12
Neil Foster, Robert Stehrer, Marcel Timmer, Gaaitzen de Vries WIOD conference, 24-26 april 2012 Groningen
Neil Foster, Robert Stehrer, Marcel Timmer, Gaaitzen de Vries
WIOD conference, 24-26 april 2012
Groningen
Local and global value chains (1st & 2nd unbundling) ◦ From made in [country] to: Made in the World
Trade statistics are revealing ◦ Growth in trade versus growth in GDP ◦ Most trade is in intermediates nowadays
Theory ◦ Trade in Activities (Feenstra and Hanson, 1996;
Grossman and Rossi-Hansberg, 2008) Why? ◦ Information and Communication Technology ◦ Declining trade barriers, regional agreements ◦ Opening up of China and India
Measures global production sharing: ◦ Over time, from 1995-2009 ◦ Viz. GTAP, relevant improvements in the WIODatabase
for measuring vertical specialization ◦ Extends production sharing measures to distinguish
production factors Finds: ◦ Increasing vertical specialization Trend is robust to a host of controls, such as regional trade
agreements and levels of economic development ◦ Regional shifts in foreign earnings ◦ Positive relation between GDP per capita and skill
content contribution in value chains
International production fragmentation requires new measures of international trade (WTO, OECD, DGTrade)
Analysis based on gross export and gross import data leads to controversial conclusions (e.g. Rodrik, 2006 vs Krugman, 2008)
Recently, trade economists have started to measure the export of value added (e.g. Trefler and Zhu, 2010; Johnson and Noguera, 2012; Bems et al., 2011) ◦ Definition: the amount of value added produced in a
given source country that is ultimately embodied in final goods absorbed abroad
Define number of countries C, industries S and Factors F
A = Intermediate input coefficients (CSxCS)
Total inputs required per unit of final demand is given by
L = (I-A) -1
Where L is the Leontief inverse, which is a matrix
indicating the output used both directly and indirectly to produce final goods
To measure the output contribution of countries in international production sharing (Johnson and Noguera, 2012):
va_exp = diag(r) (I –A)-1 cj
where r is the ratio of value added to output for each sector in each country, and cj is the
vector of final demand for country j Extension to production factors from shares in
value added
Time series analysis based on national accounts data
Intermediate use is broken down into domestic and foreign origin using detailed trade statistics
A standardised database of bilateral services flows
Socio-economic accounts for large developing countries, such as China and India
-0.05
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
0.50
Euro
pean
Uni
on 2
7
Unite
d St
ates
Rest
of t
he W
orld
Japa
n
Chin
a
Ger
man
y
Fran
ce
Unite
d Ki
ngdo
m
Italy
Spai
n
Russ
ia
Braz
il
Cana
da
Indi
a
Mex
ico
Aust
ralia
Sout
h Ko
rea
Net
herla
nds
Turk
ey
Indo
nesi
a
Pola
nd
Belg
ium
Swed
en
Taiw
an
Aust
ria
Gre
ece
Den
mar
k
Finl
and
Irela
nd
Port
ugal
1995 value added export to GDP ratio change from 1995 to 2008
30 3 14 29 12 12
32 1 21 33 4 9
28 5 7 22 7 31
36 4 13 23 2 22
21 2 11 20 10 36
22 1 23 25 5 23
21 2 8 8 12 49
24 1 16 2 9 48
16 2 18 19 10 35
18 1 28 7 6 39
54 3 5 3 9 26
59 4 7 1 10 19
10 6 16 9 12 47
6 12 23 3 12 44
6 18 10 14 52
13 25 3 13 47
0 20 40 60 80 100Percentage
8. Rest
7. BRIIMT
6. China
5. US
4. East As
3. EU12
2. EU15
1. EU27
2008199520081995200819952008199520081995200819952008199520081995
EU15 EU12 East Asia US
China BRIIMT Rest
0.1
.2.3
.4Lo
w-s
kille
d V
AX
0 10000 20000 30000GDP per capita
0.0
5.1
.15
.2H
igh-
skill
ed V
AX
0 10000 20000 30000GDP per capita
New measures of trade are needed for the modern international economy
This paper extends the export of value added in time and across production factors
Main conclusions: 1. Increasing vertical specialization 2. Regional shifts in foreign earnings 3. Positive relation between GDP per capita
and skill content contribution in value chains
Thanks [email protected]
