Does It Matter What We
Export? The Quality of Trade
D. Lederman & W. F. Maloney,
DECRG and LAC CEO, World Bank
Why Quality? It is the quality of our work that will please God, and
not the quantity”- Mahatma Gandhi
Translation: X growth, (X+M)/GDP aren’t enough
“Quality is independent of and prior to intellectual
abstractions” Robert Pirsig, Zen and the Art of Motorcycle
Maintenance
Correction: All we are saying is give IP a chance, but
base it on solid analytical principles
“I consider a bad bottle of beer to be a personal insult
to me”- Freddy Heineken
Translation: It ain’t just the what, but the how
Today’s Presentation
The “good” as the unit of analysis
Quality as price heterogeneity within goods
Should a good be the unit of analysis?
Jobs and “brainy” goods
Quality as the basket of exports
THE GOOD AS UNIT OF
ANALYSIS
Why might standard price signals be
deceptive in choosing goods
Marshallian externalities
local externalities that lead productivity to rise with
the size of the industry
local industry level knowledge spillovers, input-
output linkages, and labor pooling, for instance.
Rents
Marshallian Externalities
Intervention warranted to shift to good with
externalities against price signals.
Harrison and Rodriguez-Clare (2007):
Problems
Measurement difficult
Caveat: If Colombia can exploit these, so can/did the
US--P* reflects this.(Rodriguez critique)
Caveat to caveat:
Interindustry spillovers (Tyson-Intel in Israel… )
Assymetries (…vs Silicon valley)
Rents
e.g. Increasing returns to scale (Krugman)
Also, often tough to quantify
Caveat 3:
Gov’t actions could offset gains
Baldwin- what if ME, rents are not intrinsic to
good, but how produced?
Strategy: focus on characteristics
thought correlated with “good” things
High productivity goods
Rich Country Goods (Rodrik, Hausmann)
High tech (Lall) high inter-industry ME
Natural resources
Low productivity (Smith, Matsuyama, Sachs), few
ME
Rent seeking
HIGH PRODUCTIVITY GOODS
Does It Matter What We Exports?
Hausmann, Hwang, Rodrik (2007)
Model- broadly inter-industry spillover
Country should produce the highest productivity
good within its CA
Empirics:
PRODY, EXPY
Similar to Lall (2000)
Find higher EXPY correlated with higher growth.
Caveats
Rodriguez critique?
Rents- higher where rich countries already
are?
Not generally the case- Nokia and TVs
If easy to move into these goods, then barriers to
entry/rents low
Ditto MNCs permanence
Entrepots: Highest PRODY 2001:
“Asses, mules and hinnies, live”
Actually, no neat breakdown of
rich/poor country goods
0
5000
10000
15000
20000
25000
30000
35000 PRODYs (with +/- 1 SD*)
Empirically, some support for
MODELGrowth Regressions
Base: HHR
Regressions
Including the Export
Herfindahl and the
Investment Share
With Income Average
Value
Including the Export
Herfindahl and the
Investment Share
IV GMM IV GMM IV GMM IV GMM
Log ( initial gdp) -0.0382*** -0.0203** -0.0414* -0.0177 -0.0166* -0.0177 -0.028 0.0215
(0.01) (0.01) (0.02) (0.01) (0.01) (0.04) (0.02) (0.03)
Log (expy) 0.0925*** 0.0532** 0.107 -0.00687 0.102*** 0.0504** 0.124 0.00275
(0.02) (0.02) (0.07) (0.03) (0.02) (0.02) (0.08) (0.03)
Category Log (expy) -0.0577*** -0.00566 -0.0431 -0.119
(0.02) (0.10) (0.03) (0.08)
Log (primary schooling) 0.00468* 0.00565 0.00271 0.0101 0.00394 0.00582 0.00207 0.00958
(0.00) (0.01) (0.00) (0.01) (0.00) (0.01) (0.00) (0.01)
Log (Investment Share) 0.0111* 0.0360** 0.00935 0.0566***
(0.01) (0.02) (0.01) (0.02)
Root Herfindal Index 0.0551 -0.0381 0.0615 -0.0283
(0.06) (0.04) (0.06) (0.04)
Constant -0.426*** -0.250* -0.572 0.14 -0.186* -0.199 -0.449 0.699
(0.10) (0.13) (0.44) (0.18) (0.10) (0.47) (0.40) (0.46)
Observations 285 285 285 285 285 285 285 285
Number of wbgroup 75 75 75 75
Regressions include decade dummies
Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1
A final note on Monkeys and Trees
Being a tree in a dense area is like a ME-
subject to Rodriguez critique
If easy to jump from one tree to others, then
easy to jump to, i.e, no barriers to entry and
rents
Frontier goods (high PRODY) on the edge of the
forest?
Is past a good predictor?
iPhone didn’t exist, Saab already does
Would Chilean forestry produce Saab?
CURSED GOODS: NATURAL
RESOURCES
10/2/2009
NR abundant success stories: “β” countriesLo
g G
DP
per
cap
ita 1
990
Log Natural Resources (Leamer)-11.5041 11.7949
6
7
8
9
10
Algeria
Argentin
AustraliAustria
Banglade
Benin
Bolivia
Brazil
Burkina Burundi
Cameroon
Canada
Cape Ver
Chad
Chile
China
Colombia
Comoros
Congo, D
Costa Ri
Cyprus
Denmark
Dominica
Ecuador
Egypt, A El Salva
Fiji
FinlandFrance
Gabon
Gambia,
Germany
Ghana
Greece
Guatemal
GuineaGuinea-B
Guyana
Honduras
Hong Kon
Hungary
Iceland
India
Indonesi
Iran, Is
IrelandIsrael
Italy
Cote d'I
Jamaica
Japan
Jordan
Kenya
Korea, R
Madagasc
Malawi
Malaysia
Mali
Mauritan
MauritiuMexico
Morocco
Mozambiq
NetherlaNew Zeal
Nicaragu
Nigeria
Norway
Pakistan
Panama
Papua Ne
ParaguayPeru
Philippi
Poland
Rwanda
Senegal
Sierra L
South Af
Spain
Sri Lank
Sudan
SwedenSwitzerl
Syrian A
Thailand
Togo
Tunisia
Turkey
United K
United S
Uganda
Uruguay
Venezuel
Zambia
Zimbabwe
Leamer Measure: Net
Exports of NR/Worker:
Empirically, there is no resource curse
Minerals are good
Davis (1995), Sala-i-Martin et al. (2004), Stijns
(2005), Brunnschweiler (2008, 2009)
Ag has higher TFP growth than manufactures
Bernard & Jones (1996), Martin and Mitra (2001):
(also Jacob Viner and Douglass North)
Lederman & Maloney (2007, 2009)
Sachs and Warner results easily overturned
Trade proxies for endowments not clear
Sparse forest, or Monkeys with low HC?
Conditional curse?
No question about political economy
However, Sierra Leon does not negate experience
of Australia, Canada, Finland, Sweden, US…..
But central tendency is not a curse…it’s an issue
Diversification a problem
QUALITY AS PRICE
HETEROGENEITY
Export quality gaining interest in the
trade and development fields Productivity can’t explain export performance at the firm
level Brooks (2006), Hallak and Sivadasan (2009):
Need another dimension/factor (Caliber?)
Unit values rise with GDP/capita Hummels and Klenow (2005)
Variance within HS10 more important than across products-challenge to trade theory? Schott (2004)
Hwang (2007): LDCs grow slowly because they’re at the top of their respective (short) ladders
There is convergence within good categories: Hwang (2007), Krishna and Maloney (2009)
Products may matter for growth Unit Values: Drift and Standard Deviation, Country Fixed Effects, HS-1 1990-2001
woodprods (44-49)
vegetables (06-15)
transp (86-89)
textiles (50-63)
stoneglass (68-71)
plasticrubber (39-40)
misc (90-97)
minerals (25-27)
metals (72-83)
machinelec (84-85)
leatherprods (41-43)
foothead (64-67)
foods (16-24)
chemallied (28-38)
animals (01-05)
0.02
0.04
0.06
0.08
0.1
0.12
0.14
0.16
0.18
-0.2 -0.1 0 0.1 0.2 0.3
Standard Deviation (Product Dummies)
Dri
ft (
Pro
du
ct
Du
mm
ies
)
Krishna and Maloney (2009)
Unconditional divergence at country level
Figure 3: Quality Growth by Region 1990-2001
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
OECD
(high-
income)
EASIAP
(high-
income)
LAC MENA EASIAP
(low -
income)
SASIA EUROPE
(non-
OECD)
SSAFRICA CASIA
Region
Med
ian
Qu
ali
ty G
row
th
Krishna and Maloney (2009)
However, how matters as much or
more than what. Figure 4: Quality Growth by Region 1990-2001 (Product Fixed Effects
Included)
-0.014
-0.012
-0.01
-0.008
-0.006
-0.004
-0.002
0
0.002
0.004
OECD
(high-
income)
LAC MENA EASIAP
(high-
income)
SASIA SSAFRICA EASIAP
(low -
income)
EUROPE
(non-
OECD)
CASIA
Region
Med
ian
Qu
ali
ty G
row
th
Krishna and Maloney (2009)
IS A GOOD A GOOD A GOOD?
Table 2 China: 10 Exports with the Lowest Domestic Value Added
Electronic computer 4.6
Telecommunication equipment 14.9
Cultural and office equipment 19.1
Other computer peripheral equipment 19.7
Electronic element and device 22.2
Radio, television, and communication equipment 35.5
Household electric appliances 37.2
Plastic products 37.4
Generators 39.6
Instruments, meters and other measuring equipment 42.2
China: 10 Exports with the Highest Domestic Value
Added
Agriculture, forestry, animal husbandry and fishing
machinery 81.8
Hemp textiles 82.7
Metalworking machinery 83.4
Steel pressing 83.4
Pottery, china and earthenware 83.4
Chemical fertilizers 84.0
Fireproof materials 84.7
Cement, lime and plaster 86.4
Other non-metallic mineral products 86.4
Coking 91.6
Source: Koopmans, Wang, and Wei (2008).
Does China really export the iPOD?
“..the electronic components we
make in Singapore require less
skill than that required by
barbers or cooks, involving
mostly repetitive manual
operations”
Goh Keng Swee, Minister of
Finance Singapore (1972)
Technological Sophistication Embodied in
Products: Influential Views
“…Technology intensive exports imply greater development benefits to exporting countries. Therefore there is considerable interest in analyzing the technology structure of exports in developing and developed countries.” Lall, Weiss and Zhang (2005)
“…Ignoring such specialization can cloud our thinking about the responses of wages to globalization. It also interferes with our ability to identify other determinants of production such as cross-country differences in technology.” Schott (AER 2003)
The Global Computer Industry:
Is It “Sophisticated” Everywhere? A Bit of DataCountry Net-Exports Ranking K/L Ranking Skilled-L/L Ranking
Top
Net-exp
orters
China 1.24E+07 1 1.45E+07 51 38.4 37
Malaysia 1.18E+07 2 5.76E+07 27 50.5 25
Singapore 1.05E+07 3 2.03E+08 3 59.1 17
Korea Rep. 9187286 4 2.42E+08 1 75.3 5
Philippines 6350562 5 1.61E+07 48 53.6 23
Ireland 5953102 6 1.04E+08 21 64.1 15
Japan 5000000 7 1.85E+08 5 71.9 8
Mexico 4675278 8 4.48E+07 29 40.3 36
Indonesia 2329506 9 1.61E+07 49 26.8 50
India 33958 10 7649168 58 22.2 56
Costa Rica 21775 11 1.99E+07 44 29.9 44
Top
Net-im
po
rters
Austria -1029426 63 1.65E+08 6 70.1 11
Denmark -1196473 64 1.44E+08 12 68.1 12
U.K. -1200000 65 1.11E+08 20 58.2 18
Sweden -1592865 66 1.32E+08 16 80.3 3
Spain -1613921 67 1.13E+08 18 46.9 30
Switzerland -2773254 68 2.03E+08 2 71 9
Australia -3062108 69 1.48E+08 10 73.4 6
France -3942278 70 1.52E+08 9 55.7 20
Italy -4117605 71 1.53E+08 8 46.7 31
Canada -5744931 72 1.40E+08 14 79.6 4
U.S.A -3.11E+07 73 1.60E+08 7 89.7 1
Skilled-L/L Ranking Net-exports
1-10 Korea Rep.
Japan
U.S.A
Sweden
Canada
Australia
Switzerland
11-20 Ireland
Singapore
Austria
Denmark
U.K.
France
21-30 Philippines
Malaysia
Spain
31-40 Mexico
China
Italy
Source: Cusolito and
Lederman (2009)
Who Are the “Sophisticated” Computer Suppliers?
Factor Intensities Depend on Endowments and
IPRs
Industry Regime Factor Intensities Exporting Countries
Global computer
industry
1 capital (1.90E-06) and skill (55.3) Phillipines
2 capital (1.66 E-06) China, Costa Rica, India, Indonesia
3
skill (4.05E+02) and unskill
(1.05E+02)
Ireland, Japan, Korea Rep. Mexico, Malaysia,
Singapore
Final goods computer
industry
1 skill (1.04E+03) China
2
capital (5.33 E-05) and unskill
(7.25E+02) Indonesia, Phillipines
3
capital (3.61E-04) and skill
(7.67E+04) Malaysia, Mexico
4 skill (8.65 E+04) Ireland, Japan, Netherlands, Singapore
Skill=workers with secondary; Capital=capital stock in $; Unskill=workers without secondary.
Source: Cusolito and Lederman (2009)
Country and Industrial Differences in the Skill Premium
(Returns to Schooling) of 5 Million Workers
Source: Brambilla, Carneiro, Lederman & Porto (2009, in progress)
“Brainy” Goods in LCR?
Source: Brambilla, Carneiro, Lederman & Porto (2009, in progress)
For Policy: A Simple Analysis on Countries versus Industries
Source: Brambilla, Carneiro, Lederman & Porto (2009, in
progress)
What’s Behind the Dummies?
Skills, Exports and the Wages of 5 Million Workers
National skill endowments
Income per capita
Industrial exports The incidence of exports (industries versus countries?)
Weak evidence on scope for product differentiation
Some of the
Estimations:
Source: Brambilla, Carneiro, Lederman & Porto (2009, in progress)
Export Portfolios and Volatility
0
5
10
15
20
25
30
0 10 20 30 40 50 60 70 80 90 100
Concentration Index
Terms of trade volatility
Source: Gamberoni and Newfarmer, 2009 based on authors calculation
based on World Bank, World Development Indicators, as shown in Canuto
(2009, Presentation at CEIP).
The Role of the Commodity Trade Balance (and
Natural Resources): Some Estimations
(1) (2) (3)
Dependent Variable: Terms-of-Trade Volatility Export Concentration Terms-of-Trade Volatility
Estimator: OLS OLS First Stage Second Stage
Export-revenue
concentration [Root of the
Herfindahl Index] 0.281*** 0.348***
(0.000) (0.000)
Net exports of natural
resources per worker 0.006* 0.031*** 0.004
(0.081) (0.000) (0.157)
Log(Labor Force in 1980) 0.012*** -0.044*** 0.015***
(0.001) (0.000) (0.001)
Log(GDP per capita in
1980) -0.005 -0.071***
(0.415) (0.000)
Observations 102 102 102
Notes: ***, ** and * represent statistical significant at the 1, 5 and 10 percent levels.
P-values appear inside parentheses
The results correspond to cross-sectional estimates for 1980-2005.
Source: Authors' calculations based on data described in the Appendix.
Picking Winners and Export Portfolios:
The Power (Law) of Exports
Easterly, Reshef & Schwenkenberg (2009)
Manufacture exports are highly concentrated
Relationship between the share of each product’s exports in total exports follows a power law:
The probability of hitting it diminishes exponentially with the size of the hit!
Smart and successful old-style IP would yield high export concentration
Not a desirable outcome from the Portfolio View
IP: Back to Drawing Board
Not just the what but the how
Technologically sophisticated goods might be passé IP as technology policy
Bridging the gap between private and social returns to schooling Education is not enough
Further research: Exports-Skills complementarities
Diversification of exports Not your grandma’s IP…
Fin
More on the Determinants of Volatility: Does Malik and
Temple (2009) trump Acemoglu & Zilibotti (1997)? NO.(1) (2) (3)
Dependent Variable:
Terms-of-Trade
Volatility
Export
Concentration
Terms-of-Trade
Volatility
Estimator: OLS OLS First Stage Second Stage
A. Testing Acemoglu & Zilibotti (1997)
Export-revenue concentration [Root of the
Herfindahl Index] 0.281*** 0.348***
(0.000) (0.000)
Net exports of natural resources per worker0.006* 0.031*** 0.004
(0.081) (0.000) (0.157)
Log(Labor Force in 1980) 0.012*** -0.044*** 0.015***
(0.001) (0.000) (0.001)
Log(GDP per capita in 1980) -0.005 -0.071***
(0.415) (0.000)
Observations 102 102 102
Instrument Relevance (partial R-squared) 0.218
A. Testing Acemoglu & Zilibotti (1997) with Malik & Temple’s (2009) Geography
Export-revenue concentration [Root of the
Herfindahl Index] 0.351***
(0.000)
Net exports of natural resources per worker 0.028*** 0.004
(0.000) (0.160)
Log(Labor Force in 1980) -0.057*** 0.015***
(0.000) (0.001)
Log(GDP per capita in 1980) -0.065***
(0.000)
Frankel & Romer’s (1999) Constructed
Trade Share -0.002
(0.132)
Observations 101 101
Instrument Relevance (partial R-squared) 0.236
Sargan Over Identification Test (p-values) 0.838
Notes: ***, ** and * represent statistical significant at the 1, 5 and 10 percent levels.
P-values appear inside parentheses
The results correspond to cross-sectional estimates for 1980-2005.
Source: Authors' calculations based on data described in the Appendix.
Source: Lederman and Xu (2009, in
progress)