This article makes the following arguments:

• Pension products are complex.

• Many people make bad choices both while building up their pension savings and during drawdown.

• The policy strategy that follows is twofold: pension design should assist choice by people who wish to make choices about saving and retirement; but the pension system should work well for people who make no choice – and making no choice should be an acceptable option.

Section 1 sets out some observed facts about the behaviour of savers and section 2 why those findings are unsurprising. Section 3 considers some resulting lessons for improving pension design. Section 4 concludes.

1 Some facts

Many people make choices that do not suit themselves or their families.

Saving too little. An OECD study of Chile, France, the Netherlands, the UK and the USA found that between 29% of individuals (the USA) and 42% (Chile) have not saved enough, that is, ‘have a present value of pension income at retirement lower than the average pension incomes of current retirees’ (OECD 2014, p. 95).

Delayed choice or no choice. In Sweden workers are supposed to choose from over 800 private mutual funds. In 2011 over 98 per cent of first-time choosers made no choice and were therefore enrolled in the default fund. Notwithstanding that some new entrants make a choice later, in 2016, 45 per cent of all participants were in the default fund. A US study (Sethi-Iyengar et al. 2004) found that offering more options can reduce choice.

Making an unsuitable choice. There are many examples. Some people will choose actively-managed funds when cheaper index funds might be more appropriate. Some investors will hold an inadequately-diversified portfolio. Many investors do not understand the powerful effect of administrative charges on their accumulation, for example a 1 per cent annual management charge over a full career reduces a person’s accumulation by about 20 per cent compared with what it would be without the charge.

Choices by individuals during drawdown display parallel problems.

Annuitising too little. Buying an annuity insures the pensioner against longevity risk. Relying fully on drawing down pension savings forgoes this insurance. Behavioural economics gives insights into why in a voluntary system people do not annuitise, or do not annuitise enough – known as the ‘annuity puzzle’. Annuitising too little does not mean that requiring full annuitisation is optimal. Bequest motives and uncertainty about future expenditures, e.g. long-term care, give reasons for keeping some non-annuitised wealth. 

Making a bad choice of annuity provider. The issues are the same as the choice of fund manager during accumulation. A UK study (UK Financial Conduct Authority 2014, p. 14) estimated that, ‘80% of those purchasing an annuity from their existing pension provider would benefit from shopping around and switching’.

Drawing down too fast or too slowly. A pensioner may spend too much too soon, particularly if he or she lives longer than anticipated. Choices by an individual may take insufficient account of other family members. Or the pensioner may spend too little, either for fear of running out of money or from a reluctance to ‘spend the children’s inheritance’.

2 Some explanations

These observed outcomes are not surprising, as explained by a large literature – the subject of multiple Nobel prizes – on the economics of information and behavioural economics.

Complexity. Choices about saving and drawdown over the course of a lifetime are inherently complex because of uncertainty and because financial products are not simple. A worker making choices about saving faces multiple uncertainties:

• Future needs depend on how long a person and their family will live; by the duration of a marriage and the options after a divorce or death of a spouse; by the number, ages and experiences of children; and by medical expenses.

• Future earnings: people face uncertainty over the level and time path of future earning opportunities.

• Future asset returns are uncertain.

• Future annuity prices will be heavily influenced by future interest rates.

Limited capacity to address complexity.

Making choices requires a good understanding of financial concepts.

Many workers and pensioners end up with products that are not a good fit for themselves or their families because of low financial literacy. In a survey, 50% of Americans did not know the difference between a stock and a bond. A famous study by Lusardi and Mitchell (2014) asked three very basic questions about finance. In the USA, 35% of respondents answered all three correctly, in many OECD countries only about 25%. The implication is that many people do not understand simple interest, let alone compound interest, let alone the concept of present values which underpins sensible decisions about annuities. 

Lack of effort.

Even where an individual is financially literate, it does not follow that they will devote the necessary attention to managing their pension affairs. 

To see the extent of inertia, one has only to observe how many people fail to move money to a bank account that pays a higher interest rate.  Note that managing one’s pension arrangements is a continuing process over working life and retirement, not a one-off event like buying a car.

Behavioural biases.

A large recent literature discusses behavioural biases (for non-technical discussion, see Thaler and Sunstein 2008, Kahnemann 2011, Thaler 2015, and Valero 2019, 2020). Such biases can usefully be divided into two sorts.

Bounded rationality arises where a problem is too complex for someone to know what they should do, even when presented with the relevant information (many medical problems have this characteristic).  Manifestations include:

• Procrastination: people delay making decisions.

• Inertia: people stay where they are; in theory it should make no difference whether the pension plan is opt in or opt out – in practice, automatic enrolment leads to higher participation.

• Immobilisation: faced by complex choices, many people do nothing.

Bounded will-power arises where a person knows what he or she should do, but does not do it, the clearest example being not saving enough (or not losing weight, not getting more exercise, etc.).

A large but usually unasked question. These explanations are well known. What is rarely discussed is an equally important issue: even if none of the problems outlined above is significant, choice and regular monitoring of pension savings take time and energy, raising the question of whether a pension plan that requires such efforts by each individual is welfare-enhancing.

3 Getting it right

3.1 Implications for pension design

The preceding discussion suggests a number of pointers for good design of defined-contribution pensions.

1. Make pension saving mandatory, or use automatic enrolment, thus turning inertia to people’s advantage: once automatically enrolled, most people will stay in the plan.

2. Keep choices simple, for example offering only a few clearly differentiated funds. Offering less choice can increase participation. In complex areas, offering less choice can be a deliberate and helpful design feature.

3. Include a single default for workers who make no choice. A good default will automatically reduce exposure to risk as a worker approaches pension age.

4. Keep administrative costs low by decoupling account administration (i.e. record keeping) from fund management (i.e. decisions about the choice of assets in a pension fund).

In employer plans a further option is to design policy so that people commit now to action in the future, thus using procrastination to assist policy (see Thaler and Benartzi 2004). The essence of the arrangement is that people commit to save a given fraction of their salary, but starting only with their next pay increase.

Barr and Diamond (2017, 2018), in evidence to an Australian Inquiry, advocate a single default, a simple choice architecture within the default, one pension account per person (to avoid ‘lost accounts’ and erosion of small accumulations via administrative charges), centralised record keeping, and a level playing field for competition between a centrally-run default and multiple private providers.

Having a single default avoids the problem of diverse returns. provides a simpler picture for a worker of the consequences of not making a choice, and reduces costs by avoiding complexity in policing multiple defaults.

3.2 NEST Pensions

The UK offers many examples of bad pension design. An important exception is NEST pensions ((http://www.nestpensions.org.uk/) based on a design suggested by the Turner Commission (UK Pensions Commission 2005), which took heavy account of the findings of behavioural economics.

The elements of the plan exactly match the pointers for good design outlined above.

Automatic enrolment. Since 2012, on a phased basis, employers without any other approved pension plan have been required to enrol workers in NEST, the idea being that employers should have access to a high quality, low charge path through which to discharge their obligations.

A default. The default is a target-date fund, using state pension age as the target. The fund has with three phases:

• A foundation phase (the first 5 or so years) is a novelty in pension design. Losses in early years are profoundly discouraging, so the strategy during this phase seeks to avoid any reduction in the value of the nascent accumulation. 

• A growth phase: NEST aims at a long-run average annual net real return of 3 per cent. 

• A consolidation phase starts to de-risk the portfolio as a worker approaches state pension age.

 Limited choice. Alongside the default, a worker has the option to choose from five funds: a  higher risk fund, i.e. potentially higher growth; a lower risk fund; an ethical fund; a Sharia fund; and a  pre-retirement fund. And a worker can choose to have part of his/her pension savings in the default fund and part in (say) the higher risk fund.

In addition, a worker can choose a target date different from the default. A worker may choose a later date, either because of an intention to delay the start of benefit or to slow down the move from equities to bonds, thus choosing to take more risk in the hope of higher returns. Or a worker could choose an earlier date, thus de-risking sooner.

Centralised account administration. NEST maintains all individual records.

Wholesale fund management. NEST decides in-house on overall exposure to building block funds and asset classes, manages some parts of the portfolio itself, outsources other parts to the private sector, and publishes quarterly updates (see NEST 2020).

The NEST approach respects lessons from the economics of information and behavioural economics

• The design simplifies choice for workers, including working well for a worker who makes no choice.

• It keeps administrative costs low.

• It locates competition in the right place. The important purpose of fund management is to allocate savings to productive investment. That task is carried out more effectively with competition in the right place, i.e. between fund managers and NEST plan managers, hence with both sides of the market well-informed, rather than between well-informed fund managers and less well-informed individual workers.

4 Conclusion

Pension design cannot simply rely on markets and individual choice.

Pensions are more like pharmaceutical drugs than like smartphones – the outcomes are highly important and the choices complex and difficult to get right. The findings from the economics of information and behavioural economics justify the argument that good pension design will deliberately include a choice architecture that limits choice for the great majority of workers and pensioners.

References

Barr, Nicholas and Diamond, Peter (2010), Pension Reform: A Short Guide, New York and Oxford: Oxford University Press.

Barr, Nicholas and Diamond, Peter, (2017), ‘Designing a default structure: Submission to the Inquiry into Superannuation: Assessing Efficiency and Competitiveness’, Australia Productivity Commission, September.  

Barr, Nicholas and Diamond, Peter, (2018), ‘Response to Superannuation: Assessing Efficiency and Competitiveness: Productivity Commission Draft Report’, Australia Productivity Commission, Submission DR154, July.  

Kahnemann, Daniel (2011), Thinking, Fast and Slow, New York: Farrar, Straus and Giroux.

Lusardi, Annamaria, and Olivia S. Mitchell (2014), ‘The Economic Importance of Financial Literacy: Theory and Evidence.’ Journal of Economic Literature, 52(1): 5-44.

NEST (2020), NEST quarterly investment report, At end December 2020, https://www.nestpensions.org.uk/schemeweb/dam/nestlibrary/61537-QIR-Q4-2020-210223.pdf 

OECD (2014), Pensions Outlook 2014, Paris: OECD.

Sethi-Iyengar S., Huberman, G., & Jiang, W. (2004), ‘How Much Choice is Too Much? Contributions to 401(k) Retirement Plans’, in Mitchell, O. S. & Utkus, S. (Eds.), Pension Design and Structure: New Lessons from Behavioral Finance (83-97). Oxford: Oxford University Press.

Thaler, Richard H. (2015), Misbehaving: The making of behavioural economics, W. W. Norton.

Thaler, Richard H., and Benartzi, Shlomo (2004), ‘Save more tomorrow: Using behavioral economics to increase employee saving’, Journal of Political Economy 112 (1, part 2): pp. 164-87.

Thaler, Richard H. and Sunstein, Cass R. (2008), Nudge: Improving Decisions About Health, Wealth, and Happiness. Yale University Press.

UK Financial Conduct Authority (2014), Thematic Review of Annuities, TR14/2, London: Financial Conduct Authority, February, http://www.fca.org.uk/news/tr14-02-thematic-review-of-annuities

UK Pensions Commission (2005), A new pension settlement for the twenty-first century: Second report of the Pensions Commission. London: The Stationery Office. www.webarchive.org.uk/pan/16806/20070802/www.pensionscommission.org.uk/publications/2005/annrep/annrep-index.html

Valero, D. (2019). ‘Previsión social complementaria: un enfoque conductual’, in Herce, J.A. (Ed), Pensiones del futuro, Instituto Santa Lucía. Madrid

Valero, D. (2020). ‘Economía del comportamiento en el ahorro a jubilación’, in Rodríguez-Pardo, J.M. (Ed) El envejecimiento como riesgo empresarial. Wolters Kluwer.