 0.147 0.240
(4.47) (4.48) (13.15)
-
-
-0.420 -0.413
(1.72) (1.698)
-
-
0.036*
(0.30)
-
-0.730 -0.733 -1.463
(2.79) (2.80) (10.00)
SE
0.019
0.018
0.019
0.018
0.019
0.056
0.056
0.052
0.125 0.125 0.138
R2
# Countries
0.22
96
0.25
88
0.31
34
0.25
0.67
0.34
34
0.33
39
0.33
39
0.44
74
0.55
35
Note: t -values appear within parentheses below the coefficients. * Not significant at the 5
per cent level in a one-tail test.
0.55 0.72
35 69
Regression Results
Table 2. Cross-section Results
Economic Growth , 1978-92
(1)
(2a)
(3a)
(4a)
(5)
Initial GDP
Investment
Secondary
education
SOE5
SOE2
SOE11
Constant
-0.003*
(1.19)
0,172
(4.74)
-
-0.008
(2.11)
0.130
(3.60)
0.030
(2.10)
-
-0.011
(1.73)
0.186
(3.07)
0.021*
(0.83)
-0.045*
(1.07)
-
-0.008 -0.009*
(-1.83) (1.55)
0.167
0.186
(3.79) (3.20)
0.019* 0.021*
(1.13) (0.87)
-0.034*
(0.82)
-0.034
(2.13)
-0.083*
(1.62)
0.003* 0.038* 0.056* 0.041* 0.051*
(0.16) (1.47) (1.29) (1.37) (1.23)
Investment
(2b)
(3b)
(4b)
0.024
(1.73)
-
0.024
(1.70)
-
0.053
(7.99)
-
-
-
-
-0.265 -0.277
(2.56) (2.67)
-
-
0.078
(1.73)
-
-0.017* -0.112* -0.272
(0.15) (0.10) (-5.11)
Education
(2c) (3c)
(4c)
0.147 0.147 0.240
(4.47) (4.48) (13.15)
-
-
-0.420 -0.413
(1.72) (1.698)
-
-
0.036*
(0.30)
-
-0.730 -0.733 -1.463
(2.79) (2.80) (10.00)
SE
0.019
0.018
0.019
0.018
0.019
0.056
0.056
0.052
0.125 0.125 0.138
R2
# Countries
0.22
96
0.25
88
0.31
34
0.25
0.67
0.34
34
0.33
39
0.33
39
0.44
74
0.55
35
Note: t -values appear within parentheses below the coefficients. * Not significant at the 5
per cent level in a one-tail test.
0.55 0.72
35 69
Regression Results
Table 2. Cross-section Results
Economic Growth , 1978-92
(1)
(2a)
(3a)
(4a)
(5)
Initial GDP
Investment
Secondary
education
SOE5
SOE2
SOE11
Constant
-0.003*
(1.19)
0,172
(4.74)
-
-0.008
(2.11)
0.130
(3.60)
0.030
(2.10)
-
-0.011
(1.73)
0.186
(3.07)
0.021*
(0.83)
-0.045*
(1.07)
-
-0.008 -0.009*
(-1.83) (1.55)
0.167
0.186
(3.79) (3.20)
0.019* 0.021*
(1.13) (0.87)
-0.034*
(0.82)
-0.034
(2.13)
-0.083*
(1.62)
0.003* 0.038* 0.056* 0.041* 0.051*
(0.16) (1.47) (1.29) (1.37) (1.23)
Investment
(2b)
(3b)
(4b)
0.024
(1.73)
-
0.024
(1.70)
-
0.053
(7.99)
-
-
-
-
-0.265 -0.277
(2.56) (2.67)
-
-
0.078
(1.73)
-
-0.017* -0.112* -0.272
(0.15) (0.10) (-5.11)
Education
(2c) (3c)
(4c)
0.147 0.147 0.240
(4.47) (4.48) (13.15)
-
-
-0.420 -0.413
(1.72) (1.698)
-
-
0.036*
(0.30)
-
-0.730 -0.733 -1.463
(2.79) (2.80) (10.00)
SE
0.019
0.018
0.019
0.018
0.019
0.056
0.056
0.052
0.125 0.125 0.138
R2
# Countries
0.22
96
0.25
88
0.31
34
0.25
0.67
0.34
34
0.33
39
0.33
39
0.44
74
0.55
35
Note: t -values appear within parentheses below the coefficients. * Not significant at the 5
per cent level in a one-tail test.
0.55 0.72
35 69
Regression Results
Table 2. Cross-section Results
Economic Growth , 1978-92
(1)
(2a)
(3a)
(4a)
(5)
Initial GDP
Investment
Secondary
education
SOE5
SOE2
SOE11
Constant
-0.003*
(1.19)
0,172
(4.74)
-
-0.008
(2.11)
0.130
(3.60)
0.030
(2.10)
-
-0.011
(1.73)
0.186
(3.07)
0.021*
(0.83)
-0.045*
(1.07)
-
-0.008 -0.009*
(-1.83) (1.55)
0.167
0.186
(3.79) (3.20)
0.019* 0.021*
(1.13) (0.87)
-0.034*
(0.82)
-0.034
(2.13)
-0.083*
(1.62)
0.003* 0.038* 0.056* 0.041* 0.051*
(0.16) (1.47) (1.29) (1.37) (1.23)
Investment
(2b)
(3b)
(4b)
0.024
(1.73)
-
0.024
(1.70)
-
0.053
(7.99)
-
-
-
-
-0.265 -0.277
(2.56) (2.67)
-
-
0.078
(1.73)
-
-0.017* -0.112* -0.272
(0.15) (0.10) (-5.11)
Education
(2c) (3c)
(4c)
0.147 0.147 0.240
(4.47) (4.48) (13.15)
-
-
-0.420 -0.413
(1.72) (1.698)
-
-
0.036*
(0.30)
-
-0.730 -0.733 -1.463
(2.79) (2.80) (10.00)
SE
0.019
0.018
0.019
0.018
0.019
0.056
0.056
0.052
0.125 0.125 0.138
R2
# Countries
0.22
96
0.25
88
0.31
34
0.25
0.67
0.34
34
0.33
39
0.33
39
0.44
74
0.55
35
Note: t -values appear within parentheses below the coefficients. * Not significant at the 5
per cent level in a one-tail test.
0.55 0.72
35 69
Regression Results
Table 2. Cross-section Results
Economic Growth , 1978-92
(1)
(2a)
(3a)
(4a)
(5)
Initial GDP
Investment
Secondary
education
SOE5
SOE2
SOE11
Constant
-0.003*
(1.19)
0,172
(4.74)
-
-0.008
(2.11)
0.130
(3.60)
0.030
(2.10)
-
-0.011
(1.73)
0.186
(3.07)
0.021*
(0.83)
-0.045*
(1.07)
-
-0.008 -0.009*
(-1.83) (1.55)
0.167
0.186
(3.79) (3.20)
0.019* 0.021*
(1.13) (0.87)
-0.034*
(0.82)
-0.034
(2.13)
-0.083*
(1.62)
0.003* 0.038* 0.056* 0.041* 0.051*
(0.16) (1.47) (1.29) (1.37) (1.23)
Investment
(2b)
(3b)
(4b)
0.024
(1.73)
-
0.024
(1.70)
-
0.053
(7.99)
-
-
-
-
-0.265
(2.56)
-0.277
(2.67)
-
-
-
0.078
(1.73)
-
-0.017* -0.112* -0.272
(0.15) (0.10) (-5.11)
Education
(2c) (3c)
(4c)
0.147 0.147 0.240
(4.47) (4.48) (13.15)
-
-
-0.420 -0.413
(1.72) (1.698)
-
-
0.036*
(0.30)
-
-0.730 -0.733 -1.463
(2.79) (2.80) (10.00)
SE
0.019
0.018
0.019
0.018
0.019
0.056
0.056
0.052
0.125 0.125 0.138
R2
# Countries
0.22
96
0.25
88
0.31
34
0.25
0.67
0.34
34
0.33
39
0.33
39
0.44
74
0.55
35
Note: t -values appear within parentheses below the coefficients. * Not significant at the 5
per cent level in a one-tail test.
0.55 0.72
35 69
Regression Results
Table 2. Cross-section Results
Economic Growth , 1978-92
(1)
(2a)
(3a)
(4a)
(5)
Initial GDP
Investment
Secondary
education
SOE5
SOE2
SOE11
Constant
-0.003*
(1.19)
0,172
(4.74)
-
-0.008
(2.11)
0.130
(3.60)
0.030
(2.10)
-
-0.011
(1.73)
0.186
(3.07)
0.021*
(0.83)
-0.045*
(1.07)
-
-0.008 -0.009*
(-1.83) (1.55)
0.167
0.186
(3.79) (3.20)
0.019* 0.021*
(1.13) (0.87)
-0.034*
(0.82)
-0.034
(2.13)
-0.083*
(1.62)
0.003* 0.038* 0.056* 0.041* 0.051*
(0.16) (1.47) (1.29) (1.37) (1.23)
Investment
(2b)
(3b)
(4b)
0.024
(1.73)
-
0.024
(1.70)
-
0.053
(7.99)
-
-
-
-
-0.265
(2.56)
-0.277
(2.67)
-
-
-
0.078
(1.73)
-
-0.017* -0.112* -0.272
(0.15) (0.10) (-5.11)
Education
(2c) (3c)
(4c)
0.147 0.147 0.240
(4.47) (4.48) (13.15)
-
-
-0.420 -0.413
(1.72) (1.698)
-
-
0.036*
(0.30)
-
-0.730 -0.733 -1.463
(2.79) (2.80) (10.00)
SE
0.019
0.018
0.019
0.018
0.019
0.056
0.056
0.052
0.125 0.125 0.138
R2
# Countries
0.22
96
0.25
88
0.31
34
0.25
0.67
0.34
34
0.33
39
0.33
39
0.44
74
0.55
35
Note: t -values appear within parentheses below the coefficients. * Not significant at the 5
per cent level in a one-tail test.
0.55 0.72
35 69
Summary of Results
1. An increase in SOE employment reduces
investment and education and thereby also
economic growth indirectly.
 An increase in SOE labor share by 10% reduces
growth by 1% (t = 2.1).
2. An increase in SOE output reduces
economic growth directly.
 An increase in SOE output share by 10%
reduces growth by 0.3% (t = 2.1).
3. An increase in SOE debt reduces growth.
Conclusion
Across countries, SOE activity is inversely
related to
 Relative productivity of labor in SOE sector
 Investment
 Education
Economic growth 1978-1992 varies inversely
with the size of the SOE sector.
 An increase in SOE labor share by 10% reduces
growth by 0.5% to 1% across countries, cet. par.
So, privatization may be good for growth.
 But it needs to be implemented with care.