University of Southern California Dissertations and Theses

Structural transformation and globalization

(USC Thesis Other)

Structural transformation and globalization

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content STRUCTURAL TRANSFORMATION AND GLOBALIZATION
by
Murat
¨
Ung¨ or
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(ECONOMICS)
August 2010
Copyright 2010 Murat
¨
Ung¨ or
Dedication
to all the people I let down
ii
Acknowledgments
I wish to express my gratitude to a number of individuals who helped bring this
dissertation into existence and improve it.
I am deeply indebted to Ays ¸e
˙
Imrohoro˘ glu for her constant guidance, support, and
advice. Her passion for research and for understanding the world have been an invalu-
able example throughout my PhD experience. I have been lucky enough to benefit from
her infinite knowledge.
I owe special thanks to Caroline M. Betts. She provided not only the basic motiva-
tion for the thesis, but also lots of insightful comments throughout the progress of the
research. Her constant support, guidance and continual encouragements cannot I find
proper words to fully appreciate. This work would have not been completed without
her contributions.
I cannot overstate my debt to Robert Dekle, for his precious help and insightful com-
ments. I thank Yong Kim whose door has always open for me, and who has provided
me with an invaluable and honest advice on any matter in which I consulted with him.
I am also grateful Jeffrey Nugent and John Strauss for their support throughout the
job market process. Special thanks also go to Young Miller and Morgan Ponder for help-
ing me navigate the sometimes intricate administrative requirements at the University
of Southern California.
I have also given a large number of research seminars on my papers in this disserta-
tion. The comments and criticisms from the audiences on these occasions have helped
iii
me a great deal. The individuals are too numerous to list. However, there are per-
sons who have been especially helpful with their comments and suggestions on specific
papers, the manuscript for this dissertation, or in my collaboration with them. Their
contribution deserves a specific mention: Ahmet Akyol, Rahul Giri, Selo
˙
Imrohoro˘ glu,
Timothy J. Kehoe, Ellen R. McGrattan, Edward C. Prescott, Kjetil Storesletten, Rubina
Verma, and Kei-Mu Yi.
Last, but not least, I wish to thank my parents for putting up with my absence
abroad, and at home, and hope that they will one day look on this endeavor as worth-
while!
If this dissertation can stimulate discussion and contribute to a clearer perception of
the problems with which it deals, it will have served its purpose. I am solely responsible
for the errors that remain.
iv
Table of Contents
Dedication ii
Acknowledgments iii
List of Tables vi
List of Figures vii
Abstract ix
Chapter 1: Industrialization in Post-1978 China 1
1.1 The Structural Transformation of China, 1978-2005 . . . . . . . . . . . . . 1
1.2 Sectoral Growth Accounting and Economic Reforms . . . . . . . . . . . . 14
1.3 China’s Foreign Trade after 1978 . . . . . . . . . . . . . . . . . . . . . . . 20
Chapter 2: De-industrialization of the Riches and the Rise of China 31
2.1 Structural Transformation in Closed Economy . . . . . . . . . . . . . . . 34
2.1.1 Results for the U.S. Economy . . . . . . . . . . . . . . . . . . . . . 36
2.2 Structural Transformation in Open Economy . . . . . . . . . . . . . . . . 39
2.2.1 Quantitative Analysis . . . . . . . . . . . . . . . . . . . . . . . . . 43
2.3 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
Chapter 3: Productivity and Labor Reallocation: Latin America vs. East Asia 54
3.1 A Six-Sector Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3.2 Quantitative Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
3.3 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Bibliography 80
Appendix A 91
Appendix B 94
v
List of Tables
1.1 Growth Accounting, China, 1978-2005 . . . . . . . . . . . . . . . . . . . . 6
1.2 Total Employed Persons, Official versus Revised, million persons . . . . 8
1.3 Sectoral Growth Accounting, China, 1978-2005 . . . . . . . . . . . . . . . 18
1.4 Chinese Trade Balance in current billion US$ . . . . . . . . . . . . . . . . 25
2.1 Parameter Values for the Benchmark Economy . . . . . . . . . . . . . . . 44
3.1 Calibration for the United States . . . . . . . . . . . . . . . . . . . . . . . 64
3.2 Calibrated Sectoral Productivity Levels in 1974 . . . . . . . . . . . . . . . 66
3.3 Average Annual Growth by Sector (%), 1974-2003 . . . . . . . . . . . . . 67
3.4 Direct and Indirect Protection of Agriculture . . . . . . . . . . . . . . . . 75
A.1 Sectoral Shares of Employment and Value Added, G7 Countries . . . . . 91
A.2 Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
B.1 Country-Specific Calibration . . . . . . . . . . . . . . . . . . . . . . . . . . 96
B.2 Relative Contribution of Different Sources (%) . . . . . . . . . . . . . . . 105
B.3 Sectoral Contribution (%) . . . . . . . . . . . . . . . . . . . . . . . . . . . 106
vi
List of Figures
1.1 Growth Rates of GDP per Capita (%) : 1950-2008 . . . . . . . . . . . . . . 2
1.2 GDP per Capita as a Percentage of the U.S. . . . . . . . . . . . . . . . . . 3
1.3 Employment Shares, China, 1978-2005 . . . . . . . . . . . . . . . . . . . . 9
1.4 Real GDP Shares, China, 1978-2005 . . . . . . . . . . . . . . . . . . . . . . 11
1.5 Capital-Output Ratios, China, 1978-2005 . . . . . . . . . . . . . . . . . . . 14
1.6 TFP by Sector, China, 1978-2005 . . . . . . . . . . . . . . . . . . . . . . . . 17
1.7 China’s Share of World Manufactured Exports, 1980-2006 . . . . . . . . . 22
1.8 Effective Tariff Rates, China, 1952-2005 . . . . . . . . . . . . . . . . . . . . 23
1.9 Stages of China’s Trade Structure, 1984-2005 . . . . . . . . . . . . . . . . 26
1.10 Composition of Exports: China to the Rest of the World . . . . . . . . . . 28
1.11 Exports: China to Rest of the World, 1985-2005 . . . . . . . . . . . . . . . 28
1.12 Nominal Exchange Rate of the RMB, 1981-2005 . . . . . . . . . . . . . . . 29
2.1 Sectoral Employment Shares in the Closed Economy . . . . . . . . . . . . 38
2.2 Sectoral Employment Shares in China . . . . . . . . . . . . . . . . . . . . 45
2.3 Sectoral Employment Shares in the U.S. . . . . . . . . . . . . . . . . . . . 47
2.4 Trade Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
2.5 Counterfactuals: U.S. De-industrialization . . . . . . . . . . . . . . . . . . 49
2.6 Sensitivity Analysis in the Open Economy . . . . . . . . . . . . . . . . . . 51
3.1 GDP per Capita as a Percentage of the U.S. . . . . . . . . . . . . . . . . . 55
3.2 Labor Productivity as a Percentage of the U.S. . . . . . . . . . . . . . . . 56
3.3 Sectoral Employment Shares, 1974-2003 . . . . . . . . . . . . . . . . . . . 57
vii
3.4 Sectoral Employment Shares, U.S.: 1974-2003 . . . . . . . . . . . . . . . . 65
3.5 Employment Share in Agriculture: 1974-2003 . . . . . . . . . . . . . . . . 69
3.6 Employment Share in Mining and Quarrying: 1974-2003 . . . . . . . . . 69
3.7 Employment Share in Manufacturing: 1974-2003 . . . . . . . . . . . . . . 70
3.8 Employment Share in Utilities: 1974-2003 . . . . . . . . . . . . . . . . . . 70
3.9 Employment Share in Construction: 1974-2003 . . . . . . . . . . . . . . . 71
3.10 Employment Share in Services: 1974-2003 . . . . . . . . . . . . . . . . . . 71
3.11 Experiment: Employment Share in Agriculture . . . . . . . . . . . . . . . 73
3.12 Experiment: Employment Share in Manufacturing . . . . . . . . . . . . . 73
A.1 Employment Share of Agriculture: 1978-2007 . . . . . . . . . . . . . . . . 93
A.2 Employment Share of Industry: 1978-2007 . . . . . . . . . . . . . . . . . . 93
B.1 Sectoral Employment Shares, Hong Kong: 1974-2005 . . . . . . . . . . . 97
B.2 Sectoral Employment Shares, Korea: 1963-2005 . . . . . . . . . . . . . . . 97
B.3 Sectoral Employment Shares, Singapore: 1970-2005 . . . . . . . . . . . . 98
B.4 Sectoral Employment Shares, Taiwan: 1963-2005 . . . . . . . . . . . . . . 98
B.5 Sectoral Employment Shares, Argentina: 1950-2005 . . . . . . . . . . . . 99
B.6 Sectoral Employment Shares, Bolivia: 1950-2003 . . . . . . . . . . . . . . 99
B.7 Sectoral Employment Shares, Brazil: 1950-2005 . . . . . . . . . . . . . . . 100
B.8 Sectoral Employment Shares, Chile: 1950-2005 . . . . . . . . . . . . . . . 100
B.9 Sectoral Employment Shares, Colombia: 1950-2005 . . . . . . . . . . . . . 101
B.10 Sectoral Employment Shares, Costa Rica: 1950-2005 . . . . . . . . . . . . 101
B.11 Sectoral Employment Shares, Mexico: 1950-2005 . . . . . . . . . . . . . . 102
B.12 Sectoral Employment Shares, Peru: 1960-2005 . . . . . . . . . . . . . . . . 102
B.13 Sectoral Employment Shares, Venezuela: 1950-2004 . . . . . . . . . . . . 103
viii
Abstract
Since the pioneering empirical works of Colin Clark, Alan Fisher, and Simon Kuznets,
economists have agreed that the achievement of sustained economic growth, and a per-
manently higher level of income per capita, is strongly associated in the data with a
structural transformation. In this transformation, a substantial and permanent shift
occurs in the composition of income, output, and employment away from agriculture
and towards industry and services. This thesis comprises three essays on structural
transformation in a global world.
Chapter 1 studies the modern economic growth and structural transformation in
China. The dramatic growth of China’s role in the world economy is one of the cen-
tral topics of the current research in international macroeconomics. China’s economic
reform, which started in 1978, has driven a rapid transition of the economy from a
central planning system toward a market-oriented system integrating with the world
economy. This chapter examines the industrialization of China since 1978 and docu-
ments some of stylized facts for the structural transformation of the Chinese economy
between 1978 and 2005. I lay out a sectoral growth accounting exercise and link sectoral
productivity growths to the institutional reforms throughout period, which have had
impact on the structural transformation of China. I find that the productivity growth
in industry has been the ”engine” of aggregate productivity for the Chinese economy
since 1978.
Chapter 2 studies the impact of the industrialization of China on the industrial
employment share observed in the U.S. between 1978 and 2005. As the current wave
ix
of globalization intensifies, industrialized nations are exposed to new competition in
domestic and foreign markets. Reductions in trade costs and widespread economic
reforms may cause shifts of comparative advantage across nations which effect domes-
tic and international reallocations of production factors. A two-country, three-sector
model is developed, in which countries trade industrial goods due to the assumptions
that production of different industrial goods is country specific; and goods are differ-
entiated by the productivity with which they are domestically produced. These fea-
tures of the model and the degree of substitutability in preferences between home and
foreign produced units of industrial good endogenously determine a country’s equilib-
rium pattern of production and trade.
A comparison of the predictions of open and closed economy models suggests that a
common explanation of de-industrialization in the literature, which is based on increased
productivity in industry relative to services in a closed economy setting, is not com-
pelling. My benchmark results suggest that the closed economy model accounts for 38.1
percent of the declines in the U.S. industrial employment share while the open economy
accounts for 68.0 percent of the de-industrialization in the U.S. between 1978 and 2005.
Moreover, the open economy model has more explanatory power to explain the secu-
lar changes in the U.S. industrial employment share in the post-1990 period. The open
economy model accounts for 85.1 percent of the de-industrialization while the closed
economy accounts for 37.4 percent of the de-industrialization in the U.S. between 1992
and 2005. Counterfactual experiments show that if the Chinese economy had experi-
enced productivity in industry equal to that of the U.S., then the role of openness would
have been diminished. The higher the elasticity of substitution between home and for-
eign industrial goods is, the more accelerated structural transformation in the U.S.
Chapter 3 tries to answer the following question: Can differences in sectoral pro-
ductivity growth rates account for the different structural transformation experiences
of Latin America and East Asia? In the comparative analysis of economic performance
in terms of structural change and the sectoral productivity differences, the contrast
x
between East Asia and Latin America is striking. The contribution of this chapter is
developing a simple, but relatively detailed, six-sector general equilibrium model and
applying it to a nine Latin American and four East Asian countries using a new data
set. Over the period 1974 to 2003, Latin American countries exhibit much slower de-
agriculturalization than East Asian countries, while the manufacturing employment
share has been almost stagnant in Latin America but exhibits a hump shaped pattern
among East Asian countries.
A multi-sector general equilibrium model, treating sectoral productivity growth
rates as exogenous, accounts well for the differing sectoral reallocations of labor in Latin
America and East Asia over the sample period. Several counterfactual experiments are
conducted. Had Bolivia experienced the same productivity growth in agriculture that
Korea has, then the agricultural employment share in Bolivia would have been 12.0
percent in 2003 rather than 27.4 percent. Similarly, if Argentina had experienced the
same productivity growth in the manufacturing sector as Korea, then the manufactur-
ing employment share in Argentina would be 4.3 percent in 2003 instead of 11.5 percent.
An exploration of policy and institutional factors which account for sectoral productiv-
ity differentials across Latin America and East Asia is presented.
xi
Chapter 1
Industrialization in Post-1978 China
This chapter contributes to the growing literature which analyzes the emergence of
China in the world economy. Much of the research on China’s economic development
is microeconomic and institutional in emphasis. One of the major problems associated
with an analysis of economic growth and trade in China is lack of appropriate data.
This chapter provides a detailed sectoral data analysis to have a better understanding
of the Chinese development experience.
1.1 The Structural Transformation of China, 1978-2005
Aggregate Growth and Productivity: Using the new PPP estimates of GDP , the United
States remains the largest economy in the world by 2005, with a world share of 22.5
percent, followed by China with 9.7.
1
Figure 1.1 plots annual average growth rates of
GDP per capita in 1990 Geary-Khamis dollars over the period 1978-2008 against annual
average growth rates over the period 1950-1977 for one hundred and one countries.
2
1
http://siteresources.worldbank.org/ICPINT/Resources/icp-final.pdf
2
The countries are selected such that there is no missing observation in the sample. GDP per capita for
a given country is measured in millions of U.S. dollars (converted at Geary-Khamis PPPs). For all coun-
tries, data for the years 1950-2008 are from the Conference Board, Total Economy Database, January 2009.
The sample consists of 101 countries: Albania, Algeria, Angola, Argentina, Australia, Austria, Bahrain,
Bangladesh, Barbados, Belgium, Bolivia, Brazil, Bulgaria, Burkina Faso, Cambodia, Cameroon, Canada,
Chile, China, Colombia, Costa Rica, Cote d’Ivoire, Cyprus, Denmark, Dominican Republic, DR Congo,
Ecuador, Egypt, Ethiopia, Finland, France, Ghana, Greece, Guatemala, Hong Kong, Hungary, Iceland,
India, Indonesia, Iran, Iraq, Ireland, Israel, Italy, Jamaica, Japan, Jordan, Kenya, Kuwait, Luxembourg,
Madagascar, Malawi, Malaysia, Mali, Malta, Mexico, Morocco, Mozambique, Myanmar, Netherlands,
New Zealand, Niger, Nigeria, Norway, Oman, Pakistan, Peru, Philippines, Poland, Portugal, Qatar, Roma-
nia, Saudi Arabia, Senegal, Singapore, South Africa, South Korea, Spain, Sri Lanka, St. Lucia, Sudan,
Sweden, Switzerland, Syria, Taiwan, Tanzania, Thailand, Trinidad and Tobago, Tunisia, Turkey, Uganda,
United Arab Emirates, United Kingdom, United States, Uruguay, Venezuela, Vietnam, West Germany,
Yemen, Zambia, and Zimbabwe.
1
China grows at an annual average rate of 2.6 percent during 1950-1977 and then at a
rate of 7.1 percent during 1978-2008. China, the world’s most populous country, is also
the fastest growing country in the sample during 1978-2008. China sustains an annual
average rate of growth of GDP per capita more than 5 percentage points higher than
that of the United States.
3
6
8
China
4
6
0
2
8-2008
-2
-8 -6 -4 -2 0 2 4 6 8
1978
-6
-4
-8
1950-1977
Figure 1.1: Growth Rates of GDP per Capita (%) : 1950-2008
Figure 1.2 displays the GDP per capita relative to the United States in a set of coun-
tries between 1950 and 2008.
4
In 1950 Chinese GDP per capita is 3.6 percent of that in
the United States. By 2008, it increases to 19.1 percent. Relative income in Japan starts
about 20 percent of the U.S. level, increases to above 84 percent in 1991, and declines
to 74 percent in 2008. During the 1990s Japan’s rapid growth is replaced by protracted
3
Perkins and Rawski (2008) anticipate that China’s economy can achieve real GDP growth at average
rates of 6-8 percent per annum between 2005 and 2025.
4
Western Europe consists of Austria, Belgium, Denmark, Finland, France, Germany, Italy, Netherlands,
Norway, Sweden, Switzerland, and the United Kingdom. Latin America consists of Argentina, Bolivia,
Brazil, Chile, Colombia, Mexico, Peru, Uruguay, and Venezuela. Asian countries are Hong Kong, Singa-
pore, South Korea, and Taiwan.
2
economic stagnation.
5
Asian Dragons in 1950 start at a GDP per capita level that is
about 10 percent of the United States and they reach to about 70 percent of the U.S.
level by 2008. Latin American countries, on the other hand, show relative stagnation, if
not deterioration.
1950 1970 1990 2010
0
10
20
30
40
50
60
70
80
%90
Latin America
East Asia
China
Japan
Western Europe
Figure 1.2: GDP per Capita as a Percentage of the U.S.
When the People’s Republic of China was founded in 1949, more than 80 percent
of the population was in agriculture. China, under the socialist government, chose
the heavy-industry oriented development strategy as the so-called “engine” of eco-
nomic development with distorted factor and product prices to “leap forward” the
nation. China started industrialization within a socialist camp with the leaning-on-
one-side policy, which placed a heavy reliance upon trade with and the assistance from
the U.S.S.R and industrialization was virtually synonymous with economic develop-
ment. Soviet aid to China ceased in 1961. For the purpose of mobilizing resources
for heavy industries, the Stalinist planned system was implemented in the Chinese
5
As Hayashi and Prescott (2002) note, Japan in the 90s, after steady catch-up for more than three
decades, not only stopped catching up but lost ground relative to the industrial leader of the 20th cen-
tury, the United States and have faced a prolonged recession.
3
economy. Great emphasis was put on investment and the rapid development of heavy
industry.
6
Naughton (2007, p. 55) labels this development strategy ”big push industrializa-
tion”. Heavy industry makes intensive use of capital, yet capital was very scarce in the
beginning of the leap-forward development strategy. The viability of heavy industries
in China was, therefore, even more tenuous than in the Soviet Union in the 1920’s (Lin
and Tan, 1999). To generate and allocate resources for heavy industrial development,
China relied on mechanisms such as investment licenses and import quotas rather than
a price mechanism working through markets.
Lin (1994, pp. 28-29) summarizes the key characteristics of this heavy-industry
development strategy as consisting of low interest rates, an overvalued exchange rate,
low wage rates, and low prices for raw materials and living necessities. Under the
central plan raw material prices were kept low, and final good prices high, generating
substantial surpluses in manufacturing and processing industries, which funded the
government budget (Young, 2000).
The resulting misallocation of resources through government planning, from 1949 to
1978, did not bring sustainable economic development to China, i.e., low aggregate TFP
growth. Chow (1993) finds that growth is almost entirely capital accumulation driven
during 1952-1980 and TFP growth is absent. Borensztein and Ostry (1996) estimate the
TFP growth is negative at about -0.7 percent average rate during 1953-1978. Rosenberg
(1994, pp. 105-106) argues that the Chinese government before 1978 did not give impor-
tance to the role of technological innovation in the attainment of an efficient industrial
6
According to the Chinese Statistical Yearbooks heavy industry refers to the industry which produces
capital goods, and provides various sectors of the national economy with necessary material and technical
basis. It consists of the following three branches according to the purpose of production or the use of
products: (1) Mining, quarrying and logging industry refers to the industry that extracts natural resources,
including extract ion of petroleum, coal, metal and non-metal ores and logging. (2) Raw materials industry
refers to the industry that provides various sectors of the national economy with raw materials, fuels
and power. It includes smelting and processing of metals, coking and coke chemistry, chemical materials
and building materials such as cement, plywood, and power, petroleum refining and coal dressing. (3)
Manufacturing industry refers to the industry that processes raw materials. It includes machine building
industry which equips sectors of the national economy, industries of metal structure and cement products,
industries producing means of agricultural production, such as chemical fertilizers and pesticides.
4
society and the preoccupation with ”big-ness” in industry was hostile to technological
innovation. Brandt and Sutton (2008) discuss that noneconomic policy objectives, weak
institutions, and poor incentives are the underlying causes of productivity stagnation.
On the other hand, the aggregate TFP is the single most important factor behind the
aggregate output growth in the Chinese economy for the sample period between 1978
and 2005. To show this I perform a simple aggregate growth accounting framework. I
decompose the factors that contribute to GDP per working-age population as follows:
Y
t
=N
t
=TFP
1=(1)
t
(K
t
=Y
t
)
=(1)
(E
t
=N
t
); (1.1)
where Y
t
is the aggregate GDP , N
t
is the economically active population, K
A
and E
t
are the quantities of capital and labor employed at time t and the capital share is given
by. TFP represents total factor productivity and the power 1=(1) represents the
magnification effect of TFP . An increase in TFP generates a proportionate increase in the
capital stock, so the capital intensity factor, (K
t
=Y
t
)
=(1)
, represents only the part of
capital accumulation not induced by TFP growth.
Aggregate labor, capital, and output are obtained as the summation of the sectoral
figures that I discuss in the related sections. Bai, Hsieh, and Qian (2006) discuss the
changing nature of the importance of capital and labor in the aggregate economy, and
argue that the average labor share between 1978 and 2005 is 51.3 percent. I measure
population as those who are aged 16 and over who are capable to work, rather than the
total population based on the definitions of CSY.
According to the results in Table 1.1, over the period 1978-2005, GDP per econom-
ically active person is grown at 7.45 percent per year, which is completely accounted
for by a 9.67 percent growth rate in TFP factor, which implies that the average annual
growth in TFP between 1978 and 2005 is 4.85 percent. These results suggest that most of
the fluctuations in output per working-age person are due to changes in the TFP factor,
rather than to changes in the capital-output ratio or in the employment rate.
5
Table 1.1: Growth Accounting, China, 1978-2005
Average annual rate of growth in percents
Sources of Growth 1978-1984 1984-2005 1978-2005
GDP per economically active person,Y
t
=N
t
6.05 7.85 7.45
TFP Factor,TFP
1=(1)
t
8.92 9.89 9.67
Capital Intensity Factor, (K
t
=Y
t
)
=(1)
-2.65 -1.06 -1.42
Employment Rate,E
t
=N
t
0.01 -0.80 -0.62
Blanchard and Giavazzi (2006, Table 4) and Cao, Ho, Jorgenson, Ren, Sun, and Yue
(2009, Table 1) report the estimates of TFP growth computed by various studies. These
studies cover different periods and employ different methods, and a complete account
of their differences and a reconciliation of their results with my observations are beyond
the scope of this study. Most of the studies find that TFP growth ranges around or
above 3 percent per year since the beginning of the economic reforms and my findings
are consistent with these studies.
7
Reallocation of Labor from Agriculture: The most striking feature of the structural
transformation of the Chinese economy is seen on shifts in the pattern of employment.
As China becomes more developed, there is a steady transfer of labor from rural to
urban areas, and the percentage of the labor force engaged in agriculture falls dramat-
ically. However, agriculture is still the dominant sector in China in terms of employ-
ment. I use two sources for employment data: CSY and Holz (2006).
Labor is measured in stock terms as the number of persons employed in a sector
at the end of each year. This measure, changes in labor input measured by persons
employed, does not reflect changes in hours worked per full-time and part-time worker,
or changes in the shares by sector of total employment. However, I do not have time-
series data on hours worked as the labor input measure. A better measure would be
total number of hours worked, broken down by type of labor input. Data on the number
7
Heston and Sicular (2008) argue that the studies with higher estimates of China’s TFP growth do not
make adjustments for the contribution of human capital and other such factors. On the other hand, Perkins
and Rawski (2008) obtain TFP growth of 3.8 percent per year for 1978 -2005 net of the contribution of rising
education levels.
6
of persons employed or number of jobs, by sector can generate biased measures of
productivity if hours per person change (OECD, 2001).
The Chinese economy is divided into three broad sectors: agriculture, industry, and
services. They are also known as primary, secondary, and tertiary, respectively. Primary
refers to agriculture, forestry, animal husbandry and fishery. Secondary refers to mining
and quarrying, manufacturing, production and supply of electricity, water and gas, and
construction. Tertiary refers to all other economic activities not included in primary or
secondary industry.
Holz (2006, p. 57) and Brandt, Hsieh, and Zhu (2008) discuss the problems regarding
the total and sectoral employment series reported in CSY. Brandt, Hsieh, and Zhu fol-
low Holz’s method to get the revised sectoral employment data. Holz (2006, Appendix
13) reports the revised employment values (end-year), where he revises the period 1978-
1989. Holz adjusts pre-1990 sectoral employment values for 1978-1989 are obtained by
applying the shares of the individual sectors in official total employment to the adjusted
pre-1990 total employment values.
Holz’s method is explained as follows. Prior to 1990, the published economy-wide
number of laborers constituted the sum of laborers across industrial sectors. Since 1990,
the economy-wide number of laborers exceeds the sum across industrial sectors signif-
icantly in each year, but continues to, as in all reform years, equal the sum across eco-
nomic sectors. Since the economy-wide number following the new time series for the
years after 1990 is the one compiled according to international definitions of employ-
ment, the economy-wide number of laborers in the years prior to 1990 was adjusted
following the population censuses of 1982 and 1990 (later-year official values rely on
population census data).
Table 1.2 presents the discrepancies between the official statistics and the revision
made by Holz. Holz’s data set for the period 1990-2005 coincides with that of the CSY.
Sectoral employment shares are the same both in official statistics and in revised data.
7
Table 1.2: Total Employed Persons, Official versus Revised, million persons
CSY (2006) Holz (2006)
Year Primary Secondary Tertiary Primary Secondary Tertiary
1978 283.2 69.5 48.9 330.4 81.0 57.1
1979 286.3 72.1 51.8 334.8 84.4 60.5
1980 291.2 77.1 55.3 339.6 89.9 64.5
1981 297.8 80.0 59.5 347.6 93.4 69.4
1982 308.6 83.5 60.9 358.5 97.0 70.8
1983 311.5 86.8 66.1 363.0 101.2 77.0
1984 308.7 95.9 77.4 357.4 111.1 89.6
1985 311.3 103.8 83.6 359.2 119.8 96.5
1986 312.5 112.2 88.1 360.5 129.4 101.6
1987 316.6 117.3 94.0 364.4 134.9 108.1
1988 322.5 121.5 99.3 369.4 139.2 113.8
1989 332.3 119.8 101.3 381.7 137.6 116.4
Figure 1.3 displays the evolution of sectoral employment shares based on revised
employment data in China between 1978 and 2005. Agricultural employment share
falls rapidly in the early stages of economic reforms, at low levels of income, giving
rise to rapid increases in the share of non-agricultural sector in total employment. Even
though the importance of primary industry in China’s economy falls, it is still a large
sector, accounting for more than 40 percent of employment in 2005.
8
Between 1978 and 2005, 17.3 percent of the employment is in industry in 1978 and
the share is 23.8 percent in the year 2005. The employment share of the service sector
climbs from 12.2 percent in 1978 to 31.3 percent in 2005. The employment share of the
industrial sector does not rise as fast as that of the service sector during the period
1978-2005. Banister (2005) argues that manufacturing employment in China increases
during the 1980s and early 1990s, peaks in about 1995-1996, declines during the late
1990s until 2000-1, and increases again 2002. Brandt, Rawski, and Sutton (2008) argue
that the stagnation of employment in the secondary sector is due to the result of the
massive state-owned enterprise (SOE) layoffs since the mid-1990s.
8
De-agriculturalization fosters the urbanization and off-farm migration facilitates the development of
the nonagricultural sectors in the Chinese economy. The share of rural population in the total is falling,
but remains high at 57.0 percent in 2005.
8
1978 1984 1990 1996 2002 2008
0
10
20
30
40
50
60
70
%80
Agriculture
Industry
Services
Figure 1.3: Employment Shares, China, 1978-2005
The Measurement of Sectoral Output: I start with the analysis of the composition of
GDP at current prices (yuan) during the period 1978 through 2005 based on the official
statistics. According to this data, the production structure of the economic activity in
China changes significantly for the years 1978-2005. The primary sector’s share of GDP
declines from 27.9 percent in 1978 to 12.6 percent in 2005. On the other hand, GDP
share of the tertiary sector increases from 24.2 percent in 1978 to 39.9 percent in 2005.
The secondary sector dominates the production structure with the average share of this
sector in total GDP is 45.3 percent over the sample period.
9
For output statistics in China, the deflators used to measure sectoral real output are
the major points of the discussion. There has been a recent discussion on the reliability
of the official Chinese GDP numbers as well as the implicit sectoral deflators since,
other things being equal, an overstatement (or an understatement) of sectoral output
growth could invalidate any productivity estimate. Ruoen (1997, p.122) and Young
9
Xianchun (2002) studies the estimates of GDP at current prices and argues that there are five main
problems in China’s GDP estimates: the measurement of housing services; fiscal subsidies; welfare services
provided within enterprises; rural industrial statistics; and livestock products.
9
(2003) compare the sectoral implicit GDP deflators with the independent survey based
price indices and they suggest alternative price indices instead of the implicit deflators.
They choose the index for ”overall farm and sideline products purchasing price” as an
alternative for a primary industry index.
Ruoen chooses ”industrial products producer index” to serve as the deflator for the
secondary industry. Young compares Ruoen’s choice with two other possible alterna-
tives: the industrial products rural retail price index and the retail price index. Young
argues that the Ruoen’s choice is a superior deflator. For tertiary industry, Ruoen uses
the index for services from the overall residents’ consumer price indices. Young has a
similar approach and combines urban service price index and the overall service price
index. Chow (2004) argues that Young’s method leads to serious errors, and that his
findings contradict the alternative estimates of the rates of growth for the periods 1978-
1998 and 1988-1989 provided by Young.
10
Holz (2006) offers the following approach: The output series rely on the post-
economic census benchmark revision data as far as the revisions reach back. Holz uses
real growth rates calculated from the first published implicit deflator and nominal val-
ues whenever feasible. First, nominal values are post-economic census values across all
sectors after 1993, all other nominal values are not revised, and the earlier published
nominal values are used in those instances. Second, the output values are in constant
year 2000 prices, which imply applying real growth rates to year 2000 (post-economic
census) nominal value added in order to obtain time series of constant price output.
First published implicit deflators are available for the primary, secondary, and tertiary
sector after 1987.
Figure 1.4 displays the evolution of value added shares at constant 2000 prices
(yuan) during the period 1978 through 2005. In the year 1978, agriculture captures
40.0 percent of the Chinese value added, while in 2005 it has the lowest contribution to
10
Dekle and Vandenbroucke (2009) follow Young’s methodology to choose the sectoral deflators. On the
other hand, Bosworth and Collins (2007) prefer to use the official output data for the primary and tertiary
sectors and the alternative (the ex-factory industrial price index) only for the secondary industry.
10
Chinese total value added, with 11.5 percent. The value added share of the secondary
(tertiary) sector increases from 31.8 (28.1) percent in 1978 to 46.0 (42.5) percent in 2005.
China’s real output in the secondary (tertiary) sector grows at an average annual rate of
11.6 (11.8) percent since 1978.
1978 1984 1990 1996 2002 2008
0
10
20
30
40
%50
Agriculture
Industry
Services
Figure 1.4: Real GDP Shares, China, 1978-2005
Capital by Sector: The Chinese official statistics provide no standard estimation of the
capital stock at any industry level or by any category. The sectoral gross fixed capital
formation (GFCF) data are available only at the provincial level and only for 1978-95,
where the total sectoral provincial GFCF accounts for an average of 78.86 percent of the
annual value of national GFCF.
I follow Holz (2006) to construct the sectoral GFCF data for the period 1978-95.
GFCF is divided into the three main economic sectors using sectoral share values avail-
able for the individual provinces in GFCF 1978-95. Provincial sectoral shares are shares
in the provincial sum-across sector-GFCF value. Holz (2006) uses GDP shares of the
three sectors to allocate the nation-wide GFCF into these sectors. However, these shares
seem to overstate investment in agriculture, since agriculture’s share in production is
very high compared to the capital formation rate. To avoid such a bias, I use the shares
11
of the year 1995 that I compute from Hsueh and Li (1999) for the remaining years, 1996-
2005.
An alternative methodology described in Dekle and Vandenbroucke (2009), follow-
ing Young (2003), is to construct the sectoral GFCF data as follows: they pursue a sim-
ilar approach for the period of 1978-95, except that they use the sum of provincial total
GFCF value for the nation-wide GFCF. For the years after 1995, they obtain the distri-
butional GFCF data from the individual Provincial Statistical Yearbooks, and aggregate
across the provinces. They use the sectoral distribution reported in Hsueh and Li (1999)
to allocate overall national GFCF between the three sectors.
Hsueh and Li (1999, p.137) define their methodology as follows: “According to the
type of industry, gross fixed capital formation can be divided into the investment by the
primary, secondary and tertiary industrial sectors. The principle behind this division
according to industrial sectors is the nature of the production activities undertaken after
the construction units have completed the projects or have handed them to be put into
production, or the particular nature of production that other social economic activities
take in the process of production.”
The ideal index to deflate nominal investment figures is the price index of invest-
ment in fixed assets. The CSY, however, began to provide this index only after 1993.
Jefferson, Rawski and Zheng (1996) estimate China’s price index of investment in fixed
assets between 1979 and 1992. Zhang (2006, p. 290) argues that the estimates of Jeffer-
son, Rawski and Zheng are consistent with the official figures, since both sources esti-
mate their indices by averaging the deflators of construction/installation and machin-
ery/equipment purchases; and Zhang constructs China’s price index of investment in
fixed assets between 1978 and 2000. Using the CSY, I extend this index to the year 2005.
Zhang (2006) does not report the value of the index for the year 1978. I take the 1979’s
value as the corresponding observation for the year 1978.
Sectoral capital stock series are calculated using the perpetual inventory approach
with 10 percent depreciation rate following Bai, Hsieh, and Qian (2006), who argue that
12
10 percent is a plausible number for the period 1978-2005. The initial capital stock series
in each sector is calculated by the formulaI
j0
=(g
j
+), whereI
j0
is the first year of the
sectoral real investment series,g
j
is the average growth of the sectoral investment in the
first five years of the sectoral real GFCF series, and is depreciation rate. This approach
ensures that the 1978 values of the capital stocks are independent of the 1978-2005 data
used in our analysis. Moreover, given the relatively small capital stocks in 1978 and
the high levels of investment, the estimates for later years are not sensitive to the 1978
benchmark values of the capital stocks. All real series are valued at 2000 prices.
Figure 1.5 plots the capital-output ratios for the whole economy, and by sector. The
average capital-output ratio for the entire economy is 1.40 for the sample period. Bai,
Hsieh, and Qian (2006) estimate that the average nominal capital-output ratio between
1978 and 2005 is 1.46. The real capital stock of the primary sector grows by 3.69 percent
per annum on average during 1978-2005. The corresponding figures for the secondary
and the tertiary sectors are 7.12 percent and 11.90 percent, respectively.
The capital-output ratios of the primary and the tertiary sectors have demonstrated
a U-shape pattern since 1978. The capital-output ratio of the primary sector (the tertiary
sector) increases after 1995 (1993). In contrast, there is a reduction in the capital-output
ratio in the secondary sector, starting with value of 2.28 in 1978 and falling to about 0.81
in 2005, reflecting the higher efficiency of capital in secondary industry.
The capital intensity and the investment rate increase after 1997, consistent with
the fact that the capital share of aggregate income increases steadily. The investment
rate increases from 31.80 percent in 1997 to 41.49 percent in 2005, whereas the period
average is 31.68 percent during 1978-2005. I use the expenditure components of GDP
as investment rate and follow Young (2003) by expressing it as the ratio of gross fixed
capital formation to nominal GDP (see CSY, Table 3.12).
13
1978 1984 1990 1996 2002 2008
0
1
2
3
4
5
Agriculture
Industry
Services
Aggregate
Figure 1.5: Capital-Output Ratios, China, 1978-2005
Capital formation increases the capital stock which, in turn, expands production
capacity. Bai, Hsieh, and Qian (2006) interpret this observation as a gradual restructur-
ing of China’s industrial sector, in favor of more capital-intensive industries, requiring
higher aggregate investment rates in the steady state.
1.2 Sectoral Growth Accounting and Economic Reforms
Framework: I assume that capital and labor are the two primary production factors
in the generation of final output. Moreover, land is the third input in the production
of the agricultural goods, it is a nonreproducible factor, constant and also it does not
depreciate. Agricultural land has been almost a fixed quantity since 1952 in China. The
average annual growth rates of cultivated land area and total sown land area were 0.35
and 0.18 percent during 1952-2005, respectively.
11
. Since it is constant, its contribution
is submerged in the TFP of the agricultural sector.
11
China Data Online, Production condition for agriculture of China, http://chinadataonline.org/
14
I specify the technologies at the sector level and employ the Cobb-Douglas func-
tional form. The production function for the agricultural sector is
Y
At
=TFP
At
K

At
E

At
: (1.2)
K
At
andE
At
are the quantities of capital and labor employed in the primary sector
at time t. Y
At
is the sectoral output produced in this sector andTFP
At
is the agricul-
tural TFP at time t. The share parameters for capital and labor in are given by and
, respectively ( + < 1). The agricultural production function is consistent with a
literature on cross-country agricultural production functions (Vollrath 2009).
Non-agricultural goods are produced using two factors of production, capital and
labor, combined in constant returns to scale technology in secondary and tertiary sec-
tors,
Y
jt
=TFP
jt
K

j
jt
E
1
j
jt
; j2fsecondary;tertiaryg: (1.3)
HereK
jt
andE
jt
are the quantities of capital and labor employed in sector j at time t.
Y
jt
is sectoral output,TFP
jt
is sectoral TFP at time t, and
j
denotes the capital share
of sector j.
I need to determine the sectoral factor shares to conduct sectoral growth accounting
exercises. Holz (2006) reports the time series for the sectoral labor shares in the pri-
mary, secondary, and tertiary sectors in the Chinese economy during 1978-2002; and
calculates the labor share by sector as the share of labor remuneration in the sum of
labor remuneration, depreciation, and operating surplus. Net taxes on production are
split proportionally between labor and capital, where capital’s share is measured by the
sum of depreciation and the operating surplus. Since there is no national data on these
sources of renumeration, all shares are based on the sum of provincial values.
Holz also notes that all values are pre-economic census values; revised values have
so far not been released, and are unlikely to be forthcoming. Since the sum provin-
cial pre-economic census value added comes very close to the post-economic census
15
national value added, these provincial pre-economic census values may be quite accu-
rate (Holz, 2006, Appendices 32 and 33). Holz calculates that the average labor income
shares in the primary, secondary, and tertiary sectors are 0.884, 0.475, and 0.502, respec-
tively between 1978 and 2002. I use these figures as the sectoral labor shares for the
entire period.
Young (2003) argues that the labor share in the non-agricultural sector increases
steadily between 1978 and 1995 and is slightly below 0.5. I use identical capital and land
shares in agriculture following Dekle and Vandenbroucke (2009). Turning to sectoral
capital input in the secondary and tertiary sectors I take the share of capital by sector to
be simply one minus the share of labor.
Sectoral TFPs and Growth Accounting: I observe that agricultural TFP growth is not
affected significantly in the presence of land. That’s why I exclude the land, for the
rest of the analysis, from the production function so that the agricultural production
function looks like the other two sectors’ production functions.
Figure 1.6 shows the actual path of sectoral TFPs in China between 1978 and 2005.
TFP growth in Chinese agriculture averages 4.55 percent per annum between 1978 and
2005. The average growth rate of TFP in the secondary sector is the highest of the three
sectors. TFP growth in the secondary sector averages 6.11 percent per annum and TFP
growth in the tertiary sector averages 2.91 percent per annum between 1978 and 2005.
All sectors experience declines in TFP growth rates during the late 80s. This coin-
cides with the violent repression of the student movement at Tiananmen Square in June
of 1989, which puts a temporary end to the steady liberalization of the Chinese econ-
omy and leads to temporary recentralization of many economic activities (Naughton
1995, p. 4).
In order to measure the contribution made by factors of production relative to that
made by TFP , I conduct a growth accounting exercise at the sectoral level.
Y
jt
=E
jt
=TFP
1=(1
j
)
jt
(K
jt
=Y
jt
)

j
=(1
j
)
; j2fprimary;secondary;tertiaryg: (1.4)
16
The first term on the right of (1.4) is the sectoral TFP factor in sector j. The second
term measures the sectoral capital intensity factor. The power 1=(1
j
) represents the
magnification effect of sectoral TFP that an increase in TFP generates a proportionate
increase in the sectoral capital stock, so the capital intensity factor represents only the
part of sectoral capital accumulation.
1978 1984 1990 1996 2002 2008
0
1
2
3
4
5
1978=1
Agriculture
Industry
Services
Figure 1.6: TFP by Sector, China, 1978-2005
Table 1.3 reports the average annual growth rate of sectoral GDP per worker and
its factors shown for post-reform China. Real GDP per worker of the primary sector
grows by 5.88 percent average annual growth rate during 1978-1984. Rapid growth rate
in the wake of the early post-1978 reforms then falls to an annual average rate of 4.73
percent for the years between 1984 and 2005. For the primary sector, there is no increase
in capital intensity: between 1978 and 1984, the capital-output ratio declines.
A high per-worker GDP growth rate of 5.88 percent is brought completely by a very
high TFP (more than 6.88 percent growth in TFP factor). This observation is consistent
with earlier findings. For example, Stavis (1991) views that technological change is
the engine of agricultural growth for the period 1978-1984. What causes the high TFP
growth rate, including the early reform period, in the primary sector?
17
Table 1.3: Sectoral Growth Accounting, China, 1978-2005
Average annual rate of growth in percents
Economic Activity / Sources of Growth 1978-1984 1984-2005 1978-2005
Agriculture
Output per Worker 5.88 4.73 4.98
Capital Intensity Factor -0.94 0.04 -0.18
TFP Factor 6.88 4.68 5.17
Industry
Output per Worker 3.37 9.76 8.31
Capital Intensity Factor -3.87 -4.55 -4.40
TFP Factor 7.53 14.99 13.29
Services
Output per Worker 3.75 6.66 6.00
Capital Intensity Factor 0.54 0.00 0.12
TFP Factor 3.19 6.65 5.87
China adopted a strategy of gradual economic transformation that maintained the
existing system and created new economic activities on top of it. Between 1978 and
1984 significant developments took place agriculture. In the early reform period (1978-
1984), the household responsibility system (HRS), which replaced the production team
system as the unit of production and income distribution, significantly increased agri-
cultural productivity transferring the collective agricultural production system to indi-
vidual farms by contracting land-use rights to individual rural households, price and
marketing reforms improving the peasants’ work incentives (see, for example, Lin 1988,
1994; Lin, Cai, and Li 2003, Chapter 5; and Naughton 1995, Chapter 4 for details of the
HRS).
Lin (1988) argues that the failure of the collectivization period is not due to its social-
ist nature but it is because of the difficulties inherent in supervising agricultural work.
Farmers are residual claimants in the HRS. Since the end of the year 1978 the HRS has
gradually replaced the commune system. By the end of 1983 less than 3 percent of
households had not adopted the responsibility system. This suggests that the institu-
tional transformation from a collective to the HRS of farm management was essentially
completed by the mid 1980s. The decline in the growth rate, according to this view, has
18
been associated with the completion of one-off effects of the HRS since the institutional
reforms were one time only events. Huang and Rozelle (1996) argue that earlier studies
may have over-estimated the impact of decollectivization.
There are important reforms in the non-agricultural sectors: gradual reduction of
centralized controls on prices, inputs and outputs, and the rising share of production
outside of the state enterprise sector; and the freedom of townships and villages to
establish industrial enterprises outside of the central plan (Jian, Sachs, and Warner,
1996). China’s real output per worker in the secondary sector grows at an average
annual rate of 8.31 percent since 1978. For the secondary sector, there is no increase in
capital intensity: between 1978 and 2005, the capital-output ratio declines significantly.
One of the major institutional features of the Chinese economy is the coexistence
of state-owned enterprises (hereafter, SOEs) and non-state sector. The non-state sector,
including private enterprises, joint ventures, urban collectives, and township and vil-
lage enterprises (TVEs), has crowded out SOEs in many markets. TVEs are economic
units which are either collectively owned by local residents in the rural areas of China or
mainly owned and controlled by the peasants (Fu and Balasubramanyam, 2003). Brandt
and Zhu (2001) make clarification for the definition of the non-state sector: although it
does include private enterprises and joint ventures, until recently the non-state sector
was primarily made up of urban collectives and TVEs.
Brandt, Rawski, and Sutton (2008) divide reforms in the secondary sector into two
periods: reforms concentrated on incentives and market mechanism to prevent resource
misallocation problem. Beginning in the mid-1990s, the privatization and subsequent
stock market listing of SOEs have been integral parts of China’s state enterprise reforms.
In the beginning of the reform period, SOEs dominated the industrial structure.
In the secondary sector, policies were introduced to increase the autonomy of enter-
prise managers, to reduce the dominance of planned quotas, and to allow enterprises
to produce and sell goods in the market. More market-oriented polices have emerged
with the growing importance of the urban private sector, as SOEs are being downsized
19
and the real sector of the economy has been liberalized substantially and goods and
factor markets have become increasingly competitive.
Huang and Duncan (1997) discuss that studies of TFP in the state sector have differ-
ent results: some studies find no TFP growth during the post-1978 period; some other
studies argue that state sector has positive TFP growth during the reform period. Liu
and Otsuka (2004) show statistically that in order for SOEs to compete with TVEs, a
major ownership reform of SOEs is essential.
Real GDP per worker of the tertiary sector grew by 6.00 percent average annual
growth rate during 1978-2005. The production of services is likely to become increas-
ingly important to China’s overall economic development over the coming decades.
Being a World Trade Organization (WTO) member since 2001, increased market access
has opened new economic opportunities for China, with an expected favorable impact
on trade and investment for years to come. Services created 27.15 million new jobs,
which was 85 percent of all employment creation, during the 9th Five-Year Plan, 1995-
2000 (OECD 2003).
1.3 China’s Foreign Trade after 1978
For many years, the development of China remained largely indigenous, mainly
because of China’s isolation from other countries. However, China has become an
increasingly important part of the global trading system, especially over the last decade.
China has followed a new pattern of industrialization characterized by very rapid
growth of industrial sector and increasing participation in the international economy.
Figure 1.7 shows that China grows from having a negligible role in world trade to being
one of the world’s exporting powers accounting for 11 percent of global export share in
manufactured goods in 2006 (WTO Statistics).
From 1949 to 1978, China was a planned economy, and trade flows were entirely
controlled by the state. Since 1978, China has been moving toward a market economy
20
and, thus, a more liberal trade system. In 1980-81 four special economic zones (Shantou,
Shenzhen, Xiamen, and Zhuhai) were established. These zones were new areas created
in localities far away from the power center and concentrated on the southern coast of
China. Some argue that China’s rapidly growing coastal provinces have benefited from
the proximity to the Chinese-speaking economies of Hong Kong and Taiwan (see, for
example, Goodfriend and McDermott 1998 and Naughton 1996).
China has attempted different reforms in three broad areas to liberalize foreign
trade: the gradual elimination of central plans and the introduction of market com-
petition in the tradable sectors; the reduction of barriers to trade including both tariff
and non-tariff restrictions; and the reform of the foreign exchange regime. The rate of
decreases in trade barriers has increased especially in the 1990s.
The second phase of liberalization started in 1992. It was not until 1992 when China
declared its intention to establish a so-called socialist market economy that it began
to lower tariffs. Many major changes started in 1992, i.e., preferential policies, such
as generous tax holidays and credit matching, were given to foreign investors. Foreign
direct investments surged to US$ 11.3 billion in 1992 from US$ 4.6 billion the year before
(Bao, Chang, Sachs, and Woo 2002).
There were significant reductions starting from 1992 in tariff rates and the removal
of nontariff barriers. The government reduced the number of export goods subject to
quota-license regulation from 212 to 183 in 1992 and eliminated import-quota license
requirements for 16 classes of goods. Import license requirements for 9 classes of goods
that comprise 283 goods were eliminated in 1993. Next, the government stopped issu-
ing mandatory plans for imports and exports in 1994 as well as the elimination import
license requirements for 195 goods. It was followed by another elimination of import
license requirements for 120 goods in 1995. Lastly, in 1996, 30 percent of the remaining
quotas were eliminated (see Shuguang, Yansheng, and Zhongxin 1998, Table 2.1 for the
details in tariff and nontariff changes).
21
1980 1993 2006
0
16
32
48
%64
G7
Canada
France
Germany
Italy
Japan
United Kingdom
United States
China versus G7
1980 1993 2006
0
10
%20
Dragons
Hong Kong
Singapore
S. Korea
Taiwan
China versus Dragons
1980 1993 2006
0
10
%20
Latin America
Argentina
Brazil
Chile
Colombia
Costa Rica
Mexico
Peru
Venezuela
China versus Latin America
1980 1993 2006
0
1
6
%11
China
China versus India
Figure 1.7: China’s Share of World Manufactured Exports, 1980-2006
Figure 1.8 shows the evolution of the effective tariff ratio in China during 1952-2005.
The effective tariff rate is defined as the ratio of tariff revenue to total imports. The data
are from Kanbur and Zhang (2005) for the period 1952-77 and the CSY (2006, Tables
8.3 and 18.3) for the period 1978-2005. The effective tariff rate was 12.8 percent in 1952
and 2.0 percent in 2005. The figure is divided into two parts indicating pre-and post-
liberalization episodes and it makes a peak in 1977, just before the reforms that were
started in 1978 towards a transition from rigid central planning toward a market-based
economy. The effective tariff rate has been always less that 10 percent after the year
1986.
22
1950 1978 2006
0
5
10
15
%20
Pre-Liberalization Post-Liberalization
Figure 1.8: Effective Tariff Rates, China, 1952-2005
China officially started its WTO membership application in 1986 and she formally
became a member of the WTO on 11 December 2001. WTO accession gives China
greater access to the world’s markets. As a result of the negotiations, China has agreed
to undertake a series of important commitments to open and liberalize its regime in
order to better integrate in the world economy and offer a more predictable environ-
ment for trade and foreign investment in accordance with WTO rules.
China’s Changing Comparative Advantage: The composition of international trade has
changed considerably in the post-1978 period. The declining role of agriculture in total
employment and output is accompanied by a declining share of primary goods’ trade
in China’s total commodity trade. Primary goods include food and live animals used
chiefly for food; beverages and tobacco; non-edible raw materials; mineral fuels, lubri-
cants and related materials; animal and vegetable oils, fats and wax. In 1980, primary
goods’ share in total commodity trade was 42.15 percent. By 2005, it declined to 13.84
23
percent (CSY 2006, Tables 18.5 and 18.6). On the other hand, the share of manufactures
in total exports increased.
Table 1.4 breaks down the Chinese nominal trade balance (in current billion U.S.
dollars) in 1984 and in 2005 into industries using one-digit SITC codes. Certainly, SITC
industries at the one-digit level are rather broad, and some important details about
changes in the structure of Chinese trade are likely to be obscured. I investigate this
issue below. The table shows that China imports raw materials and China imports raw
materials and chemicals (SITC industries 2, 3 and 5) and exports manufactured goods
(SITC industries 6, 7 and 8). Similarly, decomposing China’s real export growth since
1992, Amiti and Freund (forthcoming) finds that there has been a significant decline
in the share of agriculture and soft manufactures, such as textiles and apparel, with
growing shares in hard manufactures, such as consumer electronics, appliances, and
computers.
Dean and Lovely (forthcoming) study the trends in the composition of China’s trade
and find significant changes in the sectoral composition of Chinese trade between 1995
and 2005. For example, in 1995, textiles and apparel accounted for the largest shares
of Chinese exports to the world. These shares fell by about a third by 2004, while the
export share of office and computing machinery grew by a factor of five, and that of
communications equipment more than doubled. The largest shares of Chinese imports
in 1995 were attributable to textiles and machinery. These shares fell by about 70 per-
cent and 40 percent, respectively, by 2004, while import shares in office and computing
machinery and in communications equipment more than doubled. Dean and Lovely
(forthcoming) aggregate the Chinese trade data to HS (6-digit) and then converted to
ISIC Revision 3 using the official Chinese concordance.
In a companion to the analysis of Table 1.4, I conduct an exercise in which I com-
pute the specialization indices of China using one-digit SITC codes. Kwan (2002, pp.
15-17) argues that the revealed comparative advantage of a country can be shown by
calculating the specialization indices for its major industries. For a particular industry,
24
the specialization index is defined as its trade balance divided by the volume of two-
way trade, with a higher value implying stronger international competitiveness for the
industry concerned.
Table 1.4: Chinese Trade Balance in current billion US$
SITC Code Industry 1984 2005
0 Food and Live Animals 0.9 13.1
1 Beverages and Tobacco -0.01 0.4
2 Crude Materials, Except Fuels -0.1 -62.7
3 Mineral Fuels, Lubricants and Related Materials 5.6 -46.7
4 Animal and Vegetable Oils, Fats and Waxes 0.1 -3.1
5 Chemicals and Related Products -2.7 -41.2
6 Manufactured Goods Chiefly by Materials -2.1 49.2
7 Machinery and Transport Equipment -5.7 61.6
8 Miscellaneous Manufactured Article 3.3 132.0
9 Commodities Not Classified Elsewhere -0.6 -0.4
Source: United Nations Statistics Division, Commodity Trade Statistics Database
Following Kwan (1994), a country’s comparative advantage structure (as revealed
by its trade structure) can be classified into one of four categories based on the rel-
ative magnitude of the specialization indices of the country’s primary commodities
(SITC Rev. 2, sections 0 - 4), machinery (SITC Rev.2, section 7, a proxy for capital-
and-technology-intensive products), and other manufactures (SITC Rev. 2, sections 5,
6, 8, 9, a proxy for labor-intensive products). Each stage can be characterized by a
typology, to which the countries in that stage are alleged to conform. A country typi-
cally passes from one category to another in the following sequence: (i) the developing
country stage, with primary commodities more competitive than other manufactures
and machinery; (ii) the young NIE (newly industrialized economy) stage, with other
manufactures becoming more competitive than primary commodities, which maintains
its lead over machinery; (iii) the mature NIE stage, with machinery overtaking primary
commodities while other manufactures maintain their overall lead; and (iv) the indus-
trial country stage, with machinery overtaking other manufactures, which maintain
their lead over primary commodities.
25
Figure 1.9 shows the development of China’s trade structure between 1984 and 2005
in terms of broad sectors of merchandise trade. The figure exhibits that China became
a young NIE in 1991, when the specialization index of other manufactures surpassed
that of primary commodities. Subsequently, it attained the mature NIE stage in 1999,
when the specialization index of machinery also overtook that of primary commodi-
ties. In 1984, the specialization indices for China indicated that machinery and trans-
port equipment (SITC 7) were the lowest, while the highest was for the miscellaneous
manufactured articles (SITC 8). It is clear from the figure that, if the current trend in
the sectoral specialization indices continues, the specialization index of machinery will
overtake that of other manufactures in the near future, when China will be classified as
an industrial country.
1983 1991 1999 2007
-1.0
-0.8
-0.6
-0.4
-0.2
0.0
0.2
0.4
0.6
Other Manufactures
Primary
Machinery
Figure 1.9: Stages of China’s Trade Structure, 1984-2005
Trade in New Products: Another dimension of the changing nature of China’s compar-
ative advantage is the observation of the emergence of previously non-exported prod-
ucts. Hummels and Klenow (2005) and Kehoe and Ruhl (2009) decompose the growth
of individual countries’ trade into that part due to countries exporting new products -
26
what they call the ”extensive margin” - and that part due to countries exporting more
of the same products - the ”intensive margin”.
I perform the following new goods in trade exercise, based on Kehoe and Ruhl
(2009). I take four-digit SITC (Revision 2) bilateral trade data obtained from United
Nations Commodity Trade Statistics Database for the years between 1985 and 2005.
There are 786 categories of goods in these data. First, I rank categories in order of base-
year exports, from categories with the smallest amount of trade to the categories with
the largest amount. Second, I form ten sets of ”10 percentile” export groups by cumu-
lating export product categories - the first 677.08 categories account for 10 percent of
exports, for example; the next 50.56 categories account for 10 percent of exports; the
next 15.72 categories account for 10 percent of exports; and so on. Third, I calculate the
share of exports in subsequent years accounted for by each set of categories.
Figure 1.10 shows that the largest increases in the share of exports occur for those
sets of categories that accounted for the smallest amount of trade in 1985. The 677.08
smallest categories of exports from China to the rest of the world accounted for 10 per-
cent of exports in 1985, but in 2005 these same 677.08 categories accounted for 68.50
percent of exports.
Figure 1.11 depicts the evolution over the period 1985-2005 of the export shares of
the set of categories least traded in 1985. The share of the least-traded goods in total
exports has increased gradually and continuously over time. In other words, the goods
exported from China that were the least traded in 1985 account for a disproportionate
portion of growth in trade. Broda and Weinstein (2006) note that China exported 710
different goods to the United States in 1972 as opposed to 10,315 in 2001. The find-
ings suggest that the goods exported from China that were the least traded in 1985
account for a disproportionate portion of growth in trade, and document the expansion
in export varieties from China due to the acceptance into the World Trade Organization
after the year 2001 (see, also, Kehoe and Ruhl 2009 for similar findings).
27
Figure 1.10: Composition of Exports: China to the Rest of the World
Figure 1.11: Exports: China to Rest of the World, 1985-2005
28
Exchange Rate: Figure 1.12 shows the nominal exchange rate of the RMB in Japanese
yen and in U.S. dollars between 1981 and 2005. This figure reflects the purchasing
power of RMB relative to the main convertible currencies. The name of the Chinese
currency is the renminbi (RMB) and its unit is the yuan. I use them interchangeable. I
observe a stable parity around 8.3 Yuan / US$ after a major devaluation of the RMB in
1994. During most of the period between 1994 to the present day, the U.S. has had a
substantial trade deficit in Sino-American trade.
1981 1987 1993 1999 2005
0
2
4
6
8
10
Yuan / US$
Yuan / (100*Japanese Yen)
Figure 1.12: Nominal Exchange Rate of the RMB, 1981-2005
The large trade imbalance in Sino-American bilateral trade has been blamed on the
undervalued RMB. On the other hand, McKinnon and Schnabl (2006) argue that this
common presumption of RMB undervaluation is wrong, and its appreciation need not
reduce China’s trade surplus but would cause serious deflation in China. On 21 July
2005, the government of China finally revalued the RMB against the dollar-though by
only 2.1 percent-and announced the implementation of a new managed floating system
(Frankel 2006).
During a time of economic catch-up and rapid financial transformation, fixing the
exchange rate is the preferred way of anchoring the domestic price level. Groenewold
29
and He (2007) estimate the effect on the US - China trade balance of a revaluation of the
RMB and present a range of computations but, likely, changes in the value of the RMB
are not predicted to make much inroad into the trade imbalance between the US and
China - a 10 percent revaluation is likely to improve the trade balance by less than 10
percent.
Another related argument is that the RMB’s undervaluation and its peg to the dol-
lar prevent other countries’ currencies from rising against the dollar, since apprecia-
tion would damage their competitive power relative to China. Zhang (2001) argues
that exchange rate reform is a vital supply-side factor in China’s export growth for the
period between 1978 and 1993, the period that most major events of China’s exchange
rate reforms happened.
Some argue that China’s exchange rate policy artificially holds down the value of the
yuan to the detriment of U.S. manufacturing output and employment in both import-
competing and exporting industries (Goldstein 2004, Holtz-Eakin 2003, Hua 2007).
12
Corden (2007) calls this “exchange rate protection”, i.e., industries producing tradables
are protected at the expense of producers of nontradables.
Trefler (2005) argues that China cannot forever keep the yuan undervalued, and the
Chinese currency will rise to the point where China’s low yuan denominated wages
are eradicated by the currency conversion by the logic of comparative advantage. As
the current wave of globalization intensifies, industrialized nations are exposed to new
competition in domestic and foreign markets. Chapter 2 studies one example of this
phenomenon investigating impact of the industrialization of China on the industrial
employment share observed in the United States (the industrial leader of the world)
between 1978 and 2005.
12
There is some empirical evidence that changes in exchange rate, in terms of appreciation and depreci-
ation of and volatility in exchange rates, have an important influence on domestic employment. Gourin-
chas (1999) investigates empirically the pattern of job creation and destruction in response to real exchange
movements in France between 1984 and 1992, using firm level data and finds that traded sectors are very
responsive to real exchange rate movements. Klein, Schuh, and Triest (2003) find strong evidence that
movements in real exchange rates significantly affect gross job flows in U.S. manufacturing.
30
Chapter 2
De-industrialization of the Riches
and the Rise of China
In terms of employment, China’s manufacturing industry is the largest in the world,
employing more manufacturing workers than the G7 countries combined. In 2007,
China was the second leading exporter and third largest importer of merchandise trade
in the world. Moreover, the U.S. is the largest export market for Chinese-made products
(followed by Hong Kong and Japan). This chapter studies the impact of the industrial-
ization of China on the U.S. industrial employment share between 1978 and 2005.
1
The reallocation of resources across the broad economic sectors agriculture, indus-
try, and services is a prominent feature of economic development since the pioneering
works of Fisher 1935, Clark 1940, and Kuznets 1966. In this transformation, a sub-
stantial and permanent shift occurs in the composition of income, output, and employ-
ment away from agriculture and towards industry and services. Decline in agricultural
employment in early stages of development is well-established.
2
The share of agricul-
ture in total employment, which was initially very large, has undergone a continuous
decline throughout the entire path of economic development. For example, agricultural
employment share in the U.S. fell from about 74 percent in 1800 to about 2 percent in
2000.
3
1
Coleman (2007) argues that in countries like Japan, South Korea, and Taiwan towards the end of the
20th century, the emergence of China seems to be associated with falling terms of trade and overall growth
slowdown.
2
See, for example, Caselli and Coleman 2001 and Dennis and
˙
Is ¸can 2009.
3
http://myweb.dal.ca/tiscan/research/data/account.xls
31
On the other hand, decline (rise) in industrial (services) employment share is a rel-
atively recent issue.
4
The share of industrial employment has been declining for more
than three decades in today’s most advanced economies, a phenomenon that is referred
to as ”de-industrialization”.
5
For example, industrial employment share fell from about
33 percent to 17.6 percent, while services employment share increased from 58 percent
to about 81 percent in the U.S. between 1950 and 2005.
Existing econometric studies have argued that the de-industrialization is partly
explained by factors that are internal to the advanced economies. Among these inter-
nal factors are changing expenditure patterns, the faster growth of productivity in the
industrial sector than in services, and the resulting decline in the relative price of man-
ufactures.
6
In addition to the econometric studies, there are calibrated multi-sector general
equilibrium models to understand the sources of the sectoral reallocation and to quan-
tify the impact of these shifts on aggregate growth and productivity. Most of these stud-
ies utilize two (agriculture and the rest of the economy) or three (agriculture, industry,
and services) sector models and generate structural transformation as a result of the
differences in productivity growth across sectors (Ngai and Pissarides 2007) or based
on the sectoral differences in income elasticities of demand (Kongsamut, Rebelo, and
Xie 2001) or both (Duarte and Restuccia 2010).
7
The majority of the existing studies on structural transformation examine the expe-
rience of a country in isolation failing to take into account the interactions between
4
See Fuchs 1968 and Singelmann 1978 for early studies and
˙
Is ¸can 2010 for a recent investigation of the
rise of the service employment in the U.S.
5
Job losses in the industrial sector are part of a long trend as all the G7 countries increasingly become
service economies. In the G7, the share of employment in industry reached a peak in 1969 (38.3 percent)
and the share was 23.7 percent in the year 2006 (OECD Labor Force Statistics).
6
See, for example, Alderson 1999; Rowthorn and Ramaswamy 1999; Rowthorn and Coutts 2004; Nick-
ell, Redding, and Swaffield 2008; and Kollmeyer 2009.
7
See Foellmi and Zweim ¨ uller 2008; Buera and Kaboski 2009a,b; Matsuyama 2009 and the references
therein for recent studies of structural transformation.
32
countries. There are a few very recent studies investigating the relationship between
openness and the different aspects of structural transformation. For example, Stefan-
ski (2010) analyzes the impact of structural transformation on global oil prices (rather
than employment shares) and finds that structural transformation in China and India
accounts for up to 24 percent of the increase in the oil price in the OECD over the 1970-
2007 period.
Teignier (2010) studies the effect of international trade on de-agriculturalization in a
growth model with two sectors, agriculture and the rest of the economy and finds that
trade accelerates the structural transformation, i.e., if the United Kingdom had been in
autarky, the agricultural employment share in 1800 would have been around 80 percent
instead of 35 percent, and during the 19th century the agricultural employment share
would have been 1.5 times higher on average.
Yi and Zhang (2010) develop a two-country three sector model (agriculture and
manufacturing are tradable sectors and the service sector is non-tradable) and show
that in an advanced economy, the manufacturing sector will decline at a faster rate, and
the services sector will grow at a faster rate, in an open economy relative to the closed
economy.
I develop a two-country, three-sector model, in which countries trade industrial
goods. Goods are differentiated by the productivity growth with which they are domes-
tically produced. These features of the model and the degree of substitutability in pref-
erences between home and foreign produced units of industrial goods endogenously
determine a country’s equilibrium pattern of production and trade. I study the quanti-
tative predictions of the model economy over the sample period spanning 1978 - 2005
with sectoral data from China and the U.S.
A comparison of the predictions of open and closed economy models suggests
that a common explanation of de-industrialization in the literature, which is based on
increased productivity in industry relative to services in a closed economy setting, is
not compelling.
33
My benchmark results suggest that the closed economy model accounts for 38.1
percent of the declines in the U.S. industrial employment share while the open economy
accounts for 68.0 percent of the de-industrialization in the U.S. between 1978 and 2005.
Moreover, the open economy model has more explanatory power to explain the secular
changes in the U.S. industrial employment share in the post-1990 period. The open
economy model accounts for 85.1 percent of the de-industrialization while the closed
economy accounts for 37.4 percent of the de-industrialization in the U.S. between 1992
and 2005.
Counterfactual experiments show that if the Chinese economy had experienced pro-
ductivity in industry equal to that of the U.S., then the role of openness would have
been diminished. The higher the elasticity of substitution between home and foreign
industrial goods is, the more accelerated structural transformation in the U.S..
2.1 Structural Transformation in Closed Economy
Households: The economy is populated by an infinitely-lived representative house-
hold of constant size over time. The population size is normalized to one, without loss
of generality. I assume that the household is endowed with one unit of productive time
that it supplies inelastically to the market and consumption is the only determinant of
the instantaneous utility function, which is given by:
U(

A;C) =

A + log(C) (2.1)
The instantaneous utility is defined over the agricultural good (

A) and the composite
consumption good (C), which is derived from the industry and services:
C = (
1
I

+ (1
)
1
S

)
1=
; (2.2)
34
whereI is the consumption of the industrial good, andS is the consumption of the ser-
vices. The elasticity of substitution between industrial goods and the services is given
by 1=(1). The weight
influences how non-agricultural consumption expenditure
is allocated between industry and services.
8
At each date, and given prices, the household chooses consumption of each good to
maximize his lifetime utility subject to the budget constraint,
p
A

A +p
I
I +p
S
S =!; (2.3)
wherep
j
is the price of good-j output and! is the wage-rate in the economy.
Firms: There are three goods produced. The production function for sector j is given
by
Y
j
=
j
L
j
; (2.4)
whereY
j
is output of sector j,L
j
is labor allocated to production, and
j
is sector j’s labor
productivity. Since I abstract from capital and fixed factors in production, differences
in labor productivity implicitly incorporates differences due to capital as well as due to
the institutional differences across sectors. Firm j problem is given by
max p
j
Y
j
!L
j
s:t: Y
j
=
j
L
j
; L
j
> 0: (2.5)
Competitive Equilibrium: Given a set of pricesfp
A
;p
I
;p
S
;!g, a competitive equilib-
rium consists of consumption decisions that are the household’s allocationsf

A;I;Sg,
8
The utility function belongs to the following general type of utility function:
U(A;C) =
(

A; ifA<

A,
log(C) +

A; ifA

A.
This specification of preferences implies that the economy specializes in agriculture until the subsistence
level

A is reached. Moreover, the economy will never produce more agricultural good than

A. Once

A
is reached, the representative household will supply labor to the non-agricultural sectors. Technological
progress and this specification of preferences cause structural change, with the economy shifting from a
preponderance of agricultural production to marginalization of the same sector (see Laitner 2000, Stokey
2001, Gollin, Parente, and Rogerson 2002, 2004, 2007).
35
and factor allocations for the firmsfL
A
;L
I
;L
S
g such that given prices, the firm’s allo-
cations solve its profit maximization problem, the household’s allocations solve the
household’s utility maximization problem, and factor and product markets clear:
L
A
+L
I
+L
S
= 1; (2.6)

A =Y
A
; I =Y
I
; S =Y
S
: (2.7)
Sectoral employment shares at a certain date are given as follows.
L
A
=

A=
A
; L
I
=
(1 (

A=
A
))
1 +
; L
S
= 1L
A
L
I
; (2.8)
where (
=(1
))(
I
=
S
)
=(1)
. Employment share in agriculture is determined
solely by the subsistence constraint and labor productivity in agriculture, i.e., increases
in the level of agricultural productivity push labor out of the agricultural sector. When
=( 1) < 1, faster productivity growth in industry leads to Baumol’s prediction
for services with labor moving of industry, i.e., labor goes to the slow-growing service
sector (Baumol 1967 and Nordhaus 2008).
2.1.1 Results for the U.S. Economy
Calibration: I calibrate the model economy to the U.S. data for the period from 1950
to 2005. Sectoral data for the U.S. are from Groningen Growth and Development Cen-
tre (GGDC) 10-sector database. Agriculture includes agriculture, forestry, and fishing.
Industry includes mining and quarrying, manufacturing, utilities, and construction.
The services aggregate is constructed by summing wholesale and retail trade, hotels
and restaurants; transport, storage, and communication; finance, insurance, and real
estate; community, social, and personal services; and government services.
All time series are de-trended using the Hodrick-Prescott filter with a smoothing
parameter of 6.25 before any ratios are computed. I normalize productivity levels across
36
sectors to one in 1950. I use data on sectoral labor productivity growth to obtain the time
paths of sectoral productivity between 1950 and 2005.
I calibrate subsistence in agriculture so that the equilibrium of the model matches
the share of employment in agriculture for 1950. This suggests that

A=0.0875. I calibrate

to match the industrial employment share in 1950 and obtain
=0.3652.
Both

A and
are calibrated independent of the elasticity of the substitution param-
eter. The recent literature provides a range of estimates for 1=(1). Rogerson (2008),
Bah (2009), and Duarte and Restuccia (2010) study a similar closed economy multi sec-
tor models and calibrate an elasticity of 0.44, 0.45, and 0.40, respectively, studying the
U.S. data for the post-1950 period. Ngai and Pissarides (2004, 2008) cite the empirical
literature and argue that the elasticity of substitution lies between 0.1 and 0.3. I study
three values of 1=(1): 0.1, 0.3, and 0.45.
Results: Panel (a) in Figure 2.1 plots the agricultural employment share in the model
and the data. The model generated agricultural employment share seems to capture
the secular movements in the data reasonably well. For example, during the 1950-
1998 period, the model predicts a decline in the agricultural employment share of 7.0
percentage points, which is almost all of the actual decline in the data.
9
The agricultural
employment shares generated by the model are smaller than those in the data after 1998
and the model underpredicts the magnitude of the agricultural employment share by
10.5 percent on average between 1999 and 2005.
10
Panel (b) in Figure 2.1 shows that the model overpredicts the industrial employment
share for almost all the years during the sample period. The model predicts a decline
9
Between 1950 and 1977, the model predicts a decline in the agricultural employment share of 4.8 per-
centage points, which is 82 percent of the actual decline in the data. The model overpredicts the magnitude
of the agricultural employment share by 24 percent on average between 1978 and 1985 and 4 percent on
average between 1986 and 1998.
10
Incorporating land into the agricultural production function does not change the qualitative nature
of the results. I let agricultural production function be YA = A(LA)

, where is the income share of
labor in agriculture and I normalize land to be one. Employment share in agriculture is now given by:
LA = (

A=A)
1=
. I follow Restuccia, Yang, and Zhu (2008) and the references therein and set = 0:7. The
agricultural employment shares generated by the model with land are slightly smaller than those in the
model without land.
37
in the industrial employment share of 5.1 percentage points between 1950 and 2005,
which is only 32 percent of the actual decline in the data when 1=(1) = 0:45. The
model predictions get close to match the data as the elasticity of substitution decreases,
i.e., the model with = 0:1 fits better, relatively, compared to the other two cases. The
intuition is that preferences over the non-agricultural goods approach a Cobb-Douglas
when 1=(1) approaches 1 so that the substitution effect vanishes regardless of the
magnitude of the differences between sectoral productivity differences.
11
1950 1970 1990 2010
1
3
5
7
%9
(a) Agriculture

Model
Data
1950 1970 1990 2010
16
22
28
%34
(b) Industry
1950 1970 1990 2010
57
62
67
72
77
%82
(c) Services

Data
1/(1- )=0.10
1/(1- )=0.30
1/(1- )=0.45
Figure 2.1: Sectoral Employment Shares in the Closed Economy
The gap between the model and the data is increasing steadily. Especially, there
is a large gap between the model and the data in the last three decades. The model
overpredicts the magnitude of the industrial employment share by 27 percent on aver-
age between 1975 and 1990 and 48 percent on average between 1990 and 2005 when
1=(1) = 0:45. The reason that there is a considerable gap between the model and
the actual data for the U.S. industrial employment share in the last three decades is
that productivity gains in industry relative to services are not high enough to move the
11
For example, the model predicts a decline in the industrial employment share of 9.5 percentage points
between 1950 and 2005, which is 60 percent of the actual decline in the data when 1=(1) = 0:1.
38
labor out of the industrial sector. This problem is robust under different values of the
elasticity of substitution between industrial goods and the services.
My findings are consistent with some recent studies. For example, Obstfeld and
Rogoff (2002, Chapter 4) look at the relationship between productivity change and
manufacturing employment and argue that domestic differential productivity gains in
manufacturing may not necessarily explain de-industrialization in today’s riches. Mat-
suyama (2009) presents a simple analytical example, without any quantitative analysis,
demonstrating how misleading writing down a closed economy model can be in the
context of productivity-based theory of manufacturing employment decline. These two
studies do not take the implications of the theory to the data.
Buera and Kaboski (2009b), taking the implications of calibrated closed economy to
the historical data for the structural transformation experience of the U.S. between 1870
and 2000, find that the model cannot account for the steep decline in manufacturing
and rise in services in the later data. Similarly,
˙
Is ¸can (2010) studies a three-sector closed
economy model and finds considerable gaps between the calibrated model and the
actual data, i.e. the data exhibits a relatively sharp decline in manufacturing employ-
ment share over the last three decades of the 20th century, whereas the predicted series
are rather flat.
2.2 Structural Transformation in Open Economy
The world consists of two countries, denoted by i=1,2. Each country is inhabited by
a continuum of identical and infinitely lived households that can be aggregated into a
representative household. The three sectors (goods) are the agricultural sector, denoted
\A", the industrial sector, denoted \I", and the service sector, denoted \S". Goods
are differentiated by the labor productivity with which they are domestically produced
and trade is allowed only in industrial sector. Industrial good produced in different
countries is imperfectly substitutable in consumption. This feature of the model and
39
the degree of substitutability in preferences between home and foreign produced units
of each good endogenously determine a country’s equilibrium pattern of production
and trade.
Households: There is a representative agent in each country who consumes all three
types of good and works. Labor is inelastically supplied. The preference structure of an
agent in country i is given by the following period utility function
U(

A
i
;C
i
) =

A
i
+ log(C
i
): (2.9)
The composite consumption good in country i:
C
i
=

(
i
)
1
(I
i
)

+ (1
i
)
1
(S
i
)

1=
; (2.10)
where 1=(1) is substitution elasticity between industry and services and
i
is share
of industry in non-agricultural consumption in country i. Industrial good consumption
in country i is given by following aggregation:
I
i
=

(
i
)
1
(I
i
i
)

+ (1
i
)
1
(I
i
k
)

1=
; (2.11)
where 1=(1) is the elasticity of substitution between domestically produced and
imported industrial goods, and
i
is home-product consumption bias, i.e.,
i
2 (0:5; 1)
is the weight that households place on their own country’s industrial good. This can
also be interpreted as a stand-in for the explicit introduction of trade costs for goods and
services, which are omitted from the present model. Industrial products from different
countries are imperfect substitutes; thus, demand for products is distinguished by place
of production, i.e.,I
i
k
is the countryi consumption of the industrial goods produced in
countryk.
40
Firms: The production technologies in all sectors are assumed to be linear in labor,
constant returns to scale technologies and given by, for country i=1,2 and j=A,I,S,
Y
i
j
=
i
j
L
i
j
: (2.12)
Output of each sector by country i is given by Y
i
j
. Labor is mobile among the three
sectors and perfect competition prevails in the labor market.L
i
j
represents the employ-
ment in country i sector j and the sectoral productivities are given by
i
j
. The problem
confronted by sector j in country i is the problem of maximizing profits subject to the
production technology;
max p
i
j
Y
i
j
!
i
L
i
j
s:t: Y
i
j
=
i
j
L
i
j
; L
i
j
> 0; (2.13)
wherep
i
j
is the producer price of the good j in country i, and!
i
is real wage in country
i.
Social Planner Problem: I assume complete international financial markets so that I
can solve for the Pareto-optimal allocation, and hence for the Walrasian equilibrium, by
maximizing the social welfare function with strictly positive welfare weights
i
for each
country such that
1
+
2
= 1.
12
The world social planner faces the following problem
feasibility constraints:
The agent in each country is endowed with one unit of time in every period and
allocates his time across alternative employment activities in the domestic country.
L
i
A
+L
i
I
+L
i
S
= 1: (2.14)
12
An allocation is Pareto efficient if and only if it solves the planner’s problem for some welfare weights
in this economy. Since the planning problem involves a maximization of a continuous and concave objec-
tive function subject to a convex and compact constraint set, the solution to the planning problem is unique
and by the Second Welfare Theorem, the competitive equilibrium exists and is unique.
41
Sectoral feasibility constraints for each sector in each country are given by:
Y
i
A
=
i
A
L
i
A
=

A
i
; Y
i
I
=
i
I
L
i
I
=I
i
i
+I
k
i
; Y
i
S
=
i
S
L
i
S
=S
i
: (2.15)
The employment share in agriculture is determined solely by the subsistence con-
straint and labor productivity in agriculture in each country as we observe in the closed
economy solution:
L
i
A
=

A
i
=
i
A
: (2.16)
Industrial employment shares and bilateral exports,fL
1
I
;L
2
I
;I
1
2
;I
2
1
g, in each country are
given by following system of equations (combined with the feasibility conditions):

1

2

1

2

1

1
1
2

1
=

I
1
1
I
2
1

1

I
1
I
2

C
1
C
2

; (2.17)

1

2

1

2

1

1
1

2

1
=

I
1
2
I
2
2

1

I
1
I
2

C
1
C
2

; (2.18)

1
1
1

1

1

1

1
I

1
S

=

I
1
1
I
1

1

I
1
S
1

1
; (2.19)

2
1
2

1

2

1

2
I

2
S

=

I
2
2
I
2

1

I
2
S
2

1
: (2.20)
If p
1
I
and p
2
I
are the prices of the industrial goods produced in each country, then
the terms of trade can be defined asq p
2
I
. In equilibrium, this relative price can be
computed from the marginal rate of substitution in the industrial good consumption
aggregators. If the industrial good in country 1 is chosen as numeraire, i.e., (p
1
I
1),
then the equilibrium sectoral prices are characterized by the following equations:
p
2
I
=

2
1
2

I
2
1
I
2
2

1
=

1
1

1

I
1
1
I
1
2

1
: (2.21)
42
Since labor is freely mobile across sectors within a country and the production functions
of the three final goods sectors are assumed to be linear in labor I have the following
relative prices for agricultural goods and services in each country:
p
1
A
=
1
I
=
1
A
;p
1
S
=
1
I
=
1
S
; p
2
A
=p
2
I
(
2
I
=
2
A
);p
2
S
=p
2
I
(
2
I
=
2
S
): (2.22)
The import ratio is defined by I
1
2
=I
1
1
in country 1. Similarly, the import ratio is
defined byI
2
1
=I
2
2
in country 2. Sectoral productivity differences across countries affect
the sectoral import ratios. In addition, the variability of the terms of trade is influ-
enced by the elasticity of substitution between foreign and domestic industrial goods.
If industrial goods produced in each country are perfect substitutes, i.e., ! 1, then
terms of trade does not vary.
2.2.1 Quantitative Analysis
Calibration: All sectoral productivities are taken from the data, where I measure
productivity as value added per worker. Sectoral data for the Chinese economy are
discussed in Chapter 1 and sectoral data for the U.S. are the same data used in the
closed economy model above.
All time series are de-trended using the Hodrick-Prescott filter with a smoothing
parameter of 6.25 before any ratios are computed. Choosing values for the productivity
levels
i
j
amounts to choosing units; therefore, I normalize those to 1 in 1978. The levels
of sectoral labor productivity for the year 1978 together with data on growth rates of
sectoral value added per worker in local currency units imply time paths for sectoral
labor productivity in each country between 1978 and 2005.
13
13
International comparisons of labor productivity are constrained by a number of measurement issues
and measuring output for calculating the productivity is an issue of debate. Golub (1999, p.51) argues that
the most widely used measure of output for calculating the productivity of labor is constant price value
added since intermediate inputs should be deducted. International trade increasingly takes the form of
trade in components, with the assembly and production of parts taking place at different locations. Notice
that I allow trade in final goods. When all components can be imported, international competitiveness
depends on the value-added price rather than on the output price. Another advantage of using real value
43
Social planner’s weights,
1
and
2
, are calibrated to match the GDP ratio between
the U.S. and China in 1978.
14
. I set the elasticity of substitution between tradables
(industry) and nontradables (services) as 1=(1) = 0:45. I set the elasticity of substitu-
tion between home and foreign goods as 1=(1) = 1:5. It is common in many applied
macroeconomic models to choose values of the elasticity of substitution between 1 and
1.5.
15
I report the results of the model under alternative values of and where I
present sensitivity analysis.

A
i
is calibrated to match employment share of agriculture in 1978 in each country. I
calibrate
i
and
i
to match the average export to output ratios in industry and to match
the industrial employment shares in 1978.
1617
Table 2.1 summarizes the parameter val-
ues.
Table 2.1: Parameter Values for the Benchmark Economy
U.S. China

A Subsistence level of consumption in agriculture 0.0280 0.7080

Utility weight on industrial goods 0.2966 0.5646
Home-bias 0.9516 0.9917
1=(1) Elasticity of substitution between industry and services 0.45 0.45
1=(1) Elasticity of substitution between home and foreign goods 1.5 1.5
China’s Structural Transformation: Panel (a) in Figure 2.2 shows the model predictions
for the agricultural employment share in China. The model overpredicts the transition
added is that it is readily available on a national accounts consistent basis for a wide range of economies
over time.
14
I use the series “Total GDP , in millions of 2009 US$ (converted to 2009 price level with updated 2005
EKS PPPs)” from The Conference Board Total Economy Database Output, Labor and Labor Productivity
Country Details, 1950-2009.
15
See Bodenstein 2010 for a very recent discussion.
16
I use bilateral trade data where the U.S. is reporter country (exports to and imports from China). The
data are from the OECD International Trade by Commodity Database. The SITC rev. 2 codes as proxy for
industry are: 5, 6, 7, 8 minus 68.
17
Alternatively, I can calibrate the home-bias parameters to match the initial year data on imports and
domestic traded goods. In this case, I obtain results very close to the closed economy predictions because
of the fact that China had very low trade volumes in 1978.
44
of labor out of agriculture and the differences between the model’s predictions and the
actual agricultural employment share is large. The model predicts a more dramatic de-
agriculturalization in China than observed in the data, i.e., between 1978 and 2005, the
agricultural employment share falls 35 percent in China but 73 percent in the model.
1978 1988 1998 2008
18
36
54
%72
(a) Agriculture

Data
Model
1978 1988 1998 2008
17
24
31
38
%45
(b) Industry
1978 1988 1998 2008
12
19
26
33
%40
(c) Services

Data
Autarky
Open w/o wedge
Open with wedge
Figure 2.2: Sectoral Employment Shares in China
The only mechanism at work to explain the changes in the agricultural employ-
ment share is productivity growth in agriculture. However, there are mechanisms at
work in the actual Chinese economy that I do not model. For example, there have been
institutional constraints for labor moving out of the agricultural sector such as the resi-
dential registration (hukou) system, under which official sanction was required for any
change of residence. Restrictions on migration confined the majority of the labor force
to the primary sector. This can be one explanation for model’s prediction of dramatic
de-agriculturalization.
18
18
According to Srinivasan (1988, p. 7) labor markets are likely to be segmented and imperfect in early
stages of development when more than two-thirds of the labor force is employed in agriculture. This was
the case in China in the beginning of the economic reforms where the fraction of the labor force engaged
in agriculture was more than 70 percent in 1978. In addition, Young (2000) presents evidence that China
has become fragmented internally while opening up internationally. See also Dekle and Vandenbroucke
2009 for a model of the Chinese structural transformation.
45
Since the purpose of this paper is to examine the effect of the structural transfor-
mation of China on the de-industrialization of the U.S., I introduce time-varying agri-
cultural wedge for China into the model economy feeding the disparity between the
model predicted employment share in agriculture and the actual data to account for all
the possible factors affecting the labor moving out of agriculture.
Panels (b) and (c) in Figure 2.2 show that open economy model predictions with-
out wedge are very close to the autarky series. On the other hand, I capture most of
the annual changes in the non-agricultural employment shares when I introduce agri-
cultural wedge for China. For example, open economy model with wedge predicts an
increase in the service sector share of 12.7 percentage points between 1978 and 2005,
which is 66 percent of the actual increase in the data.
19
U.S. Structural Transformation: Figure 2.3 presents the results for the benchmark open
economy as well as the autarky case and compares with the U.S. data between 1978 and
2005. In terms of industrial employment share, the U.S. data shows a 37.4 percent drop
(from around 28 percent to around 17 percent). The closed economy model (or autarky)
predicts a 14.3 percent decline (from around 28 percent to around 24 percent), while
the open economy predicts a 25.4 percent decline (from around 28 percent to around 21
percent). Hence, the closed economy model accounts for 38.1 percent of the declines in
the U.S. industrial employment share while the open economy accounts for 68.0 percent
of the de-industrialization in the U.S. between 1978 and 2005.
Moreover, the open economy model has more explanatory power to explain the
secular changes in the U.S. industrial employment share in the post-1990 period. For
example, the U.S. industrial employment share data shows a 17.6 percent drop (from
about 21 percent to around 17 percent) between 1992 and 2005. The closed economy
model predicts a 6.6 percent decline while the open economy predicts a 14.9 percent
decline. Thus, the open economy model accounts for 85.1 percent of the declines in the
19
The results for the rest of the paper are based on the model predictions with the so-called time-varying
agricultural wedge for China.
46
U.S. industrial employment while the closed economy accounts for 37.4 percent of the
de-industrialization in the U.S. between 1992 and 2005.
1978 1988 1998 2008
0.8
1.5
2.2
%2.9
(a) Agriculture

Data
Model
1978 1988 1998 2008
17
20
23
26
%29
(b) Industry
1978 1988 1998 2008
69
72
75
78
%81
(c) Services

Data
Autarky
Open
Figure 2.3: Sectoral Employment Shares in the U.S.
The open economy model predicts that the service sector employment share
increases by 12.9 percent (from 69.4 percent to 78.4) while in the data the increase is 16.7
percent (from 69.4 percent to 81.0 percent). Thus, the open economy model accounts for
77.5 percent of the rise in the service sector employment share between 1978 and 2005.
The closed economy predicts that the service sector employment share increases by 8.5
percent (from 69.4 percent to 75.3). This suggests that closed economy model accounts
for 50.7 percent of the rise in the service sector employment share between 1978 and
2005.
Trade: The first two panels in Figure 2.4 show “Trade Balance to GDP” and “Trade
Volume to GDP” ratios for the U.S. The U.S. trade balance (volume) to GDP is equal
to industrial exports to China minus (plus) industrial imports from China divided by
GDP , with all variables in U.S. dollars.
The model does a good job reproducing the shapes of trade balance and trade vol-
ume to GDP . On the other hand, the model predicts higher trade openness compared to
47
the data since model predicted “Trade Volume to GDP” is bigger than the actual data
for each year between 1978 and 2005.
The last panel in Figure 2.4 compares the real exchange rate in the data, with real
exchange rate simulated form the model between 1981 and 2005. For the data, the
market real exchange rate (using the nominal yuan-dollar exchange rate) is defined as
the ratio of the weighted GDP deflators in the two countries in terms of dollars.
20
In the model, the real exchange rate is defined as the ratio of the weighted average
of the output predicted of three sectors in China and in the U.S., in terms of a common
numeraire (industrial sector in the United States). The model predicted series for the
real exchange rate capture the behavior of the time series observed in the data.
1978 1993 2008
-3.2
-2.0
-0.8
%0.4
TB to GDP
1978 1993 2008
0
1.1
2.2
%3.3
TV to GDP
1978 1993 2008
-0.4
0.0
0.4
0.8
1.2
1.6
2.0
2.4
log(RER)
RER

Data
Model
Figure 2.4: Trade Variables
Counterfactuals: I conduct several counterfactual experiments in order to understand
the impact of different sectoral productivity growth rates across countries on the de-
industrialization of the U.S. Figure 2.5 plots the following series: The line “Data” is the
industrial employment share in the U.S. between 1978 and 2005. The line “Autarky”
20
Sectoral value added data in terms of dollars are from the United Nations Statistics Division, National
Accounts Main Aggregates Database and the nominal yuan-dollar exchange rate data are from the Chinese
Statistical Yearbooks.
48
represents the closed economy prediction for the U.S. industrial employment share and
the line “Open” represents the benchmark open economy prediction using the param-
eters reported in Table 2.1.
21
1978 1988 1998 2008
17
20
23
26
%29

Data
Autarky
Open
All sectors
Industry only
Services only
Figure 2.5: Counterfactuals: U.S. De-industrialization
The line “All sectors” plots the result of the experiment where the differences in
productivity growth rates between the U.S and China are shut down in all sectors set-
ting the sectoral productivity growth rates in China equal to the ones observed in the
U.S. Then, I focus on the experiments where the differences in productivity growth rates
between the U.S and China are shut down only in one sector setting that sector’s annual
productivity growth rates in China equal to the U.S. productivity growth rates in the
same sector. The question that guides these experiments for each sector is: What would
happened if the Chinese economy had productivity in agriculture/industry/services as
the U.S.? In all these counterfactual experiments, all the sectoral productivity growth
rates for the U.S. are kept as in the benchmark case.
22
21
These three lines are the exact series used in Panel (b) of Figure 2.3.
22
I do not plot “Agriculture only” series where I set the agricultural productivity growth rate in the U.S.
to the Chinese agricultural productivity growth since the results are exactly the same as the benchmark
case because of the strategy that I match the agricultural employment share in China each year.
49
I find a small wedge between autarky and the counterfactual that shuts down pro-
ductivity differences in all three sectors (see the line “All sectors”). This experiment
predicts a decline in the industrial employment share of 4.6 percentage points between
1978 and 2005, which is only 44 percent of the actual decline in the data. This result
is very close to the autarky prediction, a decline in the industrial employment share
of 4.0 percentage points between 1978 and 2005. This experiment, shutting down pro-
ductivity differences in all three sectors, close down a mechanism of the model that the
countries trade due to the sectoral productivity differences. However, there still is a
role of trade since the stand-in household in a country has tendency to consume some
imported goods governed by the preferences as long as there is no perfect home bias
(no foreign goods consumed). This explains the wedge between the lines “Autarky”
and “All sectors”.
There are hardly any differences between the line “All sectors” and the line “Indus-
try only”. The experiments “Agriculture only” and “Services only”, mimicking the
benchmark open economy, predict a decline in the industrial employment share of 7.1
percentage points, which is 68 percent of the actual decline in the data. These exper-
iments suggest that sectoral productivity differences between China and the U.S are
important for the de-industrialization in the U.S. More importantly, the differences in
industrial productivity growth rate between China and the U.S. are, mainly, responsible
for the changes in the U.S. industrial employment share.
Sensitivity: I am interested in different values for the elasticities in the model. I
vary the elasticity with regard to the imported industrial goods, which is governed by
, and the elasticity between industrial goods and services, governed by. Figure 2.6
presents the sensitivity analysis with respect to these two parameters. I vary only one
elasticity parameter at a time, while keeping other elasticity parameter at the value used
in benchmark economy.
50
The first elasticity parameter I consider is the elasticity of substitution between
traded (industry) and nontraded goods (services). Stockman and Tesar (1995) esti-
mate the elasticity of substitution between traded and nontraded goods in a sample of
30 countries (industrialized and developing countries) regressing the nontraded-good
expenditure share on the price index for nontraded goods (including per capita GDP
to pick up income effects). They find that the elasticity is 0.44. Mendoza (1995) esti-
mates that the elasticity is 0.74 for industrialized countries following the procedure of
Stockman and Tesar (1995).
My benchmark choice for 1=(1) is 0.45. I consider alternative elasticities of 0.30
and 0.74. A higher value of increases the elasticity between industry and services, i.e.,
the stand-in household is more willing to substitute between non-agricultural goods.
For example, the experiment where 1=(1 ) = 0:3 predicts 9.3 percent fall in the
industrial employment share in the U.S., where the actual fall is 10.4 percentage points.
1978 1988 1998 2008
17
20
23
26
%29
Sensitivity for 

Data
1/(1- )=0.30
1/(1- )=0.45
1/(1- )=0.74
1978 1988 1998 2008
17
20
23
26
%29
Sensitivity for 

Data
1/(1- )=1.2
1/(1- )=1.5
1/(1- )=2.0
Figure 2.6: Sensitivity Analysis in the Open Economy
The second elasticity parameter I consider is the elasticity of substitution between
home and foreign goods. This trade elasticity is one of the most debatable parameters in
the empirical international trade literature. An elasticity of 1.5 is the value mostly used
51
by the International Business Cycle (IBC) models. According to Backus, Kehoe, and
Kydland (1994) and Chari, Kehoe, McGrattan (2002), the most reliable studies seem to
indicate that for the U.S. the elasticity is between 1 and 2, and values in this range are
generally used in empirical trade models.
23
My benchmark choice 1=(1) is 1.5. I vary the trade elasticity in both directions
and consider two alternative elasticities 1.2 and 2. A higher value of increases the
elasticity between imported and domestic industrial goods. For example, the experi-
ment where 1=(1) = 2 predicts 37 percent fall in the industrial employment share
in the U.S. Labor in industry falls more than in the benchmark. The more substitutable
are imports, the more accelerated de-industrialization in the U.S.
2.3 Discussion
Several sectoral studies provide a breakdown of convergence and catch-up arguing
that there are large and systematic differences in sectoral productivity across countries
and these sectoral differences are important for understanding movements in aggregate
income and productivity without exploring the role of openness in this catching-up
process.
Bernard and Jones (1996) study the role of sectors in aggregate productivity move-
ments in 14 OECD countries between 1970 and 1987 finding convergence in services
and report that convergence does not hold for the manufacturing sector. Duarte and
Restuccia (2010) find that differences in labor productivity levels between rich and poor
countries are larger in agriculture and services; over time, productivity gaps have been
substantially reduced in agriculture and industry but not nearly as much in services.
23
Ruhl (2008) states that the IBC models need small values of the elasticity to generate the volatility of
the terms-of-trade and the negative correlation between the terms-of-trade and the trade balance that are
found in the data. In contrast, growth models need large Armington elasticity to explain the growth in
trade volumes that result from a change in tariffs (see, e.g., Yi 2003).
52
My contribution is to explain the role of the differences in sectoral productivity
growth rates in China and in the U.S. to explain the secular declines in industrial
employment share in the U.S. I show that differences in sectoral productivity growth
rates across time and across countries (China and the U.S.) play a major role in account-
ing for annual changes in the U.S. industrial employment share, especially in the post-
1990 period.
53
Chapter 3
Productivity and Labor Reallocation:
Latin America vs. East Asia
The development experiences of East Asia (Hong Kong, Korea, Singapore, and Tai-
wan) and Latin America (Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Mex-
ico, Peru, and Venezuela) sharply contrast over the last several decades. To see the
relative performance of these two country groups I analyze how much progress they
made after WWII in catching up with the United States.
Figure 3.1 displays their per capita GDP (measured as 1990 Geary-Khamis dollars)
relative to the United States for the 1950-2008 period. Similarly, Figure 3.2 compares
their path of development between 1960 and 2008, where my measure of development
is GDP per person employed (measured as 1990 Geary-Khamis dollars), which is what
I mean when I say productivity.
1
East Asia starts at a GDP per capita level that is about 11 percent of the United States
in 1950. At that time Latin American GDP per capita is about 28 percent of the United
States. The remarkable catching up of the Asian Dragons is visible where they reach
to about 70 percent of the U.S. level by 2008.
2
For example, the real per capita GDP
in Taiwan grows by a factor of twenty-four from $924 in 1950 to $22,241 in 2008 corre-
sponding to an average annual growth rate of 5.6 percent. With its real GDP per capita
of $30,407 in 2008, Hong Kong is now one of the richest economies in the world. Latin
American countries, on the other hand, show relative stagnation, if not deterioration.
1
The Conference Board, Total Economy Database, January 2009.
2
See, e.g., Hsieh 2002, Vogel 1991, and Young 1995 on the East Asian growth experience.
54
Figure 3.2 displays that the productivity of Latin American countries increased from
33.6 percent of the U.S. level in 1960 to 38.9 percent in 1980. However, Latin America
has experienced a relative deterioration from 1980 on as the productivity of the group
shrunk to 26.6 percent of the U.S. level in 2008. Although Asian countries continued to
grow at a steady pace during the 1980s, economic growth in Latin America declined.
3
1950 1979 2008
0
10
20
30
40
50
60
70
%80
Latin America
East Asia
Figure 3.1: GDP per Capita as a Percentage of the U.S.
Cole, Ohanian, Riascos, and Schmitz (2005) argue that Latin America is a “devel-
opment outlier” since it is the only group of Western countries that have not gained
significant ground on U.S. income levels in the last 50 years. The failure to attain higher
levels of relative income represents what Restuccia (2008) calls the development prob-
lem of Latin America.
Among the fundamental facts is that economic development involves structural
transformation (see, e.g, Kuznets 1973). In this transformation, a substantial and per-
manent shift occurs in the composition of income, output, and employment away from
3
Many observers have called the decade of the 1980s the lost decade of development for Latin America.
See, for example, Krueger (1993, p.1).
55
agriculture and towards industry and services. Figure 3.3 shows the changes in the
employment share of the six sectors in East Asia, Latin America, and the United States:
agriculture, mining and quarrying, manufacturing, utilities, construction, and services.
The source for all the sectoral data I use in this paper is the Groningen Growth and
Development Centre (GGDC) 10-sector database (see Appendix B.1 for details).
1960 1984 2008
0
10
20
30
40
50
60
%70
Latin America
East Asia
Figure 3.2: Labor Productivity as a Percentage of the U.S.
I observe a significant fall in the employment share of agriculture from 33.9 percent
in 1974 to 17.7 percent in 2003 in Latin America. However, while share of employment
in agriculture is proportionally decreasing and reaching very low levels in East Asia
and the United States, the agricultural employment share is still comparatively very
high in Latin America. This observation implies that the relatively low rate of transi-
tion from agriculture to non-agricultural activities as being one of the facts of the Latin
American economic development. The share of services grows from 41.6 percent to 63.3
percent in Latin America as a group. It is evident that the decline in agriculture’s share
of employment has been mirrored in an increase in services share of employment, while
56
manufacturing’s share of employment has been nearly constant. The share of manufac-
turing in employment starts to fall in the 1990s after remaining relatively stable in the
1980s in Latin America.
1974 1984 1994 2004
0
13
26
%39
Agriculture
1974 1984 1994 2004
0
0.4
0.8
%1.2
Mining and Quarrying
1974 1984 1994 2004
8
16
24
%32
Manufacturing
1974 1984 1994 2004
0
0.4
0.8
%1.2
Utilities
1974 1984 1994 2004
4
6
8
%10
Construction
1974 1984 1994 2004
24
44
64
%84
Services

Latin America
East Asia
United States
Figure 3.3: Sectoral Employment Shares, 1974-2003
Why does Latin America not follow the East Asian structural transformation? Why
do we observe higher employment share in agriculture in Latin America compared
to the East Asian countries? Why does manufacturing employment share stay almost
57
constant in Latin America? I develop a multi-sector general equilibrium model of struc-
tural transformation to provide quantitative answers for such questions. The model
abstracts from issues of capital accumulation and international trade and it is effec-
tively a sequence of static resource allocation problems to understand the role of the
domestic sectoral productivity changes on the structural transformation of Latin Amer-
ica and East Asia. The reallocation of labor is solely due to the interaction of productiv-
ity increases with the type of preferences adopted in the model to focus attention on the
following question: Can differences in sectoral productivity growth rates account for
the sectoral reallocation of labor in development outliers (Latin America) and growth
miracles (East Asia)?
Previous sectoral studies provide a breakdown of convergence and catch-up argu-
ing that there are large and systematic differences in sectoral productivity levels and
growth rates across countries and these sectoral differences are important for under-
standing movements in aggregate income and productivity. For example, Bernard and
Jones (1996) study the role of sectors in aggregate productivity movements in 14 OECD
countries between 1970 and 1987.
4
They find convergence in services and report that
convergence does not hold for the manufacturing sector. Services is by far the most sig-
nificant sector accounting for one-third of aggregate convergence to the United States
during the period. Duarte and Restuccia (2010), hereafter DR, find that differences in
labor productivity levels between rich and poor countries are larger in agriculture and
services; over time, productivity gaps have been substantially reduced in agriculture
and industry but not nearly as much in services.
Comparisons of productivity across countries at the sector level are not easy to
make. The crucial point for this comparison is the derivation of sectoral conversion fac-
tors to convert output and/or value added from one national currency unit to another
4
Bernard and Jones (1996) employ data on total factor productivity for (a maximum of) fourteen OECD
countries and six sectors over the period 1970 to 1987. The fourteen countries are Australia, Belgium,
Canada, Denmark, Finland, France, Italy, Japan, Netherlands, Norway, Sweden, United Kingdom, United
States, and West Germany. The six sectors are agriculture, mining, manufacturing, utilities (electric-
ity/gas/water), construction, and services.
58
since exchange rates are poor indicators of relative price levels between sectors, i.e.,
prices of goods and services vary greatly among countries, and the commercial mar-
ket exchange rates used to compare sectoral output do not reliably indicate relative
differences in prices. What are needed are purchasing power parities (PPPs), that is,
the number of foreign currency units required to buy goods and services equivalent to
what can be purchased with one unit of U.S. or another base currency, must be used
for meaningful international comparisons of the relative purchasing power of sectoral
output produced.
DR study a three-sector model of the structural transformation that is calibrated to
the growth experience of the United States, where the United States has been the world
productivity leader and sets the world productivity standards. I measure sectoral labor
productivity differences across countries at a point in time following the approach of
DR because of the lack of comparable PPP-adjusted sectoral output data across a large
set of countries. The sectoral productivity levels in 1974, together with data on growth
in sectoral labor productivity, imply time paths of sectoral labor productivity for each
country.
Specifically, I calibrate and run the model for the United Stated for the years between
1974 and 2003. I choose this sample period since it covers information for all the coun-
tries being studied in this paper. I use the calibrated model economy to measure sectoral
labor productivity differences across countries in 1974. Then, I run the model economy
for each of the Latin American and East Asian countries using country-specific sectoral
productivity series between 1974 and 2003. I compare the model predictions with the
actual data regarding the allocation of employment to the individual sectors in each
country.
Then, I perform two counterfactual experiments to explore why Latin American sec-
toral employment shares in agriculture and manufacturing did not follow East Asian
structural transformation. These experiments suggest that productivity growth in agri-
culture in Latin America has not been high enough to release labor from agriculture as
59
we observe in East Asian countries. Similarly, productivity growth in manufacturing in
Latin America has not been high enough to avoid the stagnation of the manufacturing
sector.
3.1 A Six-Sector Model
I develop a six-sector general equilibrium model, where the sectors are given by
agriculture, mining and quarrying, manufacturing, utilities, construction, and services.
My sectoral classification is consistent with the study of Bernard and Jones (1996).
Instead of studying a two-sector (agriculture and non-agriculture) or a three-sector
(agriculture, industry, and services) model, studying a six-sector model allows for pos-
sible sectoral heterogeneity in industrial activities, introducing mining and quarrying,
construction, and utilities into the economic environment of the model. However, I do
not allow for heterogeneity in the service sector.
5
In the model economy the structural transformation of employment result both
from sectoral differences in productivity growth and from non-homothetic prefer-
ences. Specifically, in each sector, productivity grows exogenously and I impose a
non-homotheticity in preferences affecting the demand for the agricultural good in the
model economy introducing a subsistence level of consumption in agriculture.
Households: The economy is populated by an infinitely-lived representative house-
hold of constant size over time. The population size is normalized to one, without loss
of generality. I assume that the household is endowed with one unit of productive time
that it supplies inelastically to the market and consumption is the only determinant of
the instantaneous utility function, which is given by:
U(

A;C) =

A + log(C) (3.1)
5
In Appendix B.2, I present a model with nine-sectors without doing any aggregation at the sector level
for the services. Moreover, I calibrate the model for each country using the country-specific information
for each country’s sample period.
60
The instantaneous utility is defined over the agricultural good (

A) and the composite
consumption good (C), which is derived from the industry and services:
C = (
1=
1
C
(1)=
1
+
1=
2
C
(1)=
2
+::: +
1=
5
C
(1)=
5
)
=(1)
; (3.2)
where C
1
;:::;C
5
are the consumption of the non-agricultural goods. The weights

1
;:::;
5
influence how non-agricultural consumption expenditure is allocated among
five sectors. The parameter is the (constant) elasticity of substitution among the
non-agricultural goods and it underlies the magnitudes of price responses to quan-
tity adjustments. Lower substitution elasticity implies that sharper price changes are
needed to accommodate a given change in quantities consumed.
At each date, and given prices, the household chooses consumption of each good to
maximize his lifetime utility subject to the budget constraint,
p
A

A +p
1
C
1
+p
2
C
2
+::: +p
5
C
5
=!; (3.3)
wherep
j
is the price of good-j output and! is the wage-rate in the economy.
Firms: At each date there are six goods produced. The production function for sector
j is given by
Y
j
=
j
N
j
; (3.4)
where Y
j
is output of sector j, N
j
is labor allocated to production, and
j
is sector j’s
labor productivity. Since I abstract from capital and fixed factors in production, differ-
ences in labor productivity implicitly incorporates differences due to capital as well as
due to the institutional differences across sectors. Firm j problem is given by
max p
j
Y
j
!N
j
s:t: Y
j
=
j
N
j
; N
j
> 0: (3.5)
Competitive Equilibrium: Given a set of pricesfp
A
;p
1
;p
2
;:::;p
5
;!g, a competitive
equilibrium consists of consumption decisions that are the household’s allocations
61
f

A;C
1
;C
2
;:::;C
5
g, and factor allocations for the firms,fN
A
;N
1
;N
2
;:::;N
5
g such that
given prices, the firm’s allocations solve its profit maximization problem, the house-
hold’s allocations solve the household’s utility maximization problem, and all product
and factor markets clear:
1. The demand of labor from firms must equal this exogenous supply at every date:
N
A
+N
1
+N
2
+::: +N
5
= 1: (3.6)
2. Since there is no international trade or capital accumulation the following condi-
tions hold at each date implying that the market must clear for each goods and
services produced:

A =Y
A
; C
1
=Y
1
; :::; C
5
=Y
5
: (3.7)
Employment share in agriculture is determined solely by the subsistence constraint
and labor productivity in agriculture:
N
A
=

A=
A
: (3.8)
Increases in the level of agricultural productivity push labor out of the agricultural
sector. High agricultural productivity is a necessary condition for industrialization
and economic development since poor countries have a much larger fraction of their
employment in the agricultural sector than rich countries to be able to produce enough
food to satisfy their subsistence requirements. This is consistent with the view in the lit-
erature that the agricultural revolution preceded the industrial revolution. For example,
Nurkse (1961, p. 52) argues that the introduction of the turnip and other improvements
in agriculture led the rise in agricultural productivity and then caused the expansion of
the industrial sector (see Johnston and Mellor 1961 and Johnston 1970 and the references
therein).
62
Employment share in a non-agricultural sectorj is given by:
N
j
=

j

1
j
(1

A=
A
)

1

1
1
+
2

1
2
+
3

1
3
+
4

1
4
+
5

1
5
; j = 1; 2;:::; 5: (3.9)
Productivity increases in sectorj lead to flows of labor out of sectorj, i.e.,@N
j
=@
j
< 0
as long as< 1.
3.2 Quantitative Analysis
Calibration and Predictions of the Model for the United States: The model economy is
calibrated to the year 1974 for the United States. In particular, the model is parameter-
ized so that it matches sectoral labor allocations in 1974. All time series are de-trended
using the Hodrick-Prescott filter with a smoothing parameter of 6.25 before any ratios
are computed. I normalize productivity levels across sectors to one for the initial year.
I use data on sectoral labor productivity growth to obtain the time paths of sectoral
productivity for the sample period.
6
I calibrate subsistence in agriculture so that the equilibrium of the model matches
the share of employment in agriculture for the initial year. Specifically,

A =N
A;1974
: (3.10)
Next, I calibrate
j
to match the share of employment in sector j for the initial year.
Specifically, the weights
1
;:::;
5
are calibrated as follows:

j
=N
j;1974
=(1

A); j = 1;:::; 5: (3.11)
6
The average annual growth in labor productivity between 1974 and 2003 has been highest in agri-
culture (3.59 percent), second in manufacturing (3.47 percent), third in utilities (2.06 percent), fourth in
services (0.89 percent), fifth in mining and quarrying (0.82 percent), and lowest in construction (-0.79 per-
cent).
63
Finally, I choose to match (roughly) the share of employment in manufacturing over
time.
7
Calibrated parameters for the United States are reported in Table 3.1.
Table 3.1: Calibration for the United States
Parameter Value

A 0.0315

1
0.0081

2
0.2177

3
0.0068

4
0.0560

5
0.7114
0.1000
Figure 3.4 shows the model predicted sectoral employment shares and compare
with the actual data for the United States between 1974 and 2003. Notice that the
model predicted agricultural employment share is independent of. By construction,
the model matches exactly the sectoral shares of employment for the initial year used
in the calibration.
The model captures the sectoral trends in each sector’s employment share through-
out the sample period. For example, during the 1974-2003 period, the model accounts
for 85.3 percent of the agricultural employment share on average in the United States.
The model predicts a decline in the agricultural employment share in the United States
of 2.0 percentage points between 1974 and 2003. The actual fall is 1.5 percentage points.
The model almost mimics the 9.9 percentage point decrease in manufacturing from 1974
to 2003.
During the sample period, the model accounts for 97.3 percent of the service sector
employment share on average implying an increase in the share of labor in services from
68.9 percent in 1974 to 76.1 percent in 2003, while the data increases to 80.4 percent in
2003. The model slightly overpredicts the employment shares in mining and quarrying,
utilities, and construction.
7
Ngai and Pissarides (2008) argue that the elasticity of substitution lies between 0.1 and 0.3.
64
1974 1984 1994 2004
1
2.1
%3.2
Data
Model
Agriculture
1974 1984 1994 2004
0.2
0.8
%1.4
Mining and Quarrying
1974 1984 1994 2004
10
16
%22
Manufacturing
1974 1984 1994 2004
0.3
0.7
%1.1
Utilities
1974 1984 1994 2004
5
7.5
%10
Construction
1974 1984 1994 2004
68
75
%82
Services
Figure 3.4: Sectoral Employment Shares, U.S.: 1974-2003
Relative Sectoral Productivity Levels and Time Paths: The calibrated benchmark econ-
omy imposes discipline on the parameters. Given the value of the parameters, I follow
DR and use the model to solve the levels of labor productivity in each sector relative to
those in the United States for the year 1974. Specifically, for each country, I choose the
five labor productivity levels in 1974 to match the sectoral employment shares in each
65
country for the year 1974 and the aggregate labor productivity relative to that of the
United States in 1974.
8
Table 3.2 shows the calibrated levels of labor productivity in each country for the
year 1974. The United States is the productivity leader in 1974 compared to the other
countries in each sector except a few cases. For example, in agriculture, Hong Kong
is the only close competitor at more than 99 percent of the U.S. level. Singapore is
in the second place with more than 85 percent of the U.S. productivity. These results
are not surprising since the magnitude of the agricultural employment share in total
employment in Hong Kong and Singapore are very similar to that of the United States.
Argentina comes in third place with about 20 percent of the U.S. level. Agricultural
productivity levels in other countries are less than 20 percent of the that in the United
States. Bolivia has the worst performance in 1974 at 5.8 percent of the U.S. level, and
performance of Korea is not much better.
Table 3.2: Calibrated Sectoral Productivity Levels in 1974
Country Agriculture Mining Manufac. Util. Construction Services
Hong Kong 0.9945 4.5805 0.2004 0.5209 0.4908 0.6971
Korea 0.0648 0.3754 0.2279 0.4554 0.2848 0.5043
Singapore 0.8524 0.8607 0.3130 0.1356 0.3777 0.4740
Taiwan 0.0953 0.1536 0.1849 0.4121 0.2346 0.5889
Argentina 0.2004 0.8979 0.5224 0.2940 0.3289 0.7360
Bolivia 0.0581 0.0254 0.3515 0.8322 0.2122 0.4649
Brazil 0.0745 0.4226 0.3283 0.1621 0.1790 0.4274
Chile 0.1289 0.1040 0.3838 0.3150 0.2574 0.6915
Colombia 0.0765 0.3166 0.6089 0.3894 0.3935 0.4844
Costa Rica 0.0896 0.9034 0.4458 0.1809 0.2045 0.5758
Mexico 0.2004 0.7625 0.4436 0.2497 0.2793 0.6250
Peru 0.0713 0.1186 0.4630 0.8461 0.2846 0.6673
Venezuela 0.1663 0.5329 1.2768 0.4341 0.5120 1.1054
8
I use The Conference Board, Total Economy Database to get the aggregate labor productivity (for each
country) relative to that of the United States in 1974. I use the series of labour productivity per person
engaged in 1990 US$ (converted at Geary Khamis PPPs).
66
These findings, for Latin America, are consistent with the (internationally compara-
ble) data for the year 1985 constructed by Restuccia, Yang, and Zhu (2008). For exam-
ple, GDP per worker in agriculture relative to the United States is very low except in
Argentina, and Bolivia has the lowest GDP per worker in agriculture relative to the
United States in 1985 among nine Latin American countries, which is similar to my
calibrated productivity levels in agriculture for 1974.
The levels of sectoral labor productivity implied by the model for the year 1974
together with data on growth rates of sectoral value added per worker in local currency
units imply time paths for sectoral labor productivity in each country between 1974
and 2003. Table 3.3 reports the average annual growth labor productivity growth rates
by country and sector for the period 1974 to 2003. I observe very high average annual
growth rates for the East Asian economies.
For example, while in 1974 relative sectoral productivity in Korea are all below
US level, Korea and Taiwan experience higher annualized rates of labor productivity
growth in all of the six sectors. Hong Kong and Singapore show similar higher average
annual growth rate in labor productivity except agriculture in Hong Kong and mining
and quarrying in Singapore.
Table 3.3: Average Annual Growth by Sector (%), 1974-2003
Country Agriculture Mining Manufac. Util. Construction Services
Hong Kong 0.81 6.88 6.99 7.76 1.84 1.89
Korea 5.35 4.08 7.21 10.88 2.30 1.09
Singapore 3.87 0.40 4.47 9.16 -0.22 3.50
Taiwan 4.22 6.00 5.10 6.09 1.35 4.78
Argentina 2.82 3.29 1.02 4.86 0.34 -0.79
Bolivia 2.19 3.62 -1.37 -0.85 -4.25 -1.96
Brazil 4.27 2.93 0.28 6.05 0.39 -0.82
Chile 5.25 5.02 2.94 3.17 1.35 0.34
Colombia 1.63 2.00 -0.31 3.12 -2.17 -0.45
Costa Rica 2.17 3.06 0.67 -0.63 0.32 -0.70
Mexico 2.82 3.29 1.02 4.86 0.34 -0.79
Peru 1.50 3.04 0.34 2.10 2.24 -2.05
Venezuela 0.52 -2.86 -0.13 -4.42 13.09 -2.45
U.S. 3.59 0.82 3.47 2.06 -0.79 0.89
67
Patterns of Structural Transformation Across Countries: Figure 3.5 - 3.10 show the
model predicted sectoral employment shares and compare with the actual data for each
country. Given the calibrated value for subsistence level of consumption in agriculture,
labor productivity growth in this sector implies a share of employment in agriculture
in the model that turns out to be remarkably close to the time series data for each of the
East Asian and the Latin American countries except a few cases.
There are some periods in which the model economy cannot explain the fall in the
agricultural employment share. For example, the model fails to capture the declines
in agricultural employment share for Hong Kong in the post-1991 period. Agricultural
labor productivity grows at an annual average rate of 4.0 percent during 1974-1991 and
then at a rate of -3.5 percent during 1992-2003 in Hong Kong.
Since the agricultural productivity growth is the sole reason of the de-
agriculturalization (the fall in the share of agricultural employment) in the model, the
model keeps silent to explain the movement of labor out of agriculture in these coun-
tries for the periods we observe negative growth rates in the agricultural productivity.
There is a close match between the model and the data in non-agricultural sectors
(given the simplicity of the model). For example, the model almost mimics the 3 per-
centage point decrease in manufacturing employment from 1974 to 2003 for Bolivia.
The model implies an increase in the share of employment in services from 40.0 percent
in 1974 to 55.8 percent in 2003 for Costa Rica, while the data increases to 59.0 percent in
2003. Data and model predictions for all countries and sectors can be compared simi-
larly.
68
1974 1984 1994 2004
0
2
%4
Data
Model
Hong Kong
1974 1984 1994 2004
0
25
%50
Korea
1974 1984 1994 2004
0
2
%4
Singapore
1974 1984 1994 2004
0
17
%34
Taiwan
1974 1984 1994 2004
0
8
%16
Argentina
1974 1984 1994 2004
0
28
%56
Bolivia
1974 1984 1994 2004
0
22
%44
Brazil
1974 1984 1994 2004
0
13
%26
Chile
1974 1984 1994 2004
0
21
%42
Colombia
1974 1984 1994 2004
0
19
%38
Costa Rica
1974 1984 1994 2004
0
10
%20
Mexico
1974 1984 1994 2004
0
24
%48
Peru
1974 1984 1994 2004
0
10
%20
Venezuela
Figure 3.5: Employment Share in Agriculture: 1974-2003
1974 1984 1994 2004
0
0.05
%0.1
Data
Model
Hong Kong
1974 1984 1994 2004
0
1
%2
Korea
1974 1984 1994 2004
0
0.5
%1
Singapore
1974 1984 1994 2004
0
1
%2
Taiwan
1974 1984 1994 2004
0
0.3
%0.6
Argentina
1974 1984 1994 2004
0
4
%8
Bolivia
1974 1984 1994 2004
0
0.5
%1
Brazil
1974 1984 1994 2004
0
2
%4
Chile
1974 1984 1994 2004
0
1
%2
Colombia
1974 1984 1994 2004
0
0.2
%0.4
Costa Rica
1974 1984 1994 2004
0
0.3
%0.6
Mexico
1974 1984 1994 2004
0
1
%2
Peru
1974 1984 1994 2004
0
1.5
%3
Venezuela
Figure 3.6: Employment Share in Mining and Quarrying: 1974-2003
69
1974 1984 1994 2004
0
23
%46
Data
Model
Hong Kong
1974 1984 1994 2004
10
20
%30
Korea
1980 1990 2000
20
25
%30
Singapore
1974 1984 1994 2004
20
30
%40
Taiwan
1974 1984 1994 2004
10
17
%24
Argentina
1974 1984 1994 2004
0
8
%16
Bolivia
1974 1984 1994 2004
10
16
%22
Brazil
1974 1984 1994 2004
10
17
%24
Chile
1974 1984 1994 2004
8
12
%16
Colombia
1974 1984 1994 2004
12
17
%22
Costa Rica
1974 1984 1994 2004
10
17
%24
Mexico
1974 1984 1994 2004
8
12
%16
Peru
1974 1984 1994 2004
8
13
%18
Venezuela
Figure 3.7: Employment Share in Manufacturing: 1974-2003
1974 1984 1994 2004
0
0.5
%1
Data
Model
Hong Kong
1974 1984 1994 2004
0
0.25
%0.5
Korea
1974 1984 1994 2004
0
1
%2
Singapore
1974 1984 1994 2004
0
0.25
%0.5
Taiwan
1974 1984 1994 2004
0
1
%2
Argentina
1974 1984 1994 2004
0
0.25
%0.5
Bolivia
1974 1984 1994 2004
0
1
%2
Brazil
1974 1984 1994 2004
0
0.5
%1
Chile
1974 1984 1994 2004
0
0.5
%1
Colombia
1974 1984 1994 2004
0
2
%4
Costa Rica
1974 1984 1994 2004
0
1
%2
Mexico
1974 1984 1994 2004
0
0.25
%0.5
Peru
1974 1984 1994 2004
0
4
%8
Venezuela
Figure 3.8: Employment Share in Utilities: 1974-2003
70
1974 1984 1994 2004
0
5
%10
Data
Model
Hong Kong
1974 1984 1994 2004
0
5
%10
Korea
1974 1984 1994 2004
0
8
%16
Singapore
1974 1984 1994 2004
0
9
%18
Taiwan
1974 1984 1994 2004
0
5
%10
Argentina
1974 1984 1994 2004
0
7
%14
Bolivia
1974 1984 1994 2004
0
5
%10
Brazil
1974 1984 1994 2004
0
6
%12
Chile
1974 1984 1994 2004
0
5
%10
Colombia
1974 1984 1994 2004
0
5
%10
Costa Rica
1974 1984 1994 2004
0
5
%10
Mexico
1974 1984 1994 2004
0
5
%10
Peru
1974 1984 1994 2004
0
8
%16
Venezuela
Figure 3.9: Employment Share in Construction: 1974-2003
1974 1984 1994 2004
40
62
%84
Data
Model
Hong Kong
1974 1984 1994 2004
25
50
%75
Korea
1974 1984 1994 2004
50
60
%70
Singapore
1974 1984 1994 2004
30
45
%60
Taiwan
1974 1984 1994 2004
50
65
%80
Argentina
1974 1984 1994 2004
25
40
%55
Bolivia
1974 1984 1994 2004
32
48
%64
Brazil
1974 1984 1994 2004
40
55
%70
Chile
1974 1984 1994 2004
40
50
%60
Colombia
1974 1984 1994 2004
30
45
%60
Costa Rica
1974 1984 1994 2004
50
65
%80
Mexico
1974 1984 1994 2004
30
45
%60
Peru
1974 1984 1994 2004
50
60
%70
Venezuela
Figure 3.10: Employment Share in Services: 1974-2003
71
Counterfactual Experiments: I design two counterfactual experiments to understand
the role of the sectoral productivity differences in Latin America and East Asia for the
observed sectoral employment shares. Specifically, I ask the following two questions
to guide numerical experiments: (1) What would have happened to the agricultural
employment share in Latin America if the Latin American countries had had the same
productivity in agriculture as Korea? (2) What would have happened to the manufac-
turing employment share in Latin America if the Latin American countries had had
productivity in manufacturing equal to that of Korea? I choose Korea to represent the
East Asian productivity experience in all of my experiments.
Figure 3.11 shows the data, model prediction, and the experiment where I use
Korean agricultural productivity growth in each of Latin American countries’ agricul-
tural sector. This experiment suggests that productivity growth in agriculture has not
been high enough to release labor from agriculture as we observe in East Asian coun-
tries. For example, if Bolivia had experienced productivity growth in agriculture equal
to that of Korea, then the agricultural employment share in Bolivia would have been
12.0 percent in 2003 instead of 27.4 percent.
Figure 3.12 shows the data, model prediction, and the experiment where I use
Korean manufacturing productivity growth in each of Latin American countries’ man-
ufacturing sector. This experiment suggests a steady decline in each of Latin American
countries’ manufacturing employment share. For example, if Argentina had had the
same productivity growth in the manufacturing sector as Korea, then the manufactur-
ing employment share in Argentina would have been 4.3 percent in 2003 instead of 11.5
percent.
72
1974 1984 1994 2004
0
8
%16
Data
Model
with Korean agricultural
productivity growth
Argentina
1974 1984 1994 2004
0
28
%56
Bolivia
1974 1984 1994 2004
0
22
%44
Brazil
1974 1984 1994 2004
0
13
%26
Chile
1974 1984 1994 2004
0
21
%42
Colombia
1974 1984 1994 2004
0
19
%38
Costa Rica
1974 1984 1994 2004
0
10
%20
Mexico
1974 1984 1994 2004
0
24
%48
Peru
1974 1984 1994 2004
0
10
%20
Venezuela
Figure 3.11: Experiment: Employment Share in Agriculture
1974 1984 1994 2004
0
12
%24
Data
Model
with Korean manufacturing
productivity growth
Argentina
1974 1984 1994 2004
0
8
%16
Bolivia
1974 1984 1994 2004
0
11
%22
Brazil
1974 1984 1994 2004
0
12
%24
Chile
1974 1984 1994 2004
0
8
%16
Colombia
1974 1984 1994 2004
0
11
%22
Costa Rica
1974 1984 1994 2004
0
12
%24
Mexico
1974 1984 1994 2004
0
8
%16
Peru
1974 1984 1994 2004
0
9
%18
Venezuela
Figure 3.12: Experiment: Employment Share in Manufacturing
73
Sectoral Policies: Sector-specific and economywide policies may affect the sectoral
productivities. One of the policy issues yielding different outcomes in different coun-
tries is the direct and indirect government policies affecting agricultural prices. In an
attempt to better understand the interaction of economics and politics of agricultural
pricing policies in developing countries, the World Bank undertook a research project,
A Comparative Study of the Political Economy of Agricultural Pricing Policies, to provide
systematic estimates of the degree of price discrimination against agriculture within
individual countries and to explain how it changed over time; to determine how this
intervention affected such key variables as foreign exchange earnings, agricultural out-
put, and income distribution; and to gain further insight into the political economy of
agricultural pricing policy.
The core of this study is an in-depth analysis of the economic effects of agricul-
tural economic interventions in eighteen developing countries during the period from
1960 to 1985. The countries included in the project are Argentina, Brazil, Chile, Colom-
bia, Cˆ ote d’Ivoire, Dominican Republic, Egypt, Ghana, Republic of Korea, Malaysia,
Morocco, Pakistan, Philippines, Portugal, Sri Lanka, Thailand, Turkey, and Zambia.
The synthesis of findings regarding the economic effects of agricultural pricing policies
can be found in Schiff and Vald´ es (1992). The World Bank’s research project on “Dis-
tortions to Agricultural Incentives” provides the latest work on this topic.
9
. Dennis and
˙
Is ¸can (2010) find that the rate of structural transformation is slow in those countries that
discriminated against their agriculture; and document that in countries with relatively
heavy taxes on their agriculture (See also Fulginiti and Perrin (1993, 1999) for statistical
studies on agricultural economic interventions).
By “agricultural pricing policies ”is meant the entire array of governmental poli-
cies that affect agricultural incomes relative to what they would be in the presence of
a laissez-faire system affecting production incentives by making agriculture more or
9
www.worldbank.org/agdistortions
74
less attractive than other sectors of the economy. These policies include direct interven-
tions to determine agricultural prices, such as government policies determining output
prices, subsidies to inputs such as fertilizers and pesticides, and policies affecting the
costs of transportation and marketing. Indirect interventions affect the prices of agri-
cultural tradables relative to nontradables (through their impact on the real exchange
rate) or to other tradables (as a result of industrial protection).
Table 3.4 shows the direct, indirect, and total rates of nominal protection in agricul-
ture (negative nominal protection rates indicate taxation of agriculture) for Argentina,
Brazil, Chile, Colombia, and Korea during the period from 1960 to 1985. Total interven-
tions taxed agriculture in four Latin American countries. When all price interventions
are considered, the average tax on agriculture of 39 percent is higher than two times the
18 percent tax from direct price interventions in Argentina. That means that farmers in
Argentina received 39 percent less for their output than they would have received in
the absence of total price interventions.
Table 3.4: Direct and Indirect Protection of Agriculture
Period averages in percentages
Indirect Tax due to industrial Direct Total
Country Period protection protection protection protection
Argentina 1960-84 -21.3 -39.5 -17.8 -39.1
Brazil 1969-83 -18.4 -21.4 10.1 -8.3
Chile 1960-83 -20.4 -37.4 -1.2 -21.6
Colombia 1960-83 -25.2 -37.8 -4.8 -30.0
Korea 1960-84 -25.8 -26.7 39.0 13.2
Source: Schiff and Vald´ es 1992, Table 2.1.
In Korea, protection through direct price interventions was larger than taxation
through indirect price interventions. Direct price interventions taxed agriculture in
Argentina, Chile, and Colombia and protected agriculture in Brazil and Korea. The
average reduction in farm prices relative to nonfarm prices because of indirect inter-
ventions was 21.3 percent in Argentina, 18.4 percent in Brazil, 20.4 percent in Chile,
25.2 percent in Colombia, and 25.8 percent in Korea.
75
The average tax on agriculture from industrial protection policies was larger than
the average indirect tax in all five countries implying that industrial protection policies,
i.e., tariffs and quantitative restrictions, had a greater effect on the indirect tax than did
overvaluation of the real exchange rate. The tax on agriculture due to industrial protec-
tion polices was above 20 percent in all five countries and was highest in Argentina.
10
During the process of structural transformation the composition of the manufactur-
ing sector has changed considerably in East Asian countries. In the 1970s and 1980s,
Hong Kong, Korea, Singapore, and Taiwan greatly accelerated the growth of their man-
ufacturing output. In these countries, the growth rate of the manufacturing sector
expanded at double digit growth rates during all periods before the 1980s. On the other
hand, Latin American growth rates never exceeded 10 percent (except in Venezuela
between 1950 and 1960). Moreover, the growth rate of the manufacturing sector was
negative in Argentina, Bolivia, Brazil, Peru, and Venezuela between 1980 and 1990. The
region as a whole saw a sharp drop in the average annual growth rates of manufactur-
ing value added during the 1980s in the context of a significant slowdown in overall
economic growth.
11
The International Comparisons of Output and Productivity (ICOP) Industrial
Database (1987 Benchmark) presents time series on comparative levels of value added
per person employed in manufacturing in Brazil, Korea, Mexico, and Taiwan relative to
the United States. The data indicate that the productivity gap between the Latin Amer-
ican countries and the United States widened, in particular in the 1980s. In the 1990s,
Brazil managed to stabilize the productivity differential, whereas Mexico continued to
loose ground relative to the United States, i.e., by 1998 labour productivity in Brazil
10
Restuccia, Yang, and Zhu (2008) argue that a high share of employment and low labor productivity in
agriculture are mainly responsible for low aggregate productivity in poor countries. Specifically, Restuc-
cia, Yang, and Zhu (2008) employ a two-sector general-equilibrium model and show that differences in
economy-wide productivity, barriers to modern intermediate inputs in agriculture, and barriers in the
labor market generate large cross-country differences in the share of employment and labor productivity
in agriculture.
11
Here, I examine the growth rates of the manufacturing sector of the countries for different subperiods
in constant (national) prices based on non-filtered data.
76
(Mexico) stood at 35.1 (25.8) percent of the U.S. level. On the other hand, Korea and
Taiwan have achieved rapid catch-up between 1963 and 1999. Comparative productiv-
ity levels in manufacturing increased from 11.3 (11.9) percent of the U.S. level in 1963
to 47.3 (30.3) percent in 1999 in Korea (Taiwan).
12
Manufacturing labour productivity
catch up with the U.S. accelerated in Korea in the 1990s compared to the 1970s and
1980s.
The U.S. Bureau of Labor Statistics (BLS) provides international comparisons of
manufacturing productivity and unit labor costs trends for 17 advanced economies
(Australia, Belgium, Canada, Denmark, France, Germany, Italy, Japan, Korea, Nether-
lands, Norway, Singapore, Spain, Sweden, Taiwan, the United Kingdom, and the
United States). According to the latest report available, manufacturing labor productiv-
ity increased in 2007 in 14 of the 17 economies compared.
13
. Korea and Taiwan had the
largest productivity increases of 8.7 percent each. On the other hand, Singapore had the
steepest decline (-4.0 percent) of the three economies where productivity declined. Over
the 2000 - 2007 period, of the 17 economies studied, only Korea, Taiwan, and Sweden
had greater productivity growth in manufacturing than the United States.
Restuccia (2008) argues that barriers to formal market entry, regulation and barriers
to competition, trade barriers and employment protection may be at the core of pro-
ductivity differences between Latin America and the United States. Removing these
barriers could lead to an increase in long-run relative GDP per worker in Latin America
of a factor of four. Cole, Ohanian, Riascos, and Schmitz (2005) document Latin Amer-
ica’s high protectionism, high costs to starting a new business, government ownership
of banks, and stifling labor market regulations at the sector level.
Cole, Ohanian, Riascos, and Schmitz (2005) provide case studies of nationalization
of Venezuelan oil and iron ore, the reversal of nationalization of Chilean copper, the
reversal of the ban on imports of PCs into Brazil (competing with the Brazilian PC
12
All the data are available at http://www.ggdc.net/databases/icop87.htm
13
http://www.bls.gov/news.release/pdf/prod4.pdf
77
industry), privatizing Brazilian iron ore, and the privatization of Mexican and Argen-
tinean state-owned enterprises. In all cases, productivity decreased following national-
ization or quota imposition, and increased following privatization or lifting of quotas.
Brazil began privatization after the mid-1990s. Mexico experienced trade and financial
liberalization partly required by its joining of GATT and NAFTA from 1986 to 1994, and
political liberalization after 1988.
3.3 Discussion
In the comparative analysis of economic performance in terms of structural change
and the sectoral productivity differences, the contrast between East Asia and Latin
America is striking. This Chapter’s contribution is developing a simple, but relatively
detailed, six-sector general equilibrium model and applying it to a nine Latin American
and four East Asian countries using a new data set.
Timmer and de Vries (2009) study the same data set employing the modified shift-
share analysis, which takes account of surplus labour in agriculture and accounts for
the contribution to growth from expanding sectors to decompose GDP per worker, for
19 countries in Asia and Latin America and find that growth accelerations are largely
explained by productivity increases within sectors, which is consistent with my find-
ings.
14
The main policy message is that we need to look deeper into policies that have
affected sectoral productivities over time. I argue that the quantitative findings are
promising for the future research and I discuss two possible avenues of research to
extend the model of this paper. First, labor mobility across the sectors of the economy
is important for an efficient allocation of resources and the barriers to the sectoral labor
movements may have important development consequences.
14
I develop a shift-share analysis of productivity growth and show the individual effects of the separate
branches on aggregate labour productivity growth for each country in Appendix B.3.
78
Vollrath (2009), using data covering the period 1970-1990 that includes sector-
specific measures of physical and human capital, finds that factor market efficiency can
explain nearly 80 percent of the variation in aggregate TFP between countries, and up
to 40 percent of the variation in income per capita. Hayashi and Prescott (2008) employ
a two-sector growth model with barriers to labor mobility between sectors, which
accounts for the Japanese stagnation between 1885 and 1940. Hayashi and Prescott
discuss that were it not for this employment barrier, Japan’s prewar GNP would have
been close to half of the United States’ versus the 36 percent that it actually was.
Hsieh and Klenow (2009), using micro data on manufacturing plants, investigate
the possible role of factor misallocation in China (1998-2005) and India (1987-1994) com-
pared to the United States. When capital and labor are hypothetically reallocated such
that the gaps in marginal products of labor and capital across plants is similar to that
observed in the United States, they calculate manufacturing TFP gains of 30-50 percent
in China and 40-60 percent in India.
De Vries (2009) applies the Hsieh-Klenow model to study allocative efficiency in
Brazil’s retail sector during the period from 1996 to 2006 using a detailed census dataset
of retail firms and finds large potential productivity gains from the reallocation of
resources toward the most efficient retailers. Lagakos (2009) provides a framework on
productivity differences in retail trade to understand aggregate differences. Lagakos
argues that low measured productivity in developing-country retail largely represents
the optimal choice of technology adoption.
Second, the role of trade policy is worth studying since reductions in trade costs and
trade liberalization may cause shifts of comparative advantage across nations which
effect reallocations of production factors. For example, Connolly and Yi (2009) study
the importance of trade reforms in explaining Korea’s growth in output per worker
and trade between 1962 and 1995 and find that the broad Korea tariff reduction can
explain up to 32 percent of Korea’s catch-up in manufacturing output per worker.
79
Bibliography
[1] Alderson, Arthur S. 1999. “Explaining Deindustrialization: Globalization, Fail-
ure, or Success?” American Sociological Review, 64(5): 701-21.
[2] Amiti, Mary, and Caroline Freund. Forthcoming. “The Anatomy of China’s
Export Growth.” In China’s Growing Role in World Trade, ed. Robert Feenstra and
Shang-Jin Wei, The University of Chicago Press.
[3] Ark, Bart van. 1996. “Sectoral growth Accounting and Structural Change in Post-
War Europe.” In Quantitative Aspects of Post-War European Growth, ed. In Bart van
Ark, and Nicholas Crafts, 84-164, Center for Economic Policy Research.
[4] Ark, Bart van, and Marcel Timmer. 2003. “Asia’s Productivity Performance and
Potential: The Contribution of Sectors and Structural Change.” Paper presented
at RIETI-KEIO Conference on Japanese Economy: Leading East Asia in the 21st
Century?, Tokyo.
[5] Armington, Paul S. 1969. “A Theory of Demand for Products Distinguished by
Place of Production.” International Monetary Fund Staff Papers, 16(1): 159-78.
[6] Backus, David K., Patrick J. Kehoe, and Finn E. Kydland. 1994. “Dynamics of the
Trade Balance and the Terms of Trade: The J-Curve?” American Economic Review,
84(1): 84-103.
[7] Bai, Chong-En, Chang-Tai Hsieh, and Yingyi Qian. 2006. “The Return to Capital
in China.” Brookings Papers on Economic Activity, 2: 61-88.
[8] Bao, Shuming, Gene H. Chang, Jeffrey D. Sachs, and Wing T. Woo. 2002. “Geo-
graphic Factors and China’s Regional Development under Market Reforms, 1978-
1998.” China Economic Review, 13(1): 89-111.
[9] Banister, Judith. 2005. “Manufacturing Employment and Compensation in
China.” The U.S. Department of Labor, Bureau of Labor Statistics Research
Report. http://www.bls.gov/fls/chinareport.pdf
[10] Baumol, William J. 1967. “Macroeconomics of Unbalanced Growth: The
Anatomy of Urban Crisis.” American Economic Review, 57(3): 415-26.
80
[11] Bernard, Andrew B., and Charles I. Jones. 1996. “Productivity Across Industries
and Countries: Time Series Theory and Evidence.” Review of Economics and Statis-
tics, 78(1): 135-46.
[12] Blanchard, Olivier, and Francesco Giavazzi. 2006. “Rebalancing Growth in
China: A Three-Handed Approach.” China & World Economy, 14(4): 1-20.
[13] Borensztein, Eduardo, and Jonathan D. Ostry. 1996. “Accounting for China’s
Growth Performance.” American Economic Review Papers and Proceedings, 86(2):
224-28.
[14] Bosworth, Barry, and Susan. Collins. 2007. “Accounting for Growth: Comparing
China and India.” National Bureau of Economic Research Working Paper 12943.
[15] Brandt, Loren, and Xiadong Zhu. 2001. “Soft Budget Constraint and Inflation
Cycles: A Positive Model of the Macro-Dynamics in China during Transition.”
Journal of Development Economics, 64(2): 437-57.
[16] Brandt, Loren, and John Sutton. 2008. “China’s Great Economic Transforma-
tion.” In China’s Great Economic Transformation, ed. Loren Brandt and Thomas G.
Rawski, 1-26. Cambridge: Cambridge University Press.
[17] Brandt, Loren, Chang-Tai Hsieh, and Xiaodong Zhu. 2008. “Growth and Struc-
tural Transformation in China.” In China’s Great Economic Transformation, ed.
Loren Brandt and Thomas G. Rawski, 683-728. Cambridge: Cambridge Univer-
sity Press.
[18] Brandt, Loren, Thomas G. Rawski, and John Sutton. 2008. “China’s Industrial
Development.” In China’s Great Economic Transformation, ed. Loren Brandt and
Thomas G. Rawski, 569-632. Cambridge: Cambridge University Press.
[19] Broda, Christian, and David E. Weinstein. 2006. “Globalization and the Gains
from Variety.” Quarterly Journal of Economics, 121(2): 541-85.
[20] Buera, Francisco J., and Joseph P . Kaboski. 2009a. “The Rise of the Service Econ-
omy.” http://www.econ.ucla.edu/fjbuera/papers/
[21] Buera, Francisco J., and Joseph P . Kaboski. 2009b. “Can Traditional Theories of
Structural Change Fit the Data?” Journal of the European Economic Association, 7(2-
3): 469-77.
[22] Cao, Jing, Mun S. Ho, Dale W. Jorgenson, Ruoen Ren, Linlin Sun, and Ximing
Yue 2009. “Industrial and Aggregate Measures of Productivity Growth in China,
1982-2000.” Review of Income and Wealth, 55(1): 485-513.
[23] Caselli, Francesco, and Wilbur John Coleman II. 2001. “The U.S. Structural
Transformation and Regional Convergence: A Reinterpretation.” Journal of Politi-
cal Economy, 109(3): 584-616.
81
[24] Chari, V . V ., Patrick J. Kehoe, and Ellen R. McGrattan. 2002. “ Can Sticky Price
Models Generate Volatile and Persistent Real Exchange Rates?” Review of Eco-
nomic Studies, 69(3): 533-63.
[25] Chow, Gregory C. 1993. “Capital Formation and Economic Growth in China.”
Quarterly Journal of Economics, 108(3): 809-42.
[26] Chow, Gregory C. 2004. “Another Look at the Growth of Output in China.”
http://www.princeton.edu/ gchow/China output growth.htm
[27] Clark, Colin. 1940. The Conditions of Economic Progress. London: McMillan.
[28] Cole, Harold L., Lee E. Ohanian, Alvaro Riascos, and James A. Schmitz. 2005.
“Latin America in the Rearview Mirror.” Journal of Monetary Economics, 52(1): 69-
107.
[29] Coleman, Wilbur John II. 2007. “Accommodating Emerging Giants.”
http://faculty.fuqua.duke.edu/ coleman/web/ColemanAEG.pdf
[30] Connolly, Michelle, and Kei-Mu, Yi. 2009. “How Much of South Korea’s Growth
Miracle Can be Explained by Trade Policy?” Federal Reserve Bank of Philadelphia
Research Department Working Paper 09-19.
[31] Corden, W. Max. 2007. “Those Current Account Imbalances: A Sceptical View.”
World Economy, 30(3): 363-82.
[32] De Vries, Gaaitzen J. 2009. “Productivity in a Distorted Market: The case of
Brazil’s Retail Sector.” Groningen Growth and Development Centre Research
Memorandum GD-112, Groningen: University of Groningen.
[33] Dean, Judith M, and Mary E. Lovely. Forthcoming. “Trade Growth, Production
Fragmentation, and China’s Environment.” In China’s Growing Role in World Trade,
ed. Robert Feenstra and Shang-Jin Wei, The University of Chicago Press.
[34] Dekle, Robert, and Guillaume Vandenbroucke. 2009. “A Quantitative Analysis
of China’s Structural Transformation.” http://guillaumevdb.net/
[35] Dennis, Benjamin N., and Talan B.
˙
I¸ scan. 2009. “Engel versus Baumol: Account-
ing for Structural Change using Two Centuries of U.S. Data.” Explorations in Eco-
nomic History, 46(2): 186-202.
[36] Dennis, Benjamin N., and Talan B.
˙
I¸ scan. 2010. “Agricultural Distortions, Struc-
tural Change, and Economic Growth: A Cross-Country Analysis.”
http://myweb.dal.ca/tiscan/research/papers/distort.pdf
[37] Dornbusch, Rudiger, Stanley Fischer, and Paul A. Samuelson. 1977. “Compara-
tive Advantage, Trade, and Payments in a Ricardian Model with a Continuum of
Goods.” American Economic Review, 67(5): 823-39.
82
[38] Duarte, Margarida, and Diego Restuccia. 2010. “The Role of the Structural Trans-
formation in Aggregate Productivity.” Quarterly Journal of Economics, 125(1): 129-
73.
[39] Fisher, Alan G.B. 1935. The Clash of Progress and Security. London: McMillan.
[40] Foellmi, Reto, and Josef Zweim ¨ uller. 2008. “ Structural Change, Engel’s Con-
sumption Cycles and Kaldor’s Facts of Economic Growth.” Journal of Monetary
Economics, 55(7): 1317-28.
[41] Frankel, Jeffrey. 2006. “On the Yuan: The Choice between Adjustment under a
Fixed Exchange Rate and Adjustment under a Flexible Rate.” CESifo Economic
Studies, 52(2): 246-75.
[42] Fu, Xiaolan, and V . N. Balasubramanyam. 2003. “Township and Village Enter-
prises in China.” Journal of Development Studies, 39(4): 27-46.
[43] Fuchs, Victor R. 1968. The Service Economy. New York: National Bureau of Eco-
nomic Research.
[44] Fulginiti, Lilyan E., and Richard K. Perrin. 1993. “ Prices and Productivity in
Agriculture.” Review of Economics and Statistics, 75(3): 471-82.
[45] Fulginiti, Lilyan E., and Richard K. Perrin. 1999. “ Have Price Policies Damaged
LDC Agricultural Productivity?” Contemporary Economic Policy, 17(4): 469-75.
[46] Goldstein, Morris. 2004. “Adjusting China’s Exchange Rate Policies.”
http://www.iie.com/publications/wp/wp04-1.pdf
[47] Gollin, Douglas, Stephen L. Parente, and Richard Rogerson. 2002. “The Role of
Agriculture in Development.” American Economic Review Papers and Proceedings,
92(2): 160-64.
[48] Gollin, Douglas, Stephen L. Parente, and Richard Rogerson. 2004. “Farm Work,
Home Work and International Productivity Differences.” Review of Economic
Dynamics, 7(4): 827-50.
[49] Gollin, Douglas, Stephen L. Parente, and Richard Rogerson. 2007. “ The Food
Problem and the Evolution of International Income Levels.” Journal of Monetary
Economics, 54(4): 1230-55.
[50] Golub, Stephen S. 1999. Labor Costs and International Trade. Washington, D. C.:
American Enterprise Institute for Public Policy Research.
[51] Goodfriend, Marvin, and John McDermott. 1998. “Industrial Development and
the Convergence Question.” American Economic Review, 88(5): 1277-89.
[52] Groenewold, Nicolaas, and Lei He. 2007. “The US - China Trade Imbalance: Will
Revaluing the RMB Help (Much)?” Economic Letters, 96(1): 127-32.
83
[53] Groningen Growth and Development Centre. 2009. The Conference Board, Total
Economy Database. Groningen: Groningen Growth and Development Centre.
http://www.conference-board.org/economics/database.cfm
[54] Gourinchas, Pierre-Olivier. 1999. “Exchange Rates Do Matter: French Job Alloca-
tion and Exchange Rate Turbulence, 1984-1992.” European Economic Review, 43(7):
1279-1316.
[55] Hayashi, Fumio, and Edward C. Prescott. 2002. “The 1990s in Japan: A Lost
Decade.” Review of Economic Dynamics, 5(1): 206-35.
[56] Hayashi, Fumio, and Edward C. Prescott. 2008. “The Depressing Effect of Agri-
cultural Institutions on the Prewar Japanese Economy.” Journal of Political Econ-
omy, 116(4): 573-632.
[57] Heston, Alan, and Terry Sicular. 2008. “China and Development Economics. ” In
China’s Great Economic Transformation, ed. Loren Brandt and Thomas G. Rawski,
27-67. Cambridge: Cambridge University Press.
[58] Holtz-Eakin, Douglas. 2003. “The Chinese Exchange Rate and U.S. Manufactur-
ing.” Congressional Budget Office Testimony.
[59] Holz, Carsten A. 2006. “Measuring Chinese Productivity Growth, 1952-2005.”
http://ihome.ust.hk/ socholz/
[60] Hsieh, Chang-Tai. 2002. “What Explains the Industrial Revolution in East Asia?
Evidence from the Factor Markets.” American Economic Review, 92(3): 502-26.
[61] Hsieh, Chang-Tai, and Peter J Klenow. 2009. “Misallocation and Manufacturing
TFP in China and India.” Quarterly Journal of Economics, 124(4): 1403-48.
[62] Hsueh, Tien-tung, and Xiaochao Li. 1999. “Chinese National Economic Account-
ing System.” In Chinas National Income, 1952-1995, ed. TientungHsueh and Qiang
Li , 65-152. Boulder, Colorado: Westview Press.
[63] Hua, Ping. 2007. “Real Exchange Rate and Manufacturing Employment in
China.” China Economic Review, 18(3): 335-53.
[64] Huang, Yiping, and Ron Duncan. 1997. “How Successful Were China’s State Sec-
tor Reforms?” Journal of Comparative Economics, 24(1): 65-78.
[65] Huang, Jikun, and Scott Rozelle. 1996. “Technological Change: Rediscovering
the Engine of Productivity Growth in China’s Rural Economy.” Journal of Devel-
opment Economics, 49(2): 337-69.
[66] Hummels, David, and Peter J. Klenow. 2005. “The Variety and Quality of a
Nation’s Exports.” American Economic Review, 95(3): 704-23.
84
[67]
˙
I¸ scan, Talan B. 2010.“How Much Can Engel’s Law and Baumol’s Disease Explain
the Rise of Service Employment in the United States?”
http://myweb.dal.ca/tiscan/research/papers/unify.pdf
[68] Jaimovich, Nir, and Henry E. Siu. 2009. “The Young, the Old, and the Restless:
Demographics and Business Cycle Volatility.” American Economic Review, 99(3):
804-26.
[69] Jefferson, Gary H., Thomas G. Rawski, and Yuxin Zheng. 1996.“Chinese Indus-
trial Productivity: Trends, Measurement Issues, and Recent Developments.” Jour-
nal of Comparative Economics, 23(2): 146-80.
[70] Jian, Tianlun, Jeffrey D. Sachs, and Andrew M. Warner. 1996.“Trends in
Regional Inequality in China.” China Economic Review , 7(1): 1-21.
[71] Johnston, Bruce F., and John W. Mellor. 1961. “The Role of Agriculture in Eco-
nomic Development.” American Economic Review, 51(4): 566-93.
[72] Johnston, Bruce F. 1970. “Agriculture and Structural Transformation in Develop-
ing Countries: A Survey of Research.” Journal of Economic Literature, 8(2): 369-404.
[73] Kanbur, Ravi, and Xiaobo Zhang. 2005. “Fifty Years of Regional Inequality in
China: A Journey through Central Planning, Reform, and Openness.” Review of
Development Economics, 9(1): 87-106.
[74] Kehoe, Timothy J., and Kim J. Ruhl. 2009. “How Important is the New Goods
Margin in International Trade.” Federal Reserve Bank of Minneapolis Research
Department Staff Report 324.
[75] Klein, Michael W., Scott Schuh, and Robert K. Triest. 2003. “Job Creation, Job
Destruction, and the Real Exchange Rate.” Journal of International Economics, 59(2):
239-65.
[76] Kollmeyer, Christopher. 2009. “Explaining Deindustrialization: How Affluence,
Productivity Growth, and Globalization Diminish Manufacturing Employment.”
American Journal of Sociology, 114(6): 1644-74.
[77] Kongsamut, Piyabha, Sergio Rebelo, and Danyang Xie. 2001. “ Beyond Bal-
anced Growth” Review of Economic Studies , 68(4): 869-82.
[78] Krueger, Anne O. 1993. Political Economy of Policy Reform in Developing Countries.
Cambridge, Massachusetts: The MIT Press.
[79] Kuznets, Simon 1966. Modern Economic Growth. New Haven, CT: Yale University
Press.
[80] Kuznets, Simon. 1973. “Modern Economic Growth: Findings and Reflections.”
American Economic Review, 63(3): 247-58.
85
[81] Kwan, C. H. 1994. Economic Interdependence in the Asia-Pacific Region: Towards a
Yen Bloc. New York: Routledge.
[82] Kwan, C. H. 2002. “The Rise of China as an Economic Power: Implications for
Asia and Japan.” In Japan and China: cooperation, competition, and conflict, ed. Hanns
Gunther Hilpert and Rene Haak, 12-31. New York: Palgrave.
[83] Lagakos, David. 2009. “Superstores or Mom and Pops? Technology Adoption
and Productivity Differences in Retail Trade.” Federal Reserve Bank of Minneapo-
lis Research Department Staff Report 428.
[84] Laitner, John. 2000. “Structural Change and Economic Growth.” Review of Eco-
nomic Studies, 67(3): 545-61.
[85] Lin, Justin Yifu. 1988. “The Household Responsibility System in China’s Agri-
cultural Reform: A Theoretical and Empirical Study.” Economic Development and
Cultural Change, Supplement: Why Does Overcrowded, Resource-Poor East Asia
Succeed: Lessons for the LDCs, 36(3): S199-S224.
[86] Lin, Justin Yifu. 1994. “Chinese Agriculture: Institutional Changes and Perfor-
mance..” In Agriculture and Trade in China and India: Policies and Performance since
1950, ed. T. N Srinivasan, 23-72. California: An International Center for Economic
Growth Publication.
[87] Lin, Justin Yifu, Fang Cai, and Zhou Li. 2003. The China Miracle: Development
Strategy and Economic Reform. 2nd ed. Hong Kong: The Chinese University Press.
[88] Lin, Justin Yifu, and Guofu Tan. 1999. “Policy Burdens, Accountability, and the
Soft Budget Constraint.” American Economic Review Papers and Proceedings, 89(2):
426-31.
[89] Liu, Deqiang, and Keijiro Otsuka. 2004. “A Comparison of Management Incen-
tives, Abilities and Efficiency between SOEs and TVEs: The Case of the Iron and
Steel Industry in China.” Economic Development and Cultural Change, 52(4): 759-80.
[90] Matsuyama, Kiminori. 2008. “Structural Change.” In The New Palgrave Dictionary
of Economics, ed. Lawrence Blume and Steven N. Durlauf, Palgrave Macmillan.
[91] Matsuyama, Kiminori. 2009. “Structural Change in an Interdependent World: A
Global View of Manufacturing Decline.” Journal of the European Economic Associa-
tion, 7(2-3): 478-86.
[92] Maudos, Joaqu´ ın, Jos´ e Manuel Pastor, and Lorenzo Serrano. 2008. “Explaning
the US-EU Productivity Growth Gap: Structural Change vs. Intra-Sectoral Effect.”
Economic Letters, 100(2): 311-13.
[93] McKinnon, Ronald, and Gunther Schnabl. 2006. “China’s Exchange Rate and
International Adjustment in Wages, Prices and Interest Rates: Japan Dj Vu?”
CESifo Economic Studies, 52(2): 276-303.
86
[94] Mendoza, Enrique G. 1995. “The Terms of Trade, The Real Exchange Rate, and
Economic Fluctuations.” International Economic Review, 36(1): 101-37.
[95] National Bureau of Statistics of China. 2006. China Statistical Yearbook, Beijing:
China Statistics Press.
[96] Naughton, Barry. 1995. Growing out of the Plan: Chinese Economic Reform, 1978-
1993. New York, NY: Cambridge University Press.
[97] Naughton, Barry. 1996. “China’s Emergence and Prospects as a Trading Nation.”
Brookings Papers on Economic Activity, 2: 273-337.
[98] Naughton, Barry. 2007. The Chinese Economy : Transitions and Growth. Cambridge:
MIT Press.
[99] Nickell, Stephen, Stephen Redding, and Joanna Swaffield. 2008. “The Uneven
Pace of Deindustrialisation in the OECD.” World Economy, 31(9): 1154-84.
[100] Nurkse, Ragnar. 1961. Problems of Capital Formation in Underdeveloped Countries.
Oxford University Press: New York, (Orig. pub. 1953).
[101] Ngai, Rachel L., and Christopher A. Pissarides. 2004. “Structural Change in a
Multisector Model of Growth.” Centre for Economic Policy Research Working
Paper 4763.
[102] Ngai, Rachel L., and Christopher A. Pissarides. 2007. “Structural Change in a
Multisector Model of Growth.” American Economic Review, 7(1): 429-43.
[103] Ngai, Rachel L., and Christopher A. Pissarides. 2008. “Trends in Hours and Eco-
nomic Growth.” Review of Economic Dynamics, 11(2): 239-56.
[104] Nickell, Stephen, Stephen Redding, and Joanna Swaffield. 2008. “The Uneven
Pace of Deindustrialisation in the OECD” World Economy, 31(9): 1154-84.
[105] Nordhaus, William D. 2008. “Baumol’s Diseases: A Macroeconomic Perspec-
tive.” The B.E. Journal of Macroeconomics, 8(1) (Contributions), Article 9.
[106] Nurkse, Ragnar. 1961. Problems of Capital Formation in Underdeveloped Countries.
Oxford University Press: New York, (Orig. pub. 1953).
[107] Obstfeld, Maurice, and Kenneth Rogoff. 2000. “The Six Major Puzzles in Inter-
national Macroeconomics: Is There a Common Cause?” NBER Macroeconomics
Annual, 15: 339-89.
[108] Obstfeld, Maurice, and Kenneth Rogoff. 2002. Foundations of International
Macroeconomics. 5th ed. Cambridge, MA: The MIT Press.
[109] Obstfeld, Maurice, and Kenneth Rogoff. 2007. “The Unsustainable U.S. Cur-
rent Account Position Revisited.” In G-7 Current Account Imbalances: Sustainability
and Adjustment, ed. Richard Clarida, 339-66, Chicago: The University of Chicago
Press.
87
[110] Organization for Economic Co-operation and Development (OECD). 2001. Mea-
suring Productivity: Measurement of Aggregate and Industry-Level Productivity. Paris:
OECD. http://www.oecd.org/dataoecd/59/29/2352458.pdf
[111] Organization for Economic Cooperation and Development (OECD). 2003. Asia
and Europe: Services Liberalization. Paris: OECD.
[112] Organization for Economic Co-operation and Development (OECD). 2008.
Employment and Labour Market Statistics Summary Tables. Paris: OECD.
[113] Perkins, Dwight H., and Thomas G. Rawski. 2008. “Forecasting China’s Eco-
nomic Growth to 2025.” In China’s Great Economic Transformation, ed. Loren Brandt
and Thomas G. Rawski, 829-86. Cambridge: Cambridge University Press.
[114] Ravn, Morten O., and Harald Uhlig. 2002. “On Adjusting the Hodrick-Prescott
Filter for the Frequency of Observations.” Review of Economics and Statistics, 84(2):
371-76.
[115] Restuccia, Diego. 2008. “The Latin American Development Problem.” University
of Toronto Working Paper 318.
[116] Restuccia, Diego, Dennis Tao Yang, and Xiaodong Zhu. 2008. “Agriculture and
Aggregate Productivity: A Quantitative Cross-Country Analysis.” Journal of Mon-
etary Economics, 55(2): 234-50.
[117] Rogerson, Richard. 2008. “Structural Transformation and the Deterioration of
European Labor Market Outcomes.” Journal of Political Economy, 116(2): 235-59.
[118] Rosenberg, Nathan. 1994. Exploring the Black Box: Technology, Economics, and His-
tory. Cambridge: Cambridge University Press.
[119] Rowthorn, Robert, and Ramana Ramaswamy. 1999. “Growth, Trade and De-
industrialization.” International Monetary Fund Staff Papers, 46(1): 18-41.
[120] Rowthorn, Robert, and Ken Coutts. 2004. “De-industrialization and the Balance
of Payments in Advanced Economies.” Cambridge Journal of Economics, 28(5): 767-
90.
[121] Ruhl, Kim J. 2008. “The International Elasticity Puzzle.”
http://www.kimjruhl.com/research-papers/
[122] Ruoen, Ren. 1997. China’s Economic Performance in an International Perspective.
Paris: OECD.
[123] Schiff, Maurice, and Alberto Vald´ es. 1992. The Political Economy of Agricultural
Pricing Policy. Volume 4: A Synthesis of the Economics in Developing Countries. Balti-
more: Johns Hopkins University Press.
[124] Shuguang, Zhang, Zhang Yansheng, and Wan Zhongxin. 1998. Measuring the
Costs of Protection. Washington, D.C.: Institute for International Economics.
88
[125] Singelmann, Joachim. 1978. From Agriculture to Services: The Transformation of
Industrial Employment. Beverly Hills: Sage Publications.
[126] Srinivasan, T. N. 1988. “Introduction.” In Handbook of Development Economics, ed.
Hollis B. Chenery and T.N. Srinivasan , 3-8, Amsterdam: Elsevier Science Pub-
lishers B. V .
[127] Stavis, Benedict. 1991. “Market Reforms and Changes in Crop Productivity:
Insights from China.” Pacific Affairs, 64(3): 373-83.
[128] Stefanski, Radoslaw 2010. “Structural Transformation and the Oil Price.”
http://www.economics.ox.ac.uk/members/radek.stefanski/
[129] Stockman, Alan C., and Linda L. Tesar. 1995. “Tastes and Technology in a Two-
Country Model of the Business Cycle: Explaining International Comovements.”
American Economic Review, 85(1): 168-85.
[130] Stokey, Nancy. 2001. “A Quantitative Model of the British Industrial Revolution,
1780-1850.” Carnegie-Rochester Conference Series on Public Policy, 55(1): 55-109.
[131] Teignier, Marc 2010. “The Role of Trade in Structural Transformation.”
http://home.uchicago.edu/ marctb/
[132] Timmer, Marcel P ., and Gaaitzen J. de Vries. 2007. “A Cross-Country Database
For Sectoral Employment And Productivity In Asia And Latin America, 1950-
2005.” Groningen Growth and Development Centre Research Memorandum GD-
98, Groningen: University of Groningen.
[133] Timmer, Marcel P ., and Gaaitzen J. de Vries. 2009. “Structural Change and
Growth Accelerations in Asia and Latin America: A New Sectoral Data Set.” Clio-
metrica, 3(2): 165-90.
[134] Trefler, Daniel. 2005. “Service Offshoring: Threats and Opportunities.” Brookings
Trade Forum, 35-60.
[135] United Nations. 2008. National Accounts Main Aggregates Database. New York:
United Nations. http://unstats.un.org/unsd/snaama/Introduction.asp
[136] Ventura, Jaume. 1997. “Growth and Interdependence.” Quarterly Journal of Eco-
nomics, 112(1): 57-84.
[137] Vogel, Ezra F. 1991. The Four Little Dragons: The Spread of Industrialization in East
Asia. Cambridge, MA: Harvard University Press.
[138] Vollrath, Dietrich. 2009. “How Important are Dual Economy Effects for Aggre-
gate Productivity?” Journal of Development Economics, 88(2): 325-34.
[139] Xianchun, Xi. 2002. “Study on Some Problems in Estimating Chinas Gross
Domestic Product.”Review of Income and Wealth, 48(2): 205-15.
89
[140] Yi, Kei-Mu. 2003. “Can Vertical Specialization Explain the Growth of World
Trade?” Journal of Political Economy, 111(1): 52-102.
[141] Yi, Kei-Mu, and Jing Zhang. 2010. “Structural Change in an Open Economy.”
http://www-personal.umich.edu/ jzhang/
[142] Young, Alwyn. 1995. “The Tyranny of Numbers: Confronting the Statistical Real-
ities of the East Asian Growth Experience.” Quarterly Journal of Economics, 110(3):
641-80.
[143] Young, Alwyn. 2000. “The Razor’s Edge: Distortions and Incremental Reform in
the People’s Republic of China.”Quarterly Journal of Economics, 115(4): 1091-1135.
[144] Young, Alwyn. 2003. “Gold into Base Metals: Productivity Growth in the People’s
Republic of China during the Reform Period.” Journal of Political Economy, 111(6):
1220-61.
[145] Zhang, Zhichao. 2001. “China’s Exchange Rate Reform and Exports.” Economics
of Planning, 34(1-2): 89-112.
[146] Zhang, Jun. 2006. “Investment, Investment Efficiency and Economic Growth in
China.” In China, Hong Kong and the World Economy, ed. Lok Sang Ho and Robert
Ash, 273-96. New York: Palgrave Macmillan.
90
Appendix A
Appendix to Chapter 2
Table A.1: Sectoral Shares of Employment and Value Added, G7 Countries
Employment Shares (%)
Agriculture Manufacturing Industry Services
Country/Year 1970 2006 1970 2006 1970 2006 1970 2006
Canada 7.6 2.6 22.3 12.8 29.8 20.8 62.6 76.6
France 13.5 3.4 27.5 15.0 38.4 22.0 48.0 74.6
Germany 8.5 2.3 39.5 22.0 48.7 29.0 42.8 68.8
Italy 20.1 4.3 27.7 21.2 38.8 29.8 41.1 65.9
Japan 16.9 4.1 27.4 18.3 35.7 27.1 47.4 68.8
United Kingdom 3.2 1.3 34.7 13.0 43.2 21.5 53.6 77.2
United States 4.5 1.5 26.4 11.3 33.1 19.9 62.3 78.5
Value Added Shares (%)
Agriculture Manufacturing Industry Services
Country/Year 1970 2006 1970 2006 1970 2006 1970 2006
Canada 4.5 2.2 21.7 18.0 36.1 31.7 59.4 66.1
France 7.7 2.4 24.9 13.7 35.6 21.1 56.7 76.5
Germany 3.7 1.0 35.1 23.7 47.9 30.0 48.5 69.1
Italy 8.8 2.4 27.6 18.8 39.3 27.1 52.0 70.5
Japan 6.3 1.6 34.0 20.2 45.6 28.9 48.1 69.5
United Kingdom 2.9 0.9 31.7 14.0 42.6 24.0 54.6 75.1
United States 3.4 0.9 22.9 13.1 34.1 22.0 62.5 77.1
Sources: U.S. Department of Labor, Bureau of Labor Statistics and United Nations Statistics
Division, National Accounts Main Aggregates Database.
Closed Economy Results for G7 Economies
Chapter 2 explains the details of the closed economy model. Here I present the data and
results for each of the G7 countries. There are some notational changes. The composite
consumption good is given by C = (
1=
I
(1)=
+ (1
)
1=
S
(1)=
)
=(1)
, where
91
the parameter is the elasticity of substitution between industrial goods and the ser-
vices. Employment share in agriculture is given byN
A
=

A=
A
. Employment share in
industry is given by:
N
I
=
(1 (

A=
A
))
1 +
; (
=(1
))(
I
=
S
)
1
=N
I
=N
S
: (A.12)
I calibrate the model economy for each of the G7 economy between 1978 and 2007. I
use GDP by kind of economic activity at constant prices (measured in national curren-
cies) and employment by kind of economic activity to derive labor productivity (value
added per worker) series for each economy between 1978 and 2007. Sectoral value
added data are from the United Nations Statistics Division, National Accounts Main
Aggregates Database. Sectoral Employment data are from the OECD Employment and
Labor Market Statistics, Summary Tables Volume 2008, Release 1. The parameters for
each country are reported in Table A.2.
Table A.2: Calibration
Country

A
Canada 0.0546 0.3091
France 0.0916 0.4048
Germany 0.0569 0.4718
Italy 0.1548 0.4544
Japan 0.1162 0.3969
United Kingdom 0.0272 0.4027
United States 0.0367 0.3265
Figure A.1 presents the results for agriculture and Figure A.2 shows that the closed
economy model fails to match the industrial employment share in the G7 economies.
92
1978 1993 2008
2
4
%6
Canada
1978 1993 2008
2
6
%10
France
1978 1993 2008
2
4
%6
Germany
1978 1993 2008
2
9
%16
Italy
1978 1993 2008
3
7.5
%12
Japan
1978 1993 2008
0.8
1.8
%2.8
United Kingdom
1978 1993 2008
0.6
2.3
%4
United States

Model
Data
Figure A.1: Employment Share of Agriculture: 1978-2007
1978 1993 2008
21
25.5
%30
Canada
1978 1993 2008
20
29
%38
France
1978 1993 2008
29
38
%47
Germany
1978 1993 2008
29
35
%41
Italy
1978 1993 2008
27
33
%39
Japan
1978 1993 2008
21
31
%41
United Kingdom
1978 1993 2008
19
26
%33
United States

Data
=0.1
=0.3
=0.45
Figure A.2: Employment Share of Industry: 1978-2007
93
Appendix B
Appendix to Chapter 3
Data
The source for all the sectoral data used in Chapter 3 is the Groningen Growth and
Development Centre 10-sector database. Timmer and de Vries (2007, 2009) provide a
detailed discussion of the construction of the employment and value added series on a
country-by-country basis. The GGDC 10-sector database includes annual data on gross
value added at both current and constant prices from 1950 to 2005 for selected Latin
America, Asia, and OECD countries.
The database includes data on persons employed, which allows the derivation of
labor productivity (value added per worker) series. Employment in this data set is
defined as “all persons employed ”, thus including all paid employees, but also self-
employed and family workers.
The database covers the ten main sectors of the economy as defined in the Interna-
tional Standard Industrial Classification, Revision 2: (1) agriculture, forestry and fish-
ing, (2) mining and quarrying, (3) manufacturing, (4) construction, (5) public utilities,
(6) wholesale and retail trade, hotels and restaurants (7) transport, storage and com-
munication, (8) finance, insurance, and real estate (9) community, social and personal
services (10) government services. Together these ten sectors cover the total economy.
The database combines data for “Community, Social and Personal Services”and
“Government Services”(except Taiwan and the United States). The services aggregate is
constructed by summing wholesale and retail trade, hotels and restaurants; transport,
storage, and communication; finance, insurance, and real estate; community, social, and
personal services.
94
A Nine-Sector Model
Households: The economy is populated by an infinitely-lived representative household
of constant size over time. Without loss of generality I normalize the population size
to one. I assume that the household is endowed with one unit of productive time that
supplies inelastically to the market and consumption is the only determinant of the
instantaneous utility function, which is given byU(

A;C) =

A + log(C). The instanta-
neous utility is defined over the agricultural good (

A) and the composite consumption
good (C), which is given by
C = (
1=
1
C
(1)=
1
+
1=
2
C
(1)=
2
+::: +
1=
8
C
(1)=
8
)
=(1)
; (B.1)
where C
1
;:::;C
8
are the consumption of the non-agricultural goods. The weights

1
;:::;
8
influence how non-agricultural consumption expenditure is allocated among
eight sectors. The parameter is the (constant) elasticity of substitution among the
non-agricultural goods. The household’s budget constraint is given by:
p
A

A +p
1
C
1
+p
2
C
2
+::: +p
8
C
8
=!; (B.2)
wherep
j
is the price of good-j output and! is the wage-rate in the economy.
Firms: At each date there are nine goods produced. The production function for sec-
tor j is given byY
j
=
j
N
j
, whereY
j
is output of sector j,N
j
is labor allocated to produc-
tion, and
j
is sector j’s labor productivity. Since I abstract from capital and fixed factors
in production, differences in labor productivity implicitly incorporates differences due
to capital as well as due to technology adoption, regulation, etc. across sectors. Firm j
problem is given by
max p
j
Y
j
!N
j
s:t: Y
j
=
j
N
j
; N
j
> 0: (B.3)
95
Competitive Equilibrium: Given a set of pricesfp
A
;p
1
;p
2
;:::;p
8
;!g, a competitive
equilibrium consists of consumption decisions that are the household’s allocations
f

A;C
1
;C
2
;:::;C
8
g, and factor allocations for the firms,fN
A
;N
1
;N
2
;:::;N
8
g such that
given prices, the firm’s allocations solve its profit maximization problem, the house-
hold’s allocations solve the household’s utility maximization problem, and all product
and factor markets clear.
Sectoral Employment Shares: Employment share in agriculture is given by N
A
=

A=
A
. Employment share in non-agricultural sectorj (j = 1; 2;:::; 8) is given by:
N
j
=

j

1
j
(1

A=
A
)

1

1
1
+
2

1
2
+
3

1
3
+
4

1
4
+
5

1
5
+
6

1
6
+
7

1
7
+
8

1
8
: (B.4)
Quantitative Analysis: Calibration procedure is explained in Chapter 3. Calibrated
parameters are reported in Table B.1. Figure B.1 - B.13 show the model predicted sec-
toral employment shares with = 0:1 and = 0:45 and compare with the actual data.
Table B.1: Country-Specific Calibration
Country

A
1

2

3

4

5

6

7

8
Hong Kong 0.032 0.001 0.456 0.006 0.052 0.230 0.077 0.025 0.153
Korea 0.636 0.019 0.209 0.005 0.066 0.327 0.077 0.023 0.273
Singapore 0.035 0.003 0.232 0.011 0.073 0.247 0.127 0.029 0.276
Taiwan 0.509 0.044 0.283 0.010 0.055 0.193 0.091 0.021 0.303
Argentina 0.267 0.007 0.320 0.008 0.064 0.178 0.085 0.048 0.289
Bolivia 0.718 0.119 0.285 0.004 0.079 0.159 0.056 0.011 0.286
Brazil 0.631 0.016 0.334 0.021 0.095 0.186 0.084 0.048 0.215
Chile 0.323 0.079 0.256 0.016 0.069 0.154 0.065 0.035 0.327
Colombia 0.564 0.036 0.258 0.007 0.068 0.111 0.077 0.104 0.340
Costa Rica 0.583 0.007 0.260 0.013 0.092 0.184 0.078 0.039 0.328
Mexico 0.583 0.028 0.285 0.007 0.067 0.202 0.060 0.025 0.327
Peru 0.542 0.044 0.265 0.007 0.070 0.182 0.059 0.071 0.303
Venezuela 0.456 0.056 0.183 0.007 0.119 0.169 0.061 0.060 0.345
96
1974 2005
0.2
1.7
%3.2
model
data
ISIC: 01-05
1974 2005
0
.05
%.1
data
=0.45
=0.1
ISIC: 10-14
1974 2005
5
25
%45
ISIC: 15-37
1974 2005
0
.5
%1
ISIC: 40-41
1974 2005
4
7
%10
ISIC: 45
1974 2005
18
27
%36
ISIC: 50-55
1974 2005
6
11
%16
ISIC: 60-64
1974 2005
0
8
%16
ISIC: 65-74
1974 2005
14
23
%32
ISIC: 75-99
Figure B.1: Sectoral Employment Shares, Hong Kong: 1974-2005
1963 2005
6
35
%64
model
data
ISIC: 01-05
1963 2005
0
1.6
%3.2
data
=0.45
=0.1
ISIC: 10-14
1963 2005
2
15
%28
ISIC: 15-37
1963 2005
0
.2
%.4
ISIC: 40-41
1963 2005
1
5.5
%10
ISIC: 45
1963 2005
12
21
%30
ISIC: 50-55
1963 2005
1
4
%7
ISIC: 60-64
1963 2005
0
9
%18
ISIC: 65-74
1963 2005
8
29
%50
ISIC: 75-99
Figure B.2: Sectoral Employment Shares, Korea: 1963-2005
97
1970 2005
0
2.2
%4.4
model
data
ISIC: 01-05
1970 2005
0
.4
%.8
data
=0.45
=0.1
ISIC: 10-14
1970 2005
17
23
%30
ISIC: 15-37
1970 2005
.2
1
%1.8
ISIC: 40-41
1970 2005
5
11
%17
ISIC: 45
1970 2005
19
23
%27
ISIC: 50-55
1970 2005
3
8
%13
ISIC: 60-64
1970 2005
2
10
%16
ISIC: 65-74
1970 2005
18
24
%30
ISIC: 75-99
Figure B.3: Sectoral Employment Shares, Singapore: 1970-2005
1963 2005
4
28
%52
model
data
ISIC: 01-05
1963 2005
0
2
%4
data
=0.45
=0.1
ISIC: 10-14
1963 2005
12
24
%36
ISIC: 15-37
1963 2005
.2
.4
%.6
ISIC: 40-41
1963 2005
2
8
%14
ISIC: 45
1963 2005
8
17
%26
ISIC: 50-55
1963 2005
4
6
%8
ISIC: 60-64
1963 2005
0
4.5
%9
ISIC: 65-74
1963 2005
12
25
%38
ISIC: 75-99
Figure B.4: Sectoral Employment Shares, Taiwan: 1963-2005
98
1950 2005
4
16
%28
data
model
ISIC: 01-05
1950 2005
0
.4
%.8
data
=0.45
=0.1
ISIC: 10-14
1950 2005
10
19
%28
ISIC: 15-37
1950 2005
0
.8
%1.6
ISIC: 40-41
1950 2005
4
7
%10
ISIC: 45
1950 2005
11
17
%23
ISIC: 50-55
1950 2005
4
6
%8
ISIC: 60-64
1950 2005
3
6
%9
ISIC: 65-74
1950 2005
20
30
%40
ISIC: 75-99
Figure B.5: Sectoral Employment Shares, Argentina: 1950-2005
1950 2003
20
46
%72
data
model
ISIC: 01-05
1950 2003
0
7
%14
data
=0.45
=0.1
ISIC: 10-14
1950 2003
6
13
%20
ISIC: 15-37
1950 2003
0
.2
%.4
ISIC: 40-41
1950 2003
1
5
%9
ISIC: 45
1950 2003
4
14
%24
ISIC: 50-55
1950 2003
1
4
%7
ISIC: 60-64
1950 2003
0
2
%4
ISIC: 65-74
1950 2003
6
18
%30
ISIC: 75-99
Figure B.6: Sectoral Employment Shares, Bolivia: 1950-2003
99
1950 2005
8
36
%64
data
model
ISIC: 01-05
1950 2005
.1
.5
%.9
data
=0.45
=0.1
ISIC: 10-14
1950 2005
12
17
%22
ISIC: 15-37
1950 2005
.2
1.2
%2.2
ISIC: 40-41
1950 2005
3
5.5
%8
ISIC: 45
1950 2005
6
21
%36
ISIC: 50-55
1950 2005
2
3.5
%5
ISIC: 60-64
1950 2005
1
4.5
%8
ISIC: 65-74
1950 2005
6
18
%30
ISIC: 75-99
Figure B.7: Sectoral Employment Shares, Brazil: 1950-2005
1950 2005
0
20
%40
model
data
ISIC: 01-05
1950 2005
0
3
%6
data
=0.45
=0.1
ISIC: 10-14
1950 2005
6
15
%24
ISIC: 15-37
1950 2005
.2
.7
%1.2
ISIC: 40-41
1950 2005
2
7
%12
ISIC: 45
1950 2005
8
15
%22
ISIC: 50-55
1950 2005
2
5
%8
ISIC: 60-64
1950 2005
2
7
%12
ISIC: 65-74
1950 2005
20
35
%50
ISIC: 75-99
Figure B.8: Sectoral Employment Shares, Chile: 1950-2005
100
1950 2005
20
40
%60
model
data
ISIC: 01-05
1950 2005
0
2
%4
data
=0.45
=0.1
ISIC: 10-14
1950 2005
10
14
%18
ISIC: 15-37
1950 2005
0
.3
%.6
ISIC: 40-41
1950 2005
3
5
%7
ISIC: 45
1950 2005
4
21
%38
ISIC: 50-55
1950 2005
3
4.5
%6
ISIC: 60-64
1950 2005
2
5
%8
ISIC: 65-74
1950 2005
13
20
%27
ISIC: 75-99
Figure B.9: Sectoral Employment Shares, Colombia: 1950-2005
1950 2005
12
36
%60
model
data
ISIC: 01-05
1950 2005
.1
.3
%.5
data
=0.45
=0.1
ISIC: 10-14
1950 2005
10
16
%22
ISIC: 15-37
1950 2005
0
1.5
%3
ISIC: 40-41
1950 2005
3
6
%9
ISIC: 45
1950 2005
7
19
%31
ISIC: 50-55
1950 2005
0
3
%6
ISIC: 60-64
1950 2005
1
6
%11
ISIC: 65-74
1950 2005
12
20
%30
ISIC: 75-99
Figure B.10: Sectoral Employment Shares, Costa Rica: 1950-2005
101
1950 2005
14
37
%60
model
data
ISIC: 01-05
1950 2005
0
1.5
%3
data
=0.45
=0.1
ISIC: 10-14
1950 2005
11
16
%21
ISIC: 15-37
1950 2005
.1
.4
%.7
ISIC: 40-41
1950 2005
2
5.5
%9
ISIC: 45
1950 2005
8
17
%26
ISIC: 50-55
1950 2005
2
4
%6
ISIC: 60-64
1950 2005
0
3
%6
ISIC: 65-74
1950 2005
12
21
%30
ISIC: 75-99
Figure B.11: Sectoral Employment Shares, Mexico: 1950-2005
1960 2005
28
42
%56
model
data
ISIC: 01-05
1960 2005
.8
1.9
%3
data
=0.45
=0.1
ISIC: 10-14
1960 2005
9
12
%15
ISIC: 15-37
1960 2005
0
.2
%.4
ISIC: 40-41
1960 2005
2
4
%6
ISIC: 45
1960 2005
8
16
%24
ISIC: 50-55
1960 2005
2
4
%6
ISIC: 60-64
1960 2005
2
4
%6
ISIC: 65-74
1960 2005
13
20
%27
ISIC: 75-99
Figure B.12: Sectoral Employment Shares, Peru: 1960-2005
102
1950 2004
8
29
%48
model
data
ISIC: 01-05
1950 2004
1
4
%7
data
=0.45
=0.1
ISIC: 10-14
1950 2004
6
12
%18
ISIC: 15-37
1950 2004
0
4
%8
ISIC: 40-41
1950 2004
0
9
%18
ISIC: 45
1950 2004
8
23
%38
ISIC: 50-55
1950 2004
3
6
%9
ISIC: 60-64
1950 2004
3
5
%7
ISIC: 65-74
1950 2004
18
28
%38
ISIC: 75-99
Figure B.13: Sectoral Employment Shares, Venezuela: 1950-2004
Shift-Share Analysis of Productivity Growth
At the aggregate level productivity is essentially driven by two sources. The first source
of productivity growth concerns productivity growth within individual sectors. The
second source involves shifts of resources from sectors with either low productivity
levels and/or low productivity growth rates to sectors with high productivity levels
and/or high productivity growth rates. This source is referred to as structural change.
To measure the effect of the contribution of employment shifts on the overall pro-
ductivity growth, one may express the productivity for the economy as a whole as the
productivity level by sector weighted by the sectoral employment shares:
103
Y
T
L
T

Y
0
L
0
=
J
X
j=1

j0
(
Y
jT
L
jT

Y
j0
L
j0
)
| {z }
Intra-Sectoral Effect
+
J
X
j=1
(
jT

j0
)
Y
j0
L
j0
| {z }
Static Sectoral Effect
+
J
X
j=1
(
jT

j0
)(
Y
jT
L
jT

Y
j0
L
j0
)
| {z }
Dynamic Sectoral Effect
| {z }
Structural Change Effect
:(B.5)
Here
Y
T
L
T

Y
0
L
0
is the labor productivity growth between years 0 andT ,j is the sector,
and
jT
is the share of employment in sectorj in yearT (see, i.e., Ark 1996, and Mau-
dos, Pastor, and Serrano 2008 for the details of the shift-share analysis of productivity
growth).
The intra-sectoral effect shows the part of the overall productivity change which is
caused by productivity growth within the sectors. The structural change effect captures
the effect of the re-allocation of factors between sectors. The static sectoral effect (or
the net shift effect) is the part of the overall productivity change which is caused by
the changes in sectoral employment shares. It is the growth that would have occurred
if there had been no change in the productivity of any sector during the period ana-
lyzed. The dynamic sectoral effect (or the interaction effect) represents the joint effect
of changes in employment shares and sectoral productivity.
Table B.2 shows the quantitative magnitudes of the three sources of the aggregate
productivity for different countries. The intra-sectoral effect dominates the outcome in
all the countries except in Hong Kong and Mexico. For example, during the period
1970-2005, productivity growth within the 9 sectors explains 86.1 percent of the aggre-
gate growth in Singapore. Structural change explains the remaining 13.9 per cent: 19.2
percent is due to the static sectoral effect, and minus 5.3 percent is due to the dynamic
sectoral effect.
Next, I take a more detailed look at the contribution of the various sectors using the
decomposition formula above.
1
Table B.3 shows the individual effects of the separate
1
See Ark and Timmer (2003) for a similar approach for some Asian countries.
104
branches on aggregate labour productivity growth. For example, aggregate labour pro-
ductivity grows at 4.2 percent annually during the period 1963-2005 in Korea. Of this,
77.4 percent is explained by intra-sectoral productivity growth, 23.2 percent by static
shift effects and minus 0.6 percent by dynamic shift effects (see Table B.2).
Table B.2: Relative Contribution of Different Sources (%)
Country Period Intra-Sectoral Static Sectoral Dynamic Sectoral Total
Hong Kong 1974 - 2005 59.5 90.3 -49.8 100.0
Korea 1963 - 2005 77.4 23.2 -0.6 100.0
Singapore 1970 - 2005 86.1 19.2 -5.3 100.0
Taiwan 1963 - 2005 70.9 8.0 21.1 100.0
Argentina 1950 - 2005 144.2 46.7 -90.9 100.0
Bolivia 1950 - 2003 178.8 148.7 -227.4 100.0
Brazil 1950 - 2005 75.5 43.7 -19.1 100.0
Chile 1950 - 2005 136.5 26.9 -63.4 100.0
Colombia 1950 - 2005 98.1 54.5 -52.6 100.0
Costa Rica 1950 - 2005 75.8 43.5 -19.3 100.0
Mexico 1950 - 2005 53.7 54.9 -8.5 100.0
Peru 1960 - 2005 73.7 64.1 -37.7 100.0
Venezuela 1950 - 2005 2342.8 -19.1 -2223.6 100.0
The columns report the separate effects: intra, static- and dynamic-shift effects and
the total effect. For example, the agricultural sector (ISIC 01-05) contributes in total
minus 0.3 percent to aggregate labour productivity growth. In fact, it contributes posi-
tively (40.2 percent) through productivity growth in the agricultural sector itself. How-
ever, the shift effect is strongly negative over the whole period, because the labour share
of agriculture declines sharply over this period. On the other hand, the manufacturing
sector (ISIC 15-37) contributes positively to labour productivity growth (48.2 percent in
total) in Korea.
When looking at the total contribution of the nine sectors in each country, it is clear
that the manufacturing sector contributes most to the aggregate productivity growth in
most of the cases (except Argentina, Bolivia, Brazil, Colombia, and Hong-Kong).
105
Table B.3: Sectoral Contribution (%)
Hong Kong (1974 - 2005) Korea (1963 - 2005)
Intra Static Dynamic Intra Static Dynamic
Sectoral Sectoral Sectoral Sectoral Sectoral Sectoral
ISIC Codes Effect Effect Effect Effect Effect Effect
01-05 0.1 -0.9 -0.1 40.2 -5.5 -35.0
10-14 0.1 0.0 -0.1 2.7 -0.4 -2.5
15-37 28.9 -3.4 -24.4 19.6 1.7 27.0
40-41 5.1 -0.2 -1.4 2.3 0.0 1.6
45 0.5 1.4 0.3 2.6 1.6 5.9
50-55 22.1 4.2 8.8 4.9 2.3 5.8
60-64 5.9 2.4 2.2 5.8 0.6 7.6
65-74 -6.0 77.1 -37.1 -1.0 14.2 -11.1
75-99 2.6 9.8 2.0 0.2 8.6 0.2
TOTAL 59.5 90.3 -49.8 77.4 23.2 -0.6
Singapore (1970 - 2005) Taiwan (1963 - 2005)
Intra Static Dynamic Intra Static Dynamic
ISIC Codes Effect Effect Effect Effect Effect Effect
01-05 0.7 -0.6 -0.6 12.7 -2.4 -11.2
10-14 0.1 -0.1 -0.1 7.9 -0.9 -7.6
15-37 29.8 -0.3 -0.8 14.3 1.8 13.8
40-41 4.3 -0.4 -2.3 3.7 -0.1 -1.4
45 0.6 0.8 0.1 0.5 0.8 0.9
50-55 21.1 -0.9 -2.3 8.9 1.2 12.6
60-64 15.9 -0.1 -0.9 7.2 0.0 0.4
65-74 1.2 22.3 4.2 1.0 5.4 6.9
75-99 12.5 -1.5 -2.7 14.7 2.1 6.8
TOTAL 86.1 19.2 -5.3 70.8 8.0 21.1
106
Table B.3 – continued from previous page
Argentina (1950 - 2005) Bolivia (1950 - 2003)
Intra Static Dynamic Intra Static Dynamic
ISIC Codes Effect Effect Effect Effect Effect Effect
01-05 54.2 -9.9 -38.5 83.1 -23.6 -52.5
10-14 9.8 -0.4 -4.4 113.0 -34.0 -92.5
15-37 77.1 -21.1 -39.5 2.1 18.1 1.4
40-41 6.5 0.3 1.7 0.4 3.4 1.3
45 -2.3 8.2 -1.2 -1.4 12.4 -5.8
50-55 -4.7 23.7 -3.1 -11.7 83.5 -55.5
60-64 15.8 2.6 3.6 4.4 12.5 11.2
65-74 -2.3 10.8 -3.0 -3.7 46.5 -26.8
75-99 -9.8 32.5 -6.5 -7.4 29.9 -8.3
TOTAL 144.2 46.7 -90.9 178.8 148.7 -227.4
Brazil (1950 - 2005) Chile (1950 - 2005)
Intra Static Dynamic Intra Static Dynamic
ISIC Codes Effect Effect Effect Effect Effect Effect
01-05 32.8 -5.1 -23.1 27.2 -3.6 -17.9
10-14 5.0 -0.1 -1.4 47.7 -4.1 -35.1
15-37 20.1 0.6 1.4 40.8 -3.7 -16.4
40-41 8.6 -0.6 -4.6 7.9 -0.3 -2.9
45 3.3 2.0 2.0 1.0 6.5 0.9
50-55 -1.0 8.7 -2.1 3.7 6.8 4.1
60-64 4.0 0.5 1.8 6.7 1.6 4.5
65-74 0.9 15.3 2.6 -0.2 20.2 -0.9
75-99 1.6 22.5 4.3 1.8 3.5 0.3
TOTAL 75.5 43.7 -19.1 136.5 26.9 -63.4
107
Table B.3 – continued from previous page
Colombia (1950 - 2005) Costa Rica (1950 - 2005)
Intra Static Dynamic Intra Static Dynamic
ISIC Codes Effect Effect Effect Effect Effect Effect
01-05 39.0 -14.3 -22.5 40.7 -9.8 -28.6
10-14 7.8 -0.5 -1.1 0.2 0.0 -0.1
15-37 19.5 -0.6 -1.0 19.6 3.6 8.1
40-41 8.8 -0.3 -3.3 0.7 1.8 1.7
45 1.8 3.6 1.4 1.5 1.2 0.6
50-55 -8.3 58.1 -38.4 -2.3 24.3 -4.8
60-64 4.6 3.8 3.9 12.3 1.2 7.9
65-74 7.3 2.1 5.4 -1.0 17.4 -5.6
75-99 17.5 2.6 3.0 4.1 3.9 1.5
TOTAL 98.1 54.5 -52.6 75.8 43.5 -19.3
Mexico (1950 - 2005) Peru (1960 - 2005)
Intra Static Dynamic Intra Static Dynamic
ISIC Codes Effect Effect Effect Effect Effect Effect
01-05 25.5 -8.4 -18.6 19.4 -8.5 -7.3
10-14 5.0 -0.8 -3.1 18.3 -6.0 -7.5
15-37 14.2 4.8 6.7 29.8 -5.2 -5.9
40-41 1.6 0.2 1.1 5.3 -0.1 -0.3
45 -0.5 6.4 -1.1 9.2 -0.2 -0.3
50-55 -2.0 29.3 -3.7 -13.9 51.3 -20.2
60-64 6.6 5.6 7.1 1.6 10.5 1.6
65-74 0.2 5.7 0.7 6.9 5.7 3.6
75-99 3.2 12.1 2.3 -3.0 16.5 -1.5
TOTAL 53.7 54.9 -8.5 73.7 64.1 -37.7
108
Table B.3 – continued from previous page
Venezuela (1950 - 2005)
Intra Static Dynamic
ISIC Codes Effect Effect Effect
01-05 671.2 -136.6 -496.2
10-14 -207.6 -974.2 105.8
15-37 560.3 -19.0 -33.6
40-41 1.6 62.9 25.5
45 1485.1 -306.5 -1334.4
50-55 -282.5 897.0 -551.0
60-64 -17.2 201.1 -18.9
65-74 36.1 70.3 33.5
75-99 95.7 186.0 45.8
TOTAL 2342.8 -19.1 -2223.6
109

Abstract (if available)

Abstract Since the pioneering empirical works of Colin Clark, Alan Fisher, and Simon Kuznets, economists have agreed that the achievement of sustained economic growth, and a permanently higher level of income per capita, is strongly associated in the data with a structural transformation. In this transformation, a substantial and permanent shift occurs in the composition of income, output, and employment away from agriculture and towards industry and services. This thesis comprises three essays on structural transformation in a global world.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Empirical essays on trade liberalization and export diversification

PDF

Growth, trade and structural change in low income industrializing economies

PDF

Three essays on the credit growth and banking structure of central and eastern European countries

PDF

Monetary policy and the term structure of interest rates

PDF

Growth and development in Africa: macroeconomic and microeconomic perspectives

PDF

Essays in international economics

PDF

Empirical studies of monetary economics

PDF

Effects of eliminating the unfunded social security system in an economy with entrepreneurs

PDF

Structure, agency, and the Kuznets Curve: observations and implications for sustainability planning in U.S. cities

PDF

Investigation of various factors behind non-deaccummulation of housing and wealth with aging

PDF

Inter-temporal allocation of human capital and economic performance

PDF

Trade, training, employment, and wages: evidence from the U.S. manufacturing industry

PDF

Legal institutions of the market economy: private commercial arbitration as transaction costs reducer in Colombia

PDF

Globalization, curricular elements, organizational structures and perceived student outcomes in California schools

PDF

Mater of transformation. Poiesis and autopoiesis in Ovid's Metamorphoses, Benvenuto Cellini's Vita and Pier Paolo Pasolini's Petrolio

PDF

Urban spatial transformation and job accessibility: spatial mismatch hypothesis revisited

PDF

Cost structure and regulation in the telecommunications industry

PDF

The planning, production, and perception of prosodic structure

PDF

Political economy of private commercial arbitration in Colombia

PDF

Legal origin and bilateral and multilateral dispute settlement mechanisms: the case of patent protection legislation

Asset Metadata

Creator Ungor, Murat (author)

Core Title Structural transformation and globalization

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Economics

Publication Date 07/09/2010

Defense Date 05/11/2010

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag china,deindustrialization,East Asia,industrialization,international macroeconomics,Latin America,OAI-PMH Harvest,sectoral productivity differences,structural transformation

Place Name China (countries), East Asia (region), Latin America (region), USA (countries)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Betts, Caroline M. (committee chair), Dekle, Robert (committee member), Imrohoroglu, Ayşe (committee member)

Creator Email muratungor99@hotmail.com,ungor@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m3181

Unique identifier UC1196695

Identifier etd-Ungor-3916 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-364600 (legacy record id),usctheses-m3181 (legacy record id)

Legacy Identifier etd-Ungor-3916.pdf

Dmrecord 364600

Document Type Dissertation

Rights Ungor, Murat

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu