A review of international experience

Colombia Mexico 
Peru 
ratio of
contributors/labor force (%)
before reform 
2010
SOURCE: National supervisors.
NOTE: Year of reform is 1993 for Colombia, 1996 for Mexico, and 1992 for Peru.
The willingness of individuals to contribute voluntarily to a pension system is a 
significant, albeit less readily quantifiable, factor in determining the observed coverage 
levels in these countries. Assuming that people at all levels of income have some capacity 
to consume less and allocate resources for old-age savings, participation in any retirement 
savings system will to a large extent depend on the features of the product offered and 
individuals’ perceptions of the need to create a formal (as opposed to a family-based) 
source of support when they are elderly. The willingness to participate in a pension system 
is also known to increase with age.
An important factor in determining the demand for pension products may also 
be the type or level of immediate monetary incentives, the most prevalent of which are 
matching defined contributions (MDCs). These incentives can come from either public 
or private sources. They represent an important policy instrument, in combination with 
other interventions, to encourage people in groups that are currently not participating in 
pension systems to begin saving for retirement.
This chapter reviews the attempts by Colombia, Mexico, and Peru to develop 
MDC-type pension schemes. It also looks at the extent to which such schemes could be 
extended to groups that are currently not contributing to the system. The chapter begins 
by analyzing the reasons why MDC schemes may be relevant in Latin America, focusing 
on the three countries. It then reviews the experience of MDC schemes in the three coun-
tries, including projects that have been implemented, regulatory development in progress, 
and other initiatives that have been abandoned for the time being. The following section 

10.  MATCHING CONTRIBUTIONS IN COLOMBIA, MEXICO, AND PERU: EXPERIENCES AND PROSPECTS 
195
uses household survey data from Colombia, Mexico, and Peru to consider the potential 
for developing MDC-type schemes and identify groups in the labor force that could be 
the focus of such an initiative. The last section draws some conclusions.
Matching Schemes and Structural Problems behind Low Pension 
Coverage in Latin America
Latin America has low coverage in social insurance and social protection, particularly in the 
case of pensions (Rofman, Lucchetti, and Ourens 2008; Rofman and Oliveri 2012). Even 
though there is a strong correlation (about 80 percent) between the level of coverage and 
gross domestic product (GDP) per capita (Forteza, Luchetti, and Pallares-Miralles 2009), 
with the exception of Chile, Latin American countries exhibit pension coverage rates that 
are below those of other countries with similar per capita GDP levels (figure 10.2). 
FIGURE 10.2  Pension coverage as a percentage of the labor force and GDP per capita in selected 
countries, early 2000s
Mexico 
Colombia 
Peru 
Norway 
Ireland 
United States 
0 
20 
40 
60 
80 
100 
0  10,000 20,000 30,000 40,000 50,000 60,000 70,000 80,000 
coverage
(% of labor force) 
per capita GDP ($)
Chile 
SOURCES: Holzmann, Robalino, and Takayama 2009; World Bank Development Indicators database. 
THE CHALLENGE OF INFORMALITY
A number of factors lie behind the problem of low pension coverage in the region. One 
is the capacity to save, which is closely related to per capita income, among other fac-
tors (Costa and others 2011; Ribe, Robalino, and Walker 2010; Robalino and Holzmann 
2009; Tuesta 2011a). A second is government inefficiency in various areas (Carranza, 
Chávez, and Valderrama 2006). Behind the problem of low income generation are poor 
economic management and ill-conceived growth strategies, a deficient institutional struc-
ture, and inadequate policies for dealing with poverty, health, education, and sex (Acosta 
and Ramírez 2004). A third, particularly serious, problem within the institutional arena 
is the extent of the informal economy, which makes it infeasible to impose a mandate 
for pension contributions because the state cannot enforce compliance (Costa and others 
2011; Levy 2008; Tuesta 2011b). As shown in figure 10.3, there is a negative correlation 

196 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
between the level of labor informality and pension coverage rates. Peru is ranked among 
the economies with the highest levels of labor informality in the world.
FIGURE 10.3  Labor informality and pension coverage in selected countries, early 2000s
Mexico 
Colombia 
Peru 
0 
20 
40 
60 
80 
100 
informality (% GDP) 
coverage (% of labor force)
Chile 
SOURCES: Holzmann, Robalino, and Takayama 2009; Schneider, Buehn, and Montenegro 2010.
Several factors explain the high levels of labor informality in Colombia, Mexico, and 
Peru (for a survey of the empirical evidence on labor informality in Latin America, see 
Perry and others 2007):
 
• Low institutional quality and burdensome regulation. Expensive and com-
plicated procedures for creating and winding up firms, difficulties in obtaining 
licenses and permits, registration problems, and inadequate protection of prop-
erty rights create an environment that is not favorable for doing business or estab-
lishing new companies (Breceda, Rigolini, and Saavedra 2008; Loayza, Serven, 
and Sugawara 2009; World Bank 2012).
 
• Rigid and complex labor legislation. The combination of minimum wages, 
relatively high employment benefits, low labor productivity, and restrictions on 
hiring and firing is conducive to labor informality (Breceda, Rigolini, and Saa-
vedra 2008; Carranza 2012; Chong, Galdo, and Saavedra 2007; Loayza, Serven, 
and Sugawara 2009).
 
• Imperfect land and water markets. In areas where the land market does not 
operate properly, it is impossible to accumulate land. In areas where water rights 
are not transparently assigned, landownership is concentrated among farmers 
with small holdings or in farming communities. The result is inefficient farming 
operations with very low productivity and without the capacity to generate levels 
of income that enable workers to participate in a social security system (De Soto 
1989; Loayza 2007; Oviedo 2009).
 
• Recent history of high macroeconomic volatility. Macroeconomic problems 
include volatile prices, unexpected changes in the tax system, erratic demand, and 
restrictions on capital flows and foreign exchange regimes that limit the devel-
opment of financial markets. All these factors have meant that companies can-
not easily or predictably comply with their social security or tax obligations. The 

10.  MATCHING CONTRIBUTIONS IN COLOMBIA, MEXICO, AND PERU: EXPERIENCES AND PROSPECTS 
197
greater macroeconomic stability achieved over the last decade has improved aver-
age incomes, developed capital markets, and led to a steady reduction in labor 
informality, but formalizing the economy remains a very slow process (Loayza 
2007).
 
• Explicit or implicit incentives to informality. The significant number of people 
who remain in the informal economy generates the political temptation to offer 
financial assistance and subsidies, either explicit (such as public purchase pro-
grams, special tax schemes, subsidies, or programs to write off agricultural debt 
for small producers) or implicit (exemptions from labor and tax inspection, spe-
cial application of regulations to physical establishments where informal activities 
are being carried out) (Carranza 2012; Loayza 2007; Oviedo 2009).
This high rate of labor informality, particularly in rural areas, has generated the fol-
lowing structural characteristics:
 
• A two-tier business structure, in which many very small companies with low labor 
productivity account for a high percentage of jobs and a few very large companies 
with high labor productivity account for a high percentage of output
 
• Small companies that have informal relations with their workers, because low 
labor productivity does not allow them to pay minimum wages or provide other 
types of required benefits
 
• Large companies that have formal relations with their workers but encounter 
problems when they attempt to expand to take advantage of economies of scale 
in sectors in which they compete with small informal enterprises (services, retail 
trade)
 
• Employment legislation that allows temporary formal employment without 
benefits; this form of employment tends to be popular among both employers 
(because of the potentially lower labor costs and more flexible ability to termi-
nate workers) and workers (who usually choose to receive higher cash income in 
the short term by not participating in social security systems, even when there is 
typically greater overall value over the longer term in establishing eligibility for 
the benefits), and affects the level of social security coverage and the employer’s 
incentives to invest in training.
Although macroeconomic growth and stability in Latin America are generating con-
ditions conducive to formalizing the economy—and Colombia, Mexico, and Peru have 
been examples of stability and growth for more than two decades—formalization is a very 
gradual process. Moreover, growth alone is not enough to achieve it. Structural reforms 
are required to introduce greater flexibility to factor markets, establish uniform and simple 
tax systems, and make it easier to do business. These reforms may be delayed or simply 
not implemented for political economy reasons.
Formalization is important because, like many industrial economies, Colombia, 
Mexico, and Peru have established systems that use employers as contractual agents to 
ensure social security coverage, mandating participation as a condition of employment 
and collecting contributions through deductions from wages. Informal firms—or firms 
that maintain informal relations with their employees—do not act as agents of the state 

198 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
in enforcing compliance with the law. As Latin America ages, the high level of informality 
is leaving a growing number of people without social protection (in terms of pensions, 
health, and unemployment).
CAN A MATCHING SCHEME INCREASE COVERAGE?
Given these structural difficulties, some countries have given up on looking for schemes 
that oblige or create incentives for workers to contribute to a social protection scheme 
during their active lives. Instead, they have opted for ex post solutions aimed at providing 
means-tested noncontributory pensions for the elderly or people who do not otherwise 
qualify to receive a pension.
4
Such solutions are problematic for several reasons. First, if the rate of informality 
remains high and life expectancy continues to rise, public expenditure will face significant 
political pressures in the future. Second, ex post solutions reinforce the economic incen-
tives for not moving into the formal economy and impose rigid thresholds that could cre-
ate disincentives for saving by people who might otherwise be capable of and inclined to 
do so. Third, to the extent that there is a social program with broad groups that are actual 
or potential beneficiaries, there will be considerable political pressure for new initiatives 
and additions to these programs, which will extend coverage and increase benefits. Devel-
opment of these social pensions is thus not a long-term solution for the problem of cover-
age related to the capacity and desire to save during one’s active working life (Levy 2008).
The most adequate way to extend coverage in the long term is to find new mecha-
nisms that can increase participation in saving for old age during the period of active 
employment. Mandatory contribution schemes could play a role, although they have clear 
limitations in a context of high levels of informality that are likely to be meaningfully 
reduced only over the long term. A smarter proposal could therefore be to promote volun-
tary saving by a broad range of groups that are not currently saving by trying to identify 
the factors that could encourage them to participate.
In a context of limited fiscal resources, particularly in medium- and low-income 
countries such as Colombia, Mexico, and Peru, the aim should be to identify groups that 
require government assistance for saving for old age and that are more likely to respond to 
incentives. Among people who currently do not participate in mandated pension schemes, 
some independent workers with high income have sufficient resources to make contribu-
tions. For such workers, it is probably a good idea to develop strategies or implement 
changes to the law to make participation mandatory.
The situation differs for low- and middle-income independent workers who have 
low productivity and unstable relationships with informal enterprises, which makes it dif-
ficult for them to generate income, thus limiting their possibility to save. This large group 
will probably represent the biggest challenge for the government in terms of designing 
incentives, considering that the incentives to remain in the informal sector and to not 
contribute are great and difficult to counteract.
Some workers in this group are probably in a better position than others and may be 
able to save for old age if the incentives are appropriate and well designed. MDC-type sys-
tems that have significant government support and provide meaningful economic incen-
tives that complement the savings of workers who decide to contribute voluntarily are 
an option. It is important that incentives be sufficiently attractive to encourage pension 

10.  MATCHING CONTRIBUTIONS IN COLOMBIA, MEXICO, AND PERU: EXPERIENCES AND PROSPECTS 
199
saving among low-income groups, where the opportunity cost with respect to other alter-
natives is usually high (Mitchell and Utkus 2004). Not only the willingness to save but 
also the adequacy and design of a financial product (that will be used only when people 
are old enough to retire) are needed. Saving for pensions tends not to be very popular, 
because it absorbs limited current resources in return for benefits at a distant point in 
time, which some people will never reach (Selnow 2004).
An MDC scheme may be relevant if it interacts with other incentives that help infor-
mal workers overcome other needs that may be more urgent in the short term. Some pen-
sion systems that have considered using MDC schemes have complemented them with 
the possibility of using pension funds to help purchase a first home or access health care. 
Other schemes have also allowed access to these savings when people suffer negative shocks 
to their short-term income or experience liquidity needs because of illness or other factors.
MDC schemes should be considered a mechanism for encouraging saving as part of 
a set of broader instruments. Particularly in the cases of Chile and Colombia, the prob-
ability of household saving increases when saving offers access to the future acquisition 
of a home, health care services, or improved educational opportunities (Cardoso 2007; 
Fuentes 2010). 
REVIEW OF MATCHING CONTRIBUTION SCHEMES IN COLOMBIA, MEXICO, AND 
PERU
Pension legislation in Colombia, Mexico, and Peru has addressed the problem of low cov-
erage through a variety of different initiatives. Each approach has to take into account the 
specific characteristics of the country and respond to the situations of workers both dur-
ing their working life (ex ante) and at the time of retirement (ex post). In general terms, 
reforms must provide a comprehensive response to social, employment, and macroeco-
nomic aspects. At the same time, although reducing labor informality should continue to 
be a goal and incentives may be introduced for this purpose, transformations must focus 
on guaranteeing adequate and sustainable pensions for the population (OECD 2010; 
Ribe, Robalino, and Walker 2010). 
Given that employment informality is widespread and old-age poverty is persistent 
in Latin America, reinforcing the solidarity pillar in the short term is an inevitable first 
step in strengthening the retirement income system. The Inter-American Development 
Bank has proposed establishing a general social protection system financed through con-
sumption taxes (Levy 2008; Pagés 2010); the Economic Commission for Latin America 
and the Caribbean proposes extending existing minimum pensions (ECLAC 2006). One 
way of strengthening this pillar could be to reduce the number of years of contributions 
required to obtain a minimum contributory pension. Another would be to introduce 
social noncontributory pensions, although this option would require the commitment of 
significantly more public resources (1–2 percent in the case of very small cash transfers to 
the elderly population in poverty according to estimates). As broad fiscal commitment to 
a basic noncontributory pension could become a strong disincentive to formality, it may 
be advisable for minimum pensions to be directly linked to contributions up to a certain 
level, as they are in Chile.
One of the few ways of providing a solution to the lack of coverage that is sus-
tainable socially and economically in the long term is to increase the savings of people 

200 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
working now. Ex ante policies thus appear to offer more room for maneuvering toward a 
reform of pension systems.
The most direct political option is to make it mandatory for self-employed work-
ers to be members of the system. Such workers represent about half of informal urban 
middle-income workers in Colombia, Mexico, and Peru (8.8 million, of which fewer than 
0.7 million have completed tertiary education).
5
 However, the case of Brazil, which has 
introduced mandatory payment, shows that coverage continues to be unequally distrib-
uted, as high-income independent workers contribute significantly more than low- and 
middle-income ones. Clearly, it is not easy to introduce this policy effectively (Costa and 
others 2011). Ensuring that workers make contributions is difficult, and requiring them 
to do so may not be the best or a sufficient way to provide for their old age, given their 
limited capacity for saving. Some countries have considered a semi-mandatory hybrid 
option, in which workers are automatically registered but may decide to quit the sys-
tem. This option may be accompanied by modifications to adapt to special cases, such as 
greater flexibility in the amount and frequency of contributions or permission to with-
draw payments in some circumstances, such as long-term unemployment or health prob-
lems (Hu and Stewart 2009).
In recent years, the debate has begun to focus on the use of matching contribu-
tions to enhance participation in defined contribution pensions by introducing transfers 
from the government that are contingent on the level of voluntary contributions made to 
individual pension accounts. Unlike minimum pensions and social pensions, MDCs can 
have a significant impact on informal sector middle-income workers (Ribe, Robalino, and 
Walker 2010). However, the likely effect will depend on the size of the coverage problem 
and the factors behind it. 
The sources of finance of these incentives may be diverse, and they may differ in 
scope and form. The experiences in Colombia, Mexico, and Peru have been specific to 
each country’s circumstances. The fiscal prudence shown by their finance ministers has 
influenced the design and limited the progress of these programs. The often contrasting 
vision of experts and policy makers has also played a role in each country. In many cases, 
matching contribution projects, and even laws already approved, have not been imple-
mented because of a lack of consensus. As a result, MDC schemes in the three countries 
have so far been fairly limited; in no case have they ended up being generally applicable.
It is important to underscore that matching schemes—or any other incentive that 
seeks to increase participation without solving the underpinnings of labor informality—
will not be enough. Structural reforms (for example, in the labor market or on the regula-
tory side) beyond pension policy will be necessary to effect real change and spur workers 
to save for old age.
Colombia
Since the Act of 1993, when structural pension reforms were implemented, the Colom-
bian pension system has implemented two core mandatory schemes. One was the new 
private mandatory defined contribution scheme known as the Individual Savings System 
with Solidarity (Régimen de Ahorro Individual con Solidaridad—RAIS). The other was 
the pay-as-you-go public scheme known as the Average Premium System (Régimen de 
Prima Media—RPM). Workers must choose and contribute exclusively to one of these 

10.  MATCHING CONTRIBUTIONS IN COLOMBIA, MEXICO, AND PERU: EXPERIENCES AND PROSPECTS 
201
schemes. They contribute 16 percent of their base income, although the percentage allo-
cated to savings for old-age pensions is different in each scheme: 11.5 percent of income 
is allocated to the individual account in the RAIS; 13 percent goes into the RPM. The 
remaining amount is allocated to the payment of administrative costs in both schemes and 
to disability and survivors’ insurance, to the Minimum Pension Guarantee Fund (Fondo 
de Garantía de Pensión Mínima—FGPM) in the RAIS and to disability and survivors’ 
pensions in the RPM. In both systems, the employer is responsible for 75 percent and the 
employee for 25 percent of the contribution. Independent workers are responsible for all 
their own contributions to the system (Llanes and Alonso 2010).
Based on this structure, three MDC pension schemes can be identified in Colombia. 
Two already work as part of the RAIS and the RPM; the third, which is in the process of 
implementation, is intended to be part of a more general scheme aimed at boosting saving 
for old age outside of the formal pension system.
The two MDC pension schemes currently in operation are associated with contri-
butions to (1) the private individual accounts system through the FGPM of the RAIS 
and (2) the Pension Solidarity Fund (Fondo de Solidaridad Pensional—FSP), which is 
supported by contributions of workers from the RAIS and the RPM. The aim of the first 
is to benefit pensioners in the RAIS who contributed to their individual accounts but 
whose accumulated account value is insufficient to obtain a minimum pension under the 
requirements of the law (frequency of contributions and age).
6
 The matching mechanism 
for RAIS members operates ex post, by activating the use of the FGPM once the pensioner 
has used up his or her individual account fund. The FGPM is funded from 1.5 percent of 
the worker’s monthly income (part of the 16 percent contribution). Because the RAIS is 
fairly new, few pensioners have accessed the minimum pension guarantee.
7
The second type of matching scheme operates through the FSP, which is divided 
into two subaccounts. The first, the solidarity subaccount (a matching scheme), tempo-
rarily subsidizes contributions to the pension system for some groups deemed to have 
financial limitations. The second, the subsistence subaccount, which has been in place 
since 1993, is intended to provide protection against poverty in old age. For this account, 
a previous contributory history is not required (this subaccount is not a matching contri-
bution scheme). These subaccounts are financed by contributions of workers to the public 
(RPM) or private system (RAIS) who earn more than four times the minimum wage. The 
subsistence subaccount is also financed by the resources of contributors who earn more 
than 16 times the minimum wage.
The solidarity subaccount is designed exclusively to subsidize the monthly contribu-
tion of workers who meet certain conditions (earning a monthly payment less than a min-
imum salary, among others) who have difficulty making required contributions and thus 
may find themselves ineligible for a pension. The matching benefit is provided ex ante; it 
makes up for the contributions that are lacking to qualify for pension benefits. In order to 
qualify for this benefit, the member is also required to have previously contributed for a 
certain number of weeks (table 10.1). 
The target population of the matching contribution from this subaccount includes 
people with very low contribution densities, including rural workers, people with disabili-
ties, and women providing care for children (madres comunitarias—community moth-
ers).
8
 To access the matching contribution, pensioners must first be over the age of 55 

202 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
(RPM) or 58 (RAIS) and have contributed to the system for a minimum of 500 weeks 
(based on the 16 percent of income contribution). In 2009, these conditions were relaxed, 
and the minimum age was reduced to 35 years. Other groups were added, such as coun-
cillors in certain local councils. The percentage of the value subsidized depends on the 
specific population group eligible for the subsidy.
To qualify for the subsistence subaccount, beneficiaries must be Colombian, have 
lived in Colombia for the past 10 years, be at least three years younger than the retirement 
age,
9
 and be classified as having a standard of living at the lower two levels.
10
 The local 
authority selects beneficiaries (after verifying their social conditions) and determines the 
level of the subsidy. In order to increase coverage, the Ministry of Social Welfare selects 
beneficiaries residing in welfare centers for the elderly, after completing notice and veri-
fication requirements. The grants covered by this subsidy include money and basic social 
services, such as food and medicine. The value of the subsidy cannot exceed 50 percent of 
the monthly legal minimum wage.
Another potentially interesting scheme, for which the implementing regulations 
have not yet been passed, is the Periodic Economic Benefits (Beneficios Económicos 
Periódicos—BEP) scheme.
11
 Once enacted, this program will provide monthly payments 
equal to less than the minimum wage to elderly people who do not meet the conditions 
required to receive a pension. Since by constitutional mandate, a pension has to be at least 
equal to the legal minimum wage, the BEP scheme is thus not a pension. The BEP scheme 
seeks to meet the needs of people working in informal low- and middle-paying jobs, who 
typically have inadequate income and few incentives to save for old age. It would provide 
incentives of around 20 percent of accumulated savings at the time of retirement (up to 
0.85 times the minimum wage) to stimulate people to make voluntary contributions to 
the scheme.
12
 The goal is to reach some of the low and middle-income groups that are not 
TABLE 10.1  Requirements for accessing the solidarity subaccount of Colombia’s Pension Solidarity 
Fund 
Population group
Requirements
Benefits 
Age
Previous 
weeks
Length of 
subsidy (weeks)
% of 
subsidy
Self-employed in rural and urban sectors
RPM: 35–55
RAIS: 35–58
250
650
75
Municipal councilors, administrative 
categories 4.5 and 6
RPM: > 55
RAIS: > 58
500
500
75
Disabled workers
n.a.
500
750
95
Community mothers
n.a.
0
750
80
Unemployed
RPM: > 55
RAIS: > 58
500
650
70
SOURCE: Departamento Nacional de Planeación 2009.
NOTE: To access the FSP, a worker must be affiliated with the General System of Social Security for Health. n.a. = not 
applicable.

10.  MATCHING CONTRIBUTIONS IN COLOMBIA, MEXICO, AND PERU: EXPERIENCES AND PROSPECTS 
203
reached through the participation mandate of the pension systems. The financing of the 
match will come from government resources rather than a special fund.
Mexico
At least two matching-type schemes in Mexico function within the general mandatory 
defined contribution pension system established in the reforms of the late 1990s. The 
first, which targets low-wage workers, is the Social Contribution (Cuota Social), which 
is part of the defined contribution scheme administered by the Mexican Social Security 
Institute (Instituto Mexicano de Seguridad Social—IMSS). The second, “solidarity sav-
ings,” is included in the defined contribution scheme of the Institute for Civil Servant 
Social Insurance and Services (Instituto de Seguridad Social y Servicios Sociales de los 
Trabajadores del Estado—ISSSTE). It targets workers who opt to increase their contribu-
tions voluntarily (table 10.2).
TABLE 10.2  Matching schemes in Mexico
Financial scheme
IMSS
ISSSTE
General contribution rate
6.5% SBC
11.3% SBC
Matching scheme
Social Contribution
Progressive scheme for 
workers up to 15 MW
5.5% MW 97
a
 indexed to 
infl ation, for all workers
Solidarity Savings (confi rmed 
by the state)
No matching contribution
3.25 times the worker’s 
contribution
b
Guaranteed pension
Mex$2,096 monthly
Mex$3,626 monthly
SOURCE: OECD 2011.
NOTE: SBC = basic contribution salary (salario base de cotización); MW = minimum wage.
A. 5.5% of minimum salary in 1997, indexed to infl ation (Mex$3.50 daily).
B. The allowed worker’s contribution is up to 2% of SBC.
The Social Contribution is a matching payment by the federal government to the 
individual accounts of lower-income workers. It consists of a government subsidy of 
5.5 percent of the minimum wage in the Federal District for each day of work, adjusted 
quarterly by the rate of inflation. The Social Contribution significantly increases the 
retirement savings of lower-income workers. When the new IMSS defined contribution 
system began operating in July 1997, the Social Contribution was introduced as a welfare 
instrument for all workers holding pension rights. However, in May 2009, the federal 
government limited payment of the Social Contribution to workers who earned less than 
15 times the minimum wage in the Federal District. The value of the contribution is pro-
gressive, decreasing by steps with multiples of the minimum wage.
13
The Social Contribution scheme can be considered a matching scheme in which the 
government complements the worker’s contribution to the defined contribution system. 
As this is a public contribution to workers who are already paying into the system, this 

204 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
assistance may be interpreted as more of an incentive to prevent workers already in the 
system from stopping their contributions rather than a scheme for attracting new con-
tributors. The scheme’s matching does not reduce the cost of formality; it increases future 
benefits for people who enter the system and contribute.
The Social Contribution provides increased pension benefits for many lower-income 
workers affiliated with the pension system. The 2009 reform establishing an income cap 
as an eligibility requirement tightened the scheme’s focus on lower-income groups. The 
policy could be better targeted by, for example, having the state contribute 11 percent of 
the 1997 minimum wage instead of the current 5.5 percent for people earning up to three 
times the minimum wage (Herrera 2010). Under such a reform, the fiscal cost of increas-
ing the Social Contribution for people with salaries up to three times the minimum wage 
would be fully compensated by eliminating benefits by workers earning more than this 
amount.
In the scheme run by the ISSSTE, the match originates with the reforms imple-
mented in 2007. The new system is based on funded defined contribution individual 
accounts for public sector workers. It includes the establishment of the National Pen-
sion Fund for Civil Servants (PENSIONISSTE), a public body that is an offshoot of the 
ISSSTE. The National Commission for the Retirement Saving System (Consar) regulates 
and supervises the fund’s operation, administration, and execution. PENSIONISSTE is 
thus a state body. It operates in a manner similar to the Afores, the private pension fund 
administrators that manage the funds for private sector workers.
PENSIONISSTE is financed through the collection of fees from member accounts. 
The law stipulates that for each Mex$1 workers contributes voluntarily to their individual 
pension accounts, the state will contribute Mex$3.25. This voluntary contribution has a 
ceiling of 2 percent on the employee contribution base and a maximum match from the 
employer that cannot exceed 6.5 percent. This matching scheme, which is attractive for 
the worker, can increase a pensioner’s replacement rate by up to 20 percentage points if it 
is maintained over a full working life. This ex ante–type matching scheme aims to increase 
the mandatory pension contributions of public sector workers.
In addition to these schemes, firms in Mexico can develop employer pension plans. 
Various types of matching contribution schemes encourage worker participation. These 
plans are registered at Consar. In 2010, more than 1,830 of these plans were operating, 
accounting for about 3 percent of GDP (Turner 2011). The plans may be defined benefit, 
defined contribution, or mixed, depending on the firm. More than half of all plans oper-
ate under schemes in which contributions are shared by employer and employee.
In 2006, Mexico proposed the Mechanism for Saving for Retirement Opportunities 
(Mecanismo de Ahorro para el Retiro Oportunidades—MAROP), as part of the Opor-
tunidades human development program (Secretaria de Desarrollo Social 2006). Under 
the scheme, Mexicans between the ages of 30 and 69 who were participating in the Opor-
tunidades program who saved Mex$20–Mex$50 a month would receive equal match-
ing contributions from the state. Seven million people would have been eligible for the 
scheme. With savings of Mex$50 a month and an equal match from the state, savers 
would have received pensions of Mex$1,000 a month at age 70. In the end, the proj-
ect was not approved, because of doubts regarding the consistency of including a saving 
objective for groups in the Oportunidades program, who lived in poverty.

10.  MATCHING CONTRIBUTIONS IN COLOMBIA, MEXICO, AND PERU: EXPERIENCES AND PROSPECTS 
205
Peru
In June 2008, Legislative Decree No. 1086 established the objective of promoting the 
competitiveness, formality, and development of micro- and small enterprises (known as 
the Spanish acronym MYPES, from micro and pequeñas empresas which are firms with up 
to 10 workers for microenterprises and between 11 and 100 workers for small enterprises).
One main component of the decree was to create a permanent regime for the support 
of MYPES. Registration requirements were made more flexible for firms with up to 100 
workers and annual sales of no more than 1,700 tax units (Unidad Impositiva Tributaria), 
which in 2012 represents sales of  S/. 6.2 million (around $2.3 million). Some 95 percent 
of all Peruvian companies fall within these limits, but most of them do not contribute 
to any pension scheme although it is mandatory. This behavior, common in high-level 
informal economies, has been explained by the lack of enforcement by the Peruvian state 
and the obvious income constraints of employees working in very low-productivity firms. 
Thus, the decree’s objective was to introduce incentives for workers and firms to follow 
the law by directly reducing the cost of labor by giving wage complements of 50 percent 
(nonsalary payments) and a third of severance payments for each worker hired. A key 
aspect for implementing this component had been to define a different minimum wage 
for these enterprises by the National Employee Council. However, this never occurred, 
and the Ministry of Finance, which promoted this law, lost interest in implementing it 
(Carranza 2012).
A second interesting component of the decree was the introduction of the Welfare 
Pension System (Sistema de Pensiones Sociales) for workers and owners of microenter-
prises, which was never implemented. This component of the 1998 decree was taken up 
in the recent Private Pension Reform Law No. 29903 of July 2012.
14
 The Welfare Pension 
System is now obligatory for workers in microenterprises who are less than 40 years old 
and voluntary for those who are older. The workers in these firms will contribute accord-
ing to a graduated contribution rate (to be defined by regulation) of up to 4 percent of 
the legal minimum wage. The government will then match that contribution by giving 
a percentage of it (also to be defined by regulation) through a recognition bond to be 
redeemed when the worker reaches retirement. This matching is only provided to workers 
who receive less than 1.5 times the legal minimum wage. Workers can make additional 
contributions exceeding the matching limit, but those contributions will not be matched. 
For microenterprises, the formalization process has not gained momentum, for two rea-
sons. First, the National Employment Council has not created a different minimum wage 
for microenterprises, a move that would allow for a more market-based approach. Second, 
no pension contributions have been collected by the Welfare Pension System, because 
the National Pension Office (Oficina de Normalización Previsional) has not been able 
to implement the operational system to manage individual contributions of independent 
workers.
Assessment of Matching Schemes in Colombia, Mexico, and Peru 
Of the three countries analyzed in this chapter, only Colombia and Mexico have imple-
mented matching schemes for pensions. The schemes in Colombia—the FGPM and the 
solidarity subaccount—have been aimed largely at increasing the savings of groups of con-
tributors to the system who could end up ineligible for a minimum pension or receive a very 

206 
MATCHING CONTRIBUTIONS FOR PENSIONS: A REVIEW OF INTERNATIONAL EXPERIENCE
limited benefit in old age. The financing source for the matching payments has been higher-
income savers, without any support from employers or the state in the case of Colombia. In 
Mexico, the target population is the same (low-density, low-income affiliates), but the fund-
ing of the more general schemes, such as the Social Contribution of the Retirement Savings 
System and the solidarity saving of PENSIONISSTE comes from the state.
Participation in these programs was limited by two factors. First, fiscal resources 
for implementing more ambitious programs were scarce, because of the transition from 
the old public pay-as-you-go systems to the new funded private systems. Second, the 
matching programs implemented in Colombia and Mexico were designed for people who 
participated in mandatory schemes; they failed to take account of the high barriers to 
participating in the formal economy (Levy 2008). Where these costs are high, matching 
schemes are not enough to bring “outsiders” (informal workers) into the system.
More ambitious matching schemes have not succeeded in advancing beyond the 
planning stage, and some laws—including the MAROP in Mexico, the BEP scheme in 
Colombia, and the Welfare Pension System in Peru—have not been implemented. It is 
difficult to speculate as to whether these matching programs would have provided enough 
incentives to increase pension participation.
It is important to put in context the importance of MDC schemes in incentivizing 
pension participation. Given the very large size of the informal sector in these economies 
(almost 70 percent in Peru), it is difficult to imagine that a monetary incentive to con-
tribute could solve a core structural socioeconomic problem. Bringing workers who are 
currently not covered by pension systems into the system requires creating incentives that 
really matter to them. Money (a matching program) could be an incentive, but other fac-
tors could provide “nudges” (Thaler and Sunstein 2009) to contribute. Countries such as 
China, Germany, and New Zealand achieved success by including other incentives, such 
as allowing partial withdrawals from pension savings in case of liquidity constraints or the 
financing of a first mortgage.
In developing a policy, perhaps the first step is to determine which workers in the 
informal sector are in the best position to save and to tailor a program for them. The next 
section examines this point.
FOCUSING ON THE MIDDLE CLASSES
Among the groups of workers with the greatest potential to respond to voluntary MDC-
style savings schemes are middle-income workers, defined as people with per capita income 
of 50–150 percent of the average.
15
 These workers have some capacity to save, and many 
lack adequate pension coverage. According to household survey data, just 43 percent of 
workers age 14–64 in Colombia make pension contributions. This percentage is even 
lower in Mexico (37 percent) and Peru (17 percent).
16
 These three countries have among 
the lowest coverage levels in Latin America (Rofman, Lucchetti, and Ourens 2008; Rof-
man and Oliveri 2012). The proportion of middle-income households that make regular 
pension contributions is 39 percent in Colombia, 33 percent in Mexico, and 9 percent in 
Peru (figure 10.4).
The low affiliation and contribution rates partly reflect informality, which is com-
mon among even among the urban middle class. Only a third of middle-income work-
ers in Colombia and Mexico, and slightly more than 20 percent in Peru, are formally 

10.  MATCHING CONTRIBUTIONS IN COLOMBIA, MEXICO, AND PERU: EXPERIENCES AND PROSPECTS 
207
employed (defined as having a written employment contract) (figure 10.5).
17
 The vast 
majority of middle-income workers in the three countries are dependent workers without 
a contract or independent workers, with about equal numbers employed in each group.
Informality not only implies significantly lower rates of affiliation and contribution, 
it can also have a regressive impact on the pension system. Inequality in Latin America has 
fallen somewhat since 2000, as a result of the increase in social expenditures, particularly 
conditional cash transfers programs, and a decline in the returns from education (López-
Calva and Lustig 2010). But inequality still remains very high. In the absence of further 
pension reforms, the contributory social insurance systems will continue to be regressive, 
as (formal sector) workers with higher incomes are more likely to participate in the system 
(and by definition contribute at higher levels) than low- and middle-income workers in 
the informal sector.
Among the 33.3 million middle-class workers in Colombia, Mexico, and Peru, 
19.8 million are informal (4.4 million in Colombia, 11.9 million in Mexico, and 3.5 mil-
lion in Peru). Many workers alternate frequently between periods of employment and 
unemployment as well as between formal and informal employment. In Mexico, for 
example, the probability of remaining in the same employment situation between 2002 
and 2005 was 63 percent for independently employed men (it was much lower among all 
types of workers analyzed) (Jütting and de Laiglesia 2009).
Among informal middle-income workers, independent workers who completed 
tertiary education have higher coverage rates than other informal workers (figure 10.6). 
These rates are still low, however: 20 percent in Colombia, 1 percent in Mexico, and 
5 percent in Peru. They are far below those of formal workers in the same countries: 
78 percent in Colombia, 81 percent in Mexico, and 56 percent in Peru. Except in Mexico, 
where the coverage rate of informal nonagricultural workers is 18 percent, informal work-
ers rarely contribute to a pension system.
FIGURE 10.4  Pension coverage in Colombia, Mexico, and Peru, by household income level
0  10 20 30 40 50 60 70 
underprivileged
middle income
well off
Columbia (2009)
Mexico (2006)
Peru (2010)