Mr. Robert Nicholson:

I thank the committee. I will try to bring the committee through the proposals in a little depth. That is needed because if automatic enrolment is introduced, it will fundamentally reform the provision of supplementary pensions in Ireland. That will not just be for now but for generations to come and for the broader population. It is a significant reform, probably the most fundamental reform of pensions policy in a generation or two. I will lay out, first, the elements of automatic enrolment already confirmed by the Government. As Deputy Joan Collins just asked, the Government has committed to beginning automatic enrolment in 2022. It is intended that it will supplement the State pension and complement existing private provision. It is a different population of individuals and employees being targeted, those that have not saved previously. It will be an earnings-related workplace system where the employee can choose to opt out. They will be automatically enrolled but they choose if they do not want to avail of the benefit being provided to them. It will be a defined contribution model with personal accounts. People will own their personal accounts.

For the first time, employees, employers and the State will each be required to contribute. That will be cast under legislation. There is no mandatory requirement at the moment for employers to contribute to supplementary pensions. Broadly, the system is intended to facilitate choice but not to demand member choice. That is particular to the type of employees we are talking about and we will talk more about that as we go on. I will turn now to the elements of automatic enrolment consultation that are up for discussion. These are the broad operational structure and governance of the system. We lay out a proposal in the straw man which we think is an appropriate line of flight. The target membership is another element consisting of who will be enrolled and why will they be enrolled. On the employee and employer contribution rates, we make a proposal but it is up for discussion in respect of the impact on employers, employees and the economy more generally. I refer also to the shape and colour of the State financial incentive. What is the most appropriate incentive for the type of individual we are talking about?

Also considered are the number of providers, the type of savings options being provided and the default options of what happens where an individual does not make choices, the conditionality of member opt-out, how wide or thin should that be, the conditions on re-enrolment if a person opts out. Should he or she be re-enrolled? Should it be possible to take periods out of saving if there are more demands on a person's limited income? Finally, there are the conditions on income drawdown. What sort of shape should the income provide over the course of the person's retirement years? That is what is covered in the straw man.

I will open with the policy side of why automatic enrolment is needed. We have a multi-pillar pension system like most developed economies where the pillars are intended to share and diversify risk between the State, employers and employees. The State pension, which we will call pillar one, is intended to deliver on a narrow interpretation of adequacy. It is to provide a minimum level of income so that individuals avoid poverty in retirement. At the moment, 1.5% of those over the age of 66 are in consistent poverty, so resource are, rightly, targeted at that group. Beyond that, the State pension largely delivers on its objective, as it stands, of avoiding poverty. As I said, it is 1.5% for those aged over 66 compared to an average of 8.3% for the rest of the population in consistent poverty.

There are challenges over the long term for the sustainability of the State pension. Our demographics highlight that the number of workers for the employed population was going to drop from about 4.9: 1 to 2.3:1 in the next 40 years or so. The numbers contributing to the system, therefore, compared to the numbers drawing from it creates funding challenges. The cost of State pensions is going to go up by approximately €1 billion every five years, there will be an extra 17,000 pensioners every year and the number of pensioners will double from approximately 600,000 to 1.2 million over that 40 year time horizon. Our longevity in retirement means pressure to maintain the State pension, as is, will increase rather than decrease. The capacity to go beyond that minimum level of provision to avoid poverty will come under increasing pressure instead of less pressure. There will be less capacity to move, and even to stand still.

After pillar one largely achieves its first objective, we then move to supplementary pensions. In supplementary pensions, the Government incentivises people to save for a level of income adequate to themselves. What is adequate to themselves is very much an open question. We talk about personal adequacy, that is what people need in retirement so as not to suffer a reduction in living standards. That is what most people target. We do not want to hit retirement and not have enough income over those years of retirement to do that. In regard to that adequacy, only 35% of the working population actually have supplementary pensions. The rest will rely entirely on the State pension and secondary benefits. As a result, the OECD, in reviewing the Irish pension system, has said that the single biggest recommendation for reform is to introduce some sort of mandatory earnings-related retirement savings system. We are one of only two of 37 OECD countries without a mandatory earnings-related element to our supplementary retirement saving. The other is New Zealand, which has a fairly sophisticated backup.

The Citizens' Assembly in June 2018 voted 87% in favour of some form of mandatory system being introduced. Over the past year and a half, this Government, as did many previous Governments, indicated a desire to move towards an increase in supplementary pensions. We have, broadly speaking, three choices. Individuals save more through tax or personal savings, they work longer or they suffer a reduction in living standards in retirement. For the past 25 years in Ireland, the provision of supplementary pensions has tried to increase coverage of people with supplementary pensions. It has not managed to do that and it has remained more or less constant at the level it is now for the past 25 years, despite financial incentives. The adequacy of those retirement benefits are also in question regarding the goal of trying to maintain living standards. There has, in effect, been a market failure. We have not done what we intended to do through public policy. Voluntarism has not worked. The Government decided with that - and, as I said, it has been a long-standing goal - to move towards some form of quasi or full mandatory contribution. That is to address the needs of the roughly 860,000 employees who do not have pensions' coverage and about 240,000 self-employed and begin to encourage them to save more. That is the broad background.

Moving to a system of mandatory contribution, where people are compelled to save to whatever extent we decide upon, we need to decide what the criteria are for telling people they should be in the system. When we are taking contributions from people's pay and putting it into a personal account for them, we need a good reason to explain to them why that needs to happen. It is a system of mandatory contributions.

We have carried out extensive research using CSO data on household spending. We have analysed household spending in the period up to retirement and if one looks at the green line, that is the replacement rate for incomes provided by the State pension. At the lower end, the State pension, at approximately €13,000 provides a high replacement rate, in and around 80% for somebody on €60,000. As one moves up the income scale that reduces. What grows is the savings gap that individuals have, to be able to achieve that goal of not having a reduction in their living standards at retirement. That population in between those two lines, are broadly speaking who should be in automatic enrolment, AE, if one agrees with policy objective of trying to avoid a reduction of living standards in retirement.

Trying to identify what exactly what one should have in retirement is a notoriously controversial exercise, as there are all sorts of demands like tax coming in, childcare costs, housing costs and whether one saves for a pension or not - costs that will not arise during retirement. For the generality of the population however we are confident that this replacement rate assessment is suitable.

We carried out this exercise independent of work that was done in the UK by the pensions commission which set its target group for their automatic enrolment system, and it was broadly similar in terms of the outcomes. We are confident enough that the much-vaunted 50% replacement rate does not suit for the generality of the population. Lower income earners need higher replacement rates, higher income earners need lower replacement rates. That revolves around the element of discretionary income one has in retirement.

Moving on to the broad population of automatic enrolment savers, we have to highlight that it is a different group from current savers. It generally comprises lesser incomes and different cohorts of employee types by employment sector. If one looks at the chart one can see that the higher the income represented by the blue line, the more likely it is that the person has a pension. The red line represents those who do not have a pension. The lower income earners, to a greater degree, do not have pension provision. They are in the typical sectors that one might suspect - accommodation and food services; retail and administration; and, support services. It also includes categories of employment where people frequently change jobs and have periods out of the workforce and so on. They are also the sectors that are less commercially viable for large providers to deliver pensions to compared to higher earners.

With that cohort we have what we call the "pot follows member" approach, which I will talk a little bit about later on, but that is what is informing that approach to frequent job changers and people who take frequent time out of the workforce. We do not want people with multiple pots. The "pot follows member" approach, which we have in the straw man, is based on that.

Moving on to the employer type, this is a very different population from current savers. What one sees with the green line is occupational pensions coverage or personal accounts. The larger the employer, the more likely it is that one is going to have supplementary savings set aside for oneself. The smaller the type of employer, conversely, the less likely it is that one is going to have cover. It is also true that smaller employers tend not to have the same level of support in terms of financial experience, human resources, industrial relations and administrative capacity to support individuals. That is really important. A different type of employer will have a different type of employee.

What are we trying to achieve with automatic enrolment? If we reflect on what I have just discussed, it concerns mainly low and medium income earners with smaller employers to adopt a system not just now, but that will exist in the coming decades, which will fundamentally reshape Irish pension provision, as it stands. We need to address the challenges that are out there for this particular population. We need the right product for the population. We need to remove true barriers to saving that have prevented us from getting into the type of coverage adequacy that we have searching for. We need to ensure that we maintain the competition in choice but address that inertia that stops people from saving for retirement. It is that old chestnut "I will get a pension or I will change my bank account tomorrow". It is a behavioural economics policy reform which relies effectively upon people's inertia.

We also have a population that does not know too much about pensions, so we need to make choices for them, but also at the same time to allow them to make their own choices if they so choose. We also need to build trust and confidence in the system. We are all aware that there is a limited trust and confidence in the wider pension system that exists.

For the first time, employers, employees and the State will be compelled to make a contribution, at least initially. Throughout automatic enrolment, however, the employee will retain the choice. If they choose not to avail of it, they can opt out. The reason for this - linking back to some of our conversations with likely members of an AE system and the wider population - is that they do not to want it perceived as a tax or as a mandatory system. They want to have the option to have the control in their own hands. Frequently they will not use that control but they want to ostensibly have it.

The reasons we want to fundamentally reform the existing system for AE are due to the following challenges facing Irish employees preparing for retirement: Saving late and inconsistently or not at all - due to shortsightedness or myopia where the here and now always trumps the long term, even if the long term becomes more important at some point; and, comparatively high fees where, although not in all cases, there is nonetheless an expensive fragmentation in the Irish model of pension provision, generally speaking. Using a figure I am frequently criticised for quoting, we have 1% of the population of the EU and 50% of the pension schemes in the EU. Even when single member schemes are stripped out, the average Irish defined contribution scheme has 24 members. We are looking to create pension schemes with 100,000 members, and the economies of scale that come with that. We are an outlier, as it stands, as to how we provide for supplementary pensions. Not utilising or maximising State tax relief incentives, provided at 20% and 40% at the moment, which is a very generous top-up to anybody's pension, is another feature of the current position where some 65% of the country's population do not use it at all. Some of these may not pay tax but a large body of people who do, do not utilise it. Making poor uninformed information decisions that can be based on discomfort with the pensions world, but with an element of being blinded by choice, is also a feature. There is such a proliferation of products out there that individuals often feel uncomfortable making decisions. There is also a lack of confidence and trust that can lead to a disengagement; that permeates the whole system at some level. There needs to be an improvement in our confidence and trust levels on pensions. Finally, we need to find suitable products.

We know that the population of savers for automatic enrolment are to a greater degree lower earners, and the charges and associated costs that will come with the commercial process of selling to them will impact on their savings. We need to address that. How would automatic enrolment do that? It makes savings easier and automatic and all of the international evidence - we have looked extensively at every country who has introduced similar systems - suggests that the most effective and efficient vehicle for improving retirement savings, and one that leads to better outcomes in the round for the individual, are these types of systems. It makes saving easy and automatic. One has default products and auto-pilots that make good decisions for individuals if they choose not to make them themselves. One can communicate better because one has a captured population. One can improve understanding. One can introduce easy-to-use IT systems and give a greater level of customer service. In terms of scale, one can see reduced member charges due to larger numbers. All of our evidence incontrovertibly highlights, all things remaining equal, that scale improves governance, administration costs and investment returns to the individual. Using a single figure which we quote, 0.5% of an individual's assets would be the charge for the system. That would see the ultimate pot of an individual's savings over their lifetime reduced by about 10%. PRSI in Ireland, for example, is approximately 1%, which is effectively double the cost of what we are bringing forward in our proposal. Many smaller schemes would have higher charges. A really crucial element of the reform is to drive down charges.

Looking at the population of frequent changers who do not understand pensions too well and do not want multiple benefit statements and multiple accounts and who have periods outside of savings, in every country we have talked to who have not introduced a "pot follows member" approach - where the pot is taken by the person from job to job - they have said the single biggest regret they have is not introducing that as an element of the system. It is a key for us in targeting this population. We can imagine a situation where we have on average 11 jobs per lifetime; if one has a different system and a different fund with every employer one has, one has 11 pots to keep an eye on. The idea then is to amalgamate these and to carry one's pot onward to further employments.

The saving persistency feature of individuals also improves. It suits job changers. It prevents high administration costs on small pots, which eat away at the individual's savings.

I have already spoken about professionally selected investment choices. Effectively, this is about introducing bespoke options for individuals in order that they do not have to make decisions for themselves. We have suggested that the system should include some sort of fiduciary responsibility on the providers - I will speak about them in a while - in order to ensure that there is a legal duty on them to act in the best interests of members at all times.

Broadly speaking, there are two system in operation in Ireland. Contract systems are delivered by banks and insurance providers, with the contract directly between two parties, and then there are trust-based systems, with a board of trustees legally obliged to make all decisions in the best interests of a party. If it does not do so, legal recourse is available. We have suggested a kind of hybrid of the two systems, and we will explain why. At the bottom of it, there should always be a fiduciary duty on trustees. Keeping those principles in mind and the idea that there has been market failure to date in terms of what we are trying to achieve on pensions coverage inadequacy, what could we suggest? When we introduce "mandation" or statutory enrolment, we are saying that it cannot be left to market forces alone and that we must mitigate the failures that have existed. We need to address the complexities and inefficiencies in the system. We are suggesting a central processing agency, CPA, that would effectively be an independent and arm's-length State body to introduce a framework to set standards and harness the expertise that exists in the private sector. It is not an expertise that exists in the public sector or more broadly. This CPA would be statutorily independent; it would not be a regulator. It might be regulated, along with the schemes it would provide, and its role would effectively be to act on behalf of consumers generally.

The proposed agency would tender for registered providers which could deliver services over a central hub to employees generally, providing a limited choice to allow an individual to compare one against another and choose a fund. Where an individual does not choose a fund, he or she would be put through a kind of a carousel to an appropriate default fund designed for the individual. Internationally, we know that somewhere between 90% and 99% of those enrolled will not make a choice. In the UK, 11 million employees have been enrolled, with 6 million in its large NEST fund, with 90% of those not making a choice and being defaulted into a fund. It is not a bad thing but we would always facilitate choice where the individual decides he or she wants to make the decision.

This CPA would act as a central hub. It is would be digital first, utilising developing financial technologies and providing online platforms where I could make comparisons and which would charge in a consistent and comparable way so I could make a choice. It would also provide basic required information for the employer and employees, acting as a contribution clearing house. We would use payroll systems to facilitate contributions from payroll to the CPA and on to the provider. It would be similar to a system in New Zealand and also used in Sweden and a couple of other countries. It is basically the harnessing and managing of the service providers in the private sector.

The CPA might make saving easy and automatic for employees, effectively acting as an agent on behalf of members. Its personnel would be highly professionalised and skilled, with an understanding of the pensions environment. The model should, in itself, engineer an improved system governance, administration costs, communication and investment outcomes by building in this scale. It would limit the number of registered providers through a tender process and also provide choice and competition. It would deliver economies of scale to bring 0.5% charges at the absolute maximum, meaning the charges effectively do not exist for the generality of employees in the market now. As already stated, there would be a default position when an employee is unwilling or unable to choose.

The pot-follows-the-member capacity is critical and would happen through the likes of a personal public service number, MyGovID or some sort of unique identifier that would allow an individual working in, for example, the catering sector to be in a scheme. Typically, one waits six months to go into a pension scheme because of the administrative burden. With frequent job changes, that can lead to a long time out of the savings process. If we use this structure, I could work for Burger King for six months and on day one my employer would make a payroll deduction to my pension account. I could move to McDonald's seven months later and start exactly the same thing without any breach in my savings pattern. I could go to another retailer six months after that and do the same thing. We are addressing a fundamental shortfall that currently exists in terms of savings persistence. The limited number of providers would help individuals make choices and reduce the kind of choice paralysis that exists when people are blinded or confused by the pensions market.

We have taken what might be considered a novel approach in removing the employer from the relationship. Typically, there is a tripartite relationship in pensions provision at an industry level, which equates to the employer, provider and the individual. We are saying we will remove the employer from that process to a large degree. I will explain why in a moment. We have spoken to all the employer groups and received feedback over the past year and a half or so. Employers legitimately highlight that there is a concern around cost base, with direct costs because of the new contribution that needs to be made, and often there is an indirect cost. That is certainly the cost currently with respect to investment and scheme advice, administration time and cost involved with setting people up, changes to the payroll system and so forth.

We spoke to employers at the lower level who do not currently provide pensions and they argued they have no human resources, industrial relations or other administration experience in this regard. In some circumstances, there is no payroll service and the parties involved do not understand pensions. They do not want to make choices for individuals working for them when they have as little knowledge as they do about pensions. They do not want workers coming to them in a number of years saying the pension that was chosen did not perform and it is the responsibility of the employer. Employers want support in that space and they are also concerned about the wider economic impact, as the large-scale introduction of automatic enrolment would have an impact on the economy more generally in the short term. It would have a more beneficial effect over the long term but making long-term decisions can be difficult.

There are 249,000 enterprises in Ireland and 1.4 million employees in this area. Of these, 93% are in the micro sector and have between one and ten employees. An earlier slide indicated that the largest gap in pension provision is among those employers. The structure we have is built around accommodating that. We intend to minimise the administrative burden wherever possible on the employer, and that brings in the CPA, which will take responsibility and remove the employer from some scheme responsibilities. We will use scale and technology to reduce service costs for these employers. We will have to introduce statutory duty, and it will the responsibility of the employer to enroll employees and make whatever contributions are considered appropriate. To a large extent, that is where the significant burden ends. Contributions would be made via payroll to registered providers using a unique identifier. There would absolutely have to be strong regulatory supports both from the CPA and, in all likelihood, from the Pensions Authority, as regulator. The contributions made would be deductible by the employer for corporation tax purposes. We know, through evidence building elsewhere - the other countries that have introduced this - that it is crucial all the way through to ensure we continue to engage with employers and understand their difficulties. They will have challenges, particularly small employers. Overall, the critical test is if it is easy to operate for employers. We hope our design proposal will help to achieve that.

I have a quick word on registered providers. We want to ensure there is a fiduciary duty to act in the best interests of members. This will provide an opportunity for providers with scale, and it will have to involve large-scale providers currently in the market or an amalgamation of those providers. Internationally, it is frequently the case that a consortium decides to get involved in the space to build service. These would need the capacity and expertise and we see four as the right number, approximately. We have looked at master trusts or large schemes internationally. To deliver the type of scale we are talking about while maintaining competition, 100,000 members per scheme is an appropriate number to start with. That give us approximately four schemes. We are open to it being either side of that but it should be close to that number.

We can look at some schemes in the UK such as, for example, NEST and The People's Pension, which have 6 million and 2 million members, respectively. This is a really big scale, and it is crucial that I get that point across. By international standards, we are an outlier and to scale is the message we learned from everywhere. It is about increasing scheme sizes. It is a fundamental intent of the pensions regulator as it stands on the same basis as I suggest.

If we achieve that scale, account administration, investment management and so on are made easier and more effective. In terms of choice for the individual, each registered provider could offer three choices and the parameters would be set by the proposed CPA in order that an individual could look at a provider and see low, medium or higher-risk options. We have set the number of options at a low level due to what is known as choice paralysis, where any more choice is just too much for individuals and they do not make choices. The strategy is to keep the number and the costs low but have a default option when the individual does not make a choice.

I will now deal with the policy side - who will be in the scheme, why, how much will it cost and so on. In regard to the target membership, under the straw man proposal, we have said that employees aged between 23 years and 60 years and earning €20,000 plus should be in the system. That goes back to the policy objective of not having a reduction in living standards at retirement. We definitively believe that those in the group in question will suffer a reduction in their living standards if they do not save for retirement. They are not currently doing so and, as a result, they should be in. We also need to assess membership against current expenditure. There would be many people who might suffer a reduction in their living standards at retirement and who would also save but who do not have the capacity to do so at present. How does one find a sweet swap between those two things, the objective but also short-term affordability? We also need to ensure that we do not enroll the wrong type of people and then have mass opt-outs early on. That might have a contagion effect across the system and encourage others to leave. We must ensure that we have enrolled the right people in the scheme.

We picked the age of 23 years based on the fact that it allows for 45 years of savings. Some 52% of people below the age of 23 years are in education or in low-earning jobs. There is frequent job churn among those below the age of 22. Job churn - people moving from job to job - tends to decrease by the age of 23. Based on the administrative burden and so on, we have stated that 23 is the appropriate age. On the basis of the contributions system as it stands, that group will then be able to achieve the goal relating to the system. At present, there are approximately 410,000 employees in the cohort aged between 23 and 60 and earning salaries of €20,000 plus. That is the starting-day population. It would also prevent over-saving and limit the level of economic opt-out and shock.

What we have said for other employees is that they would be able to opt in if they do not fall into that group. That opt-in number, namely, those earning less than €20,000 or under 23 years or over 60 years, is approximately 450,000. Those who are on salaries that are below the State pension number in the region of 255,000. Can members imagine enrolling people who, over the long term, are earning less than the State pension and informing that they will be compelled to save from very small incomes now in order to try to achieve incomes in retirement that will be supplementary to the State pension but that said pension will give them more than they are earning now? An individual would ask why he or she should save from such a limited income when he or she knows that the State pension will provide more later on. This is the reason those in that low-earnings group are not included.

We also have to look at the self-employed. There are 240,000 self-employed people who do not save for retirement. We have looked all over the world and we have failed to identify an automatic enrolment system which could, as a starting-day operation, accommodate the needs of the self-employed population. Typically, the self-employed see their businesses as their pensions. They try to keep capital as a backstop for their businesses and they do not, on the whole, like to be interfered with - obviously, we would say supported - to the same extent as employees. We have stated that the self-employed should be able to opt in rather than being automatically enrolled. This is on the basis that a successful business is not always achievable for a self-employed person and that when he or she retires, his or her capital is, in terms of experience, himself or herself. That capital cannot be sold on. In the context of the straw man proposal, we have asked if there are any suggestions as to how the system could be better designed to accommodate the self-employed.

We also need to carry out a further analysis of the population - there is a significant caveat in this regard- in the context of gender impact, the types of savers we have in terms of non-nationals and where they are on the income scale, carers and the self employed. There are particular impacts on certain groups that we need to have a better understanding of as it stands in order to allow us to make better decisions.

We have stated that the employer and employee contribution rate would start in 2022 at a low base of 1% of up to €75,000 and auto-escalate on an annual basis up to 6% of a maximum in 2027. That would be a statutory obligation on employers to match the individual contributions. Individuals could opt out. If a person opts out, his or her employer would not have to contribute. If, however, he or she remains in the scheme, the employer would have to match his or her contribution all the way up to 6%. Fundamentally, that would be a new development in the Irish context. These are ambitious contribution rates.

We were asked to develop a system to allow individuals not to have to undergo a reduction in living standards in retirement. We have laid out those ambitious contribution rates in order to ask whether, from a public point of view, it would be reasonable in ten years' time - in 2027 or 2028 - to have the type of contribution rates we know are required in order to deliver meet the objective that has been set. Again, that matter is open for discussion.

In terms of a State bonus, we are looking at a different population of automatic enrolment employees - many individuals who may be in the system but who do not pay tax and who, we know, do not understand the value of tax relief. We have stated that a State saver bonus - as a matching contribution - should be considered for these people. Under such a scheme, every €3 an individual saves would be matched by a contribution of €1 by the State. It would be a like-for-like contribution up to a prescribed cap. That is fundamentally different from the existing system and there has been a lot of conversation in respect it. However, we state that the automatic enrolment population is different from current savers. All of those products would, as Mr. Tim Duggan mentioned, be underpinned at the outset by an economic impact analysis. We know that an increase in contribution rate for this many people would have an economic impact in terms of GDP, business expansion and employment. We are working with the ESRI to model that under various scenarios. That will obviously inform Government decisions. What we would absolutely say is that automatic enrolment straw man proposal must set realistic expectations from the outset. The lower the contribution, the lower the outcome for individuals. We must match ambition and temper it with realism in terms of what can be achieved.

I have gone through the investment options, namely, four providers and three standard fund options. We have stated publicly that some of the consultation forums that the Department is by no means a centre for pensions investment specialism. That work will be done later on in the context of deciding how these funds should be structured by low, moderate and medium risk. Typically speaking, default funds internationally operate on the basis of the earlier one is in one's working life, the more one invests in equities and risk-based assets and the closer one gets to retirement, the more the risk is moderated. We have stated clearly that it is not necessarily a second-order issue. It is an issue that will be decided later on but, basically, those forms will need to examine the risk-reward profile. They will be obliged to ask "What is the appropriate level of benefit we are targeting?", "What is the competitive cost for delivering a fund?" and "What is the acceptable level of risk?" There is always going to be a trade-off in that regard because there is an element of risk in direct-contribution retirement saving.

What we know from the experience in the United Kingdom is that there are 6 million members in the NEST fund and that 37% of them wanted no risk whatsoever with their retirement savings. What we need to do is understand that and accept that individuals have free choice. However, we must also highlight that risk is two-ended - one may have safety in the fact that the money one puts into a retirement savings pot is risk free but that attaches itself to risk at the far end. In other words, when a person hits retirement, his or her money will not have worked for him or her and he or she will have far less to see him or her through the 20 years of retirement he or she had anticipated. What is the appropriate risk-reward demand profile for individuals? As I have stated, 90% of people will go into the default scheme. If, therefore, the State is involved to the extent we are suggesting, it will be absolutely critical that the default scheme is designed correctly.

I will now deal with opt-outs and re-enrolment. This is crucial in terms of the window to opt out for individuals. Due to the fact that we are the last in class to introduce this system, we have benefited because we have gained from the lessons learned by other countries. We know, for example, that opt-out windows which are of short duration and which are delayed a little in order to allow individuals see their statements and the benefits to be gained prevent knee-jerk opt-outs. If I have put in €100 and I get a statement which shows that my employer has put in a €100 and that the State has put in €33, I can see that I have €233 as opposed to €100. When people get that sense of understanding, it increases the rates of retention in the scheme so the number of opt-outs is low. In the United Kingdom, there are 11 million people enrolled and over 1 million employers involved. Some 90% of those, have stayed in the system. These are people who are saving now and who would not have done so previously.

Within the automatic enrolment, AE, straw man, the opt out can be used in the seventh or eighth month after someone gets the initial statement and sees the benefit. They can then opt out. If they do opt out, they get back their benefits. We also say that someone would then be re-enrolled every three years. Looking internationally, we have seen that people make temporary choices due to short-term demands. The environment will change. AE acceptability will broaden, there will be a peer group acceptance of AE and people will decide they need to be in because everyone else is in. We also know that of the 10% in the UK who opted out and were re-enrolled after three years, more than 50% stayed in so, effectively, it is worth it. In terms of savings suspension periods, some systems abroad look at allowing individuals periods where they do not save. For example, people may have child care or healthcare costs or unforeseen costs that make them want to stop for six months or whatever period of time is desired and then restart. We say within the document that we would like to hear views on that. What we need to be very careful about is compromising our goals in terms of allowing people opt out of the system who then through inertia do not go back in. We say that if they are considered, there should be a nudge back into the system but we should be also very careful with regard to them. In New Zealand, which allows long periods of savings suspension of up to five years, 40% of its AE-enrolled are on periods of savings suspension so it can lead to unintended behavioural changes; our watchword on it is caution.

With regard to benefits and the pay-out phase, one of the key international lessons we have learned is that our objective is around an adequate and sustainable pension so that a person has a payment over a period of their entire retirement - a traditional pensions idea. This is why our policy objective contains a provision of income for the duration of the retirement years. What sort of structure could provide that? The international lesson based on every system we have looked at is that we should make sure we get the accumulation stage, as they call it, where someone is drawing their income right from the outset and that it is coherent with the accumulation - the savings phase. It has been a major problem in other jurisdictions. It rolls off the tongue easily but it is a really challenging thing to do. This will require us to look again at the options that are available to people in retirement. At the moment, we have a lump sum, an improved retirement fund a person can draw down at their own behest and annuities. We know now that the majority of savers outside defined benefit schemes - defined contribution savers - are opting for approved retirement funds rather than annuities because annuities are costly. However, approved retirement funds bring investment risk and other risks, most particularly longevity risk, for example, if I spend all money before I hit my most vulnerable years. Someone might hit 75 or 80, have increased healthcare costs or other costs, may be less able cognitively to make decisions about their financial future and bomb out their money. What sort of structure should we have to provide security on that income stream over the long term? Internationally, it would involve things like deferred annuities where a person purchases an annuity that they draw down only when they are 80. This is much more cost competitive or cost efficient and the person knows that when they hit the most vulnerable or a more vulnerable stage in their life, they have an income backstop. In short, there is a lot of work involved to do that. We have largely just invited views on it. The Department of Finance is undertaking a review in terms of the provision of approved retirement funds, the degree to which they are fit for purpose and whether they should be changed. Hopefully, this will inform our own processes.

To a degree, that is it. That is a quick run-through. We could speak forever about almost any element of that straw man. It is a policy proposal. As Mr. Duggan said, we wanted to anchor the debate around themes we know are crucial for the system. We do not believe it is perfect but we do believe it is approximately the right line for a system such as this. Members can see the timelines in terms of our discussions. The consultation on the straw man will close on 4 November. We are travelling the country to deliver regional consultation fora and have had all sorts of other meetings and briefings with interest groups, representative groups, employers, trade unions and so on so a large degree of consultation is taking place. There is a lot of work to do once we finish that.