A review of international experience

participation rate (%)
SOURCE: NBS 1995–2011.
Concerns soon began to emerge about the operation of local schemes and the lack 
of a sound governance framework; the Asian financial crisis in 1997 underscored the need 
to rethink the program (see Chen 2002; Liang 1999; Ma 1999; and Shi 2006 for a dis-
cussion of the shortcomings of previous rural pension schemes). Responsibility for rural 
pensions was subsequently switched from MOCA to the Ministry of Labor and Social 
Security (now the MHRSS) in the 1998 administrative restructuring.

220 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
Contraction and Stagnation: 1999–2002
As a result of concerns about the effectiveness and sustainability of rural pension schemes, 
a sharp policy shift took place in the late 1990s to limit their expansion. By 2001, the shift 
in policy contributed to a steep decline in coverage to just under 60 million participants, 
with the number stabilizing in the low to mid-50 millions through 2007 in roughly 1,900 
counties (see figure 11.1). Despite this contraction, new participants continued to enter 
the system, and accumulated funds from existing schemes continued to increase, more 
than doubling between 2000 and 2007. The number of pension beneficiaries approached 
4 million by 2007 as existing schemes matured.
Renewal: 2003–09
Renewed impetus toward a new rural pension system emerged with new guidelines from 
the Ministry of Labor and Social Security in 2003 (“Notice on Seriously Improving Work 
on Current Rural Pensions”). The new rural pension schemes established during this 
period fall broadly into three types: social pooling plus individual accounts, flat universal 
pensions in combination with individual accounts, and individual accounts only. The 
schemes differed in terms of the financing role of the government at the accumulation 
and payout stages. The Chinese government reviewed the local pilots before introducing 
the new national pilot.
3
 Although no rigorous evaluation was undertaken, a number of 
deficiencies in the rural pension system during this phase undermined the achievement of 
policy objectives and influenced the design of the local and national pilot programs that 
followed.
National Guidelines and Nationwide Coverage: 2009–Present
The introduction of a nationwide framework for the NRPS in 2009 marked a milestone 
in the expansion of pension system coverage in rural areas. The framework has rapidly 
expanded the participation of rural residents and, for the first time, put all localities on a 
more sustainable financial footing through the injection of significant central funding. By 
the end of 2010, the number of contributors in the NRPS reached 103 million, account-
ing for 22 percent of rural employment, a significant increase over 2008. Coverage contin-
ued to expand rapidly in 2011 with the addition of 2,343 rural counties and county-level 
cities and districts (more than three-quarters of the total), including 258 million contribu-
tors and almost 100 million beneficiaries (Wen 2012).
4
PENSION PROVISION IN URBAN AREAS
Earlier pension reforms targeted urban enterprise workers and employees. Based on pilot 
experiments, in 1998 the government set national guidelines for China’s urban social 
insurance programs. These guidelines provided for five types of social insurance for urban 
formal sector workers: pensions, health care, maternity, unemployment, and workplace 
injury. The contributory urban pension insurance program had two components, a basic 
redistributive defined benefit pension (“social pooling”) and a state-managed defined con-
tribution funded individual account scheme.
The urban pension program witnessed steady coverage expansion among wage earn-
ers, including within private and foreign enterprises. However, the program excluded the 

11.  CHINA’S PENSION SCHEMES FOR RURAL AND URBAN RESIDENTS 
221
vast majority of people working for themselves in the informal sector, or for small private 
businesses, as well as the unemployed (collectively known as “urban residents” in the social 
insurance system). Civil servants and public service unit workers continued to have dis-
tinct schemes.
Around 1997, local governments—in Beijing, Chengdu, Guangzhou, Shanghai, 
and elsewhere—began to establish pilot programs for voluntary contributory and basic 
benefits for nonwage urban residents.
5
 Shanghai’s program provided old-age poverty pro-
tection for all urban elderly with no other source of income, financed solely by the local 
government. Guangzhou also set up an urban pension program for the elderly with no 
other fixed sources of income.
6
 The design adopted in Beijing in 2009 was similar to that 
of the urban old-age insurance scheme for wage-based workers, with a monthly basic pen-
sion of Y 280 financed by the local government.
7
 These pilot experiences generated useful 
inputs for the design of the 2011 national framework for the pilot URPS.
In June 2011, the government established nationwide guidelines for the URPS 
(“Guiding Opinions on Piloting Social Pension Insurance for Urban Residents, State 
Council”), with the objective of extending minimum basic income protection to all urban 
elderly, including the unemployed and people without support from social insurance. The 
broad design of the scheme mimics the NRPS (table 11.1), which provides a harmonized 
policy framework to facilitate subsequent integration of the pension schemes for rural and 
urban residents. The program is expected to cover at least 50 million urban residents not 
otherwise covered or insufficiently covered by the scheme for urban workers. After only 
six months of operation, there were more than 13 million participants, including 6.4 mil-
lion pension recipients (Wen 2012). The authorities have set an ambitious goal of expand-
ing the program to all urban areas by the end of 2012.
Several pilot programs established during the 2000s transcended the urban and rural 
distinction by providing an integrated pension scheme for rural and urban citizens within 
individual prefectures.
8
 In rapidly expanding areas, such schemes have merged urban and 
rural schemes to achieve an integrated pension scheme for all residents, in a number of 
cases with benefit levels that are equivalent for people with urban and rural hukou. Like 
the NRPS and URPS, such initiatives are useful vehicles for expanding pension coverage 
to the nonwage sector and promoting the vision laid out by national policy makers of 
rural-urban integration. Almost all of these local programs adopted the national NRPS/
URPS guidelines or some close variant, though there were some exceptions. In Chengdu, 
for example, the city adjusted its local program broadly in line with the NRPS framework 
but retained differences in the details of the scheme design in both the NRPS and the 
URPS (Wang, Chen, and Gao 2011).
Design Features of Pension Schemes 
The old rural pension scheme was voluntary and financed mainly from individual con-
tributions of Y 2–Y 20 a month, with a matching contribution of Y 2 by the collective. 
Retirement ages were to be 60 for both men and women, and elderly farmers were entitled 
to receive a monthly pension benefit paid from the bank until death. Administration was 
by the local bureau of civil affairs (typically at the county level, resulting in highly frag-
mented scheme management); administrative costs were paid for by contributions (with 

222 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
a cap of 3 percent, set subsequently). Oversight and regulation of the scheme sat largely 
with the implementing agency itself, subject to the internal controls of the Ministry of 
Finance and internal auditing.
Several design problems undermined achievement of the government’s policy 
objectives:
 
• Coverage was highly imbalanced geographically, with four coastal provinces 
accounting for about 45 percent of total participation and 64 percent of total 
accumulations. Matching contributions from collectives were often not made, 
and their incidence was highly skewed toward a small number of richer provinces.
TABLE 11.1  Comparison of old and new pension schemes for rural and urban residents of China
Feature
Old rural pension scheme
NRPS
URPS
Principles
Basic protection
Basic protection, broad 
coverage, fl exibility, and 
sustainability
Same as NRPS
Coverage
Rural residents age 20 
and older 
Rural residents age 16 
and older, except students 
Urban residents age 16 
and older, except students
Financing
Individual contribution 
plus collective subsidy 
Individual contribution plus 
government subsidies 
and/or subsidy from rural 
collectives
Individual contribution 
plus government subsidy
Individual 
contribution
10 levels, Y 2–Y 20 a 
month
5 levels, Y 100–Y 500 
a month; family binding 
(parents collect benefi ts 
only if all adult children 
are contributing to pension 
scheme)
10 levels, Y 100–Y 1,000 
a month
Government 
subsidy
Y 2 a month from 
collectives; tax reduction 
for township and village 
enterprises and collective 
economy
Y 30 matching to 
individual account 
annually; Y 55 a month 
basic pension benefi ts
Same as NRPS
Qualifying 
conditions
Pensionable at age 60; 
immediate vesting 
Pensionable at age 60; 
vesting period: 15 years
Same as NRPS
Benefi ts
Accumulation divided 
by 120, but benefi ts 
maintained until death
Accumulation divided by 
139 plus Y 55 a month 
basic pension
Same as NRPS
Fund 
management
Specifi c account at county 
level
Specifi c account at county 
level
Specifi c account at city 
level
Portability
Very limited
In theory, portable within 
NRPS and between 
NRPS and URPS; little 
portability to and from 
urban workers pension 
scheme 
Same as NRPS
SOURCE: Authors’ compilation, based on policy documents. 

11.  CHINA’S PENSION SCHEMES FOR RURAL AND URBAN RESIDENTS 
223
 
• Pension benefits were very low, and even those benefits could not be paid in full 
in at least 200 counties (Wang 2000). Returns to investment were also low.
 
• The administrative cost was as high as 3 percent of contributions, and supervi-
sion was weak. By the end of 2000, about 20 percent of accumulations had been 
invested in unauthorized assets, including real estate, stocks, enterprise bonds, 
and nonbank financial agencies. With county-level management, risk pools were 
highly fragmented.
In 2009, the State Council issued policy guidelines for the RPPS (now the NRPS); 
in July 2011, it issued guidelines for the URPS. The guidelines scaled up multiple local 
and provincial pilot programs into a national framework. The diverse experience with 
rural pension schemes at the subnational level offered important lessons for policy makers 
at the national level that have been reflected in significant measure in the design of the 
national guidelines.
The underlying design principles of the guidelines were basic insurance and wide 
coverage with flexibility and sustainability. Key features of the URPS design include the 
following (table 11.1):
 
• Participation is voluntary, with incentives. All rural residents over the age of 16 
are eligible to participate if they are not already covered by a contributory urban 
scheme. Participation incentives include matching subsidies; 15-year vesting 
requirements to receive the basic benefit; and family binding provisions, whereby 
the pension eligibility of contributors already over retirement age is determined 
by whether the contributor’s spouse and all adult children contribute to the new 
system (family binding).
 
• The scheme design has two components: individual pension accounts with 
matching contributions and a basic flat pension for workers who have contrib-
uted for 15 years or meet other qualifying conditions. The initial value of the 
basic pension under the scheme is Y 55 a month, which can be topped up by local 
governments from their own revenues. Individual accounts have a rate of return 
equal to the one-year deposit interest rate; benefits will be computed by dividing 
the accumulation at age 60 by 139 (as in the urban scheme). Indexation is some-
what vague, to be set in accordance with “economic development and changing 
prices.”
9
 
• Participants become eligible for benefits at age 60. People over 60 at the time 
the scheme is introduced can receive the basic pension benefit if their children 
are contributing to the scheme (family binding). People with less than 15 years 
left before reaching age 60 should contribute during their working lives and then 
make lump-sum contributions to make up any shortfall on the vesting period of 
15 years of contributions.
 
• Financing of the scheme comes from a combination of central subsidies to sup-
port the basic pension (in full for central and western regions and 50 percent for 
eastern regions); individual contributions (of Y 100–Y 500 a year determined 
by the worker); a partial match on the individual contribution by local govern-
ments of at least Y 30 a year (independent of the contribution level chosen by the 

224 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
worker) or at a higher rate as locally determined; and collective subsidies, which 
are encouraged but not mandated, with no level specified. 
 
• Fund management for individual account accumulations will begin at the county 
level, with the aim of shifting responsibility to the provincial level as quickly as 
feasible. Supervision of funds would be by the local offices of the MHRSS. The 
guidelines suggest the importance of gradually raising the level of management 
with the scaling up of the programs and indicate the importance of establishing 
systems for accumulating individual account entitlements and facilitating benefit 
portability.
The URPS shares almost the same design as the NRPS, with exactly the same basic 
pension level (determined by the central government), the same financing design, and the 
same matching contribution provided by local authorities. The scheme aims to provide 
minimum voluntary pension savings arrangements to cover the unemployed, other urban 
workers without employment contracts, and urban retirees without alternative sources 
of retirement income. Significantly, the URPS is open only to people with local urban 
hukou; migrant workers from rural or other urban areas are eligible to participate, but 
only in their area of hukou registration.
10
 One characteristic that is distinct for the urban 
scheme is that contribution tiers can range up to Y 1,000 a year.
Assessment of Schemes and Initial Experience
COVERAGE 
The authorities have set a goal of achieving full geographic coverage for both the rural and 
urban resident schemes by the end of 2012.
11
 Given the voluntary nature of the scheme, 
full geographic coverage does not automatically imply full coverage of all individuals, but 
the authorities are committed to maximizing individual participation in the scheme, espe-
cially for workers uncovered by other forms of old-age income protection. Geographic 
coverage is ensured largely by offering the schemes throughout China and supporting 
registration and contributions with an administrative apparatus.
The authorities appear on track for meeting this geographic coverage target. By the 
end of the first quarter of 2012, 376 million people were participating in the rural and 
urban resident pension schemes, with 107 million receiving pension benefits. Sixteen 
provinces had achieved full geographic coverage, and 10 provinces had integrated their 
rural and urban resident schemes (China Labor and Social Security News 2012). Early 
starters—such as Beijing, Hainan, Jiangsu, Ningxia, Qinghai, Shanghai, Tianjin, Tibet, 
and Zhejiang—generally have high participation rates. In Jiangsu, for example, which 
started to implement the RPPS very early, coverage is reported to have reached 97 percent 
(Huang 2010) (table 11.2).
Notable differences are evident in the participation rates in pilot counties between 
rates reported in administrative data and rates estimated on the basis of survey data, with 
estimated rates usually lower. Reported local coverage or participation rates therefore need 
to be interpreted with some caution. Fieldwork by the authors reveals that local officials 
are prone to calculate coverage in nonstandard ways—including only people over 45 in the 

11.  CHINA’S PENSION SCHEMES FOR RURAL AND URBAN RESIDENTS 
225
denominator (due to the 15-year vesting rule), for example, or including people receiving 
pensions in the total program coverage figure, potentially blurring the line between cur-
rent contributors and pension recipients. In figure 11.1, the coverage rate calculation uses 
total rural employment as the denominator.
PARTICIPATION INCENTIVES
Incentives for participation of the elderly reflected the existence of earlier pilot programs, 
the policy design, financing structure, and pressure on local social security authorities to 
offer the schemes. The age profile of participants and the choices of the contribution level 
have also been affected by the scheme design.
Both the 15-year contribution requirement for vesting into the basic benefit pension 
and the low rate of return on individual contributions would suggest that workers would 
begin contributing to the scheme at about age 45 (somewhat earlier if they expected peri-
ods of unemployment). Limited empirical studies from several provinces of the new rural 
scheme provide initial evidence of this effect. For example, a 2010 survey of Chengdu in 
the early stages of the rural pension pilot suggested that people age 50–59 had the highest 
participation rate and that participation across the life cycle rose (figure 11.2). A survey of 
rural Guangdong in 2011 finds a similar pattern, though neither survey reveals a spike in 
TABLE 11.2  Participation rates in selected local rural and urban schemes
Location
Participation rate (from surveys or 
administrative data) (%)
Source
Rural schemes
Jiangsu Province
97.0 
Huang (2010)
Jiangxi (11 pilot counties/districts) 
73.6 among enrollees in local RPPS 
Mu and Lu (2010)
Shenmu and Yao Counties, 
Shaanxi Province, Jimo District of 
Qingdao City 
75.0 Wu 
(2011)
Guqiao Town, Henan Province 
96.5 
Li (2011)
Urban schemes
Chengdu
27.8 
Wang, Chen, and 
Gao (2011)
Pilot counties, Anhui Province 
Average just above 50; lowest 20.1 
 Luo (2011)
1,942 rural households in 68 
villages in 68 pilot counties in 20 
provinces
Average 57.6 
Rural Research 
Center of the 
Central China 
Normal University 
(2010)
Pilot counties, Xian City, Shaanxi 
Province
54.9 among rural migrants
Liu (2011)
SOURCE: Cheng 2012.

226 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
participation at exactly age 45 (perhaps because of family binding considerations) (Wang, 
Chen, and Gao 2011; Wang, O’Keefe, and Thompson 2011). A survey of pilot coun-
ties in Anhui Province suggests that 49 percent of participants are age 45–49, with only 
6.7 percent under age 29 (Luo 2011). The survey of pilot counties in Zhejiang suggests 
that in some counties, 85 percent of participants were over age 45 and only about 1 per-
cent were under age 35 (Feng 2010).
Contributors need to contribute only at the lowest level of Y 100 a year in order 
to satisfy the vesting requirements for the basic benefit pension, and matching subsidies 
are generally limited to the lowest contribution level. In this way, the rates of return on 
contributions above the lowest contribution level are limited to the rate of return on pen-
sion assets, which is specified as the one-year bank deposit rate (Li 2011). There is thus 
a strong incentive to choose the lowest contribution level (Y 100). Overall returns on 
contributions made at the mandated minimum for the 15-year vesting period (including 
matching subsidies) make for a very high 16 percent internal rate of return on contri-
butions (Wang, O’Keefe, and Thompson 2012). If family binding is also in play, the 
internal rate of return for the pensioner is even higher. Almost 75 percent of benefits 
(93 percent of the subsidy) is provided through the basic benefit of Y 55 a month after 15 
years of contributions, suggesting that unless the parameters of the scheme are adjusted, 
the incentives for workers to contribute more than 15 years or more than Y 100 a year 
will be very low. 
The matching subsidy of Y 30 for annual contributions of Y 100 may be too low to 
incentivize workers to contribute beyond the 15-year vesting period for the basic benefit. 
It is lower than in other developing countries, such as India, where a 1:1 match is more 
common. The flat match and absence of a match above the threshold also act as a weak 
incentive for making contributions above the Y 100 minimum, although it has merit from 
an equity perspective. It will be important to monitor participation to see whether the 
FIGURE 11.2  Rural pension system coverage in Chengdu and Guangdong
b. Guangdong, 2011
a. Chengdu, 2010
0 
20 
40 
60 
80 
coverage (%) 
age
20
40
60
80
100
20
40
60
80
100
age
0 
20 
40 
60 
80 
coverage (%) 
SOURCES: Wang, Chen, and Gao 2011, based on the 2010 Chengdu Rural Pension Survey 2010; Wang, O’Keefe, and 
Thompson 2012, based on the Guangdong Social Insurance Survey 2011. 
NOTE: “Coverage” includes both people currently contributing to a rural pension scheme and people receiving benefi ts.

11.  CHINA’S PENSION SCHEMES FOR RURAL AND URBAN RESIDENTS 
227
Y 30 match is sufficient to incentivize higher individual contributions and contributions 
beyond 15 years.
The empirical findings support the assertion that contributors have very weak 
incentives to contribute at levels above the minimum of Y 100 per year. According to a 
2010 survey in Chengdu, 46 percent of participants from pilot counties chose the lowest 
contribution rate, and only 8 percent chose the highest rate (Y 500 a year). A survey of 
pilot counties in Anhui Province shows that more than two-thirds of participants chose 
the lowest contribution rate (Luo 2011).
The current design concentrates incentives on the ex post subsidy (that is, the 
financing of a basic pension benefit) and has the advantage of simplicity. For mobile rural 
populations, however, ex ante subsidies (that is, the matching of individual account con-
tributions) can be more useful. Once the system matures, rural workers who enroll in an 
urban scheme upon migration—or intent to move—would not benefit as much from the 
incentive effect of the ex post subsidy under the current design, where portability of enti-
tlements to other schemes is still unclear. Portability may be an important consideration 
with an increasingly mobile and urbanizing population. Increasing the ex ante subsidy by 
increasing matching would reduce this possible disincentive effect.
An obvious question raised by a shift in the balance of public subsidies from ex post 
to ex ante is the impact on the poverty alleviation objective of the basic rural pension. If 
greater public subsidies were shifted ex ante, maintaining a neutral fiscal impact would 
require a lower basic pension. The Y 55 benefit is already well below the rural poverty line; 
reducing it leaves the basic benefit below the average per capita dibao threshold, which 
could have additional negative incentive effects.
12
This problem could be dealt with in at least two ways. First, it may be possible 
to effect a partial benefit reduction for individuals above a certain income threshold to 
ensure a higher benefit level for poorer elderly people. Second, local authorities may be 
able to top up the basic benefit to ensure that it exceeds the local dibao threshold.
Four stylistic mathematical projections of individual accumulations and benefits are 
summarized in table 11.3. Scenario 1 evaluates the current monthly contributions and 
matching subsidies, projected basic and total benefits, and the net present value of cen-
tral government subsidies and total central and local government subsidies. Scenario 2 
increases the matching (ex ante) subsidy from 30 percent to 100 percent for the first Y 
100 in contributions while leaving the subsidized ex post basic benefit at Y 55 per month. 
Under this scenario, the net present value of total central government subsidies increases 
somewhat, with a stronger incentive created for contributors and a small increase in pro-
jected monthly benefits. Scenario 3 would remove the ex post basic benefit entirely and 
replace it with a substantial increase in the ex ante matching subsidy aimed at achieving 
the same monthly benefit as the current scheme for a worker who contributes for 15 
years. This arrangement could have the effect of reducing the government’s cost in pres-
ent value terms by about 10 percent (depending on the real discount rate), as a result of 
the return on the matching subsidy which would be invested during the accumulation 
period (assuming a 3 percent real rate of return, which has not been consistently the case 
in China). Alternatively, if the central government authorities retained the same level of 
subsidy in present value terms but shifted the subsidy from the basic pension benefit to a 

228 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
matching subsidy during the accumulation phase (Scenario 4), the monthly benefit would 
increase by about 43 percent, depending on the interest rate assumed.
Public awareness of the schemes has a substantial impact on enrollment. Farmers 
still have very limited knowledge about the NRPS, as suggested by a survey of pilot coun-
ties in Jilin, which showed that 45 percent of local farmers did not have a clear knowl-
edge about the policy and 6 percent had never heard about it (Liu, Wu, and He 2011). 
A survey of pilot counties in Hebei revealed that only 27 percent of farmers knew that 
the government provides subsidies for the RPPS (Geng 2011). Farmers were not familiar 
with the NRPS in pilot or nonpilot counties at the beginning. In order to increase their 
awareness, the central government asked local governments to disseminate this new policy 
through various channels, including broadcasting, television, posters, brochures, and vil-
lage campaign activities. Thanks to the media and administrative campaigns, farmers are 
largely informed of the NRPS program.
TABLE 11.3  Stylized Examples of Matching Subsidy Options
Y PER INDIVIDUAL
Scenario
1
2
3
4
Individual annual contribution
100
100
100
100
Local annual matching subsidy 
(accumulation phase)
30
30
30
30
Central government annual 
matching subsidy (accumulation 
phase)
70
548
825
Monthly individual account benefi t in 
retirement
 13.8 
 21.2 
 68.8 
 98.2 
Monthly basic benefi t in retirement
 55.0 
 55.0 
Total monthly benefi t in retirement
 68.8 
 76.2 
 68.8 
 98.2 
NPV local matching subsidy
 358 
 358 
 358 
 358 
NPV lifetime central government ex 
ante matching subsidy
 836 
 6,184 
 9,491 
NPV lifetime central government 
subsidy for basic benefi t
 9,490 
 9,490 
NPV total government subsidies
 9,848 
 10,684 
 6,542 
 9,849 
NOTE: Scenario 1: Current NRPS/URPS parameters and subsidies; Scenario 2: matching ex ante subsidies provided by 
the central government at 70% of contributions increasing the total central/local subsidy to 100% of contributions; Scenario 
3: matching ex ante subsidies provided by the central government at 548% of contributions with central government–
fi nanced basic benefi t eliminated; Scenario 4: matching ex ante subsidies provided by the central government at 825% of 
contributions with central government–fi nanced basic benefi t eliminated. In all scenarios, the authors projected individual 
contributions of Y 100 a year and local government matching subsidies of Y 30 a year. A 3 percent real interest rate and 
real discount rate was assumed in all scenarios. Other assumptions included 15 years of contributions at eligibility age, an 
eligibility age of 60, and life expectancy at age 60 of 19.1 years for both men and women. NPV = net present value.

11.  CHINA’S PENSION SCHEMES FOR RURAL AND URBAN RESIDENTS 
229
ADEQUACY OF BENEFITS
Benefit adequacy—how effective the benefit is in ensuring that the elderly are shielded 
from absolute or relative poverty—is difficult to measure. Two metrics often considered 
are the locally determined poverty line and an international metric such as income.
13
 At 
the individual level, there is a natural tension between benefit adequacy and affordability. 
A benefit that is completely effective at ensuring the elderly against poverty will likely 
come at a high cost if shouldered entirely by the individual and his or her family. More-
over, a benefit that is barely adequate for some people will prove inadequate for others. 
Benefit adequacy also depends not only on the benefit level but on the adjustment of such 
levels over time, in line with prices, wages, or per capita income. Bearing in mind these 
trade-offs, setting a target benefit level is challenging in China, where local economic con-
ditions vary substantially.
Benefit levels vary greatly by locality, as different counties and cities offer different 
contribution levels above the national maximum, different matching subsidies, and dif-
ferent levels of basic benefits. Moreover, the benefit accruing to individuals will depend 
on the level and duration of contributions, the rate of return, the matching subsidies and 
supplementary financing provided by localities, and the annuity factor used to calculate 
the benefit. Based on several basic assumptions and bearing these caveats in mind, the 
authors calculated that a minimum benefit of about Y 73 a month would be obtained by 
a worker making a Y 100 contribution a year for 15 years.
14
 This projected benefit would 
only partially achieve the unstated objective of protecting against poverty in old age. In 
2011, this amount represented 13 percent of net rural per capita income, 30 percent of 
the absolute poverty criterion of about Y 20/day, 51 percent of the national average dibao 
threshold, and 38 percent of the national rural poverty line.
Before the introduction of the national scheme, pilot local programs often adopted 
a rule of thumb that pension benefits should be at or slightly above the average dibao 
threshold in the locality. Together the anticipated basic benefit and funded individual 
account accumulations should yield a benefit that is expected to raise total pensions nota-
bly above the dibao level. The level of benefits will vary based on local fiscal capacity (for 
example, Beijing sets its flat pension portion at 35 percent of rural average income).
With respect to indexation, guidelines established for the rural and urban schemes 
note the importance of adjustments to the basic pension, but they do not specify criteria 
or parameters. The basic benefit of Y 55 a month, put in place in September 2009, has not 
been adjusted, despite annual inflation in 2010 and 2011 of 5–6 percent.
Survey data provide limited insights into actual benefit levels. A study of 18 counties 
in Hebei Province indicates that as many as 55 percent of rural pensioners claim only the 
Y 55 basic pension; less than 15 percent of pensioners had a monthly pension of more 
than Y 100 (Geng 2011). A 2010 survey of Chengdu suggests that despite the steady 
increase in the benefit levels of local programs, which are higher than the national aver-
age, about three-quarters of pensioners still receive a monthly benefit of less than Y 150, 
a replacement of rural per capita net income of 31 percent. A survey of pilot counties in 
Anhui Province reveals that the average monthly benefit is Y 60.4 and that 71 percent of 
the elderly believe that the current benefit level is inadequate to meet their essential needs 
(Luo 2011). Farmers cannot depend solely on their income from the NRPS; they have to 

230 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
rely on other income sources, such as family members, farming income, property rental 
income, and support from communities.
The RPPS was scaled up only in 2010; too little time has passed to evaluate the 
scheme’s impact on the incomes and well-being of elderly people in rural areas.
15
 More-
over, benefit levels of Y 55 a month would likely have a very modest impact on all but the 
poorest rural elderly.
Despite the low level of benefits, survey data suggest that rural recipients of pen-
sion benefits in earlier schemes indicated that they felt a stronger sense of security and 
self-regard (Zhang and Tang 2008). Scheme benefits help them with living expenditures, 
increase economic stability, and reduce some of their reliance on other sources of income. 
A 2010 study of Chengdu suggests that more than half the elderly claim their pension on 
a monthly basis, some rely largely on pension income support, and 71 percent indicate 
that the most important uses of their pension benefits were food, followed by health care, 
production materials, and savings.
FISCAL COSTS AND WORKER AFFORDABILITY
Five sets of fiscal costs are associated with the NRPS and URPS: 
 
• Central government financing of the basic benefit, initially set at Y 55 a month 
per person
 
• Local government financing of matching subsidies, which are at least Y 30 per 
year per contributor
 
• Local government financing of individual contributions for dibao recipients and 
possibly others deemed locally to be entitled to a contribution subsidy
 
• The subsidy required to finance the difference between the annuitized benefit 
using the current annuity factor of 139 and a more appropriate and actuarially 
fair annuity factor
 
• The payment to survivors of any remaining balance in a deceased worker’s indi-
vidual account.
Presumably, the local authorities will bear the costs of the fourth and fifth items, 
which will come payable only over time. 
An unresolved issue is the future fiscal affordability of the new schemes. Initial calcu-
lations suggest that the central government’s commitments are affordable, although as sug-
gested above, such affordability may come at the price of insufficient adequacy for many 
beneficiaries. The cost to the central government of the portion of the basic benefit for 
which it is responsible is low and likely to remain so despite a growing elderly population. 
Simulations using different system parameters find that spending on pensions consumes 
1.0–2.5 percent of central general revenues (Cheng 2011).
16
 Providing Y 55 a month to 
everyone in China over age 60 would cost the central government about 0.26 percent 
of gross domestic product (GDP).
17
 These costs are projected to grow over time, as the 
elderly population increases in China. However, if the benefit level were to grow at the 
same rate as GDP growth and the central authorities were to finance 100 percent of the 
basic benefit, an estimate of the cost would still be less than 0.75 percent of GDP in 2050. 
Keeping such costs low and affordable over the coming decades will inevitably depend on 

11.  CHINA’S PENSION SCHEMES FOR RURAL AND URBAN RESIDENTS 
231
the expansion of the importance of the NRPS/URPS saving provisions as well as other 
saving options for nonwage citizens.
Fiscal affordability at the county and municipal levels is more uncertain, as the level 
of matching contributions for individual accounts and the fiscal situation varies greatly 
across localities. The current matching is evenly shared by provincial, city, and county-
level governments. The cost to local governments of minimum matching subsidies is 
very low relative to GDP, although it depends on local revenue-generating capacity.
18
 The 
minimum fiscal commitment therefore should be affordable in the aggregate, although it 
remains to be seen whether the required commitment will be met in every locality. The 
obvious challenge is to avoid a situation in which poor counties fail to match individual 
account contributions (resulting in lower accumulations for the poor) while still maintain-
ing enough local interest in the scheme to encourage accountability at the county level.
A third subsidy is implicit in the annuity factor used to calculate benefits for the 
individual account pension. The guidelines specify that the coefficient used to determine 
the benefit is 139 months (the same coefficient used to calculate benefits from workers’ 
individual accounts in the urban workers scheme), based on a life expectancy calculation 
at age 60 and a 4 percent expected rate of return on funds during retirement. Unisex life 
expectancy tables published by the World Health Organization for China indicate that 
life expectancy at age 60 in 2009 was 19.1 years (229 months), suggesting that the 139 
figure is much too low.
19
 There is thus a substantial subsidy embedded in the annuity fac-
tor for the individual account pension of about 65 percent of the benefit amount (Wang, 
O’Keefe, and Thompson 2012). This subsidy would be the responsibility of the authori-
ties administering the scheme, which in the short term would be local authorities.
No simple metric measures worker affordability. For workers themselves, minimum 
contributions of Y 100 a year represented only 1.4 percent of rural per capita income in 
2011. They should therefore be affordable for all but the lowest deciles of rural workers. 
Of course, many rural workers will be liquidity constrained or concerned about the risk 
that the benefits received after 15 years of contributions may turn out to be far different 
from intended at the outset. Such a contribution would be more affordable for urban 
workers, who generally have much higher income levels.
FUND MANAGEMENT AND GOVERNANCE
The initial approach under the new schemes of allowing fund management at the county 
level has a range of drawbacks, including investment risk and the risk that funds are used 
for other pressing purposes, leaving accounts that are in practice empty. Localized man-
agement also complicates the portability of account balances for rural workers who move 
beyond their home counties. In addition, demographic trends in rural areas suggest that a 
sizable reserve fund will likely be necessary, which is best managed at higher levels.
Raising the level of risk pooling—which can occur only gradually—would have sub-
stantial benefits in terms of both the soundness of the system design and implementa-
tion capacity.
20
 Gradual steps should probably be considered to upgrade the pooling level 
from county to prefecture (as has already happened in the urban workers scheme and 
in some prefectures for all schemes); from prefecture to province; and from province to 
the national level. Some cities or provinces, such as Beijing, Hunan, and Shanghai, have 
already been managing the funds at the provincial level.

232 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
The experience of local pilot programs with respect to scheme administration and 
fund management demonstrates a modest degree of capacity and continuity, but regula-
tion demands greater attention. Administration of new schemes has remained with the 
Social Security Bureau, with roles for villages (for collection of contributions and in some 
cases payment of benefits), townships (for consolidation of collections, approval of ben-
efits and registration), counties (for general oversight and scheme design), and banks (for 
payment of contributions and delivery of benefits in many cases). There also appears to 
have been a general improvement in scheme information systems, in particular with the 
issuance of standard software by the MHRSS for the scheme.
Within this general administrative structure, there are a variety of local innovations 
in administration. For example, farmers in Guangdong and Suzhou have been able to 
make contributions directly through banks rather than village officials, and post offices 
(including mobile facilities) are being used in areas with lower banking penetration. Some 
schemes also allow for the seasonality of farmer incomes, so that schemes such as the one 
in Baoji collect payment only once a year.
In addition to the drawbacks of subprovincial fund management, the use of the 
one-year deposit rate of interest as the rate of return for individual accounts is likely to 
prove problematic over time (as it has for urban schemes), because it has typically pro-
vided a negative real rate of return, which affects both the return on contributions and the 
adequacy of benefits. Although very secure, the low rate virtually guarantees a low indi-
vidual account balance at retirement, thereby weakening participation incentives for rural 
workers. Other approaches, such as nonfinancial or notional defined contribution plans, 
can in principle provide benefit promises based on higher rates of return, but such returns 
would require additional fiscal subsidies. Another option would be to invest reserves in a 
much riskier pool of assets, which would subject members to additional risk and require 
substantial additional institutional infrastructure and oversight.
PORTABILITY OF PENSION RIGHTS
Portable pension rights reduce pension losses for individuals who move across space and in 
and out of formal sector employment. Portability is therefore important to the overall par-
ticipation incentives in the rural and urban resident pension schemes as well as to address-
ing gaps in the overall set of pension instruments, including the urban old-age insurance 
system. Portability of vested rights and individual account balances—both across rural 
areas and between rural and urban schemes—is important given China’s increasingly 
mobile labor force. Conceptually, it is much easier to achieve portability of pension rights 
within rural and urban resident pension schemes, as they would have the same essential 
design and compatible institutional infrastructure; it is more difficult achieving portabil-
ity between separate pension schemes, such as between the NRPS/URPS on the one hand 
and the urban workers pension scheme on the other.
The NRPS/URPS guidelines anticipate but do not address the portability of pen-
sion rights under the schemes. In 2009, the Chinese government established an initiative 
for pension portability within the urban worker’s pension scheme.
The next policy priority will focus on making pension rights portable within and 
across programs; MHRSS is developing regulations along these lines, which the govern-
ment plans to introduce in 2013. Ensuring portability will require the rapid development 

11.  CHINA’S PENSION SCHEMES FOR RURAL AND URBAN RESIDENTS 
233
of systems to reliably transfer information and funds across localities, integrating rural and 
urban schemes and integrating both schemes with urban social insurance. Provinces such 
as Guangdong are already working on development of the provincial management infor-
mation systems needed to do so.
The lack of a national policy has contributed to pension system fragmentation and 
weak incentives for participation. Fragmentation between the urban and rural pension 
schemes has become an important barrier for coordinated urban-rural development and 
urbanization (Deng and Liu 2011; Wang 2006). A number of provinces and cities have 
integrated pension schemes for rural and urban residents. What is needed is a national 
policy to guide the transfer and portability of pension rights within and between pension 
schemes.
Distinct local policies have been adopted to address this issue. Suzhou, for example, 
initiated a simple 2:1 rule for farmers wishing to transfer their social pooling rights from 
the rural to urban scheme; transferring their rights is simple, because the rural contribu-
tion base has been exactly half that of the urban system. The 2:1 rule means farmers 
will receive half of their social pooling benefits when transferring into the urban scheme 
because their contributions are exactly half that of their urban counterparts. Beijing has 
made provisions for portability, albeit only at the point of retirement. Former farmers who 
have accumulated enough years in the urban system at retirement will receive an urban 
pension (with an allowance for lower rural contributions); their urban contributions will 
be credited with a rural pension scheme if the accumulation period in the urban system 
is less than 15 years. In principle, the funded portion of all schemes could easily be made 
portable.
The clear direction of policy is toward eventual integration of the schemes for rural 
and urban residents; some prefectures have already done so.
21
 The Social Insurance Law 
promulgated in July 2011 spells out unequivocally: “The government needs to set up 
a robust pension program for urban residents. Provincial government[s] or the govern-
ments of municipalities directly under the central government may, in line with their local 
conditions, establish an integrated pension program for urban and rural residents.” Such 
integration is an important objective for China’s pension reform.
INSTITUTIONAL CHALLENGES
Little information is available about institutional capacity, although several key challenges 
are clear: 
 
• Local capacity needs to be increased, particularly at the county level and below. 
The massive and very rapid expansion of the system places demands on local 
implementation and delivery capacity. The government intends to introduce 
rural social security service centers at least down to (and ideally below) the county 
level. However, existing staffing ratios imply service loads in an expanding system 
well above levels observed in similar schemes in other countries. Experiences from 
provincial partnerships with the banking sector may assist in helping spread the 
administrative burden of managing client contributions and basic recordkeeping. 
For example, Guangdong has a generalized service agreement with the Postal Sav-
ings Bank, which accepts contributions, is the vehicle for payment directly into 

234 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
pensioners’ accounts, and has an information system linked to the Department of 
Human Resources and Social Security.
 
• Collection and payment systems need improvement. Incomplete penetration of 
financial and banking services in some rural areas likely affect implementation, 
by constraining contribution collections and payments. The MHRSS is working 
actively on this issue, in cooperation with national banks such as the partnership 
with the Postal Savings Bank, but even banks with wide coverage do not have 
branches in all townships.
 
• Information systems and links to related programs—such as to the New Coop-
erative Medical Scheme—and other localities need strengthening. The govern-
ment’s stated intention is to extend the systems to the grassroots level; doing so 
will require increasing system capacities and training personnel. If the system is 
not closely managed, information systems could be fragmented (the standardized 
software from the MHRSS should help promote greater coherence than has been 
observed in urban schemes).
Lessons for Other Countries
China’s adoption of national pension programs for rural and urban residents in 2009–11 
may provide insights for other countries considering similar reforms. Lessons include the 
need for high-level political commitment, iterative and gradual policy development, an 
incentive structure for voluntary participation, policy design to achieve rural-urban inte-
gration, and assignment of financing responsibilities of central and local governments in a 
decentralized environment. At the same time, it is important not to be overly mechanical 
in interpreting these lessons, as they reflect idiosyncratic characteristics of China’s labor 
market, political, and fiscal systems.
As China moves from a middle-income to a high-income country, the acceleration 
of pension programs aims to narrow coverage gaps between rural and urban areas and 
across regions and promote more integrated development. This effort has been driven by 
China’s ambitious policy goal to provide basic social protection for all by 2020. Such high-
level political commitment is key for developing countries seeking to approach universal 
coverage and allow all people to share the fruits of economic development. The innovative 
design of China’s matching defined contribution plus basic benefit scheme has already 
dramatically increased the registration of rural and urban residents. Only time will tell if 
such high voluntary participation is sustained.
The process and mechanism of policy reform in China has followed an iterative 
and gradual paradigm. China has conducted subnational piloting, accumulating experi-
ence and lessons for policy design at the national level and subsequent roll-out. For a 
large country, local experiments are useful to accommodate diverse conditions and decen-
tralized environments. Few countries are as fiscally and administratively decentralized as 
China (DRC and World Bank 2011; Lou and Wang 2008). Many countries may be suf-
ficiently centralized that they can prescribe more uniform designs and financing for such 
programs. Even in these cases, however, countries could benefit from piloting designs and 
“learning by doing,” as China as done.

11.  CHINA’S PENSION SCHEMES FOR RURAL AND URBAN RESIDENTS 
235
China is now trying to coordinate and integrate its systems for urban and rural 
workers. In other countries, it may be possible to avoid the initial split between rural and 
urban schemes for informal workers, which in China is a legacy of a broader social policy 
framework of separate programs for rural and urban people linked in part to the hukou 
system.
The main difference between China’s old and new rural pension schemes is that 
the ex ante and ex post financing from central and local governments has helped assure 
participants of the credibility of the scheme. The ex ante subsidy to individual accounts 
is financed by local governments (province, city, and county); the ex post subsidy to the 
basic pension is financed totally by central governments for inland provinces and shared 
by central and provincial governments for coastal provinces. The financing responsibilities 
of central and local governments are clearly defined. Both central and local financing have 
been important because many Chinese lost confidence in earlier rural pension programs 
that depended entirely on individual responsibilities and local collectives. The lesson for 
other countries is that a significant and credible public subsidy may be necessary to incen-
tivize the voluntary participation of nonformal workers. The commitment of the central 
authorities to a significant subsidy has also added credibility to the pension commitment. 
In other countries, government support may not be sufficient to give individuals the con-
fidence to contribute.
A related question is the appropriate balance between ex ante and ex post public 
subsidies and the extent to which the ex post subsidy is best “bundled” with a match-
ing defined contribution ex ante subsidy (as in the NRPS and the URPS) or poten-
tially delinked through a noncontributory social pension approach. The initial success 
in achieving substantial registration in rural China can be attributed to the structure of 
incentives of government subsidies, vesting requirements, and family binding provisions. 
Family binding provisions may prove of particular importance in China, where the pro-
portion of rural migrants of working age is substantial and many of the rural elderly have 
remained in rural areas.
Linkages between voluntary matching defined contribution schemes and a basic 
defined benefit pension represent an innovative design that deserves consideration by 
countries with comparable needs, conditions, and the administrative apparatus to carry 
such a design forward. At the same time, preliminary empirical findings show that rural 
young people are less likely to participate in the NRPS, suggesting that the incentives may 
not be strong enough. Given the relatively low participation of younger rural workers to 
date, it may be worth considering rebalancing the public subsidy from an ex post toward 
an ex ante subsidy.
A bigger question for China and other developing countries is whether it may be 
worth going even farther and converting the basic pension benefit into a full-blown 
noncontributory social pension largely delinked from the matching defined contribution 
portion of the schemes. To reduce elderly poverty, China could consider a social pension 
(World Bank forthcoming).
China’s pension schemes for rural and urban residents appear to be financially sta-
ble, because they focus on fiscal and worker affordability while providing only a very 
modest benefit. This situation will be changing with the rapidly aging population. Fiscal 
affordability will also depend on the future levels at which the authorities provide ex ante 

236 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
subsidies, the returns on individual account accumulations they commit to, and future 
adjustments to the level of basic pension benefits. Whether the fiscal costs of ex ante and 
ex post subsidies are affordable and appropriate for other countries will depend on their 
characteristics, including the working age and elderly income distribution and household 
composition.
The limited adequacy of benefits may leave a substantial portion of the elderly 
insufficiently protected against poverty. Potential remedies include increasing the ex 
ante matching subsidy, which induces individuals to save for their own retirement, and 
strengthening the linkages between the pension and dibao programs.
Other challenges China faces include pooling and managing funds, strengthening 
institutions and capacity for service delivery, and managing funds while rapidly increas-
ing the geographic coverage of the program. These challenges are common in develop-
ing countries. A key lesson of earlier schemes in China is that overly decentralized fund 
management, pooling, and regulation pose real risks that need to be managed, prefer-
ably through higher-level pooling and scheme management. Institutional penetration 
in remote areas of China—including the presence of finance bureaus, labor and social 
security bureaus, banks and other funds transfer agents, and a strong system of identifica-
tion and registry—was another important asset. Not all developing countries have this 
administrative and financial sector infrastructure, raising questions about the possibility 
for achieving such wide coverage expansion so quickly and the relative roles of govern-
ment and nongovernmental partners in the process. Even in China, partnering between 
government and financial sector players has proven important in achieving accelerated 
scheme expansion. Although China’s experience demonstrates the potential of a voluntary 
and incentive-driven approach to expanding coverage, its specific political economy has 
probably been a significant factor in the impressive initial performance. Although the 
NRPS and the URPS are voluntary, China retains a degree of moral suasion to encourage 
participation that may not be present in many developing countries. These enabling con-
ditions—and not just the design characteristics—may have supported these pension pro-
grams. At the same time, such suasion was present in earlier rural schemes, which failed 
to achieve significant penetration, suggesting that the NRPS and URPS design features of 
are likely to be important factors in the early success of the schemes.
Conclusions
China’s recent establishment of national pension schemes for its rural and uncovered 
urban populations represents a major step toward its ambitious objective of universal pen-
sion coverage by the end of 2012 and provision of basic social security for all by 2020. 
The policy designs support the principle of broad coverage, in part by providing modest 
basic benefits. The urban and rural designs have already achieved substantial momentum 
in increasing the registration and participation of previously uncovered populations. The 
common policy framework has provided a channel for an integrated pension scheme for 
rural and urban residents. Such a framework also appears able to accommodate the wide 
variation in economic needs and circumstances across China.
Although the NRPS and the URPS have several innovative features and represent 
milestones in social policy in China, there remains scope for improvement and refinement 

11.  CHINA’S PENSION SCHEMES FOR RURAL AND URBAN RESIDENTS 
237
of some policy parameters if China is to fully achieve its objectives. Looking forward, 
whether China will be able to achieve universal pension coverage at the individual level 
will depend in large part on the incentives for sustained participation, the authorities’ 
ability to enhance portability and eventually integrate the programs, and their ability to 
strengthen institutions and capacity for service delivery and management. 
The authorities have established a benefit level and financing plan that can achieve 
broad coverage, but the benefit may not be sufficient to protect substantial numbers of 
elderly beneficiaries from poverty. In setting the benefit level and guaranteeing the financ-
ing of the basic benefit for most provinces, the authorities have balanced benefit adequacy 
against a very modest fiscal commitment while at the same time enabling local authorities 
to supplement benefits where local fiscal situations and priorities permit. One means of 
promoting a higher replacement rate could be to increase the matching subsidy in the 
pension savings scheme, including perhaps providing financing by the central government 
to enhance the match. Doing so would both increase incentives to contribute more than 
15 years and improve benefit adequacy. But such a move would need to be accompanied 
by gradual relaxation of portfolio rules to allow investment of accumulations in vehicles 
that ensure a positive real rate of return, which is not currently the case.
The linkages between the vesting requirements and family binding provisions for 
the savings scheme on the one hand and the basic benefit on the other likely provide 
strong incentives for participation while at the same time creating a significant adminis-
trative challenge for a very modest benefit. Further evaluation is needed of the costs and 
challenges of linking the two components of the scheme. Increasing the matching subsidy 
(and target replacement rate) could enhance the benefits of linking the two elements.
Decentralized and fragmented fund management in China cannot achieve risk pool-
ing, which is essential for ensuring benefits over the long term. Measures are needed to 
raise the level of fund and risk management to at least the provincial level in order to 
achieve risk pooling and portability, two important objectives for China’s increasingly 
mobile labor force and diverse local fiscal circumstances. A higher level of participation 
in fund management and regulation can increase professionalism and system robustness 
within given capacity constraints.
Notes
1.  This section draws extensively on Cai and others (2012) and World Bank (forthcoming). 
2.  The term “resident” has a specific meaning in the context of social insurance, because resi-
dency in China is defined based on a person’s hukou (household registration) status and type 
of employment. In this chapter, the term refers to all rural workers (except local officials) and 
self-employed, informally employed, and nonworking people with local hukou in urban areas. 
Residents are contrasted with workers, who are covered under the urban workers pension 
scheme.
3.  There were five major variants in local schemes during this renewal phase: (1) flat pensions 
plus individual accounts with government financing at the payout stage only; (2) flat pen-
sions plus individual accounts with government financing by matching contributions to indi-
vidual accounts or at the payout stage through the financing of flat pensions from general 

238 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
revenues; (3) individual accounts with social pooling, with government financing in the 
accumulation phase; (4) individual accounts combined with social pooling, with government 
financing by matching contributions to individual accounts or at the payout stage; and (5) 
individual accounts only, with government matching for contributions to such accounts. This 
simple design lacks risk pooling of any form. For descriptions of the systems in Beijing, Baoji, 
Suzhou, and Yantai, see Wu (2009). For a detailed discussion of the Baoji pilot rural pension 
experience, see Zhang and Tang (2008).
4.  China’s definitions of scheme coverage and participation are not consistent with international 
norms, as total figures for scheme participation tend to include both working-age people still 
contributing and people already receiving pensions and no longer contributing, as discussed 
later in this chapter.
5.  Other cities launched their own pilots before the establishment of national RPPS guidelines in 
2009. They include Lianshan, in Shandong Province; Zhengzhou, in Henan Province; Yulin, 
in Shaanxi Province; and Wuhu, in Anhui Province. These urban programs generally adopted 
a design of matching defined contribution savings arrangements and basic benefits (social 
pooling), but they varied greatly in terms of the financing and contribution rate; the central 
government did not provide any financial support. There was significant variation in benefit 
levels resulting from differences in local economic circumstances and the fiscal position of 
local governments.
6.  This system combines a basic pension and individual account, with a flat individual contribu-
tion rate of Y 2,400 a year and immediate vesting. Participants could claim a monthly pension 
of Y 400 the same year they began to make contributions. Contributions during working life 
were waived for dibao household members (people participating in the minimum living guar-
antee social assistance program).
7.  Currency values in this chapter are given in Chinese yuan; Y 1 = $0.1591.
8.  Pilots were run in Zhejiang Province, Chongqing and Chengdu in Sichuan Province, Qingdao 
City in Shandong Province, and Taizhou City in Jiangsu Province.
9.  This is the same method used in urban schemes. It reflects the assumptions of a notional life 
expectancy at age 60 and an assumed draw-down interest rate of 4 percent.
10.  Survey evidence from selected major cities in China for 2010 confirms that a proportion 
of migrant workers reporting participation in pension schemes were in fact contributing 
to schemes in locations other than that of their surveyed residence (Giles, Wang, and Park 
forthcoming).
11.  Geographic coverage refers to offering the scheme to all workers and the elderly people in 
certain geographic areas. Full geographic coverage refers to offering the scheme to workers and 
the elderly throughout China.
12.  Dibao refers to the minimum living guarantee social assistance program for China. The dibao 
threshold is the subsistence level below which household incomes are topped up.
13. Ideally, an elder assistance benefit should be sufficient to cover the poverty gap for such 
individuals.
14.  Assumptions include a contribution of Y 100 a year, a matching subsidy of Y 30 a year, and 
retirement at 60 based on life expectancy in 2012. The benefit in 2012 for 15 years’ accu-
mulations would be about Y 18 a month, and the basic benefit would be Y 55 a month. The 
adequacy of benefits provided by individual account pensions will depend on the contribution 
level, the rate of return on individual accounts, and the annuity factor used to determine the 
pension benefit. 

11.  CHINA’S PENSION SCHEMES FOR RURAL AND URBAN RESIDENTS 
239
15.  Studies show that households with social security tend to have a low savings rate, suggesting 
that pension provision could reduce saving and boost household consumption (Cai and others 
2012).
16.  The assumptions underlying these simulations merit note. Chen uses a projection model 
based on data from Jiangsu Province that assumes that one-quarter of the increase in general 
revenues would be needed for rural pension subsidies to support a universal pension for men 
at age 60 and women at age 55. Assuming that this figure represents 2.5 percent of general 
revenues, a universal farmers’ pension of Y 825 a year could have been provided in 2010 com-
pared with a minimum basic benefit of Y 660 under the national scheme.
17.  The 0.26 percent of GDP figure is likely to be overstated, for several reasons. First, the central 
government is responsible for only 50 percent of the basic benefit for beneficiaries in coastal 
provinces. Second, recipients of the urban contributory old-age pension are not eligible for the 
URPS. Third, the NRPS or the URPS is not likely to cover all elderly people.
18.  In this case, the estimated cost of providing a minimum annual pension of Y 30 per person 
for all working people age 18–59 would be 0.4 percent of GDP in 2012. This figure would 
gradually decline, as the projected working-age population is projected to decline over the 
coming years.
19. See http://www.worldlifeexpectancy.com/your-life-expectancy-by-age.
20.  The authorities indicated that “in the pilot phase, the new rural pension funds can be man-
aged by counties…and with the rolling out of the pilot, management can be transferred to 
higher levels; provincial management can be practiced in places with the required conditions.” 
Funds were “to be deposited into the fiscal account of social security funds, and managed in 
a way of separating revenue from expenditures with separate book-keeping and auditing and 
inflation-proofing and appreciation according to relevant regulations” (State Council 2009). 
21.  The 2009 State Council decision states: 
The way the new rural pension scheme can be dovetailed with the urban workers basic pension 
scheme and other pension schemes should be decided by the Ministry of Human Resources and 
Social Security together with the Ministry of Finance. There is also a need to dovetail the new rural 
pension scheme with other policies and systems, such as social security of farmers whose land has 
been requisitioned; policies for reservoir resettlement; preferential policies toward households prac-
ticing family planning in rural areas; Wubao [the five guarantees program for the elderly, people 
with disabilities, and youth who lack the ability to work, sources of income, and sources of sup-
port, such as assistance from family members; about 5.5 million people have been covered by such 
benefits in recent years] in rural areas; social special care; and dibao. The specific methods will be 
studied and decided by the Ministry of Human Resources and Social Security and the Ministry of 
Finance, together with other relevant departments.
References
Cai, Fang, John Giles, Philip O’Keefe, and Dewen Wang. 2012. The Elderly and Old Age Support in 
Rural China. Directions in Development. Washington, DC: World Bank.
Chen, Ping. 2002. “Establishing a Unified Social Security System Is a Shortsighted Policy.” China 
Reform 4: 16–17. (In Chinese.)
Cheng, Jie. 2011. “RPPP Fiscal Burden Estimates: Possibility of Universal Coverage during the 
12th Five-Year Plan Period.” Social Security Studies 1: 57–66. (In Chinese.)
—. 2012. “Overview of China’s New Rural Pension and Urban Residents’ Pension Scheme.” 
World Bank, Washington, DC. (In Chinese.)

240 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
China Labor and Social Security News. 2012. Issue 4883, April 27.
Deng, Dasong, and Yuanfeng Liu. 2011. “Risk Analysis on the Social Security System: Close-up 
Analysis of the RPPP.” Journal of Chinese Economic and Business Studies 4: 84–89. (In Chinese.)
DRC (Development Research Center), and World Bank. 2011. China 2030: Building a Modern, 
Harmonious and Creative High-Income Society. Beijing and Washington, DC: World Bank.
Feng, Jing. 2010. “The RPPP: Policy Design and Implementation Effect.” World Economy, World 
Economic Situation 8: 14–19. (In Chinese.)
Geng, Yongzhi. 2011. “Follow-up Survey on RPPP Pilots: Rural Households in 18 Counties.” 
Research on Financial and Economic Issues 5: 125–28. (In Chinese.)
Giles, John, Dewen Wang, and Albert Park. Forthcoming. “Expanding Social Insurance Coverage 
in Urban China.” World Bank Policy Note, Washington, DC.
Huang, Hongfang. 2010. “Jiangsu Province Has Realized the Full Coverage of the New Rural Pen-
sion Scheme.” Xinhua News Daily, October 29. (In Chinese.)
Li, Wei. 2011. “Survey on the Implementation of RPPP.” Economic Review 6: 80–83. (In Chinese.)
Liang, Hong. 1999. “Economic Analysis of Current Rural Community Security of China,” Ph.D. 
diss., University of Fudan. (In Chinese.)
Liu, Junwei. 2011. “Analysis with Rational Choice Theory on the Determinants for the Enrollment 
Behaviors of Farmers into the RPPP.” Zhejiang Social Sciences Journal 4: 77–83. (In Chinese.)
Liu, Shangkui, Zhiqihui Wu, and Shengcai He. 2011. “Issues Arising from the Implementation 
of the RPPP and Countermeasures—A Survey on 5,000 Rural Households in Jilin.” Survey 
World 2: 30–33. (In Chinese.)
Lou, Jiwei, and Shuilin Wang, eds. 2008. Public Finance in China: Reform and Growth for a Harmo-
nious Society. Washington, DC: World Bank.
Luo, Xia. 2011. “Empirical Study on the RPPP: Case Study on Sixian City of Anhui Province.” 
Social Security Studies 1: 67–73. (In Chinese.)
Ma, Limin. 1999. “Slowing Down Rural Pension.” Exploration and Debate 7: 11–12. (In Chinese.)
Mu, Liangze, and Ren Lu. 2010. “Policies Beneficial for Farmers and Restructuring for the Rural 
Management Resources—Analysis on the RPPP.” Theory and Reform 6: 71–74. (In Chinese.)
NBS (National Bureau of Statistics). 1995–2011. China Labor Statistical Yearbook. Beijing: China 
Statistics Press. (In Chinese.)
Rural Research Center of the Central China Normal University. 2010. Survey. http://www.
juejingushi.com/znews/vvfnews-106026. (In Chinese.)
Shi, Shih-Jiunn. 2006. “Left to Market and Family—Again? Ideas and the Development of the 
Rural Pension Policy in China.” Social Policy & Administration 40 (7): 791–806.
State Council. 2009. “Guidelines of Developing the New Rural Pension Scheme Pilot.” (In 
Chinese.)
Wen, Jiabao. 2012. “Report on the Work of the Government.” Delivered at the 5th Session of the 
11th National People’s Congress, March 5. http://news.xinhuanet.com/english/china/2012-
03/15/c_131469703.htm.
Wang, Dewen. 2006. “China’s Urban and Rural Old Age Security System: Challenges and 
Options.” China & World Economy 14 (1): 102–16.
Wang, Dewen, Jiaze Chen, and Wenshu Gao. 2011. “Social Security Integration: The Case of 
Rural and Urban Resident Pension Pilot in Chengdu.” World Bank, Washington, DC.

11.  CHINA’S PENSION SCHEMES FOR RURAL AND URBAN RESIDENTS 
241
Wang, Dewen, Philip O’Keefe, and Lawrence Thompson. 2012. “China’s Pension System Integra-
tion: The Case of Guangdong.” World Bank, Washington, DC.
Wang, Guojun. 2000. “Defects of Current Rural Pension System and Reform Thinking.” Quarterly 
of Shanghai Academy of Social Science 1: 120–27. (In Chinese.)
World Bank. Forthcoming. “China 2011. A Vision for Pension Policy Reform.” World Bank, 
Washington, DC.
Wu, Yufeng. 2011. “Empirical Studies on Social Interactions and Enrollment Behaviors of the 
RPPP.” Huazhong University Journal of Science and Technology 4: 105–111. (In Chinese.)
Wu, Yyuning. 2009. “Rural Pensions.” Background paper for the World Bank report on the rural 
elderly, Washington, DC.
Zhang, Wenjun, and Dan Tang 2008. “The New Rural Social Pension Insurance Program of 
Baoji City Shaanxi Province.” HelpAge International–Asia/Pacific, Chiang Mai, Thailand. 
www.helpage.org/download/4c48ab08674c1/. (In Chinese.)

243
CHAPTER 12