6th Global Pension and Savings Conference Looks at 20 years of Pension Policies

April 7, 2014

The following are proceedings from the 6th Global Savings and Pensions Conference in Washington, D.C.

Loïc Chiquier, Director for the Capital Markets Practice and Middle East and North Africa Region and Arup Banerji, Director for Social Protection and Labor at the World Bank Group welcomed the Conference delegates.

They stressed the importance of the core themes of learning lessonsdeveloping solutions and improving outcomes if pension reform was to help eradicate poverty and boosted shared prosperity sustainable. They welcomed over 150 delegates from nearly 50 countries - a mix of government, regulators and supervisors, private sector and academics practitioners and regional and international organizations including the World Bank, IMF, OECD, Asian Development Bank and Inter-American Development Bank. The conference aimed to use this powerful global expertize explore the ‘iron triangle’ of sustainability, coverage and adequacy alongside efficiency and security that together made up the key outcomes of pension systems.

Session 1: Setting the Scene: 20 years of Global Pension Policy

The opening session heard from the lead author of ‘Averting the Old Age Crisis’ Estelle James, Professor SUNY and former Lead Economist, World Bank. She returned to the fundamental insight of using different tools to meet separate objectives for income redistribution and saving. 

Redistribution via a DB pillar can take many forms but would require direct financing from general revenue. But this poverty alleviation focus would not provide adequate income in retirement for all, so additional saving was needed – with mandatory DC schemes a key recommendation, along with voluntary saving. Over 40 countries had undertaken (country-specific) reforms along these lines – because the demographic challenges outlined in the book made reform essential. Many criticisms of the book came when people had not read its nuanced – public and private, DB and DC message. Policy reversals had occurred in some countries where political consensus was absent but most countries were sticking to their reforms. In some cases huge implicit pension debts had not been tackled sufficiently. A major legacy from the book has been the extensive research into the practical aspects of pensions reform, including behavioral economics and payouts.

Robert Palacios, World Bank Group, provided a broad survey of pensions over the previous centuries. The Pay As You Go DB model had expanded massively before its critical sustainability issues were acknowledged. But the shift to add DC pillars (as in Averting) rightly did not mean getting rid of DB but reforming it. A second shift was from exclusive reliance on payroll tax/contributions towards general taxation and ‘social pensions’ to improve coverage.

Dalmer Hoskins, U.S. Social Security Administration chaired a lively discussion including whether Notional Defined Contribution systems could operate well in developing economies with volatile employment and wages, the importance of labor market incentives, the need for clarity on the long run costs and benefits in pensions and the importance of diversified solutions - not public or private - but both.

Session 2: The Sustainability and Adequacy Trade-off as Countries Age

Anita Schwarz, World Bank Group, examined the experience in Europe and Central Asia drawing on the recently published ‘The Inverting Pyramid. The demographics in some countries had been even worse than expected as aging and falling birth rates were exacerbated by emigration. There had already been significant reform but often increases in coverage and generosity of benefits as well. The presentation showed a policy options exercise highlighting the need to focus state budgets on preventing old age poverty given fiscal capacity and to expand saving for income replacement – using a wide range of options from traditional ‘mandatory’ savings to the use of auto-enrolment alongside longer working lives. ‘Active aging’ policies could ensure older workers continued to play a productive role.

Ambrogio Rinaldi, COVIP, Italy and Chair of the OECD Working Party on Private Pensions, highlighted a range of reforms that had been popular and successful globally to achieve a better balance of coverage, adequacy and sustainability. These included extending working lives, improving efficiency through lower costs, improving diversification through the use of multiple pension pillars or tiers and improving the security of pension systems through regulation and supervision. Notional DC systems should be considered where appropriate. Clarity was needed on intended and unintended redistribution between generations. In the EU the Stability and Growth Pact treated a specific negative incentive towards ‘hiding’ pension liabilities in implicit debt instead of improving sustainability through funded pensions.

Agnieszka Chlon-Dominczak, Warsaw School of Economics, gave the history of the Polish reforms 15 years ago and their implementation. The system had gained scale and had relatively good underlying performance. But pension contributions diverted to private pensions were not replaced in the public budget by other tax revenue. Privatization proceeds were not high enough, or were used for other priorities. And the original system was made more generous than initially intended. So tackling fiscal issues during the transition was a critical lesson. In the future labor markets needed to support longer working lives. However, trust in the system was now low.

Donghyun Park, Asian Development Bank, spoke about the challenge of rapid aging in Asia. Coverage, adequacy and indeed equity are lacking, while public pension costs are expected to rise sharply. Urgent structural responses to ageing populations are needed; increasing the credibility and reach of funded pension systems, investment professionalization, cost reduction, and removing serious barriers to labor market participation by older workers. Despite the very rapid speed of developing and demographic transition there was so far less reform than would be expected. Leaders needed to seize the challenges urgently. In discussions the session chair. 

Heinz Rudolph, World Bank Group, observed that policy makers need to act now while a trade-off is possible rather than wait until drastic action was needed. Other comments included the need to use general taxation to reduce payroll taxes for state pensions to allow contributions to private pensions, the need to focus more on politics and communication, promoting strong evidence that longer working lives do not mean fewer jobs for the young and the continuing role of intra-family support.

At lunch George Symeonidis, Hellenic Actuarial Authority, Greece spoke on the Greek experience. Unsustainable growth in PAYG costs and promises had finally to be tackled after the 2008 crisis. The EC’s policy of increasing working lives and retirement ages as well as reducing excessive employer contributions and the generosity of PAYG promises have been largely adopted. But the changes were very tough because the issues had been ignored for too long. Proactive reform is essential.

Session 3: Challenges and Innovations in Expanding Coverage

Maximizing coverage, or financial inclusion, is an essential outcome for pensions.

Brigitte Madrian of Harvard University presented the latest evidence on effective policies and interventions to increase coverage in voluntary systems. She showed financial incentives are often much less effective than ‘nudge’ mechanisms like auto- enrollment. For governments wanting to maximize the efficiency of spending on retirement, she presented a broad menu of optionsSome, like auto-enrolment, have sizable effects; others, like making plan simple to understand and join, may have less dramatic impacts, but are still very cost effective. Politics and views on paternalism can also influence design choices – but this will be very country-specific.

Mariano Bosch, Inter-American Development Bank, presented results from ‘Better Pensions, Better Jobs’ on Latin America and the Caribbean. The labor market is at the ‘epicenter’ of the problems and solutions due to the challenges of informality at all income levels and across many professions. It was critical to understand the impact of income, occupation and firm size and design policies that would target the specific issues. Simple subsidies or tax relief may be ineffective. Creative solutions – using the media or social pressure for example to increase coverage among domestic workers could work. But always it was important that each intervention was part of a coherent overall pension system.

Guatam Bhardwaj, Invest India Micro Pension Services, followed with a case study on increasing coverage in India. There are many challenges in providing pensions in an economy with 450 million workers but only 30million tax payers. Over 300million workers are excluded from the formal pension programs. In addition to informality there are issues of cost, access, literacy and geography to overcome. The solution uses existing payments and mobile phone infrastructure which has transformed the access to the poor in remote locations with very irregular income (see video).

Charles Counsell, The Pensions Regulator, UK, explained how auto-enrolment developed from concept to over 3.2 million more pension savers. Employers must offer, contribute to and put workers into pensions automatically. Workers can then choose to opt out. This has reversed the declining trend in coverage. Opt-out rates so far under 10%. He contrasted the focus in the UK on the employer to choose a pension and compliance via the Pension Regulator with the New Zealand approach of default funds and individual choice and compliance by the tax office. In the UK there has been a big focus on educating employers and the supply chain, such as payroll software providers, as well as on individual to deliver high awareness and low opt- outs.

Edward Odundo, CEO, Retirement Benefits Authority, Kenya and President IOPS moderated an interesting discussion that included how to establish an integrated eco-system of suppliers and savers; whether auto-enrollment is more than just a second-best to mandatory schemes; appropriate models for emerging countries with limited but expensive existing coverage; whether sustainable programs exist for the self- employed; and the mechanics of how the Indian and UK systems work in practice.

Session 4: Global and Regional Reflections Roundtable

Ana Revenga, Acting Vice President of Poverty Reduction and Economic Management, World Bank Group moderated a roundtable providing external reflections.

Benedict Clements, Division Chief IMF, set out the current and future costs of aging from the IMF’s ‘Equitable and Sustainable Pensions’Pension expenditure was set to increase by 2.5% of GDP by 2050 with health spending going up even more. Many countries had already reformed but there was more to do. A range of policies are needed to ensure cuts in benefit levels do not take all the strain as this would increase risks of poverty.

Edward Odundo, CEO, Retirement Benefits Authority, Kenya and President IOPS, highlighted innovations in targeting the informal sector. As in India mobile phone money transfer was central for access and costs. The funds could then be managed by fund managers with custodians to ensure security of accounts. 

Carlos Ramirez, President, CONSAR, Mexico reminded people about the successes of funded pensions in Latin America, coupled with continuing coverage challenges and the importance of individuals being educated to be more responsible for their pension provision.

Uluc Icoz, Head of Private Pensions, Turkish Treasury, outlined recent changes to replace tax incentives with matching contributions. In the first year of operation more people were contributing, average contributions had increased and withdrawals had decreased. Other reforms were focusing on lowering fees, improving investments and exploring auto-enrollment.

Junichi Sakamoto, Chief Adviser, Nomura Research Institute highlighted the strong political opposition in Japan to any ‘privatization’ of social security. There had been a series of changes to contributions and benefit levels across the pension system which ranged from the (world’s largest) social security fund through to corporate DB and DC plans. Although it had been much discussed defining inter-generational equity remained difficult.

Finally, Edward Whitehouse, Axia Economics retuned to some of the key modelling in Averting the Old Age Crisis. Understandably in some areas the demographic projects from 1994 were too optimistic but in others aging had been somewhat slower. Multiple reforms had helped to improve the situation compared to that which would have been faced. But there was still much to do. It was critical to create diversified systems with broad based but affordable public provision matched by increased private saving.

A range of questions and comments then followed. Some worried increased retirement ages would reduce youth employment but in fact the evidence did not support this fear. More attention was needed on politics and simple communications as well as the technical demands of the IT and infrastructure at the other end of the spectrum. Tax subsidies for higher income savers seemed poor value for money. Overall, there was a need to improve the clarity on ultimate objectives and a focus on tools to target different problems as well as a call for more focus on the most challenging country cases.

Session 5: Delivering Better Costs and Investment

Improving net of fee returns improves the efficiency of a system and can contribute to easing the adequacy-sustainability-coverage issues discussed on the first day.

Mike Heale, CEM Benchmarking, used their unique long run data base to show what drives costs and returns. Simply put high costs do not lead to higher investment performance. The best long run predictor of investment returns net of fees is the level of fees themselves. This is because it is possible, particularly in large organizations, to combine both low costs and good investment performance. This can be from buying power with external managers or by having the scale to do investment in-house. The wide cost dispersion of similar services shows the scope to improve outcomes for many funds.

Will Price, World Bank Group, then considered how market structure, scale, expertise and governance act as drivers of improved costs and investments. He argued it was critical to pay attention to the full value chain of a pension system – to demand and supply and distribution channels. Good systems focus on all aspects. These include strong governance in member interest, large scale (which can be done even in small countries), rigorous expertize and sensible direct controls of fees. Evidence of good practice ranged from state, employer based and private systems in Sweden, Denmark and (parts of) the US and UK through to Kosovo, Peru and Malaysia.               

Annika Sunden, Swedish International Development Cooperation Agency, provided a case study on Sweden. The public pension and mandatory DC ‘second pillar’ share an administration platform that keeps costs low. The second pillar uses a blind accounts or central clearing house model so that fund managers receive bulk transfers to invest but do not manage individual member accounts or know their individual customers. Fund managers have to provide fee discounts that rise with assets under management – ensuring scale economies are passed on to members. And a default fund ‘AP7-Safa’ was established with a low cost diversified investment focused on long run returns. It has outperformed the market average after costs and operates at around 12 basis points compared to over 30 basis points for the average fund.

Michel Canta, Superintendent of Insurance and Pensions, SBS, Peru, considered the impact on costs and investment of auctions in Peru and Chile. In both countries the auctions had reduced costs. In Chile the cost reductions apply only to new entrants to the market. In Peru the cost reduction must be applied to all members of the successful firm. But, in addition, even unsuccessful bidders must offer their existing members the (lower) fees at which they bid into the auction. The full impact in Peru has been limited by the existence of the public Pay As You Go system as a default ‘choice’ (even though it is unsustainable) so fewer people have private pensions.

The discussion chaired by Carlos Ramirez, CONSAR Mexico probed issues including: the lack of persistence in pension fund performance; how to reduce costs of UCITs providers, and ensure lower cost are passed to consumers; the role of voluntary savings and fee caps in Sweden; the power of pension funds internalizing fund management to drive down costs; how to achieve increased scale; and how to cope where lack of scale in funds or markets is inherent. Participants heard examples from India where good design allowed access to professional fund management at low cost.

Session 6: Investment: Global Trends to Practical Infrastructure Projects

This session continued the theme of improving the efficiency of pension systems by looking at investments in practice.

John Rogers, CFA Institute delivered a broad ranging overview of investment practice and challenges pre and post crisis. He explained how sophisticated investors are moving away from policies built on fixed allocations to narrowly defined asset classes. It was essential to manage individual behavioral flaws such as suspending re-balancing during extreme periods and herding. Overall he saw a return to an era of Fiduciary Capitalism based on long run owners investing for the long term.

Fiona Stewart, World Bank Group, then started a focus on the potential role of pension funds in infrastructure investment. She explored the wide range of different project types with different risk-return profiles which are suited to some country circumstances but not others. She advanced a decision tree of questions that could be used to determine which approach might work depending on country circumstances. This work was aimed at helping countries at all levels of development assess how best to develop good quality projects.

Kirstine Damkjaer, Global Infrastructure Fund, IFC, then spoke about building the capability to develop infrastructure investments. She noted that the infrastructure funds market is still at an early stage of development and may not meet institutional investor needs. She highlighted that most funds are in developed countries. There are many ways countries can package investment with different amounts of risk transference and potential liquidity. The IFC’s Global Infrastructure Fund had one of the longest track records and provides features pension funds should be looking for in such investment.

Clemente Del Valle, Financiera de Desarrollo Nacional, Colombia, provided a case study on developing investment in transport infrastructure. The Colombian Government responded with an ambitious public private partnership model that learnt from past failures. Critical to its design was a clear regulatory framework, clarity about risk sharing and improvements in the capacity of the organizing institutions. The investment instruments will have strong governance and investor protection mechanisms to make investment secure enough for pension fund participation.

The ensuing discussion was moderated by Satheesh Sundararajan, World Bank Group. There was much focus on how the daily pricing required in many DC pension funds and IFRS rules on market to market valuation make it difficult for pension funds to invest in infrastructure. There were examples of Government guarantees assisting infrastructure investment but concerns that such investment would not assist diversification in portfolios already heavily exposed to sovereign debt risk. There was a call for some best/international standards for infrastructure instruments. Issues of capacity building, governance and good regulatory design were also covered. A key takeaway was the need to build better data and have a rigorous and systematic approach to ensuring infrastructure was a good investment for pension funds.

At lunch, Stephen Goss, US Social Security Administration, set out the funding challenges to the US Social Security pillar. He showed that longevity was responsible for only 20% of the funding pressures. Far more important were the very significant falls in the birth rate. Falling fertility rates have caused a permanent rise in retirement cost per worker that requires lower benefits paid or higher contributions. The issue was politically very contentious and there was no sign of a ‘grand bargain’ yet. But the issue was too important to leave so the debate would continue.

Session 7: The Payout Phase: Products, Providers and Structures

This session highlighting how many of the insights into improving the accumulation phase for members are directly applicable for the payout phase.

Edward Whitehouse, AXIA Economics, focused on the size and nature of the retirement income challenge. Using the replacement rate he showed the variation in outcomes across OECD countries using the latest system parameters. He then highlighted the need to look at both average contribution rates and average number of years of contributions in order to find a combination that would deliver a decent retirement income. Clearly if people contributed for fewer years their average contributions had to rise. Echoing earlier contributions he highlighted how tax incentives are often expensive ways to improve coverage and contributions to fill a pension gap – and auto-enrollment and mandating enrolment may be more effective.

Craig Thorburn World Bank Group, noted that the design of the payout phase needs to be integrated with the goals and targets of the accumulation phase. This emphasized the message from other speakers about the need for clarity about long term objectives – and ensuring different parts of the system delivered both poverty reduction and ensured adequate income smoothing in retirement. He reviewed different payout options and who could provide them across the public and private sectors. He then considered distribution and advice issues and noted how simple it was for advisers to manipulate projected payout illustrations.

Annika Sunden, Swedish International Development Cooperation Agency, explained the automatic balancing in the Swedish system to adjust for lower investment returns or increased longevity. This was not risk sharing so much as risk smoothing as individuals ultimately saw the impact of investment returns or rising longevity in their pension payouts. A recent temporary change in the default option at retirement had seen a massive temporary change in the ‘choice’ of retirement product – emphasizing again the power of defaults to direct consumers to a particular option.

Carmen Hoyo, BBVA, then outlined the role of private sector providers in delivering annuities in developing and developed markets in Latin America. She contrasted a range of features of the Chilean, Peruvian, Colombian and Mexican markets from capital rules to mortality assumptions. She noted the need to raise the levels of voluntary saving and hence financial education, and considered the use of intelligent pricing.

The ensuing discussion, chaired by Roberto Rocha, World Bank Group, focused on: having the right menu of payout choices for country circumstances; intelligent pricing; the lessons from behavioral economics given the greater complexity of payout products; the merits of risk sharing as opposed to age-indexing of pensions; the negative implications of recently announced changes to UK payout rules; and the poor track record of reverse mortgages which highlighted they were not a magic solution.

Session 8: Bringing it all together: Lessons, Solutions and Outcomes

This session focused on the overall themes of learning lessons, developing solutions and delivering outcomes that matter.

Richard Jackson, Senior Associate Center for Strategic and International Studies (CSIS), warned global ageing would challenge the ability to provide decent living-standards for the old without a crushing burden on the young. The CSIS’ ‘Global Ageing Preparedness Index’ is based on projections of total government benefit spending and total household income by age. Most countries appear to be making worrisome trade-offs between adequacy and sustainability. Of 20 countries in the index, only Australia, Canada, Chile and Sweden score well on both dimensions of “ageing preparedness.” (See video)

Solange Berstein, former Superintendent of Pensions, Chile and Chair of IOPS Technical Committee, returned to the core purpose of pensions as savings or insurance. The Bismarckian model had been adopted by many countries as insurance for the unlikely event of a worker outliving their productive capability. Now the aim was that everyone could have a retirement period. So savings had become a dominant driver – but without the full costs always being appreciated. She summarized relevant experience in Chile, including the negative impacts of affiliate portfolio choice where most people who switch performed badly.

Will Price and Robert Palacios of the World Bank Group highlighted the importance of a broad based and diversified approach to building strong pension systems. There was a need to bring together the best mix of public and private provision to get the best deal for current and future pensioners. We had learnt a lot in the past 20 years. Social pensions had a clear role but the costs needed to be controlled – following clear lessons from the health sector on how to achieve universal coverage. Instead of a blue print for all countries there was a need for rigorous review of objectives and conditions followed by carefully design and implementation to deliver objectives in a politically and technically sustainable way.

Discussion then focused on: the challenges of economies where 60-80% population make pension contributions for less than 20 weeks a year; the inability to separate informality and ageing challenges and the importance of good labor markets; the importance of improving compliance with taxation and pensions, calls for standardized reporting of unfunded pension liabilities; the scope to use pension or income transfers through social pension or social assistance. A key conclusion was that there is no one recipe for every country.

Closing this part of the Conference, Loïc Chiquier, World Bank Group noted the persistent theme of clarity on the long run objectives of sustainability, adequacy and coverage with efficiency and security. It is a tricky equation but a clear message was the need to use all parts of the toolkit. A strong political economy focus was essential to build a lasting consensus. So was a detailed focus on implementation, including mobile and IT solutions that could drive access and inclusion at low cost. The roundtable to come would continue the exchange of practical reform examples from countries across the income scale seen so far in the conference. The 7th Conference would showcase successful solutions and collaborations with partners. Better pensions for all generations would provide an important contribution to eradicate poverty and boost shared prosperity sustainably.

Day Three: April 4 Roundtable on Regulation and Supervision.

The final day of the conference moved to a seminar format to focus on improving supervision and strategy, investments and data. It focused on issues and solutions encountered in multiple examples of practical reform programs.

Introducing it, Fiona Stewart, World Bank Group and Edgar Robles, Superintendent SUPEN, Costa Rica and IOPS Vice-President, noted that low investment yields since the 2008 crisis and the administrative costs of pension funds remain big issues.

Compliance costs make systems so expensive that there are high barriers to entry and less competition. So there was a constant need to do more with less and to ensure that supervisors allocated their scarce resources effectively. Limitations of individual choice and the need to deliver pensions in all countries were also critical.

In Part 1 Michael Hafeman, Chair, Toronto Centre Insurance and Pensions Advisory Board, presented work conducted with Will Price and John Aschcroft on an Outcomes Based Assessment Framework (OBA) and the arguments for Risk Based Supervision (RBS).

The OBA approach aimed to address the issues raised in the first 2 days of the conference of how to set a clear long run objective and then build a system that brought international best practice tailored to a country’s specific circumstances. He then showed how RBS should be seen in this broader context as a way for supervisors to increase their capacity to drive change, strengthen their ability to deal with challenges and leverage off the knowledge of others. RBS was not about eliminating all compliance and equally it was not about light-touch regulation. He highlighted the range of tools to implement RBS approaches. His own institution - the Toronto Centre – worked with regulators and supervisors across the world. The Centre focused on giving participants the strategic and technical tools to implement RBS, emphasizing the importance of improving governance across the industry as a critical way to reduce risks to members and the importance of not using any tools or rules mechanistically.

In discussions that followed delegates from the Albanian, Egyptian, Hong Kong, Mexican and Turkish supervisory authorities and John Ashcroft, senior consultant, then shared experiences. These were of countries considering or already implementing RBS. Insights included the importance of capacity building and understanding that a risk-based approach is about strengthening capacity to protect members by ensuring that new risks or behaviors will be tackled as they arise. Peter Penzes from Slovakia summarized some EU work on consumer protection in which they have been playing a leading role. The member or consumer needed to have confidence in the long run products that were essential for retirement income. Further points made in the discussion included: the need for RBS to focus on reversing the damaging short-term bias of pension investment managers; the big differences in the design of RBS between countries with large numbers of heterogeneous or occupational funds and those with mandatory systems and a small number of funds; how RBS can apply to governmental pension institutions; and fraud risks in long term investment products.

Part 2 of the Roundtable considered the investment governance and benchmark strategies, introduced by Heinz Rudolph. 

He noted how the open fund model, that assumed effective competition on investment returns, had in practice resulted in a focus on the short term rather than strategic asset allocation. He explained that it was essential for pension funds to have a clear long run objective to deliver a future replacement rate target since this was what the member ultimately cared about. Benchmarks could be designed that would target this long run outcome and against which pension fund performance would be assessed rather than short term year on year changes.

Solange Berstein, Chile reviewed the investment experience in Chile. She noted that as pension funds increased in size diversified portfolios become increasingly important, as does ensuring that there is good investment governance in the best interests of affiliates. Delivering a long run outlook was essential to good retirement income provision.

Luis Mario Hernandes, Consar, Mexico, explained latest developments in their supervisory approach. They had moved away from the previous focus on value at risk (VaR). This was because the aim of controlling short-run volatility could have negative consequences for building long run portfolios. The VAR was now only used to control the use of derivatives.

Discussion then ranged over: the importance of combining a focus on governance with the use of benchmarks; educating politicians on what constitutes effective long term investment; whether the herding that would accompany benchmark portfolios is desirable; and the scope for financial education to drive better competition – on which views differed.

Part 3, moved to the important technical issues involved in recordkeeping and payment infrastructure, introduced by Oleksiy Sluchynsky, World Bank Group.

He noted that the core business is liability management whether in DB or DC – and pension funds need data to establish accurately their liabilities. He reviewed the problems associated with obtaining, validating and preserving data of the necessary quality, and the approaches adopted to address these issues. The core principle was that it is best to start with clean data than to try to clean it later.

The contrasting experiences of Tanzania, Kenya, Turkey, the UK, New Zealand, Poland and Egypt were then shared. Despite the differences there were also a surprising number of common themes. These included the need for good tax or ID numbers, ensuring employers contributions were accurately and properly paid (utilizing the tax authority can be effective), and how to deal with duplicate and dormant accounts. The UK experience in building compliance for the new auto-enrolment reforms highlighted that even if a tax authority was in theory the right organization in theory (and had been in New Zealand as explained by Darren Ryder) it was critical to assess if its current capacity and other tasks would allow it to take on an important new function. Developing countries could have an advantage over those with long standing pension systems that were originally designed for paper because they could start with a modern IT enabled approach. The experience in Kenya and India showed that much of the payments infrastructure might be able to use new mobile and IT platforms – removing the need for costly distribution networks and error prone record keeping systems. The range of the discussion on the final day mirrored one of the key messages from the first two days – with a need to consider issues of the ultimate long run outcomes and the political economy of pensions through to the essential practical issues of data, information and IT that underpin a well-built system.

Closing the Roundtable Fiona Stewart welcomed the lively interaction and many country experiences shared. She thanked everyone and wished them well in future projects that would put the discussions at the conference into practice and be material for the next Conference.