Suggestions for Pension Re-Reform in Peru

This essay proposes the need for an integral re-reform of the Peru pension system, preceded by broad social dialogue, in order to overcome this system’s major flaws. As well as a comprehensive model of re-reform, it recommends specific measures that: extend the low level of coverage of the labor force and the elderly, expand Pensión 65, guarantee the minimum pension, improve benefits, unify the national and private pension systems, increase competition and reduce the high administrative costs and profits of private pension funds (AFP), create a public AFP, strengthen financial and actuarial sustainability, raise capital returns, and shift the disability-survivors premium to employers. This proposal also favors mandatory rather than voluntary affiliation to the AFP.


INTRODUCTION
Between 1980 and 2001, eleven Latin American countries implemented structural pension reforms that resulted in the total or partial privatization of their public systems.The pioneer, Chile, replaced its system with one based on individual retirement accounts overseen by pension fund administrators (administradoras de fondos de pensiones, AFPs).The second reform in the region, implemented in Peru (1992), did not eliminate the public system but created a parallel private one based on the Chilean model with some restrictions, while permitting transfers between the two (Mesa- Lago 2008).
The World Bank (1994) adopted the Chilean model and promoted it throughout the region, overlooking the significant socioeconomic differences between countries and accelerating the erosion of public systems.Positive effects included the unification and standardization of superior conditions of access and provisions across multiple separate schemes, reinforcement of the relationship between contribution and pension level (equivalence principle), creation of individual accounts and periodic account statements for pensioners, and generation of sufficient returns, at least to begin with.
However, most of the promises associated with structural reform were not delivered.Coverage of the economically active population (EAP) was supposed to expand given the incentive of individual account ownership, but it shrank.Administrative expenses decreased due to competition, but in Bolivia and El Salvador there are only two AFPs while oligopolies prevail in most of the other countries in the region; market concentration in the hands of two or three major AFPs has increased and transfers from one firm to another are minimal.
The costs associated with the Peruvian system are the second-highest of all private systems in Latin America.The payment of contributions was meant to improve due to the incentive of individual account ownership, but only around half of all affiliates in the region contribute on average, and even less do so in Peru.Private pensions were to be much larger than public ones, but in actuality the replacement rates are low -even in Chile.The presumption that capitalization funds would grow has taken root, but real returns, which were high to begin with, decreased across all systems and turned negative during the global financial crisis, although they later recovered (Mesa- Lago 2008, 2010).State and political interference were to come to an end, but the private systems of Argentina and Bolivia were closed without the affiliates being consulted. 1The World Bank has acknowledged that a number of such promises were not kept (Gill et al. 2005).
1. Bolivia, Chile, and Peru abolished employer contributions, infringing the International Labor Organization's (ILO) standard that workers must not pay more than 50% of total contributions.In 2010, Bolivia reintroduced the employer contribution, and in 2015 Chile's Presidential Commission on Pensions proposed its restoration.
2. For a comparison of the three re-reforms as of September 2015 with references to Peru, see Mesa-Lago (2015).3.In addition to the authors of the report, the team also included Favio Leiva, Carla Moreno, Jorge Rojas, María Alejandra Zegarra, and Richard Webb. 4. The referee of the present article pointed toward what were, in his or her opinion, problems with the report's proposals, but the very limited time I had available left me unable to obtain a response from the proposal's proponents with respect to these purported problems; thus, I have opted not to address them here.
As a result of the aforementioned problems, between 2008 and 2010, three countries implemented "re-reforms" of their private pillars or systems: Argentina and Bolivia did away with theirs, replacing them with public pay-as-you-go (PAYG) systems (Bolivia guaranteed existing individual accounts) while Chile retained its private system but improved it considerably in terms of the principles of social security (Mesa-Lago ) 3 provides a detailed analysis of the system's problems and proposes solutions.It is well-documented and professional and features sophisticated analysis, statistics that are valuable for researchers and policy-makers, and policy proposals that are reasonable but for some exceptions. 4The report generated a public discussion about the rereform, which is important and healthy.It is hoped that the parties involved will present their own respective studies in order to provide solid foundations on which to expand the debate.
My visit to Lima between September 22 and 24, 2015, sponsored by UP, was very fruitful in ascertaining the progress made regarding the 2014 report's proposal.My extensive and in-depth discussions with the entire team, the lunch meeting I attended alongside 14 experts chaired by the dean of the UP Faculty of Economics, Professor Gustavo Yamada, and my presentation in a crowded Aula Magna sparked a positive exchange of opinions, enriched my knowledge, and demonstrated what has to be a university's primary function: the exchange of academic ideas through respectful dialogue.Also useful were my interviews with the Peruvian presidential candidate, Pedro Pablo Kuczynski; congressman Jaime Delgado; the ex-Superintendent of Banking, Insurance and AFPs, Lorena Masías; and the ex-President of the Central Reserve Bank of Peru, Richard Webb.
5. All responsibility for the content of this article is mine alone, but I thank María Amparo Cruz-Saco for her useful comments, the UP team for its valuable information, and the anonymous referee for his or her comments and corrections, most of which have been incorporated.
In this paper, I offer the team and Peru a number of suggestions for the re-reform as a product of this worthwhile visit.Because the time I had to refresh my knowledge of the Peruvian system was very short, there may be errors or inadequate interpretations; consequently, comments or criticism are welcome.

RE-REFORM MODEL
I propose four pillars: a) a basic state-financed solidarity pillar, which would be an extended form of Pensión 65; b) a pillar offering a guaranteed minimum pension across the entire (unified) system, financed in part by the Treasury and subject to certain conditions; c) a mandatory capitalization pillar, financed by contributors themselves but with key reforms, such as lower commission, more competition, greater accrual in individual accounts, and better pensions; and 4) a voluntary individual-saving pillar financed by contributors with optional voluntary employer contributions. 5

Extension of Pensión 65
Twelve countries in the region have introduced non-contributory pension systems.In Peru, the Pensión 65 scheme disburses a monthly pension of 125 soles (equivalent to US $40, as in Bolivia) to individuals over the age of 65 from households identified as extremely poor (according to the Household Targeting System [Sistema de Focalización de Hogares, SISFOH]); these individuals must not receive an income from another source and are required to pass a means test.At the end of 2014 there were 450,000 beneficiaries at a cost of 675 million soles (US $238 million), equivalent to 0.12% of GDP.There are two options with respect to the extension of this pension scheme: a) targeted disbursement to non-extremely-poor individuals based on the SISFOH means test.The advantages of this approach, already put to the test in other countries, is that the pension reduces the rate of poverty, favors women more than men (greater gender equity), increases the system's social solidarity, and has a cost far below the universal pension; or b) disbursement to all citizens of retirement age, irrespective of income.The advantages of this option are that it circumvents the stigma associated with targeting the poor, and it is simpler in that it does not require means testing.However, it requires a test of "affluence" or wealth to prevent high-income groups from collecting the pension, which also has cost implications and would be difficult to implement.It is crucial for cost projections to be performed for both options.

Guarantee of a minimum pension
At present, 65-year-olds receive a minimum pension when they have contributed for at least 20 years under Peru's national and private pension systems (Sistema Nacional de Pensiones, SNP; and Sistema Privado de Pensiones, SPP, respectively).In the case of the latter system, the minimum pension is subject to more restrictions than in any other private pension system in the region: affiliates born before 1945 with at least three years of contributions have only received it since 2002 (ten years after the reform); the state only assumes payment once the funds amassed in the individual account plus the recognition bonus have been exhausted; and it must be less than the annualized minimum pension provided under the SNP.I propose that under the unified system, second-pillar contributors receive a minimum pension with uniform requirements.In order not to discourage affiliation and contribution, Pensión 65 has to be a good deal less than the minimum pension, and this in turn, much less than the average pension in the system.The fiscal cost of any new minimum pension would have to be estimated.

Mandatory or voluntary AFP affiliation
Several members of the UP research team are in favor of eliminating obligatory AFP affiliation, as is currently required, and making affiliation voluntary.This would go against the prescriptions of the ILO, which are an essential underpinning of social security throughout the region and the world. 6Thus, going down this route would mean venturing into uncharted territory with high risks of negative effects.For instance, many current affiliates would likely withdraw from the system and be left unprotected in old age, since they would be ineligible to receive Pensión 65 (renouncing this pension would be a condition for AFP disaffiliation).Thus, it would be better to concentrate on correcting the shortcomings of the current individual capitalization system than to contribute to dismantling it.The UP team proposes the retention of AFPs but, by reducing the number of affiliates, there would be a decrease in the number of AFPs and probably only two would remain -a duopoly without competition, as in Bolivia and El Salvador.

Voluntary savings pillar
This pillar has met with little success in those countries where it exists, including Chile, but it would not go amiss to propose its future implementation in Peru.One incentive for affiliation in such a pillar would be for employers, in negotiation with their employees, to offer a contribution so as to encourage company productivity while promoting social security.The employer's contribution would be deductible from tax on earnings.

CONDUCTING SOCIAL DIALOGUE AND SURVEYS
Social dialogue, with extensive participation by the relevant sectors, must precede the re-reform.The Argentine and Bolivian processes failed to do so, thereby detracting from their legitimacy.Conversely, the Chilean re-reform of 2008 encompassed broad-based social participation and 90% of its recommendations were approved by Congress, while the second re-reform, overseen by the Presidential Commission on the Pensions System, involved 25 members who represented all relevant sectors of the country, of which seven were international experts in whose number I was privileged to be included.Peru ought to replicate this highly fruitful experience.The UP team proposed an opinion survey similar to that conducted by Chile's Presidential Commission; a social protection survey of the kind already conducted in several countries in the region would also be a worthwhile, in that it would yield specific data in relation to many important variables as well as ascertaining the population's diverse points of view regarding key aspects of the re-reform.

Of the EAP
Calculation of coverage should not be based on affiliates, as many individuals will no longer be contributors due to having left the workforce, moved from formal to informal employment, etc.In Chile, coverage measured by affiliates is 114% of PEA and in Costa Rica it is 113% (FIAP 2015), which attests to vast over-estimation.In Peru, only 44.4% of affiliates contributed in 2014, a decrease from 46% in 1998 (SBS 2014; FIAP 2015).Based on contributors to both systems, 26% of the EAP is covered,7 which makes Peru one of the countries with the lowest coverage in the region. 8The fundamental problem of low coverage is the sizeable informal sector; moreover, a legal provision introduced in 2012 for the gradual coverage of independent workers was subsequently repealed.Because of the difference between regions of the country and groups of independent workers, there is a need to devise different strategies to extend coverage within this very important sector.

Of the older adult population
Coverage of the population aged 65 and over was estimated at 52% in 2015 (Asociación de AFP 2015).My calculation based on official figures as of mid-2015 is 56%, but this is an over-estimation because it also includes those in the SNP and SPP who are below retirement age but have pension entitlement -namely, the disabled, survivors, and orphans; 9 the percentage distribution is 47.2% in the SNP, 40% under Pensión 65, and just 12.8% in the SPP.In comparison with other countries in the region, total coverage falls in the middle, below Bolivia, with 97%; Argentina and Uruguay, each with 90%; Brazil, with 86%; Chile, at 84%; and Costa Rica, with 57%.As I have pointed out, Pensión 65 could be extended to all of those classified as poor -thereby expanding coverage in this group -or to the entire population.Argentina and Uruguay facilitated the affiliation of independent workers and micro-entrepreneurs through an integrated tax scheme -known as monotributo in both countries -that simplifies the payment of taxes and contributions.The 2012 Peruvian reform provides for lower contributions by micro-entrepreneurs, which as I understand it consists of an annual contribution of a fixed sum in soles rather than a percentage because of the difficulty of proving the income of independent workers and micro-entrepreneurs; in any case, this sum should be adjusted for inflation.

IMPROVING PENSION PAYMENTS AND FUND FLEXIBILITY
In 2014, it was estimated that the average pension in the SPP was 984 soles per month (US $308), paid out on condition that affiliates have at least 3,000 soles in their individual accounts, while the average pension in the SNP was 676 soles (US $211), 31% less (Asociación de AFP 2015).

Determining the replacement rate and performing re-estimations
The team's report estimates that the replacement rate is 40% in the SPP and 30% in the SNP, calculated on the basis of the latest paycheck received (Cruz-Saco et al. 2014).It would be necessary to use workers' average salaries over the last ten or 20 years or, better still, the average salary/income over an entire working lifetime.The ILO and the Organisation for Economic Co-operation and Development (OECD) set a minimum replacement rate of 45% based on the average salary over a working lifetime.

Establishing the Fondo Cero
There are only three multi-funds in Peru, which is better than Chile's five, which the Presidential Commission has recommended be reduced, also to three. 10The Fondo Cero fixed-income fund has been approved and regulated for those affiliates nearing retirement age, with the aim of preventing any possible future economic crises from reducing funds in individual accounts, but this fund only came into effect in until April 2016.As a result of the current economic slowdown, the value of the fund fell by 5% between July and October 2015 (Campodónico 2015).In Chile, affiliates are assigned by default to the least risky, or intermediate, fund, and this default fund was found to perform better than those selected by affiliates from among the five multi-funds in existence.

Compensating women for child-rearing
Argentina, Bolivia, Chile, and Uruguay disburse (financed by the Treasury) a maternity bonus for each live born child or, alternatively, reduce the number of years required for retirement in order to compensate women's unpaid work bringing up their children.As well as improving women's pensions, this serves as an incentive to affiliate in the system.The costs associated with these schemes should be estimated. 11

Modifying the annuity and programmed withdrawal
In the case of annuities, the intermediary can receive a payment of up to 10% from the affiliate (to encourage acceptance, it used to be common to offer a car), and commissions are very high.The amount received through programmed withdrawal is decreasing, and may be insufficient to cover the final years of a pensioner's life.Accordingly, in Chile, the Presidential Commission recommended the abolition of programmed withdrawal.This matter was discussed by the UP team and further research into alternatives is required to prevent affiliates from being left unprotected.

Possibility of withdrawing part of the fund
It has been claimed that Peru is a peculiar case in the region in terms of the percentage of the population that has savings.Indeed, Webb (2014) estimates that this applies to 24% of the population, which compares favorably to the figures for OECD countries.Webb recommends that affiliates be permitted to withdraw a proportion of the savings in their retirement funds to build a house, invest in a micro-enterprise, or fund education or high non-covered healthcare expenses; this has already been proposed by a World Bank study (Gill et al. 2005). 12his might make sense in Peru and could help small and medium companies (SMEs) as well as individuals who wish to build a house, but robust precautionary measures would have to be implemented, such as a requirement to have accrued a certain minimum amount in 11.All provisions added or improved by the Chilean reform of 2008 is 0.4% of gross domestic product (GDP), while the additional cost of the improvements proposed by the Presidential Commission in 2015 is 0.7%.12.A dilemma may arise in the case of young affiliates with limited resources between, on the one hand, investing in a home with the debts that this would incur; and on the other, saving for a pension in the private system, since the returns on the savings are not enough to offset the payment of the debt (Cruz-Saco et al. 2014).
the individual account; and payment of a loan of an amount linked to that accumulated in the account, to be repaid in a given period, for use in the construction of a primary home, investment in a micro-enterprise, education, or healthcare.In the first two cases, the home or the business could serve as collateral, though this would not apply in the latter two cases. 13

IMPLEMENTING SYSTEMIC AND ADMINISTRATIVE REFORMS
A re-reform of the current system is necessary to increase competition between AFPs, reduce commissions and premiums, improve efficiency, and increase deposits to individual accounts, which would certainly lead to better pensions.Moreover, it is necessary to unify both systems and standardize their access conditions.

Unifying the SNP and the SPP
The 1992 structural reform of pensions in Peru followed a "parallel" model that retained the SNP -the pre-existing publicly-administered pay-as-you-go (PAYG) system, which was characterized by multiple funds that were later consolidated by the Pension Normalization Office (Oficina de Normalización Previsional, ONP) -and placed it in competition with a new, fully-funded privately-administered system of individual accounts, the SPP.New affiliates are free to choose between the two systems.This model, the only one of its kind in Latin America and the world except for the Colombian system that followed it, has given rise to a number of problems; for example, many SNP affiliates overlook that fact that they are supposed to contribute a minimum amount for a 20-year period and that this must be documented, which is often not possible, with the result that they are left without pensions and cannot get back what they have paid in.Unlike in other countries, the Peruvian structural reform failed to make implicit pension debt completely explicit, given that the PAYG system (SNP), which receives substantial and ever-increasing fiscal transfers, was retained.Since there is no guaranteed pension across the entire system, the sole fiscal cost is the recognition bond disbursed to those affiliated in the private system who had amassed contributions in the public system at the time of the reform.In 2004 and 2008, the actuarial deficit in the SNP was estimated at between 23% and 25% of GDP; this had declined to 21% by the end of 2013 given that economic growth exceeded the deficit that year (Cruz-Saco et al. 2014).Of the total number of affiliates in 2014, 62% were affiliated with the SPP and only 38% with the SNP, while the affiliation figures for 2012 and 2013 showed a growth in the SNP; however, in 2014 the rate went down (based on figures provided by the UP team).
13.After this study was completed, the Congress approved provisions for affiliates of the SPP system to withdraw 95.5% of the funds in their individual accounts (there was no time was available to review this law); I understand that the President is currently reviewing the draft of this law.
I propose the closure of the SNP and the standardization, as far as possible, of access conditions, pension calculations, and contributions. 14Ideally, the entire system would be done away with at once.Another option would be to close off access to new contributors.According to the World Bank (2004), the latter option would be less costly than the former, but it would not make explicit part of the implicit liabilities in the SNP.Moreover, it would perpetuate diversity and create the temptation for irresponsible politicians to increase disbursements under the SNP with a consequent increase in fiscal costs.Current SNP affiliates would transfer to the SPP, but to an updated version rather than the system in its current form, as I have proposed in this article.One obstacle is that there are no individual accounts under the SNP; however, since 1999, employer contributions to the ONP have been accounted for, and these contributions are now digitized.In addition, earlier contributions are currently being digitized, but it is not known when this task will be completed.These tabulated contributions would be certified by the ONP or allocated to the individual account of each affiliate who transfers to the SPP, constituting their initial capital.As is the case in all structural reforms in the region, the state would fund the recognition bond.Moreover, US $5 billion belonging to SNP affiliates has been deposited in a Consolidated Reserve Fund (Fondo Consolidado de Reserva) at the Central Bank, which should have earned 6% annual interest year.This sum could be deposited and invested in a special independent fund, which would gain income and help offset the fiscal expenditure on recognition bonds.Estimation of the fiscal cost of these bonds is essential.
The special pension schemes with generous access conditions and payouts -the so-called "living decree" (cédula viva), for senior public officials and teachers -have been gradually brought into line with the rest of the system.This required a constitutional reform -an unusual measure in the region.In the SNP, the special regime for miners endures, whereby these workers can retire at 45 years of age following ten years of service; this carries a very high financial cost and is actuarially unsustainable.While it is logical that these workers are subject to special conditions given the high-risk nature of their work, the scheme ought to be reviewed in order to balance it out.Outside the SNP, the armed forces enjoy very generous pension entitlements and receive tax credits; these should be integrated with the SNP or reformed and actuarially balanced.

Reducing AFP costs and profits by increasing competition
Peru has the second highest cost, an average of 2.78% (1.55% in net commission and 1.23% on premiums), of the nine private pension systems in the region. 15Colombia occupies first 14.A similar reform was proposed by Olivera (2010, 2016).15.The others are: Chile, Colombia, Costa Rica, El Salvador, Mexico, Panama, Dominican Republic, and Uruguay.
place, but its coverage of disability and survivorship risks include employer contributions.
According to Barr and Diamond (2012), an administrative fee of 1% over an affiliate's working life reduces accumulation in the individual account by 20%; thus, in the case of Peru, the rate would be 30%.Hence, there is a need to reduce the cost of the Peruvian system and the profits of AFPs 16 by increasing competition.The 2012 law introduces a number of measures to this end, such as the net commission tender process for AFPs in which the company that offers the lowest rate will gain access to new entrants in the job market.This rate of commission also applies to the winning AFP's pre-existing affiliates.
To date, two tender processes have been conducted.
In Chile, in the 2014 procurement process, winning AFP a net commission rate of 0.47% (compared with 1.47% in Peru).However, this AFP only has 20% of affiliates while the remaining 80% are concentrated in the larger AFPs, all of which charge 1.5%.Despite the fact that these AFPs charge more, they do not necessarily obtain greater returns.This phenomenon is the result of two factors: lack of knowledge on the part of individual affiliates, and the intense publicity and use of marketing power by AFPs with superior resources, which stress the "security" of their corporations and the risks associated with transfer to a cheaper competitor.There is a need for greater education of affiliates and for the publication in the media of simplified comparative information regarding commissions, premiums, and real net returns across the AFPs.The Chilean Presidential Commission also recommended that the cheapest AFP be assigned not only new affiliates in addition to those it already has, but also a proportion of affiliates from other AFPs (unless they expressly opt to remain in the more expensive ones).In Mexico, new affiliates are automatically assigned to the pension fund management company (Administradora de Fondos para el Retiro, AFORE) that offers the highest returns.
The 2012 Peruvian law established collective SPP insurance, which provides for periodical tender processes for the right to offer a collective disability and survivorship premium to SPP affiliates (rather than being determined by each of the AFPs, which are usually associated with an insurance company).The objective of the law is to create a single company that covers all affiliates by way of a single premium while promoting cost reductions due to the larger pool of contributors.The premium was set at 1.23%, representing a reduction between 2012 and 2015 of only 0.08% (SBS 2015a).It is necessary to investigate why a greater reduction has not occurred.
16.According to Delgado and Fuertes (2010), in the period 1993-2009, the AFPs' private returns were almost seven times greater than the real net returns of the pension funds.Cruz-Saco et al. (2014) show that between 1998 and 2013, AFP returns (net income on net intangible assets) were several times the value of affiliate returns, except for the period 2005-2007.
Another issue of note is the transfers between AFPs, which in the case of Peru have been very limited and much less frequent than in Chile.In the past, transfers were very complicated and prompted long lines of affiliates and delays in the process.Now, because of the internet, transfer is much faster, but there has not been a large increase in the percentage of affiliates who transfer annually and most processes continue to be conducted by the AFP sales forces compete with one another for affiliates, not necessarily to the benefit of the individuals concerned.Much of the AFPs' operating costs correspond to marketing, such as publicity and sales staff, which can be deduced.The elimination or the introduction of a cap on such expenditures as means of reducing commissions is something that ought to be considered.17 At present, the money saved from reductions in commissions and premiums in Peru are not deposited into individual accounts, as should be the case, particularly since the 10% that is currently discounted from salaries is insufficient for an adequate pension.Such deposits would help to increase the value of pensions.

Creating a public AFP
It is extremely difficult for new AFPs to enter the Peruvian market due to the very high costs involved and the relatively low number of contributing affiliates.In the survey conducted by the Chilean Presidential Commission (2015), 79% of those interviewed were in favor of the creation of a public AFP, and 69% said that they would affiliate with one.With a large majority of votes, the Presidential Commission recommended the creation of such an AFP, but subject to the same rules as private AFPs, with autonomous equity and total independence from the state.In this regard the successful experience of Uruguay sets a precedent, where República, the state-run pension fund management company (Administradora de Fondos de Ahorro Previsional, AFAP), reduced net commissions and attracted a majority of affiliates to the system, thus prompting the other, private AFAPs to reduce their commission levels.In Peru there is considerable distrust of state entities, exacerbated by the PetroPerú scandal; however, the Central Bank is ranked first in terms of public trust and the National Institute for the Defense of Competition and the Protection of Intellectual Property (Instituto Nacional de Defensa de la Competencia y de la Protección de la Propiedad Intelectual, INDECOPI) is well regarded by the population.Both are models from which inspiration ought to be drawn for a public AFP.18

Prioritizing commission on balance instead of remuneration
Commission on remuneration (flow) assures a fixed inflow to AFPs that is not related to their performance and contributes to the "bandwagon effect," as with similar investment portfolios.In 2012, a mixed commission scheme was established: on flow and on balance accumulated in individual accounts; the latter is intended to reward good AFP performance -for instance, those with greater profitability.Current affiliates can choose between the two commission schemes while new entrants are assigned to the mixed scheme (which progressively gives way to commission on balance), so that in the long run all affiliates will pay commission on balance. 19Following extensive public debate, the mixed scheme was broken down as follows: 30% on flow, and 70% on balance. 20In deciding between the two commission schemes, affiliates' interests and accumulated capital played an important part: those with very little preferred commission on balance, while those who have accrued larger amounts of capital opted for commission on flow. 21tween December 2012 and June 2015, commission on flow decreased in three of the four AFPs (-0.19, -0.41, and -0.42), and remained the same in the case of the fourth.Commission on balance did not change over the period (approximately 1.22%) and remains very high (SBS 2015a).In Mexico, commission on balance decreases as the individual fund increases.It has not been possible to determine whether there has been a transfer of affiliates between AFPs with higher and lower commission on flow, but the difference is very small (only 0.22).

Making information about AFPs more transparent and accessible
The aforementioned report by the UP team identifies problems related to the availability of information in the Peruvian AFP system (Cruz-Saco et al. 2014). 22In general, surveys conducted in a number of countries with private systems attest to a lack of knowledge on the part of affiliates about fundamental system elements.For example, in Chile, only 11% of affiliates know the amount of the commissions charged by AFPs, 15% the solidarity payment made by the state, 16% the percentage of income subject to payment of contributions, and 18% how pensions are calculated (Comisión Presidencial sobre el Sistema de Pensiones  2015).The relevant information needs to be provided by FPAs in a clear and transparent manner, with simple comparisons between the key variables (commission, real net returns, etc.), and published in the media.Moreover, the periodic individual account statements should be simplified, and AFPs ought to engage in the educating of affiliates.The 2008  Chilean re-reform established a Pension Education Fund (Fondo de Educación Previsional) to better inform affiliates and the population at large.In addition, the Presidential Commission has recommended that courses on financial and pension education be provided at schools.The 2012 Peruvian reform stipulated the creation of an education fund, but I was unable to access information regarding the outcome of this provision.

Re-establishing the Pension Superintendency with treasury financing
Peru used to have an independent Superintendency of Pensions, but it was merged with the Superintendency of Banking, Insurance, and ADPs (Superintendencia de Banca, Seguros y AFP, SBS) as a cost-cutting measure;23 this entity is not fit for the purpose, in that it groups together three sectors with diverse interests.As such, it would be better to revert to a separate Superintendency of Pensions -which would also comprise the ONP until the system was unified -staffed by technocrats specializing in pensions, thus preventing any possible conflicts of interest.However, any such division would render a new superintendency costlier to the pension sector in that it would only oversee and draw finances from the four AFPs while losing out on the funding that the SBS currently receives from banks, financial institutions, insurance companies, etc.In my view, it would not be healthy for the AFPs to finance such a superintendency due to the attendant risk of dependency; a better alternative would for the Treasury to fund it, as in Chile and other countries, while safeguarding its autonomy and independence from both the state and the AFPs.

Worker participation
Workers own their pension fund accounts but have no say whatsoever in their management, and nor are they able to monitor them.The re-reforms in Argentina and Chile introduced committees to oversee implementation.In Chile, these committees issue annual reports that evaluate progress made and problems encountered, as well as proposing measures to address them.Chile's Presidential Commission recommended the creation of a directorate with worker participation.In Peru, it has been proposed that an affiliate representative participate in the AFP Investment Committee (Delgado and Fuertes 2010).The 2012 Peruvian reform provided for the creation of a commission with direct citizen participation, but I was unable to access information regarding the outcome.There is a need to assess possible channels for effective social participation in Peru.

UNDERPINNING FINANCIAL AND ACTUARIAL SUSTAINABILITY
The aforementioned changes would count for nothing if the system's financial and actuarial stability were not assured (based on the model proposed) in the long run.

Increasing fund profitability
The fund's real gross annual returns over the period 1992-2014 were 7.5%, but they have been on a downward trend; in 2013 and 2014, they stood at 4.4% (SBS 2015c).This is a concern, since future pensions are funded by profitability to a proportionally greater extent than by pensions (Barr and Diamond 2012).To obtain a more precise figure with regard to profitability as well as a medium-and long-term perspective of these investment options, it would be necessary to prepare a historical series of net returns (with commission discounted), along with a percentage breakdown of the portfolio by instrument and the real returns of each one. 24The rate of investment in government securities was 18% in 2014, probably the lowest in the region; 25 this is positive, because it is indicative of the independence of the state, but it could be increased by way of profitable instruments (see below).Of this investment total, 41% corresponded to foreign issuances (the greatest percentage after Chile), and authorization has been granted for the proportion to be increased further (SBS 2015c), which will be very risky for affiliates. 26 2014, the capital accumulated in individual funds totaled US $37.962 billion (the fourth greatest of the nine private systems in the region; FIAP 2015), the equivalent of 19% of GDP.The Lima Stock Exchange is relatively small and incapable of absorbing this enormous volume of funds.In order to diversify the portfolio and contribute to the development of Peru, I recommend channeling part of the investment to SMEs -so as to reinforce these companies that are so important to the country's economy -through bonds issued by a number of banks that provide small loans. 27Investment could be made in the country's infrastructure through bonds tradable on the stock exchange, which would contribute to development by way of roads, bridges, potable water, sewage systems, irrigation, and so on.Investing in housing plans is not advisable because the historical experience in the region is very negative, although it could take the form of investment in mortgage bonds tradable on the stock exchange (there is a small percentage invested in these instruments).
The SBS should play a more active role in promoting the diversification of portfolios and the generation of new profitable instruments.
Finally, the law establishes minimum returns for all AFPs of a rate equal to the industry average over the preceding three years.This requirement is better than a single year, which was the stipulation when the private systems were first created, because it stimulates include: a) the cost of the operating deficit in the SNP (the experience of other countries suggests that it is declining, but it remains the largest component of the transition cost while the operating deficit lies in the SNP, whose implicit liabilities have not been made explicit); b) the cost of the recognition bond and the guaranteed minimum pension; c) the fiscal costs of the new pension entitlements under the re-reform; d) the aggregate income from reduced fiscal costs and possible new income; e) the long-term actuarial balance of the unified system; and f) the impact of the suggested changes on the average pension under a unified system and the replacement rate on average income over a working life.

CONCLUSION
The most recent structural reform of the Peruvian pension system was carried out 25 years ago, and the numerous amendments that this reform has since undergone have failed to address its fundamental failings.But it is now time to embark on a wholesale re-reform to introduce a better system that fulfills the basic principles of social security.The President of the Peruvian Association of Insurance Companies (Asociación Peruana de Compañías de Seguros) argues that these principles need not apply to individual savings schemes because "the AFP system is diametrically opposed to a social security system" ( Morón 2015). 31But the last structural reforms in the region promised to improve the old public PAYG social security system by expanding population coverage, disbursing adequate pensions, and reducing administrative costs through increased competition, as well as other aforementioned principles.These promises have not been kept in the vast majority of countries.At present, El Salvador is considering a possible re-reform and other countries will probably go down this same route.
Broad-based social dialogue and the contribution of professional and documented proposals by other actors is vital as part of a serious, thorough, and respectful debate.Chilean AFPs have shown themselves to be somewhat flexible in the pursuit of consensus.Their Peruvian counterparts ought to recognize that such an attitude is necessary if the system is not to be abolished, as has occurred in Argentina and Bolivia.In the current electoral campaign, the presidential candidates should include this crucial matter in their policy platforms, promising to appoint a commission representing all sectors involved to evaluate re-reform proposals, and issue a report as part of a precise agenda, whose recommendations would then be debated by Congress in a responsible manner so as to avoid politicization.If this were to be achieved, all Peruvians could enjoy decent pensions at the end of their working life.
2014, 2016; Becker and Mesa-Lago 2013; Mesa-Lago and Bertranou 2016; Bertranou et al. 2016).To address outstanding problems, a Presidential Commission on the Pension System in Chile issued its report on September 15, 2015 (Comisión Presidencial sobre el Sistema de Pensiones 2015). 2n Peru, the 2012 parametric reform, influenced by that of Chile, has not brought about substantial improvements.In Chile, independent workers have been incorporated gradually since 2012, but in Peru this measure was suspended, partly due to the opposition of this group of workers.
The 2014 report prepared by a team of researchers at the Universidad del Pacífico, UP (Cruz-Saco et al.2014