China's Housing Provident Fund: Its Success and Limitations

Full text

HOUSING FINANCE INTERNATIONAL – March 2006
In recent years, home ownership has
become a key item on the agenda of
China’s housing reforms. As a consequence
of this policy direction, outlined in State
Council decisions in 1994 and 1998, the
issue of housing finance has gained
prominence in the minds of prospective
home-owners. Although the Chinese
government urged state-owned banks in
1998 to expand mortgage lending, the main
policy tool to enhance housing affordability
for urban residents has been the Housing
Provident Fund (HPF). This system was first
introduced in Shanghai in 1991, soon
followed by other cities, and was then
propagated as national policy from 1994.
Since the Housing Provident Fund1system
has been in force over ten years in Chinese
cities, it seems appropriate to evaluate the
progress and limitations of this policy. In
addition, there have been few detailed
studies on the functioning of the Housing
Provident Fund. This article aims to remedy
that shortcoming2.
The article begins with a study of income
levels in urban China and it compares
people’s earnings with prices on the
housing market. While Chinese statistics are
problematic with regard to their accuracy
and reliability, we can safely conclude that
there is a gap between people’s earnings
and the money needed to buy an
apartment. The second section traces the
beginning and evolution of the Housing
Provident Fund policy, and it explains its
basic principles. As shown in this section,
the policy aims of the HPF have shifted over
time. While the initial focus lay on financing
housing construction, in 1998 these policy
loans were scaled down and were replaced
by an emphasis on loans for households.
The third section evaluates to what extent
city-level HPFs have attained the goal to
secure monetary contributions from
enterprises and workers. It is shown that
fund collection efforts have been relatively
successful, but that implementation is
uneven across regions and vis-à-vis
different enterprises. Problems of fund
collection are linked to legacies of the
planned economy but are also due to
obstacles arising from an economy in
transition. The fourth section examines the
usage of HPF funds with a special focus on
individual housing loans. It shows that fund
usage has been limited and that only a small
share of the contributors has benefited from
the HPF system. The final section examines
the social equity aspects of the Housing
Provident Fund. It is argued that this
employment- and income-based system of
housing finance may have discriminatory
effects on low-income earners, temporary
staff, employees in non-state enterprises
and in the informal economy, and laid-off
workers.
Earnings and household purchasing
power
The importance of wage levels for people’s
well-being is a new phenomenon in China.
During the planned economy – 1949 to 1978
– money and wages had a less prominent
position in people’s minds and in their daily
lives. To begin with, it was an economy of
constant shortages, and available goods
and services were mainly allocated through
administrative methods rather than money
transactions (Naughton 1996: 26-31). Thus,
the wages of urban workers and staff were
kept low, and occasional wage adjustments
were the result of bureaucratic decisions.
However, for workers in state-owned and
collective enterprises, and for staff in
government agencies, their work units
(danwei) were major providers of welfare
benefits, goods and services3. These
welfare benefits – combined with job
CHINA’S HOUSING PROVIDENT FUND
China’s Housing Provident Fund:
Its Success and Limitations
Mattias Burell (e-mail: mattias.burell@ibf.uu.se)
Department of Government, Uppsala University, Sweden
1Two notable exceptions are Lee (2000), and Wang (2001).
2This study is based on data collection during three field trips in China – winter 2003, summer 2004, and autumn 2004. Interviews were conducted with
government officials at the Ministry of Construction, and with officials at Housing Reform Offices and HPF Management Centers in Tianjin, Beijing, Jiangxi
and Guangdong provinces. Other sources of information consist of published and unpublished (internal) materials provided to me during the interviews,
Chinese newspaper items, and secondary sources.
3Workers in state-owned enterprises (SOEs) could be expected to be taken care of from cradle to grave. Large factories – with maybe 50.000 – 60.000
workers – would have their own day-care centers for children, schools, hospitals, banks, post offices, transportation services, and subsidized food supplies.
38

HOUSING FINANCE INTERNATIONAL – March 2006
security – compensated for the low wage
levels, and lessened the importance of
money. Chinese enterprises were not simply
production units but also welfare providers
and institutions of social control (Walder
1986: 28-29).
As part of the urban welfare package in the
planned economy, public sector housing
was provided virtually free of charge to the
employees. Rents were kept artificially low
and represented less than 1-2% of a
worker’s monthly wage (Interviews BJ 2003,
BJ 2004a). As a consequence, China’s
public housing system became seriously
under-financed, and this was a key factor
behind the housing reforms which started in
1980. The main orientation of China’s
housing policy over the last twenty years
has been to move away from the traditional
system of welfare allocation (fuli fenpei), or
material allocation, toward a system of
monetized allocation (huobi fenpei) of
housing benefits. The basic aim of this
policy – articulated already in the 1980s and
stressed with even more emphasis during
the 1990s – is that urban residents should
spend a larger share of their incomes on
housing consumption. This policy goal
would either be attained by mandatory rent
increases or by encouraging people to buy
apartments and become home owners
(Interviews BJ 2003, BJ 2004c)
By the end of the 1990s, the housing
system in China had dramatically changed.
Due to a massive construction boom in the
1980s, the previous problems of inadequate
living space had been resolved for many
urban residents (Wang and Murie 1999:
103-113). In addition, a large share of public
housing had been sold off to sitting tenants.
Many of them had bought their rental
apartments at discounted prices from their
work units, sometimes at very low prices
such as 100-200 yuan per square meter. In
Beijing, 75 per cent of public housing had
been privatized by the end of 2004 and the
situation was similar in other cities
(Interview MoC 2004b, BJ 2005a, Wang
2003). In addition, a significant share of
newly constructed housing had been
purchased by work units and re-sold at
discounted prices to their employees until
this practice was banned in 1998 (Zhang
2000: 347, Interview MoC 2004b).4
These changes in housing policy –
privatization of public housing, a shrinking
rented sector, and the monetization of
housing benefits – placed new entrants on
the housing market in a vulnerable position.
They mainly had to rely on their own
earnings and try to save enough money to
purchase a home. Although salary levels in
Chinese industry increased rapidly in the
1990s and tangibly improved living
conditions (Burell 2001: 210), the inflation
rate remained high and limited people’s
purchasing power. The sale prices of new
housing also kept rising in this period, due
to higher construction costs. The price for
commercial housing in urban China was set
between 500 and 2,000 yuan per square
meter in 1992 (Wu 1996: 1614), but in the
late 1990s most large cities had reached the
level of 3,000 yuan per square meter.
Naturally, house prices in very large cities
and economic centers as such Beijing,
Shanghai or Guangzhou tend to be higher
and rise more rapidly than in other cities.
Since economic development, price levels
and people’s incomes vary across China,
the Ministry of Construction (MoC)
suggested a set of regional guidelines for
the sale of commercial housing in different
parts of the country (Table 1). As shown in
the table, house prices vary considerably in
line with regional characteristics and
housing standards. As a basic rule,
however, medium-standard housing in most
regions was sold at 2,500 to 4,000 yuan per
square meter in 1998.
CHINA’S HOUSING PROVIDENT FUND
4In March 1998, the State Council announced that all welfare allocation of housing from work units to their employees would be prohibited from mid-1998
and then quickly phased out (Beijing Review 1998).
Sources: FBIS-CHI-98-147 (Xinhua News Agency, 27 May 1998), FBIS-CHI-98-348 (Xinhua News Agency, 14 December 1998), and CNA (1998: 6).
Note: According to the MoC announcement in 1998, China’s cities are divided in four categories according to their levels of economic development
and the income of local residents.
Table 1. Regional housing prices suggested by the Ministry of Construction (1998)
Category of city Low housing Medium housing High housing
Prices Prices Prices
1,500-2,000 above 4,000 8,000-10,000
yuan per m2yuan per m2yuan per m2
from 1,200 to 8,000
yuan per m2yuan per m2
1. The first group consists of Beijing, Shanghai, Guangzhou
and Shenzhen: the four main economic centers.
2. The second group is composed of regional economic
centers, i.e. Tianjin, Hangzhou, Chongqing, Dalian, Wuhan
from 800 to 4,000
yuan per m2yuan per m2
3. The third group are other provincial capitals such as
Chengdu, Changsha and Lanzhou.
from 800 to 1,200
yuan per m2yuan per m2
4. The fourth group is composed of medium-size and small
cities at the prefectural level and below.
39

HOUSING FINANCE INTERNATIONAL – March 2006
In China, a distinction is made between the
construction space (jianzhu mianji) and the
living space (shenghuo mianji) of an
apartment. When buying an apartment you
have to pay for the construction space, but
the actual living area is about 20 per cent
smaller (Interview GZ 2003). For a couple to
have a decent living space, they need to
buy an apartment of about 90-100 square
meters. That will give them a living space of
70-80 square meters. Assuming that the
sale price is 3000 yuan per square meter,
the Chinese couple has to deliver a down
payment of 60,000 yuan (20% of the total)
and pay 300,000 yuan in total for this
apartment. In view of current wage levels in
China, is this price level affordable for most
urban residents? This question is not so
easy to answer due to the mediocre quality
of Chinese official statistics. These statistics
suffer from problems of inaccuracy and
under-reporting (Banister 2005), inadequate
attention to people working in the informal
economic sector, and the fact that people
may have additional sources of income.
Taking these statistical problems into
account, we can still make a few general
conclusions about China’s housing
affordability by taking a look at the annual
earnings of urban employees (see Table 2).
CHINA’S HOUSING PROVIDENT FUND
Table 2. Urban manufacturing earnings (2002) –
by province. Average annual per capita and household earnings and their housing purchasing power.
Data source: SSB 2003: 171-179. Author’s calculations. Note: The down payment of 60.000 yuan equals 20% of the full price of the apartment, i.e. 300,000
yuan. 1 yuan equals 0.125 USD. We make the assumption that a couple can set aside a maximum of 30% of their earnings to save for the down
payment and house purchase. In fact, savings ratios will be lower for low-income couples, and higher for high-income couples.
Provinces Individual
average annual
earnings
(yuan)
Couple
average annual
household
earnings
(yuan)
Time needed for
a 60,000 yuan
down payment.
Entire HH
earnings
(years)
Time needed
for a 60,000
yuan down
payment. 30%
of HH earnings
(years)
Time needed to
purchase a
300,000 yuan
apt. Entire HH
earnings
(years)
Time needed to
purchase a
300,000 yuan
apt. 30% of HH
earnings
(years)
Henan 7,795 15,590 3.8 12.6 19.2 64.0
Shanxi 7,892 15,784 3.8 12.6 19.0 63.3
I. Mongolia 8,135 16,270 3.6 12.0 18.4 61.3
Jiangxi 8,261 16,522 3.6 12.0 18.1 60.3
Anhui 8,326 16,652 3.6 12.0 18.0 60.0
Heilongjiang 8,755 17,510 3.4 11.3 17.1 57.0
Hebei 8,810 17,620 3.4 11.3 17.0 56.7
Hubei 8,876 17,752 3.4 11.3 16.8 56.0
Shandong 8,898 17,796 3.4 11.3 16.8 56.0
Guizhou 9,279 18.558 3.2 10.7 16.2 54.0
Shaanxi 9,340 18,680 3.2 10.7 16.0 53.3
Ningxia 9,503 19,006 3.1 10.3 15.8 52.7
Hainan 9,672 19,344 3.1 10.3 15.5 51.7
Sichuan 9,816 19,632 3.0 10.0 15.3 51.0
Hunan 9,825 19,650 3.0 10.0 15.3 51.0
Guangxi 9,860 19,720 3.0 10.0 15.2 50.7
Gansu 10,047 20,094 3.0 10.0 14.9 49.7
Chongqing 10,202 20,404 2.9 9.7 14.7 49.0
Xinjiang 10,231 20,462 2.9 9.7 14.7 49.0
Jilin 10,231 20,462 2.9 9.7 14.7 49.0
Tibet 10,258 20,516 2.9 9.7 14.6 48.7
Liaoning 10,553 21,106 2.8 9.3 14.2 47.3
Qinghai 10,717 21,434 2.8 9.3 14.0 46.7
Fujian 11,627 23,254 2.6 8.7 12.9 43.0
Jiangsu 11,731 23,462 2.5 8.3 12.9 42.7
Yunnan 11,752 23,504 2.5 8.3 12.7 42.3
Zhejiang 13,435 26,870 2.2 7.3 11.1 37.0
Tianjin 14,781 29,562 2.0 6.7 10.1 33.7
Guangdong 14,958 29,916 2.0 6.7 10.0 33.3
Beijing 18,157 36,314 1.6 5.3 8.3 27.7
Shanghai 21,957 43,914 1.4 4.7 6.8 22.7
National 11,152 22,304 2.7 9.0 13.4 44.7
40

HOUSING FINANCE INTERNATIONAL – March 2006
As shown in Table 2, there are large regional
variations in annual per capita earnings. A
working couple in Shanghai earn almost
three times as much as a couple in Henan
province. These wage differences have
repercussions for their ability to save
enough money for a down payment or pay
the full price of an apartment. A working
couple in Shanghai, where wages are high,
will reach the down payment target of
60,000 yuan in 1.4 years if they devote all
their earnings on this. But people cannot
spend all their incomes on housing, so with
a savings ratio of 30% it takes them almost
five years to reach that sum. For a full
payment of the apartment, the Shanghai
couple needs to set aside 30% of their
earnings for a period of nearly 23 years. It
should immediately be clarified that
although wage levels in Shanghai are
reasonably high, housing costs are higher
than in other provinces, and 300,000 yuan
does not allow you to buy a large
apartment.
At the other extreme, we see that a working
couple in Henan province needs almost four
years of their full earnings to reach the down
payment target of 60,000 yuan, and it will
take them over twelve years with a savings
ratio of 30%. If they set aside one-third of
their combined earnings for a period of 64
years, they will be able to pay off the
300,000 yuan apartment. To be sure,
housing prices are lower in Henan province,
but in provincial capitals and all major cities
they are not lower than 3,000 yuan per
square meter. The calculation is thus
indicative of the relationship between
people’s incomes and the cost of becoming
a home owner. It is demonstrated in Table 2
that there the cost of new housing – without
subsidies or mortgage finance – represents
a heavy burden on the household budget for
employees in the manufacturing industry.
However, a few cities and provinces form a
special group, due to their higher wages.
Manufacturing workers in Beijing, Shanghai
and Tianjin, and in Guangdong province,
have annual earnings between 15,000 and
22,000 yuan. This wage level seems to give
slightly better conditions to save for a down
payment and buy an apartment. We may
ask ourselves if manufacturing workers in
China are unusually badly paid compared to
other occupational groups, but that is not
the case. Factory workers are in a medium
position of the wage hierarchy (see Table 3).
CHINA’S HOUSING PROVIDENT FUND
Table 3. Annual average per capita earnings (2002) –
by occupational sector – in Beijing, Tianjin, and Jiangxi province
Sources: SSB 2003: 173-225. Author’s calculations and rank-ordering of occupational wage levels. Note: 1 yuan equals 0.125 USD. Employee “earnings”
(baochou) include basic wages, overtime and supplemental salary, bonuses, allowances, benefits, and social insurance payments. (Cf. Banister
2005: 24-27 for a review of the concept of earnings in Chinese official statistics).
Occupation Beijing Tianjin Jiangxi
1. Retail trade 15,620 ** 10,202 ** 5,495
2. Highway transport 15,333 10,934 8,742
3. Hotels 18,524 11,172 7,811
4. Catering services ** 12,611 11,512 5,747
5. Civil engineering 14,215 14,561 7,697
6. Manufacturing 18,157 14,781 8,261
7. Construction 14,497 15,190 8,153
8. Wholesale trade (food, beverages, etc.) 18,963 15,219 6,419
9. Education 24,447 16,799 9,752
10. Public institutions 25,761 16,928 9,870
11. Storage services 16,440 17,479 7,795
12. Railway transport 18,378 18,206 16,564
13. Government and Party agencies 26,291 18,209 10,493
14. Healthcare institutions * 30,923 18,668 11,205
15. Post and telecommunications 26,341 21,892 ** 12,760
16. Supply of gas, electricity and water 28,007 * 22,093 10,724
Average annual wage 21,861 16,223 9,151
** Lowest annual wage 12,611 10,202 5,495
* Highest annual wage 30,923 22,093 12,760
41

HOUSING FINANCE INTERNATIONAL – March 2006
Table 3 shows that wage levels in Beijing are
higher than in Tianjin and in Jiangxi province
for nearly all occupational groups. This is a
reflection of the relative degree of economic
development and the living costs in these
locations. It also signals their levels of
political importance. Jiangxi is at the lowest
rank of the ladder and quite a distance from
both Beijing and Tianjin. But within each
locality, the wage hierarchy is similar, with
the highest earnings directed towards
employees in government agencies, public
institutions, healthcare, and state
companies maintaining a monopoly
position. By contrast, people working in
retail trade, hotels, catering services,
transportation and manufacturing have low
or medium-level earnings. Such
occupations tend to be closer to the private
sector and more exposed to market forces.
In view of their lower wages, these workers
will find it more difficult to make the savings
needed for home ownership.
Several scholars have noted the problem of
housing affordability in China. This issue
became more salient after the restructuring
of state industry, in the 1990s, which left
many lay-offs and widespread job insecurity
in its wake. In other words, it is not only
people’s propensity to save for home
ownership, but also their ability to find
stable employment – and income security –
in a changing economy which pose a
challenge to housing reforms. The key
problem, however, is the discrepancy
between people’s incomes and housing
prices.
The biggest hurdle to increasing home
ownership among urban workers is
affordability […] In most countries,
homes priced at three to five times
annual household income are considered
affordable, while housing prices in
China are often 10-20 times family
income. Chinese reformers hope to
achieve a 4:1 ratio between housing
prices and family income, but reaching
this goal appears to be a long way off
(Rosen and Ross 2000: 79).
Development of Housing Provident
Funds
To bridge the gap between people’s
incomes and the price for housing, the
Chinese government has introduced a
series of policies aimed at both the supply
side and the demand side of housing
provision. On the supply side, there have
been attempts to bring down housing
construction costs. Low-cost housing
projects were planned and implemented
during the 1990s under the “comfortable
housing” (anju) project. This project involved
policy loans by the central government,
matched by tax exemptions and subsidized
allocation of land by local authorities and
profit caps imposed on real estate
developers. This policy was slightly
adjusted in 1998 and continued under the
label “economic housing” (jingji shiyong
fang). The key idea was to produce low-cost
housing for sale to medium-income
households. Sale prices of these
apartments were set lower than open
market prices to make them affordable to
households with limited earnings (Interview
MoC 2004b, Wang and Murie 2000, Wang
2001: 631-32, CNA 1994, CNA 1996: 7,
Rosen and Ross 2000: 80-81).
The Housing Provident Fund (HPF) system
has been an important supplement to low-
cost housing, and it is targeted to improve
the demand side of housing consumption. It
aims to enhance people’s housing
purchasing power through a system of joint
savings – with mandatory contributions
from employees and work units – and
placement of the funds into individual
accounts. The savings in these HPF
accounts allow workers to apply for low-
interest housing loans (Interviews BJ 2004c,
TJ 2004).
Shanghai was the first city to introduce a
Housing Provident Fund on a large scale. It
was part of a larger housing reform package
adopted by Shanghai’s government, and it
was endorsed by the State Council in May
1991. This policy innovation was quickly
announced in news media and many
provincial and city governments sent
delegations to Shanghai for inspection
visits. Similar HPF schemes were
established in Beijing, Guangzhou and
Tianjin in 1992, and soon other cities across
China followed suit. In 1993, the third
national conference on housing reform was
held in Beijing. In its wake, the State
Council’s Leadership Group on Housing
Reform issued the Decision on Deepening
the Urban Housing Reform, and this policy
document urged major cities to set up
Housing Provident Funds. This was the
national government’s effort to
institutionalize the HPF system across
China (Interviews BJ 2004a, MoC 2004b, TJ
2004, GZ 2004a, Lee 2000: 65, Wang 2004:
90). As a consequence, nearly 200 cities
introduced HPF schemes in the 1990s, and
in 2004 almost all medium-sized cities had
adopted this policy (Wang and Murie 1999:
160, Interview MoC 2004b).
Eventually, after the State Council’s housing
reform decision in 1998, it was decided the
HPF system needed a solid legal basis. This
policy was now reaching the end of its
experimentation phase and needed to be
consolidated. Hence, in March 1999, the
State Council issued the Housing Provident
Fund Management Provisions, as a legal
tool to standardize HPF decision-making
procedures and fund management. From
this point onward, all cities (above county
level) were required to set up HPF schemes
and it was stipulated that all enterprises
(including private firms, joint ventures and
private enterprises), government agencies,
public institutions and social organizations
with employees on their payrolls had to take
part in the HPF system (State Council 1999).
Subsequently, in March 2002, the Amended
Housing Provident Fund Management
Provisions were issued by the State Council.
The amended text extended the policy
scope to include township enterprises, and
it clarified the legal clauses on fund usage,
financial auditing and supervision, and
penalties that would be imposed on
enterprises, agencies or individuals who
violated these regulations (SC 2002).
Apart from this effort of legal
standardization, the State Council made
another major policy shift in 1998-99. From
that point onward, HPF funds could no
longer be used as policy loans for house
construction, but should be exclusively
CHINA’S HOUSING PROVIDENT FUND
42

HOUSING FINANCE INTERNATIONAL – March 2006
used to finance individual loans for home
ownership (Lee 2000, Interview BJ 2004a,
MoC 2004b). It can therefore be said that
although the HPF system was first
established in 1991-92, it is only in recent
years (1998-2005) that it has been re-
oriented to grant house loans to Chinese
households. In the following sections, I
examine how HPF schemes have operated
– with respect to fund collection and loan
allocation – in a few cities.
Fund collection
The collection of HPF contributions is
supervised by local HPF management
centers which operate under the authority of
Housing Committees appointed by the city
government. The Housing Committees set
HPF contribution ratios – in light of local
economic conditions – and approve the
annual plans for collection and usage of
HPF funds. They also perform year-end
reviews of the execution of these plans,
along with financial auditing by Finance
Bureaus. The HPF management center is
responsible for all routine operations and is
directly involved in collection, financial
supervision, and loan management of HPF
funds (SC 2002, Wang 2001: 635, Interview
TJ 2004).
CHINA’S HOUSING PROVIDENT FUND
Sources: Interview in China (NC 2004) and written materials provided during this interview. NSB 2004: 31. Note: Author’s calculations. Figures have been
rounded to the nearest million yuan. These statistics exclude households and HPF loans for staff in province-level units and the railway industry,
which are handled separately. N.d. = no data.
Table 4. Beijing City Housing Provident Fund (2004)
Table 5. Nanchang City Housing Provident Fund (2004)
Year Contributing
work units
Workers and
staff covered
Accumulated
collection of HPF
funds (million yuan)
Households
granted HPF
loans
Annual grants of
individual HPF loans
(million yuan)
1998 9,602 1,928,000 6,780 7,710 631
1999 n.d. 2,224,000 10,590 8,077 1,150
2000 n.d. 2,316,000 15,330 20,657 3,900
2001 n.d. 2,324,000 21,350 39,520 7,460
2002 n.d. 2,344,000 28,650 29,517 5,560
2003 n.d. 2,344,000 37,420 25,726 5,210
2004 n.d. 2,589,000 45,400 32,420 7,230
Total:164,999 31,300
Year Contributing
work units
Workers and
staff covered
Accumulated
collection of HPF
funds (million yuan)
Households
granted HPF
loans
Annual grants of
individual HPF loans
(million yuan)
1998 n.d. 110,000 90 388 3
1999 n.d. n.d. 235 1,144 27
2000 n.d. n.d. 431 1,300 45
2001 n.d. n.d. 661 1,496 90
2002 n.d. n.d. 930 1,989 154
2003 3,100 300,000 1,289 3,515 459
2004 3,100 330,000 1,689 3,953 530
Total:13,785 1,308
Sources: Interviews in China (BJ 2004c, BJ 2005b) and written materials provided during the interviews. Wang (2001: 637). Note: Author’s calculations. N.d.
= no data.
43

HOUSING FINANCE INTERNATIONAL – March 2006
How successful have HPF management
centers been in building up HPF funds
based on regular monetary contributions
from enterprises and individuals? The
answer to this question depends on our
standards of evaluation. Collection of HPF
funds – and their rate of increase – depend
on stipulated contribution rates, the number
and size of enterprises and other work units
in the locality, the wage levels and financial
strength of these work units, their
willingness to comply with HPF policy, and
the diligence of the local HPF centers in
making sure that their regulations are
universally enforced.
Tables 4 and 5 demonstrate two extreme
cases of HPF fund collection in two
localities. On the one hand, there is Beijing,
which started in 1992 with a contribution
rate of 5% of the employee’s monthly salary
and matched by another 5% by his/her
employer. The contribution rate was
gradually raised to 8% in 1998 and further
to 10% in 2003. By 2004, about 50-60% of
Beijing’s 4.8 million employees were
covered by the HPF policy (Interview BJ
2004c, MoC 2004b, SSB 2003: 173). In
combination with fairly high wage levels in
Beijing, this generated 5-8 billion yuan of
funds per year, and an accumulated 45
billion yuan of HPF funds in 2004 (see Table
4). In this respect, we can say that Beijing’s
HPF management center successfully
carried out its mandate.
By contrast, Nanchang is the capital of
Jiangxi province, a region which has
lingered in the backwaters of economic
reforms and suffers from a lack of foreign
investment. Wage levels in both the public
and private sectors are among the lowest in
China (cf. Tables 2 and 3). According to
interviews with local officials, there are
some 500,000 workers and staff in
Nanchang city, but only 330,000 of these
employees are covered by the HPF scheme
(66%). In 2004, there were 3,100 work units
which had signed up for the HPF scheme,
but only 2,200 of these enterprises and
agencies made regular payments.
Trustworthy contributors were mostly
government agencies and public
institutions. In fact, there is an estimated
total of 6,000 work units in Nanchang, if all
unregistered small non-state enterprises are
taken into account, and the majority of
these firms did not even bother to sign up
for the HPF scheme. In the experience of
Nanchang HPF officials, it was very
troublesome, and hardly worthwhile, to go
after private enterprises, foreign enterprises
or joint-ventures and ask them to comply
with HPF regulations (Interview NC 2004).
As a result of these adverse factors, the
annual collection of HPF funds was modest,
ranging between 200 and 400 million yuan,
and in 2004, the total HPF collection in
Nanchang had reached about 1.7 billion
yuan.
While both Beijing and Nanchang are
extreme cases, at opposing poles of a
spectrum, the general situation in other
parts of China gives a mixed picture. The
contribution rates to HPF funds varied
greatly throughout the 1990s, and in some
places it was as low as 1-2% of local
employees’ monthly salaries. Eventually,
this was standardized to a minimum ratio of
5% of the enterprise wage sum, thanks to
the State Council’s Provisions on HPF
Management in 1999 (Interview MoC 2004b,
SC 1999). Another problem has been the
slow inclusion of all enterprise forms into
the HPF scheme, both in a regulatory sense
and in actual practice. In cities where there
is a strong presence of joint-ventures and
foreign firms that employ temporary
workers – HPF centers are reluctant to
enforce policy in these enterprises, and
content themselves to impose it on public
institutions, state enterprises, and
government agencies (Interviews GZ 2004a,
2004b, 2004c, YN 2005). In this situation,
the HPF policy coverage becomes very
limited as compared to the number of
enterprises and workers that de facto are
located in the city.
HPF participation rates seem to be highest
in large cities, such as Beijing, Tianjin and
Shanghai (98%), and in some cities located
in wealthy provinces such as Jiangsu and
Zhejiang (90%), but in most cities the policy
coverage is less than 50% (Zhang 2000:
343). According to the Ministry of
Construction, which supervises the HPF
system at the national level, there were 80
million workers and staff registered in 2004
but, due to the financial difficulties of many
firms, only about 60 million workers made
regular contributions to the HPF funds. The
remaining 20 million employees were
labelled as “intermittent” or “erratic”
contributors. And, of course, laid-off staff
simply drop out of the system (Interview
MoC 2004b). It should also be mentioned
that temporary or migrant laborers –
numbering about 150 to 200 million persons
in urban China – are not included in the
system. According to current regulations,
only those workers who sign an
employment contract for one year or longer
are required (entitled) to take part in the HPF
scheme. If the narrow definition is used – ie
on-post urban workers and staff – this
amounts to a total of about 100 million
people. So the overall participation rate in
the HPF scheme in urban China is either
60% or 80%, depending on how you count
(Interview MoC 2004b). A less narrow
definition of “urban employees” would
probably generate a figure of 30-40%,
which is much less impressive.
Keeping these shortcomings in mind, the
collection of HPF funds in recent years has
been fairly impressive. The four cities
examined here – Beijing, Guangzhou,
Tianjin and Nanchang – managed to steadily
increase HPF fund collection over the years.
By 2004, Beijing had accumulated 45 billion
yuan, followed by Guangzhou and Tianjin
with 29 and 24 billion yuan, respectively.
Nanchang, with its less developed economy
smaller population, and seemingly weak
enforcement, trailed far behind with only 1.7
billion yuan (Table 6).
This mixed picture is supported by other
sources of information. According to the
Ministry of Construction, collection of HPF
funds varies considerably between cities.
Large economic centers such as Shanghai
and Beijing are able to raise 10-30 billion
yuan in HPF funds per year, whereas
medium-sized cities can collect 3-4 billion
yuan, and small cities only can raise 20-40
million yuan (Interview BJ 2004b). This may
be seen as a reflection of the uneven
economic development in China today.
While there is no national-level Housing
Provident Fund, the Ministry of Construction
is in charge of aggregate HPF statistics. For
CHINA’S HOUSING PROVIDENT FUND
44

HOUSING FINANCE INTERNATIONAL – March 2006
China as a whole, the accumulated HPF
amount rose from 11 billion yuan in 1994 to
630 billion yuan in 2004. A crucial question,
however, is how these funds have been
used.
Fund usage and extension of housing
loans
At its inception in 1991-92, the Housing
Provident Fund was mainly seen as a
vehicle to raise funds for housing
construction, and for several years large
amounts of local HPF funds were used as
policy loans for work units and real estate
developers. From 1994 to 1998, the HPF
scheme was used as a financial tool in
producing affordable housing under the
state-sanctioned anju-project (Interview
MoC 2004b). During the 1990s, there were
also conflicts over the use and authority
over HPF funds. Several bureaucratic actors
– including Construction Bureaux, Finance
Bureaux, and banks – claimed that they
were most suitable institutions to handle
HPF funds. Naturally, they were all acting in
self-interest but tried to couch their
arguments in terms of economic efficiency
and the common good (Interview BJ
2004b). Finally, these bureaucratic conflicts
were resolved by the State Council and,
starting in 1998, HPF funds were to be
exclusively used as housing loans to
individual households (SC 1999).
CHINA’S HOUSING PROVIDENT FUND
Sources: Interviews in China (MoC 2004b, BJ 2004c, BJ 2005b, GZ 2004a, NC 2004, TJ 2004) and written materials provided during these interviews. Other
sources: Gateway (2006), FBIS-CHI-1999-0823, FBIS-CHI-2000-0223, CNA (1996: 8), THR (2002). Note: Author’s calculations.
Sources: IInterviews in China (BJ 2004c, BJ 2005b, GZ 2004a, NC 2004, TJ 2004) and written materials provided during these interviews. Note: Author’s
calculations. Figures have been rounded to the nearest million yuan. (a) The total loan amount for households in Tianjin was provided during the
interview (TJ 2004). N.d. = no data.
Table 6. City comparison of accumulated HPF funds (million yuan)
Table 7. Annual HPF housing loans (million yuan) to individual households
BeijingYear Guangzhou Tianjin Nanchang National
1992 - 88 270 - -
1993 - 215 520 - -
1994 - 380 770 - 11,000
1995 - 714 1,230 - 11,000
1996 - 1,276 2,090 - 39,300
1997 - 2,059 3,390 - 79,900
1998 6,780 3,362 5,110 90 123,100
1999 10,590 4,889 7,090 235 140,900
2000 15,330 6,086 9,570 431 140,900
2001 21,350 - 12,640 661 332,600
2002 28,650 - 16,560 930 413,100
2003 37,420 22,297 - 1,289 556,300
2004 45,400 28,800 24,000 1,689 630,000
BeijingYear Guangzhou Tianjin (a) Nanchang
1992 - - > 1 -
1993 - - 2 -
1994 - - 6 -
1995 - - 19 -
1996 - - 101 -
1997 - - 497 -
1998 631 - 1,330 -
1999 1,150 414 1,220 3
2000 3,900 636 2,240 27
2001 7,460 1,160 2,000 45
2002 5,560 1,524 2,420 90
2003 5,210 2,422 n.d. 154
2004 7,230 2,556 n.d. 459
Total (million yuan)31,300 8,712 17,000 530
Total households 164,999 58,714 150,000 13,785
45

HOUSING FINANCE INTERNATIONAL – March 2006
According to the State Council’s Provisions
on HPF Management in 1999 (amended
2002), the HPF funds should be kept in
individual accounts – one for each
employee – in a designated local bank, and
can only be withdrawn for certain
purposes5, subject to approval by the HPF
management center. Since 1998, HPF
management centers have focused on the
task of granting loans to individual
households. The maximum size of the loan
depends on the sum deposited in an
individual’s HPF account. In Beijing, it is
possible to borrow a sum that is ten times
the deposited amount. Hence, if a person
holds 10,000 yuan in her account, she can
borrow 100,000 yuan. For individuals with
large deposits in their accounts, however,
the maximum limit was set at 300,000 yuan.
Similar limits on HPF loans were set in the
other localities: 200,000 yuan in Tianjin,
200,000 yuan in Nanchang, and 250,000
yuan in Guangzhou (Interview BJ 2004c, TJ
2004, GZ 2004a, NC 2004).
Table 7 illustrates the progress of individual
housing loans in these four cities. Tianjin
started to grant individual loans already in
1992, but on a modest scale. The real shift,
however, took place in 1998, and since then
HPF loans have risen quickly. By the end of
2004, Beijing had granted over 31 billion
yuan in HPF loans to almost 165,000
households. Guangzhou and Tianjin had
granted 9 and 17 billion yuan respectively,
and Nanchang had granted 530 million
yuan. It should be noted, however, that
these amounts were small in relation to the
total assets kept in the HPF system. In both
Nanchang and Guangzhou, HPF loans
represented only 30% of their local HPF
assets. In 2001, vice-minister of
construction, Liu Zhifeng, analyzed the
problem of idle HPF funds. He noted that
the national HPF loan ratio was 30% of the
planned figure and this was deeply
unsatisfactory. Some provinces had not
even begun to grant HPF loans. Other
provinces had only attained 5-10% of the
planned figures (Liu 2001).
CHINA’S HOUSING PROVIDENT FUND
5The State Council (2002) revised Provisions on HPF Management, Art 24, state that legitimate withdrawals from HPF accounts are those used for (1)
Purchasing, building, renovating, or repairing residences; (2) Leaving work posts due to old age or retiring from work; (3) Physically unsuitable for work with
employer-employee relations terminated; (4) Resettling abroad; (5) Repaying the principal or interest of housing loans; (6) Rent exceeding the prescribed
proportion of wages.
Sources: Interviews in China (MoC 2004b, BJ 2004c, BJ 2005b, GZ 2004a, NC 2004, TJ 2004) and written materials provided during these interviews. Note:
Author’s calculations, with aggregate loan amounts divided by the number of loan beneficiaries.
Table 8. Average level of HPF housing loans (yuan)
BeijingYear Guangzhou Tianjin Nanchang National
1998 - - - - -
1999 - - - - -
2000 188,797 86,000 - 34,615 -
2001 188,765 122,000 82,304 67,160 -
2002 188,366 137,000 93,245 77,425 65,838
2003 202,518 174,000 - 130,583 71,688
2004 223,010 191,000 - 134,075 -
We may also wonder if the HPF loans are
really sufficient to bridge the gap between
individual incomes and housing prices. If we
take 300,000 yuan as the standard price for
a medium-sized apartment in large cities, it
seems as though HPF loans are at least
partly helpful in facilitating home ownership
(see Table 8). The average size of HPF loans
in Beijing has hovered around 200,000 yuan,
and in Guangzhou they have risen from less
than 100,000 yuan to almost 200,000 yuan
in recent years. In Tianjin and Nanchang the
size of HPF loans is lower, but this may be
counterbalanced by slightly lower housing
prices in these two cities. Nevertheless, it is
obvious that prospective home buyers have
to bring up additional funds – from 100,000
to 150,000 yuan – in order to purchase an
apartment. That sum must be based either
on individual savings, informal loans from
friends and relatives, or commercial bank
loans. It goes without saying that this will be
a significant stumbling-block for low- or
even medium-income households in China.
Social equity dilemmas
Scholars have noted certain biases
incorporated into China’s housing policies,
and the HPF scheme is no exception. Zhou
(2004: 8) argues that urban housing policies
only are targeted at certain social groups,
and that they have tangible discriminatory
effects on other groups. The intended
beneficiaries of the HPF scheme, for
example, are only persons with stable and
full-time employment – mainly those
working in public sector work units and in
financially strong enterprises. Other groups
are clearly excluded, ie, laid-off workers,
people with temporary employment, the
self-employed, and migrant workers who
live and work in a semi-legal status in
Chinese cities.
Another equity issue concerns persons who
actually are included into the HPF system,
and who contribute to its maintenance and
growth, but for whom the benefits are low or
uncertain. These are mainly workers and
staff with low incomes – usually below
15,000 to 20,000 yuan per year – and for
whom HPF individual savings accumulate
very slowly. For them, it will take a very long
time to qualify for a down payment and
become eligible for a sufficiently large HPF
loan to finance a home purchase. In fact,
46

W
hen we examine how many of HPF
contributors who actually benefit – in the
sense that they are granted HPF loans – the
results are striking and disappointing. In the
four cities in our sample, less than 7% of the
contributors became beneficiaries of this
scheme. Admittedly, the share of people who
can benefit has increased slowly, from 0.4%
to 6.4% in Beijing between 1998 and 2004,
but the figure is still disappointingly low.
Similar patterns can be seen in Guangzhou,
Tianjin and Nanchang (Table 9). For China as
a whole, the average figure is 5.4%.
If so many people contribute to the Housing
Provident Fund, why are the beneficiaries so
few and who are they? According to Wang
(2001: 642), a large number of low-income
employees make small contributions to the
HPF system but their individual deposits are
too low to make them eligible for loans. By
contrast, a small percentage of high-income
earners make larger deposits and can then
draw on the vast resources of the HPF
system to apply for – and obtain – low-
interest housing loans. If this analysis is
correct it entails a re-distribution of
resources, but in the wrong direction, ie,
from low- to high-income households. This,
however, can only be a preliminary
conclusion and must be analyzed on the
basis of more detailed data on HPF lending
practices and the economic characteristics
of HPF loan beneficiaries.
Conclusions
The affordability gap between people’s
incomes and housing prices is clearly seen
in China. Wage increases for the majority of
urban households have not kept up with the
rise in housing costs. Even though the rate
of home ownership has reached 70-80% in
most cities, this is not the result of people’s
purchasing power, because it is mainly a
legacy of the privatization of former public
housing during the 1990s. Due to a rapid
shrinkage of the rental sector, new entrants
on the housing market have few options but
to make savings for home ownership in the
newly constructed housing sector. If they
are lucky, they may be eligible to buy
“economic housing” which is sold at slightly
lower rates. Most people, however, must
expect to pay 3,000 yuan per square meter.
For people with low- and medium-level
incomes these housing prices are too
expensive. In order to alleviate these
problems, the Chinese government
introduced the HPF system in 1991, and in
1998 it had been established in practically
all cities. In addition, State Council
decisions in 1999 and 2002 ensured that
HPF funds would exclusively be used to
promote home ownership among individual
households.
While local Housing Provident Funds have
been successful in accumulating funds, the
policy coverage among enterprises and
employees remains a problematic issue.
Many private and foreign enterprises evade
the HPF contributions – and financially weak
enterprises are unable to make regular
payments. The victims of such shortcomings
are the employees in these firms. Depending
on how we view official statistics, and our
definition of who should be counted into the
urban workforce, the policy coverage can be
as high as 60% or as low as 30%.
Moreover, HPF management centers are
only partially successful in promoting home
ownership via HPF loans. Although the
number of individual HPF loans has
increased over the years, and the average
size of HPF loan has risen, a large share of
HPF assets remain idle in the banks. This is
a waste of resources in view of the fact that
so many households are in need of housing
loans. In addition, the percentage of
contributors who actually obtain a HPF loan
is ridiculously low, on average less than 6%.
Finally, it can be argued that the HPF policy
is exclusionary vis-à-vis low-income
families, laid-off workers, temporary workers
and migrants. It is even possible that the
HPF scheme entails an economic re-
distribution from low- to high-income
households and thereby accentuates
housing inequalities. In future studies, a
closer analysis of HPF lending practices can
shed light on this issue.
HOUSING FINANCE INTERNATIONAL – March 2006
some scholars argue that the HPF loans
mainly benefit high-income households,
thus accentuating rather than reducing
housing inequalities in China (Mostafa et al.
2003, Wang and Murie 2000: 406-09). In
Wang’s (2000: 857) analysis of Beijing, over
80 percent of the HPF lending went to high-
income families. Based on survey research
in Chongqing and Shenyang, Wang (2003:
182, 2004: 100-17) explains that a very low
percentage of poor households had
participated in, or were aware of the
existence of, Housing Provident Funds.
HOUSING MORTGAGE & HOUSING TRANSACTION IN CHINACHINA’S HOUSING PROVIDENT FUND
Sources: Interviews in China (MoC 2004b, BJ 2004c, BJ 2005b, GZ 2004a, NC 2004, TJ 2004) and written materials provided during these interviews. Note:
Author’s calculations. The national percentages are based on the estimate of 60 million (active) HPF contributors – i.e. workers and staff – in all cities
of China, in 2004 (Interview MoC 2004b).
Table 9. Share of the HPF contributors who obtain HPF loans (%)
BeijingYear Guangzhou Tianjin Nanchang National
1998 0.4 % - - 0.4 % 0.7 %
1999 0.7 % 3.2 % - 1.0 % -
2000 1.6 % 0.9 % - 1.9 % -
2001 3.3 % 1.7 % 6.2 % 2.2 % -
2002 4.5 % 2.6 % - 3.1 % -
2003 5.6 % 3.5 % - 3.3 % 5.4 %
2004 6.4 % 4.4 % - 4.2 % -
47

HOUSING FINANCE INTERNATIONAL – March 2006
References:
Banister, Judith. 2005. “Manufacturing
Earnings and Compensation in China.”
Monthly Labor Review (August): 22-40.
Beijing Review, 1998. “Welfare Housing
Allocation Will Become History.” Beijing
Review, 2-8 November 1998: 14-16.
Burell, Mattias. 2001. The Rule-Governed
State: China’s Labor Market Policy,
1978-98. Doctoral dissertation,
Department of Government, Uppsala
University.
CHSR. 1996. Zhongguo zhufang zhidu
gaige [China’s Housing System Reform].
Edited by China Research Association
for Urban Housing System Reform.
Beijing: Gaige chubanshe.
CNA. 1991. “Housing Reform: What is
New?” China News Analysis, 1 April
1991 (No. 1432).
CNA. 1994. “Urban Housing Reform.” China
News Analysis, 1 September 1994 (No.
1517).
CNA. 1996. “Housing Reform in the Cities.”
China News Analysis, 15 October 1996
(No. 1570).
CNA. 1998. “The New Look for Housing.”
China News Analysis, 1 October 1998
(No. 1619).
Gateway. 2006. “Urban Housing Reform.”
Online article accessed at Development
Gateway, http://www.chinagate.com.cn/
english/161.htm (Accessed 2006-02-05).
GCHR. 2000. Guangzhou shi zhufang zhidu
gaige huigu yu zhanwang [Guangzhou
City Housing System Reform: Looking at
Past and Future]. Unpublished report
edited by Huang Xinjing and Zhang Yan,
Guangzhou Housing Reform Office.
Lee, James. 2000. “From Welfare Housing
to Home Ownership: The Dilemma of
China’s Housing Reform.” Housing
Studies 15(1): 61-76.
Liu, Zhifeng. 2001. “Strengthen the
Management of Housing Provident
Funds.” Speech at national housing
conference, Beijing, 28 October 2001.
Accessed on-line (2006-02-06) at:
http://www.cin.gov.cn/meeting/01g&x/2
001110602.htm.
Mostafa, Anirban, Francis K.W. Wong and
Eddie C.M. Hui. 2003. “A Study on
Housing Provision System towards
Housing Affordability in Shanghai.”
Paper presented at the Department of
Building and Real Estate, the Hong Kong
Polytechnic University.
Naughton, Barry. 1996. Growing Out of the
Plan: Chinese Economic Reform 1978-
1993. Cambridge Cambridge University
Press.
NSB. 2004. Nanchang Economic and
Statistical Yearbook [in Chinese], edited
by Nanchang Statistics Bureau. Beijing:
Zhongguo tongji chubanshe.
Rosen, Kenneth and Madelyn C. Ross.
2000. “Increasing Home Ownership in
Urban China: Notes on the Problem of
Affordability.” Housing Studies 15(1): 77-
88.
SSB. 2003. China Labour Statistical
Yearbook 2003. Beijing: China Statistics
Press.
SC. 1999. Housing Provident Fund
Management Measures, issued by the
State Council’s Standing Committee on
3 March 1999 (Decree no. 262).
SC. 2002. Amended Housing Provident
Fund Management Measures, issued by
the State Council on 24 March 2002.
THR. 2002. 10th Anniversary of Tianjin
Urban Housing System Reform: 1992-
2002, edited by Sun Huaqi. Tianjin
Housing Reform Office.
Walder, Andrew. 1986. Communist Neo-
Traditionalism: Work and Authority in
Chinese Industry. Berkeley and Los
Angeles: University of California Press.
Wang, Ya Ping. 2000. “Housing Reform and
its Impacts on the Urban Poor in China.”
Housing Studies 15(6): 845-864.
Wang, Ya Ping. 2003. “Progress and
Problems of Urban Housing Reform.”
Pp. 177-190 in Social Policy Reform in
China, edited by Catherine Jones Finer.
Burlington, VT: Ashgate.
Wang, Ya Ping. 2004. Urban Poverty,
Housing and Social Change in China.
London and New York: Routledge.
Wang, Ya Ping and Alan Murie. 1999.
Housing Policy and Practice in China.
London: MacMillan Press.
Wang, Ya Ping and Alan Murie. 2000.
“Social and Spatial Implications of
Housing Reform in China.” International
Journal of Urban and Regional Research
24:2 (June): 397-417.
WNC. [World News Connection] Various
articles from 1997-2005. The WNC is a
database with translations of foreign-
language news items and radio
broadcasts from around the world.
References for WNC news items are
labelled with the prefixes WNC or FBIS.
The WNC is funded by the U.S. Dep. of
Commerce.
Wu, Fulong. 1996. “Changes in the
Structure of Public Housing Provision in
Urban China.” Urban Studies 33(9):
1601-1627.
Zhang, Xing Quan. 2000. “The restructuring
of the housing finance system in China.”
Cities 17(5): 339-348.
Zhou, Yun Qing. 2004. An Analysis of the
Factors of Housing Stratification. Paper
presented at the APNHR conference in
Hong Kong, 5-6 February 2004.
CHINA’S HOUSING PROVIDENT FUND
48

HOUSING FINANCE INTERNATIONAL – March 2006
Interviews:
BJ 2003. Interview with Head of Housing
and Social Security Division, Beijing
Housing Reform Office, 12 December
2003.
BJ 2004a. Interview with the former Head of
State Council Housing Reform Office
(1993-97), Beijing, 7 July 2004.
BJ 2004b. Interview with the former Head of
State Council Housing Reform Office
(1993-97), Beijing, 21 July 2004.
BJ 2004c. Interview with the Director of
Beijing Housing Provident Fund
Management Center, Beijing, 23 October
2004.
BJ 2005a. Interview with Head of Housing
and Social Security Division, Beijing
Housing Reform Office, 31 March 2005.
BJ 2005b. Interview with the Director of
Beijing Housing Provident Fund
Management Center, Beijing, 4 April
2005.
GZ 2003. Interview with the Director of
Guangzhou Housing Reform Office, 23
December 2003.
GZ 2004a. Interview with Director of
Guangzhou Housing Provident Fund
Management Center, Guangzhou, 2
November 2004.
GZ 2004b. Interview with Section Chief at
Zhuhai Housing Provident Fund
Management Center, Zhuhai
(Guangdong), 3 November 2004.
GZ 2004c. Interview with Director of
Dongguan Housing Provident Fund
Management Center, Dongguan
(Guangdong), 5 November 2004.
MoC 2004a. Interview with the Director of
the Real Estate Department, Ministry of
Construction, Beijing, 17 July 2004.
MoC 2004b. Interview with the Director of
the Housing Provident Fund Division,
Ministry of Construction, Beijing, 23 July
2004.
NC 2004. Interview with the Director of
Nanchang Housing Provident Fund
Management Center, Nanchang
(Jiangxi), 28 October 2004.
TJ 2004. Interview with the Director of
Tianjin Housing Provident Fund
Management Center, Tianjin, 20 October
2004.
YN 2005. Interview with former Head of
Yunnan Housing Provident Fund
Management Center, Kunming (Yunnan),
18 August 2005.
CHINA’S HOUSING PROVIDENT FUND
49