2015 China’s Dangerous Property Compromise HOW THE LAND LEASE SYSTEM CURRENTLY IN USE IN CHINA MAY LEAD TO SERIOUS ECONOMIC PROBLEMS BRIAN SCHUELER
2015
China’s Dangerous Property Compromise
How the land lease system currently in use in China may lead to serious economic problems
Brian Schueler
Brian Schueler Course Paper- Econ 4240 15/4/2015
Introduction
In English, the term communism is not necessarily self-explanatory, however in China the
translation for Communism, 共产主义 (gong chan zhu yi) has a clear meaning of “commonly held
property philosophy”. For today’s Chinese Communist ruling party, this word’s lack of ambiguity has led
to awkward compromises, especially in property rights. Full ownership in land remains with the
governments, or with communes that are controlled by the government. However, land ‘purchases’ and
development are permitted through the issuance of private leases, effectively granting the leaseholder
private property rights for a certain amount of time, normally between 30-70 years depending on the
desired usage. This paper will analyze how this lease system has and will influence aspects of the
Chinese economy, focusing on overdevelopment, banking, and government finances.
Background
Before Mao Zedong, private property holdings were quite normal in China, with private holdings
of land accounting for roughly 77% of all landholdings in the city of Beijing by 1949 (Meyer 2009).
However, with the Communist revolution, private holdings of land were nationalized. Meyer estimates
that by 1978, private holdings of land in Beijing had fallen to 10%. Across China, although near total
nationalization of land had occurred by the time that Mao died, it wasn’t until 6 years after his death
that Deng Xiaopeng, the new chairman, and later the architect of many market-oriented reforms,
completed full nationalization.
Deng did away with the previous constitution and implemented a new document which revised
Mao’s place in official Chinese history, and made the country as a whole more driven by the rule of law,
at least compared to previously. The 1982 constitution made firm that any land within the cities was
owned by the state, and that all other lands and natural resources were the property of either the
communes or the state (Constitution of the People's Republic of China 1982). However, this
arrangement was not meant to last. By 1987, the government approved the first land-leasehold
Brian Schueler Course Paper- Econ 4240 25/4/2015
agreement, for a 50 year lease in Shenzhen, a city across the border from Hong Kong (Anglin, et al.
2014). It wasn’t until 1988 that the 1982 constitution was amended to allow for the lawful transfer of
land usage rights, and the amendments were specific in clarifying that only the temporary usage right
was transferred.
It is through these land leases that much of modern China has developed. However, as
explained by Gregory Stein, a professor of law at the University of Tennessee, these land use leases are
not like a traditional Western style lease. The entire fee for the land use right must be paid in full before
the land use right may be transferred, and if that land parcel is not developed within 2 years, the owner
of the land use right must either pay an extension fee, or relinquish possession of the land use right.
Importantly, the holder of the land right must also own the buildings on that land. Transfer rights are
also more restricted than in a traditional western lease (Stein 2007).
Today, a significant amount of development also occurs through another legal mechanism for
land acquisition provided in the 1982 constitution. Cities have grown not only in population, but also in
size, making necessary the transition of agricultural, communally held land to developable city land.
The ability of the state to use eminent domain in China faces few restrictions, and although
amendments in 2004 now require the state to “make compensation for the land expropriated or
requisitioned,” in practice this does not approach Western-style fair value standards (Constitution of the
People's Republic of China 1982) (Randolph 2007). Because the state acts as the only intermediary for
these transactions, a large chunk of money can be extracted for the state, and sometimes illicitly for
officials as well. Almost all of city expansion has been done outward, onto this requisitioned and resold
land.
Analysis
In an attempt to analyze the effects of using the Chinese land right lease system, it is necessary
to construct a basic analytical model to provide a framework for the economic arguments to follow. In
real life, property markets are affected by many complex variables, including many variables whose
Brian Schueler Course Paper- Econ 4240 35/4/2015
value is unlikely to be known. This model will analyze consumer demand for properties as driven by age,
location (relative to the city), length of lease, upkeep, and taxes. There are some things that have been
left out for simplicity, and below I will address these briefly.
Consumer tastes are largely left out of this model, and we also ignore the effect that other cities
have on housing demand. Both of these are reasonable assumptions, because in contrast to the height
of Communism in China, property developers now have a decent amount of leeway in their choice of
designs and styles. We can assume that property developers would choose the optimal stylistic
approach to stimulate demand.
The effect of other cities is likely a present effect, especially due to the network effects that
some of these larger cities have created. The standard of living can be much higher in some cities in
China compared to others. Citizens of Chengdu, a city in central China, could expect to make the
equivalent of around $6,000 per year in 2010. By contrast, citizens of Beijing or Shanghai made on
average over $11,000 per year, and in Shenzhen, just across the border from Hong Kong, citizens earned
on average close to $14,000 in 2010 (National Bureau of Statistics of China 2011). However, the cost of
living increases proportionally with income, and so the difference in purchasing power parity between
the cities is not as large. The model that I will be using will mitigate some of these problems by
comparing prices within cities, rather than setting real values and attempting to normalize values across
provinces or the whole nation. Additionally, the Chinese household registration system makes it more
difficult for most citizens to switch locales, reducing the economic benefits of moving to a different city.
Other effects that most certainly affect housing market outcomes are pollution, and local
environment. One of the hypotheses of my model, and a conclusion made by the Anglin paper on
Chinese city growth patterns is that the land lease system leads to overdevelopment of the periphery of
the city. However, considering that much of the pollution that affects Chinese cities is a regional effect
much more like the weather than some perceive, we can assume that the intra-city effects have less
effect on prices than the variables we are analyzing, and can be plausibly omitted without creating
significant bias in the results.
Brian Schueler Course Paper- Econ 4240 45/4/2015
Basic Model of property value
Before laying out the specifics of my model, it should be noted that the fundamental concept of
my property value model is taken from the model that Paul M. Anglin, David Dale-Johnson, Yanmin Gao,
and Guozhong Zhu used to estimate the value of a residence in their paper “Patterns of growth in
Chinese cities: Implications of the land lease”. Their model of value uses a real estate developer as the
actor, and analyzes the level of value that a developer can realize from a property that is already owned,
and evaluates the effect of choosing to redevelop land, or not redevelop land. I will not replicate their
conclusions when it comes to analyzing how the land-lease system affects redevelopment or city growth
patterns, as their analysis is already at a higher level than I would be independently capable of.
However, as their model applies only to residential housing developments, I will alter it, and simplify it
some to better suit my goals of analyzing the whole effect of the land-lease system.
In the Anglin et al. paper the model starts with the assumption that our town is circular, with a
center at x=0, and that there are N locations, so that x ϵ {1,2…,N}. N is the farthest reaches of our land.
Each distance x is essentially a ring with a center at x=0, and radius of x, with a width of 1. It is assumed
that development will occur uniformly, so that all the land in the ring x=4 will be leased (flipped from
agricultural land to city land through the government, and sold to developers) before land in ring x=5
will begin to be leased. This is because the model assumes that distance from the city center will result
in diminishing prices of land.
The Anglin et al. model assumes that the utility of a residential property at any time (t) will
depend on the size of the residence (house) (h) the age of the property (a) and the distance from the
city center (x). This is modeled as:
st=s (ht , a , x )= ψ a
f (x)ht
Here ψ is a constant between 0 and 1 which represents the discount rate on the service of a
residence due to the age of the house. In ψa as (a) becomes larger, as long as psi is less than one, it will
Brian Schueler Course Paper- Econ 4240 55/4/2015
decrease. F(x) represents a function of distance to a residence. This likely contains many variables
which are frequently changing, and difficult to quantify, such as traffic, price of owning a vehicle,
efficiency of public transportation, the value of time, etc. We can make a usable model by simply
assuming that as x becomes larger, we will see F(x) also increase, in effect lowering the utility of our
residence.
The Anglin et al. paper then simplifies the model to model the utility of a unit of residence (one
could think of this as standardizing for size, as a model of the utility of one square meter of residence).
This model also sets the service of all units as a function of the utility of the best possible unit, which we
define in the model as a new unit (age = 0) and at the center of the city (x=0). The rent for this best
possible unit of residence will then be defined as:
Qt=ψ0
f (0)∗1
Where ht=1to normalize for one unit of housing size. Utility is a very ambiguous measure, and
we can assume that utility and rent are highly correlated, as it would be irrational for an actor to pay
more for a property which returned less benefit. We then can determine the (economic) rent for a unit
of housing based on these assumptions as:
qxat =Q t∗ψ
a
f (x)
The Anglin et al. model only applies to residential housing, and so some of these assumptions
must be altered to have the model apply to all forms of property. First, it should not be a difficult
argument to state that the optimal location which would minimize transport distances and distance to
other services is not the same for industrial actors as it is for residential actors. These centers of activity
are likely more congruent when comparing commercial and residential areas, but it shouldn’t matter
much for our conclusion. Although the true geographic centers of industry, commerce, and residential
land developments may differ some, we will assume that all properties share a common x=0 central
Brian Schueler Course Paper- Econ 4240 65/4/2015
point. This is permissible, because our model doesn’t actually require a real geographic central location,
but simply states that actors will find more value in property when the distance to desired services and
locations is shorter, regardless of what those desired services and locations are. Additionally, even
though when (x) is large for one sector, it may be smaller for another, we would expect that any area
more valued by one sector than another would be purchased by that first sector before it would be
purchased by the latter, and the first would naturally be willing to pay more. As portions of one sector’s
rings are taken up by other industries, we would see a decrease in overall land availability in each ring x.
This could be easily negotiated in our theoretical model by changing the width of each ring. For our
theoretical model, the actual dimensions of each ring are less important than the relationship of the
center to the periphery, and so including multiple industries is permissible. To simplify for our model,
we just aggregate society’s central point down to x=0. If we modify the Anglin et al. model to apply to all
properties, we can describe it as:
pxat =
Pt∗ψ pt
f ( x )
Wherepxat is the rental value of a unit of property, ψ p
t is the coefficient of depreciation for
property in general (the average coefficient of depreciation of property) , and f(x) remains a function of
distance to the central location.
This model should work for any kind of property, regardless of the purchasing or ownership
system. It also can be converted to various time periods, based on the time period that the optimal
property unit-rent Pt is set to. If we assume the model is set for yearly rents, then we can find the total
remaining value of a land-lease contract at time t with the function:
Pv=∑i=1
n−t
¿¿
This equation assumes that the value of a property is equal to the sum of future rents minus the
sum of future expenses. In this equation, n represents the term of the lease, and t is the time which we
Brian Schueler Course Paper- Econ 4240 75/4/2015
start, compared to the lease. For a residential lease, the term is 70 years, and for commercial and
industrial uses it can last up to 40-50 years (Randolph 2007). In a western economy, where property is
owned outright, this model would specify ¿∞ . (Or a sufficiently high number) Forψa+i, the age (a) is the
age of the property at time (t), and so (a+i) simply refers to the total age of the property for each year.
Despite talk of property taxes being established in China, currently no form of property tax
exists, and many observers doubt that a widespread property tax will be implemented any time soon, as
national registers of property do not exist, and among the wealthy and well connected, many want to
keep their holdings of property secret (Vanderklippe 2014). However, the existence of a tax would
certainly change the calculus for a property’s value, and so it should be included to better reflect the
return that property owners accrue. Taxes per property unit are represented as T p , and is calculated
for each period from the time t, to the end of the lease.
As any property holder knows, upkeep costs are a function of age, and will increase as a
property becomes older. u (a+i ) is simply a cost function that increases as the age of a property
increases.
In an initial analysis, one can see that the value of a property to a landholder is likely less when
there are limits on the time-frame of land holding. This holds as long as ψ is large enough (close enough
to 1) that as the age of the property approaches n, the term limit, if the quantity ψa is still significantly
greater than 0, then value is lost to the landholder by capping the lease limit. However, a leaseholder
will experience no loss in value if the property’s depreciation rate is strong enough that the property
loses value to the point the property cannot offset costs before, or at the expiration of the lease.
Additionally, one should consider that there may be situations where a time limit may affect the
decision to redevelop a destitute property. The Anglin et al. paper discusses this point well, and will be
addressed later in this section.
Finally, in this model we should establish how renewal of leases is likely to impact property
holders. This is the most ambiguous aspect of the Chinese land-lease system. Because the first land-
Brian Schueler Course Paper- Econ 4240 85/4/2015
lease agreement in China was initiated only as recently as 1987, only a few special leases in the
Shenzhen area have expired. In 2002 some of the experimental leases granted in Shenzhen expired,
which created a minor panic. The city government eventually created a law stating that property holders
with expired leases could either pay 35% of the face value of their property for a 40 year extension, or
lose the property outright (Chinese Economic Review 2013). Other experts believe that the government
plans that future renewals will be of shorter periods, and most certainly would include a significant
renewal fee (Randolph 2007).
We can explain the expected value of a property with an extension of the model of property
value, set for China’s lease renewal circumstances. Here we have the property value function,
PV=∑i=1
n−t
¿¿
This equation may not be complete. We can assume that with current laws passed in 2007,
which state that leases will be automatically renewed upon the expiration of the contract for residential
properties, holding the property even at the expiration of the contract may have some value as a claim
on renewal of the lease (Chinese Economic Review 2013). With renewal, property holders are not
certain to renew, as they are still subject to the whims of the government as a landlord. As such, the
sum of the next lease’s rents minus the costs, and the renewal price should all be multiplied by the
probability of renewal.
E(P¿¿R)= (Prob .Renew )∗¿¿
In this equation, we find the expected value of a property renewal. The probility of renewal is
multipled by the expected value of the new lease. It should be noted that when the age of a property is
considered, we count the age at the intital t, the remaining time on lease 1, and then can start summing
upwards from that point, so we get the term a+ (n−t )+i. Our new period is set as nR, and the renewal
price is set as RPrice. We can thus establish the value of a property over two lease periods as
Brian Schueler Course Paper- Econ 4240 95/4/2015
PV+PR=PT where PT represents the total value, not to be confused with Pt which is the optimal
property rent at time t.
So in this basic model of property values, we see that for a property owner to want to renew,
the age of the buildings on the property will play a role. As the property is older, loses value, and has
higher upkeep costs, we can see that there is less incentive to try and renew the property lease.
It may not be intuitive, but from our model we can determine that the land lease system creates
some perverse incentives for development and for redevelopment, could have impacts on how property
is used for collateral, and may also impact local government finance.
Model ParametersSome of the parameters in this model are adopted from the model used in the Anglin et.al.
paper, which had been calibrated to match with observed tendancies in Chinese Cities. ψ is drawn from
the Anglin paper, and is 0.99, as well as the structure for the difference in transportation costs. In the
Anglin model, f (0 )=1 ,∧ f (N )=2 where N is
the farthest periphery of the city. However,
considering that many smaller towns and small
cities have a less pronounced effect on
property values, in my model the value of
f (N )=1.75.
In Chinese law, residential properties
have lease lengths of up to 70 years, and this
seems to be the norm for a lease length. Lease
lengths are 50 years for commercial uses, and 40
years for industrial property use (Chinese Economic Review 2013). These lengths will be calculated, and
weighted within the model by the intensity of which land is utilized. I will use rounded estimated 2012
rates, of 11% commercial, 45% industrial, and 23% residental as benchmarks.
Figure 1 Share of main urban land use, 2005-2012 (The World Bank and the Development Research Center of the State Council, P.R. China 2014)
Brian Schueler Course Paper- Econ 4240 105/4/2015
Upkeep prices levels have been estimated based upon data from U.S. apartments holdings.
While it is most likely true that various other
land utilizations, particularly for industrial
usages may require more land upkeep as a
percentage of total land value, land upkeep
costs on industrial, or commercial applications
as a whole were not available. Using
apartment upkeep rates to model average
property upkeep over time should provide a
conservative estimate for total upkeep
costs. From my estimations, the intial cost of upkeep at year 0 should be equal to 0.14% of initial price,
and increase by 0.0092% each year.
I have set initial price at 1, which sets
establishes the 1 year optimal rent at 1 (rent in new
property at optimal location). This means our value
units from this model are not set in currency, but
instead in units of 1 year optimal rent equivalents. The
probability of renewal, and renewal price are also
exogenous, but difficult to pin down. Patrick Randolph
Jr. in Probate and Property writes that it is his
understanding that unless there are policy changes, the
government is highly unlikely to renew for periods
which are as long as they are currently, and very likely
will not allow renewals to occur for free. (Randolph 2007).
0 5 10 15 20 25 30 35 40 450
0.050.1
0.150.2
0.250.3
0.350.4
0.450.5
f(x) = 0.00923964968152866 x + 0.140294585987261R² = 0.907542399950403
Upkeep trend
Series2Linear (Series2)
Age
Perc
ent o
f Pro
pert
y Va
lue
Figure 2 Upkeep Cost Estimation – Data from (Liu 2005)
Baseline Parameters
Years since Lease 0lease Length 50Distance Effect 1.75Initial Price 1Tax level 0upkeep initial price % 0.1400
%upkeep increase/year 0.0092
%psi 0.99probability of renewal 0.5renewal price 0.25distance from city center (0-1)
0
Renewal term 20
Brian Schueler Course Paper- Econ 4240 115/4/2015
Redevelopment rates, City expansion, and overdevelopment
One of the criticisms leveled at China’s economy lately has been overdevelopment by property
firms. Huge amounts of money went into the development of oversized housing developments and
business districts. Large amounts of Western capital, searching for higher yields elsewhere entered into
the Chinese property market. Billions of dollars was invested, and for the majority of the 2000’s,
property prices were skyrocketing (Barboza 2015). However, as the real estate market slowed in the
early 2010s, many firms were left with large amounts of over-development. A survey by the
Southwestern University of Finance and Economics found that more than 1 in 5 homes in Chinese cities
were empty, which is equivalent to 49 million homes. At the time of the survey, another 3.5 million
homes remained unsold. The excess supply of housing could create liabilities for the financial sector,
with an estimated 674.33 billion dollars in outstanding loans on vacant homes. The study estimates that
11.2% of these homes would be underwater on their mortgages with just a 30% fall in home prices
(Fung 2014).
The Anglin et al. paper specifically evaluates how the land-lease system affects the choice to
redevelop properties or develop new properties, and how this effect changes how Chinese cities will
grow. By using their model, which I have based my model upon, and by calibrating the variables to fit
with observed patterns of growth in Chinese cities, their model demonstrates some of the effects that
the land-lease system has. They find
that redevelopment of land occurs less
frequently regardless of the age of a
property when it is leased in the
Chinese system rather than owned.
Under the land-lease system, excess
land is used, the average age of
properties is older, and the average
values are lower. About 5.7% extra land is used under the lease system, which when applied to China’s
Figure 3 - Rate of Redevelopment - (Anglin, et al. 2014)
Brian Schueler Course Paper- Econ 4240 125/4/2015
residential developments in 2014, is roughly 700 square kilometers. The study continues that if the
effect is the same for industrial and commercial developments, which is an argument that this paper
supports, the total excess developed land due to the land lease system would be about 2,300 square
kilometers (Anglin, et al. 2014).
We could see how this might occur, if we look at the equation of the renewal value of a
property, and compare it with purchasing a new property on the periphery. If a developer decides that
renewing an existing property, that has already aged a few decades is less profitable than purchasing
ostensibly cheaper land on the periphery of the city, and building a property that may also have lower
returns, the developer will allow their first property’s lease to
lapse, and purchase the second. This would result in a new
lease being given at a lower price because the new owner
must chose to stay with an old building with higher costs and
lower rents, or pay for demolition of the old building and
reconstruction of a new building. Either way, because the
resale value of the property at the end of the lease is only
equal to the value of the right of first refusal when re-leasing,
assuming that existing owners will even be granted that right,
owners may face situations where they can better maximize
benefit by not redeveloping their old land, and instead
investing that money on other, newer properties.
According to the Anglin et al. paper, if China used a
land ownership model rather than the land lease system, the
price of homes would be 7.3% lower, the supply would be
7.5% higher, and buildings would be 8.3% younger (Anglin, et al.
2014). Because the length of the lease is shorter for commercial
and industrial uses, and the higher levels of responsiveness to cost
Figure 4 Quantity of property in Anglin, et al. model. Top graph represents equilibrium price and quantity. The middle graph represents a revenue maximizing government, and the lower graph simulates the population doubling by time=50.
Brian Schueler Course Paper- Econ 4240 135/4/2015
changes that corporations have, we could see an even larger effect for the property market as a whole.
It may seem counterintuitive that excess property would also result in higher prices, but the prices do
have a tendency to increase nearing the 50 year mark in the Anglin et al. model because an initial excess
of housing developments leads to a contraction in the housing market, resulting in more area being
used, overall, but because the center regions are not redeveloped, or are redeveloped cheaply, total
supply of housing space is actually less.
One important aspect from the Anglin, et al. paper is the prediction made about how property
development under a lease-holding and full ownership structure differ. Under ownership, we see
development rates remain continually increasing, but under a lease-holding structure, we see initial fast
growth followed by a stagnation in growth rates. In fact, under the models that were provided, as lease
terms came closer to expiration, we see some predictions of negative growth in developments. (See
Fig.4 above) These models show the total quantity of properties over time, and seem to suggest that in
later years there is more demolition than there is construction. Later in this paper, these findings are
used to help forecast new construction
per year into the future. These forecasts
will attempt to remain on the
conservative side by simulating declining
levels of positive new construction
rather than rates dipping negative.
Analysis of how the resale value
of a property changes over time through
my model can show how in a system
with unlimited property rights, the
resale value intuitively declines at a
much slower rate. Even when you
apply a tax of 30% optimal yearly
0 10 20 30 40 50 60 70 800
5
10
15
20
25
30
35
40
Value of Property over Time
Residential - ChinaCommercial - ChinaIndustrial - ChinaUnlimited Property RightUnlimited Property Right with 1.2% property Tax
Year
Opti
mal
1 Y
ear R
ent E
quiv
alen
t
Figure 5 Depreciation based upon Property lease term. Source – Author’s Calculations
Brian Schueler Course Paper- Econ 4240 145/4/2015
rents – which is roughly equal to a 1.2% property tax, assuming that the tax rate is levied on the current
market value, we still see significantly less depreciation.
Impacts on lending markets
Although China’s financial markets have only been opened up to the world economy for a little
less than 30 years at this point, the size of China’s financial system has grown massive. Of the largest 100
banks in the world by total assets in 2013, China has 17 of them, with the Industrial and Commercial
Bank of China as the largest in the world, holding roughly $3.126 trillion dollars in assets. The United
States by comparison has only 10 of the top 100 largest banks in the world, with JP Morgan Chase as the
largest bank by assets in the United States with 2.476 trillion dollars in assets, barely edging out China’s
3rd largest bank, the Agricultural Bank of China for 6th place in world rankings (relbanks 2013). All told,
the total value of assets in banks in China is 15.37 trillion dollars at current exchange rates (March 2015)
(Martin 2012).
Although China’s banking and financial sector is enormous, it also is highly exposed to the
mortgage market, with 38% of all large bank’s outstanding loans being loans to developers and
individuals using property as collateral (Kan and Wu 2014). In a market where property rights were
indefinite and inviolable this might be a small concern, but with the land-lease system, there are more
serious concerns. In 2030 the 40 year leases of industrial and hotel properties will begin to come due.
With the level of uncertainty that lease renewal would create, we can assume that the maximum value
that a lease-holding could be worth would be its value remaining at the time that the loan was issued,
and potentially less. As collateral, the value of a property would likely be estimated using an accounting
method rather than with the economic value method previously used.
PAV=(∑i=1n−t PMP
n )+(V x )
Brian Schueler Course Paper- Econ 4240 155/4/2015
This shows that the accounting value of a property PAV is equal to the sum of the amortized
market price of the property PMP over the remaining years left in the lease, plus the value of having the
right of first refusal to renewing the lease, in a sense the value of the location, or VX . VX shouldn’t
become large unless the terms of a renewal are much less than the costs of purchasing a new lease
where a land lease has already expired. For this to be valuable to a lending institution, these favorable
terms would also need to be transferable. This would mean that although favorable renewal terms
secured through political connections might make renewal an obviously beneficial economic choice,
because these connections are most likely un-transferable, this still would have little effect on that
property as collateral. There is very little discrepancy between the model described earlier in this paper,
and the accounting version of the model as long as upkeep costs remain at reasonable levels, and taxes
are not too high. As long as these aspects remain within these assumptions, we would not expect
consumers and financial institutions to value properties much differently over time.
From previous models that show how the land lease model results in excess property
development for the first half of the lease’s term, followed by a slowdown in new construction in the
second term, we can assume that a significant proportion of land leases will have similar expiration
dates, with expirations starting as soon as 2030, increasing to high levels of industrial and hotel land
lease expirations in 2045, followed by a boom of commercial land lease expirations in 2055, and
residential lease expirations increasing around 2075. The peak dates for lease expirations, and when
they will begin to increase to high levels is difficult to determine, because as Vanderklippe stated,
national registries of leases do not exist, and in some places only paper records exist (Vanderklippe
2014). However, using our model of property value, and some limited estimates of new construction in
urban areas, we can make a rough estimate of how property values will change.
In a publication put forth by the Reserve Bank of Australia, estimates are made for urban
residential floor space construction, dating back to 1990, and projecting forward until 2030. I have based
assumptions for total property construction on these estimates, and although using residential data to
predict industrial and commercial construction is not ideal, it would be reasonable to expect that
Brian Schueler Course Paper- Econ 4240 165/4/2015
residential, commercial, and industrial construction is relatively well correlated given that individuals
tend to live near where they work,
and buy more goods and services as
their ability to purchase homes
increases.
In the graph shown is a set
of projections for new construction
of residential housing. It is important
to analyze not simply new property
purchased, but instead new
construction, to better take into account the value of the Chinese property market. The Reserve Bank of
Australia factored in demolition and reconstruction into their projections which yields a more accurate
estimation of value than would assuming that properties will depreciate constantly. Understanding that
it is quite difficult to estimate future economic conditions, I have provided three scenarios, one with
high economic growth, and new construction at levels seen in the mid to late 2000’s, when China had
GDP growth in the double digits. I have also included an estimate with low growth, where new
construction falls sharply, which would reflect stagnation, population decreases, and overall economic
malaise. It would seem relatively likely that actual growth in construction will be between these two
assumptions. We are unlikely to see growth continue to increase year-on-year, as Chinese property
supply seems to have now caught up with demand, which will likely fall in the future as demographic
trends take hold. By 2050, China’s population of individuals 15-64 is projected to decrease by 11.4%,
with the median age increasing from 34.5 in 2010 to 48.7 in 2050 (Economist 2012). . In fact, when the
above forecasts are compared to the projections in the Anglin, et al. paper, these assumptions all are
relatively optimistic.
Without changes to the current system, and assuming that the current proportions of
residential, industrial and commercial property as a part of China’s total property market remain the
1980 2000 2020 2040 2060 2080 21000
200400600800
10001200140016001800
New Construction Projections
Low Growth High GrowthMedium Growth
Year
Cons
truc
tion
Mill
ion
Squa
re M
eter
sObserved Forecast
Brian Schueler Course Paper- Econ 4240 175/4/2015
same, we can estimate how
the total value of the
Chinese property market
may change over time. The
calculated units of value
technically are in yearly
optimal rents multiplied by
million square meters. This is
a unit that is not very useful
for conceptualizing the size
of the market, so it has been left off in the graphical form.
1980 2000 2020 2040 2060 2080 2100
High Growth Outlook
Residential (70yr) Commercial (50yr) Industrial (40yr)
Year
Tota
l mar
ket v
alue
Brian Schueler Course Paper- Econ 4240 185/4/2015
As one would
expect, the length of
the lease not only
affects when the
property market
peaks, but also how
high that peak is,
although in a
congruent manner.
Remembering the
proportions of each
sector, we can aggregate these results into a total market. In these projections we see a very interesting
result. Even though the projected growth rates of construction were quite different, the model predicts
the value of the property market to begin to decline as a whole within a 10 year window. The low
growth model predicts decreases to begin in 2042, while the high growth model predicts decreases to
begin in 2052. By aggregating residential, industrial and commercial property markets together, overall
the market value changes less rapidly, but the overall trend does not change. Given that property values
are such a large asset class compared to the economy as a whole, decreases in property market
valuations across the board would cause consumers to behave as if they have fewer assets, reducing
aggregate demand, and likely depressing the economy. To compare the Chinese market with a
hypothetical, more western counterpart, I simulated unlimited property rights by setting lease length to
500 years. To make a hypothetical comparator property market more realistic, I also set tax rates to 0.43
yearly rents, which is approximately a 2% levy on the total value of the property. It was pleasing to find
that the model valued property with this tax at around 20 optimal yearly rents, because in many cities
the price to rents ratio sits between 15-25 (CNN Money 2009). This seems to indicate that my model is
also decently calibrated to estimate property valuations with unlimited lease length, and taxes, albeit
roughly, as we are comparing economic rents to lease rents. It is quite clear as demonstrated in Figure
7 that there is significant amounts of foregone value to property holders when artificial limits are made
1980 2000 2020 2040 2060 2080 2100
Aggregated Property Market Values, Projected
Total, High Growth Total Medium Growth Total, Low Growth
Tota
l Mar
ket V
alue
Figure 8 Projections of aggregate market values in various scenarios
Brian Schueler Course Paper- Econ 4240 195/4/2015
Even if one has doubts about forecasted construction, which would not necessarily be irrational,
as construction is relatively volatile, there still remains a significant difference in the market value with
unlimited property rights verses the land lease system. As leases start to expire at the same rate that
new properties are constructed, we see the value of the market flat-line. Although China’s property
market has been quite buoyant during the last few decades, even continuing growth at the current pace
seems quite unlikely.
These effects will likely not have homogenous impact across China. In larger cities, these effects
will occur earlier, and in smaller cities which experienced growth spurts later in the 2000’s, these effects
will begin to occur later. A study using satellite imagery to chart city growth found that large eastern
cities, had spurts of growth in the late 90’s or early 2000’s with growth rates slowing through the later
2000’s, while medium sized cities, and large western (western China) cities experienced more growth in
the later 2000’s (Schneider and Mertes 2014). We can make a rough guess that these booms also
correlated with increased property development, especially because population estimates in the study
were determined from visible urban sprawl.
As these loans begin to approach expiration, we would expect to see banks more and more
unwilling to accept these properties for loan collateral, or value those properties less, requiring more
and more property to be used as collateral. As many loans made in China currently use property as
collateral, a decrease in the amount of collateral available will make it more difficult to access credit,
particularly for property owners looking to take out loans to renew leases. The high economic growth
model predicts that between 2050 and 2100, the market value would fall by 14.5%. For the medium
growth model it predicts that between 2044 and 2100 market value would fall by 30%, and for the low
growth model the market’s value decreases by almost half (48.2%) between 2040 and 2011. For
comparison, between the beginning of the US recession starting in December 2007, until January 2012,
when the Case-Shiller U.S. National Home price index bottomed out, housing market values had fallen
by about 22.5% (Federal Reserve Bank of St. Louis 2015). While the timeline certainly is longer, the level
Brian Schueler Course Paper- Econ 4240 205/4/2015
of decline predicted in China compared to the recent levels of property market declines from the US’s
most recent recession should illustrate the significance of these figures.
Government Finance
A recession, or even a long slump that could occur as the property market artificially lost value
preceding lease expirations is something that no government would desire, and it is possible that the
Chinese government may desire to alter the land-lease mechanism before lease expirations occurred to
attempt to mitigate the negative effects a wave of lease expirations would have on credit markets.
However, leniency is unlikely to occur because of the incentives that local governments have to extract
revenue from the land-lease system.
In 2010, China’s central government collected around 51.1% of all government revenues across
China, but accounted for only 17.8% of all government expenditures, meaning essentially that local
governments have become responsible for more than 80% of all government expenditures in China,
while receiving less than 50% of all revenues (Ye and Wu 2014). This leaves local governments
struggling to make up the difference. A system that has suited many local governments has been to use
land financing to make up the difference. In some cities, the taxes and profits made from the sale of the
lease to developers accounted for over 50% of government revenues (Zhang and Barnett 2014). In fact,
in the study done by Ye and Wu, it was shown that across the 286 cities surveyed, between 2004 and
2009 the average percentage of revenue from land fees ranged between 43.3% to 62.7% (Ye and Wu
2014).
With such high percentages of government revenue accruing from property development,
Chinese local governments have large incentives to keep the growth of property development high,
even as property markets become saturated. As a result, local governments have created mechanisms
to finance further property development and raise money through loans, and bonds without violating
Beijing’s prohibition on local governments issuing bonds. These mechanisms in general are called Local
Government Financing Vehicles (LGFVs). These financing vehicles have been created with government
Brian Schueler Course Paper- Econ 4240 215/4/2015
capital and subsidies in order to raise money for various projects, using properties transferred to these
corporations by the government as either collateral for bank loans, or to increase credit ratings in order
to issue corporate bonds. Between 2012 and 2013, LGFV’s accounted for roughly 20% of the total
corporate bond issuances, and by 2010, bank loans accounted for 80% of total local government debt
(Zhang and Barnett 2014). Up until January of this year, the only way that local governments can access
debt is through these LGFVs, which use property as collateral. Since the beginning of 2015, the central
government has allowed for local governments to issue a limited amount of bonds to help cover
repayments of prior debt, which will help extend solvency, but likely will not solve underlying problems
(The Economist 2015). Currently, as stated in the article, Chinese local governments hold almost 18
trillion yuan in debts, ($2.88 trillion). As long as the underlying gap between revenues and expenditures
fails to close, local government debt will continue to grow.
While the underlying cause of high local government debt is the lack of revenue, the structure of
the land lease, and the mechanism to take commune held agrarian land and flip it into city land for
profit has led local governments down a dangerous path of simultaneously inflating a property bubble
and jeopardizing their future revenue streams. Additionally, much of the debt held in China is held by
companies, with construction and property management and development industries in China reaching
levels of debt to equity at 170% for construction firms, and 150% for Property management and
development firms by 2013 (The Economist 2014). With a slowdown in property markets, as predicted
by the Anglin et al. model to occur in the land-lease model after 20-30 years, the solvency of many of
these firms could be brought into question. Local governments could be stuck with a depressed
economy and serious questions as to their solvency and ability to pay back loans. This would be
exacerbated if the current trend of local governments bailing out creditors of moribund companies
continues, as has been the case with recent near defaults of large construction and infrastructure
companies. Recent examples have been investment products such as Credit Equals Gold #1, as well as
Huatong Road and Bridge Group, and Chaori Solar (The Economist 2014).
Brian Schueler Course Paper- Econ 4240 225/4/2015
Local governments should be concerned about the effect that the land-lease system has on their
ability to generate revenue, and the boom-bust cycle that the lease terms can cause. Even though
China’s urbanization rate remains lower than some other similarly developed nations, we cannot expect
demand for urban properties to continue to rise at the rates that occurred in the late 1990s and 2000s.
In a land lease system where the payment for leases is completely up-front, and the only revenues
garnered by the state are from renewals or newly granted leases. Even if we assume that the large
economic slowdown predicted by the property redevelopment and value models does not occur, the
absence of truly private property makes revenue generation difficult for localities. The revenues that
could be generated by the land lease system are highly problematic. Currently, with long lease periods,
we will see less rapid depreciation of property value, and as the lease length becomes longer and longer,
the property value becomes more and more similar to true privately owned property. However, with
long lease values, government revenues fall, or lease prices must be incredibly high.
Government revenues could be higher if lease periods are shorter, because individual lease
renewals would occur more often. This process could also prevent boom and bust cycles resulting from
mass lease issuance and expirations happening in similar time periods. However, this may be an even
worse situation, because this would decrease the value of each lease across the board by limiting the
time period over which rents may be accrued, which could result in an overall long term decrease in
household wealth, as property holdings lost value.
Although most analysts view China’s current wealth and control over banks as hedges against
sudden financial crisis, structural faults if left unsolved could over-time drag down economic prosperity
in China. Additionally without solving the structural deficits that affect local governments, debts will
grow to unsustainable levels. With progressively higher debts, the effect of a property crash or
economic slowdown would be dangerous. It also should be noted that during the 2008 recession, a
significant portion of stimulus funding were delegated to be raised by local governments. Higher debts
in the future may prevent adequate response to recessions and depressions.
Potential Solutions
Brian Schueler Course Paper- Econ 4240 235/4/2015
The first and foremost concern with the Chinese land lease system is the more rapid decrease in
the value of property assets that the system creates, because we might expect this to cause economic
stagnation and recession. The other two issues feed into these effects. The way that housing
development adjusts to the land-lease system causes high levels of development early, and stagnation in
construction as time goes on, with cities becoming older, more expensive, and less efficient as a result.
Governments accelerate these effects because the current revenue model creates perverse incentives
which exacerbate the problem of overbuilding in the short term which we would expect to cause worse
economic outcomes in the longer term.
The first, and perhaps most important change that could be done would be to prevent
landholdings from losing value due to lease expiration. The periodicity of land holdings is the core issue,
affecting redevelopment and concentration of cities, and creating the potential for serious credit crises
in the future and questions of solvency for local governments.
The easiest way to prevent this periodicity would be to simply privatize land, and guarantee
indefinite right to that land similar to how
most Western systems of property
ownership operate. With adequate property
taxes, revenue could be generated to fund
local governments. As long as property rights
were well protected, the value of a property
as collateral would be secure. We might also
see various other improvements in city-wide
development described in Anglin et al.
However, as long as China remains
technically a communist nation, these
changes will be anathema. More creativity
1980 2000 2020 2040 2060 2080 2100
Limited and Unlimited Term Property Rights
Residential Medium GrowthCommercial Medium GrowthIndustrial Medium GrowthUnlimited Property Rights - Medium Growth Unlimited Property Rights With 2% Tax - Medium Growth
Prop
erty
Mar
ket V
alue
Brian Schueler Course Paper- Econ 4240 245/4/2015
may be needed to reconcile the necessity of eliminating the negative effects of the current land-lease
system with political necessity of communist party policy makers to at least pay lip service to doctrine.
The current incentives for flipping unnecessary amounts of agricultural land and converting it
into urban development are perverse and can lead to dangerous outcomes. Ultimately, the process of
developing agricultural land should be determined by the market rather than by bureaucrats looking to
boost the government’s revenue, or potentially their own pocketbooks.
The effects of overdevelopment should be taken seriously. The models that I used to evaluate
and forecast new construction in the Chinese economy do not assume any dramatic decreases in
construction. If current levels of construction are highly abnormal, and indicative of a property bubble, a
correction to lower levels would create a much sharper decline in housing market values that would
arrive sooner, and become a much more severe drag on economic progress.
To prevent the inflation of a property
bubble, local governments need to have access to different revenue streams, and should not have the
ability to profit off of the requisition and resale of agricultural land to construction companies.
Most simply, this could be done by privatizing agricultural land, strengthening property rights
against the government by reducing the ability of the government to acquire land through eminent
domain powers, and placing property taxes on agricultural lands as well.
Some steps concerning privatization of agricultural land have already been taken, which may
signal steps in the right direction. Recently, China has pushed for more loans and credit to be extended
to the agricultural sector, and has pushed banks to increase lending, however banks are reluctant to
invest their assets with no held land as collateral (Reuters 2015). A pilot project in Liaoning province
(Northeast China) has begun to allow farmers to use their farmland use rights as collateral for loans,
which seems to indicate a willingness in government to cede some aspects of property ownership right
to individuals. In the scheme, farmers could place some of the land that they are entitled to cultivate up
as collateral, for a period of time. One farmer put his use right on 2.3 hectacres for 10 years up as
Brian Schueler Course Paper- Econ 4240 255/4/2015
collateral for a loan of just over $7,000. However a loan of that amount seems to be an outlier, with the
average loan less than $200 (Qiu and Yang 2009).
This policy, while groundbreaking for agricultural China, falls far short of a privatization scheme,
and fails to create a system where government revenues could be collected, or a system that could
grant more economic freedom to rural citizens.
Recommendations:
Policy recommendations are useless if they are infeasible. Because of China’s ideological
constraints, solving current problems that arise from the land-lease and other land-holding structures in
China will be relatively difficult. However, there are some potential compromises that could help solve
many economic problems while not appearing to be “too capitalist”.
Although full privatization of property would be optimal, other changes could be made that
preserve the state as the ultimate holder of property, and property values to more accurately reflect
economic conditions. Because of the
political constraints that bind the Chinese
Communist Party, we can probably rule
out total privatization. However, there
may be room to modify the Land-Lease
system so that periodic effects are less
noticeable. One way would be to spread
lease payments out so that they are
payable yearly. As long as the payments
to the government are less than the
economic rents which are accrued each
period, and are relatively low, these
lease payments would function
1980 2000 2020 2040 2060 2080 21000
5
10
15
20
25
30
Adjusting Tax and Steady Fee - Total Remaining Value over time
Adjusting Tax (0.3 property value at time)Steady Fee (.15 Optimal Rent)
Opti
mal
Yea
rly R
ents
Figure 11 -- Comparison of Steady level fees and continually assessed taxes based on current property value. By my estimations, a tax of 30% current property value is similar to a 2% tax on the total value of a property when assessed over a longer period of time.
Brian Schueler Course Paper- Econ 4240 265/4/2015
similarly to a tax. In my model, these two systems raise similar amounts, with the steady fee model
yielding somewhat smaller revenues.
In figure 11, we calculate the total
remaining value assuming very long lease
length (500 years was the variable used in
the model). If a precedent was set that
lease terms automatically renew without
an additional (or a relatively small) fee,
and were quite long, we start to approach
a situation where the negative effects of
the current land-lease system would be
mitigated. As long as property owners, or
prospective buyers did not worry about a
looming land lease deadline where
property right would be either lost, or
become much more expensive, we
would likely see property values look
similar to how they would in a more western system.
If we calculate the effect of a steady fee on our property markets, we can see that there is a
slightly more negative effect later in the forecast than a tax would have, but not nearly as bad as the 70
year lease limit effect. The main difference between a steady fee system and a tax is that the steady fee
does not fluctuate with the resale price of the property, and so will be a higher proportion of the yearly
economic rent later in the property’s life than a tax which would adjust to assessed values. Overall, a
steady fee system with long lease terms has high potential to mitigate many economic problems that
may occur under the current land-lease system.
1990 2010 2030 2050 2070 2090
Aggregated Property Values by System - Medium Outlook
Unlimited Property Rights With 2% Tax - Medium Growth 70 year leasesUnlimited Leases 10% steady FeeUnlimited Leases Steady Fee 15%
Year
Valu
e
Figure 12 Comparing Steady Fee system to Western tax system and 70 year leases
Brian Schueler Course Paper- Econ 4240 275/4/2015
As China’s stellar GDP growth rate of the 90’s and early 2000’s has started to come back down
to earth, there will be more pressure on Chinese leadership to recharge the economy. Property markets
would be a good place for leaders to look. By fixing the current system, Chinese leaders could not only
prevent future crises, but immediately improve the economic conditions for many citizens while
providing more stability to local governments. For years, the Chinese Communist Party has preached on
the social stability that communally held land would provide. Leadership in Beijing will be faced with a
question of whether to follow doctrine or data while planning for the future. For the sake of the 18% of
people in the world who are Chinese, and the billions more who are affected by the health of the
Chinese economy, we must hope that progress prevails.
References
Anglin, Paul M., David Dale-Johnson, Yanmin Gao, and Guozhong Zhu. 2014. "Patterns of growth in
Chinese Cities: Implications of the Land Lease." Journal of Urban Economics 87-107.
Barboza, David. 2015. "In China, A Building Fenzy's Fault Lines." New York Times, March 13.
Chinese Economic Review. 2013. "If Beijing is your landlord, what happens when the lease is up?"
Chinese Economic Review, June 17.
1982. "Constitution of the People's Republic of China." People's Daily Online. December 4. Accessed
March 21, 2015. http://en.people.cn/constitution/constitution.html.
Fung, Esther. 2014. "More than 1 in 5 homes in Chinese Cities are Empty, Survey says." The Wall Street
Journal014, June 11.
Kan, Huo, and Hongyuran Wu. 2014. "Slowing Property Market is Banking Risk, CBRC Offical Says."
Caixin online, May 14.
Brian Schueler Course Paper- Econ 4240 285/4/2015
Martin, Michael F. 2012. China's Banking System: Issues for Congress. CRS Report for Congress,
Washington D.C.: Congressional Research Service.
Meyer, Michael. 2009. The Last Days of Old Beijing: Life in the Vanishing Backstreets of a City
Transformed. Walker and Co.
National Bureau of Statistics of China. 2011. National Data: National Bureau of Statistics of China.
Accessed March 22, 2015. http://data.stats.gov.cn/english/.
Randolph, Patrick A. Jr. 2007. "The new Chinese Property Law: a real estate practitioner's perspective."
Probate and Poperty 14.
relbanks. 2013. relbanks.com. December 31. Accessed March 22, 2015.
http://www.relbanks.com/worlds-top-banks/assets-2013.
Schneider, A, and CM Mertes. 2014. "Expansion and Growth in Chinese Cities, 1978-2010."
Environmental Research Letters.
Stein, Gregory M. 2007. "Mortgage Law in China: Comparing Theory and Practice." Missouri Law Review
1315-1352.
The Economist. 2015. "China's Local Government Debt: Defusing a Bomb." The Economist, March 11.
—. 2014. "Chinese Debt: A Moral Deficit." The Economist, October 18.
Vanderklippe, Nathan. 2014. "China Moving quickly to roll out property taxes nationwide." The Globe
and Mail, September 1.
Ye, Lin, and Alfred M. Wu. 2014. "Urbanization, Land Development, and Land Financing: Evidence from
Chinese Cities." Journal of Urban Affairs.
Brian Schueler Course Paper- Econ 4240 295/4/2015
Zhang, Yuanyan Sophia, and Steven Barnett. 2014. Fiscal Vulnerabilities and Risks from Local
Government Finance in China. Working Paper, IMF.