technological upgrading in china and india: what do we ... · jaejoon woo international monetary...

UNU-WIDER conference, Hanoi, June 29-30, 2014

Technological Upgrading in China and India:

What Do We Know?†

Jaejoon Woo

International Monetary Fund and

DePaul University † Presentation is based on author’s paper, “Technological Upgrading in China and India: What Do We

Know?”, OECD Development Centre, No. 308. Views expressed here are solely those of the author, and

should not be attributed to the IMF or to the OECD.

Introduction

Emergence of China and India as major forces in the world economy is one of the most important developments in the early 21

st century.

(Until very recently) China’s economy has grown at an unprecedented rate of near 10% per year over the last 30 years. Also, average annual growth rate of India has been 6% over the same period.

Focus on the technological upgrading behind their spectacular economic growth.

From a developing economy’s perspective, technological upgrading depends on the extent of adoption/assimilation of advanced technologies that are in use in the advanced countries – technology diffusion.

Channels of technology diffusion include import of capital goods that embody

technologies, foreign direct investment (FDI), export activities, and international production network (IPN).

Take two-pronged approaches: (i) sources of technological change, and channels of tech. diffusion in a cross-country perspective; (ii) each channel in Chinese and Indian contexts (including FDI, imported capital, role of China as a final assembler in IPN, R&D and patenting activities).

0

2

4

6

8

10

12

14

161

98

0

19

81

19

82

19

83

198

4

19

85

19

86

19

87

19

88

19

89

19

90

19

91

19

92

19

93

19

94

19

95

19

96

19

97

19

98

19

99

20

00

20

01

20

02

20

03

20

04

20

05

20

06

200

7

20

08

20

09

20

10

20

11

20

12

20

13

20

14

Gro

wth

Rat

e (%

, per

yea

r)

Real GDP Growth in China and India, 1980-2014

China

India

Emerging and Developing Countries

China: average annual growth rate of real GDP in 1980-2007 is 10% , and in 2008-14, 8.8% .India: average annual growth rate of real GDP in 1980-2007 is 6%, and in 2008-14, 6.3%.Source: IMF, World Economic Outlook 2014

Technology (TFP): China and India in a Cross-Country Perspective

As a measure of technology level, let us begin with total factor productivity (TFP).

Standard aggregate production function: Y=AK(HL)1-, where =capital income

share, K=physical capital, H=human capital, A=TFP (technology)

Technological change is an important determinant of TFP. This was Robert Solow

(1957)’s original view and that of many economists (Guinet et al., 2009).

Endogenous growth models Two sources of technological change: (i) Innovation through R&D (ii) Technology diffusion via assimilating and adapting advanced foreign

technology

Recent debates on factor accumulation versus TFP:

More than half of the cross-country variation in income per capita and growth

results from differences in TFP and its growth (Hulten and Isaksson, 2007; Caselli,

2005; Parente and Prescott, 2001).

Growth Accounting and Development Accounting

Aggregate production function: Y=AK(HL)1-

ΔY/Y= ΔA/A + (ΔK/K) + (1-)( ΔL/L+ΔH/H) (1)

In terms of per worker, y=AkH1- , where y=Y/L, k=K/L

Output per worker in country i, relative to the counterpart of USA: yi/yUSA = Ai/AUSA (ki /kUSA)

α (Hi /HUSA)

1-α (2)

K (non-residential capital stock) is estimated using the perpetual inventory method:

Kt = It + (1-)Kt-1 , where It is the investment and δ is the depreciation rate.

H (human capital) is estimated using the Mincerian approach (Caselli, 2005;

Klenow & Rodriguez-Clare 1997):

H=exp((E)), where E is average years of schooling, and the function (E) is piece

linear with slope of 0.134 for E ≤ 4, 0.101 for 4 < E ≤ 8, and 0.068 for 8 < E.

-3

-2

-1

0

1

2

3

4

5

6

7

8

-3 -2 -1 0 1 2 3 4

Gro

wth

of

Lab

or

Pro

du

ctiv

ity

(Y/L

), a

vera

ge a

nn

ual

rat

e o

ver

19

70

-20

11

Growth of Total Factor Productivity (TFP), average annual rate over 1970-2011

Figure 1a. TFP Growth and Labor Productivity Growth, 1970-2011

China

Korea

MaltaBotswana

Hong KongIndia

Mauritius

MalaysiaSingapore

Taiwan

Source: Author's calculation based on Penn World Table 8

Thailand

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 0.2 0.4 0.6 0.8 1 1.2 1.4

Leve

l of

lab

or

pro

du

ctiv

ity

in 2

01

1 (

USA

=1)

Level of TFP in 2011 (USA=1)

Figure 1b. TFP and Labor Productivity Level in 2011 (USA=1)

Norway

Ireland

China

India

Botswana

Hong Kong

Korea

Mauritius

Taiwan

Luxembourg

Singapore

Source: Author's calculation based on Penn World Table 8

Thailand

Comparison of Productivity Level for Selected Countries in 2011 (USA=1)

CountryOutput per

worker (Y/L)TFP (A)

Human Capital

(H)

Capital Stock per

worker (K/L)

OECDa 0.77 0.86 0.86 0.95

Asia 0.42 0.60 0.81 0.51

Japan 0.67 0.71 0.90 1.02

Singapore 0.82 0.92 0.76 1.06

Hongkong, China 0.81 0.90 0.83 1.00

Chinese Taipei 0.69 0.92 0.89 0.61

Korea 0.59 0.68 0.92 0.80

Malaysia 0.30 0.50 0.82 0.38

Thailand 0.15 0.41 0.67 0.20

China 0.14 0.37 0.71 0.20

Philippines 0.10 0.33 0.75 0.12

Indonesia 0.10 0.38 0.57 0.10

India 0.10 0.46 0.53 0.07

Latin America 0.25 0.62 0.73 0.21

Chile 0.34 0.65 0.82 0.31

Argentina 0.39 0.64 0.78 0.46

Costa Rica 0.21 0.54 0.74 0.14

Mexico 0.34 0.72 0.76 0.29

Brazil 0.18 0.43 0.68 0.22

Colombia 0.20 0.48 0.69 0.23

Sub-Sahara Africa 0.09 0.42 0.56 0.09

South Africa 0.23 0.58 0.73 0.18

Notes: a. The average of 23 OECD member countries as of 1990 (excluding Turkey that is still

considered as emerging market): Australia, Austria, Belgium, Canda, Denmark, Finland,

France, Germany, Greece, Iceland, Ireland, Italy, Japan, Netherlands, New Zealand, Norway,

Portugal, Spain, Sweden, Switzerland, United States, United Kingdom.

Source: Author's calculation using data from Penn World Table 8.0.

Decomposition

Technology (TFP): China and India in a Cross-Country Perspective

While TFP levels are still low, China and (to a lesser degree) India have posted

strong TFP growth in the last decade.

China:

Output growth = 9.98% (2000-07), up from 8.38% (1970-99)

TFP growth = 3.48% (2000-07), up from 2.72% (1970-99)

Physical capital growth contribution =5.54% (2000-07), up from 3.56% (1970-99)

While the industry- and capital-intensive growth pattern is well-known, strong TFP

growth is a key feature of the Chinese economy, in contrast with the earlier growth

patterns of NIEs (Young 1995).

India:

Output growth = 6.98% (2000-07), up from 4.57% (1970-99)

TFP growth = 1.47% (2000-07), up from 0.56% (1970-99)

Physical capital growth contribution =3.83% (2000-07), up from 1.52% (1970-99)

Growth Accounting for Selected Countries: 2000-07

Country

Growth of

Output (Y) (%

per annum)

TFP (A)

Physical

Capital Stock

(K)

Labor (L) and

Human Capital (H)

Growth of output

per worker (Y/L)

(% per annum)

2000-07 Period

OECDa 2.84 0.24 1.33 1.26 1.28

Asia 4.82 1.54 2.21 1.07 3.15

Japan 1.50 0.78 0.63 0.10 1.77

Singapore 6.11 2.08 2.69 1.33 3.70

Hongkong, China 5.11 2.45 1.78 0.88 4.08

Chinese Taipei 4.30 0.77 2.48 1.05 3.14

Korea 5.05 0.95 2.80 1.31 3.24

Malaysia 5.40 1.78 2.09 1.53 3.30

Thailand 5.13 2.32 1.37 1.44 2.73

China 9.98 3.48 5.54 0.96 9.02

Philippines 4.78 1.30 2.43 1.05 2.55

Indonesia 4.93 1.22 2.69 1.02 3.58

India 6.98 1.47 3.83 1.68 4.70

Latin America 3.93 0.60 1.74 1.58 1.30

Chile 4.63 -0.52 3.38 1.78 1.42

Argentina 3.19 1.38 0.94 0.86 1.35

Costa Rica 4.73 0.31 1.87 2.56 0.88

Mexico 2.99 -0.44 2.25 1.19 1.16

Brazil 3.47 0.16 1.12 2.19 0.64

Colombia 4.23 0.88 1.75 1.59 1.60

Sub-Sahara Africa 3.69 0.31 1.37 2.01 0.94

South Africa 4.22 0.86 1.57 1.78 1.43

Notes: a. The average of 23 OECD member countries as of 1990 (excluding Turkey that is stil l





Contributed to output growth by

Growth Accounting for Selected Countries: 1970-99

Country

Growth of

Output (Y) (%

per annum)

TFP (A)

Physical

Capital Stock

(K)

Labor (L) and

Human Capital (H)

Growth of output

per worker (Y/L)

(% per annum)

1970-99 Period

OECDa 2.98 0.60 1.33 1.05 2.03

Asia 5.63 0.54 3.35 1.70 3.13

Japan 4.30 0.94 2.59 0.77 3.61

Singapore 7.81 0.46 5.32 2.04 4.00

Hongkong, China 6.37 1.13 3.53 1.71 3.75

Chinese Taipei 7.71 0.89 4.81 2.00 5.15

Korea 8.05 1.51 4.17 2.37 5.42

Malaysia 7.11 0.39 3.96 2.77 3.76

Thailand 7.11 1.21 4.66 1.23 5.24

China 8.38 2.72 3.56 2.10 5.83

Philippines 3.39 -1.28 2.89 1.77 0.31

Indonesia 6.05 0.36 4.02 1.67 3.33

India 4.57 0.56 1.52 2.48 2.07

Latin America 3.35 -0.51 1.99 1.86 0.65

Chile 3.79 -0.01 2.43 1.36 1.58

Argentina 1.98 -0.25 1.25 0.98 0.80

Costa Rica 4.49 -0.33 1.80 3.03 0.93

Mexico 3.87 -0.79 2.47 2.18 0.15

Brazil 4.32 -0.23 2.39 2.15 1.75

Colombia 3.91 -0.33 2.20 2.04 0.96

Sub-Sahara Africa 3.29 -0.55 1.44 2.19 -0.04

South Africa 2.24 -1.65 1.41 2.48 -0.84

Notes: a. The average of 23 OECD member countries as of 1990 (excluding Turkey that is stil l





Contributed to output growth by

Technological Structure of Export

Technology employed in the production manifests in quality and variety of goods and services, which may be observable from the export structure.

What is striking about the impressive growth of China and India is that they export products associated with a high productivity level that is much higher than a country at their income levels (Rodrik, 2006; Schott, 2008).

China has been increasingly diversifying its exports into complex, capital- and

technology-intensive products.

China has been pursuing a two-pronged strategy, rather than pursuing an export-growth strategy based on its seeming comparative advantage in low-skill and labor-intensive products.

While capitalizing on its abundant labor by promoting job-creating labor-intensive manufactures, they also pursue rapid upgrading of their economy by producing and exporting higher-technology products.

China

Costa Rica

Hong Kong, China

India

Ireland

Korea

Malta

Mexico

Philippines

Singapore

Thailand

United States

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

4 5 6 7 8 9 10 11 12

Hig

h-T

ech

Exp

ort

(sh

are

in t

ota

l exp

ort

) in

200

8

Log of Real GDP per capita in 2008 (constant 2000 US$)

Figure 2. High-Tech Export and GDP per capita in 2008

Source: Author's calculation based on UN COMTRADE (2009) and WDI (2009)

Technological Classification of Export

A popular methodology to classify products by technology level (Lall, 2000; OECD 1994) is distinguish (a) primary products, (b) resource-based manufactures, (c) low-tech manufactures, (d) medium-tech manufactures, and (e) high-tech manufactures. Provide valuable information on different levels of technology used in disaggregated export activities and their upgrading over time.

We use export data from UN COMTRADE at the 3-digit SITC, Rev.3. Some drawbacks: classification at this 3-digit level sometimes puts together

products at the different levels of technological complexity under the same category. e.g., telecommunication apparatus includes highly advanced mobile telephone technology as well as a simple plastic telephone set. Plus, a rise of IPN makes the technological classification complicated. Aggregate trade statistics does not tell about the process involved in the IPN across different locations (return to this later).

World Leading 15 Exporters of High- Medium- and Low-Technology Products

High-Tech Medium-Tech Low-Tech

Year 2007

ranking

Country Export (US$ mil)

Share (%)


Share (%)


Share (%)

1 China 409663 15 Germany 610066 15 China 384474 19 2 United States 311634 12 United States 411103 10 Germany 170388 9 3 Germany 236260 9 Japan 394413 10 Italy 133030 7 4 Hong Kong 154828 6 China 280454 7 United States 114216 6 5 Japan 149454 6 Italy 214286 5 Hong Kong 95087 5 6 Singapore 141202 5 France 203565 5 France 79670 4 7 Korea 123216 5 United Kingdom 150728 4 Belgium 66352 3 8 France 110759 4 Korea 149775 4 United Kingdom 60152 3 9 Taiwan 105678 4 Belgium 142932 4 Japan 59785 3 10 Netherlands 103404 4 Canada 122496 3 Netherlands 45874 2 11 United Kingdom 79066 3 Netherlands 110402 3 Taiwan 43639 2 12 Malaysia 72529 3 Spain 104734 3 Spain 40519 2 13 Belgium 69451 3 Mexico 97099 2 Korea 40239 2 14 Mexico 63951 2 Hong Kong 69831 2 Turkey 38144 2 15 Switzerland 50764 2 Taiwan 64103 2 India 37797 2 Total Above 2181860 81 Total Above 3125988 77 Total Above 1409365 71 World Total 2688522 World Total 4066613 World Total 1973202

1995 1 United States 164090 17 Germany 241379 15 Germany 78242 9 2 Japan 138202 15 Japan 224550 14 Hong Kong 74554 9 3 Germany 80115 8 United States 195911 12 Italy 74387 9 4 Singapore 63115 7 France 100861 6 China 69525 8 5 United Kingdom 62163 7 Italy 92041 6 United States 60473 7 6 France 52763 6 United Kingdom 75903 5 France 44412 5 7 Hong Kong 38437 4 Canada 67109 4 Japan 37887 4 8 Korea 38391 4 Belgium/Luxembourg 61319 4 United Kingdom 35485 4 9 Taiwan 36899 4 Korea 43920 3 Taiwan 33272 4 10 Netherlands 31743 3 Netherlands 42723 3 Belgium/Luxembourg 30653 4 11 Malaysia 30235 3 Hong Kong 41892 3 Korea 27599 3 12 Italy 22542 2 Spain 38651 2 Netherlands 22557 3 13 China 19350 2 Switzerland 34340 2 Spain 15521 2 14 Canada 18716 2 Mexico 31506 2 Thailand 15192 2 15 Mexico 16366 2 Taiwan 30220 2 Canada 14794 2 Total Above 813127 86 Total Above 1322324 84 Total Above 634555 75 World Total 944710 World Total 1570713 World Total 846480

Source: Author’s calculation based on UN COMTRADE (2009), SITC Rev 3. See Box 2 for technological classification.

Leading 15 Exporters of High- Medium- and Low-Technology Products

among Non-Advanced Economies

High-Tech Medium-Tech Low-Tech

Year 2007

ranking


Share (%)


Share (%)


Share (%)

1 China 409663 31 China 280454 22 China 384474 38 2 Hong Kong 154828 12 Korea 149775 12 Hong Kong 95087 9 3 Singapore 141202 11 Mexico 97099 7 Taiwan 43639 4 4 Korea 123216 9 Hong Kong 69831 5 Korea 40239 4 5 Taiwan 105678 8 Taiwan 64103 5 Turkey 38144 4 6 Malaysia 72529 6 Poland 55468 4 India 37797 4 7 Mexico 63951 5 Singapore 54225 4 Poland 30738 3 8 Thailand 37989 3 Czech Rep. 52068 4 Mexico 28344 3 9 Philippines 31946 2 Thailand 45513 4 Czech Rep. 24897 2 10 Hungary 27328 2 Brazil 39666 3 Thailand 23176 2 11 Czech Rep. 24525 2 Turkey 38038 3 Vietnam 18168 2 12 Poland 13418 1 Hungary 36376 3 Singapore 17502 2 13 Brazil 11516 1 Russia 31536 2 Indonesia 17076 2 14 Slovak Rep. 9825 1 Malaysia 28608 2 Malaysia 17058 2 15 India 8771 1 Slovak Rep. 25844 2 Brazil 14050 1 Total Above 1236320 95 Total Above 1068560 82 Total Above 830390 82

1995 1 Singapore 63115 23 Korea 43920 15 Hong Kong 74554 21 2 Hong Kong 38437 14 Hong Kong 41892 14 China 69525 20 3 Korea 38391 14 Mexico 31506 10 Taiwan 33272 9 4 Taiwan 36899 13 Taiwan 30220 10 Korea 27599 8 5 Malaysia 30235 11 China 27687 9 Thailand 15192 4 6 China 19350 7 Singapore 23540 8 India 11798 3 7 Mexico 16366 6 Malaysia 14304 5 Mexico 11702 3 8 Thailand 13790 5 Brazil 12094 4 Turkey 10474 3 9 Israel 3599 1 Thailand 9423 3 Indonesia 10249 3 10 Philippines 2852 1 Czech Rep. 7166 2 Singapore 8576 2 11 Indonesia 1790 1 Poland 5566 2 Poland 7626 2 12 Czech Rep. 1785 1 Indonesia 4806 2 Malaysia 7396 2 13 Brazil 1611 1 South Africa 4578 2 Brazil 6951 2 14 India 1433 1 Turkey 4066 1 Czech Rep. 6713 2 15 Hungary 1410 1 Israel 3956 1 Pakistan 6043 2 Total Above 271063 97 Total Above 264724 88 Total Above 307671 87

Note: Advanced economies are defined to include the OECD member nations as of 1990, excluding Turkey that is conventionally classified as an emerging economy. Thus, non-advanced economies include some of the current OECD members such as Korea, Mexico, Czech Rep, and Slovak Rep., etc. Source: Author’s calculation based on UN COMTRADE (2009), SITC Rev 3. See Box 2 for technological classification.

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

Shar

e in

To

tal E

xpo

rt

Source: Author's calculation based on the UN COMTRADE database 2009

Figure 3a. Composition of Export Products in China

1995 2007

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

Shar

e in

To

tal E

xpo

rt

Source: Author's calculation based on the UN COMTRADE database 2009

Figure 3b. Composition of Export Products in India

1995 2007

Technological Classifications

To what extent is the technological sophistication of export products associated

with TFP?

We construct an index of technological sophistication (ITS), which is higher the

greater the percentage of each country’s exports in the more technologically

advanced categories.

Specifically, the index is obtained by assigning a value to each category as follows:

1 for primary products (PP),

2 for resource-based manufactures (RB1, RB2),

3 for low-technology products (LT1, LT2),

4 for medium technology (MT1, MT2, MT3),

5 for high-technology (HT1, HT2).

The percentage of exports in each category is multiplied by the assigned value, and

these are summed and divided by 100. The ITS ranges from 1 to 5.

Index of Technological Sophistication (ITS) for Selected Countries in 1995 and 2007

Rapid rise in China’s export of medium- and high-technology products is reflected by an increase in

the ITS score to 3.75 in 2007 from 3.13 in 1995. By contrast, export products of India are still much

less sophisticated, and ITS score has little changed over 1995-2007.

Country

Index of Technological Sophistication in 1995

Index of Technological Sophistication in 2007

OECD 2.92 2.96 Asia (except Japan) 3.09 2.95

China 3.13 3.75 Hong Kong 3.53 3.95 India 2.50 2.61 Indonesia 2.19 2.22 Japan 3.98 3.69 Korea 3.78 3.88 Malaysia 3.58 3.47 Philippines 1.93 4.11 Singapore 3.98 3.68 Taiwan 3.80 3.94 Thailand 3.16 3.34

Latin America 1.98 2.16 Argentina 2.05 2.06 Brazil 2.53 2.49 Chile 1.55 1.58 Colombia 1.81 2.07 Costa Rica 1.66 3.11 Mexico 3.37 3.25 Peru 1.45 1.53

Sub-Saharan Africa 1.62 1.82 Mauritius 2.74 2.75 South Africa 1.82 2.44

Source: Author’s calculation based on UN COMTRADE (2009) database.

Relatively strong positive correlation between ITS and TFP. But for China and

India, TFP levels are much lower than the ITS scores would suggest!

Chile

India

Mauritius

Costa Rica

Mexico

Ireland

Thailand

United States

Malaysia

Singapore

China

Korea, Rep.

Taiwan

Hong Kong, China

Philippines

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5

Leve

l of

TFP

in 2

00

7 (

USA

=1

)

Index of Technological Sophistication of Export in 2007

Figure 4. Index of Technological Sophistication of Export and TFP, 2007

Fitted line: TFP=0.07+0.17*ITS, where t-statistics are in parenthes, R2 = 0.23.(0.78) (4.96)

Source: Author's calculation

Technology Diffusion via FDI

FDI is considered as an important way to access advanced foreign technology.

Beyond adding more capital to a host country, it can be a conduit to the production

technology, cutting edge of R&D, and management expert, while boosting market

competition and generating spillovers to local firms in the host economy.

China has attracted a large amount of FDI inflows since its opening to the world in

1979, and is one of the world’s largest recipients of FDI inflows.

Also, FDI into India has begun to accelerate more recently.

Primary goal of China’s FDI policies is to address its technological backwardness by promoting technology transfer to China and is trading market access in return for technology (Guiding Directory on Industries Open to Foreign Investment first promulgated in 1995). Since the mid-1990s, China has been encouraging FDI to flow into cutting-edge, technology-oriented industries such as electronics, information technology, bioengineering, new materials, and aviation and aerospace, as well as establishing local R&D centers (technology or industry parks).

0

1

2

3

4

5

6

7

0

20

40

60

80

100

120

US

do

llars

in B

illio

ns

(co

nst

an

t 2

00

9 p

rice

s)

Years after growth takeoff

Figure 5. FDI Inflows to China and India Since Growth Takeoff

China (US dollars, bn)

India (US dollars, bn)

China (% of GDP)

India (% of GDP)

Note: The growth takeoff year for China is 1979 and for India 1982China: FDI inflow (as % of GDP) peaked at 5.8% n 1994, and averaged at 3.6% over 1995-2007.India: FDI inflow (as % of GDP) averaged 1% over 1994-2007, and climbed to 3.6% in 2008.

Source: UNCTAD, Foreign Direct Investment Database (2014)

Empirical Evidence on FDI and TFP Growth

Popular view: positive effect of FDI on income growth is only contingent on the recipient country’s capability to absorb foreign technology, e.g., human capital (Borensztein et al., 1998); outward-looking trade policy (Balasubramanyan et al., 1996); financial development (Alfaro et al., 2004).

Blonigen and Wang (2005) argue that inappropriate pooling of developed countries

with developing countries is responsible for estimation of insignificant effects of FDI on income growth in earlier studies, and find positive significant effects of FDI on per capita GDP growth in developing countries only.

However, little study on the effects of FDI on technological change (measured by

TFP) at the macro level, with notable exception being Woo (2009) that presents evidence of positive direct effect of FDI on TFP growth for period of 1970-2000.

New evidence for the 1970-2007 period in a panel of 90 countries

Baseline specification:

TFPgrowthit = constant + ln(initial TFP relative to US)it + ln(human capital)it

+ ln(population)it + (government share)it + (FDI)it + Xit + ηi +it

India

Botswana

China

Mexico

Thailand

Chile

Hong Kong, China

Singapore

Mauritius

Ireland

-5

-4

-3

-2

-1

0

1

2

3

4

5

0 5 10 15 20

TFP

gro

wth

rat

e (%

per

yea

r) in

197

0-20

07

Inward FDI as a percent of GDP, average for 1970-2007

Figure 6a. TFP Growth and FDI Inflows, 1970-2007

Source: Author's calculation based on PWT 6.3 (2009) and International Finance Statistics (2009)

India

Botswana

China

Korea, Rep.

Mexico

Thailand

Chile

Hong Kong, China

Singapore

Mauritius

-5

-4

-3

-2

-1

0

1

2

3

4

5

0 1 2 3 4 5

TFP

gro

wth

rat

e (%

per

yea

r) in

197

0-20

07

Inward FDI as a percent of GDP, average for 1970-2007

Figure 6b. TFP Growth and FDI Inflows from OECD DAC Countries, 1970-2007

Source: Author's calculation based on PWT 6.3 (2009) and International Finance Statistics (2009)

Panel Regression of TFP Growth on FDI for Period of 1970-2007 (ten-year panel) Dependent Variable: TFP growth rate (% per annum)

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)

Explanatory Variables

OLSa

Robustb

Fixed-Effects Panel

c

System

GMMd

OLS Robust Fixed-Effects Panel

System GMM


System GMM

Full sample

Full sample

Full sample

Full sample

Non-advanced

Non-advanced

Non-advanced

Non-advanced

Non-advanced

Non-advanced

Non-advanced

Non-advanced

Initial TFP relative to USA (log)

-1.175*** (0.317)

-1.124*** (0.23)

-6.039*** (0.651)

-1.508*** (0.54)

-1.162*** (0.397)

-1.134*** (0.291)

-5.862*** (0.74)

-2.770*** (0.809)

-1.418*** (0.412)

-1.264*** (0.311)

-6.488*** (0.698)

-3.501*** (0.813)

Initial years of schooling (log)

0.963** (0.377)

0.797*** (0.267)

0.912 (0.876)

0.972 (0.675)

0.873* (0.459)

0.775** (0.343)

1.030 (0.960)

0.182 (0.828)

1.008** (0.488)

0.967*** (0.361)

-0.022 (1.039)

1.16* (0.644)

Initial population (log) 0.178 (0.139)

0.16 (0.104)

-2.055* (1.092)

-0.011 (0.486)

0.218 (0.180)

0.278** (0.141)

-2.273* (1.165)

0.241 (0.508)

-0.097 (0.189)

-0. 049 (0.136)

-0.901 (1.217)

-0. 332 (0.664)

Initial government size (% of GDP)

-0.024 (0.019)

-0.027 (0.017)

0.004 (0.038)

-0.131** (0.058)

-0.024 (0.019)

-0.029 (0.021)

-0.014 (0.039)

-0.164** (0.066)

-0.024 (0.022)

-0.035 (0.022)

0.005 (0.037)

-0.203** (0.079)

Terms of trade growth (percent)

0.011 (0.057)

0.027 (0.032)

0.042 (0.043)

-0.035 (0.052)

0.012 (0.063)

0.048 (0.04)

0.036 (0.050)

-0.000 (0.049)

0.005 (0.066)

0.042 (0.041)

0.032 (0.053)

0.034 (0.052)

Inward FDI (% of GDP)

0.308*** (0.074)

0.277*** (0.062)

0.338*** (0.077)

0.251* (0.131)

0.347*** (0.098)

0.358*** (0.082)

0.398*** (0.112)

0.351*** (0.118)

Inward FDI from OECD (% of GDP)

0.387** (0.182)

0. 339* (0.177)

0.37* (0.214)

0.455* (0.246)

Arellano-Bond test for AR(2), p-value

0.83 0.48 0.15

Hansen Test of Joint Validity of instruments

0.12 0.57 0.64

No. of Instruments 45 45 45 No. of Obs. 338 338 338 338 244 244 244 244 254 254 254 254 No. of countries 90 90 90 90 65 65 65 65 68 68 68 68

Note: The panel is comprised of four 10-year periods for each country, if data permit. Heteroskedasticity and country-specific autocorrelation consistent standard errors are reported in parentheses. Levels of significance are indicated by asterisks: *** 1 percent, ** 5 percent, * 10 percent. An intercept term is included in each regression. See Appendix for the list of countries included in the sample. a: Pooled OLS. Regional dummies (Asia, Latin America, Sub-Saharan Africa) and time period dummies are included. b: Robust estimation (to address leverage points and outliers). Iterated re-weighted least squares regression in which the outliers are dropped (if Cook’s distance >1) and the observations with large absolute residuals are down-weighted. c: Fixed-effects (within) panel regression. Country-specific fixed effects are controlled for (through within transformation). d: System GMM dynamic panel estimation (Arellano and Bover, 1995; Blundell and Bond, 1998). Country-specific fixed effects are controlled for (differenced out).

Micro-Data Evidence from Other Studies

A large number of micro-data (firm- or plant-level) studies on FDI, technology transfer and productivity (e.g., Harrison and Rodriguez-Clare, 2009; Moran et al., 2005; Keller, 2004).

They tend to focus on three aspects of technology transfer (spillovers) from FDI:

(i) Own-plant effect – whether firms with foreign equity participation systematically

have higher productivity or TFP than other domestic firms, (ii) Horizontal spillovers – whether foreign ownership in a sector positively affects

the productivity of domestic firms in the same sector. Such spillovers can occur through demonstration effect, labor turnover and competition effect.

(iii) Vertical spillovers (backward versus forward) – whether positive externalities

are stemming from the relationships of foreign enterprises with domestic suppliers or customers. Backward spillovers can occur if domestic suppliers to downstream foreign firms benefit from contacts with the foreign firms to increase productivity. Forward spillovers can occur if foreign firms (that are located) domestically supply inputs that embody new technologies or processes.

Micro-Data Evidence based on Chinese firm-level data

Hu and Jefferson (2002) use firm-level data from survey of large and medium

enterprises by Chinese National Statistical Bureau for 1995 and 1999. Find strong

positive own-plan effects, but insignificant (or negative) horizontal spillovers on

domestic firms within the same industry.

Lin, Lui, and Zhang (2009) use firm-level panel data for manufacturing sector for

1998-2005. Report strong positive backward and forward spillovers. As for

insignificant (net) horizontal spillovers, they distinguish between FDI from Hong

Kong, Macao, Taiwan (HMT) and FDI from non-HMT countries (mostly OECD

countries). Find HMT-invested firms generate negative horizontal spillovers, while

non-HMT foreign invested firms tend to bring positive horizontal spillovers, which

seem to cancel each other at the aggregate level.

Similarly, Du, Harrison, and Jefferson (2011) find evidence of positive FDI spillovers

via backward and forward linkages, using panel data on manufacturing firms in

China for 1998-2007.

Xu and Sheng (2012) use data set at firm level for the manufacturing sector in

2000-03, and find that FDI provides significant positive spillovers for the productivity

of firms in the same industry and largely within the same region.

Technology Diffusion via Imported Capital Goods

Technological advances, in the form of production of capital equipment and R&D activity, are highly concentrated in a handful of AEs. Most of DCs import the bulk of their machinery and equipment.

Imported capital goods that embody new technology can be a crucial mechanism for transmitting knowledge spillovers across countries.

Some earlier evidence that import of capital goods is a significant source of

technology diffusion. (e.g., Coe and Helpman, 1995; Eaton and Kortum, 2001; Woo, 2004; Almeida et al., 2007).

New evidence in a panel of 91 countries for the 1970-2007 period

Baseline specification:

TFPgrowthit = constant + ln(initial TFP relative to US)it + ln(human capital)it

+ ln(population)it + (government share)it + (Imported capital goods)it

+ Xit + ηi +it

India

China

Chile

Mauritius

Korea

ThailandTaiwan

MalaysiaHong Kong

Singapore

-8

-6

-4

-2

0

2

4

6

0 5 10 15 20 25 30 35 40

TFP

gro

wth

(% p

er y

ear)

, 197

00-2

007

Capital Goods Import from the OECD Countries (% of GDP), averaged for 1970-2007

Figure 7. Import of Capital Goods and TFP Growth

Source: Author's calculation based on PWT 6.3 (2009), Feenstra et al. (2005), and UN COMTRADE (2009)

Panel Regression of TFP Growth on Import of Capital Goods for Period of 1970-2007 Dependent Variable: TFP growth rate (% per annum)

(1) (2) (3) (4) (5) (6) (7) (8)

Explanatory Variables OLSa

Robustb

Fixed-Effects Panel

c

System

GMMd


System GMM

Full sample Full sample Full sample Full sample Non-advanced

Non-advanced

Non-advanced

Non-advanced

Initial TFP relative to USA (log) -1.333*** (0.349)

-1.152*** (0.222)

-6.548*** (0.778)

-1.819*** (0.508)

-1.111*** (0.399)

-1.105*** (0.294)

-6.109*** (0.828)

-3.15*** (0.828)

Initial years of schooling (log) 1.023*** (0.384)

0.82*** (0.260)

0.622 (0.966)

1.841*** (0.439)

0.889* (0.465)

0.841** (0.346)

0.519 (1.034)

1.275** (0.591)

Initial population (log) 0.111 (0.142)

0.111 (0.106)

-1.316 (1.173)

0.027 (0.433)

0.09 (0.201)

0.150 (0.151)

-0.983 (1.261)

0.011 (0.689)

Initial government size (% of GDP)

-0.019 (0.022)

-0.021 (0.017)

0.023 (0.045)

-0.122** (0.057)

-0.020 (0.022)

-0.032 (0.021)

-0.001 (0.043)

-0.145 (0.093)

Terms of trade growth (percent) -0.018 (0.058)

-0.007 (0.031)

0.029 (0.046)

-0.032 (0.055)

-0.004 (0.067)

0.010 (0.04)

0.032 (0.052)

0.015 (0.066)

Import of capital goods from OECD (% of GDP)

0.105*** (0.025)

0.107*** (0.031)

0.147** (0.064)

0.105** (0.044)

0.093*** (0.031)

0.099** (0.040)

0.109* (0.064)

0.211** (0.094)

Arellano-Bond test for AR(2), p-value

0.71 0.98

Hansen Test of Joint Validity of instruments

0.10 0.12

No. of Instruments 45 45 No. of Obs. 347 347 347 347 248 248 248 248 No. of countries 91 91 91 91 65 65 65 65

Note: The panel is comprised of four 10-year periods for each country, so that data permitting, each country has four observations. Heteroskedasticity and country-specific autocorrelation consistent standard errors are reported in parentheses. Levels of significance are indicated by asterisks: *** 1 percent, ** 5 percent, * 10 percent. An intercept term is included in each regression. See Appendix for the list of countries included in the sample. a: Pooled OLS. Regional dummies (Asia, Latin America, Sub-Saharan Africa) and time period dummies are included. b: Robust estimation (to address leverage points and outliers). Iterated reweighted least squares regression in which the outliers are dropped (if Cook’s distance >1) and the observations with large absolute residuals are down-weighted. c: Fixed-effects (within) panel regression. Country-specific effects are controlled for (through within transformation). d: System GMM dynamic panel estimation (Arellano-Bover 1995). Country-specific effects are controlled for (differenced out).

Astonishing increase in capital equipment imported into China from OECD countries over 1995-2008, during which the technological structure of export shifted dramatically towards high-technology categories. It increased by 368% between 1995 and 2008 (i.e., 16

th to 29

th year since growth takeoff).

For India, it went up by 205% in 1995-2008.

China became the largest buyer of industrial robots in 2013 (60% up from 2012). On average, 36% increase a year from 2008 to 2013 (International Federation of Robotics, 2014).

0

20

40

60

80

100

120

140

160

180

Billi

ons

of U

S do

llars

(in

2005

pric

es)

Note: The nominal figures were deflated by using GDP deflator for equipment (2005=1)Source: Trade Flow Data from Feestra et al. (2005) and UN COMTRADE (2009)

Figure 8. Import of Capital Equipment from OECD Countries Since Growth Takeoff

Japan Korea

China India

Years after the Growth Takeoff

Intra-Industry Trade and International Production Network (IPN)

The strongest export growth of China has been in high-tech products. How much of the rapid technological sophistication of export structure in China is

real? Is China merely assembling imported parts and components for re-export? What is the role of foreign-invested firms in the shift toward high-tech products?

Explanation has to do with the emergence of IPN (unbundling of production stages).

China has played a primary role as a final product assembler, using parts and components and semi-finished goods imported from AEs.

Processing exports account for more than 50% of China’s exports every year since 1996, and imported inputs account for around 50-80% of the value of processing exports (Koopman et al. 2008; Krugman 2008; Dean et al. 2007) Foreign-invested firms are dominant in processing exports and tend to produce much more sophisticated products than domestic firms. Their share in China’s total exports has continuously increased from 31.5% in 1995 to 58.2% in 2006 (Wang and Wei, 2010).

-40

-30

-20

-10

0

10

20

30

40

50

60

HongKong,China

EU 15 UnitedStates

Singapore Australia Vietnam Indonesia Thailand Malaysia Japan Korea Taiwan

USD

in b

illio

ns

Source: Author's calculation based on UN COMTRADE (2009) data

Figure 9. China's Trade Balance with Trade Partners on High-Tech Products (HT1), 2007

-15%

-10%

-5%

0%

5%

10%

15%

20%

Brazil China Germany Hong Kong,China

Indonesia India Japan Korea, Rep. Singapore Taiwan,China

UnitedStates

% o

f M

anu

fact

uri

ng

Trad

e

Source: Author's calculation based on UN COMTRADE (2009) data

Figure 10. Contributions to Manufacturing Trade Balance (% of Manufacturing trade), 2007

Low-Technology Medium-Technology High-Technology

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

Ch

ina

Ind

on

esia

Ph

ilipp

ines

Thailan

d

Malaysia

Singap

ore

Ko

rea

Taiwan

Ind

ia

Japan

USA

Can

ada

Mexico

Germ

any

France

Ireland

Hu

ngary

Czech

Rep

ub

lic

Perc

ent

of

Tota

l Man

ufa

ctu

res

Exp

ort

/Im

po

rt

Source: UN COMTRADE (2009)

Figure 11a. Parts and Components - Export and Import in Selected Countries, 2007

Exports Imports

0%

10%

20%

30%

40%

50%

60%

70%

Ch

ina

Ind

on

esia

Ph

ilipp

ines

Thailan

d

Malaysia

Singap

ore

Ko

rea

Taiwan

Ind

ia

Japan

USA

Can

ada

Mexico

Germ

any

France

Ireland

Hu

ngary

Czech

Rep

ub

lic

Perc

ent

of

tota

l man

ufa

ctu

res

exp

ort

/im

po

rt

Source: UN COMTRADE (2009)

Figure 11b. Finished Goods - Export and Import, 2007

Exports Imports

Upgrading Technological Capacity: R&D Efforts and Human Capital

A rapid increase in R&D intensity is consistent with the goal of upgrading the quality

and skill content of products. In China, it rose to 1% in 2000 and 1.84% in 2011

(OECD average of 2%). In India, it is getting close to 1% (2007).

The number of Chinese foreign-oriented patent families has increased rapidly (average annual growth rate of 40% in 2000-05, and of 23% since 2005), leaving behind other fast-growing economies (Brazil, Russia, India, South Africa).

China’s focus on tertiary education differs from other countries that stressed primary and secondary education at similar stages of development. The number of undergraduate and graduate students in China has been grown at approximately 30% per year since 1999 (Li et al., 2008). China has the second largest stock of human resources in science and technology (HRST) in the world, just after the United States (having pulled ahead of Japan in 2000). Nonetheless, the overall level of tertiary education attainment is still low. China and India lag behind advanced economies in terms of overall educational attainment. The number of researchers per person employed is also low, reaching only about one-tenth of Finland’s level, the highest in the world. This is also true of India.

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0 10000 20000 30000 40000 50000 60000 70000 80000 90000 100000

R&

D E

xpe

nd

itu

re, %

of

GD

P

GDP per capita (constant 2011 international dollars)

Figure 12. R&D Expenditure (% of GDP) and GDP per capita, 2011

For India, the R&D expenditure figure is for 2007.Source: World Development Indicator 2014

Israel

Japan

Korea

Germany United States

Singapore

LuxembourgIreland

China

IndiaHong Kong

Figure 13. International comparison of foreign-oriented patent families, 1970-2009

Source: WIPO (2014), International Patenting Strategies of Chinese Residents, April 23, 2014.

0

10,000

20,000

30,000

40,000

50,000

60,000

70,000

80,000

1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 2006 2009

China USA Japan Republic of Korea Germany India Brazil Russia

0

2

4

6

8

10

12

14

1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010

Ave

rage

yea

rs o

f sc

ho

olin

g o

f p

op

ula

tio

n o

f ag

e 15

+

Figure 14. Years of Schooling of Population of Age 15 and Over

Germany Japan USA Korea

India China Taiwan

India

China

Data: Barro and Lee (2010)

Researchers in R&D and Students in Science and Engineering

Country

Researchers in

R&Da

Researchers in

R&D per Million People

Enrollment Ratio in Natural Science &

Engineeringb

China 1,294,670 963 53 India 149,892 137 23.9 Japan 656,534 5,151 62.2 Korea 269,331 5,451 43

Source: World Development Indicators (2014), Science and Engineering Indicators (2014). Note: a. The figure for China is in 2011, Japan and Korea in 2010, and India in 2005. b. Ratio of undergraduate and post-graduate enrolment in natural science & engineering to the total in 2006.

Concluding Remarks

Technological level in China and India is still low, regardless how it is measured

(e.g., TFP level, domestic value-added of technically sophisticated products, or domestic skill contents of exports), indicating substantial scope for catching-up in the future. Yet, technological upgrading is taking place at a rapid pace in China and (to a lesser degree) India.

Our new evidence from a panel regression on TFP growth confirms the importance of FDI and import of capital goods in the technology diffusion.

It is beyond doubt that openness to FDI and foreign trade have played a significant role in the stellar economic performance in China and (to a lesser degree) India where the reform started a decade later.

China’s extraordinary export performance in medium- and high-technology products is closely linked to the emergence of IPN in Asia.

China has been playing a primary role as a final product assembler. This appears to be an important explanation for the puzzle of why overall technological level is much lower than technological structure of export would suggest.

Concluding Remarks

Technological upgrading is not an automatic outcome of opening to trade and FDI. China’s pattern of production and exports would have looked different if China pursued an export-growth strategy based on its seeming comparative advantage in low-skill and labor-intensive products. Deliberate efforts were made to promote technological progress through government policies: e.g. China has established a number of special economic zones (SEZs) including high technology industry zones. SEZs offer lower tax rates, simplified administrative and customs procedures, duty-free import of components and supplies and subsidized utilities. China and India have emphasized the capability to absorb technology and innovate by encouraging investment in human capital and R&D activities.

Broad lessons: diversify and upgrade domestic production structure; overcome

barriers for innovation (skills, infrastructure, financing); and encourage the private

sector’s participation in global knowledge networks and markets.

technological upgrading in china and india: what do we ... · jaejoon woo international monetary...

Documents