A review of international experience

% of contributors or affiliates among workers age 14–64
SOURCES: Costa and others 2011 for Mexico; authors’ calculations for Colombia and Peru, based on data from national 
household surveys.
NOTE: Figures includes affiliates who are paid a wage (sueldo) in Mexico and contributors in Colombia and Peru.

208 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
FIGURE 10.6  Pension coverage of middle-class nonagricultural workers in Colombia, Mexico, and 
Peru, by type of employment
% of workers age 14–64
0 
10 
20 
30 
40 
50 
60 
70 
80 
90 
formal independent, 
completed 
tertiary education 
informal
nonagricultural 
independent
nonagricultural 
Colombia (2009)
Mexico (2006)
Peru (2010)
SOURCES: Costa and others (2011) for Mexico; authors’ calculations based on national household surveys for Colombia and 
Peru.
NOTE: Figures are for affiliates who are paid a wage (sueldo) in Mexico and contributors in Colombia and Peru. Figures for 
Mexico include only workers who also earned wages and should therefore be interpreted with caution.
Informal middle-income workers are therefore the group on which matching defined 
contributions could have the greatest impact. This group numbers about 19.8 million 
workers in these three countries. On average, the remuneration of these workers is simi-
lar to the minimum wage in Colombia and Peru and up to twice the minimum wage in 
Mexico. Their labor income alone places them significantly above the national poverty 
threshold, suggesting that they could respond positively to incentives for pension savings.
FIGURE 10.5  Nonagricultural middle-income workers in Colombia, Mexico, and Peru by 
employment category 
Colombia (2009)
Mexico (2006)
Peru (2010)
% of workers age 14–64
0 
20 
40 
60 
80 
100 
independent nonagricultural 
informal nonagricultural 
independent, completed tertiary education 
formal 
SOURCES: Costa and others 2011 for Mexico; authors’ calculations based on national household surveys for Colombia and 
Peru.

10.  MATCHING CONTRIBUTIONS IN COLOMBIA, MEXICO, AND PERU: EXPERIENCES AND PROSPECTS 
209
Conclusions
Although Colombia, Mexico, and Peru introduced major structural reforms in the mid-
1990s, they still exhibit very limited levels of social protection coverage (28 percent of the 
labor force in Colombia, 27 percent in Mexico, and 22 percent in Peru). These figures 
reflect the limited effectiveness of basing a pension scheme on withholding of contribu-
tions by employers in environments in which most employment relations are informal.
A high level of labor informality in the region is the result of various factors, includ-
ing weak institutions; excessive regulation; rigid and complex employment legislation; 
the combination of minimum wages, relatively high employment benefits, and low labor 
productivity; imperfect land and water markets; the still recent history of high macroeco-
nomic volatility; and incentives to informality, whether explicit (such as public purchase 
programs, special tax regimes, subsidies, or programs to write off the debt of small farm-
ers) or implicit (exemption from employment, tax, and building inspections). This com-
bination of factors has had several worrisome consequences. It has led to a two-tier firm 
structure, made up of a large number of very small companies, which employ workers 
with low productivity, and a small number of very large companies, which employ work-
ers with high productivity. It has also incentivized small companies to maintain informal 
relations with their workers, because low labor productivity does not allow them to pay 
the minimum wage and provide benefits. It has made it more difficult for large companies 
to compete with informal enterprises (services, retail trade). As some employment legisla-
tion allows formal hiring temporarily without benefits (for example, the law promoting 
nontraditional exports in Peru), employers tend to choose this method of employment 
relation, reinforcing the low level of coverage of social security and the lack of employer 
investment in training employees.
Addressing the lack of social protection cannot rely solely on the formalization pro-
cess, which will be gradual. Several programs in the three countries aim to support savings 
for old-age income through ex post or ex ante incentives and subsidies. Ex post policies 
provide direct subsidies when benefits are calculated on the basis of the value of savings 
accounts at retirement or establish a noncontributory pension for people who reach the 
retirement age without having participated in a pension system. Rather than attempt to 
provide universal coverage, these kinds of programs could target specific groups. Imple-
mentation of ex post schemes should take account of their fiscal effects, the distortions 
they could generate in assigning resources, and the disincentives to saving that could result 
for groups that are currently saving or who have the potential to do so.
Ex ante schemes provide a preferable alternative, not only because they reduce dis-
tortions and are more manageable from the fiscal point of view but also because they 
address the basic problem of limited participation and long-term savings within the pen-
sion system. In order to design an adequate reform, it is important to identify the limita-
tions of the existing mandatory schemes, study the level and types of informality, and take 
into account the inherent economic conditions potential savers face. These conditions 
include low productivity, high exposure to economic shocks, difficulty in accumulating a 
continuous employment history, the need to cover basic shorter-term demands (for hous-
ing, education, and health services), and others.

210 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
MDC schemes could be appropriate for some workers who are not currently making 
contributions to the pension system, provided that the monetary transfer by the state is 
sufficiently attractive to encourage saving. An MDC scheme could be of significant value 
if it includes certain features, such as allowing funds to be withdrawn to purchase a first 
home, access health services, or weather negative shocks to short-term income.
Colombia, Mexico, and Peru have very limited experience with matching-type 
schemes, which have focused mainly on increasing the savings of groups of workers who 
are already participating in some part of the pension system. These schemes do not seem 
attractive enough to people who are not contributing or are not even affiliated.
Future reform efforts should target groups that have the capacity to save and might 
respond to government incentives. The emerging middle class, among whom pension 
coverage remains very low, is beginning to show some capacity for saving. The nearly 
20 million informal middle-income workers in Colombia, Mexico, and Peru could well 
benefit from MDC schemes.
Notes
1.  In Mexico, workers belonging to the old pay-as-you-go system until 1997 who were trans-
ferred to the new funded system could choose the form in which their pension was calculated, 
based on either the rules of the pay-as-you-go system or the savings they had accumulated in 
their individual accounts.
2.  For a complete review of the origin of the reforms in Latin America, see Gill, Packard, and 
Yermo (2004); Holzmann and Hinz (2005); Tuesta (2011a); and World Bank (1994).
3.  The term informality covers a range of employment, from illegal activity to exchanges that are 
outside formal or contractual environments, such as mutual aid between neighbors.
4.  One example of this type of scheme is the recently approved Pensión 65 plan in Peru, aimed at 
very poor people age 65 and older. Bjeletic and Tuesta (2010) review other schemes.
5.  Independent workers in Colombia are required to contribute to both pension and health 
insurance schemes.
6.  The minimum pension in Colombia is the legal minimum wage, which is annually discussed 
in a tripartite commission made up of representatives from the government, unions, and 
employers. If they do not reach agreement, the wage is set by a government decree. 
7.  The public defined benefit pay-as-you-go scheme (RPM) does not have a similar fund. The 
reserves of the social security were depleted in 2004. When a worker completes the number 
of weeks of contribution required and reaches the age of retirement, the government pays the 
minimum pension through a transfer from the budget.
8.  Community mothers are women hired by the Welfare Family Institute to provide child care 
in a community.
9.  The age for retirement is 62 for men and 57 for women.
10. According to the Identification System of Potential Beneficiaries of Social Programmes 
(Sistema de Identificación de Potenciales Beneficiarios de Programas Sociales), a tool that clas-
sifies individuals according to their living standard. It is used to objectively select potential 
beneficiaries of social programs managed by the state.
11.  Legislative Act 01 of 2005 and Decreto 4944 of December 18, 2009.

10.  MATCHING CONTRIBUTIONS IN COLOMBIA, MEXICO, AND PERU: EXPERIENCES AND PROSPECTS 
211
12.  Among the incentives policy makers are considering is the possibility of allowing the funds to 
be used as a guarantee for obtaining credit, including for housing.
13.  The progressive incremental scheme starts with a daily amount equivalent to Mex$57 ($4.35) 
for workers earning up to 1 minimum wage (MW); Mex$54 ($4.16) for workers earning 
1–4 MWs; Mex$52 ($3.98) for workers earning 4–7 MWs; Mex$50 ($3.80) for workers earn-
ing 7–10 MWs; and Mex$47 ($3.62) for workers earning 10–15 MWs.
14.  Law No. 29903 takes into account various modifications of the private pension system, such 
as industrial organization, investment regime, corporate governance, and law enforcement, 
among others. The Welfare Pension System is one of the issues considered by this law.
15.  Empirical studies on poverty and income distribution often use half the average income as the 
poverty line. Middle-classes workers represent 41 percent of workers in Peru, 50 percent in 
Colombia, and 55 percent in Mexico.
16.  The figure for Colombia (the number of contributors) comes from the 2009 Great Integrated 
Household Survey, which covered 13 cities. The figure for Mexico (number of affiliates receiv-
ing a wage) comes from the 2006 National Household Income and Expenditure Survey. The 
number for Peru (number of contributors who declared having made contributions during 
the year) comes from the 2010 National Household Survey. See Costa and others (2011) 
for a detailed description of the methodology and an analysis of pensions and informality in 
Bolivia, Brazil, Chile, and Mexico.
17.  Using this criterion to define formality facilitates international comparisons, as it reflects a 
form of regulation that is common to Latin American countries: the obligation to formalize 
and register an employment relationship.
References
Acosta, O., and J. Ramírez. 2004. “Las redes de protección social: modelo incompleto.” Develop-
ment Finance Series, Economic Commission for Latin America and the Caribbean.
Bjeletic, J., and D. Tuesta. 2010. “Pension Reform in Peru.” In Pension Reforms in Latin Amer-
ica: Balance and Challenges Ahead, ed. Eduardo Fuentes, Alicia García Herrero, and José Luis 
Escrivá. Madrid: BBVA. http://www.bbvaresearch.com/KETD/fbin/mult/Pensionreformsin-
latinamerica_tcm348-238550.pdf?ts=2292012.
Breceda, K., Rigolini, J., and Saavedra, Jaime. 2008. “Latin America and the Social Contract: Pat-
terns of Social Spending and Taxation.” Policy Research Working Paper Series 4604, World 
Bank, Washington, DC.
Cardoso, Miguel. 2007. “Pension Systems and Incentives for Independent Workers: An Analysis 
for the Chilean Economy.” Presentation prepared for the Latin American and Caribbean Eco-
nomic Association and Latin American Meeting of the Econometric Society, parallel meeting, 
Bogota, October 5.
Carranza, L. 2012. “La Ley Mypes 2008 en el Perú.” Draft note. Universidad San Martin de 
Porres-USMP.
Carranza, L., J. Chávez, and J. Valderrama. 2006. “The Political Economy of the Budget Process: 
the Peruvian Case.” IADB Publications 15178, Washington, DC, Inter-American Develop-
ment Bank.
Chong, A., J. Galdo, and J. Saavedra. 2007. “Informality and Productivity in the Labor Market: 
Peru 1986–2001.” RES Working Papers 4526, Inter-American Development Bank, Research 
Department.

212 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
Costa, R. Da, J. R. de Laiglesia, E. Martinez, and A. Melguizo. 2011. “The Economy of the Pos-
sible: Pensions and Informality in Latin America.” Working Paper 295, Organisation for Eco-
nomic Co-operation and Development Development Center, Paris.
Departamento Nacional de Planeación. 2009. “Requisitos de Acceso al Programa Subsidiado de 
Aporte a la Pensión Financiado con los Recursos de la Subcuenta de Solidaridad del Fondo de 
Solidaridad Pensional.” Documento Conpes 3605, Consejo Nacional de Política Económica y 
Social República de Colombia, Bogotá.
De Soto, H. 1989. The Other Path: The Invisible Revolution in the Third World. Harpercollins.
ECLAC (Economic Commission for Latin America and the Caribbean). 2006. “La protección 
social de cara al futuro: acceso, financiamiento y solidaridad.” ECLAC, Santiago.
Forteza, A., L. Luchetti, and M. Pallares-Miralles. 2009. “Measuring the Coverage Gap.” In Closing 
the Coverage Gap: The Role of Social Pensions and Other Retirement Income Transfers, ed. R. Hol-
zmann, D. A. Robalino, and N. Takayama. Washington, DC: World Bank.
Fuentes, E. 2010. “Creating Incentives for Voluntary Contributions to Pension Funds by Indepen-
dent Workers: An Informal Evaluation Based on the Case of Chile.” 10/12, BBVA Research 
Working Papers. Madrid: BBVA.
Gill, I., T. Packard, and J. Yermo. 2004. “Keeping the Promise of Social Security in Latin America.” 
World Bank, Washington, DC.
Herrera, C. 2010. “Towards Stronger Pension Systems in Mexico: Vision and Proposals for 
Reform.” In Pension Reforms in Latin America: Balance and Challenges Ahead, ed. Eduardo 
Fuentes and others. Madrid: BBVA.
Holzmann, R., and R. Hinz. 2005. “Old Age Income Support in the 21st Century.” World Bank, 
Washington, DC.
Holzmann, R., D. A. Robalino, and N. Takayama. 2009. Closing the Coverage Gap: The Role of 
Social Pensions and Other Retirement Income Transfers. Washington, DC: World Bank.
Hu, Y., and F. Stewart. 2009. “Pension Coverage and Informal Sector Workers: International Expe-
riences.” Working Papers on Insurance and Private Pensions 31, Organisation for Economic 
Co-operation and Development, Paris.
Jütting, J. P., and J. R. de Laiglesia, eds. 2009. “Is Informal Normal? Towards More and Better Jobs 
in Developing Countries.” Organisation for Economic Co-operation and Development, Paris.
Levy, S. 2008. Good Intentions, Bad Outcomes. Social Policy, Informality and Economic Growth in 
Mexico. Washington, DC: Brookings Institution Press.
Llanes, M., and J. Alonso. 2010. “Confidence in the Future: Proposals for an Improved Pension 
System in Colombia.” In Pension Reforms in Latin America: Balance and Challenges Ahead, ed. 
Eduardo Fuentes and others. Madrid: BBVA.
Loayza, N. 2007. “The Causes and Consequences of Informality in Peru.” Working Papers 2007-
018, Banco Central de Reserva del Perú.
Loayza, N., L. Serven, and N. Sugawara. 2009. “Informality in Latin America and the Caribbean.” 
Policy Research Working Paper Series 4888, World Bank, Washington, DC.
Lopez-Calva, L. P., and N. Lustig, eds. 2010. Declining Inequality in Latin America: A Decade of 
Progress? Baltimore: Brookings Institution Press and United Nations Development Programme.
Mitchell, O., and S. Utkus. 2004. “Lessons from Behavioral Finance and Retirement Plan Design.” 
In Pension Design and Structure: New Lessons from Behavioral Finance, ed. Olivia Mitchell and 
Stephen Utkus. Oxford: Oxford University Press.

10.  MATCHING CONTRIBUTIONS IN COLOMBIA, MEXICO, AND PERU: EXPERIENCES AND PROSPECTS 
213
OECD (Organisation for Economic Co-operation and Development). 2010. Latin American Eco-
nomic Outlook 2011: How Middle-Class Is Latin America? Paris: OECD Development Center.
—. 2011. “Pensions at a Glance 2011: Retirement-Income Systems in OECD and G20 
Countries.” www.oecd.org/els/social/pensions/PAG.
Oviedo, A. M. 2009. “Economic Informality: Causes, Costs, and Policies: A Literature Survey of 
International Experience” Draft background paper prepared for Turkey Country Economic 
Memorandum—Informality: Causes, Consequences, Policies. World Bank, Washington, DC. 
Pagés, C. 2010. The Age of Productivity: Transforming Economies from the Bottom Up. New York: 
Inter-American Development Bank and Palgrave Macmillan.
Perry, G., W. Maloney, O. Arias, P., Fajnzylber, A. Mason, and J. Saavedra-Chanduvi. 2007. Infor-
mality: Exit and Exclusion. Washington, DC: World Bank.
Ribe, H., D. A. Robalino, and I. Walker. 2010. From Right to Reality: Achieving Effective Social Pro-
tection for All in Latin America and the Caribbean. Washington, DC: World Bank.
Robalino, D. A., and R. Holzmann. 2009. “Overview and Preliminary Policy Guidance.” In Clos-
ing the Coverage Gap: The Role of Social Pensions and Other Retirement Income Transfers, ed. 
R. Holzmann, D. A. Robalino, and N. Takayama. Washington, DC: World Bank.
Rofman, R., L. Lucchetti, and G. Ourens. 2008. “Pension Systems in Latin America: Concepts 
and Measurements of Coverage.” Social Protection and Labor Discussion Paper 0616, World 
Bank, Washington, DC.
Rofman, R., and M. L. Oliveri. 2012. “La cobertura de los sistemas previsionales en América 
Latina: Conceptos e indicadores.” Serie de Documentos de Trabajo sobre Políticas Sociales 
No. 7, World Bank, Washington, DC.
Schneider, F., A. Buehn, and C. Montenegro. 2010. “Shadow Economies All over the World: New 
Estimates for 162 Countries from 1999 to 2007.” Working Paper No. 322, University of 
Chile, Department of Economics.
Secretaria de Desarrollo Social. 2006. “Decreto por el que se establece el Mecanismo de Ahorro 
para el Retiro Oportunidades.” March 27, 2006, Diario Oficial.
Selnow, G. 2004. “Motivating Retirement Planning: Problems and Solutions.” In Pension Design 
and Structure: New Lessons from Behavioral Finance, ed. Olivia Mitchell and Stephen Utkus. 
Oxford: Oxford University Press.
Thaler, Richard, and Cass Sunstein. 2009. Nudge: Improving Decisions about Health, Wealth, and 
Happiness. New York: Penguin Books.
Tuesta, D. 2011a. “A Review of the Pensions Systems in Latin America.”11/15 BBVA Research 
Working Papers.
—. 2011b. “Matching Arrangements in Mexico, Colombia, and Peru.” Presentation, MDC 
Conference, World Bank, June 6.
Turner, A. 2011. “Planes privados de pensiones en México: situación actual y perspectivas.” Presen-
tation by Instituto Tecnológico Autónomo de México, Ninth Conference of the International 
Center for Pension Research, April.
World Bank. 1994. Averting the Old Age Crisis. Washington, DC: Oxford University Press.
—. 2012. Doing Business in a More Transparent World. Washington, DC: World Bank.

PART IV
Developing 
Country Experience

217
CHAPTER 11
China’s Pension Schemes for 
Rural and Urban Residents
Mark C. Dorfman, Dewen Wang, Philip O’Keefe, and Jie Cheng
China’s recent establishment of national pension schemes for its rural and urban residents 
represents a major step toward its ambitious objective of universal pension coverage. The 
design includes a locally financed matching subsidy linked to a flat defined benefit pen-
sion after a vesting period or through “family binding” provisions requiring retirees’ adult 
children to contribute. The policy design supports broad coverage, in part by providing 
modest basic benefits and a channel for an integrated scheme that can accommodate the 
wide variation in economic needs and circumstances across China. The benefit level and 
financing plan balance benefit adequacy against a very modest fiscal commitment while 
enabling local authorities to supplement benefits. The design may be able to broaden cover-
age, but its benefits may not be sufficient to provide adequate financial protection in old age 
for substantial numbers of beneficiaries. Looking forward, universal pension coverage at the 
individual level will depend in large part on the incentives for sustained participation and 
the authorities’ ability to enhance portability and eventually integrate pension programs 
across schemes and across space, and pool risk.
S
ince economic liberalization in the early 1980s, China has faced substantial dispar-
ities in pension coverage between urban and rural areas and across regions, and it 
has implemented multiple reforms of its pension system. As in many former command 
economies, labor force coverage of China’s contributory pension scheme declined sharply 
in the 1980s. After experimenting with a number of pilot programs in the early 1990s, 
the government established a national framework for urban contributory pensions, albeit 
one allowing for regional variation. In the last few years, the authorities have articulated 
principles for reform, including broad coverage, basic protection, multilayering (urban 
systems), flexibility (rural systems), and sustainability. These principles underpin the com-
mitments made at the 17th Party Congress toward a comprehensive and integrated social 
security system by 2020.
As part of its ongoing social protection reforms, in September 2009 the authori-
ties established a national framework for rural pensions, the Rural Pension Pilot Scheme 
(RPPS), now the National Rural Pension Scheme (NRPS); in July 2011, it introduced 
an Urban Resident Pension Scheme (URPS). Both schemes fall under the broader 
umbrella of the 2010 Social Insurance Law. These voluntary schemes include a match-
ing defined contribution element and a heavily subsidized flat basic pension. Both 
schemes are being rapidly rolled out, with a target of full national geographic coverage 
by 2013. Early numbers on participation are impressive: by early 2012, the NRPS had 
more than 250 million contributors, and more than 100 million people were receiving 
basic pensions.

218 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
China’s introduction of the NRPS and URPS has arguably been one of the world’s 
most ambitious voluntary pension saving and minimum elderly assistance schemes in 
a low- or middle-income country. The sheer scale and speed of coverage expansion has 
already exceeded the initial targets of the Chinese authorities.
The design supports broad coverage as a core objective. Consistent with this objective, 
minimum contributions are low, benefits are very modest, and fiscal costs are low for all 
but the poorest local communities. The design links a basic defined benefit to a minimum 
contribution history. In a unique variation, it also includes a “family binding” provision, 
whereby the pension eligibility of a contributor already over pensionable age is determined 
by whether the contributor’s adult children contribute to the new system. The financing 
structure provides the flexibility for communities with additional resources to provide more 
generous benefits while guaranteeing a minimum level of basic benefits through central 
government financing. Although initial participation levels suggest considerable momen-
tum for registration, only time will tell if such participation levels are sustained.
This chapter is organized as follows. The first section provides a historical context 
for the national guidelines established in 2009 for urban and 2011 for rural areas. The 
second section outlines the design features of the reforms. The third section assesses the 
design elements and initial experience of the reforms, including benefit adequacy, worker 
and fiscal affordability, and institutional issues. The last section evaluates the lessons from 
the Chinese experience and considers their applicability to other countries.
Evolution of Pension Provision for Rural and Urban Residents
PENSION PROVISION IN RURAL AREAS
The pension system in rural areas has experienced two milestones. The first was the old 
rural pension scheme, introduced by the Ministry of Civil Affairs (MOCA) in 1992. The 
second is the new rural pension scheme, introduced by the Ministry of Human Resources 
and Social Security (MHRSS) in 2009.
1
 Both schemes built on a typical Chinese policy 
process of bottom-up local experimentation, which informed the development of national 
guidelines. The national guidelines in turn attempted to impose greater standardization 
of scheme design and management, with only partial success in the case of the old system. 
Perhaps the major distinction between the old and new schemes is the significant financ-
ing commitment of the central authorities in the new scheme, which for the first time 
provides a strong incentive to local authorities to follow a nationally harmonized scheme 
design.
The old pension scheme experienced steady, if unspectacular, expansion in the 
1990s, before coverage growth stagnated and the scheme design fragmented beginning in 
the late 1990s before the establishment of the new rural pension scheme. This evolution 
unfolded in various stages.
Initiation and Expansion of Rural Pensions: 1986–98 
The initial policy directions on rural pensions were set out in the Seventh Five-Year Plan 
(1985–90), in which the Chinese government committed to exploring ways to provide 

11.  CHINA’S PENSION SCHEMES FOR RURAL AND URBAN RESIDENTS 
219
basic pensions for rural residents.
2
 The plan was built upon through county-level pilots 
led by MOCA in the Beijing and Shanxi provinces, which emphasized mixed financing 
responsibility by government, collectives, and individuals, with most of the responsibility 
falling on individuals.
There was a clear policy push in the early 1990s to expand rural pension schemes, 
with the large majority following a simple defined contribution design with a small sub-
sidy from the collective in some areas. The initial pilots were followed in 1991 by mea-
sures assigning MOCA the lead role on rural pensions and the establishment of pilots in 
five counties of Shandong Province. The various pilots contributed to a major MOCA 
policy document in 1992—”The Basic Approach of Rural Pension Schemes at County 
Level (Trial)”—that outlined a set of principles for rural pension schemes while allow-
ing for regional variations. As a result of this policy initiative, coverage of rural pen-
sions expanded significantly during the 1990s, with programs established by the end of 
1998 in 31 provinces and 2,123 counties (just under three-quarters of all rural coun-
ties). This period was also the high point of coverage before the NRPS, with more than 
80 million rural workers contributing and more than 600,000 people receiving pensions 
(figure 11.1). 
FIGURE 11.1  Rural pension participation and coverage, 1994–2010
0
5
10
15
20
25
30
0
20
40
60
80
100
120
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 
coverage (millions)