This chapter reviews Paraguay’s main social assistance programmes, with a focus on design, coverage, targeting performance and fiscal costs. It examines three flagship programmes – Tekoporã, Adulto Mayor, and Hambre Cero – and analyses reform scenarios to enhance poverty reduction and cost-effectiveness. The chapter also presents detailed costing of the proposed reform options.
Financing Social Protection in Paraguay
The Role of General Revenues and Social Security Contributions
Abstract
Tekoporã is Paraguay’s main conditional cash transfer programme, designed to reduce poverty and vulnerability among beneficiary households. The programme combines direct income support with family and community-based support services delivered by field staff. It targets families identified as monetary or multidimensionally poor (Box 4.1) and provides additional allowances for children, pregnant women, older persons without pension entitlements, and persons with disabilities.
Adulto Mayor is a non-contributory social pension that guarantees a minimum income for Paraguayans aged 65 and over who are not covered by contributory pension schemes. Persons with severe disabilities can access the benefit from the age of 60 while indigenous people are eligible from age 55. The transfer is equivalent to 25% of the minimum wage.1
Hambre Cero is a national school meals programme that provides free meals in public schools and some subsidised private schools2 with the objective of ensuring that all children in initial and primary schools have access to adequate nutrition during the school day. The programme also covers middle schools in the regions Central, Capital and Presidente Hayes. In public schools, coverage rates are 97% in initial education, 95% in basic education, and 19% in middle education, with a total coverage rate of 81% of enrolled children. Meals are provided for approximately 180 school days per year. The programme also supports formalisation by requiring contracted service providers to employ Tekoporã beneficiaries as formally registered workers.
Selected social assistance programmes in Paraguay, 2024 or latest year available
Source: Administrative data provided by the Paraguayan Ministry of Social Development (MDS).
Hambre Cero operates through a public procurement process. The government sets reference prices for school meals prior to each tender, and bids submitted by providers cannot exceed these reference prices by more than 5%. Private enterprises compete for contracts lasting two years and are required to have relevant experience. Contracts also include local development requirements: at least 10% of supplies must be sourced from regional farmers and 5% from small local businesses. While firms may be based outside the region, they must maintain a logistics base within 100 kilometers of the delivery area. The financing of the Hambre Cero programme through non-tax revenues from Paraguay’s two hydropower dams is only secured until 2027, raising concerns about its sustainability and long-term impact.
Paraguay’s social assistance programmes extend beyond cash transfers to include training, community-based support, and productive development components (Table 4.1). Tenonderã, for example, links micro-credits with capacity-building for entrepreneurship, while Tekoporã incorporates community guidance and leadership initiatives.
Coverage of the programmes remains incomplete3
Some of Paraguay’s main social assistance programmes are still in the process of being rolled out. For programmes such as Hambre Cero and Adulto Mayor, the currently modest coverage levels largely reflect their partial implementation rather than shortcomings in eligibility determination alone (see Chapter 1). As a result, coverage indicators should be interpreted with caution, taking into account both administrative capacity constraints and available fiscal space.
Tekoporã targets poor households, but still fails to reach all those eligible. According to the analysis of the 2024 Encuesta Permanente de Hogares Continua (EPHC) survey data, only 17% of eligible families are covered by the programme (Table 4.2).4 Nevertheless, the programme has made significant progress in reaching particularly vulnerable groups: for example, coverage among indigenous families reaches 92%. This reflects the government’s initial decision to prioritise the rollout of the programme in communities facing the highest levels of poverty and vulnerability. At present, however, no data are available to assess the programme’s effectiveness in reaching persons with disabilities.
Coverage of selected social assistance programmes in Paraguay, 2024 or latest year available
Notes: Data for Tekoporã and Adulto Mayor is based on EPHC survey data from 2024 available in INE’s website in 2025. Households within this table are defined as people living under the same dwelling. The underlying survey data does not include the departments of Boquerón and Alto Paraguay, nor indigenous communities. The coverage of Tekoporã for indigenous households is calculated using administrative data. There is no data available to calculate the coverage for disabled people under Tekoporã. Data for Hambre Cero is based on administrative data provided by the MDS from July 2025.
Source: OECD based on (EPHC, 2024[1]); INE, Censo Nacional de Población y Viviendas, 2022; MDS, 2025.
The outreach of Tekoporã’s support staff, known as family guides, varies considerably across regions. Family guides provide assistance on issues such as vaccination and school enrolment. As of October 2025, 54% of Tekoporã beneficiary families had not been assigned a family guide, pointing to budget constraints in the delivery of complementary support services.
Approximately 40% of beneficiaries receive benefits even though they are not eligible. These so-called inclusion errors mostly occur among families in the middle of the income distribution, particularly those in the fourth to sixth income deciles (Figure 4.1, Panel A). This is likely due to the way households are classified using proxy means tests, which sometimes identifies families as deprived even if they are above the monetary poverty line or are not multidimensionally poor.
Adulto Mayor achieves high coverage among poorer old-age adults, with coverage declining steadily as income increases. In 2024, 62% of the eligible population aged 65 and over was covered by Adulto Mayor. Coverage exceeds 85% in the first income decile and remains close to 80% in the second income decile, before declining progressively across higher income deciles (Figure 4.1, Panel B). This pattern reflects the programme’s rollout strategy, which initially prioritised enrolment among the poorest older adults.
Coverage rates and targeting errors by income decile, 2024
Notes: Tekoporã beneficiaries are considered eligible if they are monetarily or multidimensionally poor. Multidimensional poverty exists – albeit at declining levels – across all income deciles, and thus eligible individuals are also present in the higher income deciles. The rural poverty line corresponds to an income level around the 2nd income decile, while the urban poverty line corresponds to an income level around the 3rd income decile. This explains why some rural households in the 2nd income decile are classified as non-poor, while urban households in the 3rd income decile may still be considered poor. Eligibility for Tekoporã is determined based on income before transfers. If transfers lift a household out of poverty, this is not classified as an inclusion error. Adulto Mayor’s coverage rates are shown for individuals aged 65 and over and does not include indigenous or disabled people who are eligible for the programme at lower ages. The data do not include the departments of Boquerón and Alto Paraguay, indigenous communities, collective housing and live-in domestic workers.
Source: OECD based on (EPHC, 2024[1]).
The Adulto Mayor and Hambre Cero programmes are not subject to income-related inclusion errors, as they are not means-tested programs. Adulto Mayor is designed as a universal programme and, as such, its coverage extends beyond poor households. Among individuals aged 65 and over who meet Adulto Mayor’s eligibility criteria, only around 13% are classified as multidimensionally poor. Similarly, Hambre Cero also extends to households that are not necessarily poor. Its benefits are tied to school enrolment and apply to both public and certain subsidised private schools, thereby covering families across the income distribution. Exclusion remains a challenge as well, in part because the programme is still being rolled out, leaving a significant share of children in middle school without coverage (Figure 4.2).
The Ministry of Social Development (MDS) applies a sequential targeting process to determine eligibility for Tekoporã. The first stage relies on a geographic filter, whereby communities are selected based on the concentration of poverty, using a combination of monetary poverty and multidimensional poverty indicators derived from the nationally representative household survey (EPHC). Once a community is selected, field teams conduct household visits to collect detailed information using a registration questionnaire. The variables included in this form are selected based on their predictive power for household income in the EPHC, consistent with a proxy-means testing methodology.
The household assessment questionnaire collects a broad set of variables used as proxies for income and living standards. These include 15 questions on housing conditions (such as the type of walls, floor, roof, and access to water and sanitation), 20 questions on ownership of durable goods and services (ranging from refrigerators and vehicles to internet access and waste disposal) and approximately a dozen questions on household composition, education, and work status. In total, the form incorporates more than 50 variables to construct a comprehensive profile of household living conditions. These data are processed through a Proxy Means Test to estimate household income and assess multidimensional poverty, enabling the classification of households as poor or non-poor in the absence of reliable income information.
Eligibility is cross-checked against social security records. Prior to enrolment in Tekoporã, the MDS coordinates with the social security institution (IPS) to verify whether household members hold formal employment or receive pensions. This verification aims to prevent the inclusion of households with formal incomes exceeding the programme’s eligibility threshold.
Source: MDS.
Coverage of the Hambre Cero programme varies across regions and school cycles (Figure 4.2). In July 2025, coverage in Inicial and Básica schools was close to universal, exceeding 90% in all regions except for Asunción. By contrast, coverage in Media schools remains limited, reaching only 19%, with several regions not yet covered. These differences reflect the phased roll‑out strategy adopted by the MDS, which prioritised Inicial and Básica schools before extending the programme to Media education. While coverage within each school cycle shows relatively limited regional variation, disparities are notably larger for Media schools, where implementation has not yet been completed, except for the regions Central, Capital and Presidente Hayes. Coverage of subsidised private schools is also currently concentrated in a small number of regions (e.g. Boquerón, Alto Paraguay and Ñeembucú), where the availability of public schools is particularly low.
Note: The maps show the regional distribution of students in public schools (Básica, Inicial, Media) benefitting from the Hambre Cero programme.
Source: OECD based on administrative records.
Programmes are effective in reducing poverty
Paraguay’s social assistance system achieves a modest reduction in monetary poverty and inequality. Overall, existing programmes reduce poverty by 4.5 percentage points and inequality by 1.6 Gini points, closing approximately one quarter of the pre-transfer poverty gap (PGE) (Table 4.3). Around one-third of total transfers contribute directly to reducing the poverty gap (PRE), while 45% of social spending reaches households that are poor prior to transfers, as measured by vertical expenditure efficiency (VEE).
Distributive impact of selected social assistance programmes in Paraguay, 2024
Note: The poverty gap estimate (PGE) measures the proportion of the total pre-transfer poverty gap that is filled (or eliminated) by social transfers. It captures the extent to which transfers reduce the depth of poverty among those who were poor before receiving transfers. The poverty reduction efficiency (PRE) measures the share of total transfers that reduces the poverty headcount by lifting individuals above the poverty line. Vertical expenditure efficiency (VEE) measures the proportion of total programme spending that is received by households who were poor before transfers. Spillover measures the share of transfers received by pre-transfer poor households that exceeds the amount required to close their poverty gap (i.e. transfers that raise beneficiaries above the poverty line), based on Beckerman (1975[2]).
Source: OECD based on (EPHC, 2024[1]).
Adulto Mayor has the largest impact on reducing monetary poverty, reflecting its comparatively large budget and broad beneficiary coverage. The social pension programme lowers poverty by 3.34 percentage points and inequality by 1.06 Gini points and closes about 16% of the pre-transfer poverty gap (PGE).
The distributive impact of Hambre Cero is increasing as the programme scales up. The present analysis relies on 2024 data, when approximately 360 000 students benefited from school meals. Coverage expanded substantially in 2025, reaching more than one million children, around 72% of all students enrolled in public and private schools. As a result, poverty‑reduction effects estimated on the basis of 2024 data are likely to underestimate the programme’s current impact. Simulations suggest that extending coverage to all publicly enrolled students would more than triple the programme’s poverty‑reducing effect, while modestly improving spending efficiency, as a larger share of resources would accrue to poorer households.
Tekoporã is the most efficient programme in terms of reducing poverty and inequality, despite its relatively limited overall impact. Tekoporã delivers the largest reduction in the poverty gap, estimated at 49%. However, its aggregated effects are constrained by modest budget allocations and limited coverage, resulting in a reduction in poverty by only 0.38 percentage points and inequality by 0.19 Gini points.
Within Tekoporã’s transfer components, the additional child benefit and the basic family benefit are the most effective in reducing poverty. Transfers directed to children exhibit the highest efficiency in reaching poor households, while the family benefit plays an important role. By contrast, the additional transfer to older adults has only a limited direct impact on poverty reduction, reflecting the fact that households with older beneficiaries are, on average, less likely to be among the poorest segments of the income distribution. Poverty rates rise with household size, particularly among families with multiple children. This pattern suggests that reforms, such as increasing the child allowance or shifting from a flat family transfer to a per-person family allowance, could yield larger gains for households at the bottom of the income distribution.
Tekoporã faces trade-offs between efficiency and inclusion that are common in targeted social assistance programmes. Targeted programmes are generally the most cost-efficient instrument for reducing poverty and inequality. However, increasing the precision of targeting criteria also raises the risk of exclusion errors, particularly for households located just above eligibility thresholds. Conversely, expanding the programme to reach a broader share of the vulnerable population inevitably increases inclusion errors, as a higher number of non-poor households become beneficiaries. Evidence suggests, nevertheless, that most ineligible but beneficiary households are not well-off; rather, they tend to be clustered just above the monetary poverty line and are likely to experience forms of multidimensional poverty. Striking an appropriate balance between inclusion and exclusion errors in poverty‑targeted programmes is ultimately a policy choice for national authorities. This decision should take into account not only administrative capacity and feasibility, but also the presence and adequacy of well‑funded and widely rolled‑out categorical programmes, which can help mitigate the adverse consequences of exclusion by reaching a substantial share of the poor, particularly in contexts such as Paraguay, where poverty remains closely linked to lifecycle‑related vulnerabilities.
Categorical programmes with broad coverage, such as Adulto Mayor and Hambre Cero, have a significant impact on poverty reduction despite being less efficient. Assessing these programmes solely through efficiency metrics understates their broader role, as their scale delivers the largest aggregate reductions in poverty and inequality and generates tangible benefits for a wide share of the population. For Paraguay, an effective social assistance system is therefore likely to combine these complementary approaches: broad, categorical programmes that provide basic income security for population groups without a labour-market substitute, including children, older persons and people with disabilities, alongside targeted, poverty-focused programmes such as Tekoporã to reach households who fall outside categorical eligibility yet remain poor.
As expected, lower-income households benefit the most from social assistance transfers. The largest income gains accrue to households in the bottom income decile with an increase in monthly income of around 21%. Gains amount to 12% and 9% in the second and third deciles, respectively (Figure 4.3). By contrast, the impact declines markedly at higher income levels, with income gains falling to 3% or less from the sixth decile onward.
Pre and post transfer average monthly household income measured as absolute and relative income gains
Notes: Bars indicate the average monthly per capita household income before and after transfers received through the social assistance programmes Tekoporã, Hambre Cero and Adulto Mayor. The line graph measures the relative income gains expressed in percent. Results reflect only first-round, static impacts of the social assistance transfers; behavioural responses, multiplier effects, or broader general equilibrium impacts are not considered.
Source: OECD based on (EPHC, 2024[1]).
Social assistance programmes reduce overall inequality slightly by providing proportionally larger income gains for poorer households. On average, social transfers increase household income by around 2% across the population, with these gains concentrated almost entirely among the bottom three income deciles. This result is reflected in a post-transfer Lorenz curve that shifts slightly closer to the line of equality, indicating a narrowing of income disparities (Figure 4.4). While households in the top income decile experience virtually no change in income after transfers, the bottom 30% record meaningful improvements. However, given that middle‑ and higher‑income households account for a large share of total income, changes in their incomes exert a stronger influence on the Gini coefficient than relatively small gains at the lower end of the distribution.
Income distribution before and after the transfer of selected social assistance programmes, 2024
Note: The Lorenz curve presents the distribution of income within a population, showing the proportion of total income earned (vertical axis) by cumulative percentages of the population (horizontal axis). Results reflect only first-round, static impacts of social assistance transfers. Behavioural responses, multiplier effects, or broader general equilibrium impacts are not considered. The selected social assistance programmes include Tekoporã, Hambre Cero and Adulto Mayor.
Source: OECD based on (EPHC, 2024[1]).
Significant improvements in Tekoporã's coverage and design can be achieved at a moderate cost
Tekoporã faces challenges related both to coverage, with gaps in reaching all eligible households, and to benefit adequacy. This section analyses the cost and poverty impact of the following reform scenarios:
Extending the programme to all eligible households using the current benefit amounts.
Extending the programme to all eligible households, increasing the amount per child to PYG 70 000.
Increasing the basic family allowance to PYG 200 000 for families currently receiving benefits.
Increasing the additional amount per child to PYG 70 000 for families currently receiving benefits.
Removing the limit on the number of children per household for families currently receiving benefits.
Modifying the basic family allowance to PYG 50 000 per person.
Analysing the cost of assigning a family guide to all families currently receiving benefits.
Extending the programme to all eligible indigenous households
Extending Tekoporã to all eligible households would cost 0.33% of GDP and could reduce the poverty gap by 21%. In 2024, expenditure on the Tekoporã programme amounted to 0.09% of GDP (Figure 4.5). Extending coverage to all eligible households would increase costs to about 0.33% of GDP, rising further to around 0.39% of GDP if the child allowance were increased to PYG 70 000. At present, Tekoporã reduces the poverty gap by 3%, while an extension to all eligible households would decrease the poverty gap by 21%, or by 24% if the child allowance were also increased.
Cost (as a share of GDP) and pre-transfer poverty gap reduction (in percent) for the status quo and selected Tekoporã expansion scenarios
Notes: The status quo cost refers to costs of benefits to Tekoporã families in 2024 based on administrative records. Disability benefits and costs of Tekoporã for indigenous households are not included in the analysis. Results reflect only first-round, static impacts of transfers. Behavioural responses, multiplier effects, or broader general equilibrium impacts are not considered.
Source: OECD based on EPHC (2024[1]) and administrative records.
Increases in Tekoporã benefit levels could be achieved at low fiscal cost. Policy options such as increasing the family or child allowance, or removing the cap on child benefits, would raise expenditure by less than 0.1% of GDP at current coverage levels (Table 4.4). However, once Tekoporã coverage is expanded, the budgetary impact of these measures would increase by around four to five times.
Expanding the number of family guides across regions and extending Tekoporã to all eligible indigenous households would entail negligible fiscal costs, even at higher coverage levels. Ensuring that all beneficiary households are supported by family guides – based on a minimum ratio of one guide per 150 families – would increase programme expenditure by an estimated PYG 31 000 million, equivalent to less than 0.01% of GDP.
Additional costs for Tekoporã reform scenarios as a share of GDP
Notes: Costing scenarios 1 to 5 do not account for households with indigenous people or families with persons with disabilities. Results reflect only first-round, static impacts of transfers. Behavioural responses, multiplier effects, or broader general equilibrium impacts are not considered. The current leakage rate or share of beneficiaries that are not eligible under Tekoporã is 39%.
Source: OECD based on EPHC (2024[1]) and administrative records.
Extending Hambre Cero to all middle school students would entail a relatively modest fiscal cost, while delivering a meaningful reduction in poverty among households with children
Expanding the Hambre Cero programme to all students enrolled in public basic and middle schools would cost an additional 0.2% of GDP. Hambre Cero costs were projected to reach 0.8% of GDP in 2025. Extending the programme to all enrolled students would increase costs to 1.0% of GDP (Figure 4.6, Panel B). By 2030, the number of school-age children eligible for Hambre Cero is projected to decline slightly, reflecting reductions infertility rates (Figure 4.6, Panel A). While programme costs are projected to increase in nominal terms by 2030 as a result of inflation, expenditures are expected to remain broadly stable as a percentage of GDP under both scenarios considered, at around 0.6% and 0.8% of GDP, respectively.
Number of beneficiaries and costs as a share of GDP for selected expansion scenarios, 2025 and 2030
Note: Two reform scenarios are presented for the years 2025 and 2030. The status quo scenario assumes that Hambre Cero coverage remains constant at 2025 levels for the different school types. The second scenario assumes that Hambre Cero is extended to all publicly enrolled students in media school, while keeping the share of private subsidized school students constant. Enrolment rates are assumed to remain constant. Results reflect only first-round, static impacts of transfers. Behavioural responses, multiplier effects, or broader general equilibrium impacts are not considered. GDP is expressed in current prices.
Source: OECD calculations using administrative records.
Hambre Cero is effective in reducing both poverty and inequality. Even during its rollout phase in 2024, the programme is estimated to have closed around 6% of the national poverty gap and reduced poverty by almost one percentage point (Table 4.5). Simulation results indicate that, once full coverage of students enrolled in public schools is achieved, the programme’s poverty-reducing impact could more than triple, with a modest improvement in spending efficiency. Extending coverage to students in both public and privately subsidised schools would yield similar results. The results highlight that – although Hambre Cero is not an income means-tested programme – its broad reach among low-income households makes it an effective and efficient instrument for reducing poverty among families with children. Further reductions in child poverty would depend on the expansion of Tekoporã or the introduction of a universal family transfer.
Distributive impacts of expansion scenarios on poverty and inequality, 2024
Notes: The status quo estimates refer to 2024 figures, when Hambre Cero was still in a roll-out phase. Enrolment rates are assumed to remain constant. Results reflect only first-round, static impacts of transfers; behavioural responses, multiplier effects, or broader general equilibrium impacts are not considered. The poverty gap estimate (PGE) measures the proportion of the total pre-transfer poverty gap that is filled (or eliminated) by social transfers. It captures the extent to which transfers reduce the depth of poverty among those who were poor before receiving transfers. The poverty reduction efficiency (PRE) measures the share of total transfers that reduces the poverty headcount by lifting individuals above the poverty line. Vertical expenditure efficiency (VEE) measures the proportion of total programme spending that is received by households who were poor before transfers. Spillover measures the share of transfers received by pre-transfer poor households that exceeds the amount required to close their poverty gap (i.e. transfers that raise beneficiaries above the poverty line).
Source: OECD based on EPHC (2024[1]) and administrative records.
Adulto Mayor is expected to remain the primary source of income support for older adults over the coming decades
Expenditure on Adulto Mayor is projected to increase from 0.8% of GDP in 2025 to between 1.0 and 1.2% of GDP by 2028, and to 1.2-1.3% of GDP by 2050 (Figure 4.7, Panel B). The largest increase is expected to occur in the initial years, reflecting the gradual rollout of the programme. Paraguay is expected to complete the gradual expansion of Adulto Mayor to all individuals aged 65 and over by 2028. These projections assume full implementation, with all eligible beneficiaries effectively receiving the transfer. Once full coverage is reached, expenditure growth is expected to slow, driven primarily by population ageing and the associated increase in the number of beneficiaries (Figure 4.7, Panel A). Two scenarios are considered for 2050 with respect to contributory pension coverage among older persons. Under the baseline scenario, the share of individuals aged 65 and over receiving a contributory pension is assumed to rise from 12% in 2025 to 18% by 2050. An alternative, more optimistic scenario assumes that contributory pension coverage reaches 21% by 2050. Under the optimistic scenario the costs of Adulto Mayor in 2050 would only be 0.05 percentage points lower than under the realistic coverage scenario (Figure 4.7, Panel B).
Number of beneficiaries and associated costs as a share of GDP for selected expansion scenarios
Notes: The realistic coverage scenario assumes that contributory pension coverage among people aged 65+ increases from 11.8% in 2025 to 17.8% in 2050. This corresponds to an annual increase of 0.48 percentage points in the share of contributors (rising from 23% in 2025 to 35% in 2050). The vesting rate of contributory pensions is assumed to increase from 43% in 2025 to 50% in 2050. The annual GDP growth rate is assumed at 3.5%. The optimistic coverage scenario assumes that contributory pension coverage increases from 11.1% in 2025 to 20.8% in 2050, with faster growth in the number of contributors (+1.0 pp/year, from 23% in 2025 to 48% in 2050) and a higher vesting rate (60%). The scenarios assume the same severe-to-mild disability ratio, where severe disability defined as having multiple disabilities. No disabled or indigenous individuals are assumed to receive contributory pensions and indigenous population growth is assumed equal to total population growth.
Source: OECD calculations using administrative records.
Adulto Mayor is expected to remain the main source of old-age income in the coming decades. Two scenarios are considered for 2050 with respect to the share of the older population receiving a contributory pension. Under both scenarios, the proportion of older adults relying on Adulto Mayor rather than a contributory pension is projected to decline only marginally, from 88% in 2025 to, at best, 79% by 2050 (Figure 4.7, Panel A). While the share of workers contributing to social security has risen gradually over the past decades, these gains have not yet translated into a higher share of old-age adults that are eligible for a contributory pension. Between 2001 and 2022, the proportion of workers contributing to social security rose from around 12% to 22%, yet the share of adults aged 65 and over receiving a contributory pension has remained broadly stable at approximately 15%. This reflects the significant time lag between increases in contributory participation during working life and higher share of contributory pension recipients in old age. Moreover, the rise in contributory participation does not appear to have been accompanied by improvements in contribution density or vesting rates – that is, the share of months in an individual’s employment history during which contributions are made. As a result, a large share of workers have not accumulated sufficient contribution histories to qualify for a contributory pension. Consequently, Adulto Mayor is expected to remain the main source of income security for the majority of older adults in Paraguay over the coming decades.
A gradual expansion of contributory pensions could reduce the reliance on Adulto Mayor over time. Reforms to the contributory pension system, such as the introduction of partial pensions for workers who do not meet full eligibility requirements (see Chapter 3), could strengthen incentives to contribute and support further formalisation. Allowing contributors to qualify for a partial pension when minimum contribution thresholds are not fully met would increase vesting rates and encourage greater continuity in contribution histories. Over time, such reforms could expand the pool of workers eligible for contributory pensions and gradually reduce dependence on Adulto Mayor. However, the effects would materialise only slowly, as sufficient years of contributions are required before workers become eligible for contributory benefits.
Once Adulto Mayor has reached full coverage of the eligible population, the programme’s poverty-reduction efficiency is expected to gradually decline. In 2024, Adulto Mayor reduced poverty among the old age population by 3.3 percentage points (Table 4.6). Once the programme has been rolled out to the entire eligible population, the programme is projected to reduce poverty by 4.1. percentage points. However, additional beneficiaries are increasingly likely to come from higher-income groups, where exclusion errors are less prevalent and the scope for further reductions in the poverty gap is more limited. As a result, while broader coverage will continue to increase aggregate impacts on poverty and inequality, the marginal poverty‑reducing return of additional spending is expected to diminish. This pattern is reflected in the decline in the poverty reduction efficiency (PRE) indicator reported in Table 4.6.
Distributive impacts expansion scenarios on poverty and inequality for 2024
Notes: The poverty gap estimate (PGE) measures the proportion of the total pre-transfer poverty gap that is filled (or eliminated) by social transfers. It captures the extent to which transfers reduce the depth of poverty among those who were poor before receiving transfers. The poverty reduction efficiency (PRE) measures the share of total transfers that reduces the poverty headcount by lifting individuals above the poverty line. Vertical expenditure efficiency (VEE) measures the proportion of total programme spending that is received by households who were poor before transfers. Spillover measures the share of transfers received by pre-transfer poor households that exceeds the amount required to close their poverty gap (i.e. transfers that raise beneficiaries above the poverty line). Results reflect only first-round, static impacts of transfers. Behavioural responses, multiplier effects, or broader general equilibrium impacts are not considered.
Source: OECD based on EPHC (2024[1]) and administrative records.
Expanding social assistance coverage would significantly reduce poverty at moderate cost
Extending the coverage of the three main social assistance programmes to their full target populations would entail additional costs of around 1.2% of GDP. Extending coverage of the three programmes to everyone that meets the eligibility requirements would almost double total social assistance spending, from 1.5% of GDP to 2.6-2.8% of GDP. Such an expansion would be associated with substantial reductions in poverty (Figure 4.8) and a significant narrowing of the poverty gap (Table 4.7). Tekoporã is the most cost‑effective programme in reducing poverty as it targets poor households. However, the reported cost‑efficiency indicator in Table 4.7 appears particularly low because the scenario assumes no additional leakage, that is, the programme is expanded exclusively to eligible households. Taken together, these results suggest that combining broad coverage through lifecycle‑based categorical programmes with well‑targeted transfers focused on the poorest households could enhance both the inclusiveness and the efficiency of Paraguay’s social protection system.
Costs as a share of GDP and poverty headcount reduction in percentage points
Notes: For Tekoporã, two scenarios are displayed: (i) extending the programme to all eligible households without modifying transfer amounts, and (ii) expanding coverage to all eligible households while increasing the child benefit to PYG 70 000. These scenarios only consider households that are non-indigenous and disability benefits are not simulated. For Hambre Cero, the scenario of expanding the programme to all eligible students in public pre-schools, primary and middle schools is presented. For Adulto Mayor, the scenario of expanding the programme to all eligible population by 2028 is presented. Results reflect only first-round, static impacts of transfers. Behavioural responses, multiplier effects, or broader general equilibrium impacts are not considered.
Source: OECD based on EPHC (2024[1]) and administrative records.
Note: The range of estimates reported for Tekoporã and the total costs reflect a lower-bound estimate assuming no further leakage and an upper-bound estimate, assuming the inclusion error or leakage is maintained at 35% of the beneficiaries. The indicator on the costs per 1% reduction in total poverty gap assumes that there are not additional inclusion errors when extending Tekoporã’s coverage. The estimates do not include additional administrative costs.
Source: OECD based on EPHC (2024[1]) and administrative records.
Tekoporã:
Prioritise the expansion of coverage to all eligible households.
Increase the number of family guides per region, with a minimum of one guide per 150 beneficiary households.
Consider parametric reforms to strengthen the poverty-reducing impact of Tekoporã, including:
Lifting the cap on the number of children per household eligible for benefits.
Increasing the levels of the child and basic family benefit.
Replacing the flat family benefit with a per-person transfer.
Indexing benefits to inflation to preserve their purchasing power.
Adulto Mayor:
Assess the introduction of a pension design that combines partial contributory pensions with the social pension, with social pension benefits being progressively reduced as contributory entitlements increase (see Chapter 3).
Hambre Cero:
Expand coverage to all public middle school students.
Further reductions in child poverty will depend on the expansion of Tekoporã or on the introduction of a universal child benefit, or a benefit targeted to children enrolled in public schools.
References
[2] Beckerman, W. (1975), “The Impact of Income Maintenance Payments on Poverty in Britain”, The Economic Journal, Vol. 89/354, pp. 261-279, https://www.jstor.org/stable/2231601.
[1] EPHC (2024), Encuesta Permanente de Hogares Continua de Paraguay, Instituto Nacional de Estadística (INE) Paraguay, https://www.ine.gov.py/microdatos/Encuesta-Permanente-de-Hogares-Continua.php.
[3] Montt, G., C. Schmidlin and V. Jorquera (2022), “Transferencias no contributivas y su aporte a los procesos de formalizaciön: Experiencias y lecciones del Cono Sur para Paraguay”, ILO technical document, p. 1000, https://www.ilo.org/es/publications/transferencias-no-contributivas-y-su-aporte-los-procesos-de-formalizacion.
[4] MTESS (2024), “Boletín Estadístico de Seguridad Social 2024 incorpora valioso comparativo entre las cajas jubilatorias”, https://www.mtess.gov.py/?p=30115.
This annex presents the methodology and costing assumptions underlying the simulated Tekoporã reform scenarios. Two main data sources are used: the 2024 EPHC and administrative records provided by the MDS. The household survey is used to simulate reforms that involve expanding programme coverage and to estimate their distributive and poverty impacts. Administrative records from the Tekoporã programme are used to cost parametric reforms affecting benefit levels (family allowance, child allowance and per-person transfers), the expansion of family guides, and the extension of coverage to indigenous families.
The EPHC defines a household as individuals residing within the same dwelling which differs from Tekoporã’s definition of family, the latter being closer to a nuclear family concept. It is possible to find more than one beneficiary family that live in the same household. However, based on survey responses, this overlap is limited: only 2.5% of beneficiaries reported living in a household with another beneficiary family. The survey has additional limitations, as it does not identify persons with disabilities, and it excludes indigenous communities.
The EPHC underestimates the number of Tekoporã beneficiaries. The 2024 EPHC identifies 105 119 families as beneficiaries, whereas administrative records report 187 106 beneficiaries (excluding indigenous people, people with disabilities, and areas not covered by the EPHC survey) as of April 2025. Therefore, it is important to acknowledge that the EPHC does not provide representative coverage of the programme’s full target population. To partially address this limitation of survey data, two beneficiaries per household were identified, when the head of household had grandchildren and when there were two women – one adult and another adult aged at least 14 years or older. Using this criterion, 16 348 households in the survey data were reclassified as containing two beneficiary families, increasing the total number of Tekoporã families identified through the EPHC to 119 157. The remaining underestimation of beneficiaries using the survey data appears to be primarily due to the exclusion of indigenous families residing in communities not surveyed and, to a lesser extent, families with disabilities who applied independently for benefits but do not live in the geographically prioritized areas of high poverty incidence. For these groups, the costing of achieving full coverage will be carried out separately using data from the Census and other surveys. The EPHC is only used to estimate the cost of extending coverage to poor households, while changes in benefit amounts are costed using administrative data.
The poverty reduction impact and costs of Tekoporã were estimated using the survey data. The EPHC reports households that receive Tekoporã but also indicates the amount of benefits received. To estimate the poverty reduction effect of Tekoporã, the current benefit amounts received were subtracted from household income to construct a counterfactual scenario. The overall programme costs were calculated as the sum of total benefits received by households.
To assess the distributive and fiscal impact of expanding Tekoporã coverage, all households classified as either monetarily poor or multidimensionally poor were identified as eligible to receive transfers. Benefits were allocated to eligible households based on household composition: family allowances, child allowances, and old-age allowances (for individuals not covered by contributory pensions). In cases where household composition suggested the presence of two or more families within one household, the family allowance was applied to both, and the cap on child allowances was multiplied accordingly.
The reform scenarios considered for Tekoporã are the following:
Extending the programme to all eligible households using the current benefit amounts.
Extending the programme to all eligible households, increasing the amount per child to PYG 70 000.
Increasing the basic family allowance to PYG 200 000 for families currently receiving benefits.
Increasing the additional amount per child to PYG 70 000 for families currently receiving benefits.
Removing the limit on the number of children per household for families currently receiving benefits.
Modifying the basic family allowance to PYG 50 000 per person.
Analysing the cost of assigning a family guide to all families currently receiving benefits.
Extending Tekoporã to all eligible indigenous families.
Scenarios (1 and 2) on coverage extension were simulated using the household survey (Table 4.A.1). In the second scenario, the child allowance was increased from 50 000 to 70 000. In both scenarios, households that were included in the EPHC as beneficiaries – even if not identified as eligible by the survey – were retained in the costing and distributive impact analysis of the reform scenarios.
Costs in billion PYG or as a share of GDP and reductions in poverty and inequality
Note: The survey data do not include indigenous communities.
Source: OECD calculations based on EPHC 2024.
The estimates of scenario 3 on Tekoporã expenditures for 2025 are derived from administrative records identifying the number of beneficiary families by department, excluding households with indigenous members or persons with disabilities. For each department, the cost per family was calculated under two benefit levels: the current transfer of PYG 112 500 and a higher family allowance of PYG 200 000.
The costing of scenarios 4 and 5 relies on administrative records of current Tekoporã beneficiary families, excluding indigenous communities and households with members with disabilities. These records provide the number of families and their distribution by number of children, allowing for precise estimation of child allowance expenditures. For reform 4, costs were recalculated by increasing the child allowance from PYG 50 000 to PYG 70 000 per child while keeping the current beneficiary base unchanged. For reform 5, the existing cap of four children per family was removed, and costs were recalculated by applying the PYG 50 000 allowance to all children.
The costing of scenario 6 is based on administrative records of current Tekoporã beneficiary families, excluding indigenous communities and households with members with disabilities. Under the status quo situation, the family allowance benefit of PYG 112 500 per family translates into an average benefit of PYG 38 769 per household member, given the average beneficiary family size of 2.9 persons. For the reform scenario, this amount was adjusted to a uniform transfer of PYG 50 000 per person across all family sizes, applied to the total number of individual beneficiaries recorded in the administrative data. This approach captures only the budgetary effect of increasing the per capita family allowance for existing beneficiary families, without considering the expansion of coverage.
The cost of expanding family guides (guías familiares, scenario 7) was estimated by comparing current coverage levels of family guides with the number of family guides required to reach all participating households. Each family guide is assumed to effectively accompany at maximum 150 families, depending on the number of beneficiaries in each department. The difference between the number of currently assigned guides and the number required under this assumption yields the number of additional family guides needed in each department. The total additional cost was then calculated by multiplying the number of additional family guides by the average annual remuneration per guide. The resulting additional cost amounts to approximately PYG 30.8 billion, equivalent to 0.009% of GDP.
The costing of expanding Tekoporã to all eligible indigenous families (scenario 8) combines administrative data on current programme expenditures for 2025 with population information from INE’s Censo Nacional de Población y Viviendas 2022. Since the census data precedes the administrative records, the number of indigenous families in 2022 was projected forward to 2025 by applying the average annual population growth rate observed between 2020 and 2024. Poverty rates by department from the census were then applied to estimate the number of indigenous families eligible for the programme. These figures were compared with administrative records on current departmental expenditures to simulate the costs of extending coverage to all eligible indigenous households, using the current benefit amounts (Table 4.A.2).
Costs in billion PYG
Source: OECD calculations based on administrative records.
The projections estimate the future fiscal cost of Paraguay’s Adulto Mayor social pension as coverage expands to all adults aged 65 and over and demographic ageing accelerates. The costing scenario assumes that the programme’s rollout will be fully completed by 2028 and that all eligible older adults will receive the transfer without implementation gaps. The costs are presented in constant prices and as a share of GDP, which is projected to grow annually by 3.5% in real terms. Sensitivity tests were also conducted assuming alternative GDP growth rates of 3% and 4%, showing only modest variation in results.
The projected number of Adulto Mayor beneficiaries is driven by demographic change and the gradual expansion of contributory pension coverage. Population projections are derived from the most recent national census data and reflect Paraguay’s expected demographic transition through 2050. The model first estimates the share of workers contributing to social security, which is assumed to increase by 0.48 percentage points per year, in line with historical trends observed between 2001 and 2022. The vesting rate – the proportion of contributors who ultimately qualify for a pension – is assumed to be 43% and is taken from ILO estimates (Montt, Schmidlin and Jorquera, 2022[3]). The vesting rate is projected to increase to 50% by 2025 and remain constant thereafter. These two parameters are used to project the share of the population aged 65 and over that will receive a contributory pension, applying a 20-year lag between contribution and retirement to capture the delay between labour market participation and eligibility for pension benefits. The number of pensioners with contributory pensions is calculated using the yearly publication from the social security office (MTESS, 2024[4]), which calculates the share of pension beneficiaries above the age of 65 for each pension fund. The number of pensioners is then divided by the total population in 2023 to calculate the coverage rate of contributory pensions.
The share of older adults without contributory pensions determines the potential beneficiary population of Adulto Mayor. Multiplying this share by the total population aged 65 and over yields the number of beneficiaries in each year. The benefit amount is held constant in real terms and projected forward using inflation assumptions to preserve purchasing power.
Additional assumptions are made about the evolution of disabled and indigenous people. Individuals with severe disabilities between the ages of 60-64 are assumed not to receive contributory pensions and hence to benefit from Adulto Mayor. The ratio of severely to mildly disabled individuals is assumed to remain constant over time, with severe disability defined as having more than one functional limitation. The indigenous population is projected to grow at the same rate as the total population. All indigenous people aged 55–64 are assumed to receive a social pension.
The analysis on Hambre Cero projects the coverage and cost of the programme until 2030 under both the status quo and expansion scenarios. The analysis combines data on demographic projections, school enrolment and administrative records of the programme to estimate the number of beneficiaries and associated costs. The number of potential beneficiaries for Hambre Cero is calculated by education level. Children, who attend public or subsidised private schools, are grouped into the following school types: Inicial (0–5 years), Básica (6–14 years), and Media (15–17 years). Demographic projections are used to estimate the total number of children in each age cohort, while enrolment rates are applied to determine the share of children attending targeted schools. The programme coverage is measured by comparing the number of authorised meals to the number of enrolled students at each school level and by department. In the expansion scenarios, these coverage rates are adjusted. Furthermore, unit costs are calculated as the cost per school day at each level, multiplied by 180 school days per year. These costs are applied to the projected number of beneficiaries (Table 4.C.1). Costs are then aggregated across departments and education levels to generate total costs at the national level, which are expressed both in absolute terms and as a share of GDP (Table 4.C.2).
Beneficiaries of Hambre Cero by school type, 2025-2030
Note: Enrolment rates and coverage rates are assumed to remain constant for the scenario of expanding Hambre Cero to all Media schools.
Source: OECD calculations using administrative records.
Costs in billion PGY by school type, 2025-2030
Note: Enrolment rates and coverage rates are assumed to remain constant for the scenario of expanding Hambre Cero to all Media schools.
Source: OECD calculations using administrative records.
← 1. The minimum wage is PYG 2 899 048 per month as of 1st July 2025.
← 2. Some religiously affiliated subsidised schools – previously covered under the Programa de Alimentación Escolar (PAE) – are also included under Hambre Cero, though not all receive financing. These schools are often located in regions such as El Chaco, where public school provision is limited and where implementation of the earlier PAE programme was inconsistent.
← 3. In March 2026, the National Institute of Statistics revised the household survey expansion factors for 2022–2025. The analysis in this report is based on the 2024 EPHC microdata released in 2025.
← 4. Estimating the total number of Tekoporã beneficiary households using the EPHC faces several limitations that are discussed in Annex 4.A. However, the EPHC is the primary data source for measuring poverty. This study assumes that the estimated share of poor households receiving Tekoporã is accurate.