Mr. Robert Nicholson:

I will now address the queries on auto-enrolment, including those raised by the witnesses from Age Action.

On the question of the target membership and who will be enrolled, that needs to be defined by the objective of the system itself. Usually supplementary second pillar pensions are concerned with consumption smoothing, so avoiding reductions in living standards at retirement. That is generally an income-based measure to prevent those reductions. If that is the target membership, one is then in the space of the level of pre-retirement earnings of the individual. For example if the State pension is €12,500 or €13,000 per annum is it legitimate to enrol somebody in a pension who is earning less than that when they will get a higher replacement rate from the State pension than they would from their earnings? It is then balancing the question of the cost of those contributions to the individual with a limited discretionary income in terms of short-term demands as against long-term gain. The question is where is the balance to be struck in that case.

I think in the reform plan, a figure of €20,000 was given as an indicator because we know definitively at that level of income that individuals will suffer a material reduction in living standards when they retire even with the State pension, so they should be enrolled. The space below that begins to be more challenging for the reasons I have mentioned but also because, typically speaking, it would proportionately affect certain groups more, for example, women, because they work part time - and as the ESRI has highlighted a greater number of women are on the minimum wage as are people from the ethnic minorities and people with disabilities. We need to look at those issues and again no decisions have been made, but it is a fundamental for auto-enrolment.

On the questions of fund administration, privatising risk and so on, traditionally in pensions policy there is the multi-pillar system, where the State, through the social insurance system guarantees a minimum level of income and poverty avoidance in terms of the State pension and then adequacy is around what the individual needs in order to maintain his or her living standards. Again that is an earnings-related indicator. The framework as to how that is to be managed in terms of an auto-enrolment system is yet to be decided. There are major questions around a public service obligation on the State structure, as in the United Kingdom or does one move towards a structure where there is State involvement but a regulation of the number of providers or does one just roll out the existing retail structure. The key question Ireland has to answer relates to population because we have a far smaller population than the UK in terms of generating economies of scale. In the United Kingdom, there is the National Employment Savings Trust, NEST, which Deputy O'Dea referred to earlier and it has 7 million members. We estimate the population in Ireland for such a scheme would be between 500,000 and 800,000, but we absolutely believe they need to go into a small number of large, efficient pension schemes. What that small number is has to be defined - is it one, two, five, ten? We do not believe that the existing infrastructure is fit for purpose. We believe it is very difficult to provide services to lower income earners at a cost that makes it economical to do so, unless one generates scale and the impact then is on the member outcomes because they pay more for the service. We know scale creates the capacity to deliver services to these individuals.

If it is to be an earnings related system where an individual in investing to work their income to generate retirement income, there will be an element of risk. That is about getting the default options right. We know that in the UK, 90% of people are defaulted into a choice quality assured investment package and that 90% of them stay in it. The design of that package and how it should be structured is crucial. However, again, that has not been looked at in detail yet.

Tax relief is a matter for the Minister for Finance in the annual budgetary process. We talked to over 50 sectorial interest groups representing all strands on auto-enrolment. There have been two broad arguments. First, it is argued tax relief, as it stands, is more beneficial to higher earners which is unfair. Second, it is claimed that the first argument does not consider the amount higher earners contribute to the social insurance system, which is then redistributed to provide State pensions to lower earners. It also does not recognise the proportionately greater level of tax that higher earners pay and tax in drawdown. I am not taking a position on these but those are the two ends of the spectrum from our discussions.

From an auto-enrolment point of view as a policymaker, the question that has to be answered is what the policy objective is. In this case, our policy objective is to enrol middle and lower income earners into a product which will generate a reasonable retirement income. Is the financial incentive, as it is structured, suitable to ensure those people stay in the system and save? That is a question for which I do not have an answer. One must look at the potential behavioural impact of change. We need to recognise that people of modest income levels, €32,800 and above, are receiving marginal rates of tax relief on those pension contributions. Effectively, a change to that system would be a pay reduction for those individuals and may result in the opposite behaviour we want. It may encourage them to opt out of pension saving rather than opt in. It is absolutely not an easy question to answer. It is fundamental to the success of auto-enrolment and a key one for our consultation process, as well as for the review the Department of Finance is undertaking.