Upload
others
View
11
Download
0
Embed Size (px)
Citation preview
Economic Diversification in Low–Income
Countries (LICs): Stylized Facts and
Macroeconomic Implications
Second Annual IMF/WB/WTO
Joint Trade Workshop
Geneva, Switzerland
June 5 - 6, 2013
Chris Papageorgiou Nikola Spatafora
Motivation
• LICs have historically been heavily dependent on a narrow
range of traditional primary products and on few export markets.
• While diversification has been a longstanding ambition for many
LICs, we have limited experience of what aspects of
diversification are important.
• Many developing economies have managed to diversify and
transform their economies especially over the past two decades.
But there remain significant differences across countries.
• What are the facts? What does diversification involve?
• Inform the policy debate on diversification and transformation.
Overview Motivation
Diversification, growth and volatility
• Cross-country growth relationship
• Diversification episodes and volatility
Stylized Facts
Cross-country patterns: Link between diversification, and GDP or time
• Trade diversification indices across products and partners
• Quality upgrading
• Country case studies
Ongoing Work & Preliminary Messages
• Messages
• Workring papers
Growth through Diversification
USAGBRAUT
DNKFRADEUITANLD
NOR
SWECHE
CAN
JPNFIN
GRCIRLPRTESP
TURAUS
NZLZAFARGBOL
BRA CHLCOLCRI
DOM
ECUSLV
GTM
HTI
HND
MEX
NIC
PAN
PRYPER
URY
VENJAM
TTO
BHRCYP
IRN
IRQ
ISR
JORKWT
LBN
OMN
SAUSYRARE
EGY
AFGBGDKHM
LKA
HKG
INDIDN
KOR
LAOMYS
NPL
PAK
PHL
SGPTHA VNM
DZA AGO
BDICMR
CAF
TCD
COG
ZAR
BENERI
ETH
GAB
GMB
GHA
GNB
GIN
CIVKEN
LBR
LBYMDG
MWIMLIMRT
MUS
MAR
MOZ
NER
NGA
ZWE
RWASEN SLESDN
TZA
TGO
TUN
UGABFA
ZMB
PNG
ARMAZE
BLR
ALB
GEOKAZ
KGZ
BGR
MDARUS
TJK
CHN
TKM
UKRUZB
CZESVK
EST
LVA
HUN
LTU
MNGHRV
SVN
MKD
BIH
POL
ROM
-50
510
Re
al G
DP
PC
gro
wth
1 2 3 4 5 6Total Theil
(mean) growthln_rgdppc Fitted values
Diversification associated with higher real GDP growth
Volatility and Diversification
3.6
3.7
3.8
3.9
4.0
4.1
4.2
4.3
4.4
4.5
Pre-episode Episode Post-episode
Output volatility diminishes after major diversification episodes
Stylized Facts: Trade Diversification Trade Diversification (across Products) vs. GDP
Export Diversification and Concentration
Along the development path, diversification increases (until threshold) 0
24
6
Div
ers
ific
atio
n I
nd
ex
0 20000 40000 60000Real GDP Per Capita
Nonparametric Quadratic
Stylized Facts: Trade Diversification Trade Diversification (across Products) vs. Time
22
.53
3.5
44
.5
Ave
rag
e T
heil In
dex
1960 1970 1980 1990 2000 2010Year
HIC MIC
LIC
LICs and SSA have high concentration, but there is a break around 1990-5
By Income Group
23
45
Ave
rag
e T
heil In
dex
1960 1970 1980 1990 2000 2010Year
East Asia & Pacific Europe & Central Asia
Latin America & Caribbean Middle East & North Africa
South Asia Sub-Saharan Africa
By Region
Stylized Facts: Trade Diversification Agricultural and Manufacturing Export Share over Time
Changes in diversification are paralleled by shifts in the relative importance of agriculture… 0
.1.2
.3.4
.5
Ave
rag
e S
ha
re o
f A
gri
cu
ltu
re E
xp
ort
s
E/S Asia Europe/Cent. Asia LAC MENA SSA
1965-1970 2006-2010
0.2
.4.6
.8
Ave
rag
e S
ha
re o
f M
an
ufa
ctu
rin
g E
xp
ort
s
E/S Asia Europe/Cent. Asia LAC MENA SSA
1965-1970 2006-2010
… and manufacturing
Stylized Facts: Trade Diversification Trade Diversification (across Partners) over Time
After 1995, Asia and Africa greatly diversified across partners
0.5
11
.5
Ave
rag
e B
etw
ee
n T
he
il
E/S Asia Europe/Cent. Asia LAC MENA SSA
1965-1970 2006-2010
02
46
Div
ersific
atio
n In
dex
0 20000 40000 60000Real GDP Per Capita
All Other Regions South and East Asia
Nonparametric Quadratic
Stylized Facts: Partner Diversification Trade Diversification (across Partners) vs. Time
2.4
2.6
2.8
33
.23
.4
Ave
rag
e T
heil In
dex
1960 1970 1980 1990 2000 2010Year
HIC MIC
LIC
After 1990, LICs and esp. SSA greatly diversified across partners
0.5
11
.5
Ave
rag
e B
etw
ee
n T
heil
1960 1970 1980 1990 2000 2010Year
East Asia & Pacific Europe & Central Asia
Latin America & Caribbean Middle East & North Africa
South Asia Sub-Saharan Africa
By Income Group By Region
Stylized Facts: Partner Diversification Trade Diversification (across Partners) vs. Time
Share of Exports to EU
.3.4
.5.6
Average S
hare o
f E
xports
to E
U
1960 1970 1980 1990 2000 2010Year
HICs MICs
LICs
0
.05
.1.1
5
Average S
hare o
f E
xports
to C
hin
a
1960 1970 1980 1990 2000 2010Year
HICs MICs
LICs
Share of Exports to China
This is partly related to emergence of China and diminishing role of EU
Quality Upgrading
• Economic development is underpinned not just by new
products and markets, but also by quality improvements
to existing products.
• Ongoing work develops a toolkit to answer:
– What is an economy’s export quality and how has it changed over
time?
– What is its present potential for quality upgrading?
– How large is the need for diversification?
• Quality estimates improve on unit values, which can be
imperfect quality proxies. Estimated quality measures
smooth volatility and dispersion inherent in unit values.
The Toolkit
• Broadest set of quality estimates to date, covering 178 countries
during 1962-2010. More than 21 million quality estimates at
‘importer-exporter-year-product-unit of measurement’ level.
• Toolkit will be made publicly available and contains exporter
country totals and 3 different breakdowns:
– SITC 4, 3, 2, 1 digit
• Over 1.5 million quality estimates available at the SITC 4-digit level (after
aggregating over importers and units of measurement)
– BEC 3, 2, 1 digit
• BEC1: Useful breakdown into intermediate products, capital goods, and
consumer goods
• BEC2: Distinguishes, for instance, between (i) primary vs. processed
varieties; and (ii) consumer durables vs. non-durables.
– 3 broad custom categories
• Manufactures, Agriculture, and Non-Agricultural Commodities
Stylized Facts: Quality Upgrading
Quality upgrading is a crucial component of development,
particularly when trying to move to upper middle-income status Some LICs need diversification, others need quality upgrading.
0.2
.4.6
.81
90th
perc
entile
= 1
0 10000 20000 30000 40000Exporter GDP per capita (2000 constant Dollars)
HIC MIC
LIC Lowess Fit
Quality across all Exports
Stylized Facts: Quality Upgrading Manufacturing and Agriculture
There seems to be potential to also quality upgrade in agriculture, though it may be more constrained by soil and climate conditions.
.2.4
.6.8
11.2
90th
perc
entile
= 1
0 10000 20000 30000 40000Exporter GDP per capita (2000 constant Dollars)
HIC MIC
LIC Lowess Fit
Quality in Manufacturing Exports
.2.4
.6.8
1
90th
perc
entile
= 1
0 10000 20000 30000 40000Exporter GDP per capita (2000 constant Dollars)
HIC MIC
LIC Lowess Fit
Quality in Agricultural Exports
Stylized Facts: Quality Upgrading Asia: considerable cross-country heterogeneity
Some countries have converged or are continuing to converge to the world frontier.
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
19
63
19
66
19
69
19
72
19
75
19
78
19
81
19
84
19
87
19
90
19
93
19
96
19
99
20
02
20
05
20
08
Japan
Korea
China
Vietnam
Qu
alit
y (
90
th p
erc
en
tile
= 1
)
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
19
63
19
66
19
69
19
72
19
75
19
78
19
81
19
84
19
87
19
90
19
93
19
96
19
99
20
02
20
05
20
08
Malaysia
Thailand
India
Indonesia
Bangladesh0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
19
63
19
66
19
69
19
72
19
75
19
78
19
81
19
84
19
87
19
90
19
93
19
96
19
99
20
02
20
05
20
08
Japan
Korea
China
Vietnam
Qu
alit
y (
90
th p
erc
en
tile
= 1
)
0.4
0.5
0.6
0.7
0.8
0.9
1
1.1
19
63
19
66
19
69
19
72
19
75
19
78
19
81
19
84
19
87
19
90
19
93
19
96
19
99
20
02
20
05
20
08
Malaysia
Thailand
India
Indonesia
Bangladesh
In other countries, convergence seem to have slowed since the mid-1990s. Fast-convergence countries Slow-convergence countries
In Africa, some countries show promisse.
Other countries’ export quality has continued to stagnate.
.4.5
.6.7
.8.9
1
Qu
ality
(9
0th
pe
rcen
tile
= 1
)
1960 1970 1980 1990 2000 2010Year
Nigeria Tanzania
Uganda South Africa
Egypt Senegal
Fast-convergence countries
.4.5
.6.7
.8.9
1
Qu
ality
(9
0th
pe
rcen
tile
= 1
)
1960 1970 1980 1990 2000 2010Year
Morocco Ghana
Cote d'Ivoire Cameroon
Kenya
Slow-convergence countries
Stylized Facts: Quality Upgrading Africa: considerable cross-country heterogeneity
Potential for Quality Upgrading Destination markets are no constraint for LICs
Quality demanded in destination markets is not an apparent constraint. Policy should focus instead on creating conditions broadly favorable to quality upgrading
0.6 0.7 0.8 0.9 1 1.1
in Africa
in Asia
Low income exporters
Lower middle income exporters
Upper middle income exporters
High income exporters
Average quality demanded in destination
markets
Quality exported
Export quality relative to destination markets in 2009 (World Frontier=1)
Average of countries for which convergence seemingly stalled:
Quality Upgrading and Growth
Countries experiencing faster quality convergence since 1995 also experienced faster growth in GDP per capita
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
LICs and Lower Middle Income (incl. China)
LICs and Lower Middle Income (excl. China)
African LICs only
Before 1996
1996-2010
Additional per capita growth in fast quality convergers relative to slow convergers (percentage points)
Drivers
Diversification and quality are driven by deep,
underlying determinants, including:
Institutions
Labor and product market rigidities
Public infrastructure, and investment efficiency
Education (human capital)
Financial development
Case Studies
Malaysia
Bangladesh
Vietnam
Tanzania
Tanzania
0
5
10
15
20
25
30
35 Exports of Goods and Services
(percent of GDP)Other
Tourism
Manufaturing
Minerals
Traditional exports
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
Geographic Distribution of Exports
China India South Africa
EAC EU Switzerland
Other
• Exports grew rapidly
• Agriculture was ‘crowded out’ by
gold and manufacturing
• Trade geography also changed
• EU’s share declined, giving way to
Switzerland and China (reflecting gold)
and regional trade
Tanzania showcases successful LIC diversifier
Tanzania
Output concentration decreased, reflecting diversification away
from agriculture to manufacturing, construction, trade, and
communications.
Potential for Both Quality Upgrading
and Horizontal Diversification Tanzania
Given its concentration in agricultural products & crude materials, Tanzania has potential for horizontal diversification but also quality upgrading in agriculture
Case Studies: Some Takeaways
• Successful diversification typically underpinned by reforms
that are general in scope, rather than industry-focused and
narrowly targeted.
• Necessary reforms come in “waves” of interrelated
measures, and depend on political economy considerations
and external conditions.
Conclusions
Diversification closely linked with structural transformation, and an important element of the development process in developing economies.
Development strategies must promote sustained resource reallocation, and encourage continued quality upgrading.
Key issues: What goods are produced and exported, and of what quality?
What are the specific bottlenecks to structural transformation
in given countries? Why are some countries much more
successful?
Are the policies needed to promote initial diversification
different from those required to then sustain quality upgrading?
Working Papers Economic Diversification in Low-Income Countries: Stylized
Facts, C. Papageorgiou and N. Spatafora (February 2013)
Export Quality in Developing Countries, C. Henn, C. Papageorgiou, and N. Spatafora, (May 2013)
Economic Diversification: Recent Experiences and Policy Lessons from Five Case Studies, A. Pitt (forthcoming)
Diversification and Structural Transformation in Frontier Asian Economies, C. Papageorgiou, Nikola Spatafora and Ke Wang (forthcoming)
Diversified Structural Transformation (model)
Growth through Diversification
Diversifying exports and Macro Stability
Export Quality and Complexity
Work in Progress
THANK YOU
Quality Upgrading: Method
• Quality estimates improve on unit values, which can be
imperfect quality proxies if:
• Significant cross-country differences in production
costs
• Selection bias in quality composition (“Shipping the
good apples out” effect)
• Estimated quality measures smooth volatility and
dispersion inherent in unit values.
Diversification Episodes Trade Data: Products
.1.2
.3.4
.5
He
rfin
dah
l in
de
x
1960 1970 1980 1990 2000 2010year
Div. episode (1969-2000) Herfindahl index
HP Break
Chile
.05
.1.1
5.2
.25
He
rfin
dah
l in
de
x
1960 1970 1980 1990 2000 2010year
Div. episode (1962-91) Div. episode (2004-08)
Herfindahl index HP
Break
Malaysia
0
.05
.1.1
5.2
He
rfin
dah
l in
de
x
1960 1970 1980 1990 2000 2010year
Div. episode (1962-89) Herfindahl index
HP Break
Thailand
.1.2
.3.4
.5.6
He
rfin
dah
l in
de
x
1960 1970 1980 1990 2000 2010year
Div. episode (1983-97) Herfindahl index
HP Break
Ghana
.05
.1.1
5.2
.25
He
rfin
dah
l in
de
x
1960 1970 1980 1990 2000 2010year
Div. episode (1977-95) Herfindahl index
HP Break
Madagascar
.2.4
.6.8
1
He
rfin
dah
l in
de
x1960 1970 1980 1990 2000 2010
year
Div. episode (1967-72) Div. episode (1978-86)
Herfindahl index HP
Break
Mauritania