Download SHOCK THERAPY VERSUS GRADUALISM RECONSIDERED

SHOCK THERAPY VERSUS GRADUALISM
RECONSIDERED: LESSONS FROM
TRANSITION ECONOMIES AFTER 15 YEARS
OF REFORMS
Vladimir Popov
This paper is the logical continuation of my earlier article:
Shock Therapy versus Gradualism: The End of the Debate
(Explaining the Magnitude of the Transformational Recession)
Comparative Economic Studies, Vol. 42, No. 1, Spring 2000, pp. 1-57.
The arguments of this earlier article, however, are reconsidered in
light of new research and evidence.
GDP change in selected transition
economies, 1989 = 100%
Poland
125
Slovenia
Central
Europe
115
Slovak
Republic
105
Hungary
95
Czech
Republic
South
Eastern
Europe
85
Rom ania
75
Bulgaria
65
FSU
(Russia,
Ukraine)
55
Russia
Ukraine
45
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
35
INITIAL LIBERALIZATION AND OUTPUT CHANGE DURING RECESSION
LIBERALISATION AND OUTPUT CHANGE IN CHINA AND RUSSIA
Fig. 3. Indices of econom ic freedom and GDP grow th in Russia and China
250
4
GDP as a % of 1994
Index of econom ic freedom , right scale
220
3,5
190
3
160
2,5
CHINA
130
2
GDP, 1993=100%, left scale
100
1,5
RUSSIA
70
1
1993
1994
1995
1996
1997
1998
1999
2000
2001
GDP change in Asian economies, 1989 = 100%
Fig. 1b. GDP change in Asian transition econom ies, 1989 = 100%
200
China
180
Vietnam
160
Central Europe
140
Uzbekistan
Mongolia
120
Turkm enistan
100
Kazakhstan
80
Russia
60
Kyrgyzstan
40
1989
1991
1993
1995
1997
1999
2001
2003
Tajikistan
Source: EBRD, World Bank.
GDP change in FSU economies, 1989 = 100%
Fig. 2. GDP change in FSU econom ies, 1989 = 100%
Central Europe
Uzbekistan
Belarus
115
Kazakhstan
105
Estonia
95
Turkm enistan
Azerbaijan
85
Latvia
75
Lithuania
Tajikistan
65
Russia
55
Kyrgyzstan
Arm enia
45
Ukraine
35
Georgia
25
Moldova
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Russia was leading in economic liberalization,
while Belarus was lagging
Private sector share in GDP, %
80
Russian Federation
70
Ukraine
60
Belarus
50
40
30
20
10
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
0
But Belarus and Uzbekistan are doing better (even though they
are net importers of fuel), not to mention net exporters like
Azerbaijan, Kazakhstan, Turkmenistan
GDP in 2006 as a % of 1989
Moldova
Georgia
Ukraine
Tajikistan
Kyrgyzstan
Russia
Bulgaria
Croatia
Lithuania
Latvia
Rom ania
Azerbaijan
Kazakhstan
Arm enia
Czech Republic
Hungary
Belarus
Uzbekistan
Slovakia
Slovenia
Albania
Estonia
Poland
Turkm enistan
40
60
80
100
120
140
160
180
Russian growth is lagging behind that of oil
exporters and some oil importers
Average annual GDP grow th rates in CIS countries in 2000-07, EBRD
estimates
Azerbaijan
Turkmenistan
Armenia
Kazakhstan
Tajikistan
Belarus
Ukraine
Georgia
Russia
Moldova
Uzbekistan
Kyrgyzstan
16
14
12
10
8
6
4
2
0
INITIAL LIBERALIZATION AND OUTPUT CHANGE DURING RECOVERY
Fig. 8. Liberalization index by 1995 and perform ance in 1994-98
1998 GDP as a % of 1994 GDP
155
LITH
145
CHINA
135
125
CIS countries
ALB
115
LATV
ROM
105
95
EST
R2 = 0,3345
POL
VIETN
SLOVAK
CZ
CROAT
SLOVEN
HU
MACED
MONG
BULG
85
75
65
0,5
1
1,5
2
2,5
3
Cum ulative liberalization index, 1995
3,5
4
4,5
LIBERALIZATION STOCK AND OUTPUT CHANGE DURING RECOVERY
Fig. 9. Liberalisation and output change in transition econom ies in 1995-2003
2003 GDP as a % of 1995
200
LIBER_94
LIBER2004
180
160
R2 = 0,0229
140
R2 = 0,0618
120
100
80
1
1,5
2
2,5
3
Cum ulative liberalisation index
3,5
4
4,5
LIBERALIZATION STOCK AND OUTPUT CHANGE DURING RECOVERY
Fig. 10. Liberalisation increase and output change in transition econom ies in 19952003
2003 GDP as a % of 1995
200
180
160
R2 = 0,0645
140
120
100
80
-1
-0,5
0
0,5
1
1,5
Increase in cum ulative liberalisation index in 1995-2003
2
2,5
Does liberalization matter?
• Vietnam and China are similar in initial conditions and in
transition results (immediate growth of output without
transformational recession) despite different reform strategies:
• Chinese reforms are the classical example of gradualism
• Vietnamese reformers introduced shock therapy treatment
(instant deregulation of most prices and introduction of
convertibility of dong) in 1989
• Differing performance of the former Soviet Union (FSU) states:
• Baltic states are the champions of liberalization and
stabilization in the region. In the Baltics, however, output
fell in the early 1990s by 36-60% and even in 2005, 10 years
after the bottom of the recession was reached, was still
below the pre-recession maximum.
• Uzbekistan is commonly perceived to be one of the worst
procrastinators. However in Uzbekistan the reduction of
output in 1990-95 totaled only 18% and the economy
started to grow again in 1996
• By 2005 only two former Soviet republics - Uzbekistan and
Turkmenistan - surpassed the pre-recession level of 1989
Reasons for recession:
(1) Supply shock - change in relative prices
Fig. 4. Change in relative prices and output in 1990-98 in Russian industry
January 1998 output as a % of
January 1990 output
90
Electric energy
Non-ferrous m etals
R2 = 0,4062
60
Steel
Fuel
Chem icals
30
Food
Petrochem icals
Wood
Machinery
Constr. m aterials
Light
0
40%
60%
80%
100%
120%
140%
Ratio of 1998 proices to 1990 prices as a % of industrial average
160%
Industrial structure and external trade
distortions
1996 GDP as a % of 1998 GDP
Fig. 7. Aggregate distortions in industrial structure and external
trade before transition and GDP change during transition
CHINA
200
VIETN
170
140
SLOVEN
110
HUNG
EST
80
MACED
50
CZECH
CROAT
ARM
20
0
10
20
30
40
50
60
70
Distortions in industrial structure and external trade, % of GDP
80
90
Reasons for recession:
(2) Collapse of state institutions - government revenues and
expenditure fell in most transition economies
Fig. 13. Consolidated government revenues as a % of GDP
Central European
countries
55
50
45
South East Europe
countries
40
Baltic states
35
25
Caucasian states
%
30
Central Asian
countries
20
-China
15
10
1989
RUSSIA
1990
1991
1992
1993
1994
1995
1996
Three major patterns of change in
government expenditures
Fig. 10. Gove rnme nt e xpe nditure , % of GDP
60
POLAND
50
USSR/ RUSSIA
CHINA
40
Debt service
30
Defence
20
Subsidies
10
Investment
"Ordinary government"
0
1985
1989
1995
1978
1985
1994
1989
1996
The decline in government revenues is
correlated with performance
Impact of initial conditions, institutions,
liberalization:1989-96
Table 1. Regression of change in GDP in 1989-96 on initial conditions, policy factors, and rule of law
and democracy indices, robust estimates
Dependent variable = log (1996 GDP as a % of 1989 GDP)
For China - all indicators are for the period of 1979-86 or similar
Equations, Number of
1,
2,
3,
4,
5,
6,
Observations / Variables
N=28
N=28
N=28
N=28
N=28
N=28
Constant
5.3***
5.4***
5.2***
5.4***
5.4***
5.5***
-.005**
-.005**
-.003
-.006**
-.007***
-.007***
Distortions, % of GDPa
1987 PPP GDP per capita, % of the
US level
War dummyb
Decline in government revenues as
a % of GDP from 1989-91 to 199396
Liberalization index
Log (Inflation, % a year, 1990-95,
geometric average)
Rule of law index, average for
1989-97, %
Democracy index, average for
1990-98, %
Ratio of the rule of law to
democracy index
Adjusted R2, %
-.009**
-.006*
-.007**
-.007**
-.009***
-.19c
-.36***
-.011***
-.16***
-.20***
.05
-.18***
-.17***
-.13***
-.008
***
-.37***
-.008***
-.011
***
-.011
***
-.02
-.13***
.03
-.14***
.008
***
-.005
***
82
7,
N=28
5.7***
-.007***
-.45***
-.003**
.07***
.07***
.06***
.05***
.05***
83
83
85
91
91
*, **, *** - Significant at 1, 5 and 10% level respectively.
aCumulative measure of distortions as a % of GDP equal to the sum of defense expenditure (minus 3%
regarded as the 'normal' level), deviations in industrial structure and trade openness from the 'normal' level,
the share of heavily distorted trade (among the FSU republics) and lightly distorted trade (with socialist
countries) taken with a 33% weight – see (Popov, 2000) for details.
bEquals 1 for Armenia, Azerbaijan, Croatia, Georgia, Macedonia, and Tajikistan and 0 for all other
countries.
c
Significant at 13% level.
90
Evaluating the impact of non-policy and
policy factors:1989-98
Dependent variable = log (1998 GDP as a % of 1989 GDP)
For China - all indicators are for the period of 1979-88 or similar
Equations, Number of
1,
2,
observations / Variables
N= 2 8
N= 2 8
Constant
5 .3 0
4 .8 8
-.01
-.01
Distortions, % of GDPa
1987 PPP GDP per capita, % of the US -.01
-.02
level
-.67
-.58
War dummyb
Decline in government revenues as a % of -.02
-.01
GDP from 1989-91 to 1993-96
Liberalization index
( .1 1 )
Log (Inflation, % a year, 1990-95,
geometric average)
Shadow economy as a % of GDP in 1994
67
69
Adjusted R2, %
3,
N= 2 8
5 .6 8
-.01
-.01
4,
N= 2 8
5 .7 3
-.01
-.01
-.38
-.01
-.37
-.01
-.15
(-.01)
-.15
80
80
5,
N= 1 7
5 .7 4
-.01
-.01
6,
N= 1 7
5 .4 3
-.01 c
-.02
7,
N= 1 7
5 .8 6
(-.00)
-.01c
8,
N= 1 7
6 .0 8
(-.00)
(-.01)
-.13
(-.06)
-.16
-.01
82
-.02
81
( .1 1 )
-.02
72
-.02
73
aCumulative measure of distortions as a % of GDP equal to the sum of defense expenditure (minus 3% regarded as the 'normal' level), deviations
in industrial structure and trade openness from the 'normal' level, the share of heavily distorted trade (among the FSU republics) and lightly
distorted trade (with socialist countries) taken with a 33% weight (see Appendix for details).
bEquals 1 for Armenia, Azerbaijan, Croatia, Georgia, Macedonia, and Tajikistan and 0 for all other countries.
cSignificant at 8% level.
Best performance: low distortions, strong institutions
Worst performance: high distortions, weak institutions
INITIAL CONDITIONS (DISTORTIONS) AND INSTITUTIONS –
CLASSIFICATION OF COUNTRIES
DISTORTIONS
INSTITUTIONAL
CAPACITY
HIGH
LOW
LOW
HIGH
CHINA,
EASTERN
VIETNAM EUROPE
ALBANIA, FSU
MONGOLIA
Explaining the magnitude of recession:
• Differences in performance during
transition depend strongly on the initial
conditions:
• The higher the distortions (militarization, overindustrialization, "under-openness" of the economy
and the share of perverted trade flows), the worse is
the performance
• The higher was GDP per capita before transition, the
greater were distortions embodied in fixed capital
stock, the more difficult it was to overcome these
distortions to achieve growth
Explaining the magnitude of recession:
• Macroeconomic stability matters a great
deal for economic performance
• Liberalization index does not appear to be
important
• Changes in the institutional capabilities of
the state have dramatic impact on
performance
By the end of the 1990s most countries were already
recovering from recession
Table 6. Regression of change in GDP in 1994-98 on non-policy and policy-related factors
(all coefficients are significant at 15% level except those in brackets)
Dependent variable = log (1998 GDP as a % of 1994 GDP)
For China - all indicators are for the period of 1984-88 or similar
5,
4,
3,
2,
1,
Equations, Number of
N=28 N=28 N=28 N=28 N=28
Observations / Variables
4.60
4.51 4.25 4.56 4.32
Constant
.003
.004 .005 .003
Distortions, % of GDPa
1987 PPP GDP per capita, % of the US level
.15
War dummyb
Decline in government revenues as a % of
-.003 c -.004
GDP from 1989-91 to 1993-96
.07
.10
.09
.12
.07
Liberalization index
.04
-.06
Log (Inflation, % a year, 1990-95,
Geometric average)
33
29
37
27
21
Adjusted R2, %
aCumulative measure of distortions as a % of GDP equal to the sum of defense expenditure (minus 3% regarded as the 'normal' level), deviations
in industrial structure and trade openness from the 'normal' level, the share of heavily distorted trade (among the FSU republics) and lightly
distorted trade (with socialist countries) taken with a 33% weight (see Appendix for details).
bEquals 1 for Armenia, Azerbaijan, Croatia, Georgia, Macedonia, and Tajikistan and 0 for all other countries.
c Significant at 21% level.
Impact of initial conditions, institutions, liberalization
Table 2. Regression of change in GDP in 1995-2003 on initial conditions, institutional capacity,
liberalization and rule of law and democracy indices, robust estimates
Dependent variable = 2003 GDP as a % of 1995 GDP
For China - all indicators are for the period of 1979-86 or similar
Equations, Number of
1,
2,
3,
4,
5,
Observations / Variables
N=28
N=28
N=28
N=28
N=28
Constant
105***
91***
99***
78***
99***
1996 GDP as a % of 1989 GDP
.33***
.45***
.46***
.24**
1987 PPP GDP per capita, % of the US level
22.9**
42.3***
32.0***
19.4*
War dummya
Liberalization index in 1995
Increase in the liberalization index in 1995-2003
Decline in government revenues as a % of GDP from
1989-91 to 1993-96a
Rule of law index, average for 1989-97, %
Democracy index, average for 1990-98, %
Adjusted R2, %
-19.9***
15.3
***
16.7
***
17.6***
17.6***
-.8***
.8**
25
38
45
1.0***
-.6***
52
*, **, *** - Significant at 1, 5 and 10% level respectively.
aEquals 1 for Armenia, Azerbaijan, Croatia, Georgia, Macedonia, and Tajikistan and 0 for all other
countries.
1.2***
-.8***
55
OBJECTION:
Speed and extent of liberalization may be endogenous
Cumulative
iberalization
indices by 1995
Fig. 3. Liberalization indices and distortions in
industrial structure and trade patterns
5
4
3
2
1
0
CIS
0
20
40
60
Distortions in the industrial structure and trade patterns
(% of GDP)
80
Economic liberalization and democratization go hand in hand
Cumulative iberalization index
by 1995
Fig. 4. Democracy index (1990-98, average) and economic
liberalization index by 1995
5
4
CIS and BALKANS
VIETN
CROAT, MACED
3
CHINA
2
1
0
0
20
40
60
Democracy index (Freedom House)
80
100
Instrumenting liberalization stock with
democracy level variable: 1989-96
Table 3. 2SLS robust estimates – regression of change in GDP in 1989-96 on initial conditions,
institutional capacity, liberalization and rule of law and democracy indices (Liberalization index
instrumented with the democracy level variable)
Dependent variable = Log (1996 GDP as a % of 1989 GDP)
For China - all indicators are for the period of 1979-86 or similar
Equations, Number of
1,
2,
3,
4,
Observations / Variables
N=28
N=28
N=17
N=17
Constant
6.4***
6.3***
6.0***
6.0***
Pre-transition distortions, % of GDP
-.01***
-.02***
-.004
1987 PPP GDP per capita, % of the US level
-.007**
-.01***
-.45***
-.29b
War dummya
Liberalization index in 1995
Decline in government revenues as a % of GDP from 1989-91 to
1993-96
Log (Inflation, % a year, 1990-95, geometric average)
Rule of law index, average for 1989-97, %
Increase in the share of shadow economy in GDP in 1989-94, p.p.
R2, %
-.18**
-.02***
-.39*
-.02***
-.19***
-.19***
-1.7***
-.22***
-.01c
-.22***
-.19***
-.02***
88
-.015***
90
86
77
*, **, *** - significant at 1, 5 and 10% level respectively.
aEquals 1 for Armenia, Azerbaijan, Croatia, Georgia, Macedonia, and Tajikistan and 0 for all other
countries.
b
Significant at 12% level.
c
Significant at 16% level.
Instrumenting liberalization change with liberalization
stock and FSU dummy variables:1995-2003
Table 4. 2SLS robust estimates – regression of change in GDP in 1995-2003 on
initial conditions, institutional capacity, liberalization and rule of law and
democracy indices
Dependent variable = 2003 GDP as a % of 1995 GDP
For China the indicator is for the period 10 years earlier.
Equations, Number of
Observations / Variables
Instruments for liberalization change in 1995-03 variable
1,
N=28
LIBER95
2,
N=28
FSU
Constant
1996 GDP as a % of 1989 GDP
War dummya
97.8***
95.8***
3,
N=28
LIBER95
and FSU
97.7***
19.5*
19.8**
19.5*
4,
N = 28
LIBER95
and FSU
79.5***
.18*
25.0**
Increase in liberalization index in 1995-2003
Decline in government revenues as a % of GDP from 198991 to 1993-96
Rule of law index, average for 1989-97, %
Democracy index, average for 1990-98, %
R2, %
18.2***
-.76***
19.2**
-.78**
18.3***
-.76***
22.9***
-.65***
1.24***
-.76***
55
1.28***
-.76***
54
1.25***
-.76***
55
1.13***
-.62***
56
*, **, *** - significant at 1, 5 and 10% level respectively.
aEquals 1 for Armenia, Azerbaijan, Croatia, Georgia, Macedonia, and Tajikistan and 0 for all other
countries.
Conclusions
• The impact of the speed of liberalization at
the initial stage of transition, i.e. during the
transformational recession, appears to be
negative, if any.
• The reason for the negative impact is most
probably associated with limited ability of the
economy to adjust to new price ratios
Conclusions
Fig. 5. Hypothetical trajectories of output (Year "0" = 100%) assum ing gradual and
instant liberalization
Reduction of
output in a
NC sector at
10% annually
Output, Year "0" = 100%
125%
120%
115%
110%
Reduction of
output in a
NC sector at
30% annually
105%
100%
95%
90%
85%
80%
0
1
2
3
4
5
6
7
8
9
10
Years
Assum ptions: size of non-com petitive sector (NC) in the initial
year = 20% of total output; net investm ent (s) = 10% of total output;
m arginal capital productivity, output increase per unit of net investm ent (a) = 1/3.
Reduction of
output in a
NC sector at
100%
annually
Conclusions
• At the recovery stage liberalization starts to affect
growth positively, whereas the impact of pretransition distortions disappears. Institutional
capacity and macroeconomic policy continue to be
important prerequisites for successful performance.
• Liberalization at the recovery stage influences
performance positively because it creates market
stimuli without causing rapid collapse of output of
inefficient
industries,
which
cannot
be
compensated fully by the rise of efficient industries
due to investment constraints.