Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
16651691, 2008
2008 Elsevier Ltd. All rights reserved
0305-750X/$ - see front matter
www.elsevier.com/locate/worlddev
doi:10.1016/j.worlddev.2007.09.013
1. INTRODUCTION
The rapid economic growth of China since
the beginning of economic reforms in 1978
has captured the imagination of Western commentators and researchers. The responses range
from outright pessimism about Chinas future
to fear of a strong China. Brown (1995) wondered who will feed China. Chang (2001) announced the coming collapse of China.
Henderson (1999) saw China as on the brink,
while Terrill (2003) wrote of the illusory nature of the market in most of the Chinese economy and that a crash looms because the
Leninist core of the regime is unchanged from
Maos construction of it in Yanan six decades
ago (pp. 329, 313).
At the other end of the spectrum are those who
envisage a strong China. Murray (1998) described China as the next superpower. A number
of authors view an all-powerful China as a threat
(Gertz, 2000, or Timperlake et al., 2002). 1 News
items imploring the devastation Chinese competitors are inicting on US industries, from
kitchenware and car tires to electronic circuit
boards and the futility of trying to match the
China price have become common fare. 2
1665
1666
WORLD DEVELOPMENT
1667
1.
2.
3.
4.
5.
6.
Price adjustment?
Aggregate
GDP
National
China coast
China 5 prov.
2011
2009
2015
2012
2033
2024
2035
2026
2040
2029
2057
2041
2024
2019
2027
2021
2041
2031
2021
2017
2023
2019
2036
2028
Base year for the projections is 2005. Annual growth rates used in the projections are the average annual real growth
rates of China and the United States in 19782005. The price adjustment factor of 3.86 is that of 2004, the most
recent year for which the PWT 6.2 provides data (The values of 200004 are: 4.217306, 4.223668, 4.264024, 4.186857,
3.859263).
China coast refers to the coastal provinces of China (in comparison to the national average in the United States); it
includes Beijing, Tianjin, Hebei, Liaoning, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong, and
Guangxi (the small province Hainan was omitted due to a lack of data for 1978 when Hainan was part of Guangdong
province). China 5 prov. refers to the ve fastest-growing coastal provinces in China (Jiangsu, Zhejiang, Fujian,
Shandong, and Guangdong). The share of the coast in Chinas population in mid-2005 is 42.5%, and in annual GDP
66.2%; the corresponding shares of the ve provinces are 25.9% and 43.3%.
Nominal and real GDP values of 1978 and 2005 are those published by the National Bureau of Statistics in China
and by the Bureau of Economic Analysis in the United States. Per capita GDP values are obtained by dividing
aggregate GDP by the mid-year population. United States population data for the year 2005 is the ocial estimate.
Chinese provincial-level population data of 1978 are based on the July 1, 1982 and the November 1, 2000 censuses,
with mid-1978 values obtained by extrapolation using the implicit average annual population growth rate of 1982
2000.
The average annual real growth rate of economy-wide GDP in China in 19782005 is 9.6358%; in the United States,
it is 2.9687%. The average annual growth rate of real GDP per capita in China is 8.3840%, and in the United States it
is 1.8821%. The average annual population growth rate in China is 1.1549% and in the United States it is 1.0665%.
The average exchange rate in 2005 is 8.1917 yuan RMB/USD.
Sources: PWT 6.2. Census 2000 Major Figures, pp. 16, 31. Statistical Yearbook 1991, p. 79; 2006, pp. 57, 60, 99, 734.
Bureau of Economic Analysis, at <https//www.bea.gov/bea/dn/home/gdp.htm> (choose Current-dollar and real
GDP, XLS; accessed on 30 Nov. 06). US population data are from <http://www.census.gov/popest/archives/1990s/
popclockest.txt> and for 200005 from <http://www.census.gov/popest/states/NST-ann-est.html> (both accessed on
November 30, 2006).
1668
WORLD DEVELOPMENT
Extrapolations do not constitute a reliable research tool. The following two sections employ
theories and tools of economics to examine
Chinas future growth and its potential determinants.
3. DEVELOPMENT THEORIES AND
ASIAN PRECEDENTS
Explanations of Chinas economic growth
tend to focus on economic transition. The success of the reform process is explained by transition facts and strategies, where success is
usually taken to imply economic growth (or a
rise in living standards). For example, Woo
(1994) listed as crucial the creation of non-state
rms in every sector of the economy, a high
savings rate, good initial conditions (such as a
small extent of central planning, unemployment
in the countryside that could be taken up by
township and village enterprises, or a limited
social security net), historical conditions, and
the Chinese Diaspora. Qian (2000, 2003) ascribed much of Chinas reform success to the
unorthodox economic policy measures adopted
by Chinas leadership; the key to Chinas economic growth was the unleashing of incentives
and competition while making reform interestcompatible for those in power.
Yet these explanations of past economic reform successes (and thus economic growth) lack
a counterfactual. Wing-Thye Woo oers crosscountry comparisons, but the number of explanatory variables appears larger than the number
of comparison countries. Qian Yingyi makes
an in part historical argument, but oers no
explicit time series evaluation that examines the
status before and after a particular reform measure was implemented. These explanations, thus,
are not as strong as one might wish them to be. 13
If one takes the view that economic transition
in total has caused Chinas past economic
growth, then one could argue that the key elements of transition have been in place since
the early 1990s (price and domestic trade liberalization, the entry of private enterprises) and
that therefore the gains from transition have already been exhausted. Consequently, economic
growth should since have slowed, which it did
not, or it could slow any time soon. Alternatively, one could argue that past transition measures impact on economic growth over an
extended period of time, or that transition is
as yet incomplete, with further measures to
go, in which cases the gains continue.
Growth patterns identied by economic theories of development and trade perhaps oer rmer ground. Economic transition could then be
viewed as the removal of constraints that prevented well-known development patterns from
unfolding. Furthermore, if Chinas reform period growth patterns match those exhibited by
other countries at stages in their economic
development similar to Chinas development level in the reform period, then Chinas future
growth patterns may also match those of the
other economies later. The argument has two
foundations: one is a focus on standard growth
patterns, the other is a cross-country comparison with countries with which a comparison is
likely to be meaningful.
The growth patterns are structural change,
catching up, and factor price equalization. 14
These three patterns are not independent of
each other, nor do they hold irrespective of
the larger economic environment; they do not
always come with perfectly clear-cut causality
(though one causality, perhaps the most plausible one, is used to present each pattern). The
patterns have the advantage that they do not
rely on individual transition measures and that
they have been identied by development economics and trade theory as of relevance. In as
far as China is at the very early stages of these
accepted development patterns, with good reasons for these patterns to continue to unfold,
future economic growth is likely.
The comparison countries are Japan, Korea,
and Taiwan. These three countries are relevant
for China only if they do not dier systematically with respect to the relationship between
key variables. While these four countries dier,
as do the domestic and international circumstances under which they experienced a particular stage of development, that does not
necessarily invalidate the assumption of a similar relationship between key variables.
The choice of countries to compare China to is
a subjective choice, motivated only by the desire
to make comparisons across a relatively homogeneous group of countries. China may share
some economic growth patterns with Japan,
Korea, and Taiwan due to cultural similarities,
geographic location, similar economic development strategies, or, in the case of Japan, relatively large size of the domestic economy.
Limiting the analysis to Japan, Korea, and Taiwan allows the careful compilation of the necessary data from each individual countrys
statistical oce, with a very few missing data
points obtained from the World Bank Develop-
1669
WORLD DEVELOPMENT
1670
50000
45000
Primary
40000
Secondary
Tertiary
35000
30000
25000
20000
15000
10000
5000
0
1978
1981
1984
1987
1990
1993
1996
1999
2002
2005
Figure 1. Sectoral labor productivity in China. Employment data are mid-year values (as average of end-year values).
All employment data (sectoral, total) during 19891990 experience a statistical break; end-year 1990 total employment
exceeds that of 1989 by 17.03%. Ocial (unadjusted) employment data are used. The secondary sector comprises
industry and construction.
China
Taiwan
Korea
-0.035
-0.030
-0.025
-0.020
-0.015
-0.010
14
Japan
13
Korea
12
11
Taiwan
10
China
9
8
7
6
5
4
3
2
1 Japan
0
-0.005 -10.000 0.005 0.010
-2
-3
-4
Figure 2. Structural change and labor productivity growth. Japan 19562005, Korea 19712005, Taiwan 19672005,
and China 19792005. (Pre-1991 employment data for China are adjusted as explained in appendix.) Two data points
are omitted in order to keep the chart compact: Korea 1972 (0.0224, 0.9448) and 1998 (0.0125, 4.6146).
than in the U.S. School leavers in rural areas almost invariably do not enter agriculture, and
this is unlikely to change. The development of
rural industries is ocially encouraged. Urban
income, including for unskilled labor, is still
0.8
0.7
Japan
Korea
Taiwan
China
1671
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1953 1957 1961 1965 1969 1973 1977 1981 1985 1989 1993 1997 2001 2005
Figure 3. Share of agricultural laborers in economy-wide employment. Employment data are average annual values (in
the case of China, mid-year values). Values for Japan only comprise agriculture (presumably farming and animal
husbandry) and forestry; they exclude shery due to the lack of time series data.
WORLD DEVELOPMENT
1672
14
13
12
11 China
10
9
8
7
6
5
4 Korea
3
2
1
0
-10.0 0.1
-2
-3
Japan
Korea
Taiwan
China
Taiwan
Japan
0.2
0.3
0.4
0.5
0.6 0.7
0.8
0.9
1.0
1.1
1.2
1.3
1.4
GDP per laborer (in USD) per U.S.GDP per laborer (in USD)
Figure 4. Catching up and labor productivity growth. Japan 19602005, Korea 19712005, Taiwan 19672005, and
China 19792005.
14
13
12
11
10
9
8
7
6
5
4
3
2
1
0.012 0.014 0.016 0.018 0.020 0.022 0.024 0.026 0.028 0.030 0.032 0.034
GDP per laborer (in USD) per U.S.GDP per laborer (in USD)
Figure 5. Catching up and labor productivity growth in China, 19792005.
1673
Japan
Korea
Taiwan
China
Taiwan
Japan
Korea
0.2
0.3
0.4
0.5
0.6
0.7
0.8
1674
WORLD DEVELOPMENT
4. GROWTH ACCOUNTING
26
1
2
dY t
dt
1 st
Y t dsdtt
Y t 1 st Y t 1 st
w
t
dP t
dt
wt
Pt
Lt
wt
Pt
Y st
ddt
dst
t
wt Lt dL
dt
dt K t dt
dt
P t Y st Lt dt P t Y st st
d
Kt
Pt
t
K t st
dt P t
:
s dt st K t
Pt Y t
Y st
Pt
Using hats to denote growth rates and abbreviating the shares of labor, depreciation, and
operating surplus in GDP less net taxes on
production by ast ; d st , and est lead to the income
growth identity
dst
dt
Yb st Yb t
1 st
bt
s w
ast b
L t d st ^
dt
at
Pt
!
bt
K
s
s
s
:
et ^st d t et
Pt
Like the behavioral growth accounting equation derived from an assumed aggregate production function, the income growth identity
represents one form of GDP decomposition,
only now in the form of a denitional identity.
Compared to the traditional growth accounting
equation, the income version newly introduces
the growth rates of the real wage rate, of the
depreciation rate, of the surplus rate (or rental
rate of capital), and of the net tax rate on production; in exchange, there is no room for a
residual. 31
In the period 19782002, nominal GDP
deated by the implicit deator as rst
published grew by 827%, and nominal GDP
less net taxes on production, deated by the implicit deator as rst published, grew by
814%. 32 The dierence of only thirteen percentage points implies that the changes in the
net rate of taxes on production are negligibly
small. The income decomposition, using the
average income shares of 19782002, that is,
weighting factor growth with the mean of the
rst and last periods income share values, is
as follows:
bt
s
s w
b
Y t at
ast b
L t d st ^
dt
Pt
!
bt
K
s
s
s
et ^st d t et
Pt
814% 0:5981 495% 0:5981 60%
0:1370 16% 0:2649 50%
0:4019 1185%
296% 36% 2%13%
476% 797%
In the period 19782002, growth in constantprice capital accounted for approximately
60% of real GDP growth (476/797) and growth
in the real wage for 37%; growth in the quantity
of labor contributed just below 5%, with the
remainder due to very small positive and negative contributions of changes in the depreciation rate and in the surplus rate.
The income growth identity only holds with
perfect accuracy in the instantaneous case when
growth rates are innitesimally small and the
factor shares at a given point of time therefore
are exactly applicable. Measuring growth rates
over 24 years introduces a discrepancy of 814%
versus 797% due to the need to apply average
factor shares. The discrepancy is small, that
is, using mid-period weights has little impact
on the outcome.
1675
!
bt
K
Pt
!
bt
K
:
Pt
The future growth rates of the quantity of labor are readily available if one is willing to
make a few, plausible assumptions. Those aged
15 in 2015 (the minimum working age) were already born at the time of the year 2000 population census, that is, forecasts through 2015 are
highly reliable, and in as far as trends in reproduction rates have been stable in recent years,
forecasts through later years are likely to only
gradually lose reliability. 33
What remains is to obtain future growth
rates of the real wage and of real capital. Ideally, these growth rates are related as much as
possible to variables whose future values are already known todaythe quantity of labor, the
quality of labor, and age measures. When the
unit of analysis is the individual, these are natural explanatory variables for the real wage. 34
1676
WORLD DEVELOPMENT
0.65
0.60
0.55
0.50
0.45
0.40
Labor remuneration
Depreciation
Operating surplus
0.35
0.30
0.25
0.20
0.15
0.10
1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002
Figure 7. Income factor shares in GDP less net taxes on production, 19782003. The values of all income components
are by necessity derived from provincial-level data published in GDP 195295, GDP 19962002, and for the years since
the mid-1990s in the corresponding years Statistical Yearbook. 2004 values are not available. For each year, the share of
each income component in GDP less net taxes on production is obtained as the sum across provinces of the values of a
particular income component divided by the sum across provinces of GDP less net taxes on production.
1677
Regression number
Constant
C. Average years of schooling
minus 5
All variables
in levels
All variables in
natural logarithms
A. Wage
I
B. Capital
II
A. Wage
III
B. Capital
IV
4387.96
4144.38
6.2426
0.7171
(12.84)
7.6163
0.2829
(3.18)
3126.44
(23.20)
1.3593
(5.02)
39066.18
(4.07)
152080.0
(18.80)
2
1, 5
OK
None
13973.42
(2.55)
238.63
(2.39)
2
1, 1
OK
H:D(10)
A: 97/1/2
F: 34/66/0
G: 1/78/21
A: 66/12/22
F: 16/81/3
G:19/60/21
B: 87/11/2
D: 19/77/5
H: 83/1/17
B: 25/70/5
D: 23/72/5
H:33/45/22
1
1, 1
C 1%, E 10%
A: E(0.1)
C: E(5)
A: 82/3/15
C: 0/94/6
E: 0/61/38
A: 65/4/31
C: 11/89/10
E:13/60/27
23
22
24
H. Number of laborers
I. Trend
Number of observations
0.0946
(15.23)
2
5, 5
OK
B:I(1)
B: 87/13
C: 8/92
B: 29/71
C: 6/94
22
Wage per laborer: labor compensation in the national income accounts divided by the number of laborers including military
personnel (in yuan per laborer); deated by the GDP deator (in 2000 prices).
Capital: measure of capitals contribution to production, in billion yuan, deated by the GDP deator (in 2000 prices).
Number of laborers is in billion.
Values in parentheses underneath coecient estimates are asymptotic t-values (with standard errors corrected for degrees of freedom). All coecients are signicant at the 1% level, except the ones of H and I in the second data column (signicant at the 5% level).
Trend: the cointegrating vector always includes a constant (as does the VAR); depending on the results of the trace test and the
maximum eigenvalue test, a trend is included or not included.
The logic behind the use of the variable average years of schooling minus ve (squared in one case) is that a primary school
education is standard, and wage dierentiation then depends only on the number of years of schooling in excess of primary school
education. While primary school lasts for six years, the 1978 average years of schooling is just below 6 years; in order to have a
positive variable throughout (necessary to take logarithms), 5 years is used as a benchmark.
Trace, maximum eigenvalue test: signicance level of the existence of one eigenvector.
Normality in VEC equations: Jarque-Bera values are used (based on Cholesky (Lutkepohl) orthogonalization). When normality is
rejected for one of the VEC equations at the 10% level, the signicance level is stated together with a letter which identied the
variable that, in rst dierences, is the dependent variable of the particular equation.
Granger causality Y:X (Z%) means that Y is Granger-caused by X (at the 0.1%, 1%, 5%, or 10% signicance level).
Variance decomposition: a Cholesky ordering of variables as listed above in the particular table column was used, and the forecast
error variance decomposed annually over periods of 3 and 10 years. Values are percentages, ordered as the variables appear in the
column. For example, if the three variables involved are A, C, and D, and the variance concerned is that of A, then A/C/D % of the
variance of A is explained by variation in A, C, and D. Percentages may not add up to 100 due to rounding.
1678
WORLD DEVELOPMENT
Table 3. Growth forecasts (average annual growth rates in %)
Individual variables
200005
200510
201015
201520
202025
200025
Cumulative
200025
GDP
Wage
Capital
Wage
Capital
Labor
Eqn. (5)
Labor share
assumed xed
(I)
(II)
(III)
(IV)
(V)
I, II, V
III, IV, V
I, V
III, V
4.15
8.18
7.76
6.36
4.63
6.20
6.98
6.37
5.63
5.40
4.14
5.70
5.18
6.29
7.81
11.79
14.22
9.01
10.91
10.85
10.75
10.71
10.54
10.75
0.83
0.71
0.20
0.80
0.62
0.06
5.80
7.84
7.06
5.50
4.06
6.05
8.09
8.83
9.42
10.85
12.02
9.83
5.01
8.95
7.98
5.51
3.98
6.27
6.05
7.05
8.03
10.90
13.51
9.07
350.04
299.69
763.49
1184.75
1.58
333.77
942.78
357.15
777.14
The numbers IIV refer to the previous table. All calculations are done annually before averaging to ve-year or 25year values. The antilogarithm was applied to the annual forecast values of the wage rate (III) and capital (IV) before
calculating growth rates.
GDP calculated via the denition of the labor share, with an assumed xed labor share in GDP including net taxes
on production (of 0.5138, year 2000), uses one of the two wage series plus the labor series to calculate annual GDP
and then the ve-year average GDP growth rate. (Alternatively, one could simply add up the wage and labor growth
rates as provided in the table.) GDP calculated using Eqn. (5), in addition, uses one of the two capital series.
When calculating the year 2001 GDP growth rate using Eqn. (5), the actual year 2000 labor share in GDP less net
taxes on production (0.5985) is used to weight the 2001 (over 2000) growth rates of the wage, of labor, and of capital
to obtain 2001 GDP. The 2001 labor share is then calculated from the 2001 labor remuneration and GDP values, and
used in the calculation of the 2002 GDP growth rate, etc. Using variables I+II, the minimum labor share in the
period 200025 is 0.5741, the maximum labor share is 0.6336; using variables III+IV, the two values are 0.4615 and
0.5985.
For comparison, the average annual GDP growth rate in 19782003 was 9.37% (using the ocial real GDP growth
rates) or 9.72% using the latest revised nominal GDP data combined with the rst published, implicit GDP deator.
Further into the future, the predictions diverge signicantly, with one GDP growth series
falling o, and the other rising. Due to the recent explosion in education, the out-of-sample
forecasts use education values that are far from
sample average values. The comparison of the
two GDP growth series shows well the degree
of uncertainty involved in predicting further
into the future than 2015.
An alternative approach to predicting GDP
growth is to assume a constant labor share.
This appears plausible given the worldwide
experience that labor shares tend to be constant
over time. 39 In China, the share of labor in
GDP (now not less net taxes on production)
has been near-constant throughout the reform
period, with a mean of 0.5210 and a standard
deviation of 0.0111 in 19782002, that is, the
coecient of variation is only 2%. Given the
denition of the labor share as labor remuneration divided by GDP, that is, as
wt
Lt
at P t ;
Yt
1679
5. HUMAN CAPITAL
Change in human capital formation has been
rapid in reform period China. A comparison
with the United States puts Chinas human capital measures into perspective, while the development of education in China over time
reveals the scale of changes underway.
Table 4 compares the educational level of the
Chinese versus US population as reported in
the population censuses of 1990 and 2000 in
both countries. 40 As a percentage of the total
population, a far higher proportion of the US
population has achieved a secondary or tertiary
level of education. For example, in 1990, 75.2%
of the US population had completed high
school (or above), but only 9.52% of the Chinese population had; 20.3% of the US population had completed a bachelors degree (or
above), but only 0.56% of the Chinese population had. But because the US population in
1990 was only 28% of the size of the Chinese
population (in 2000, 24%), the dierence
shrinks by a factor of approximately four once
total population numbers are considered.
During 19902000, China narrowed the relative gap. While in 1990 the number of US citizens with high school education was 1.65 times
the number in China, by 2000 this ratio had
fallen to 1.17 (last column shown in Table 4).
For the BA degree, the ratio fell from 7.59 to
4.33. At the level above the bachelors degree,
the ratio in 2000 was still 21.42 in favor of
the United States (with no such gure available
for 1990).
The population total of China masks drastic
dierences in education levels of dierent age
cohorts. Figure 8 shows that in 2000 more than
half the population aged 65 or above had a level of education below primary school, while of
the age cohort 2029 years, only 2.06% had; for
some form of secondary education, the two percentages are 10.55% (age 65 or above) versus
69.44% (aged cohort 2029), and for tertiary level education 1.50% versus 7.53%. The age cohorts in between reveal a smooth transition
from the less educated older generation to the
ever better educated younger age cohorts. 41
The Chinese age cohort of 2029 years in
2000, at the primary and secondary school levels, comes close to the United States population
average, but in tertiary level education still falls
far short.
The picture changes drastically if one considers current education data. In terms of currentyear new enrollment at the BA level (including
1680
WORLD DEVELOPMENT
Table 4. Educational level (or higher) of population aged 25 or older (Census data)
in %
Absolute number
(mio.)
80.4
51.8
16.5
4.3
1.4
0.1
24.4
8.9
75.2
45.2
20.3
7.2
9.52
1.51
0.56
n.a.
146.498
94.386
44.460
16.217
182.212
119.469
71.809
32.250
11.439
158.868
Ratio United
States/China
China
125.137
1.17
32.457
10.258
0.757
760.480
2.91
4.33
21.42
0.24
72.421
1.65
11.517
4.249
n.a.
571.589
6.24
7.59
n.a.
0.28
For the classication, see appendix. Advanced degree holders in the United States refer to the three degrees above
B.A., in China to Masters or Doctorate degree holders. No data on advanced degree holders in China for 1990 are
available. If Bachelors degree and above in the case of the United States were compared to College-level associate
degree and above in China, the ratio of the United States to China in 1990 is 2.80 (32.250/11.517), and in 2000 it is
1.37 (44.460/32.457).
United States data report degree obtained, while Chinese data include those who have already obtained the
particular degree as well as those who currently study for the particular degree. Since the US age classication is
limited to those aged 25 and older (and the Chinese classication has been reduced here to the US classication), the
number of Chinese who are currently studying toward a particular degree should be small, except possibly in the case
of advanced degrees.
Sources: United States: Educational Attainment 2000. China: Census 2000, pp. 593602; Census 1990, Vol. 1, p. 380,
Vol. 2, pp. 112f. and 132f.
80
60
40
20
%
No school/ lit. class
Primary school
20-29
Secondary school
40-49
Tertiary level
30-39
50-60
60-64
65+
Figure 8. Share of educational level in individual age groups, China, 2000 (in %). No school / lit. class refers to no
schooling or only basic literacy class (below primary school level). Source: Census 2000 (nationality), pp. 125127.
1681
5.5
0.55
5.0
0.50
4.5
0.45
0.40
3.5
0.35
3.0
0.30
2.5
0.25
2.0
0.20
1.5
0.15
1.0
0.10
0.5
0.05
0.0
1978
1981
1984
1987
1990
1993
1996
1999
2002
million
million
4.0
0.00
2005
1682
WORLD DEVELOPMENT
3.2
2.8
2.4
China: MA + PhD
U.S.: MA + PhD
million
2.0
1.6
1.2
0.8
0.4
0.0
1978
1981
1984
1987
1990
1993
1996
1999
2002
2005
1683
Demographics also matters in terms of market sizeand 80% of Chinas population lives
in the densely populated Eastern part of the
country. Size of the domestic market should allow unprecedented variety and economies of
scale. Economies of scale lead to low cost. A
large number of suppliers suggest a high degree
of competition and pressure on prices; eventually, decreasing prot margins create incentives
to innovate in order to either lower costs further or be able to oer a more dierentiated
product with higher prot margins. The price
wars of the late 1990s are evidence of the degree
of price pressures Chinese markets are capable
of. 49
One immediate implication of Chinas economic growth on peoples livelihood is that
one-fth of the world population will continue to experience signicant improvements
in their living standard. A share of Chinas
population that exceeds the size of the US
population will enjoy living standards close
to the level of developed countries in the near
future. In the Chinese hinterland, the rise out
of poverty continues. Consumption waves
similar to those previously experienced in
the West are unfolding in China, albeit with
some variations, such as the early introduction of the cell phone. In production, with
rst labor shortages in Guangdong province
in 2005, wages of unskilled labor have begun
to rise, and labor standards are increasingly
being enforced. From the creation of capital
markets to the establishment of a countrywide social security system, what has taken
the West centuries has been condensed into
decades in China.
Internationally, the growth of China implies
that the center of world economic activity shifts
to an Asia that besides China also includes India, South East Asia, Japan, and Korea.
Shocks to the United States and European
economies, or stagnation of these economies,
will have less and less impact on Asia that continues to integrate. In 2005, Chinas ratio of
exports of goods and services to GDP was
37%, compared to 10% for the United States. 50
Of Chinas exports of goods in 2005, 21% went
to the United States and 22% went to Europe,
but 48% went to Asia; 67% of Chinas imports
came from Asia. 51
The West plays an important entrepreneurial
role in China. In 2004, the most recent year for
which the data are available, 19% of gross
output value of industrial enterprises with annual sales revenue in excess of 5 m yuan
1684
WORLD DEVELOPMENT
RMB (USD 0.6 m) was produced by foreignfunded enterprises, a gure which excludes an
additional 11% in enterprises funded by Hong
Kong, Macao, and Taiwanese entrepreneurs, 52
that is, non-domestic entities control one-third
of Chinese industry. To the extent that Western
pension funds invest in these rms, often multinational corporations, Western citizens directly
benet from Chinas continued economic
growth.
The foreign-funded industrial enterprises
alone, with their approximately one-fth share
of industrial output in China, accounted for
58% of Chinas exports in 2005 (and 59% of
Chinas imports). 53 This suggests that Chinas
exports to the West largely reect the West
importing from itself, where the Wests
nationals, or rms incorporated in the West,
use Chinese labor in their production
processes.
Chinas economic growth aects relative
prices around the world, thereby promoting
an increasing division of labor and specialization worldwide. This implies structural
changes in countries around the world. The
restructuring extends beyond the traditional
realms of manufacturing, to services; for example, the newly rich in Chinas coastal provinces
are potential tourists for the West, while backroom banking services may gradually shift
from the West to China. Chinas current
strength in unskilled labor and mass production implies a wage constraint for unskilled
laborers employed in Western industries producing tradables.
Chinas economic growth is in some ways
a rerun of the industrial revolution in the
UK two hundred years earlier, but at a
much swifter pace. It is accompanied by
simultaneous changes in the patterns of production, ownership, and sources of income in
economies around the world. The concept of
a national economy, and, following from
that, of a new economic superpower, appears far-fetched; China is integrated into
the world economy to an extent that is similar to that of a European country within
Europe. What emerges is one world economy
organized along dimensions of comparative
advantage, where the construct of a nation
is, in the economic realm, reduced to issues
of enterprise registration, regulation, and taxation.
One may wonder if China will not, one
day, be overwhelmed by obstacles to further
1685
NOTES
1. Broomeld (2003) provided an overview over the
China threat literature. Lubman (2004) documented
the negative views of China in the US congressional
debates.
2. The quotes are from The China Price in Business
Week, 6 December 2004 (issue 3911), pp. 102.
3. Statistical Yearbook (2006, pp. 734, 1025).
4. See The Economist, November 15, 2003, p. 67,
quoting the Bank Credit Analyst, a Canadian research
rm.
5. Perkins and Rawski (2006) perhaps came closest.
They rely on the growth accounting equation derived
from the Cobb-Douglas production function, using
income shares as weights. Their labor calculations are
similar to those made here; their future capital series are
assumed series.
6. This paper is also not concerned about how meaningful the concept of GDP data is for measuring wellbeing or standard of living. One may further
question if value added created in China contributes as
much to well-being as does the value added created in
the United States, or vice versa. Since the use of GDP as
a measure of economic size, or, in per capita terms, as a
proxy for standard of living, dominates in the literature
as well as in popular discussions, this paper focuses on
GDP.
7. See Statistical Yearbook (2006, pp. 57, 734); <http://
www.bea.gov/bea/newsrel/gdpnewsrelease.htm> (accessed November 30, 2006).
8. The GDP data are those published by National
Bureau of Statistics in China (for China) and those
published by the Bureau of Economic Analysis in the
United States (for the United States). The PWT also
include GDP values; in the case of China these incorporate manipulations of the ocial values in accordance
with adjustments to ocial values suggested by Maddison. Holz (2006a) argued that Maddisons adjustments
are not justied (and questions Angus Maddisons
(2006) reply in Holz (2006c)).
9. The PWT 6.1 for the year 2000 (the most recent year
covered in the PWT 6.1) report a factor of 4.32; in the
PWT 6.2, for 2000, it is 4.22. This suggests a rather small
margin of error, except that the PWT 6.2 values seem to
incorporate very little new information.
10. One could allow the future real growth rate to dier
from the past one. For example, assuming a 7% real
growth rate for China (instead of the 197805 average of
9.64%) and a 3% real growth rate for the United States
(instead of 2.97%), Chinas GDP exceeds that of the
United States by 2015; with a 3% RMB appreciation this
changes to 2011. Per capita, reducing the growth rate by
three percentage points results in the years 2059 countrywide, 2033 for the coastal provinces, and 2027 for the
ve fastest-growing provinces (or, with RMB appreciation, 2035, 2023, and 2020). In contrast, increasing
Chinas aggregate real growth rate to 12% (or increasing
the per capita growth rate by two percentage points)
yields 2010 (2029, 2020, 2018) or, with RMB appreciation, 2009 (2023, 2017, 2015). The impact of such
changes to the growth rate is small in the short, but
substantial in the long run.
11. One particular alternative would be the projections
based on nominal values combined with assumptions
about the price adjustment factor and/or the exchange
rate.
12. An appendix with further details on extrapolations
is available, as all appendices are, at <http://ihome.ust.hk/~socholz/Growth/Appendices-Holz-ChinaGrowth-20253July07.pdf>.
13. In a dierent approach, Woo (1999) examined the
results of sector-level reforms in China, thus breaking
down the case of China into several sectoral observations (which, however, are not independent). Rawski
(1999) traced the interaction between dierent reforms;
in total, these are taken to add up to the reform success.
The World Bank (1997, p. 19) listed a variety of
strengths and advantages considered to promote
economic growth, and a variety of risks and challenges considered to hamper economic growth.
14. The rst two are standard fare in development
economics; see, for example, Cypher and Dietz (1997,
pp. 27, 267, 403). The third is a basic theorem in trade
theory as a corollary of the Heckscher-Ohlin theorem;
see, for example, Salvatore (1998), chapter 5.
15. The individual data sources, with explanations, are
documented in an appendix, which also oers some
further thoughts on the inclusion of other countries (see
note 12 for the web location of the appendix).
16. Total and sectoral employment data for China
experience a statistical break during 198990; total
1686
WORLD DEVELOPMENT
1687
1688
WORLD DEVELOPMENT
1689
REFERENCES
Abramovitz, M. (1986). Catching Up, Forging Ahead,
and Falling Behind. Journal of Economic History,
46(2), 385406.
Barro, R. J., & Sala-i-Martin, X. (1995). Economic
growth. New York, NY: McGraw-Hill.
Broomeld, E. V. (2003). Perceptions of Danger: The
China Threat Theory. Journal of Contemporary
China, 12(35), 265284.
Brown, L. R. (1995). Who will feed China? Wake-up call
for a small planet. New York: W.W. Norton & Co..
Carter, C. A., & Rozelle, S. (2001). Will China Become a
Major Force in World Food Markets?. Review of
Agricultural Economics, 23(2), 319331.
Census 1990 (1993). Zhongguo 1990 nian quanguo renkou
pucha ziliao (National population census material
1990). Compiled by the State Council population
census oce and the National Bureau of Statistics
population division. Beijing: Zhongguo tongji chubanshe.
Census 2000 (2002). Zhongguo 2000 nian quanguo renkou
pucha ziliao (National population census material
2000). Compiled by the State Council population
census oce and the National Bureau of Statistics
population, society, and science division. Beijing:
Zhongguo tongji chubanshe.
Census 2000 Major Figures (2001) 2000 nian di wu ci
quanguo renkou pucha zhuyao shuju (Major gures
from the 2000 population census of China). Beijing:
Zhongguo tongji chubanshe.
1690
WORLD DEVELOPMENT
1691