## wpiea2019149

## Source details

**Canonical URL:** [wpiea2019149](https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019149.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2019/wpiea2019149.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2019/wpiea2019149.pdf.json)

---

### Key findings on pension coverage and adequacy
- Only around 30 percent of the economically active population are contributing to statutory pension schemes; less than 10 percent of workers in the bottom income-quintile do.
- Affiliates in Peru on average contribute for 4-5 months out of a possible 12 months over their working life.
- Baseline empirical replacement rate projections (averaged across all affiliates and across the next 30 years):
  - SPP (private defined contribution system): 25-33 percent.
  - SNP (public defined benefit system): 25-31 percent.
- SNP projected trajectory (averages across cohorts):
  - Roughly 45 percent in 2020 declining to 12-20 percent in 2047; steepness depends on future adjustments to minimum and maximum pensions.
- SPP projected trajectory (averages across cohorts):
  - Around 25-28 in 2020 changing to 23-34 percent in 2047, with exact values depending on key assumptions.
- Distributional outcomes:
  - In the SNP, affiliates who reach the 20-year minimum contribution period have replacement rates of roughly 60 percent.
  - About 60 percent of all affiliates, and roughly 50 percent of affiliates who contributed over the past five years, are likely not to reach the SNP 20-year threshold and therefore have a replacement rate of zero.
  - Affiliates from the bottom income quintile are three times less likely to reach the 20-year threshold than affiliates from the top quintile, implying highly regressive redistribution.
  - In the SPP, affiliates with full contribution careers (generally high-income workers) are expected to have replacement rates around 40 percent even under conservative scenarios; low-income workers with below-average contribution densities will have replacement rates substantially below 20 percent.

### System structure and key parameters
- Contributory system components:
  - SPP: private fully-funded plan (defined contribution).
  - SNP: public defined benefit plan (largely pay-as-you-go, partly funded by an investment fund since 1996).
  - Special regimes: cedula viva (closed for new entrants in 2004) and military/police pension scheme.
- Affiliation rules:
  - Affiliation with a contributory pillar is obligatory for formal, salaried workers and voluntary for independent workers.
  - Workers choose between SNP and SPP upon entering the workforce; if no choice is made they are assigned to the SPP. Workers can move from SNP to SPP later but cannot move back.
- Market structure (SPP): four pension funds with market shares of 39, 31, 25 and 5 percent (assets under management).
- Contribution and retention rates:
  - SNP contributions: 13 percent of gross earnings.
  - SPP contributions towards old-age retirement savings: 10 percent of gross earnings.
  - Including fees (used to be around 1.5 percent but falling) and disability and life insurance (1.36 percent), total salary retention in the SPP is close to 13 percent.
- Retirement age and SNP rules:
  - Legal retirement age: 65 in both pillars.
  - SNP minimum pension: S/. 415 per payment.
  - SNP maximum pension: S/. 857.36.
  - Current legal minimum wage: S/ 930.
  - SNP vesting period (minimum contribution time) is 20 years.
- Non-contributory support:
  - Pension 65: social transfer to elderly poor; pays S/ 125 per month to over 500,000 recipients (circa 25 percent of all people over 65).
  - Pension 65 transfers currently cost slightly over 0.1 percent of GDP and the amount has not been adjusted since program creation; no explicit indexation mechanism.

### Key numbers and demographics (Peru, 2017)
- Population: 31.8 million
- Population 15-64: 20.9 million
- Population 65 and over: 2.3 million
- Labor Force (2016): 16.9 million
- Employment (2016): 16.2 million
- SNP Affiliates: 4.5 million
- SPP Affiliates: 6.6 million
- SPP Assets under Management: 47.7 billion USD (ca. 22 percent of GDP)
- SNP Expenditures: 1% of GDP
- SNP Contributions: 0.5% of GDP
- SNP State Budget Transfers: 0.2% of GDP
- SNP Other Revenues: 0.3% of GDP
- SNP Reserve Fund (FCR): 2.7% of GDP
- GDP per capita (USD, current): 6,762
- Poverty rate: 21.7%

### Fiscal sustainability: historical context and projections
- Financing needed from the general budget to cover pension deficit:
  - Around 3 percent of GDP in 2000
  - 1.2 percent of GDP in 2012
  - Around 0.2 percent of GDP in 2017 (SNP deficit)
- Multiple studies project SNP deficits by 2050 vary roughly between 0.1 and 2 percent of GDP.
- Long-term projected average fiscal cost referenced: 0.2 percent of GDP per year over the long-term.

### Contribution densities: definitions and empirical findings
- Two measures:
  - CEM (contributions at the extensive margin): whether an affiliate contributed in a given time period.
  - CIM (contributions at the intensive margin): frequency of contributions = total months accrued / maximum possible months to be accrued. Simulations use CIM.
- Data samples:
  - SNP: representative sample of 65,382 affiliates (coverage 2012-16, ~1.5 percent of overall affiliates).
  - SPP: representative sample of 2 percent of affiliates (coverage 2007-17).
- Mean contribution density:
  - SNP: 34 percent (roughly 4 months out of 12 per year).
  - SPP: about 40 percent (4.8 months).
- Contribution densities increase with income:
  - SNP: over 75 percent of bottom income affiliates have densities below 50 percent; virtually nobody contributes every month. About 25 percent of the top income quintile have full contribution lengths.
  - SPP: average contribution density in top quintile roughly double that of bottom quintile.
- Vesting and eligibility:
  - SNP 20-year vesting period implies most SNP affiliates will not be eligible for a pension.
  - Even among affiliates who contributed during 2012-16, close to 50 percent will not reach the 20-year minimum contribution period.
  - Probability of reaching the 20-year threshold in SNP (age group 30-59 in 2016) by earnings quintile:
    - 1st: 17%
    - 2nd: 51%
    - 3rd: 55%
    - 4th: 54%
    - 5th: 60%
- Trends:
  - Contribution densities rose over the past 5 years by around 10 percent (about 4 percentage points), especially among younger cohorts, but remain below regional averages.

### Methodology and projection assumptions (overview)
- Replacement rates simulated for 2018-2047 separately for SNP and SPP using representative contribution-history samples.
- Replacement rate definition: average monthly pension in first year of retirement divided by final monthly earnings prior to retirement.
- Cohort focus: cohorts aged 35 and older in base year 2017.
- Baseline macro-demographic and parameter assumptions (Table 2 highlights):
  - Inflation: 2 percent.
  - Real wage growth: 2 percent.
  - Real gross rate of return: Baseline: 4.2 percent; robustness scenario: 5.2 percent.
  - Mortality: SPP-S-2017 mortality tables.
  - Retirement age: baseline 65.
  - Contribution density: baseline unchanged relative to 2012-16 average; alternative scenario continues linear growth observed 2012-16.
- SPP-specific technical assumptions:
  - Contribution rate in base scenario: constant at 10 percent after fees.
  - Asset-based fees: 1.13 percent for mixed-fee affiliates; 0 percent for flow-fee affiliates (assumed constant).
  - Account evolution equation: AC_t,i = AC_{t-1,i} * (1 + rr_t - f_t) + C_{t,i}.
  - CRU calculation for starting pension uses discounting with i = (1 + rr) / (1 + p_index) - 1 and survival probabilities to maximum age M (109).
- SNP-specific assumptions:
  - DB accrual rate × average wages over last 3–5 years; accrual rate schedule differs by cohort; capped at 100 percent.
  - Individual contribution years observed for 2012-16; years prior to 2012 and after 2016 assumed same average individual contribution density as in 2012-16.
  - Minimum and maximum pension adjustment scenarios include CPI indexation vs no adjustment.

### Baseline and robustness results: replacement rates
- Baseline projected replacement rates: lie between 20 and 30 percent over the coming decades.
- Mean replacement rates over the full forecast horizon 2020-47:
  - SNP: projected to be 25-31 percent.
  - SPP: projected to be 25-33 percent.
- SNP dynamics:
  - Replacement rates decline from around 45 percent in 2020 to 12-20 percent by 2045; decline driven by past reforms and by how minimum/max pensions are adjusted.
  - About 60 percent of affiliates projected to get zero replacement rate due to 20-year vesting.
- SPP dynamics:
  - Time profile sensitive to real gross rate of return and contribution densities.
  - Increasing real gross rate of return from 4.2 percent to 5.2 percent raises replacement rates in 2047 by about 7 percentage points (robustness scenario).
  - Increasing contribution densities in line with past trends raises replacement rates in 2047 by around 3 percentage points.
- Sensitivity to returns and fees:
  - An increase in the real rate of return of 0.5 percentage points relative to the baseline increases replacement rates by around 3 percentage points.
  - Reducing asset-based pension fund fees by 0.5 percentage points increases replacement rates by around 1 percentage point (given current mixed regime); once all affiliates are under asset-based fees, the same fee reduction increases replacement rates by 1.5 percentage points.
  - Note: net real rate of return = gross real rate of return adjusted for fees; reductions in asset-based fees do not have the same effect as increasing gross real returns because many affiliates still operate under contribution-based fees and no asset-based fees are charged during retirement.

### Lump-sum withdrawals and risks to adequacy (SPP)
- Withdrawals allowed since 2016:
  - Up to 95.5 percent of the individual account value at retirement (4.5 percent to health insurance).
  - Up to 25 percent prior to retirement for mortgage down payments or to pay down mortgage debt.
- Take-up and magnitude:
  - As of end-January 2018, close to 60,000 affiliates had withdrawn a total of 2bn soles, or 1.2 percent of total SPP assets, for mortgage related withdrawals.
  - 89 percent of the total value of these withdrawals was used for mortgage amortization; the remainder for down-payments.
- Simulated impacts:
  - A 25 percent housing withdrawal reduces replacement rates by around 18 percent for a 55-year-old, around 14 percent for a 45-year-old, and around 8 percent for a 35-year-old.
  - A 95.5 percent withdrawal at retirement transfers full longevity and investment risk to the affiliate and can deplete resources before end of life under low returns (example: 20-year life expectancy, 1 percent real return -> depleted after 14 years).

### Fee structure, reforms and impact on adequacy (SPP)
- Pre-2013: flow-based fees; total contributions ~13 percent of gross earnings with ~1.5 percent to pension fund fees.
- Since 2013: asset-based fees for new entrants; total contributions effectively around 11.5 percent to accounts net of fees.
- Transition mechanism: mixed commissions for pre-2013 contributors with weight on stock increasing over 10 years.
- Empirical fee context:
  - As of 2016, 33 percent of affiliates and 52 percent of contributors still under flow-based commissions.
  - Flow commissions in 2017 between 1.47 and 1.69 percent of gross earnings depending on AFP.
  - Mixed system: 0.18-1.07 percent on the flow and 1.2-1.25 percent on the stock (average ranges).
  - December 2018 tender: mixed-fee members flow component charged at 0 percent; stock component 0.82 percent.
- Effects of reductions:
  - Reducing asset-based fees by 0.5 percentage points increases replacement rates by around 1 percentage point given current mixed regime; 1.5 percentage points once all under asset-based fees.

### Policy proposals, reform scenarios and fiscal implications
- SNP reform:
  - Shortening minimum contribution period from 20 to 15 years:
    - Roughly 5 percent more SNP affiliates become eligible for a pension (assumed to get 15/20 of the current minimum pension).
    - Increases replacement rates for overall affiliates by about 2 percentage points.
    - Estimated gross fiscal cost close to 0.05 percent of GDP per year in the medium-term.
  - Alternative design: de-link contribution period from vesting period required to qualify for minimum pension; allow lump-sum access for short-contributors.
- SPP parametric options:
  - Broadening contribution base to make pension contributions obligatory for the 13th and 14th monthly salaries would increase replacement rates by close to 3 percentage points.
  - Increasing contribution rates by 5 percentage points has the largest impact on replacement rates (example: increasing SPP contribution rate from 10 percent to 15 percent would raise replacement rates by 7 percentage points over the long-term), but may adversely affect labor formalization.
  - Increasing retirement age from 65 to 66 in five years and to 67 by 2026 would increase replacement rates by about 3 percentage points at the end of the projection period.
- Design recommendations:
  - Government-supported minimum pensions, matching contributions schemes, or a fully integrated non-contributory pillar are either missing or could be strengthened.
  - Large withdrawals from individual pension savings accounts should not be allowed or should be strongly limited to protect retirement adequacy.
  - Communicate expected replacement rates under several scenarios publicly and transparently to avoid false expectations and implicit fiscal claims.
  - Non-contributory pillar should be strengthened given large uncovered population.
  - Consider larger reform to restructure system and avoid competition between private (SPP) and public (SNP) plans to reduce fiscal and distributional uncertainty.

### Distributional and subgroup findings
- Income:
  - SPP top income quintile estimated replacement rate 29 percent; bottom quintile 22 percent.
  - SNP: strongly regressive redistribution due to 20-year vesting—low-income groups disproportionately fail to reach threshold.
- Gender:
  - Women projected to have slightly lower replacement rate than men in the SPP (about 1 percentage point) due to higher life expectancy lowering annuities.
  - In the SNP, projected average replacement rates are higher for women than men due to somewhat higher female contribution densities (SNP benefit formula does not account for gender mortality differences).
- Eligibility and coverage:
  - Only 20 percent of the working-age population contribute to their pension plan in any given month.
  - Close to 70 percent of economically active 20-64-year-olds did not contribute at all during 2015 or 2016.
  - Urban/rural: rural contribution rate 7 percent vs urban 37 percent.
  - Bottom income quintile contribution rate: 4 percent; top quintile: 61 percent.
  - By education: virtually 0 percent for least educated groups to close to 100 percent for the highest education group.

### Comparative context and international options
- Regional empirical replacement rates:
  - Mexico: recent IMF projections conditional on receiving a pension average replacement rates between 40 and 47 percent, but coverage gaps mean many workers will not receive pensions.
  - Chile: observed replacement rates around 40 percent including solidarity pillar, significantly lower excluding it.
- Non-contributory pillar options and fiscal consequences:
  - Strengthened non-contributory pillar justified given low contributory coverage; Pension 65 has documented positive impacts despite low nominal value.
  - Indexing eligibility age to life expectancy implies by 2050 eligibility could be over 69; keeping eligibility at 65 would add 0.1 percent of GDP annual costs by 2050.
  - Raising effective pension coverage to near 100 percent via non-contributory pillar would cost roughly 0.2 percent of GDP per year in a referenced scenario.
- Options to increase contributory coverage:
  - Matching contributions: can incentivize participation; Peru has unimplemented legislation for matching for firms with fewer than 10 employees; field experiments (50 percent and 100 percent matching) are ongoing.
  - Consumption-based contributions: proposal to complement income-based contributions with a sales-tax-like contribution to pension accounts; feasibility constrained by low rural saving rates and likely limited initial impact.

### Research gaps and recommended further work
- Evaluate cross-country best practices for private pension funds to inform Peru-specific fee and market-structure recommendations.
- More evidence on behavioral responses to matching contributions in high-informality contexts needed before scaling.
- Further simulation of reform permutations that de-link vesting and contributory thresholds and that integrate non-contributory and contributory pillars.

*Source: wpiea2019149 (IMF working paper excerpts provided).*

### References _______________________________________________________________37

### wpiea2019149 - References _______________________________________________________________37

### Key findings on pension coverage and adequacy
- Only around 30 percent of the economically active population are contributing to statutory pension schemes; less than 10 percent of workers in the bottom income-quintile do.
- Affiliates in Peru on average contribute for 4-5 months out of a possible 12 months over their working life.
- Baseline empirical replacement rate projections (averaged across all affiliates and across the next 30 years):
  - SPP (private defined contribution system): 25-33 percent.
  - SNP (public defined benefit system): 25-31 percent.
- SNP projected trajectory (averages across cohorts):
  - Roughly 45 percent in 2020 declining to 12-20 percent in 2047; the steepness depends on future adjustments to minimum and maximum pensions.
- SPP projected trajectory (averages across cohorts):
  - Around 25-28 in 2020 changing to 23-34 percent in 2047, with exact values depending on key assumptions.
- Distributional outcomes:
  - In the SNP, affiliates who reach the 20-year minimum contribution period have replacement rates of roughly 60 percent.
  - About 60 percent of all affiliates, and roughly 50 percent of affiliates who contributed over the past five years, are likely not to reach the SNP 20-year threshold and therefore have a replacement rate of zero.
  - Affiliates from the bottom income quintile are three times less likely to reach the 20-year threshold than affiliates from the top quintile, implying highly regressive redistribution.
  - In the SPP, affiliates with full contribution careers (generally high-income workers) are expected to have replacement rates around 40 percent even under conservative scenarios; low-income workers with below-average contribution densities will have replacement rates substantially below 20 percent.

### System structure and key parameters
- Contributory system components:
  - SPP: private fully-funded plan (defined contribution).
  - SNP: public defined benefit plan (largely pay-as-you-go, partly funded by an investment fund since 1996).
  - Special regimes: cedula viva (closed for new entrants in 2004) and military/police pension scheme.
- Affiliation rules:
  - Affiliation with a contributory pillar is obligatory for formal, salaried workers and voluntary for independent workers.
  - Workers choose between SNP and SPP upon entering the workforce; if no choice is made they are assigned to the SPP. Workers can move from SNP to SPP later but cannot move back.
- Market structure (SPP): four pension funds with market shares of 39, 31, 25 and 5 percent (assets under management).
- Contribution and retention rates:
  - SNP contributions: 13 percent of gross earnings.
  - SPP contributions towards old-age retirement savings: 10 percent of gross earnings.
  - Including fees (used to be around 1.5 percent but falling) and disability and life insurance (1.36 percent), total salary retention in the SPP is close to 13 percent.
- Retirement age and SNP rules:
  - Legal retirement age: 65 in both pillars.
  - SNP minimum pension: S/. 415 per payment.
  - SNP maximum pension: S/. 857.36.
  - Current legal minimum wage: S/ 930.
  - SNP vesting period (minimum contribution time) is 20 years.
- Non-contributory support:
  - Pension 65: social transfer to elderly poor; pays S/ 125 per month to over 500,000 recipients (circa 25 percent of all people over 65).
  - Pension 65 transfers currently cost slightly over 0.1 percent of GDP and the amount has not been adjusted since program creation; no explicit indexation mechanism.

### Projected drivers and sensitivities
- Lump-sum withdrawals:
  - Since 2016, lump-sum withdrawals of 95.5 percent of the individual account value are allowed at retirement.
  - Withdrawals of up to 25 percent are allowed prior to retirement for mortgage down payments and to pay down mortgage debt.
  - Large withdrawals, especially 95.5 percent at retirement, pose significant downward risks to replacement rate projections by transferring longevity and capital market risks to affiliates.
- Sensitivity to returns and fees:
  - An increase in the real rate of return of 0.5 percentage points relative to the baseline increases replacement rates by around 3 percentage points.
  - Reducing asset-based pension fund fees by 0.5 percentage points increases replacement rates by around 1 percentage point.
  - Note: the net real rate of return is the gross real rate of return adjusted for fees; reductions in asset-based fees do not have the same effect as increasing gross real returns because a significant fraction of affiliates still operate under contribution-based fees and no asset-based fees are charged during retirement.
- Contribution densities:
  - Increasing contribution densities substantially would have a significant impact on replacement rates; this effectively implies reducing labor informality.

### Policy proposals and fiscal implications
- SNP reform:
  - Shortening the minimum contribution period from 20 to 15 years would allow more people to obtain a pension and raise the average replacement rate at a relatively limited gross fiscal cost of 0.05 percent of GDP.
- SPP parametric options:
  - Broadening the contribution base to make pension contributions obligatory for the 13th and 14th monthly salaries would increase replacement rates by close to 3 percentage points.
  - Increasing contribution rates by 5 percentage points has the largest impact on replacement rates and may be unavoidable in the medium term to guarantee adequate pensions, but could adversely affect labor formalization.
- Design recommendations:
  - Elements common in other Latin American defined contribution systems—government-supported minimum pensions, matching contributions schemes, or a fully integrated non-contributory pillar—are either missing or could be strengthened in Peru.
  - Large withdrawals from individual pension savings accounts should not be allowed or should be strongly limited to protect retirement adequacy.

### Comparative context and literature
- Empirical replacement rate findings in the region:
  - Recent IMF projections for Mexico find, conditional on receiving a pension, average replacement rates between 40 and 47 percent, but about 64 percent of private sector workers and 27 percent of public sector workers will not contribute long enough to receive a pension.
  - Literature reports empirical replacement rates somewhat below 40 percent in Chile and Mexico; observed replacement rates for recent pensioners in Chile are 40 percent when including the solidarity pillar and significantly lower when excluding it.
- Existing Peruvian studies:
  - Most prior studies compute theoretical replacement rates for representative agents; only a few recent studies simulate empirical replacement rates for Peru (Alonso et al. 2014; Peruvian Social Protection Commission 2017).
- Broader inference:
  - Even with parametric reforms, replacement rates in Peru are projected to remain significantly below levels occurring in advanced countries.
  - Labor market reforms that reduce informality are critical to raise contribution densities and improve pension adequacy.

*Source: wpiea2019149 - References _______________________________________________________________37*

### 0.2 percent of GDP per year over the long-term.

### wpiea2019149 - 0.2 percent of GDP per year over the long-term.

### Key numbers and demographics (Peru, 2017)
- Population: 31.8 million
- Population 15-64: 20.9 million
- Population 65 and over: 2.3 million
- Labor Force (2016): 16.9 million
- Employment (2016): 16.2 million
- SNP Affiliates: 4.5 million
- SPP Affiliates: 6.6 million
- SPP Assets under Management: 47.7 billion USD (ca. 22 percent of GDP)
- SNP Expenditures: 1% of GDP
- SNP Contributions: 0.5% of GDP
- SNP State Budget Transfers: 0.2% of GDP
- SNP Other Revenues: 0.3% of GDP
- SNP Reserve Fund (FCR): 2.7% of GDP
- GDP per capita (USD, current): 6,762
- Poverty rate: 21.7%

### Fiscal sustainability: historical context and projections
- Reforms made SNP and special regimes less generous, improving fiscal position.
- Financing needed from the general budget to cover pension deficit:
  - Around 3 percent of GDP in 2000
  - 1.2 percent of GDP in 2012
  - Around 0.2 percent of GDP in 2017 (SNP deficit)
- Multiple studies project the fiscal cost will increase as new affiliates fail to keep pace with retirees; projected SNP deficits by 2050 vary roughly between 0.1 and 2 percent of GDP (estimates cited from Alonso et al. (2014), Bernal (2016), IMF (2018a), Social Protection Commission (2017)).
- Long-term projected average fiscal cost referenced: 0.2 percent of GDP per year over the long-term.

### Pension coverage: current status and inequalities
- Affiliates (early 2018): SNP 4.5 million, SPP 6.6 million (over 45 percent of working-age population).
- Only 20 percent of the working-age population contribute to their pension plan in any given month.
- As a percentage of economically active 20-64-year-olds, close to 70 percent did not contribute at all to a contributory pension plan during 2015 or 2016.
- Urban/rural and income/education disparities:
  - Contribution rates: rural workers 7 percent vs urban 37 percent.
  - Bottom income quintile contribution rate: 4 percent; top income quintile: 61 percent.
  - By education: virtually 0 percent for least educated groups to close to 100 percent for the highest education group.
- Regional comparison: Peru’s pension coverage ranks below the regional average and lags peers such as Colombia; relative to GDP per capita Peru is one of two regional underperformers (with Mexico).

### Pension adequacy: replacement rates for current retirees
- SNP current retiree replacement rates: around 40 percent (Cruz-Saco and others, 2014), aligning with short-term forecast.
- SPP current retiree replacement rates (administrative data since 2006):
  - Mean replacement rate: 37 percent
  - Median replacement rate: 25 percent
  - Mean and median contribution density of current retirees: 0.38
- Mean replacement rates are higher than median due to a tail of very high replacement rates (>1), associated with early retirement and early retirement due to unemployment and possible income-reporting effects.
- Since 2012, mean replacement rates with and without outliers averaged 32 percent and 28 percent, respectively—close to the short-term forecast.

### Contribution densities: definitions and empirical findings
- Two measures:
  - CEM (contributions at the extensive margin): whether an affiliate contributed in a given time period.
  - CIM (contributions at the intensive margin): frequency of contributions = total months accrued / maximum possible months to be accrued. Simulations use CIM.
- Data samples:
  - SNP: representative sample of 65,382 affiliates (coverage 2012-16, ~1.5 percent of overall affiliates).
  - SPP: representative sample of 2 percent of affiliates (coverage 2007-17).
- Mean contribution density:
  - SNP: 34 percent (roughly 4 months out of 12 per year).
  - SPP: about 40 percent (4.8 months).
- Contribution densities increase with income:
  - SNP: over 75 percent of bottom income affiliates have densities below 50 percent; virtually nobody contributes every month. About 25 percent of the top income quintile have full contribution lengths.
  - SPP: average contribution density in top quintile roughly double that of bottom quintile.
- Vesting and eligibility:
  - SNP 20-year vesting period implies most SNP affiliates will not be eligible for a pension.
  - Even among affiliates who contributed during 2012-16, close to 50 percent will not reach the 20-year minimum contribution period.
  - Probability of reaching the 20-year threshold in SNP (age group 30-59 in 2016) by earnings quintile:
    - 1st: 17%
    - 2nd: 51%
    - 3rd: 55%
    - 4th: 54%
    - 5th: 60%
- Trends:
  - Contribution densities rose over the past 5 years by around 10 percent (about 4 percentage points), especially among younger cohorts, but remain below regional averages (using 2010 Duran and Pena baseline).

### Methodology and projection assumptions (overview)
- Replacement rates simulated for 2018-2047 separately for SNP and SPP using representative contribution-history samples.
- Replacement rate definition: average monthly pension in first year of retirement divided by final monthly earnings prior to retirement.
- Cohort focus: cohorts aged 35 and older in base year 2017.
- Contribution density used in simulations: CIM measured over observation periods; for affiliates entering during observation, denominator adjusted to affiliation start date.
- SNP-specific assumptions:
  - SNP is a defined benefit (DB) system; starting pension at age 65 = accrual rate × average wages over last 3–5 years.
  - Accrual rate schedule differs by cohort; capped at 100 percent.
  - Individual contribution years observed for 2012-16; for years prior to 2012 and after 2016 it is assumed affiliates show same average individual contribution density as in 2012-16 (flat age profile for age groups 30 and older, with densities rising between ages 20 and 30).
  - Future individual earnings: affiliates aged 35 and older in 2016 remain in the relative earnings position observed in 2012-16; real wage growth and inflation assumed 2 percent per year.
  - Minimum and maximum pension adjustments matter; baseline assumes either CPI indexation or no adjustment; robustness exercise includes wage indexation.
  - Pensions paid 14 times per year in SNP; individuals with less than 20 contribution years have replacement rates of zero.
  - SNP pensions estimated only for affiliates with earnings records in 2012-16; about 22 percent of scheme members inactive in those years are neglected (may overestimate future pension levels).
- SPP-specific assumptions:
  - SPP is defined contribution (DC); observable entitlements until September 2017 include DC accounts + Bonos de Reconocimiento (indexed with CPI until retirement).
  - Pension entitlements accrued after 2017 estimated using simulations; affiliates assumed to show same mean individual contribution density as observed in 2012-17.
  - Affiliates aged 35 and older assumed to remain in the relative earnings position observed in 2010-2017.
  - Contribution rate in base scenario remains constant at 10 percent after fees; simulations exclude voluntary contributions.
  - Asset-based fees: 1.13 percent for mixed-fee affiliates; 0 percent for flow-fee affiliates. Assuming constant asset-based fees is conservative given recent downward trend.
  - Account evolution equation (per individual i):
    - AC_t,i = AC_{t-1,i} * (1 + rr_t - f_t) + C_{t,i}
  - Starting pension estimation uses CRU (Capital Requerido Unitario) calculation accounting for survival probabilities to age M (maximum age 109) and discounting with rate i that reflects nominal rate of return and pension indexation (p_index = assumed inflation = 2 percent). Equation highlights:
    - CRU_{t,r} = 12 * ( sum_{j=0}^{M-r} p_{t,r,r+1+j} / (1+i)^j - 11/24 )
    - i = (1 + rr) / (1 + p_index) - 1
  - Survival probabilities and mortality rates based on SPP-S-2017 mortality table (published March 2018). Example: man born in 1952 has remaining life expectancy 22.5 years at retirement age 65 in 2017; woman 25.8 years. General population life expectancy about 5 years shorter than SPP population.

### Simulation scope and outcomes
- Simulations will produce empirical replacement rates for the period 2018-2047 under a baseline and various reform scenarios, using observed contribution densities and the above assumptions for earnings, returns, fees, indexation, and mortality.
- The paper notes sensitivity of fiscal and adequacy projections to assumptions; prior studies produce a range of fiscal outcomes (roughly 0.1 to 2 percent of GDP by 2050 for pension deficits).

*Source: IMF working paper content provided in the PDF "wpiea2019149 - 0.2 percent of GDP per year over the long-term."*

### 2.5 years is assumed.

### wpiea2019149 - 2.5 years is assumed.

### Methodology and formulas
- Eq. 4: 푞푡,푥 = 푞2017,푥*(1 – 푎푥)(푡−2017)
- Eq. 5: 푃푡,푟,푖 = 퐴퐶푡,푟,푖 . 퐶푅푈푡,푟
- Individuals who accrued pension rights in the SNP before the SPP was established have the right to a minimum pension under certain conditions (such as age and contribution length).
- Final replacement rates: monthly starting pension divided by final monthly earnings prior to retirement. Replacement rates estimated only for affiliates with earnings information in 2010-17 (about one third of pension affiliates born between 1953 and 1982 are neglected, possibly leading to an overestimation of future replacement rates).

### Key assumptions for replacement rate projections (Table 2)
- Inflation: We assume inflation of 2 percent.
- Real wage growth: We assume real wage growth of 2 percent.
- Real gross rate of return: Baseline: 4.2 percent; robustness scenario: 5.2 percent. Replacement rates in 2047 are about 7 percentage points higher in the 5.2 percent scenario.
  - Rationale: Peru historical real annual returns 2003-2017: 8.9 percent (Peru) vs 5.5 percent (regional average). Over 2013-17, Peru 4.2 percent and regional 4.4 percent.
  - Long-run anchors: potential growth ~3.75-4 percent; comparison to “best practice” fund (Norwegian Government Pension Fund Global average annual gross real return 3.8 percent since inception); prior studies use net real returns around 3.5 percent or gross ~5 percent.
- Mortality assumptions: SPP-S-2017 mortality tables (came into effect in early 2019) — relative to previous tables this lowers replacement rates due to increased life expectancy.
- Retirement age: Baseline retirement age of 65; reform scenarios consider increases.
- Contribution density: Baseline assumes unchanged relative to 2012-16 average. Alternative scenario: contribution densities continue to grow linearly as they did over 2012-16 — raises replacement rates in 2047 by around 3 percentage points.

### Baseline results: replacement rates (Section IV.A)
- Projected replacement rates in baseline simulations: lie between 20 and 30 percent over the coming decades.
- Mean replacement rates over the full forecast horizon 2020-47:
  - SNP: projected to be 25-31 percent.
  - SPP: projected to be 25-33 percent.
- In the SNP:
  - The replacement rate for pensioners eligible because they contributed the minimum necessary years is relatively high at 60 percent.
  - About 60 percent of affiliates (and around 50 percent of affiliates who contributed over the past five years) are likely to receive a replacement rate of zero.
  - Replacement rates decline markedly over time: from around 45 percent in 2020 to 12-20 percent by 2045.
  - Decline mainly due to reforms improving fiscal sustainability by reducing replacement rates of younger cohorts; steepness depends on whether and how minimum and maximum pensions are adjusted (Figure 6 scenarios: full CPI indexation vs no adjustment).
- In the SPP:
  - Time-profile depends on assumptions (real rate of return and contribution densities).
  - Four scenarios considered: conservative baseline, high return, and growth in contribution densities; conservative scenario shows slight increase until about 2030 then drop ~3 percentage points; high return scenario shows rise until ~2030 then stabilization at a higher level.
  - Going forward authors use the most conservative of the four scenarios as baseline.
- Drivers and trade-offs in SPP:
  - Positive: increased contribution densities in recent years; decreased fees in recent years.
  - Negative: decreased rates of return in recent years; younger cohorts not eligible for minimum pensions; increasing life expectancy.
  - Net effect determines time profile shown in Figure 6.

### Distributional and subgroup findings
- Income quintiles (SPP): top income quintile estimated replacement rate 29 percent; bottom quintile 22 percent — largely due to lower contribution densities of lower income affiliates.
- Contribution density effects amplified in the SNP because of the 20-year vesting period.
- SNP distribution: close to 50 percent of affiliates projected to get a replacement rate of 0 (fail to contribute minimum 20-year period). For remaining 50 percent, majority of replacement rates between 30 and 100 percent.
- SPP distribution: smoother distribution, going from close to 0 to around 40 percent for most affiliates.
- Positive relationship between replacement rates and income in the SPP; in the SNP, among those reaching 20 years, strong negative relationship between replacement rates and income due to maximum and minimum pensions.
- Redistribution patterns in the SNP:
  - Minimum contribution period implies strongly regressive redistribution because low-income groups are disproportionately likely to not reach the 20-year floor; their contributed resources are effectively lost from their perspective and partly finance pensions of those who reached the threshold.
  - Combination of maximum and minimum pensions adds progressive, albeit intransparent, redistribution: redistribution from both high-income affiliates and lower-income affiliates to lower and middle-income affiliates with long contribution careers.
- Gender:
  - Women projected to have a slightly lower replacement rate than men in the SPP (about 1 percentage point over the forecast horizon), mainly due to higher life expectancy translating into lower SPP pension annuities.
  - In the SNP, gender-specific mortality differences are not considered in the SNP benefit formula; somewhat higher female contribution densities explain why projected average replacement rates are higher for women than men in the SNP.
  - Note: women account for significantly less than half of total affiliates in both SNP and SPP.

### Comparison with other studies
- Previous theoretical studies report a range from 71 percent to 28 percent depending on assumptions and schemes covered.
  - OECD (2014) high estimate explained by assumed contribution density of 100 percent and high indexation of maximum pensions with wages.
- Alonso et al. (2014): only comprehensive previous empirical replacement rate calculations for Peru; differences from current paper due to updated data through 2017, SPP-S-2017 mortality tables (higher life expectancy), inclusion of SNP affiliates with zero replacement rates, and the current paper’s inclusion of asset-based fees for mixed fee members equal to 1.1 percent. Alonso et al. use a gross rate of 5 percent (closer to optimistic scenario).

### Reform scenarios and policy levers (Section IV.B)
- Recent reforms that have worsened the outlook for replacement rates include:
  - (i) exempting the 13th and 14th salary from contributions,
  - (ii) moving towards a system where commissions are charged on the stock of contributions,
  - (iii) allowing lump-sum withdrawals,
  - (iv) reverting the requirement for independent workers to contribute.
- SPP reform levers modeled and simulated impacts (Figure 9):
  - Contribution rates:
    - Current contribution rate: 10 percent in the SPP.
    - Increasing the contribution rate to 15 percent would raise replacement rates by 7 percentage points over the long-term (assumption: immediate and permanent increase to 15 percent).
    - Trade-off: higher contribution rates may discourage formal labor participation in context of high labor informality, potentially lowering coverage and contribution densities.
    - Note: Chile discussion to increase contribution rates from 10 to 14 percent; regional and OECD comparisons imply 15 percent would still be broadly in line with regional averages and below the OECD average.
  - Contribution base:
    - Including the 13th and 14th salaries in pensionable income would increase replacement rates by over 2 percentage points.
    - Trade-off: base broadening may also affect coverage and requires careful analysis.
  - Retirement age:
    - Current retirement age: 65 for both men and women in SNP and SPP; voluntary to work until 70 (mandatory at 70).
    - Scenario: increase retirement age from 65 to 66 in five years and to 67 by 2026 would increase replacement rates by about 3 percentage point at the end of the projection period.
    - Authors consider current retirement age adequate in short- to medium-term but possibly warranting adjustment longer-term.
  - AFP fees and commissions:
    - Net real rate of return is a key determinant of replacement rates; net return increases when gross real return increases or when asset-based pension fund fees fall.
    - Moving from flow-based commissions to stock-based commissions (all else equal) reduces replacement rates, though there may be valid reasons for changing commission bases.

*Source: Authors’ calculations and assumptions as presented in the provided content.*

### 4.2 increases replacement rates by roughly 7 percentage points. And as Figure 5 showed,

### wpiea2019149 - 4.2 increases replacement rates by roughly 7 percentage points. And as Figure 5 showed,

### Pension fund fee structure and reforms (SPP)
- Prior to the 2013 reform:
  - Fees charged on the flow of gross earnings rather than on assets.
  - Out of roughly 13 percent of gross earnings (total contributions of affiliates):
    - 10 percentage points went to individual savings accounts.
    - 1.4 to old-age and disability insurance.
    - Around 1.5 to pension fund fees.
  - No additional fees collected on accumulated savings.
- Since 2013:
  - All new entrants are charged asset-based fees.
  - Total contributions are now only around 11.5 percent of gross earnings:
    - 10 percent go to individual savings accounts.
    - Fees charged on the stock of savings reduce the net real rate of return.
  - Transition for pre-2013 contributors:
    - Choice to migrate to new fee setup or stay with old one.
    - If migrating, a 10-year transition with “mixed commissions” (weighted average between flow and asset-based fees; weight on stock increases over time).
- Empirical fee levels (as of end-2017 / 2016 context):
  - As of 2016, 33 percent of affiliates and 52 percent of contributors were still under the flow-based commission system.
  - According to the SBS, flow commissions in 2017 were between 1.47 and 1.69 percent of gross earnings depending on the AFP.
  - In the mixed system the average fee structure was 0.18-1.07 percent on the flow of gross earnings and 1.2-1.25 percent on the stock.
  - December 2018 tender: for mixed-fee members the flow component will be charged at 0 percent while the stock component will have a fee of 0.82 percent.
- Cross-country positioning:
  - Peru shows a clear downward trend in the ratio of income from commissions to net contributions collected, but in terms of level Peru is toward the upper end among peers.
  - Market concentration and small market size likely contribute to higher fees.

### Effects of fee reductions and returns on replacement rates (simulations)
- Reducing asset-based fees by 0.5 percentage points:
  - Would increase replacement rates at the end of the forecast horizon by around 1 percentage point on average (given current mixed regime).
  - Once all affiliates are under an asset-based fee regime, reducing fees by 0.5 percentage point will increase replacement rates by 1.5 percentage points.
- Comparison with increasing net real rate of return:
  - Increasing the net real rate of return by 0.5 percentage points yields a 3 percentage points increase in replacement rates in simulations.
  - The difference arises because simulations assume no asset-based fees are paid during the deaccumulation (retirement) phase.

### Lump-sum withdrawals (SPP) — prevalence and impact
- Withdrawal options since 2016:
  - Up to 95.5 percent at the time of retirement (remaining 4.5 percent going to health insurance).
  - Early withdrawal up to 25 percent for a mortgage down-payment.
  - Early withdrawal up to 25 percent to pay down previously incurred mortgage debt.
- Take-up and amounts:
  - As of end-January 2018, close to 60,000 affiliates had withdrawn a total of 2bn soles, or 1.2 percent of total SPP assets, for mortgage related withdrawals.
  - 89 percent of the total value of these withdrawals was used for mortgage amortization; the remainder for down-payments.
- Consequences for pensions and annuity market:
  - Introduction of lump-sum at retirement led to a sharp reduction in retirement pensions paid by the SPP, suggesting majority choose withdrawal over SPP pension.
  - The annuity market (already small) has come under pressure (as of 2016 there were 6 insurance companies offering annuities in Peru).
- Risks to pension adequacy:
  - A 95.5 percent withdrawal at retirement places full longevity and investment risk on the affiliate.
  - Back-of-the-envelope example:
    - Retiree with 20-year life expectancy who withdraws the full amount, leaves it in a savings account with a 1 percent real return and pays themselves the same monthly amount they would have received from the SPP:
      - Savings would be depleted after 14 years, leaving no financial resources in the final six years of life.
  - Impact of 25 percent housing withdrawal on replacement rates (simulations):
    - For a 55-year-old, replacement rate would fall by around 18 percent.
    - For a 45-year-old, replacement rate would fall by around 14 percent.
    - For a 35-year-old, replacement rate would fall by around 8 percent.
    - For older cohorts who still benefit from the minimum pension, the impact is cushioned somewhat.

### SNP (public defined benefit) key features and scenarios
- Early retirement possibilities in the SNP:
  - At age 55 with at least 30 years of contributions (men) or age 50 with at least 25 years of contributions (women).
  - Age 55 (men) or age 50 (women) with at least 20 years of contributions in the event of a collective lay-off.
- Minimum and maximum pension indexation:
  - No explicit indexation rules; minimum and maximum pensions frozen in nominal terms since 2001.
  - If frozen, replacement rates would be noticeably lower than with CPI indexation (by about 8 percentage points in 2045).
  - Range of average replacement rates in the SNP between 12-20 percent by the end of the forecast horizon (scenario bounds: no adjustment vs CPI indexation).
  - If minimum and maximum pension increase with wage growth, replacement rates would be around 26 percent in 2045.
- Minimum contribution period adjustments:
  - Lowering minimum contribution length to 15 years:
    - Roughly 5 percent more SNP affiliates become eligible for a pension (assumed to get 15/20 of the current minimum pension).
    - Leads to an increase in replacement rates for overall affiliates by about 2 percentage points.
    - Estimated gross fiscal cost of close to 0.05 percent of GDP per year in the medium-term for reducing vesting period from 20 to 15 years.
  - Further options:
    - De-link contribution period from vesting period required to qualify for minimum pension (allowing lump-sum at retirement for those with shorter contributions).
    - Consider thresholds as low as 5 or 10 years, while keeping minimum pension eligibility at 20 years.
- Distributional and incentive problems:
  - Minimum contribution period creates highly regressive redistribution.
  - SNP redistributes from those who do not reach the 20-year period and high-income contributors to low-income long-term contributors.
  - Mechanism is in-transparent, unpredictable, and creates perverse incentives (high-income workers incentivized to leave SNP due to maximum pension; low-income groups incentivized to stop contributing if they cannot reach 20 years).

### Overall simulations and headline replacement-rate projections
- Baseline projections, averaging across all affiliates and next 30 years:
  - SPP (private defined contribution): 25-33 percent.
  - SNP (public defined benefit): 25-31 percent.
- Time-path specifics:
  - SNP replacement rates projected to decline from roughly 45 percent in 2020 to 12-20 percent in 2047 (largely due to past reforms for fiscal sustainability).
  - SPP average replacement rates over same period change from around 25-28 in 2020 to 23-33 percent in 2047 (range depends on real gross rates of return and contribution densities).
- Simulation results for policy changes:
  - Broadening contribution base to include 13th and 14th monthly salaries would increase replacement rates by close to 3 percentage points in the SPP.
  - Increasing contribution rates by 5 percentage points has the largest impact on replacement rates (specific magnitude not restated in text) but may have adverse effects on labor formalization.
  - Reducing asset-based pension fund fees by 0.5 percentage points increases replacement rates by 1 percentage point during the forecast horizon and by 1.5 percentage point in the longer-run.

### Policy implications and recommendations
- Short-term SPP priority:
  - Do not allow or strongly limit large lump-sum withdrawals since they place full market and longevity risk on affiliates.
- SNP reforms and trade-offs:
  - Lowering the minimum contribution period from 20 to 15 years reduces regressive redistribution but entails a fiscal cost close to 0.05 percent of GDP.
  - Consider de-linking minimum contribution period from minimum pension vesting to reduce losing accumulated contributions; allow lump-sum access for short-contributors.
- Systemic recommendations:
  - Communicate expected replacement rates under several scenarios publicly and transparently to avoid setting false expectations that could generate implicit fiscal claims.
  - Non-contributory pillar remains important given large uncovered population and should be further strengthened.
  - Competition between private (SPP) and public (SNP) pillars is a major weakness:
    - Introduces hard-to-predict fiscal risk for SNP depending on future affiliate choices.
    - Increases uncertainty about future pension levels and hampers retirement planning.
    - Complicates supervision, reduces transparency, and can increase intra-generational inequality after shocks.
    - Opportunity cost: a single integrated system could better hedge demographic/labor market risks (PAYG) versus capital market risks (funded).
  - A larger reform to restructure the system and avoid competition between private and public plans should be a priority; multiple reorganization options exist and public discussion can build on the 2017 Peruvian Social Protection Commission report.
- Research and analysis gaps:
  - Further work recommended to study best practices on private pension funds from a cross-country perspective to draw specific recommendations for Peru.

*Source: IMF working paper wpiea2019149 (excerpt provided).*

### APPENDIX I: A BRIEF INTERNATIONAL COMPARISON

### APPENDIX I: A BRIEF INTERNATIONAL COMPARISON

### Empirical replacement rates in regional comparison
- Mexico:
  - Defined contribution system, once matured (2050), expected replacement rates are "somewhat below 40 percent on average" (Alonso et al., 2015).
  - Recent IMF projections: conditional on receiving a pension, average replacement rates between 40 and 47 percent.
  - Coverage caveat: about 64 percent of workers in the private sector and 27 percent of workers in the public sector will not contribute long enough to receive a pension, implying a lower average replacement rate for overall affiliates (IMF, 2018b).
- Chile:
  - Mesa-Lago and Beltranou (2016) projected median replacement rates of 37 percent over 2025-2035 when including the public solidarity contribution (APS).
  - Benavides and Valdéz (2018) report observed replacement rates for recent pensioners: 40 percent when including the solidarity pillar and around 20 percent when excluding it.
  - Gender and income patterns: higher replacement rates for men than for women; strong relationship between income quintile and replacement rate, qualitatively similar to results for Peru.

### Theoretical replacement rates in international comparison
- Methodology and scenarios:
  - IMF (2018a) simulates replacement rates for a representative agent in several Latin American countries using country-specific parameters and theoretical contribution densities of 50, 75 and 100 percent.
  - Distributional elements (minimum/maximum pensions) and non-linear accrual schedules are included and partly explain non-linear increases in replacement rates between contribution density scenarios.
- Country-specific notes:
  - Costa Rica and Uruguay: mixed systems with higher replacement rates partly explained by a defined benefits component.
  - Colombia: high adequacy due to defined benefit (DB) entitlements; assumption that future new contributors opt for DB over DC. Individuals with a low contribution density of 50 percent do not fulfill the vesting period (1150 weeks) for DB annuities and therefore end up with significantly lower (DC) replacement rates.
  - Uruguay: long vesting periods matter—no pension paid out if fewer than 30 years of contributions have been accrued.
- Peru:
  - Representative agent generally in the SPP.
  - Using a theoretical contribution density of 50 percent, the representative agent in Peru is projected to have a replacement rate very close to that of Panama and higher than that of Chile in 2065.
  - Empirical contribution densities in Peru are below 50 percent and are lower than those in many peer countries.
- Behavioral assumption:
  - It is assumed that the agent chooses at the start of his career the system which provides the highest replacement rates for a given contribution density; thus the 50 percent scenario may reflect an SNP affiliate while the 75 percent and 100 percent scenarios represent an SPP affiliate.

### Concepts and measurement of contribution densities
- Conceptual diversity:
  - No unique consistent definition in the literature; most agree it measures intensity of contributions in a given time period.
  - Differences relate to definitions of "intensity" and "time period".
- Time-based metrics:
  - Most studies measure intensity in time units: months, quarters, or contribution years (examples: months in De Mesa et al., 2004; quarters in Braham and Marouni, 2016; contribution years in Melguizo et al., 2017).
  - Contribution time is generally related to "total potential periods of contributions" or "total months or years potentially active".
- Monetary metrics:
  - Some authors measure contribution densities in monetary units, e.g., Valdes-Prieto (2008): "share of (the present value of) earnings in the active phase of life on which the individual contributes".
  - Mc Gillivrary (2003) proposes using a monetary indicator in addition to time-based metrics.
- Measurement periods:
  - Theoretical models often consider the entire working career of 40 and more years (OECD, 2014; IMF, 2018a).
  - Empirical estimates often cover much shorter periods due to data limitations.
- Metric used in this paper:
  - Time-based contribution density metric measuring total contribution months accrued relative to the maximum possible months to be accrued.
  - Referred to as "contributions at the intensive margin" (CIM).

### Considerations for a non-contributory pillar and options for increasing contributory coverage

- A. The case for a strengthened non-contributory pillar
  - Rationale:
    - Tight link between level of development and contributory pension coverage implies near-complete coverage is a long-term goal; a non-contributory pillar remains essential in the meantime to tackle old-age poverty.
  - Evidence on Pension 65:
    - Impact evaluation (Bando and others, 2017) shows that despite nominal payments being less than the poverty line, the program improved health and consumption of recipients.
    - Counter-factual exercise using 2016 household survey data: subtracting Pension 65 payments from household income increases the old-age extreme poverty rate from 4 to 10 percent (12 to 29 percent in rural areas).
    - Comparable findings: beneficial effects similar to Mexico's non-contributory pension (Adultos Mayores) (Galiani and others, 2014); Lopez Garcia and Otero (2017) find only limited effects for Chile.
  - Design challenge:
    - More generous non-contributory pensions can disincentivize participation in the contributory pillar, potentially lowering contributory coverage and contribution densities.
    - Policy choices include whether to target the pension, at what value to set it, and what claw-back rate to apply.
    - The claw-back rate (not smaller than 0 and not larger than 1) is the rate at which the non-contributory pension decreases with contributory pension savings.
  - Reform option suggested:
    - Broaden the non-contributory pillar while introducing an intermediate claw-back rate between 0 and 1 to set incentives for contributory pension savings, following the Peruvian Social Protection Commission (2017).
    - Keeping the current 125 soles per month would guarantee virtually everybody with a contributive pension will receive substantially more in the contributive pillar than in the non-contributory one.
    - Eligibility age could be automatically adjusted in line with rising life expectancy.
      - Indexing eligibility to life expectancy implies that by 2050, Peruvians over 69 would be eligible.
      - Keeping eligibility constant at 65 would add another 0.1 percent of GDP in annual costs by 2050.
    - Fiscal impact:
      - Raising effective pension coverage to close to 100 percent (everybody who does not get a contributory pension is covered by the non-contributory one) would involve an additional fiscal cost of roughly 0.2 percent of GDP per year in that scenario.
    - Note: Kinks in the annual cost of the reform scenario are due to discrete increases in eligibility age tied to increased life expectancy.

- B. Options for increasing contributory pension coverage and contribution densities: Matching and consumption-based contributions
  - Matching contributions (ex ante subsidies):
    - Provide ex ante support for every dollar or sol contributed, increasing incentives to participate.
    - Examples in Latin America: Colombia, Chile and Mexico have elements of matching in their systems.
    - Theoretical advantages: could improve pension outcomes at lower fiscal cost and with less distortion than ex post subsidies.
    - Implementation challenges: behavioral impact in a context of high informality is not well understood; uncertainty about how large matching needs to be and to whom it should be targeted.
    - Peru-specific note: legislation calling for matching contributions for workers in firms with less than 10 employees exists but has not been implemented.
    - Ongoing evaluation: Bernal and others are running a field experiment in Peru to evaluate the impact of a 50 percent and a 100 percent matching contribution.
    - No explicit reform scenario for matching contributions is estimated in the main body of the paper due to high uncertainty regarding behavioral response.
  - Consumption-based contributions:
    - Proposal from the 2017 Social Protection Commission: complement traditional income-based contributions with consumption-based contributions operating like a sales tax directed to private pension savings accounts.
    - Empirical evidence is scarce; electronic payment roll-out could make this more feasible over time.
    - Likely limited initial impact: expected to be felt mostly at the margin.
    - Practical constraint: household survey data show rural saving rates close to 0 on average in Peru, particularly low for workers without pension coverage; these workers do not appear to substitute formal pension savings with other savings.
      - Attempting to obtain reasonable replacement rates from such "hand-to-mouth" workers is difficult and, if attempted via consumption-based contributions, would likely require prohibitively high rates.

*Source: APPENDIX I, II and III of the provided IMF content unit.*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019149.pdf_