A review of social protection in Latin America

Abstract

This paper reviews social protection trends and policy responses in Latin America as part of a global scoping study on social protection commissioned by Ford Foundation (See Appendix One for ToR for this paper). The paper is organised as follows (following the paper contents as described in Appendix One): Section 1 identifies the main trends in poverty and vulnerability in the region, with particular references to rural-urban and life course poverty incidence. Section 2 provides a review of social protection trends. It discusses the two main areas of change and innovation in the last two decades: the reform of social insurance provision and the rapid expansion of social assistance schemes. Section 3 identifies some potential points of engagement between Ford Foundation work and the social protection policy agenda in the region.

Full text

Open Research Online
The Open University’s repository of research publications
and other research outputs
A review of social protection in Latin America
Other
How to cite:
Barrientos, Armando and Hinojosa-Valencia, Leonith (2009). A review of social protection in Latin
America. Centre for Social Protection - IDS, Sussex.
For guidance on citations see FAQs.
c
2009 The Authors; 2009 Centre for Social Protection - IDS
Version: Version of Record
Link(s) to article on publisher’s website:
http://www.ids.ac.uk/index.cfm?objectid=4AA14C5A-EC76-A7E5-49068DFE1CC056E3
Copyright and Moral Rights for the articles on this site are retained by the individual authors and/or other copy-
right owners. For more information on Open Research Online’s data policy on reuse of materials please consult
the policies page.
oro.open.ac.uk

A Review of Social Protection in Latin
America
June 2009

A Review of Social Protection in Latin
America
Armando Barrientos and Leonith Hinojosa-Valencia
Brooks World Poverty Institute, University of Manchester
Prepared as part of a Social Protection Scoping Study funded by the
Ford Foundation
For permission to cite this paper, please contact the author.
June 2009

1
A Review of Social Protection in Latin America
Armando Barrientos and Leonith Hinojosa-Valencia,
Brooks World Poverty Institute, University of Manchester, UK
Contents
1. Poverty and Vulnerability in Latin America ............................................................ 2
Regional trends ....................................................................................................... 2
Urban and rural poverty .......................................................................................... 4
Life course poverty ................................................................................................. 5
Poverty trends and the global financial and economic crisis ................................... 6
2. Social Protection in Latin America ......................................................................... 7
A brief background to social protection institutions in the region ........................... 7
Social insurance ...................................................................................................... 9
Social assistance programmes in Latin America ................................................... 13
New forms of social assistance ......................................................................... 14
Design features of social assistance programmes .............................................. 16
Implementation ................................................................................................. 19
Funding ............................................................................................................. 21
Impact ............................................................................................................... 21
Challenges ........................................................................................................ 23
3. Social Protection in the Ford Foundation .............................................................. 26
Key features of social protection reforms .......................................................... 26
…and appropriate role of independent organisations ........................................ 27
Income and asset transfers ................................................................................ 28
Rights and social guarantees ............................................................................. 29
4. Conclusions and further research .......................................................................... 30
APPENDIX One: Terms of Reference for Scoping Study and Regional Papers . 32
References ................................................................................................................ 35
Final version: June 2009
Regional Paper prepared for the Ford Foundation Social Protection Scoping Study
Correspondence to :
Armando Barrientos, Senior Research Fellow -e-mail:a.barrientos@manchester.ac.uk
Leonith Hinojosa-Valencia Researcher, School of Environment and Development and
Faculty Associate – e-mail: leonith.hinojosa@manchester.ac.uk
Brooks World Poverty Institute, University of Manchester,
Humanities Bridgeford Street Building, Oxford Road, Manchester M13 9PL, U.K.

2
This paper reviews social protection trends and policy responses in Latin America as
part of a global scoping study on social protection commissioned by Ford Foundation
(See Appendix One for ToR for this paper).
The paper is organised as follows (following the paper contents as described in
Appendix One). Section 1 identifies the main trends in poverty and vulnerability in
the region, with particular references to rural-urban and life course poverty incidence.
Section 2 provides a review of social protection trends. It discusses the two main
areas of change and innovation in the last two decades: the reform of social insurance
provision and the rapid expansion of social assistance schemes. Section 3 identifies
some potential points of engagement between Ford Foundation work and the social
protection policy agenda in the region.
1. Poverty and Vulnerability in Latin America
Regional trends
An important drives for the the spread of social protection in developing countries,
and Latin America in particular has been the step rise in poverty and vulnerability
which followed crises and structural adjustment in the 1980s and 1990s, and which
then persisted in the recovery phase. Global estimates of poverty incidence based on
US$1 a day show a slight increase in headcount poverty in 1990 compared to 1987
(from 28.3 percent to 29 percent) and a subsequent decline to 1996 (27 percent)
(Bigsten and Levin 2005). Regional estimates of poverty incidence based on national
poverty lines show a stronger rise in poverty incidence, and a more gradual decline
than the global picture. Figure 1 below shows estimates of poverty and indigence
headcount rates for Latin America.
1
As can be observed from the Figure, there was a
sharp rise in the incidence of poverty from 40.5 percent in 1980 to 48.3 percent in
1990 reflecting the acute crisis experienced in the region in the 1980s, following by a
gradual, and unsteady, recovery. By 2005, headcount poverty was just below the 1980
level estimate, but had remained above it throughout the intervening period. The
trends reflect the impact of crises and adjustment in the „lost decade‟, and the
persistence of poverty in the 1990s and early 2000s.
The same trend lines can be observed for extreme poverty (indigence or food
poverty), a rise in the 1980s and a subsequent gradual recovery. The reduction in the
incidence of extreme poverty in 2005 appears to have been stronger than for overall
poverty. This is important because the trends in extreme poverty will reflect more
directly the impact of the rapid expansion of social assistance programmes in the
region. The key message is that after crisis and recovery, four in ten Latin Americans
are in poverty and one sixth remain in extreme poverty.
1
In Latin America households are classified as indigents if their income is below that required to
purchase a basic food basket. Households are classified as poor is their incomes fall below that required
to purchase a basic basket of food and non-food items. Roughly, the poverty line is twice the indigence
line. ECLAC poverty data provides estimates from household survey data on a consistent basis for the
countries represented (ECLAC 2008).

3
Figure 1.
Data Source: (CEPAL 2007)ECLAC 2008)
Comparable estimates of vulnerability and vulnerability trends are sadly lacking.
Vulnerability describes the chances of staying in, or falling into, poverty in the future.
Given the high incidence of poverty in the region, we could get a sense of trends in
vulnerability by focusing on the absence of protection. Estimates of the proportion of
the labour force in Latin America not contributing to social insurance schemes, for
example, could help us track trends in vulnerability. Workers failing to contribute to
social insurance schemes are primarily informal workers who are highly vulnerable.
As can be seen from Figure 2 below, the share of dependent workers not covered by
social insurance shows a rising trend in the 1990s and 2000s, consistent with a rise in
vulnerability in the region (ILO 2000b, 2003, 2006).
Figure 2.
28
29
30
31
32
33
34
35
36
37
38
percentage
Uncovered measure Male Uncovered measure Female Uncovered measure
Share of dependent workers not contributing to social protection insurance in Latin America
(weighted average of 12 countries, ILO's Panorama Laboral)
1990
2000
2002
Data Source: Panorama Laboral, several issues (ILO 2000a, 2003, 2006), ILO,
Lima.

4
The Figure is useful in identifying the trend in vulnerability, but it underestimates the
incidence of informality in the labour force as a whole as the vast majority of self-
employed and unemployed workers, not included in the calculations in the Figure, are
not covered by social insurance institutions and programmes. Social pensions and
health insurance programmes cover, on average, one third of the labour force in the
region.
Urban and rural poverty
A feature of poverty incidence in Latin America is its sectoral segmentation. One of
the main consequences of high and persistent inequality and social exclusion, both in
terms of income and opportunities (that is, access to human development assets), is
that poverty indicators reveal that extreme poverty tends to be concentrated in rural
areas and for indigenous groups. The incidence of extreme poverty is significantly
higher in rural areas, especially in areas with a concentration of indigenous groups. At
a regional level, countries from Central America and the Andes show significantly
higher shares of urban and rural poverty (see figures 3 and 4). At a national level,
regions located in the highlands and other rural areas of the rainforest seem to have
lower economic performance and therefore less opportunity to tackle poverty in
comparison to coastal and more urbanised areas. In addition and given that geography
and politics seem to show a pro-urban bias, the advances in terms of poverty
reduction tend to be concentrated in urban areas.
Figure 3 Figure 4
Evolution of urban indigence in Latin America
0 5 10 15 20 25 30 35 40 45 50
Chile
Uruguay
México
Perú
Costa Rica
Panamá
Brasil
Argentina
América Latina
Ecuador
El Salvador
Guatemala
Colombia
República Dominicana
Bolivia
Paraguay
Honduras
Nicaragua
Percentage
Evolution of rural indigence in Latin America
010 20 30 40 50 60 70 80
Chile
Costa Rica
México
Brasil
Ecuador
Colombia
El Salvador
República Dominicana
América Latina
Panamá
Perú
Guatemala
Paraguay
Nicaragua
Bolivia
Honduras
Percentage
1990-93
1999
2006
Notes:
- Colombia 1991, 2005; Chile 1998, 2004; Guatemala 1998, 2002; Mexico 1998;Nicaragua 1993, 1998, 2001;
Republica Dominicana 2000; Panama (Urban) 1991; El Salvador 2004; Bolivia 2004; Paraguay 2005; otherwise
data is for 1990, 1999 and 2006.
- No data for rural areas of Argentina, Uruguay and Venezuela
Data source: ECLAC database.

5
However, it is important to note that the increased mobility of poor populations across
urban and rural spaces and the deployment of combined livelihood strategies
2
,
together with high rates of urbanisation, make difficult to clearly identify – in practice
– who the rural and the urban poor are, where they do settle down and how they can
be reached more efficiently with social protection programmes.
Life course poverty
Life course factors have a strong influence in the distribution of poverty across age
groups. As shown in Figure 5, from the early 1990s until the beginning of the current
decade, child poverty increased at least 3 percent in eleven countries, with higher
percentages in Argentina, Brazil, El Salvador, Panama, Costa Rica, Uruguay and
Bolivarian Republic of Venezuela, where poverty affected more children than in the
early 1990s. Relative poverty levels only changed for the better in Nicaragua, Peru,
Paraguay and Chile. Furthermore, in all but three countries (Chile, Costa Rica and
Uruguay), over half of all children below the relative poverty line live in families
which do not meet their basic needs. High income inequality and late demographic
transition would be among the significant factors that explain this pattern
3
(UNICEF,
2005).
Figure 5
Changes in the incidence of relative child poverty,
1990-2002
-3.4
-2.7
-2.1
-0.3
0.6
0.6
2.4
2.7
3
3
3
3.4
3.4
3.6
10.5
-6 -4 -2 0 2 4 6 8 10 12
Nicaragua
Peru
Paraguay a/
Chile
Dominican Republic
Mexico
Ecuador b/
Bolivia
Venezuela
Uruguay b/
Costa Rica
Panama
El Salvador
Brazil
Argentina d/
Changes in percentage points
Source: UNICEF (2005) based on ECLAC data.
Notes: a/ Asuncion metropolitan area. b/ Urban areas. c/ Eight largest cities plus El Alto. d/
Buenos Aires metropolitan area. The base year is 1995 for El Salvador, 1993 for Nicaragua
and 1997 for Peru and the Dominican Republic; otherwise the base year is between 1989
and 1991.
2
See at this respect Hinojosa (2009), Escobal (2001), and Reardon et al (2001).
3
Reduction in relative child poverty did not imply absolute child poverty reduction. In Nicaragua, for
instance, the number of poor children has increased from just over 550,000 to some 700,000. In Chile,
the figure has risen by some 100,000 since the early 1990s.

6
Figure 6 below shows the incidence of poverty among people aged 60, more precisely
the percentage of people aged 60 and over who live in households in poverty
(Barrientos 2006). There are significant differences in old age poverty incidence
across countries in the region. Old age poverty ranges from 7.9 percent in Chile, to
38.4 percent in Ecuador. In four countries, Chile, Uruguay, Argentina and Brazil,
roughly one in ten older people are poor. In the next four countries, Peru, Nicaragua,
Venezuela, and Panama, around one in five older people are poor. In Honduras and
Paraguay, one in four older people are poor. For the remaining seven countries,
roughly one in three older people are poor. Overall, it can be concluded that old age
poverty is important in Latin America.
Figure 6
Source Barrientos (2006)
Poverty trends and the global financial and economic crisis
Although poverty and indigence rates have shown a decreasing trends in the Latin
American region in the period after 2002 (CEPAL, 2008), it is widely predicted that
the current world financial crisis will lead to rising poverty and vulnerability. The rise
of food prices, increasing rates of urban unemployment
4
and weak (or lack of) social
protection schemes in most low-income countries will influence these indicators
adversely. According to forecasts by ECLAC (ECLAC 2008), the crisis will affect
particularly poor households whose income and livelihood strategies are based on
informal economic activity . It is also expected that women will be among the more
vulnerable to the labour market downturn cycle.
There are also concerns that a rapid deterioration in government revenues could limit
the fiscal space for supporting existing social protection programmes and the
introduction of new programmes. The combination of a rise in poverty and limited
4
For an assessment of the impact of the financial crisis in poverty see www.eclac.cl/cgi-
bin/getProd.asp?xml=/prensa/noticias/comunicados/7/34747/P34747.xml&xsl=/prensa/tpl/p6f.xsl&bas
e=/tpl/top-bottom.xsl

7
fiscal space might have a strong impact on the more vulnerable groups, and could
wipe out the gains in terms of poverty reduction from the last decade of economic and
social recovery. ECLAC has called countries to strengthen the progress made
providing social assistance, not just to prevent the causes of chronic poverty, but to
improve the chances of vulnerable groups to enter – or continue – in a path towards
getting out of poverty.
So far, the countries in the region have reacted by announcing and implementing a
diverse range of measures, mostly centred on increased spending rather than tax cuts
or direct subsidies. Governments have prioritized public investments in infrastructure
(housing and public works, especially road building) and most countries have
reinforced their social programmes giving special attention to conditional cash
transfer programmes to minimise the effects of the crisis in the poorest areas.
However, groups of poor people in rural areas might not benefit from these measures
as much because policies focused on agriculture have been limited to the creation of
temporary jobs or the implementation of partial subsidies to production inputs.
5
2. Social Protection in Latin America
A brief background to social protection institutions in the region
The evolution of social protection institutions before the „lost decade‟ was initially
influenced by the example of southern Mediterranean countries, and later by the ILO
in the 1950s and 1960s. Its core consisted of social insurance funds, cajas de
previsión, financed by contributions from workers and their employers. These
provided a degree of protection from life course and work related contingencies.
Social insurance funds were initially established for specific groups of workers. Later
these were extended to cover a wider range of workers and contingencies. The
expectation among policy makers was that economic development would eventually
lead to the incorporation of the entire formal labour force, thus replicating the spread
of social protection in European countries „from the strong to the weak‟.
Mesa-Lago distinguishes three groups of countries in the region (Mesa-Lago 1991,
2007). In the first group, the pioneer countries, stratified social insurance funds
developed in the 1920s. This group includes Chile, Uruguay, Argentina, Cuba, and
Brazil. A second group of intermediate countries includes Colombia, Costa Rica,
Mexico, Paraguay, Peru and Venezuela, in which social insurance developed after the
1940s. A third group of mainly Central American countries, the latecomers,
developed social insurance institutions much later, in the 1950s and 1960s. The
„pioneer‟ countries from the Southern Cone managed to make significant advances
towards a European-style social insurance. Employment stability for the assumed
male bread-winner ensured that the protection carried over to his dependants. The
emphasis on social insurance and employment protection meant that social assistance
was at best residual, both in terms of contingencies and groups covered. By contrast to
5
International seminar “The Global Financial Crisis and the Rural Sector in Latin America: Options
for IFAD Engagement”, IFAD, April 23-24, 2009, Rome.

8
social insurance, social assistance was strongly maternalistic, with an emphasis on
family and in kind benefits (Barrientos 2004b; Molyneux 2006). By the end of the
1970s, in Chile, Brazil, Uruguay and Argentina the share of the labour force covered
by social insurance schemes reached over 70 percent (Barrientos 2004a). For most
other countries in the region, informality set hard limits to the development of social
insurance funds. Employment protection was essential to this model. Social protection
in Latin America was „truncated‟ at the margins of formal employment (Barrientos
2004a; Fiszbein 2005).
The political settlement associated with Import Substitution Industrialisation
strategies (ISI) is an important part of the explanation for the „truncated‟ nature of
social protection institutions and for the residual role of social assistance. Support
from formal sectors workers, especially those in the public sector, military and police,
and key export sectors, was essential to sustaining the development model. Social
protection ensuring minimum levels of security for these workers was a key element
in the political settlement which provided a measure of stability for the development
model. The focus on workers in formal employment also meant that social assistance
was not only residual, but also poorly focused on the poor and poorest. Social
protection in Latin America is largely regressive in distributional terms. Regressivity
arises both from the incidence of taxation and from the large public transfers to social
insurance.
6
A recent paper has coined the term „reverse Robin Hood‟ to describe the
distributional impact of social protection in Latin America, in that it takes income
mainly from the poor only to give it mainly to the non-poor (Lindert, Skoufias and
Shapiro 2005).
The 1980s crisis, and the structural adjustment which followed, led to extensive
liberalization of trade and factor markets. The incidence of labour market informality
is a good indicator of the impact of liberalization. According to ILO figures, the share
of employees in the urban labour force who are not covered by social insurance grew
from 33.4 percent in 1990 to 36.3 percent in 2002, thereafter declining to 25.2 percent
in 2006; and the share of the urban population in work (including employees and the
self-employed) who are not covered by social insurance grew from 48.2 percent in
1990 to 51.2 percent in 2002, declining to 39.2 percent in 2006 (ILO 2006).
7
The
upward trend in urban informality up to 2002 seriously challenges the expectations
held before the 1980s regarding the gradual extension of social protection. It also calls
into question the extent to which employment-based social protection systems in the
region could ever reach beyond the minority of the labour force, mainly public and
formal sector workers (Bourguignon 2005). In some countries, the decline in the
labour force coverage of social insurance was precipitous. Argentina is a case in
point, while in 1980 around 70 percent of the labour force were contributing to a
social insurance fund, by 2003 this figure dropped to 35 percent, thereafter recovering
to 45.4 percent in 2006 (Gill, Packard and Yermo 2004; Rofman 2005).
Emergency responses to the rise in poverty and vulnerability in the wake of the crisis
proved inadequate. The residual social assistance institutions in place before the
1980s did not have the capacity to provide an effective response to the rapid rise in
6
In Brazil for example, one of the few countries that so far has avoided fundamental reforms to the
pension component of their social insurance schemes, the subsidy from tax revenues to plugging the
deficit of social insurance pensions amounts to 4-5 percent of GDP annually.
7
Note that changes to household surveys after 2003 could affect comparability with earlier data.

9
poverty and vulnerability, and in any case structural adjustment hit public agencies
and their budgets, hardest. Some countries, with support from international
institutions, put in place emergency safety nets, mainly in the form of public works
and in kind transfers. These proved insufficient and highly divisive given the blunt
and discretionary nature of the targeting instruments employed. Emergency responses
gave way to social funds and sectoral programmes targeting „vulnerable‟ groups.
Social funds were set up to support demand led community initiatives, and involved
emergent NGOs, but they rarely reached the scale required to address widespread
poverty and vulnerability, and the creation of parallel structures to government
agencies meant that they were unlikely to make a significant contribution to
developing long term social assistance institutions and the social contracts needed to
underpin them. Some countries focused social programmes on vulnerable population
groups such as indigenous people, senior citizens, or women, which created
fragmented sectoral institutions with little flexibility to respond to the dynamics of
poverty. These did not prove successful either, as they failed to exploit opportunities
for economies of scale in programme design and implementation, and often targeted
the same population many times with separate but insufficient forms of support.
In the last two decades, the main changes to social protection institutions in the region
are the reform of social insurance, especially pensions and health insurance, and the
emergence of new large scale social assistance. These are discussed in turn below.
Social insurance
Following Chile‟s 1981 pioneering pension reform, a number of countries in the
region have reformed their pension systems. Pension reform has been implemented in
Peru in 1993, Argentina and Colombia in 1994, Uruguay in 1996, Mexico and Bolivia
in 1997, El Salvador in 1998, Costa Rica in 2001 and the Dominican Republic in
2003. Enabling legislation has been approved in Ecuador, Nicaragua and Venezuela
but implementation has not materialised. The reforms implemented share common
traits. They aim to replace unfunded defined benefit social insurance pension schemes
with funded defined contribution pension plans. However, the scope and condition of
pre-existing pension schemes and political economy factors have resulted in a
diversity of outcomes. This section briefly reviews this diversity.
The main innovation at the core of pension reforms was the introduction of individual
retirement pension plans managed by newly created private pension fund managers.
8
The new pension plans are of the defined contribution type and are fully funded. In
the new pension environment, employees in formal employment are required to
contribute a fraction of their earnings to a retirement account. Pension fund managers
compete for savings, manage workers‟ accounts, invest the pension fund, and
surrender balances at retirement. They also provide disability and survivor benefits.
At retirement, workers use their account balances to make pension arrangements,
which include purchasing an annuity from insurance providers or agreeing a
scheduled withdrawal of the balances from their pension accounts. The introduction
of individual pension accounts competitively offered by new pension fund managers
8
For a more detailed discussion of the reforms see Barrientos (1998) and Queisser (1998). Some
countries retained pay-as-you-go pension schemes, and Argentina allowed a public pension fund
manager to compete with private ones.

10
is a novel way of organising pension provision in Latin America, and the development
of a pension fund management market is perhaps the most important innovation
introduced by the reforms.
Figure 1.
Diversity in Pension Reforms in Latin America
Before the reforms
After the reforms
Bolivia Colombia Argentina Costa Rica Brazil
Chile Peru Uruguay
Mexico
El Salvador
Dominican Rep.
DB-PAYG
Defined benefit
PF  PAYG/Subsidised
Formal sector coverage
Fragmented
DC-FF
DC-
FF
DB-
PA
YG
DC-
FF
DB
-
PA
YG
DB -PAYG
DB -
PAY
G
DC-
FF
DB-
PAY
G

11
While this core feature of the reforms became common to all ten countries, its role
within the pension system, and its linkages to the pre-existing social insurance
pension schemes depended on specific country conditions (Gill, Packard and Yermo
2004; Mesa-Lago 2007). Two factors played an important part in shaping the reforms.
In countries with more developed social insurance schemes, the transition to
individual retirement plans was bound to be costlier, especially as those in
employment would be required to support those in retirement while saving for their
own retirement. In these countries, the reforms were always likely to aggregate a
greater number of opponents. Political conditions were also influential. Pension
reform in Chile was introduced with very little discussion and scrutiny mainly due to
the fact that political institutions were suspended. In other countries, political
discussion around the reforms was fierce (Mesa-Lago 1999). The development
reached by pre-existing institutions and political economy factors resulted in several
configurations of the mandatory components of the new pension systems. Figure 1
shows the different variants produced by the reforms.
In five countries, Bolivia, Chile, Mexico, El Salvador and the Dominican Republic,
the reforms simply replaced the pre-existing defined benefit, pay- as-you-go (PAYG),
social insurance schemes, with defined benefit, fully funded individual retirement
plans. New entrants to the labour market had no choice but to join the new pension
plans, while affiliates to the pre-existing pension schemes were encouraged to migrate
to the new pension plans. Entitlements under the old scheme were transferred to the
new individual retirement plans. In Chile, for example, contributors to the social
insurance scheme were provided with a „recognition bond‟ reflecting their
contribution history and redeemable at retirement. This practice also had the
advantage of spreading government liabilities into the medium term.
In two countries, Colombia and Peru, contributors to the old social insurance schemes
were given the choice to stay or to move to individual retirement accounts. New
workers were required to join the new pension plans. The two types of pension
schemes therefore run in parallel. In Argentina, were opposition to the reforms by
trade unions was most acute (Kay 1999; Ghilarducci and Ledesma 2000), the outcome
of the reform was a two pillar pension system in which the first pillar consists of a
PAYG basic pension, with a choice of either the old PAYG scheme or the new
individual retirement accounts. This is in fact a two-pillar pension system with limited
choice in the second pillar. In two countries, Uruguay and Costa Rica, a different two
pillar pension system emerged, with a first pillar consisting of a redistributive PAYG
defined benefit schemes, and a second pillar consisting of individual retirement
accounts. There is no choice in the second pillar (Martinez Franzoni and Mesa-Lago
2003; Martinez Franzoni 2005).
Finally, in twelve countries in the region individual retirement accounts were not
introduced and PAYG defined benefit schemes remained in place. In some of these
countries, parametric reforms to the social insurance scheme were introduced, of
varying degrees of importance. Brazil stands out as the parametric reforms introduced
have shifted social insurance schemes in the direction of notional defined contribution
plans, which take account of demographic change in the benefit formula (Pinheiro
2005).

12
As can be seen from the Figure, the shift to defined contribution fully funded pension
plans in the region has been a dominant feature, but it is by no means complete, and
the model of reform has many variants. Moreover, the articulation of defined
contribution pension schemes and defined benefit pension plans has shifted over time.
In those countries where a choice between these two was provided, the expectation
among the reformers was that the old social insurance schemes would gradually
wither away. This has not proved to be the case.
The outcomes from the reforms to social insurance have been debated in detail in the
region, by supporters and detractors (Mesa-Lago 2007). Our focus here is to draw out
the implications from these reforms for the shape of social protection systems in Latin
America, and the implications for poverty reduction. A fair assessment of the social
insurance reforms in the region are that they have been more successful in
establishing new financial institutions (private pension funds, annuity providers), than
in strengthening the protection of workers in old age. Figure 7 below helps to
illustrate these points. Private pension funds control a significant proportion of
financial wealth, but cover at best a fraction of the labour force. Only in Chile and
Costa Rica, pension funds proved attractive to a majority of the labour force. In Chile,
private pension funds control financial instruments equivalent to nearly 65percent of
GDP. In other countries, private pension cater for a minority of well paid workers.
Figure 7
Source: AIOS Boletin Estadistico 2007
The fact that social insurance reforms did not take root strengthens the case for
focusing on social assistance as the route to extend social protection in Latin America.
This has now been acknowledged by the World Bank, a key driver for the reforms
(Gill, Packard and Yermo 2004). Brazil is an interesting case, because it did not
privatise its pension funds but opted for parametric reforms. Around 60 percent of

13
pensioners from the social insurance funds there only manage to achieve a minimum
pension, equivalent to non-contributory pensions, while a minority of highly paid
workers and civil servants draw very generous pensions. In Brazil the social insurance
funds are an important poverty reduction programme.
The financial crisis could have large effects on private pension funds, given their
financial exposure, and implications for public resources and by extension social
assistance. In December 2008, the Argentinian government „nationalised‟ the private
pension funds, at a stroke taking over the pension liabilities of workers in them. The
measure will hugely raise the regressive nature of social insurance in the region, and
could crowd out resources for the expansion of social assistance for households in
poverty. Chile, on the other hand, implemented reforms in 2008 which strengthen the
first pillar, non-contributory, pensions. The aim of the reforms is to eventually extend
coverage to all older groups in the country. (Mesa-Lago 2008)
Social assistance programmes in Latin America
In the last decade there has been a very rapid expansion of new forms of social
assistance programmes in the region. These programmes target particular groups in
the population in poverty and/or regions and municipalities characterised by high
poverty indicators.
Income transfer programmes in middle-income countries in the region have their
earliest referent in the Uruguay‟s Family Allowances Scheme (initiated in 1914) and
non-contributory pension schemes in Costa Rica, Chile and Brazil. . However, in the
late 1990s and the 2000s there has been substantial growth in income transfer
programmes as a means to tackle poverty and vulnerability. Their success with
regards to their short term objectives in middle income countries – mainly Mexico
and to a lesser extent Brazil – has led other middle and low-income countries to adopt
similar approaches. There has been strong growth in human development transfer
programmes. These programmes were introduced initially as temporary or pilot
programmes, but , as Barrientos and others suggest, many of those programmes have
been scaled up and institutionalised as a specific route to the extension of social
protection (Barrientos and Santibañez 2009; Barrientos and Santibanez forthcoming).
Whilst initially income transfer programmes developed in middle income countries,
recent experiences show that they are feasible and effective in low income countries
too.
A feature of the recent development of social assistance programmes is that their
surge and evolution reflects both policy learning and strong civil society demand. In a
sense the spread of these programmes is a a result of policy makers‟ decisions to
respond to a challenging context of formal labor market retrenchment and upsurge of
poverty and inequality. In an important sense, changes in social policy and
implementation of social assistance programmes have also been induced by civil
society groups who along the years have campaigned and mobilized grassroots groups
making specific claims about welfare provision – in Brazil, for instance, grassroots
support to President Lula has largely been identified with his commitment to

14
consolidate Bolsa Familia (Hall 2008). The various organizations classified in the
“Third Sector” have also implemented self-help groups and support networks, which
in some cases have been adopted by the public sector as „innovative‟ experiences for
social protection design and eventual administration (for example, the Programa Vaso
de Leche in Peru (Banco Mundial 2007)). This feature introduces an additional
challenge to social protection program managers to avoid reliance on passive and
perhaps clientelistic relations with beneficiaries.
In the following sub-sections we provide a summary of social assistance programmes
in Latin America, considering their main objectives and impact. Programmes can be
classified into two groups: non-contributory pensions programmes and income
transfer programmes with focus on human development. Our analysis builds on the
programme information provided in the Social Assistance in Developing Countries
database (Barrientos, Holmes and Scott 2008).
New forms of social assistance
According to Bertranou and Grushka (2002), non-contributory pension programmes
constitute a good alternative to extend the scope of social protection to those segments
of the population that have traditionally been excluded from social security.
Argentina. Argentina started with an unconditional income transfer programme in
1994 (Pensiones Asistenciales) which established non-contributory pension
programme support to vulnerable individuals and those who have made a significant
contribution to society (war veterans, relatives of disappeared persons, scientific
achievement, etc). It then implemented two conditional income transfer programmes
in 2002 and 2004 respectively: the Jefes y Jefas („Male and Female Heads of
Household‟), which was a response to the economic crisis experienced by the country
at the end of 2001, and the Programa Familias por la inclusión social („Families for
social inclusion‟), which provides income transfers to families in extreme poverty to
support household access to health and education as well as implements community
development projects.
Bolivia. Bolivia started its first non-contributory pension programme named
BONOSOL in 1997, after the privatisation period, as a redistributive means that
would benefit to citizens reaching 60 years of age. This programme has suffered
several modifications and been re-launched as Renta Dignidad in 2008. Another
income transfer programme is the Bono Juancito Pinto which was launched in 2006
and targets all students from public primary schools.
Brazil. Brazil‟s first income transfer programme, Beneficio de Prestaçao Continuada
(„Continuous delivery of benefits‟), started in 1993 as a non-conditional non-
contributory pension programme targeting poor individuals aged 65 and over. It was
then followed by a series of pilot conditional programmes with human development
purposes: the Child Labour Eradication Programme (PETI) from 1996, the Bolsa
Escola initiated in the Campinas State in 1995 to assist poor households with children
of school age, Bolsa Alimentaçao which targeted indigent households, and Auxilio-
Gás (an unconditional income transfer subsidising poor households‟ consumption of
gas). All of these pilot programmes were sequentially integrated since 2003 into Bolsa

15
Familia, the federal programme carried on with intervention of states and
municipalities which consolidates an income transfers strategy as a means for
reducing poverty, inequality and exclusion.
In addition to these programmes, the Federal Government of Brazil has a long
standing tradition of assisting to informal workers from rural areas (for instance with
programmes such as FUNRURAL). After the 1988 Constitution, such a support was
consolidated into the Prêvidencia Rural (Rural provision) programme, which since
1991 supports rural informal workers reaching the age of 55 if women and 60 if men,
or disabled.
Chile. The first unconditional non-contributory pension programme programme, the
Programa de Pensiones Asistenciales (Assistance pensions programme), was
established in 1975 to support old or disabled individuals without other sources of
income. No other programme was registered until after the change of political regime
when the new government established in 2002 the Chile Solidario programme, which
provides integrated support to households in extreme poverty.
Colombia. Colombia‟s first income transfers programme, the Programa de
Ampliación de Cobertura de la Educación Secundaria („Programme for the Extension
of Secondary School Coverage‟) was established in 1991 to raise secondary school
enrolments among low-income groups in Colombia. Its second programme, Familias
en Acción („Families in action‟) started in 2001 as a conditional income transfer for
human development targeting poor households with children in poor areas.
Costa Rica. The Régimen No Contributivo de Pensiones por Monto Básico (Non-
contributory Basic Pensions Regime) was the earliest unconditional income transfer
programme in Latin America targeting the elder and disabled poor individuals. In
2001 the Costa Rican government established the Superémonos programme („Let‟s
get better‟) which targets children from poor households to improve school enrolment
and attendance.
Dominican Republic. The Programa Solidaridad („Solidarity Programme‟) was
launched as a conditional income transfer programme for human development in 2005
targeting households in extreme and moderate poverty.
Ecuador. The Bono Solidario („Solidarity Bonus‟) was initially launched in 1998 as a
compensation programme to poor households with children, elderly and disabled for
withdrawing price subsidies on petrol and derivatives. Since 2004 it became the Bono
de Desarrollo Humano (Bonus for Human Development), which aims to reduce
poverty and targets the same groups of the previous programme provided that no one
in the household is in formal employment or covered by social insurance.
Honduras. The Programa de Asignacion Familiar (Family allowance programme)
started as a pilot conditional income transfer programme for human development in
1990 and then re-launched in 2000 in such as a „second phase‟. It targets all
households from municipalities with lowest mean height for age z-scores.
Mexico. The Programa Nacional de Educacion, Salud y Alimentacion (PROGRESA)
was developed as a income transfer programme for rural households in extreme

16
poverty. It started in 1997 and aimed at poverty reduction and prevention in small
rural communities with a high marginality score and access to education and health
providers. With the same vision, the OPORTUNIDADES programme has extended
the coverage to urban areas and also included micro enterprise support components.
Panama. The Red de Oportunidades (Opportunities network) programme started in
2006 as a pilot income transfer programme aiming to reduce extreme poverty. It
guarantees to poor households access to basic services, household support, and
infrastructure development in order to re-integrate them into processes of economics
dynamics and social development.
Paraguay. The conditional income transfer pilot programme for human development
Red de Protección y Promoción Social (Social Protection and Promotion Network)
started in 2005 assisting poor households with children from poor communities.
Peru. The programme Juntos (Together) was established in 2005 to provide cash and
food supplements transfers to poorest households with children, conditional on health,
schooling, and registration. It targets households in extreme poverty from the nine
poorest regions.
Uruguay. The Programa de Pensiones No-Contributivas (No-contributory pensions
programme) formally started in 1986 as an unconditional non-contributory pension
programme programme, however it gives continuity to long standing similar
programmes put in place since 1914. It targets the elder or disabled poor individuals
who are excluded from social insurance schemes. In addition and in order to attack
poverty with an integrated programme, in 2005, the Government of Uruguay put in
place the Plan de Atención Nacional a la Emergencia Social (National Plan to
Address the Social Crisis) which includes income transfers, public works and micro-
enterprise development for the lowest quintile of the population below the poverty
line.
Design features of social assistance programmes
Many of the social protection programmes in Latin America arose from a political
recognition that a significant proportion of the labour force and their families are in
informal employment and do not have access to traditional social security schemes.
Furthermore, given that the exclusion from social security is a consequence of their
marginal access to formal labour markets, the protection of vulnerable and poor
population is basically based on non-contributory and social assistance programmes
(Bertranou et al 2002).
Argentina, Brazil, Costa Rica, Chile and Uruguay are among the countries which have
set a precedent for the design of non-contributory pension programmes, in the same
way as they are well known for their relatively more developed social security
systems. However, several economic crises in the last 20 years have modified the
coverage and impacts of these systems and programmes. In several cases, have been
used as a political cushion to absorb the adverse effects of significant reductions in
contributory pensions systems.

17
In Brazil the design of non-contributory pensions for elders, widows, handicapped
and mothers with more than seven children has been done in a way of protecting the
entire population (Schwarzer and Querino, quoted in Bertranou et al 2002). In Chile
non-contributory pensions would have been also designed to reach the whole
population not covered by social security schemes. The assistance programme
targeted poor population over 65 years-old, handicapped over 18 years-old and all
mental-ill individuals, and not covered by any other pension scheme.
9
In addition to
income transfer, beneficiaries of this programme have free access to public health
service, family allowances and funeral expenses (Gana 2002).
In Costa Rica , where the level of social development can largely be explained by
longstanding concern with extending social protection to a majority of the population,
social protection programmes arose as a way to compensate those who have been left
aside from the benefits of economic growth and instead have fallen in poverty. In that
way, non-contributory income transfer schemes were implemented along the years
since 1974 as a way of ensuring that social protection is part of a social development
strategy and not just a compensatory resource in times of crisis or social emergency
(Duran 2002).
Uruguay is another country with long standing social policy tradition. Expenditure on
social sectors focus largely on non-contributory pensions for elderly people over 70
years-old and handicapped are stipulated as a right in the country legal framework.
Therefore, in 2002 more than 85 percent of elder population were covered by formal
social security and 5.54 percent by non-contributory pensions (and Lorenzelli 2002).
By contrast, the Argentinean non-contributory pension programme programme
developed in the context of a social policy characterised by a high degree of
fragmentation not only between the national institutions (for example, the Social
Security Administration and the Ministry of Social Development), but also the various
levels of government (national, provincial and municipal) (Bertranou and Gushka
2002). One of the characteristics of this programme, much like other subsystems of
social protection, is that it was founded on a legal framework that developed in a
disjointed and uncoordinated fashion. Indeed, the auxiliary pensions (pensiones
graciables) granted by Congress deserve special comment. In theory, these benefits
must adhere to the objectives established for protecting vulnerable groups; however,
national legislators are responsible for selecting and managing the benefit without
intervention on the part of the institution responsible for administering the non-
contributory pension programme programme.
With regards to income transfers programmes with an aim other than pensions, those
have focused on poor households to subsidise investment in human capital by
conditioning on enrolling their children in school and making regular health checks to
ensuring child health development. Most of these programmes combine demand and
supply side components. From the „demand-side‟ their main objective is to induce
households (through income transfers and conditions associated with the receipt of
these income transfers) to make more intensive use of the existing educational and
health. From the supply side, programmes are accompanied by complimentary efforts
9
Food pensions and political-victims compensation pensions are excluded from this restriction.

18
and resources directed at strengthening the supply and quality of the educational and
health services.
Two of the most known programmes for human development are the Eradication of
Child Labour Programme (PETI) and Bolsa Escola in Brazil, and PROGRESA in
Mexico. In Brazil the idea of income transfer to poor families emerged from the
University of Brasilia and first implemented by the government of the Distrito Federal
in 1995 (Cardoso and Souza 2002). There is also a supply-side component to each
programme with resources allocated toward improving school quality and access
(e.g., more teachers, health clinic staff, higher salaries, and extensive expansion).
PETI was designed to eradicate child labour from families in a situation of extreme
poverty and social exclusion in urban and rural areas. This programme is covered by
constitutional mandate and is part of the social assistance commitment made by the
Federal Government within the scope of Program Avança Brasil. From being a pilot
programme implemented in the coal production areas of the State of Mato Grosso, it
sequentially was expanded to other municipalities and States of Brazil. (Brazilian
Court).
Similar to the PETI programme, the Federal Bolsa Escola gives income grants to poor
families with school-age children (selected on an income-means test and/or a scoring
system, but initially located in metropolitan areas). The main goal of Bolsa Escola has
been to reduce future poverty by increasing school educational attainment and to
reduce current poverty by transferring income to poor families. Following the
programme popularity, in the early 2000s more than 100 other programmes would
have been implemented in other municipalities within state and non-governmental
strategies of poverty reduction (Levinas and others 2001).
In Mexico the PROGRESA programme began as a component of an overall strategy
for poverty alleviation, as such PROGRESA works in conjunction with other
programmes that are aimed at developing employment and income opportunities.
10
PROGRESA‟s multi-sectoral focus has provided an integrated package of education,
nutrition, and health services to poor families, and requires active participation by the
recipient households in exchange for the benefits. PROGRESA gives benefits mainly
to mothers.
Following these referents, human capital-oriented income transfer programmes were
also implemented in Chile, Argentina, Peru, Guatemala, Honduras, Nicaragua,
Paraguay and Bolivia. After the change of political regime in Chile, the Chile
Solidario programme was designed as a system of social protection which combines
social assistance and community development to tackle extreme poverty. For an
estimated period of five years, beneficiaries of this system receive psycho-social
support, a protection bonus, income subsidies and have preferential access to other
programmes of community development. In Bolivia the Bono Juancito Pinto was
introduced in 2006 as part of an integrative policy of social protection and community
development named “Bolivia digna, soberana, productiva y democratica para vivir
10
Beneficiaries are selected through a three-stage targeting mechanism. First, using national census
data, geographic targeting is applied to select the most marginal communities. Second, socioeconomic
data are collected from all households in the most marginal communities. Using income and other data
(e.g., education, housing conditions, and durable goods) a discriminating analysis is used to identify
“poor” households. Finally, community feedback is used to reclassify households (Coady 2003).

19
bien”. In Peru the Juntos programme is also part of a state strategy for tackling
extreme and child poverty (the Estrategia Nacional Crecer), and is targeted to
households in situation of indigence and social exclusion. Juntos combines income
transfers with the supply of health and education. Design of the Peruvian programme
is similar to Mexico‟s Oportunidades, Brazil‟s Hambre Cero and Chile‟s Puente. All
of them were acknowledged as part of a regional strategy to eradicate poverty.
In spite of a general aim to eradicate poverty shared by all programmes from medium
and low-income LACs, in most of the cases the transfer levels are set only at a
fraction of the poverty line. Therefore, the transfers have not been set to bring
households up to the poverty line but instead to support schooling and health care
access
11
. Nonetheless, these programmes are intended to produce non-income effects.
For instance, the transfers are paid to the mother in the expectation that they will be
used to support children, and perhaps strengthen the mother‟s position within intra-
household decision making; programmes have co-responsibilities, both from
beneficiary households and from the programme agency; and some programmes
include more intensive support from programme agents (Barrientos and Santibanez,
forthcoming).
A general feature of design in most of the Latin American countries is that selection
processes rely on several criteria which combine geographical identification,
deprivation indicators, categorical selection and community validation. Indicators for
targeting include: household poverty, per capita expenditure below the poverty line,
quality of life index, assets test. Given that rural poverty expose indicators of higher
incidence, many of the programmes focus on the rural poor, however, in El Salvador
and Bolivia, the urban poor are also reached with income transfers. In other cases it is
only the „supply side‟ of human development income transfer programmes that reach
both urban and rural areas and the „demand side‟ stays focused on the rural poor.
Implementation
In most of the countries the implementation and administration of social protection
programmes have been assigned to ministries and public sector organisations which
are in charge of social assistance and social development. However, in countries such
as Costa Rica, Brazil, and Uruguay, social security public administrations are also the
ones which administer social protection programmes. That entails the advantage of
having the structure needed for managing large amounts of income transfer, but that
has also shown limitations to identify and verify the target population (Bertranou,
Solorio and van Ginneken 2002).
In Argentina, from 1996 the non-contributory pension programme was administered
by the Secretariat for Social Development (SDS). Until 1999 it was under the Office
of the President and since then has been transferred to the Ministry of Social
Development and Environment (Bertranou and Grushka 2002).
In Brazil, the PETI programme followed the principles of decentralised management,
which implied that the execution of the programme is shared by the three levels of
11
For instance, in Nicaragua the transfers were equivalent to 21 percent of household expenditure on
average, but only 3.6 percent in Honduras. In Paraguay the transfer level was set to equal the average
poverty gap among the bottom to deciles of the population. In El Salvador the combined transfers equal
around 50 percent of the poverty line (Barrientos and Santibanez forthcoming).

20
Government (federal, state and local). Initially transfers were made to municipalities,
however the administrative burden to meeting certification criteria made the central
government to opt for direct transfers to the beneficiary families. As Cardoso and
Souza note, despite the decentralised design of income transfer programmes, main
limitations in implementation can be explained by the difficulty to build a trustworthy
catalogue of poor families entitled to benefits and the hurry that has characterised the
updating of prior catalogues each time that a new municipal administration came into
power (Cardoso and Portela Souza 2003).
In Chile, where the overall coverage of non-contributory pension programme are
defined at central government level and based on budgetary criteria, distribution along
the country responds to regional differences of poverty levels and the waiting lists of
potential beneficiaries in each region. Final targeting and eligibility have been point-
based, where regions and municipalities have kept their own system (Gana Cornejo
2002).
In Costa Rica the implementation of social protection programmes has been made
through a complex network of public organisations („executive units‟), but with a
single centralised funding body (the Fondo de Desarrollo Social y Asignaciones
Familiares). That has implied problems of duplicity and increased bureaucracy, spill-
over and path-dependence effects, and constraints for adequate impact assessment.
In Uruguay, the programme is managed entirely from a special unit within the Banco
de Previsión Social (Social Prevention Bank), which works from its base in the capital
city (Montevideo) and then decentralises into BPS‟s regional offices. At that regional
level there is no particular administrative apparatus that guarantees an efficient
implementation. Main problems would have been related to non-specialised personnel
in charge of the programme and the difficulties to verifying the income basis of
potential beneficiaries.
In Chile, where the Chile Solidario system was implemented at multi-institutional
level, coordination has been a particular challenge. Evaluations of some of the
programmes involved showed that the flexibility and adaptation observed in
programmes of psycho-social support were hardly observed during implementation
phases of income and employment programmes (MIDEPLAN 2004). Among some
factors rarely considered in implementation phases are the family characteristics and
the local contexts that bound the ability of the poor to enter into labour markets.
In Peru, formulation and implementation of social policies is in charge of a multi-
sectoral commission (CIAS), created by law in 1990 and in front of the anti-poverty
strategy since 2003. The CIAS‟s core groups get the ministries of Health, Education,
Agriculture, and Women and Development together. However, given the high
political component of social protection programmes, in practice, these programmes
are directly managed from the Ministry of the Presidency and coordinated with other
state instances both at central and regional levels (such as the Decentralised Offices of
the Presidency - OPDs).

21
Funding
In Argentina, Brazil, Chile, Costa Rica and Uruguay, Bertranou et al (2002) noted that
most of the non-contributory pensions and protection programmes are funded by taxes
and government general income, consequently there is a cross-sector subsidy (from
formal to informal sectors and from urban to rural areas). Nonetheless, they also
found that in some cases funding for these programmes are semi-contributory, that is,
has roots in other forms of contributions from sectors considered by convention as
“informal”, such as agriculture
12
. Funding for non-contributory pensions programmes
in Chile is included in the national budget. In Uruguay funding for non-contributory
pension programmes comes from government income and also from workers payroll
contributions. It is noteworthy that expenditure in social sectors grew more than the
GDP and that fact was greatly related with the emphasis government put on non-
contributory cash-transfer programmes.
A common feature to most human development focused transfer programmes is the
international nature of their financing, with the World Bank and the IADB as main
financing agencies and their loans involve funds for expenditure in conditional
income transfers, technical assistance and project management, and capitalisation of
the front-end fee
13
. For medium-income countries such as Brazil, Mexico, Chile,
Costa Rica and Colombia loans range go from US$ 92.7 million to US$ 520.2
million. In addition, the Bolsa Familia programme gets World Bank funds within a
results-based management framework which triggers increases in the rates of loan
disbursements
14
.
In Nicaragua and in El Salvador their programmes were initially set up to channel
international aid to demand-driven local community development. Honduras‟ Family
Allowances Program (PRAF) has also been financed by the IDB (FSO). Few
exceptions to dependence on international loans are the PETI programme in Brazil
which is funded by the National Social Assistance Fund, the Bono Juancito Pinto in
Bolivia which is financed by taxes on hydrocarbons and the Juntos programme in
Peru which 90 percent of its finance comes from general revenue and the remaining
10 percent by international loans
15
.
Impact
Studies of the contribution of non-contributory pensions to poverty reduction showed
that such an effect in Brazil and Chile would have been to decrease indigence
incidence in a range between 67 and 95 percent and poverty incidence in a range
between 18 and 30 percent and between 21 and 24 percent in Costa Rica (Bertranou et
12
By 2002, it was found that 100 percent of non-contributory pensions and assistance in Argentina,
Uruguay and urban Brazil would have been funded by general income, and in rural Brazil 8.4 percent
of rural programmes would have been funded with taxes on primary trade of agriculture products. In
Costa Rica, 5.4 percent came from taxes to cigarette and spirits, and 46.2 percent from patrons‟
contributions. In Chile 8.4 percent would have come from individuals‟ contributions. (Bertranou et al
2002, p. 17).
13
See for more detail the WB‟s country projects at http://www.worldbank.org/infoshop
14
The Bank is also experimenting with that funding modality in other projects in Brazil and evaluating
its application to other countries (Lindert, 2007).
15
http://www.juntos.gob.pe/pdf/trans_plan_anual/PIA2009.pdf

22
al 2002)
16
. In Argentina, by 2002 the non-contributory pension programme would
have been effective at reducing poverty, among those families with a member who
receives the benefit, by 31 percent, and extreme poverty or indigence by 67 percent
(Bertranou and Gushka 2002). The impact of the non-contributory cash-transfers
programmes in Chile would have been important for poverty reduction and
improvement of income distribution. According to Gana (2002), 11.3 percent of the
lowest quintile‟s income comes from these programmes. With regards to poverty
reduction, by 2000 extreme poverty incidence would have decreased by 69 percent
and poverty in 15.3 percent. An important factor that would help to explain such an
impact in each one of the countries is the extent to which the programmes became
extensive. For instance, in Argentina between 1991 and 2000 the number of
beneficiaries more than doubled (from 158,000 to 351,000). The number of direct
beneficiaries with pensions was 350,000 persons, but if health coverage provided for
the families of some beneficiaries is included, coverage may be said to extend to
450,000 persons. The average benefit was $153, which is 57 percent of the average
benefit in the contributory system (Bertranou and Gushka op cit).
In Brazil the PETI programme increased significantly its reach since it was created.
From 3710 children assisted in 1996, it reached to 749,353 children in 2001.
Following that effect, the municipalities involved in the programme have reported
significant reduction of the student drop-out rates. By 2002 Bolsa Escola reached
almost all Brazilian municipalities, providing assistance to five million children
(Cardoso and Souza 2002) and by 2006 it would have reached to all households in
poverty (about 11 million in the country).
17
Based on a study of seven states in
Mexico, Skoufias and McClafferty (2001) showed that PROGRESA‟s impact on
education has significantly increased enrolment rates, particularly of girls, especially
at the secondary school level. The programme would also have impact on
improvements in health of both children and adults, and on food consumption.
An additional overall impact observed in several countries has been on women
empowerment. Not necessarily being part of the design, targeting showed that women
are the largest group of population who have had access to non-contributory pension
programmes (for instance, 72 percent of in Argentina and 60 percent in Chile)
18
. The
evaluation of PROGRESA also reported improvements in women empowerment due
to control of the monetary benefits which would have increased their household
decision making. Saldain and Lorenzelli showed that beneficiaries in the Uruguay
cash-transfer pensions programme were mainly women and the urban poor (Saldain
and Lorenzelli 2002). Nonetheless, they also showed that there has been a linear
relationship between the number of urban poor and the programme beneficiaries,
which implies that the programme has played a cushioning role when social
deterioration was observed, but it would not necessarily reduced poverty severity.
Similarly, given the geographical concentration in urban areas, rural poverty may not
have been properly addressed. In spite of these weaknesses, the transfer programmes
involved would have had positive redistributive effects (Saldain and Lorenzelli 2002).
16
Impact in Brazil includes the effect of contributory pensions. In Chile and Argentina the impact was
measured in terms of households, whilst in Costa Rica and Brazil it is in terms of individuals.
17
PNUD Brazil.
http://www.pnud.org.br/pobreza_desigualdade/reportagens/index.php?id01=2237&lay=pde
18
Percentages calculated by 2000. (Collaborations quoted in Bertranou et al 2002).

23
Challenges
With non-contributory pension schemes, issues of targeting and differential access by
rural and urban populations are an important challenge. In Chile most of the non-
contributory pensions programme are located in urban areas (62 percent in 2000
(Gana 2002))
19
. In Costa Rica, targeting non-contributory pensions and assistance
programmes on poor beneficiaries should have benefited primarily the rural
population, women and those with lower education levels. As it is, a majority of
beneficiaries live in urban areas, reducing the impact of the pension on poverty
reduction would have been limited. The low level of the pension benefit also limits
the impact of the programme on poverty incidence. Estimations for Costa Rica
suggest that 32 percent of beneficiaries are expected to remain in extreme poverty and
only 8.6 percent would have graduated from extreme poverty into moderate poverty
(Duran 2002).
Although a generalised assessment of the cost-effectiveness of all Bolsa Escola
programmes in the several Brazilian states has not been made yet, studies such as the
one on Brasilia DF showed that its programme increased school attendance and
reduced child labour using less than one percent of Brasilia's budget. However, in
poorer states similar results needed considerably more resources. For instance,
Salvador Bahia would have needed 20 percent of its own budget (World Bank 2002)
and according to Levinas and others (2001) the vast majority of municipal programs
in 1998-99 served a very small fraction of the poor population. Similarly, with regards
to the PETI programme drawbacks have been pointed out in terms of its impact on
child labour reduction. Given that there is no precise data on the amount of children
involved in labour activities, selection and targeting has been difficult. Furthermore,
in some municipalities the criterion that children must actually be working has been
relaxed and the programme was extended to all poor families (Bertranou and Gushka
2002).
In low-income countries impact evaluation is still an uncommon practice. Impact
assessment studies for income transfer programmes in Nicaragua, Honduras and
Paraguay are unable to confirm a significant rise in consumption among beneficiary
households, instead transfers appeared to have stabilised consumption in a context of
generalised consumption declines for the population in rural areas. These programmes
show a large impact on the poverty gap, but little impact on the poverty headcount.
They also show a significant rise in school enrolments and attendance among
beneficiary households (but impact on child labour are less clear cut), and large
improvements in primary health care utilisation (Barrientos and Santibañez,
forthcoming).
Although many human development income transfer programmes combine the
„demand-side‟ with the „supply-side‟ as an integrative way of attacking the constraints
to improving children development in poor localities, there is still ongoing discussion
about the effectiveness of each component. Comparing the cost effectiveness of
income subsidies for schooling with school infrastructure development in Mexico,
Coady and Parker (2002) show that the demand-side component (income subsidies)
19
Chile‟s population is predominantly urban, 85.9 percent in 2000, but the incidence of poverty is
higher in rural areas (Gana 2002).

24
are a much more cost-effective way of increasing education levels relative to building
additional schools.
The sustainability of social assistance programmes is an important challenge. Three
main factors appear to be central to shaping the sustainability of social assistance
programmes: government commitment, the credibility and quality of the program, and
fiscal sustainability. In particular, it has been observed that sustainability of income
transfer programmes depends on a combination of funding constraints, criteria for
establishing the period of participation in the programme and exit strategies.
Bertranou and Gushka (2002) showed that in Argentina, by 2002, the non-
contributory pension programme programme accounted for three percent of the
aggregate social security expenditure and 0.2 percent of the GDP. Although such
amount was „manageable‟ for the government – at least until the economic crisis
occurred in 2004 – it has also been acknowledged that it has been the fiscal complex
organisation (divided between three co-existing levels of government: national,
provincial, and municipal) which has adversely affected their implementation in terms
of efficiency and effectiveness. In Brazilian non-contributory pension programme,
concern was raised about the sustainability of the rural non-contributory pension
programme. In particular, the fact that the legislation included a subsidy from workers
affiliated to social insurance to finance the rural pension programmes raised concerns
from trade unions representing largely urban workers. In fact, the financing of the
rural non-contributory pension scheme comes in full from general tax revenues and
involves no subsidy from social insurance contributions (Schwarzer and Querino
2002). In Costa Rica, there are concerns that a recent increase of expenditure in
pensions might have implications for social expenditures on health, education and
social services.
A response to sustainability concerns it to aim to combine fragmented programmes
into a single large programme. For instance, Bolsa Familia in Brazil will bring
together four separate programmes, Bosal Familia, PETI, a gas subsidy programme
and a food subsidy programme.
20
However, a tendency towards integration has not
been widely adopted by policy makers. In many programmes the institutional
architecture follows presidential initiative. The Office of the President has a direct
influence over policy formulation while service delivery is in charge of a plural
network. However, a plurality of providers can constitute an obstacle to the longer
term institutionalisation of the programmes (Soares and Britto 2007). Furthermore, as
Moore (2008) illustrates for Honduras, the combination of political instability that
surrounds presidential initiatives, the influence of international agencies, and the
creation of parallel administrative structures is not conducive to the institutional
strengthening that is needed to convert pilot short-term income transfer programmes
into long term anti-poverty strategies.
In sum, there has been a rapid development of social assistance programmes in Latin
America. Whilst there is large diversity in programmes design, non-contributory
20
In discussion of social policy in Brazil, basic income policies have proved attractive as a means to
integrate social programme. Basic income policies are a response to the high income and wealth
inequality in the country. According to Senna et al. (2007) "The Brazilian minimum income programs
propose to promote the linkage with other social policies and programs, creating the possibility that, in
theory, the typical fragmentation of the Brazilian social policies will be overcome, facilitating the
adoption of intersectoral actions" (Senna et al. 2007).

25
pension schemes and human development income transfer programmes are the main
programme types in the region. The section has identified the main development and
provided a discussion of their design, impact and institutionalisation. Emerging forms
of social assistance are the main policy responses to poverty and vulnerability. They
have the capacity to make an important contribution to these objectives.
Social protection and social equity
Inequality is one of the striking features of many Latin American countries.
Inequalities in income and opportunities are widely regarded as one of the barriers to
human, social and economic development in the region. The fundamental aim of
social protection is to provide the population with protection in the event of hazards
that threaten to push people into poverty, and to strengthen exit from poverty among
those already in persistent poverty. To the extent that social protection is rightly
understood as one of the main components of a broader social equity agenda
(ECLAC 2006), it can contributing to reducing social inequalities. . The reduction of
inequality is not the primary role of social protection. Policies that can directly
redistribute assets and opportunity can have a stronger impact on reducing inequality.
In the following circumstances, social protection policies can make an important
contribution to reducing inequality.
Firstly, existing inequalities reflect people‟s degree of exposure to risk and their
ability to cope with those risks. Underprivileged groups bear the brunt of the
disadvantages associated with social exclusion and greater exposure to risk (ECLAC
2006). Social protection policies, in protecting these groups, can reduce economic
inequalities among those in poverty.
Secondly, and notwithstanding the positive trends of economic growth in many Latin
American economies prior to the current financial and economic crisis, the formal
labor market has not evolved in a inclusive way so as to produce stability for the
workforce, which is more vulnerable to unemployment, precariousness and
informality. Existing contributory social protection mechanisms have demonstrated
to be insufficient to protect the bulk of the population from the risks associated with a
possible loss of income, health, ageing, gender discrimination or other factors.
Poverty reduction programs aimed at promoting social inclusion via solidarity-based
financing mechanisms are one of the most tangible expressions of the principle of
collective equity. Without exception, new forms of social assistance in Latin America
target the poor and poorest (Barrientos and Santibañez 2009).
Thirdly, segmented and incomplete pension regimes have become a major source of
inequity (ECLAC 2002). At present, many social insurance schemes in Latin America
are regressive. Reform of pension schemes that establish more inclusionary and
comprehensive social protection systems can have positive effects on the reduction of
inequality (Arza 2008).
Finally, social rights are increasingly being invoked as a guiding principle for social
protection. Social protection can also advance and embed rights. As Piron (2004)
suggests, this can in addition have an effect on citizenship.

26
3. Social Protection in the Ford Foundation
The purpose of this short section is to speculate on the potential areas in which Ford
Foundation work and social protection initiatives in Latin America overlap. The
discussion in this section is very preliminary and the points below are not grounded
on a detailed assessment of Ford Foundation work on social protection in Latin
America.
21
Ford Foundation regional officers are in a far better position to make this
assessment, and the points discussed are offered as an input to that assessment.
Key features of social protection reforms
As noted in the previous section, social protection in Latin America and the Caribbean
shows major changes in the last two decades: the reform of social insurance pensions
and health insurance; and the rapid expansion of social assistance. These changes
have affected most countries in the region and have re-shaped social protection in the
region. The reform of social insurance has more or less completed its cycle. Countries
with unreformed social insurance institutions are not contemplating fundamental
reforms. Early reformers like Chile are engaged in reforming individual retirement
accounts to improve coverage and benefit certainty, but the second generation reforms
mainly extend the role of government in pension provision. The expansion of social
assistance, on the other hand, is still at the early stages of its cycle. In Mexico and
Brazil, there appears to be a medium term dynamics which further institutionalises
Oportunidades and Bolsa Familia. In lower income countries, income transfer
programmes have not developed the same momentum, in Nicaragua for example the
Red de Protección Social is being scaled down. In Honduras, El Salvador, and
Paraguay, the expansion of social protection has been slow and fitful.
International attention has concentrated on the innovative design of the new social
assistance programmes in Latin America, and the „conditional cash transfer‟ tag is
widely used among policy makers to brand the new programmes and similar ones
emerging in other regions. For our purposes, it is important to focus on other features.
Firstly, perhaps the most important feature of the new social assistance programmes is
their scale. Most countries in the region had by the 1990s some form of social
assistance addressed to those in poverty, but it was largely residual and fragmented.
By comparison, Bolsa Familia reaches over 11 million households, while
Oportunidades reaches over 5 million households. Chile Solidario is intended to reach
all households in extreme (food) poverty. Scale, not design, is perhaps the defining
feature of the new social assistance programmes in the region.
Secondly, contrary to the „conditional cash transfer‟ tag description of the
programmes, they combine a range of interventions, in-kind transfers, and services. In
21
This is in part due to the limited information on which the discussion below is based. The listing of
recent FF grants for the region which was made available provides only a limited insight into social
protection work. In fact, social protection was not used as a keyword in the classification of grants, so
the list provided may have significant gaps (this was confirmed by FF officers for the Southern Cone
and Andean regional office). The ToR for the Global Social Protection study does not request an in
depth examination of Ford Foundation work on social protection. The main focus is on mapping broad
social protection developments and trends.

27
most programmes, except for unconditional transfers such as pensions or family
allowances, income transfers are just one component. Many researchers have
remarked upon the fact that the non-income effects of the programmes are very
important. In fact the non-income effects, such as improvement in school attendance
and health status, are crucial to delivering long term reduction in intergenerational
poverty transmission. The combination and integration of interventions is a primary
feature of new social assistance programmes.
Thirdly, in the more successful and larger programmes, government agencies play a
dominant role in coordinating a wider range of providers. In fact, one of the
weaknesses in the institutionalisation of new social assistance programmes in lower
income countries arises from the presence of delivery agencies which are parallel to
government agencies (the experience of PRAFII in Honduras is a good example).
Combining interventions necessarily involves several agencies, and perhaps a
combination of profit and not for profit providers (in health and education), but a
feature of existing programmes is the central role of government agencies in directing
delivery.
…and appropriate role of independent organisations
These three key features pose some challenges in defining an appropriate role for
independent organisations supporting social protection. An organisation like FF has
necessarily a reduced scale focus. FF work on the IDAs in the USA, for example,
focuses on providing assistance to government and non-governmental agencies in
adopting and providing IDAs. This is done by supporting „demonstration‟
programmes, supporting technical assistance, providing initial funds to leverage other
funders, and generating and disseminating knowledge. As a USA based organisation,
with a longstanding involvement in public policy, FF was in a good position to
advance this work. The large scale of social protection programmes in Latin America,
and their governmental institutionalisation (admittedly incipient in lower income
countries) defines a different policy environment. FF capacity to influence existing
large scale social protection programmes is perhaps limited, and its potential
involvement in establishing new programmes in lower income countries in the region
would need careful consideration especially as the Inter-American Development Bank
and other multilateral and bilateral donors are in a better position to provide medium
term financial and technical support. FF does not have a long track record and
experience in core social protection in the region. An important point emerging here is
that external organisations wishing to involve themselves in social protection work in
the region would need to have a strategy for engaging and collaborating with
government agencies.
The fact that new social assistance programmes aim to integrate a range of
interventions addressing the multidimensional nature of poverty does generate spaces
for engagement for organisation such as FF. The scope of income transfer
programmes in Latin America varies across countries, from pure income transfers
such as Bono Dignidad in Bolivia to integrated poverty transfers such as Chile
Solidario in Chile. Human development transfer programmes such as Familias en
Acción in Colombia include schooling, nutrition and health interventions. The
dynamics of these programmes over time suggests a widening in the scope of the
programmes. The extension of PROGRESA to urban areas, for example, led to

28
widening of the range of interventions packaged in the programme to include training
and post school education, small enterprise skills, and others. Apart from
OPORTUNIDADES and Chile Solidario, few social assistance programmes have
given serious consideration to the type of interventions that would enhance
employment and investment in physical or financial productive assets among
beneficiary households. In some countries this is because other agencies or
programmes focus on these areas; but it has also to do with the fact that social
assistance programmes are in their early stages of development or programme cycle.
For organisations such as FF there are opportunities to engage in widening the scope
of the programmes. FF longstanding interest and expertise in micro-saving and asset
accumulation are relevant here.
Income and asset transfers
It will be useful to preface this with some brief comments on the relationship existing
between income and asset transfers. In developing countries, income transfers are a
means to raise household consumption. Income transfers are usually a fraction of
household consumption and are not intended to lift beneficiary households above the
poverty line. Non-contributory pensions, for example, target poor older people and
their households while human development transfer programmes target households
with children or expectant mothers. Asset transfers are also intended to raise
household consumption but mediated through the exploitation of the assets. Among
policy makers, and in some development policy discussions, income and asset
transfers are sometimes perceived as alternatives (as in „giving people fishing rods not
fish‟). There are important differences in the properties of typical (pure) income and
typical (physical or financial) asset transfers
22
- for example short term income
transfers might not translate into long term welfare improvements while asset
transfers may involve short term sacrifices to secure longer term welfare
improvements – they can more appropriately be looked at as complementary.
Increasingly, they are perceived in this light. Bangladesh‟s BRAC learned from their
microfinance programmes that a substantial group of borrowers in extreme poverty
would often be unable to generate sustainable welfare improvements from asset
transfers and designed a specific programme Challenging the Frontiers of Poverty
Reduction- Targeting the Ultra Poor in order to address the specific needs of this
group. This asset transfer programme includes a stipend to help stabilise consumption.
Oportunidades, on the other hand, has over time incorporated skills and small
enterprise development components to the basic transfers and services package to
help support sustainable exit from poverty. At any rate, human development transfer
programmes are specifically aimed at improving the productive assets of beneficiaries
through health, nutrition and education interventions. They can be described
appropriately as productive asset transfer programmes. The effectiveness of poverty
reduction policies requires consideration of the mix and complementarity of income
and asset transfers. The Latin American social protection policy environment provides
a context in which this issue can be fruitfully explored.
22
Other distinctions between typical asset and typical income transfer programmes are important too:
income transfer programmes can be scaled up faster and can reach households with few productive
resources.

29
Proyecto Capital is an innovative project supported by Ford Foundation and managed
by Fundación Capital and the Instituto de Estudios Peruanos aiming to support the
introduction of micro-saving components in income transfer programmes. The project
supports research into the opportunities for micro-saving instruments in five
countries: Argentina, Chile, Colombia, Ecuador and Peru; engagement with policy
makers; and the design, development and testing of relevant instruments. In its first
phase in 2009, the project is expected to produce needs-based proposals for
appropriate components. At this point in time, there is an expectation that a further
phase in the project will support the actual piloting of these components. Proyecto
Capital provides an example of how FF can pursue ways in which to engage with
emerging social assistance programmes in ways that make full use of its experience
and expertise.
It will be interesting to find out what proposals emerge from this project. In the USA,
FF has championed Individual Development Accounts and Child Savings Accounts as
saving instruments. The experience with individual retirement accounts in the region
suggests earmarked saving instruments might not be especially attractive to low
income households. As discussed in section 2 above, individual retirement accounts
introduced in several Latin American countries require earnings related contributions
from workers to retirement accounts, with the balance available only at retirement or
the death of the beneficiary. Coverage is very low, and restricted to formal
employment. Individual retirement accounts proved particularly unattractive to self-
employed workers. While the low public trust in financial institutions in the region
was probably a factor, the illiquidity of the saving instrument was often given as a
reason for not contributing among workers. IDAs in the USA provide matching
saving contributions by participants providing withdrawals are associated with higher
education, housing, or small business investment (similar to provident funds in some
developing countries). Liquidity of savings is likely to be an issue for households in
poverty in Latin America.
23
Rights and social guarantees
Another area where FF longstanding interests and expertise and the new social
protection policy environment in Latin American suggest potential synergies is the
current interest with social guarantees as a means to ensure economic and social rights
are realised. Health insurance and care in the region are strongly segmented, with a
plurality of providers: public, for profits, not for profits, etc. Chile‟s recent AUGE
programme seeks to ensure a set of basic services are provided on demand, across the
plurality of health providers. It establishes a set of basic rights to care which all
Chilean citizens can exercise regardless of their health insurance or health provider.
On a different level, human development transfer programmes ensure access to
primary health care for households in extreme poverty, with co-responsibilities for
both beneficiary households and programme agencies clearly established at the outset.
In lower income countries, human development transfer programmes include supply
side interventions to ensure these co-responsibilities. Social guarantees constitute an
interesting route to ensuring economic and social rights, especially among those in
poverty and subject to social exclusion.
23
It is also an issue for USA participants, Ford Foundation reported that 2/3 of participants in a 14-site
impact study showed un-matched withdrawals.

30
In pursuing a rights-based approach to social protection Latin American countries
need to upgrade their social contracts between the various agents of the state and civil
society. This entails designing of institutions and policies capable of generating space
for broad participation, in particular to enable the participation of the most
disadvantaged groups and sub-national territories. The FF longstanding support to
institution building could target institutional design – and participatory mechanisms –
for social protection systems. Connected to this is the necessary support for the
formation of highly skilled professionals on issues like tax structure and burden, fiscal
and social services delivery decentralization, expansion and selectivity of social
expenditure, and monitoring and impact assessment of social protection programs. On
these same lines, „lesson learning‟ supported by donors and independent
organizations has been shown to be essential to the development of evidence-based
policy. While important progress has been made to systematise poverty reduction
programmes (for instance the lesson learning program carried out by Rimisp and
FIDA (Aguirre et al 2009)), less has been done on social protection. The FF could
also extend its support to this area.
In a number of LAC countries the government revenue used to finance social
protection programs includes not only central government tax revenue but the
proceeds from the sale of natural resources obtained both at central and regional (sub
national) levels. Furthermore, in some mineral-rich countries large mining and oil/gas
companies are implementing social assistance projects as part of their corporate social
responsibility (CSR) programs. However, neither local governments nor companies
seem to be fully aware of the country‟s social protection system and how their CSR
actions can contribute to it. The FF could explore collaboration with the corporate
sector and the ways in which affected communities by mineral activity can participate
in social protection-oriented CSR measures.
4. Conclusions and further research
The incidence of poverty and vulnerability showed a step increase in the 1980s, and
halting recovery in the 1990s. Poverty is higher in rural areas, and for indigenous
groups. Life course factors are also an important determinant of poverty, but policy
responses have proved effective in reducing old age poverty in some countries. Most
recently, a downward trend in poverty incidence is under threat as a consequence of
the global financial crisis.
Social protection institutions in Latin America at the start of the 1980s can be best
described as „truncated‟, with social insurance schemes accessible to workers in
formal employment, and social assistance largely underdeveloped. In the period since
1980s two main reforms have taken place. Firstly the reform of social insurance
programmes in ten countries in the region, replacing social insurance funds with
individualised saving retirement plans. The private pension funds have attracted at
best a fraction of the labour force, reinforcing the „truncated‟ nature of social
protection institutions. Two thirds of the labour force remains unprotected in the
region. Secondly, new forms of social assistance have emerged in many countries,
primarily human development income transfer programmes focused on households in
poverty, and non-contributory pension schemes.

31
New social assistance programmes are designed to reduce poverty and vulnerability,
making poverty reduction, and in particular chronic poverty, the cornerstone of social
policies. Social assistance programmes, and especially human development income
transfer programmes, have the potential to make a significant contribution to reducing
extreme poverty in the region, but important challenges remain. First, the
effectiveness of these programmes requires attention is paid to the level of benefits,
and the duration of support. Second, it will be important to ensure that existing
programmes move in the direction of stronger institutionalisation and sustainability.
This is especially true of programmes in low income countries in the region, where
financing and delivery capacity are limited. Third, implementation and delivery of
social assistance programmes require long term partnerships under the direction of
secure and well resourced public sector agencies. In this context, it is urgent to
consider the role of non-government organisations with a commitment and expertise
in social protection in meeting these challenges.

32
APPENDIX One: Terms of Reference for Scoping Study and
Regional Papers
Social protection as development and transformation: Assessing the field
Scoping study for the Ford Foundation – Interim note
Sarah Cook
Institute of Development Studies
November 2008
The purpose of this grant is to provide a review of current approaches to social
protection as responses to vulnerability and insecurity across different regional
contexts. Currently regional reviews of social protection programmes and approaches
are underway, covering Africa, Latin America, South and South East Asia, and the
US. The reviews are being structured around life-course related vulnerabilities, the
needs of different population groups at different stages of the life-course and how
these are (or are not) being met through formal or informal social protection
interventions.
While we have not yet completed the reviews, we can see emerging variations across
regions in terms of need or vulnerabilities; social protection approaches and
responses, and the gaps or limitations of existing provisions. Among the challenges
faced in this initial stage of the work are the huge scope of the review being
undertaken and the need to set some boundaries on the exercise; and finding a
framework for analysis that allows for multiple dimensions of regional variability
while also being able to provide some generalisable lessons. This is highlighted in the
preliminary efforts to identify Ford Foundation grants on social protection. Given that
social protection has not been an articulated line of work, there is limited shared
understanding of the terminology within the Foundation. While many grants could be
defined as falling broadly within the rubric of social protection, this is often not their
primary purpose. There may be resistance to applying this label – which is often
interpreted as „welfarist‟ – to projects whose motivation embodies a social justice,
equity or rights agenda. While we will attempt to define social protection more
broadly – incorporating a developmental and transformative agenda of moving from
protecting livelihoods to empowering citizens in demanding their entitlements – this
debate will remain on-going within the community of researchers and practitioners
working in this field.
As initially proposed, our organizing framework for the reports involves a typology
based on an analysis of the intersection of three sets of issues:
 The social protection needs and vulnerabilities across different stages of the
life course from a „productive‟ standpoint. This would distinguish between
those who are going to be economically active in the future (the young), those
who are currently economically active or seeking to be active (working-age
adults) and those whose activity is declining (the elderly).
 Patterns of risk, vulnerability and exclusion experienced by these different
cohorts in different contexts, given prevailing structures of protection.

33
 A review of measures in place to address excluded and vulnerable groups and
the extent to which they incorporate a concern with a broader developmental
or transformative agenda.
A suggested outline of the report is attached below. Draft regional papers and concept
notes will be produced in December/January for discussion, review and revision prior
to organizing a workshop, planned tentatively for April.
Ford Foundation Social Protection Scoping Study
Suggested outline of report
Introductory chapter (SC/NK)
1) Definitions and concepts of social protection, key approaches and dominant
discourses (what is SP trying to achieve: protection, promotion)
2) Main instruments of social protection (conventional / unconventional)
3) Special issues – e.g institutions/actors; financing; politics and policy
processes; marginalised / hard to reach groups…
4) Organising framework for subsequent chapters – life course mapping of
programs: gender and generation (for example, see Kabeer 2008 Chapter 3)
Children (infant/pre-school/school); adults (workers/non-workers);
retirees/elderly (economically active – past/present/future)
Measures which take productive work as the raison d‟etre…
Measures which take reproductive work…
5) Analytical principles: evaluation of programmes in terms of own objectives
(protection, promotion, transformation; other criteria?)
6) Lessons in terms of broader objectives: strengths, limitations, gaps
7) Implications for Ford and other organisations.
Regional chapters (max. 15,000 words)
(USA/Europe; Sub-Saharan Africa; Latin America; South Asia; East and SE
Asia)
1) Introduction (short) – regional context and challenges
2) Situation analysis: nature of poverty, vulnerability and marginalisation,
approaches to / dominant discourses on social protection within that region
3) Life course mapping of existing measures:
Gender and generation
Social difference (race, caste, ethnicity, migrant status,) how are these
addressed
4) Analytical framework: evaluation in terms of immediate and broader
objectives plus lessons
5) Conclusions – key points – strengths / limitations; gaps; recommendations
Possible additional section or part of analysis and conclusions: Key issues / cross-
cutting themes e.g.:
 Institutions/actors
 Politics and policy processes
 Financing of SP

34
 Neglected / ignored groups or issues (relevant to specific region) – eg
housing, domestic violence…
 Implications for justice, citizenship, equity
We aim to write some short concept notes/theme papers on these (or other) themes: if
they emerge as prominent issues from the regional reviews they could be highlighted
there; and/or the information can be developed/incorporated in specific short papers.
References (overall + by region)
Appendices: review / „database‟ of programmes

35
References
Aguirre Francisco, Tania Ammour, Octavio Díaz, Eduardo Ramírez and Rafael Sarno
(2009) Aprendizajes Innovadores para la Reducción de la Pobreza Rural.
Aportes del FIDA en America Latina y El Caribe. RIMISP and
FIDAMERICA, Santiago.
Arza, Camila (2008) „Pension Reform in Latin America: Distributional Principles
Inequalities and Alternative Policy Options‟ Journal of Latin American
Studies, Vol. 40, No. 1, pp 1-28
Asociación Internacional de Organismos de Supervisión de Fondos de Pensiones
(2008), 'Boletín Estadístico AIOS - December 2007'
Banco Mundial (2007) Protección social en el Perú. Cómo mejorar los resultados
para los pobres? Banco Mundial, Lima.
Barrientos, A. (1998), Pension Reform in Latin America, Aldershot: Ashgate.
Barrientos, A. (2004a), 'Latin America: towards a liberal-informal welfare regime', in
I. Gough; G. Wood; A. Barrientos; P. Bevan; P. David and G. Room (eds.),
Insecurity and welfare regimes in Asia, Africa and Latin America, Cambridge:
Cambridge University Press, pp. 121-168.
Barrientos, A. (2004b), 'Women, informal employment and social protection in Latin
America', in C. Piras (ed.) Women at Work. Challenges for Latin America,
Washington DC: Inter-American Development Bank, pp. 255-292.
Barrientos, A. (2006), 'The missing piece of pension reform in Latin America:
Poverty reduction', Social Policy and Administration, vol. 40, no. 4, pp. 369-
384.
Barrientos, A.; R. Holmes and J. Scott (2008), Social Assistance in Developing
Countries Database version 4, Report, Manchester: Brooks World Poverty
Institute and CPRC.
Barrientos, A. and C. Santibanez (forthcoming), 'Social policy for poverty reduction
in low income countries in Latin America: Lessons and Challenges', Social
Policy & Administration, vol.
Barrientos, A. and C. Santibañez (2009), 'New forms of social assistance and the
evolution of social protection in Latin America', Journal of Latin American
Studies, vol. 41, no. 1, pp. 1-26.
Bertranou, F. and C. O. Grushka (2002), The Non-Contributory Pension programme
in Argentina: Assessing the impact on poverty reduction, ESS Paper 5,
Geneva: International Labour Organization.
Bertranou, F.; C. Solorio and W. van Ginneken (2002), Pensiones no-contributivas y
asistenciales. Argentina, Brazil, Chile, Costa Rica y Uruguay, Santiago: ILO.
Bigsten, A. and J. Levin (2005), 'Growth, income distribution and poverty: A review',
in A. Shorrocks and R. van der Hoeven (eds.), Growth, inequality, and
poverty, Oxford: Oxford University Press, pp. 251-276.
Bourguignon, F. (2005), Development strategies for more and better jobs, Washington
DC: The World Bank.
Cardoso, E. and A. Portela Souza (2003), The impacts of cash transfers on child labor
and school attendance in Brazil mimeo, Sao Paolo: Department of Economics
University of Sao Paolo.
CEPAL (2007), Panorama Social 2006, Santiago: CEPAL.

36
Coady, D. (2003), Alleviating structural poverty in developing countries: The
approach of Progresa in Mexico, mimeo, Washington DC: International Food
Policy Research Institute.
ECLAC (2008), Social Panorama of Latin America 2008, Report, Washington DC:
ECLAC.
Fiszbein, A. (2005), Beyond truncated welfare states: Quo Vadis Latin America?,
Washington DC: The World Bank.
Gana Cornejo, P. A. (2002), 'Las pensiones no contributivas en Chile: pension
asistenciales (PASIS)', in F. Bertranou; C. Solorio and W. Van Ginneken
(eds.), Pensiones no contributivas y asistenciales: Argentina, Brazil, Chile,
Costa Rica y Uruguay, Santiago: Oficina Internacional del Trabajo.
Ghilarducci, T. and P. L. Ledesma (2000), 'Union's Role in Argentine and Chilean
Pension Reform', World Development, vol. 28, no. 4, pp. 753-762.
Gill, I. S.; T. Packard and J. Yermo (2004), Keeping the Promise of Social Security in
Latin America, Washington DC: Stanford University Press and the World
Bank.
ILO (2000a), Panorama Laboral 2000, Informa 7, Lima: International Labour
Organisation.
ILO (2000b), World Labour Report 2000, Geneva: International Labour Office.
ILO (2003), Panorama Laboral 2003, Lima: International Labour Organisation.
ILO (2006), Panorama Laboral 2006, Lima: International Labour Organisation.
Kay, S. J. (1999), 'Unexpected Privatizations. Politics and Social Security Reform in
the Southern Cone', Comparative Politics, vol. 31, no. 4, pp. 403-422.
Lindert, K.; E. Skoufias and J. Shapiro (2005), Redistributing income to the poor and
the rich. Public transfers in LAC, Discussion Paper, Washington DC: The
World Bank.
Martinez Franzoni, J. (2005), Reformas recientes de las pensiones en Costa Rica:
Avances hacia una mayor sostenibilidad financiera, acceso y progresividad del
primer pilar de pensiones, miemo, San Jose: Universidad de Costa Rica.
Martinez Franzoni, J. and C. Mesa-Lago (2003), La reforma de la Seguridad Social en
Costa Rica en pensiones y salud, San José: Fundación Friedich Ebert.
Mesa-Lago, C. (1991), 'Social Security in Latin America and the Caribbean: A
Comparative Assessment', in E. Ahmad; J. Dreze; J. Hills and A. Sen (eds.),
Social Security in Developing Countries, Oxford: Clarendon Press, pp. 357-
394.
Mesa-Lago, C. (1999), 'Política y reforma de la seguridad social en América Latina',
Nueva Sociedad, vol., no. 160, pp. 133-150.
Mesa-Lago, C. (2007), Reassembling Social Security. A Survey of Pensions and
Healthcare Reforms in Latin America Oxford: Oxford University Press.
Mesa-Lago, C. (2008), 'Proteccion Social en Chile: Reformas para mejorar la
equidad', Revsita Internacional del Trabajo, vol. 127, no. 4, pp. 421-446.
MIDEPLAN (2004), Conceptos Fundamentales. Sistema de Protección Social Chile
Solidario, Report, Santiago: MIDEPLAN.
Molyneux, M. (2006), 'Mothers at the Service of the New Poverty Agenda:
Progresa/Oportunidades, Mexico's Conditional Transfer Programme', Social
Policy and Administration, vol. 40, no. 4, pp. 425-449.
Pinheiro, V. C. (2005), 'The Politics of Pension Reform in Brazil', in C. A. Crabbe
(ed.) A quarter century of pension reform in Latin America and the Caribbean.
Lessons learned and next steps, Washington DC: Inter-American
Development Bank, pp. 187-217.

37
Queisser, M. (1998), The Second-Generation Pension Reforms in Latin America,
Working Paper AWP 5.4, Paris: OECD.
Rofman, R. (2005), Social Security Coverage in Latin America, Report, Washington
DC: The World Bank.
Saldain, R. and M. Lorenzelli (2002), 'Estudio del programa de pensiones no
contributivas in Uruguay', in F. Bertranou; C. Solorio and W. Van Ginneken
(eds.), Pensiones no contributivas y asistenciales: Argentina, Brazil, Chile,
Costa Rica y Uruguay, Santiago: Oficina Internacional fdel Trabajo, pp. 221-
264.