Lecture 2. Understanding China’s Growth.
Introduction.
• Despite China’s remarkable growth, there is not much literature trying to explain its very high growth rates.
• There is much speculation about the causes of Chinese economic growth and about predictions for the future.
• Many of the data, especially from the past, are problematic and subject to criticism.
• We are at a stage where we get mostly only macroeconomic estimates.
• Yt=AtF(Kt,Lt)
• The big questions turn around • Kt (capital stock). Problem of valuation of
socialist capital.• At (total factor productivity).
• Lt also raises problems of its own (problems associated to one child policy)
Hu-Khan (1997)
• IMF team report.• Kt grew by 7% between 1979-94 but
capital output ratio has remained more or less constant (no capital deepening).
• Estimate TFP growth of 3.9% in that period (compared to 2% for Asian tigers). Explain by success of reforms.
Chow-Li (2002)
• Yt = AetKtLt
• Ln Yt = a0 + a1lnKt + a2 lnLt +a3t .6136 .4118 .0263 (.0772) (.1996) (.0025)• Trend only significant after 1978. • Suggests constant returns to scale (a1 +a2 not
different from 1). • Ln (Yt/Lt) = a0 + a1ln (Kt/Lt) +a3t .5577 .03028 (.0468) (.0040)
Chow-Li (2002)
• Suggests a 3% annual growth in TFP from 1978 to 1998.
• Growth of 9.1% can be decomposed in– 5% due to capital (.5577 X 9.1)– 3% due to TFP growth– 1.1% due to labor [(1-.5577)x2.7]
• Growth due mostly to capital accumulation and TFP growth (1/3 of growth explained by TFP growth).
• Projections of 8% annual growth until 2010. Were not proven wrong.
Young (2003)
Contrarian view critical of official statistics (1978-1998).1. Statistical Bureau reluctant to revise downward
estimates. 2. Problems with reporting nominal and constant output
by enterprises. Leads to underestimation of deflator by 1.7 to 2.5 %.
3. Proposes reconstruction of many data but with not too many differences.
1. Output per capita growth at 6.1% (instead of 7.8), 2. output per worker at 5.2% (instead of 6.9) due to rising
participation rates, 3. non agricultural output per worker at 3.6% (instead of 6.1%)4. non agricultural output per effective worker to 2.6% (instead of
5%), 5. non agricultural TFP per worker at 1.4 (instead of 3).
Perkins-Rawski (2007)
• Try to predict growth over the next two decades.
• Growth accounting:• TFP growth = gY – aKgK -aLgL
• gY , gK , gL growth rates of GDP, capital and labor.
• Shares of capital and labor aK,aL in total factor payments (in total national income).
Perkins-Rawski (2007)
• Use official statistics. Accept possible deflator problems but also underestimation of service sector. Moreover, quality adjustments not applied in existing price indices.
• Use 2000 prices to avoid inflated industrial prices from socialism.
• Capital stock estimated using annual investment outlays and depreciation rate of 9.6% (perpetual inventory method)
• Labor measured as population 16-65 and decomposing in 5 education categories (no diploma, primary, junior high, high school, college).
Perkins-Rawski (2007)
• Inconsistencies between census and data on graduation. Build data using age-specific mortality rates and graduation rates.
• L = L1 + L2 + L3 + L4 + L5
• H = w1L1 + w2L2 + w3L3 + w4L4 + w5L5
• wi are education-linked wage differentials.
• H/L measures educational attainment.
Total Number L RatioYear Age 16-65 No diploma Primary Junior High Senior High Tertiary H/L
1952 313,346,294 73.9 19.6 4.8 1.4 0.4 0.901965 372,783,230 56.8 30.2 9.2 3.0 0.8 0.981978 506,667,822 39.6 33.4 20.6 5.7 0.7 1.071990 722,042,872 27.6 33.5 25.8 12.1 1.0 1.182000 808,414,668 18.5 34.4 30.6 14.7 1.8 1.262005 849,636,604 15.1 32.8 33.6 15.7 2.7 1.31
Mincer weights from 2000 wage data 0.818 1 1.346 2.010 3.277
Source: Appendix. Working age population includes ages 16-65.
Percentage of Workers in Each Education Category
Table 1Non-student Working Age Population, 1952-2005
Headcount, Educational Attainment, Education-enhanced Total
Perkins-Rawski (2007)
Period GDP AverageFixed Raw Education TFP Fixed Education
Capital Labor Enhanced Growth Capital Enhanced TFPK L H K Labor H
1952-2005 7.0 7.7 1.9 2.6 2.1 47.7 21.4 30.91952-1978 4.4 5.8 1.9 2.5 0.5 56.3 32.7 11.01952-1957 6.5 1.9 1.2 1.7 4.7 12.7 14.9 72.41957-1978 3.9 6.7 2.0 2.7 -0.5 73.7 39.7 -13.41957-1965 2.4 5.2 1.5 2.1 -1.0 93.1 49.5 -42.61965-1978 4.9 7.7 2.4 3.1 -0.2 67.7 36.7 -4.41978-2005 9.5 9.6 1.9 2.7 3.8 43.7 16.2 40.11978-1985 9.7 9.2 3.4 4.5 3.2 40.6 26.6 32.81985-1990 7.7 6.9 2.5 2.9 3.1 38.8 21.5 39.71990-1995 11.7 9.1 1.4 1.9 6.7 33.3 9.5 57.31995-2000 8.6 10.5 0.9 1.6 3.2 52.7 10.5 36.82000-2005 9.5 12.6 1.0 1.8 3.1 57.1 10.6 32.3
Source: Appendix.
Note: Calculations assume that 1952 capital stock is two times that year's GDP; assumed depreciation rate is 9.6 percent.
Labor Input Growth Attributable to
Table 2Average Annual Growth of GDP, Fixed Capital, Labor, and TFP,
With Contributions to TFP Growth, 1952-2005 (percent)
Average Growth of Inputs Percentage Shares of GDP
Perkins-Rawski (2007)
Figure 1Sources of Annual GDP Growth, 1952-2005
Annual Change in Fixed Capital, Education-enhanced Labor, and TFP
-25
-20
-15
-10
-5
0
5
10
15
20
1952
/53
An
nua
l cha
nge
(per
cen
t)
Fixed Capital
Education-enhanced Labor ForceTotal Factor Productivity
Figure 2China: Fixed Capital Formation Proportions, 1952-2005
0
10
20
30
40
50
60
1952
1954
1956
1958
1960
1962
1964
1966
1968
1970
1972
1974
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
Per
cent 2000 prices
Current prices
• Japan’s postwar growth based on TFP growth. Half of 9% growth rate between 1950 and 1970 explained by TFP. Later slowdown associated to fall in TFP growth.
• Question: Will the growth continue?
Population and Dependency Ratio, 1990-2025
Year Population Percentage Share
(Millions) Working Age 16-65 Dependents
Include Exclude Include Exclude Working Age Students Working Age Students
1990 1,130 65.4 64.1 34.6 35.9 2000 1,263 66.7 64.5 33.3 35.5 2005 1,308 68.8 65.3 31.2 34.7 2015 1,390 70.6 66.4 29.4 33.6 2025 1,457 67.9 63.4 32.1 36.6 2035 1,489 64.1 35.9 2045 1,497 61.4 38.6
Source: Data for 1990 from official census results for that year. Dependents are defined as persons aged under 16 and over 65. High school and college students: appendix to this paper
Low dependency ratio imply robust household savings
Labor Force and Graduates
(Millions) Educational Composition of Non-student
Workforce
Year Working Age Annual
Graduates (Percent)
Non-students Senior Tertiary No Primary Junior Senior Tertiary Ratio L High Diploma High High H/L
2005 849.6 8.3 3.1 15.1 32.8 33.6 15.7 2.7 1.31 2010 885.3 9.9 5.0 11.6 29.3 36.6 17.8 4.7 1.39 2015 913.4 12.8 5.0 9.5 25.3 37.8 20.2 7.2 1.48 2020 903.6 11.5 6.9 7.4 21.9 38.2 21.7 10.8 1.58 2025 885.6 12.8 6.6 5.7 19.4 37.0 23.2 14.7 1.69
Projections of Working Age Non-Student Population Ages 16-65: Size and Composition, 2005-2025
Decline of L after 2015 but increasing human capital
Productivity Consequences of Input Projections and 6 or 9 Percent Growth to 2025
Period Annual Average Growth of Inputs % Annual Growth of GDP (%)
GDP Labor Input Average Attributable to TFP Share
Growth Fixed Raw Education TFP Fixed Education TFP of GDP
% Capital Labor Enhanced Growth Capital Enhanced Growth Growth
Labor Labor percent
K L H % K H
Version 1 2005-2015 9.0 9.8 0.7 2.0 3.6 4.2 1.1 3.6 40.4 2015-2025 9.0 8.2 -0.3 1.0 4.9 3.5 0.6 4.9 54.4 Version 2 2005-2015 9.0 9.0 0.7 2.0 4.0 3.9 1.1 4.0 44.5 2015-2025 9.0 6.6 -0.3 1.0 5.6 2.8 0.6 5.6 62.4 Version 3 2005-2015 6.0 8.1 0.7 2.0 1.4 3.5 1.1 1.4 23.1 2015-2025 6.0 5.6 -0.3 1.0 3.0 2.4 0.6 3.0 50.7 Version 4 2005-2015 6.0 7.3 0.7 2.0 1.7 3.1 1.1 1.7 28.7 2015-2025 6.0 4.0 -0.3 1.0 3.7 1.7 0.6 3.7 61.7
Memo item: average TFP growth 1978-2005: 3.8 percent (Table 2)
Versions 1 and 2 assume 9% GDP growth; Versions 3 and 4 assume 6% growth
Ratio of fixed capital formation to GDP declines linearly from the 2005 figure of 42.3 % to a terminal 2005 level of 35 percent (Versions 1 and 3) or 25 percent (Versions 2 and 4).
Country Year start Period Growth Period Growth
PPP GDP
Exch Rate
of slowing before Rate after Rate
per cap at GDP at
slowdown Before slowdown After slowdown slowdown (2005 prices)
Japan 1971 1948-1970 10.72 1971-78 2.605 13800 16500
Taiwan 1990 1980-89 8.56 1990-99 6.304 13370 10880
Korea 1992 1982-91 9.05 1992-2005 5.28 13370 8810
It would take China 10-15 years at least to reach PPP GDP per capita of $13,000 ($5,300 in 2007).
How productive is investment in China?
• China has had a very high investment rate (around 40%). This suggests that the marginal product of investment might be low (too much investment and allocative inefficiencies). However, a very high marginal return to capital might explain a high investment rate.
• Bai et al. (2006) have addressed this question.
How productive is investment in China?
• Use revised output estimates after 2004 census (covers 1978-2004).
• Assume depreciation rate of 8 % for structure and of 24 % for machinery. Use estimate of initial stock for 1952 plus gross fixed capital formation.
• Estimate share of capital income as residual of labor income.
How productive is investment in China?
• Estimates relatively high even though falling.
• Why have returns remained high?– Capital-output ratio has not increased much– Capital share has increased steadily together
with increase in the investment rate.
How much misallocation of capital in China?
• There is a strong perception that despite strong growth, there may be a severe misallocation of capital in China.
• General idea: if distortions favor certain firms, they can survive with lower TFP than firms that do not benefit from those distortions. Hsieh and Klenow (2008).
Open questions
• China’s growth will affect the global economy and there may be feedback effects.
• Energy. • Environment.