May 21, 2013 Global Economics Paper No: 218 Research Report China: More efficient cities key to a brighter growth path A more efficient use of surplus labor will more than offset the impact on growth from a modest decline of the working age population. Further urbanization will be part of future growth. Urbanization process needs to improve, in order to deliver more efficient and balanced growth through: 1) productivity gains from city agglomeration; 2) a decline in capital intensity as land prices better reflect market forces; and 3) a moderation in the household savings rate, helped by reforms to social safety nets. Using US cities as a benchmark, greater efficiency in Chinese cities has the potential to add 1ppt to real economic growth over the next decade. Investment-to-GDP share will decline and consumption- to-GDP share improve meaningfully with reforms. Urban housing construction will benefit from future income growth. The recent policy directions are promising and suggest that economic imbalances should gradually be corrected. A more concentrated urban growth model would be positive for public transportation, utilities, water management and diversified consumption, and the burden on public finances would be lower. Investors should consider this report as only a single factor in making their investment decision. For Reg AC certification and other important disclosures, see the Disclosure Appendix, or go to www.gs.com/research/hedge.html. Li Cui +852-2978-0784 [email protected]Goldman Sachs (Asia) L.L.C. The Goldman Sachs Group, Inc. Goldman Sachs Global Economics, Commodities and Strategy Research
29
Embed
Global Economics Paper No: 218 - ghsl.cn · Global Economics Paper No: 218 Research Report China: More efficient cities key to a brighter growth path A more efficient use of surplus
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
May 21, 2013
Global Economics Paper No: 218
Research Report
China: More efficient cities key to a brighter growth path
A more efficient use of surplus labor will more than offset the impact on growth from a modest
decline of the working age population. Further urbanization will be part of future growth.
Urbanization process needs to improve, in order to deliver more efficient and balanced growth
through: 1) productivity gains from city agglomeration; 2) a decline in capital intensity as land prices
better reflect market forces; and 3) a moderation in the household savings rate, helped by reforms
to social safety nets.
Using US cities as a benchmark, greater efficiency in Chinese cities has the potential to add 1ppt to
real economic growth over the next decade. Investment-to-GDP share will decline and consumption-
to-GDP share improve meaningfully with reforms. Urban housing construction will benefit from
future income growth.
The recent policy directions are promising and suggest that economic imbalances should gradually
be corrected. A more concentrated urban growth model would be positive for public transportation,
utilities, water management and diversified consumption, and the burden on public finances would
be lower.
Investors should consider this report as only a single factor in making their investment decision. For Reg AC certification and other important disclosures, see the Disclosure Appendix, or go to www.gs.com/research/hedge.html.
The Goldman Sachs Group, Inc. Goldman Sachs Global Economics, Commodities and Strategy Research
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 2
Table of Contents
Summary 3
1. Urbanization has a long way to go 4
2. The multiple links between urbanization and Chinese growth 8
A. Urbanization and high investment share: a more conservative use of land key for bringing down
the investment/GDP share 8
B. Urbanization and productivity gains: Exploring the agglomeration effects of cities 11
C. Urbanization and household savings: Migrant savings key to the rising household savings rate 14
D. Urbanization and housing: Higher income, more housing consumption 16
3. Next stage of urbanization to be driven by key policy changes: Land reforms, city agglomeration, savings 18
A. A more concentrated approach is preferable 18
B. Land policies and better infrastructure are key ingredients for more compact cities 20
C. Comprehensive social security reforms supported by fiscal reforms needed to stem the rise in
household savings 22
4. Growth and sectoral impact under different urbanization scenarios 23
5. Conclusion 26
References 27
Disclosure Appendix 28
Many thanks to Jan Hatzius, Dominic Wilson, Andrew Tilton, Goohoon Kwon and MK Tang for their insightful comments.
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 3
Summary The most pressing issues for China’s macro landscape are linked to its medium-term
growth model and whether the country will become stuck in the ‘middle-income trap’.
Investors are concerned about the country’s high investment share, elevated household
savings and constrained consumption growth, as well as the apparent decline in the
economic returns on investment.
In this paper, we argue that many of these imbalances have their roots in the urbanization
process pursued so far. Carefully considered and measured policies can go a long way
towards addressing such imbalances. The government has announced its intention to
tackle several key reforms under a “new urbanization strategy”. We examine some of the
key dimensions of urbanization and assess quantitatively how the growth trajectory is
likely to be affected under different urbanization patterns. Our analysis suggests there is
significant potential to achieve more efficient and balanced growth through a well
calibrated strategy. Land policies, city size, reforming social welfare and housing policies
are key areas that we will be monitoring closely. Although policy details have not yet been
announced, the strong focus on a number of these areas is promising.
1. Land reforms are important: A more economic use of land would lower the intensity of
investment in structures, i.e., construction-related activities such as infrastructure and
buildings for industrial needs, and as a result lower the investment/GDP share. This would
help reverse the recent trend whereby urban land expansion has outpaced population
growth, and in turn would increase the relative share of equipment in total investment. As
the latter is more directly related to future growth, this would help to boost investment
efficiency for the economy as a whole. We estimate that a more modest pace of expansion
in city construction—by 1.5 percentage points (ppt) less each year, or about one-third
slower than the current pace—can reduce the investment/GDP share by 3-6ppt in the
coming decade.
2. The ‘agglomeration effect’ in cities can be a key source of future productivity gains but this
requires careful urban planning. Although urbanization has so far succeeded in shifting
surplus resources into non-agricultural use, efficiency gains through resource
agglomeration have been absent. Such economies of agglomeration have the benefit of
greater economies of scale, a deeper job market, efficiency in land use and a lighter burden
on public finances. We estimate that, over the coming decade, if Chinese cities achieve the
economies of agglomeration present in US cities, there will be an additional gain of 1ppt in
the growth of total factor productivity (TFP) by the end of the period. Land policies and
related urban management are critical to realize such gains. In particular, adequate
infrastructure is needed to support more compact cities in order for the positive impact of
city agglomeration to offset the negative side-effects of big cities. The recent focus on
public transportation in cities is positive in this respect.
3. Hukou and social welfare reforms are critical: Increased savings by migrant workers have
been behind the decline in the consumption share. The urban savings rate is higher and is
increasing faster than in rural areas, driven in part by the high propensity of migrant
workers to save for precautionary reasons. In the absence of any reforms, further urban
migration can increase the average household savings rate each year by 0.1 ppt per year.
We expect hukou and social security reforms, accompanied by fiscal reforms, will be
important to realize the full potential of urbanization for domestic demand. We estimate
that consumption could be boosted by as much as 2-3 ppt of GDP as a result of these
reforms.
4. Housing purchase has been financed mostly by household savings rather than leverage.
Similar to consumption, it will continue to benefit from further income growth. Urban
housing supply has been mostly driven by population and income growth, and residential
housing investment currently accounts for less than 7% of GDP and 10% of total
investment. We forecast that housing investment as a share of GDP will recover somewhat
in the next few years, even under the ongoing policy controls.
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 4
1. Urbanization has a long way to go
China’s growth in the past 30 years has gone hand-in-hand with the rise of cities, driven by
rapid industrialization and integration with the world economy, and an urban population
that has more than tripled from 1980. In the past 10 years alone, the number of urban
residents has grown by 20 million each year, accounting for about 30% of the increase in
the world’s urban population. China now has the largest urban population in the world,
even though the urbanization rate is still just above 50% (Exhibits 1-2).
Growth and urbanization have also been closely associated geographically, suggesting that
urbanization is the outcome of growth and job creation. The regions along the coastal
areas have urbanized at a fast pace, with around 70% of their population already urban
residents. In contrast, in western regions such as Sichuan, the share of city dwellers
remains at around 40% of the total population, although these provinces have been
urbanizing rapidly along with brisk income growth in recent years (Exhibit 3).
Cities are likely to remain the engine of growth in China, as has been the case for economic
development in the past (Spencer, 2009). The new Chinese government has identified
urbanization as one of the four pillars of its economic mandate in the coming decade.1
Indeed, rural areas still retain about 30% of the total labor force, with an average labor
productivity that is one-sixth that in urban areas. The United Nations projects that China’s
urban population will grow by another 185 million in the coming decade, bringing the
urbanization rate to 60% by 2020, and sustained non-agricultural job creation will remain
critical.
The large amount of surplus labor is a key counterforce to the overall demographic trend,
and argues against the widely held, yet misplaced, view that China’s growth is destined to
slow as a result of a shrinking labor force. In our view, a more efficient use of the surplus
labor in rural areas and of under-employed workers in urban areas would more than offset
the impact on growth of the slight decline in the overall working age population (about
0.1% each year) in the coming decade (see Asia Economic Analyst: China: Shifting to a
lower gear, September 20 2012.) How best to bring these underutilized resources into more
productive use will therefore be critical and, as in the past, sustained non-agricultural job
creation will be the driver.
Meanwhile, a new strategy is necessary, given the imbalances in the past and more limited
scope for easy productivity gains in the future:
First, urbanization has so far been associated with some of the pressing
imbalances, including the depletion of arable land, environmental damage, air
pollution, income disparity and rising urban congestion, to name a few. This has
raised concerns that the urbanization path pursued so far is not sustainable.
Second, there is a need to search for new sources of efficiency gains as the scope
of easy productivity gains through the ‘catch-up’ effect naturally narrows. In
particular, trade growth is likely to be more modest given China’s sizeable market
share, and the scope of productivity gains through trade is also arguably smaller
than before, as a result of the rising sophistication of Chinese products.
1 Premier Li Ke Qiang laid out several key areas of consideration in promoting future urbanization, in particular the location of urban population growth, land utilization to protect arable land in the process of urbanization, hukou reforms so that rural migrant workers can become more integrated within the cities where they work, energy and environmental implications, etc.
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 5
1. China has reached its current income level at a lower urbanization rate than other
countries
Note: Exhibits 1-4 are based on the United Nations Population data. There may be national differences in characteristics that distinguish urban from rural areas.
Source: CEIC, United Nations Population Statistics, World Bank.
2. China has contributed significantly to the world’s
urban growth
3. China's urbanization process closely follows its growth
path
Note: Exhibits 1-4 are based on the United Nations Population data. There may be national differences in characteristics that distinguish urban from rural area.
Source: CEIC, United Nations Statistics.
Source: CEIC, China National Bureau of Statistics.
0
10
20
30
40
50
60
70
80
90
100
5.0 6.0 7.0 8.0 9.0 10.0 11.0
GDP per capita (PPP, constant 2005 int'l dollar, logarithm
term)
Urbanization rate (%)
*1950 to 2010 (China as of 2012):
UK China Korea
India Spain Italy
US Brazil Japan
0
1000
2000
3000
4000
5000
6000
7000
0
10
20
30
40
50
60
70
80
1950 1970 1990 2010 2030 2050
China's contribution to world urbanization rate (left scale)
Goldman Sachs Global Economics, Commodities and Strategy Research 7
Our baseline model assumes a continuation of the past trend of land expansion and
city creation, and slow reform of social welfare. There is no extra efficiency gain
through city agglomeration. Non-housing investment/GDP would remain elevated
(although moderating slightly for cyclical reasons), while the consumption/GDP share
climbs only slowly.
The concentrated and faster reform scenario assumes that land use becomes more
conservative, that there will be greater efficiency gains through ‘agglomeration’ and
that social welfare reforms have a faster impact on household savings than in the past.
The new scenario would improve growth in three ways: 1) investment/GDP share
declines relative to the baseline; 2) TFP is boosted through better urban agglomeration;
and 3) consumption/GDP share improves faster. This scenario would be the most
positive path for urbanization, in our view, as it improves growth efficiency, reduces
investment intensity and burdens on land and other resources, improves consumption
and alleviates pressure on public finances. Indeed, the positive impact is likely to be
understated as larger cities tend to provide better job prospects, which we have
assumed to be the same across all scenarios.
The dispersed and slow reform scenario assumes that land use will become even more
scattered than in the past, no efficiency gains through city agglomeration and that
social security reforms remain gradual. TFP gains and consumption/GDP shares would
therefore follow the baseline, but the non-housing investment/GDP would rise even
faster. This scenario is the least favorable, in our view, as it implies a worsening of
investment vs. consumption imbalances, an even more significant burden on land and
resources than in the past, overheating pressures in the near term owing to the
significant increase in investment demand, and a deterioration of economic returns on
investment.
The policies discussed in recent months have increased the possibility that the second
scenario may be pursued, although we look for more clarity on the policy front in the
coming months.
Section 5 looks at the implications of our analysis for the economy and associated markets.
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 8
2. The multiple links between urbanization and Chinese growth
In this section, we describe several channels through which the shape of urbanization
affects growth balances and efficiency. In particular, we link the urbanization strategy and
policies to non-housing investment, growth efficiency, consumption and savings, and
housing investment. We look at each of these channels separately and estimate the
quantitative linkages between the macro variables and urbanization-related development.
Conscious that urbanization is endogenous to growth, in our estimation we also account
statistically for the potential reverse causality where possible (see, for example, Box 2).
A. Urbanization and high investment share: a more conservative
use of land key for bringing down the investment/GDP share
It is well known that China’s investment as a share of GDP is high by international
standards. Moreover, investment as a share of GDP has risen in recent years, resulting in a
deterioration of economic returns on investment. As an example, the Incremental Capital
Output Ratio (or ICOR), calculated as the incremental change in growth relative to
investment, has deteriorated. While the low cost of funding is one of the key factors
explaining this high investment share [see Aziz and Cui (2007)], it does not explain the
recent increase or the deterioration in investment returns, in our view.
To account for the rising investment share, we believe a more detailed breakdown of
China’s investment composition is necessary. Compared with other countries, it is striking
that investment in structures—related to construction activities such as infrastructure and
buildings for industrial use—is dominant in China. Although the share of investment in
structures within the total declined during 2003-2008, the process has stalled since then
and the share of construction-related investment remains well above 40% (Exhibit 5), and is
higher than in other major economies (Exhibit 6). At the current level, China’s investment
structure remains similar to that of UK in the 1870s at a much lower income level.2
Investment in structures tends to be associated with a lower ICOR, while investment in
equipment is more directly related to technological innovation and productivity gains, and
thus a higher ICOR. De Long and Summers (1991) have demonstrated the link between
investment in equipment and economic growth for a broad group of countries. While many
countries have had high shares of investment in structures at some point during their
development, the share generally trends down over the longer term (Ball, et al, 1996);
indeed, a reduced intensity of investment in structures was related to productivity
enhancement in countries that were industrialized early, such as the UK. Therefore, we
expect a decline in the share of investment in structures to see investment efficiency rising.
2 Over the longer term, the relative share of investment in construction falls relative to equipment investment increases across OECD countries. This is associated with relative price changes: as land prices rise and equipment prices fall, businesses find it more profitable to invest in equipment while preserving the cost of land use. See Ball, et al (2001) for further discussion.
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 9
5. China's investment dominated by structures
6. China’s investment in structures is higher than in other
countries
Source: CEIC, China National Bureau of Statistics. Note: For China, investment in structures (non-housing) is derived by deducting residential
housing investment from total construction. As the residential housing investment includes the purchase of equipment, the investment in structures may be underestimated.
Source: Haver Analytics, OECD, China National Bureau of Statistics.
The strength of investment in structures in China coincides with the remarkable expansion
of the city landscape, which has outpaced urban population growth. We estimate that the
built-up urban area has expanded by near 40% over the last decade, while the population
has increased by less than 30% (Exhibit 7). Urban land expansion has been brisk in the
less-developed Western region, as well as in the export-oriented Eastern regions, while
larger cities such as Beijing saw the slowest urban land expansion (Exhibit 8). Urban
population growth, by contrast, was the fastest in Beijing and Shanghai, and was much
slower in the West and Eastern regions. Thus, much of the city expansion cannot be
explained by population growth or urbanization needs. Indeed, apart from a few big cities,
the increase of urban density has been unremarkable, and the overall density is modest by
international standards.
In our view, both the strength of investment in structures and the rapid expansion of land
use for cities in China reflect current land policies, which allow the conversion of
agricultural land into urban construction land at below-market prices (IMF 2011, Li 2008,
and Deng 2004). Rural land is collectively owned, and local governments have the
discretion to expropriate land at a low price and sell the industrial land use rights for a
period of up to 50 years (IMF, 2011). Systematic data on land expropriation prices is not
available but examples of farmland expropriated at low prices are abundant. One of the key
reforms announced is to grant explicit land ownership and certificates for farmers’
homesteads and contracted land. This year, for the first time, farmers' property rights were
included in the government report3, an encouraging sign that agricultural land
expropriation system is likely to be reformed. A more market-oriented land pricing scheme
is an important first, but arguably not sufficient, step towards more efficient use of land
and less urban sprawl.
Empirically, we estimate the links between the expansion of the urban area and the non-
housing investment share of GDP, based on a panel dataset that covers 31 provinces
during 2003-2011 (Box 1). We find that the expansion of urban land is important in
explaining the rising investment share of the economy. Our estimates suggest that slowing
growth in the amount of land used for construction by 1.5ppt each year would help to
reduce the investment/GDP ratio by 3-6ppt in 10 years, with the impact mostly back-loaded
given the time needed for the effect to accumulate.
3 In China’s No 1 government report in 2013, released in February, the central government requires a complete registration of rural collectively-owned land and that farmers’ contract rights are assigned to the collectively-owned land within five years.
0
10
20
30
40
50
60
0
10
20
30
40
50
60
1995 1998 2001 2004 2007 2010
China: Investment by types(share in total)
Residential housing
Equipment
Others
Non-residential structure
42
36
36
33
23
20
China
Japan
UK
Korea
Germany
US
Structure (non‐housing) Investment as a Share in Total (%)
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 10
7. Urban land expansion outpacing population growth 8. Urban land expansion by region
Source: China National Bureau of Statistics, CEIC. Source: CEIC, China National Bureau of Statistics.
Box 1: The impact of urban expansion on capital investment
Chinese cities have been expanding at a faster pace than
the urban population, suggesting that the urbanization
process has led to a high intensity of land use. Not only
has this raised concerns about food supply (although
arable land is now subject to strict policy controls), it also
has an impact on the investment-heavy model of growth.
According to existing laws, individual farmers do not have
ownership rights and agricultural land is collectively
owned by rural cooperatives. Local authorities have the
ability to convert agricultural land into land for urban
development at a below-market price. This has made one
of the key inputs of production—land—too cheap,
spurring investment growth, in particular in construction
activities. This can be seen from the high correlation
between the growth in the investment/GDP ratio and that
of land expansion. To quantify the impact, we link the
land size with non-housing investment growth, using a
panel dataset that covers 31 provinces during 2003-2011.
Exhibit B1: Land expansion positively relates to non-
housing investment
Source: GS Global ECS Research.
The model follows a stylized framework that Li, Yin and
Li (2008) use to analyze the impact of land size on capital
intensity.
ln (I/Y)t = a + b x ln (I/Y)t-1 +c x ln (area) t-1 +e
Where I is non-residential GFCF, Y is output and area is
the size of the land under construction. We use the built-
up area in cities. The first lag of the investment/GDP
ratio is used as an instrumental variable for all other
relevant explanatory variables. The model is then
estimated with generalized method of moments (GMM)
to overcome the statistical issues related to the lagged
dependent variables. Our analysis suggests that land
area has a strong and positive link with non-residential
investment, and that the elasticity of investment share
with respect to the land area has increased in recent
years. The estimated results based on the last five years
of the sample period are presented below (Exhibit B1).
The results suggest that slower land expansion has a
meaningful impact on investment intensity. Land
expansion that is one-third slower than the pace (i.e.
urban land size increases by 3.5% each year compared
to 5% before) could lower the investment-to-GDP ratio
by up to 3-6ppt compared with the baseline scenario
over ten years.
We also use the provincial level of land under
construction as an alternative measure of urban land
areas, and the results are very similar.
100
105
110
115
120
125
130
135
140
100
105
110
115
120
125
130
135
140
145
2003 2004 2005 2006 2007 2008 2009 2010
Construction land and population growth:
Area of built district
Urban population
Index: 2003=100 Index: 2003=100
100
110
120
130
140
150
160
170
180
100
110
120
130
140
150
160
170
180
2003 2004 2005 2006 2007 2008 2009 2010 2011
Urban built areas(2003=100)
Beijing
West and central
NE
East
Index Index
Dependent variable: ln (I/Y)
Variable Coefficient Std. Error t‐Statistic Prob.
Ln (I/Y,‐1) 0.21 0.00 61.88 0.00
Ln (AREA,‐1) 0.59 0.03 22.73 0.00
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 11
B. Urbanization and productivity gains: Exploring the
agglomeration effects of cities
Cities play two key functions in a country’s development. The first function is the
shift of surpluses into more productive use. As experienced by many initially agriculture-
based countries, a shift in surplus resources from the agricultural sector into
manufacturing and services with higher productivity helps to enhance factor productivity.
This has been the experience of many developing countries, as highlighted by Lewis (1954).
The second and more lasting advantage of cities is the ‘agglomeration’ effect. As cities
become more densely populated, city dwellers tend to become more productive through
the technology spillover and knowledge sharing permitted by close proximity (Henderson,
2004), but this requires appropriate infrastructure to offset the negative side-effects of big
cities such as traffic congestion and pollution. These agglomeration effects can occur
through businesses or consumers. Business agglomeration increases the productivity of
firms and workers, while consumer agglomeration creates demand for more diversified
consumer needs. One example of such agglomeration is the positive income effect from
larger populations evidenced in US cities (Carlino, 2011).
In China, city growth has so far been mostly about shifting resources away from the agricultural sector; agglomeration effects have been absent.
China’s urbanization has generally succeeded in shifting surplus resources towards more
productive use—including labor and land. Urban job growth has averaged 3.3% per year
over the last decade, creating 146 million urban jobs since the late 1999s, or a 40% increase
in urban employment since then. Since the productivity of rural workers remains about
one-sixth that of urban workers, shifting rural workers into urban areas should continue to
be a key driver of future productivity growth. This is in contrast to a popular view that
China’s labor surplus is now coming to an end, and that China has come close to a “Lewis
turning point” whereby the quantity of labor has become a key constraint.
Such resource shifts have continued to bring productivity gains in recent years, but the pace
has slowed. To see this trend, we rely on a growth model to decompose the contributions of
various production inputs (not including land for simplicity). There has been a marked
increase in the capital contribution to growth, and a reduction in TFP in recent years (Exhibit
9). We then further decomposed the TFP gains into those related to resource shifts
(estimated from urban job growth and urban and rural productivity differentials) and the
residual which represents the efficiency gains from other sources. It is striking that in recent
years the internal shift in resources accounted for most of the TFP gains.
9. TFP gains from resource shifting sectors have continued but other TFP gains diminished
China's growth accounting: diminishing TFP growth post GFC:
TFP from resource shifting into urban areas
TFP not related to resource shift
Contribution of capital
Percentage points Percentage points
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 12
10. Agglomeration economies across US cities…
11. …and the lack thereof in Chinese cities
Source: US Census Bureau.
Source: CEIC.
The reduced scope for ‘easy’ productivity gains suggests the need to search for new areas
of efficiency improvements. One such area is the positive efficiency gains through city
agglomeration, which has been largely absent in the urbanization process so far. A simple
(though imperfect) test of such an agglomeration effect is the comparison of per capita
income and population size. If the agglomeration effect is strong, through the positive
externality, larger cities should be associated with higher income. Across US cities, for
instance, a doubling of city population is associated with a 3%-4% increase in the average
wage (Exhibit 10).4
In China, however, the association of population size and per capita income is weak. In
contrast to US cities, where income and population size tend to be positively linked, in
China such an effect is muted (Exhibit 11). In fact, formal tests suggest that when variables
corresponding to regions, administrative levels of cities and initial conditions are controlled
for, income and population tend to be negatively correlated across cities (Box 2). If Chinese
cities exhibit the same ‘agglomeration economies’ as US cities by the end of the coming
decade, the real rate of growth could be boosted by an additional 1ppt relative to the
baseline by the end of the decade. We take this as the potential magnitude of the efficiency
boost through ‘economies of agglomeration’ in section 5.
4 Note that the slope of Exhibit 10 should not be taken as an agglomeration impact itself. If agglomeration effects are present, larger cities should provide higher income. However, there is reverse causality as well– a city with higher income attracts people and leads to the expansion of the city. Once such sources of potential overestimation are controlled for, a 1% larger population is associated with a 0.038% increase in per capita income in US cities (see Carlino [2011]).
2.5
3.0
3.5
4.0
4.5
10 11 12 13 14 15 16 17 18
log (population)
log
(city
inco
me)
US cities: Income and population size
3
4
4
5
5
6
10 11 12 13 14 15 16 17 18
log
(per
cap
ita in
com
e)
log (population)
China: population and income across cities
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 13
Box 2: Test for the ‘agglomeration effects’ in Chinese cities
One of the direct tests of the ‘agglomeration effect’ in cities is the positive association between average wages in a city
and the city’s population size. The basic argument is that if agglomeration effects are important, wages and demand for
labor should both reflect the advantages of ‘agglomeration economies’ (Carlino, 2011). The simple test of this
relationship for Chinese cities does not support a positive association of income and population size, in contrast to the
clear and strong positive association in US metropolitan areas. To be sure, such a positive association of city size and
average wage represents the upper bound of the agglomeration effects in cities. If agglomeration effects are important,
workers in large cities should be more productive than those in small cities. At the same time, more productive workers
may be attracted to large cities, exaggerating the agglomeration effects. Once these sources of potential overestimation
of agglomeration effects are accounted for, a doubling of the metropolitan population is associated with 3.8% increase
in average income in the US.
We test the links between the per capita income and population in cities by controlling for the regional income (by
including the provincial average income), the initial condition (the city’s per capita income in 2002), and administrative
importance (a dummy variable corresponding to provincial capitals). In all specifications, the level of city income is
actually negatively related to the population size. This supports the view that agglomeration effects in cities are muted
at best, or even negative (Exhibit B2).
For our simulation exercise, we assume that the impact of city agglomeration on growth efficiency in China would catch
up with that of US cities in 10 years. More specifically, a 1% larger population is associated with a negative 0.1% decline
in average income, when other factors related to the average regional income, the administrative level in the cities and
initial income are controlled for. This compares to a positive income gain of 0.038% from a 1% larger population in US
cities. We use the difference of 0.13ppt in elasticities as the ‘agglomeration effect gap’ between Chinese cities and US
cities. Assuming the same urban population growth as in the baseline, but Chinese cities ‘catch up’ to their US
counterparts in this agglomeration effect over the next decade, the average income growth would be 13% higher than
the baseline. We assume this is achieved in 10 years under the more concentrated model, and this would correspond to
about a 1ppt gain in TFP growth (cumulative) relative to the baseline by the end of the decade, or an additional TFP
growth of 0.1ppt each year.
Exhibit B2: Arable land per capita in China is low
Source: China National Bureau of Statistics, CEIC.
We also augment the analysis by estimating the links between urban population and urban income growth using
provincial level data, where data on urban population is directly available (by contrast, the city population may include
those who live in the surrounding rural areas). We run a regression of the average income growth on urban population
and population density, while controlling for other variables such as initial income, external trade and investment
growth. We find that while the cross-region growth differentials are strongly related to the initial income (via internal
catch-up, and thus a negative association) and investment growth, the urban density tends to contribute little or even
negatively. This is consistent with our findings at the city level.
Goldman Sachs Global Economics, Commodities and Strategy Research 14
Agglomeration can have positive effects on growth through knowledge spill-over, a deeper
labor market and the diversification of activities. The absence of an agglomeration effect in
urban growth suggests that the potential efficiency gains from bigger cities have been
overshadowed by the negative side-effects, such as transportation costs, congestion and
pollution. We see two likely reasons for this:
The rapid expansion of cities mentioned above: With the exception of Tier 1
cities (which are among the largest cities in the world, see Exhibit 12), for most of
the other large cities, such as Chongqing and Wuhan, the average density remains
quite low (and even declining), given the fast extension of the city limits.
Inadequate infrastructure, in particular public transportation. There is an
undersupply of critical city infrastructure, such as subways and water systems.
Greater agglomeration effects are a critical aspect of why a more concentrated model of
urban development is desirable in the case of China. We discuss this in more detail in
Section 3.
12. Chinese Tier 1 cities are among the largest in the world (2011)
Source: Domographia World Urban Areas.
C. Urbanization and household savings: Migrant savings key to the
rising household savings rate
We find the lack of access by migrant workers to essential city public services to be a main
reason for the rising household savings rate, which has been one of the key challenges to
rebalancing the Chinese economy and promoting domestic consumption. We estimate the
average urban savings rate to be 10ppt higher than that in rural areas (Exhibit 13). We
established a link between the increase in household savings and the lack of access to
adequate public services by migrant workers through an empirical test of the panel dataset
of Chinese provinces (see Emerging Markets Macro Daily: The hukou system: Holding back
China's rebalancing, May 1 2013).
0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
45,000
50,000
0
5
10
15
20
25
30
35
40
Density (RHS)
Population
persons per square kilometerMillion
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 15
13. Urban household savings rate higher than the rural rate
Source: CEIC, China National Bureau of Statistics.
A simple statistic simulation suggests that in the absence of any change, savings could
continue to drift up along with city migration. For instance, for each 14 million rural
residents moving to urban areas—corresponding to about 1% of the total population
shifting from rural to urban works each year—the average savings rate could increase by
0.1ppt per year.5 Our previous work suggested that removing hukou-related differences in
social services would boost household consumption/GDP share by up to 6ppt over 10 years.
Assuming a slower-paced reform than in our previous work, we estimate that 2-3ppt can be
added to the consumption/GDP share over 10 years.
From a macroeconomic perspective, curbing the rising household savings rate is essential.
All else equal, a rising savings rate is an underlying deflationary source given the demand
‘leakage’. According to some economists, increasing urbanization helps to support
aggregate demand, as it is boosted by spending by migrant workers. However, we think
this argument is dubious, since migrant workers will add to both supply and demand.
Unless migrant workers consume a larger share than they produce, the average savings
rate of the urban household is likely to rise rather than fall.
Therefore, hukou reforms and associated social welfare reforms provided to migrant
workers will be an essential part of reforms for sustained growth.
5 In our forecast, we expect that each year, about 1% of the total population will shift from the rural area into cities. Assuming each one migrant increases her savings rate from the average rural level (25%) to the average urban level (37%). This has the net effect of increasing the average savings rate of households by about 0.1 ppt every year, all else equal.
Goldman Sachs Global Economics, Commodities and Strategy Research 16
D. Urbanization and housing: Higher income, more housing
consumption
We also link urbanization to China’s housing supply. Our estimation based on a cross-
province panel data suggests that housing investment is driven mainly by income growth
and demographics, which implies that housing supply is broadly consistent with the rising
consumption needs of households (Box 3). These factors suggest that housing investment
as a share of GDP should hold up, even with the ongoing policy control:
Housing is part of consumption. As of 2012, the urban housing stock stood at
around 20 million square meters, or roughly 27 square meters per person. We
estimate that about 45% of urban housing was built after 2000 (housing reform
started in 1998, since when more modern residential housing has been
constructed). The newer houses are larger in size, even though household sizes
have been declining (Exhibits 14 and 15). From a growth perspective, while not
contributing to the country’s productive capacity itself, housing investment has
met the needs of urban demand.
Compared with other countries, this consumption is not leveraged…
Households have generally not borrowed to finance their homes. An estimated
20% of households borrow from banks to finance their houses. A lack of
alternative investment of household wealth is another factor driving housing
demand.
…And it will continue to benefit from urban income and population growth. Housing investment is now about 7% of national GDP in real terms. We
estimate that housing investment is likely to stay strong before its share in GDP
declines after 2015. However, the current policy control is likely to keep housing
investment more subdued, in our view.
14. Housing stock driven by rising consumption
15. Newer flats are larger
Source: CEIC, China Population Census.
Source: CEIC, China Population Census.
22
23
24
25
26
27
28
29
0
5000
10000
15000
20000
25000
1995 1997 1999 2001 2003 2005 2007 2009 2011
Urban housing stock and size per person
Urban housing stock (mln sqrm)
urbanhouse/person (sqrm, RHS)
0
10
20
30
40
50
0
25
50
75
100
built in 1970s built in 1980s built in 1990s built in 2000s
City houses by construction time
avg size per flat
share in total housing stock(RHS)
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 17
Box 3: What is driving housing investment in China?
China’s housing investment has increased substantially in
recent years, from less than 3% of total GDP in 1996 to
nearly 8% of GDP in 2012, before cooling down somewhat
following government controls introduced last year. The
strong investment growth was partly due to the low
starting point and housing reforms. New and more
modern commercial housing in cities started to replace
the dormitories prevalent until the late 1990s when China
started its housing reform, and households have been
upgrading their housing consumption.
As of 2010, about 40% of the total housing stock had been
built after 2000. The newer houses are larger, even though
household sizes have been shrinking, implying an
upgrade in terms of living space.
Across regions, the housing investment/GDP ratio has
generally climbed over the past decade, except for a
decline in Beijing and Shanghai during 2003 and 2009,
before a strong rebound occurred there too (Exhibit B3).
In general, housing investment as a share of GDP tends to
be higher in areas where population growth is faster but
the correlation is not one-on-one (Exhibit B4). This is
because income growth is just as important a driver of
housing investment.
We estimate the drivers of economic fundamentals using
a panel dataset for China’s provinces in the last 6 years.
Income level and population growth are significant,
explaining the cross-regional differences in housing
investment/GDP ratios. In particular, we estimate the
following equations:
ln(HI/Y)it =a+ b x ln(HI/Y)it-1 +c x ln(GDPPC)it-1 +d x
urbanpop_git-1 +e
Exhibit B3: The share of housing investment in GDP
varies by region
Source: CEIC, China National Bureau of Statistics.
Exhibit B4: Population growth displays diverging trends
across regions
Source: CEIC, China National Bureau of Statistics.
Where HI, Y, GDPPC, urbanpop_g represent real housing
investment and GDP, per capital income and
urbanization population growth, respectively. Again, a
GMM estimation method is employed to account for the
statistical issues introduced by the lagged dependent
variables on the RHS. The estimation results are
presented in Exhibit B5. Using the estimated
coefficients, and expected urban income and population
growth, we are also able to project the housing
investment/GDP ratio in coming year. However, in the
next few years, how housing investment/GDP develops
will also depend critically on the government’s policy
aimed at curbing property price increases (see Riding
the Chinese construction waves, October 10, 2012). The
scenarios, with or without government controls, are
depicted in Exhibit 26.
Exhibit B5: Housing investment explained by income
and population growth
Source: GS Global ECS Research.
2
4
6
8
10
12
14
16
2
4
6
8
10
12
14
16
1999 2001 2003 2005 2007 2009 2011
East ex BJ SH
Central
West
BJ & SH
China regional residential investment as share of GDP:
% share of GDP % share of GDP
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
4.5
5.0
Capital area East Central West
Urban pop growth (avg %)2005-2008
2009-2011 (exclude2010)
yoy% yoy%
Variable Coefficient Std. Error t‐Statistic Prob.
LN(HI/Y,‐1) 0.88 0.01 60.90 0.00
LN(GDPPC,‐1) ‐0.09 0.03 ‐3.28 0.00
Urbanpop_g(‐1) 0.005 0.000 11.90 0.00
Dependent variable: ln (HI/Y)
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 18
3. Next stage of urbanization to be driven by key policy changes:
Land reforms, city agglomeration, savings
A. A more concentrated approach is preferable
The debate is ongoing as to whether future urbanization should take the shape of city
clusters around large regional centers, mid-sized cities or a large number of scattered
townships. One popular view is that farmers should be ‘urbanized locally’, which means
absorbing the surplus labor into non-agricultural jobs, without further concentration of the
population. In our view, however, a dispersed model implies duplicate investment in
infrastructure and inefficient use of land and other resources. We see a number of reasons
why a more concentrated model is a preferred approach.
1. More diversified domestic demand, benefiting job growth
As a diversified and dynamic services sector will be the main driver of job growth in
coming years, the deeper labor market in larger cities or city clusters are better placed to
create jobs for both skilled and unskilled workers.
Surplus labor is still abundant. We estimate that rural workers still account for about 30%
of the total working age population, and labor productivity in rural areas remains one-sixth
that in the cities. Therefore, shifting the surplus rural labor into non-agricultural jobs
continues to be an important source of productivity gains. The higher rural birth rate (the
average birth rate in rural areas relative to cities is about 1.2:1, and this is expected to rise
given the demographics) also suggests that the natural growth of the rural population
could potentially outpace that of the urban population in coming years.
Meanwhile, the ability of the manufacturing sector to absorb surplus labor has become
more constrained than in the past, in light of the reduced role of the tradable sector in the
economy. Future surplus labor will therefore increasingly be absorbed by the services
sector. The incremental growth of manufacturing jobs has been strongly linked to export
performance in the past decade. Based on the historical relationship, and assuming exports
grow at 5%-10% a year (given the large size of China’s exports and the high market
penetration in some areas), manufacturing jobs are likely to rise by 0.5%-1% a year. This
means about 1 million new jobs can be created in the manufacturing sector (compared
with 5-7 million a year before the global financial crisis [GFC]), and the rest will need to be
created in the construction sector and other services. Thus, we expect the migrant
population to average 14 million and workers to average 9 million a year, compared with
21 million and 15 million in the past five years (Exhibit 16).
The continued need for cities to absorb surplus labor from rural areas, mostly at the lower
end of the job ladder, while creating more skilled jobs for college graduates can be better
met by a more diversified job market in clusters of larger cities.
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 19
16. We expect the urban population and employment growth to moderate in coming years
Source: CEIC, GS Global ECS Research estimates.
2. Helping to improve investment efficiency
Strong investment growth in the past few years and the slowdown in trend growth have
meant that the capital-to-output ratio, which remained largely steady during most of the
2000s, has risen sharply in the last few years, suggesting a deterioration in investment
efficiency.6 The scope to urbanize through more intensive investment has become even
more constrained. As shown in Section 2, increased land use has been associated with the
high investment intensity of growth, suggesting a more conservative use of land would
help to boost investment efficiency.
3. Exploring the agglomeration effect of cities
As illustrated earlier, although productivity gains have continued through shifting
resources into cities, productivity gains associated with other sources have been limited in
recent years. The strong advantage of cities relative to rural areas lies in the ability of cities
to organize resources, both for producers and consumers.7 This is particularly important for
China as its vast population contrasts with its limited arable land, and the scarcity of water
and energy.8 A more concentrated model can balance the manageability of cities and the
efficiency gains, and help to preserve resources.
6 To measure the capital intensity of the economy, we make an important yet often ignored adjustment to the method of estimating capital stock. Some of the increase in the recent investment growth was ‘non-productive’ in the sense that residential investment as a share of the total has increased substantially. We estimate it to account for more than 7% of national GDP, compared with about 2% in 2004. Estimating the capital stock in the economy thus needs to exclude this part of investment, which is accrued to households rather than the corporate sector. Once this adjustment is made, the rise in capital intensity and the decline in investment efficiency (suggested by ICOR) are less pronounced. Nevertheless, it remains the case that capital intensity is now rising along with the investment share, in contrast to the early 2000s when the rise in the investment share led to strong GDP growth, leaving investment returns largely unchanged. Thus, at least from a cyclical perspective, investment efficiency has declined. In particular, the K/Y ratio, which has stayed steady at around 2.5 for many years, has risen above 3. 7 Greenstone, Hornbeck, and Moretti (2010) attempt to measure the impact of agglomeration on firms’ productivity, and find that opening a new plant boosts the existing firms’ productivity by 12% in cities where the new plant is located relative to other cities. 8 The amount of arable land per capita in China is quite low by international standards. China has 9% of the world’s total arable land and 20% of its total population. China’s per capita arable land is one of the lowest globally. Urbanization in the past 20 years has reduced the amount of arable land, for two reasons. First, arable land has been converted into construction sites. Between 1993 and 2004, total arable land declined by near 10%, largely due to this land conversion. More than half of
Goldman Sachs Global Economics, Commodities and Strategy Research 20
4. Lower burden on public finances
A further increase in debt financing by government and government-related entities is
likely to be unavoidable in future urban growth, given that:
Urban development so far has been financed both by debt borrowing and land
sales. Land reforms should in principle limit the revenue source from land sales,
with the explicit ownership rights granted to farmers. This increases the need to
rely on debt financing compared with in the past.
Although the envisaged greater private-sector participation should help to bring in
private-sector expertise and discipline, public-sector guarantees or participation is
perhaps still needed given the inherent riskiness of infrastructure projects.
This comes at a time when local government finances are already under heavy scrutiny. A
more concentrated model is likely to help reduce the gap in public finances given the lower
investment intensity.
B. Land policies and better infrastructure are key ingredients for
more compact cities
1. Reforms to ensure a more efficient use of land
The recognition of farmers’ land rights is the first step towards more discipline and greater
efficiency in land use. A scheme being tried out is direct market-based pricing and
transaction, such as the auction scheme in Chongqing, where ‘land certificates’ are backed
by underutilized rural land. In principle, this allows for a better relationship between the
price of rural land and the marginal returns from converting it into urban land, thereby
reducing excessive appropriation for city development.
In our view, such a market-based system is important, but it may not be sufficient to curb
urban sprawl on its own. In fact, the system has been designed to free up more rural land
for urban use (but in a more equitable manner), rather than to limit urban land growth itself.
Containing urban expansion will require an explicit set of policies in land management,
reuse of vacant brown-field areas, and a greater focus on public transport.
2. Recent infrastructure plans are right for more compact and efficient cities
Better managed cities can deliver more visible efficiency gains and overcome the negative
side-effects of large cities. This also implies shifting needs for infrastructure:
There is a lack of city infrastructure such as water treatment plants (Exhibit 17) and
utilities, as well as environmental protection in China. There is a critical need to
correct this.
Transportation needs are changing. The length of city subways in the newly
announced metro plans is closely related to the population size in surrounding
areas, and is in line with international experience (Exhibit 18). By contrast, roads
have already been built to meet the population distribution, even though for the
nation as a whole road density remains low (Exhibits 19-20). More compact cities
suggest a preference for public transportation, rather than roads.
the land used for construction in cities (including property, transportation and water systems) during 1998-2004 was formerly arable land. Since then, the government has implemented strict rules against the improper conversion of arable land, and the encroachment of arable land has moderated. Second, the quality of the remaining agricultural land may have declined, since land used for construction is often the best arable land close to city centers.
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 21
17. China’s wastewater treatment still has a long way to
go
18. Expected length of subways closely correlated to the
expected population size
Source: CEIC, World Bank.
Source: CEIC, China NBS, and various national agencies of statistics.
19. China’s road density is mostly compatible with
where the population is located…
20. …even though for a nation as a whole road density is
low relative to other countries
Source: CEIC, China National Bureau of Statistics.
Source: CEIC, World Bank.
30
40
50
60
70
80
90
100
30
40
50
60
70
80
90
100
Korea Singapore OECD Malaysia Thailand Philippines China BRICS Indonesia India
% of Population with Access to Improved Sanitation Facilities
2010
% of population % of population
Beijing
Tianjin
Guangzhou
Nanjing
Chengdu
Shanghai
Shenzhen
Seoul
New YorkLondon
TokyoMoscow
Chicago
Frankfurt
0
200
400
600
800
0 5000 10000 15000 20000 25000 30000 35000 40000
Population including those in the suburbs (000s)
Expected length of subway by 2017 (kms)
Henan
Shandong
Jiangsu
HubeiHunan
Anhui
Guangdong
Zhejiang
Guizhou
Jiangxi
Shanxi
0
20
40
60
80
0 50 100 150 200
population and road density by province
Road density: km per 100 square km(2011)
Population density: mn person per 100 square km
UK
Italy
Spain
India
Korea
Japan
US
China
Brazil
0
10
20
30
40
50
60
0 50 100 150 200
Population and road density by country:
Population density: thous person per 100 square km
Road density: km per 100 square km
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 22
C. Comprehensive social security reforms supported by fiscal
reforms needed to stem the rise in household savings
Since last year, policies have been put in place to enable migrant workers to obtain city
residence permits, or hukou, based on certain thresholds that vary by the size and
administrative levels of cities. Other social security services—the health care, education,
and pension systems—are also being reformed to reduce the level of precautionary
savings across urban and rural areas (Exhibit 21).
The hukou reform needs to be supplemented by fiscal reforms, in our view. Local
governments are concerned about the extra public spending needed to pay for the cost of
incorporating migrant workers into cities. Although such spending is necessary to correct
the previous misrepresentation of the true costs of city and company growth, the resulting
fiscal gap will require adjustments in the tax system, for instance by extending the
Granting hukou depending on city levelsTownships: those with stable jobs and accommodation qualify immediately Small cities(<1m population): those with stable jobs/accommodation for at least three years and at least one year of social security contribution qualify Large cities: continue to restrict population
Expanding hukou coverage Feb-12
Rural medical systemNew Rural Cooperative Medical System (NRCMS);- In 2013, China raised the annual NRCMS compensation per capita from Rnb240 to Rmb280 (Rmb30 in 2003)
More social insurance to rural area residents Effective in 2013
Health care-drug pricingEssential Drug List (EDL) expansion: According to Ministry of Health (MoH), EDL is to be expanded to include more than 500 drugs from 307
Lower the drug costs for patients Dec-2012
Health care-severe illness insuranceExpansion of the coverage of healthcare insurance system to include the treatment of critical illnesses
Enhance social safety net Aug-2012
Urban pension reform
A new urban pension program was launched in July 2011. The program aims to help create full coverage for China's basic social pension system. Insured residents over the age of 60 will be able to receive a basic monthly pension of 55 yuan
Enhance social safety net
60% urban area covered in 2011; 100% urban area covered by end of 2012
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 23
4. Growth and sectoral impact under different urbanization
scenarios
In the above discussions, we estimated the impact of the urbanization process on non-
residential investment, residential investment and household savings, based on the
historical data. We also estimated the ‘agglomeration effect’ in China compared with that
of the US. We use these frameworks to simulate the future likely path of these variables
under different scenarios.
Impact of land use on investment (non-housing) intensity of growth: We rely on
the model estimated in Box 1, which links the non-housing investment/GDP to the pace of
land use. The trend model assumes the size of the cities will expand at about 5% a year, a
similar rate to that of recent years. The concentrated model assumes cities expand at a
pace that is one-third slower, or about 3.5% a year, and the dispersed model assumes an
even more rapid expansion of urban land growth of about 6% a year. These scenarios are
mapped into the paths depicted in Exhibits 22 and 23.
Impact of agglomeration on growth efficiency or TFP: We assume that under the
more concentrated model there will be additional average TFP growth of about 0.1ppt a
year, with an accumulated TFP gain of 1ppt in 10 years. This would allow the
agglomeration effect in Chinese cities to catch up with that of the US, as described in Box 2
and Footnote 4. The effect is also assumed to be back-loaded as the impact accumulates,
implying additional efficiency gains from agglomeration of about 13% in a decade. Such
effects on growth itself are mapped into the paths in Exhibit 24, and we can then derive the
investment growth scenarios in Exhibit 25.
22. Investment efficiency is higher under a more
concentrated model…
23. …as investment/GDP share moderates
Source: CEIC, GS Global ECS Research.
Source: CEIC, GS Global ECS Research.
0
1
2
3
4
5
6
7
8
0
1
2
3
4
5
6
7
8
1996 1999 2002 2005 2008 2011 2014 2017 2020
ICOR (investment efficiency)
Baseline
Disbursed
Concentrated
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.0
0.1
0.2
0.3
0.4
0.5
0.6
1996 1999 2002 2005 2008 2011 2014 2017 2020
Non-residential investment/GDP (%)
BaselineDisbursedConcentrated
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 24
24. More concentrated model associates higher
growth…
25. …with less investment
Source: CEIC, GS Global ECS Research.
Source: CEIC, GS Global ECS Research.
Impact of urbanization and policy controls on the share of housing investment: We also forecast the share of housing investment in GDP in the coming years, based on the
historical relationships between the residential investment/GDP ratio, income and
population growth estimated in Box 3. However, as government policy controls will remain
in place and these have, in our view, drawn a wedge between the model-simulated
housing investment/GDP rate and the actual rate, the forecast housing investment/GDP
share is adjusted downwards, assuming the policy impact will be most significant in the
near term (Exhibit 26).
Impact of urbanization and reforms on the share of consumption in GDP: The
household consumption/GDP rate rose in 2012 (to 35.3% from 34.9% in 2011) after a
persistent decline over more than a decade. The improvement was both cyclical (as
investment slowed more than consumption) and structural (as government policies have
supported household income growth, offsetting some of the impact of rising savings on
the consumption/GDP share). Our baseline model assumes such an improvement will
continue in the coming year as a result of policies aimed at boosting household income
growth, but the high and rising savings rate is likely to result in a modest pace of
improvement (with the consumption/GDP share rising by about 2ppt in 10 years). However,
strong reform efforts could add as much as 2-3ppt to the consumption/GDP share over 10
years (this is a more conservative estimate than our earlier work, as the reforms are likely
to be implemented gradually) (Exhibit 27).
0
2
4
6
8
10
12
14
16
0
2
4
6
8
10
12
14
16
1996 1999 2002 2005 2008 2011 2014 2017 2020
Real growth (%)
BaselineDisbursedConcentrated
0
5
10
15
20
25
0
5
10
15
20
25
1996 1999 2002 2005 2008 2011 2014 2017 2020
Investment growth (yoy, %)
Baseline
Disbursed
Concentrated
yoy% yoy%
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 25
26. Housing investment more subdued under ongoing
policy controls
27. Reforms needed to boost consumption
Source: CEIC, GS Global ECS Research.
Source: CEIC, GS Global ECS Research.
We then combine these variables to project three urbanization paths. We assume the same
job creation and housing policies in all scenarios, and focus on the impact of land and
social security reforms (Exhibit 5).
Our baseline model assumes a continuation of the past trend of land expansion and
city creation, and a slow pace of social security reform. There is no extra efficiency
gain through city agglomeration. Non-housing investment/GDP remains elevated
(although moderating slightly for cyclical reasons), while the consumption/GDP share
climbs only slowly.
The concentrated and faster reform scenario assumes that land use becomes more
conservative, that there are greater efficiency gains through agglomeration and social
security reforms deliver a faster impact on household savings than in the past. The
new scenario improves growth in three ways: 1) the investment/GDP share declines
relative to the baseline; 2) TFP is boosted through better urban agglomeration; and 3)
the consumption/GDP share improves faster. This scenario is the most positive path
for urbanization, in our view, as it improves growth efficiency, reduces investment
intensity and the burdens on land and other resources, improves consumption and
alleviates pressure on public finances. The positive impact would likely be understated
as larger cities tend to provide better job prospects, which we have assumed to be the
same across all scenarios.
The dispersed and slow reform scenario assumes that land use will become even more
dispersed than in the past, that no efficiency gains are made through city
agglomeration and that social security reforms remain gradual. TFP gains and
consumption/GDP shares would therefore follow the baseline, but the non-housing
investment/GDP would rise even faster. This scenario is the least favored case, as it
implies a worsening of investment vs. consumption imbalances, even more significant
burdens on land and resources than in the past, overheating pressures in the near term
owing to the significant increase in investment demand and a deterioration in
Goldman Sachs Global Economics, Commodities and Strategy Research 26
5. Conclusion
The urbanization process has underpinned the transformation of the Chinese economy in
the past 30 years. The shape of future urbanization, along with the necessary adjustment of
factor costs (in particular land) will be important in affecting the efficiency of resource
allocation for the economy as a whole.
Growth and inflation
Urbanization will continue to be a source of productivity growth. Besides continued labor
migration into urban activities, the ‘agglomeration effect’ in cities is another important area
of such productivity gains. The latter is an area where easy gains can be made given the
absence of such an effect so far. Meanwhile, social security reforms can deliver meaningful
benefits which could help reduce the savings of migrant workers, removing one of the
deflationary forces for longer-term growth.
Commodities and sectional implications
A more concentrated urban growth model—say, developing city clusters—combined with
the conservative use of land would reduce the resource intensity of urban growth. This
would be negative for some commodities related to infrastructure investment, but positive
for goods and services associated with public transportation, utilities, water management
and diversified consumption. This is in contrast to a more dispersed urban growth model
(say, by developing townships or small cities) where resource intensity stays high or rises,
with near-term strong demand for commodities. Policy discussions so far, including the
emphasis on land reforms, suggest a more concentrated model is likely to be chosen.
Housing-related sectors. Housing purchases have been financed mostly by household
savings rather than leverage. Urban housing supply has been mostly driven by population
and income growth across regions. Such factors will continue to support housing demand,
notwithstanding the policy controls including the introduction of property taxes. We
therefore expect housing-related sectors to hold up in coming years.
Government finance
Although we do not expect the consolidated government balance sheet to come under
stress in the coming years, concerns related to the micro risks of local government
investment arms are likely to remain. As we have argued elsewhere, government
investment in local infrastructure is a major reason for the rising credit intensity of the
economy in recent years. This is because, compared with other types of investment,
infrastructure investment relies more on debt financing (owing to its longer construction
cycles and more delayed cash flows), and the economic benefits materialize with a time lag.
In our view, the returns on local investment would improve, and concerns for local finances
and debt burdens would ease considerably, under a more concentrated urbanization model.
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 27
References
Aziz, Jahangir and Li Cui (2007), “Explaining China’s low consumption: the neglected role of household income,” IMF
Working Paper.
Ball, Michael, Tanya Morrison and Andrew Wood (1996), “Structures Investment and Economic growth: A long-term
International Comparison,” Urban Studies.
Barras, Richard (2001), “Building investment is a diminishing source of economic growth,” Journal of Property Research.
Carlino, Gerald (2011), “Three Keys to the City: Resources, Agglomeration Economies, and Sorting,” Business Review,
Philadelphia Fed, Q3 2011.
De Long, J. Bradford and Lawrence H. Summers (1992), “Equipment investment and economic growth: How strong is
the nexus?” Brookings Papers on Economic Activity.
Deng, Frederic, Youqin Huang (2004), “Uneven land reform and urban sprawl in China: the case of Beijing,” Progress in
Planning.
Greenstone, Hornbeck, and Moretti (2010), “Identifying agglomeration spillovers: evidence from winners and losers of
large plant openings," Journal of Political Economy, 2010, 118 (3): 536-598.
Henderson , J. Vernon (2004), “Urbanization and growth,” Handbook of Economic growth.
Li, HuiZhong, Feng Yin, Jianlun Li (2008), “China’s construction land expansion and economic growth: a capital-output
ratio based analysis,” China & World Economy.
Lewis, W. Arthur (1954), "Economic Development with Unlimited Supplies of Labor," Manchester School of Economic
and Social Studies, Vol. 22, pp. 139-91.
Urbanization and growth (2009), ed. by Michael Spencer, Patricia Annez, and Robert Buckley, IBRD.
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 28
Disclosure Appendix
Reg AC
I, Li Cui, hereby certify that all of the views expressed in this report accurately reflect my personal views, which have not been influenced by
considerations of the firm's business or client relationships.
Goldman Sachs Disclosures
Global product; distributing entities
The Global Investment Research Division of Goldman Sachs produces and distributes research products for clients of Goldman Sachs on a global
basis. Analysts based in Goldman Sachs offices around the world produce equity research on industries and companies, and research on
macroeconomics, currencies, commodities and portfolio strategy. This research is disseminated in Australia by Goldman Sachs Australia Pty Ltd
(ABN 21 006 797 897); in Brazil by Goldman Sachs do Brasil Corretora de Títulos e Valores Mobiliários S.A.; in Canada by Goldman, Sachs & Co.
regarding Canadian equities and by Goldman, Sachs & Co. (all other research); in Hong Kong by Goldman Sachs (Asia) L.L.C.; in India by Goldman
Sachs (India) Securities Private Ltd.; in Japan by Goldman Sachs Japan Co., Ltd.; in the Republic of Korea by Goldman Sachs (Asia) L.L.C., Seoul
Branch; in New Zealand by Goldman Sachs New Zealand Limited; in Russia by OOO Goldman Sachs; in Singapore by Goldman Sachs (Singapore)
Pte. (Company Number: 198602165W); and in the United States of America by Goldman, Sachs & Co. Goldman Sachs International has approved this
research in connection with its distribution in the United Kingdom and European Union.
European Union: Goldman Sachs International, authorized and regulated by the Financial Services Authority, has approved this research in
connection with its distribution in the European Union and United Kingdom; Goldman Sachs AG and Goldman Sachs International
Zweigniederlassung Frankfurt, regulated by the Bundesanstalt für Finanzdienstleistungsaufsicht, may also distribute research in Germany.
General disclosures
This research is for our clients only. Other than disclosures relating to Goldman Sachs, this research is based on current public information that we
consider reliable, but we do not represent it is accurate or complete, and it should not be relied on as such. We seek to update our research as
appropriate, but various regulations may prevent us from doing so. Other than certain industry reports published on a periodic basis, the large
majority of reports are published at irregular intervals as appropriate in the analyst's judgment.
Goldman Sachs conducts a global full-service, integrated investment banking, investment management, and brokerage business. We have
investment banking and other business relationships with a substantial percentage of the companies covered by our Global Investment Research
Division. Goldman, Sachs & Co., the United States broker dealer, is a member of SIPC (http://www.sipc.org).
Our salespeople, traders, and other professionals may provide oral or written market commentary or trading strategies to our clients and our
proprietary trading desks that reflect opinions that are contrary to the opinions expressed in this research. Our asset management area, our
proprietary trading desks and investing businesses may make investment decisions that are inconsistent with the recommendations or views
expressed in this research.
The analysts named in this report may have from time to time discussed with our clients, including Goldman Sachs salespersons and traders, or may
discuss in this report, trading strategies that reference catalysts or events that may have a near-term impact on the market price of the equity
securities discussed in this report, which impact may be directionally counter to the analysts' published price target expectations for such stocks. Any
such trading strategies are distinct from and do not affect the analysts' fundamental equity rating for such stocks, which rating reflects a stock's
return potential relative to its coverage group as described herein.
We and our affiliates, officers, directors, and employees, excluding equity and credit analysts, will from time to time have long or short positions in,
act as principal in, and buy or sell, the securities or derivatives, if any, referred to in this research.
This research is not an offer to sell or the solicitation of an offer to buy any security in any jurisdiction where such an offer or solicitation would be
illegal. It does not constitute a personal recommendation or take into account the particular investment objectives, financial situations, or needs of
individual clients. Clients should consider whether any advice or recommendation in this research is suitable for their particular circumstances and, if
appropriate, seek professional advice, including tax advice. The price and value of investments referred to in this research and the income from them
may fluctuate. Past performance is not a guide to future performance, future returns are not guaranteed, and a loss of original capital may occur.
Fluctuations in exchange rates could have adverse effects on the value or price of, or income derived from, certain investments.
Certain transactions, including those involving futures, options, and other derivatives, give rise to substantial risk and are not suitable for all investors.
Investors should review current options disclosure documents which are available from Goldman Sachs sales representatives or at
http://www.theocc.com/about/publications/character-risks.jsp. Transaction costs may be significant in option strategies calling for multiple purchase
and sales of options such as spreads. Supporting documentation will be supplied upon request.
In producing research reports, members of the Global Investment Research Division of Goldman Sachs Australia may attend site visits and other
meetings hosted by the issuers the subject of its research reports. In some instances the costs of such site visits or meetings may be met in part or in
whole by the issuers concerned if Goldman Sachs Australia considers it is appropriate and reasonable in the specific circumstances relating to the
site visit or meeting.
All research reports are disseminated and available to all clients simultaneously through electronic publication to our internal client websites. Not all
research content is redistributed to our clients or available to third-party aggregators, nor is Goldman Sachs responsible for the redistribution of our
research by third party aggregators. For all research available on a particular stock, please contact your sales representative or go to
http://360.gs.com.
Disclosure information is also available at http://www.gs.com/research/hedge.html or from Research Compliance, 200 West Street, New York, NY
No part of this material may be (i) copied, photocopied or duplicated in any form by any means or (ii) redistributed without the prior written consent of The Goldman Sachs Group, Inc.
May 21, 2013 Global Economics Paper
Goldman Sachs Global Economics, Commodities and Strategy Research 29
Gao Hua Securities Disclosures
General disclosures
This research is disseminated in China by Gao Hua Securities. Gao Hua Securities is licensed to conduct Securities Investment Consultancy.
This research is for our clients only. This research is based on current public information that we consider reliable, but we do not represent it is
accurate or complete, and it should not be relied on as such. We seek to update our research as appropriate, but various regulations may prevent us
from doing so. Other than certain industry reports published on a periodic basis, the large majority of reports are published at irregular intervals as
appropriate in the analyst's judgment.
Goldman Sachs Gao Hua, an affiliate of Gao Hua Securities, conducts an investment banking business. Gao Hua Securities, Goldman Sachs Gao Hua
and their affiliates have investment banking and other business relationships with a substantial percentage of the companies referred to in this
document.
Our salespeople, traders, and other professionals may provide oral or written market commentary or trading strategies to our clients and our
proprietary trading desks that reflect opinions that are contrary to the opinions expressed in this research. Our proprietary trading desks and
investing businesses may make investment decisions that are inconsistent with the recommendations or views expressed in this research.
The analysts named in this report may have from time to time discussed with our clients, including Gao Hua Securities salespersons and traders, or
may discuss in this report, trading strategies that reference catalysts or events that may have a near-term impact on the market price of the equity
securities discussed in this report, which impact may be directionally counter to the analysts' published price target expectations for such stocks. Any
such trading strategies are distinct from and do not affect the analysts' fundamental equity rating for such stocks, which rating reflects a stock's
return potential relative to its coverage group as described herein.
Gao Hua Securities and its affiliates, officers, directors, and employees, excluding equity and credit analysts, will from time to time have long or short
positions in, act as principal in, and buy or sell, the securities or derivatives, if any, referred to in this research.
This research is not an offer to sell or the solicitation of an offer to buy any security in any jurisdiction where such an offer or solicitation would be
illegal. It does not constitute a personal recommendation or take into account the particular investment objectives, financial situations, or needs of
individual clients. Clients should consider whether any advice or recommendation in this research is suitable for their particular circumstances and, if
appropriate, seek professional advice, including tax advice. The price and value of investments referred to in this research and the income from them
may fluctuate. Past performance is not a guide to future performance, future returns are not guaranteed, and a loss of original capital may occur.
Fluctuations in exchange rates could have adverse effects on the value or price of, or income derived from, certain investments.
Certain transactions, including those involving futures, options, and other derivatives, give rise to substantial risk and are not suitable for all investors.
Investors should review current options disclosure documents which are available from Gao Hua sales representatives or at
http://www.theocc.com/about/publications/character-risks.jsp. Transaction costs may be significant in option strategies calling for multiple purchase
and sales of options such as spreads. Supporting documentation will be supplied upon request.
No part of this material may be (i) copied, photocopied or duplicated in any form by any means or (ii) redistributed without the prior written consent of Beijing Gao Hua Securities Company Limited.