Piggy bank

When should individuals save for retirement? Predictions from an economic model of household saving behaviour

Published on 10 May 2021

We use a life-cycle economic model to illustrate that there are good reasons for saving rates not to be constant over working life.

Most individuals need to save privately for retirement if they are to maintain their living standards when they stop working. There has been lots of research and discussion on how much individuals need to save, and how this compares with the saving being encouraged through automatic enrolment into workplace pensions. However, there has been little discussion of when individuals should save for retirement and the appropriateness of a single default contribution rate for all.

In this briefing note, we use a life-cycle economic model to illustrate that there are good reasons for saving rates not to be constant over working life, due to predictable factors that change with age. The model is a simple approximation to real life, in that individuals face little uncertainty and can only save in one asset that has a known rate of return. Individuals choose how much to spend each year, and how much to save, with the objective of smoothing their living standards over their life cycles. While necessarily simple to be computationally tractable, this model yields important conclusions with implications for the design of real-world policies.

Key findings

  • Most individuals expect some earnings growth over their working lives. If an individual is aiming to smooth their spending over their lifetime, then they should save a greater proportion of their earnings for retirement at later ages when earnings are higher, rather than saving at a constant rate throughout working life.  

  • Households with children are typically assumed to require higher spending to achieve the same standards of living as those without. Given this, most parents aiming to smooth their living standards over their lifetime should save relatively more for retirement before their children arrive and/or after they have left home.

  • Many recent graduates hold student loans that will be written off after a certain period of time (30 years after graduation for those entering higher education from 2012). Graduates aiming to smooth their living standards over time should increase their pension saving by the amount of their previous loan repayments when loans are written off.

  • Employer pension contributions that are only made if the employee also contributes incentivise individuals to contribute throughout working life, even in years when earnings are relatively low or there are children in the household (or if, for some other reason, living expenses are greater). However, individuals are likely to want to contribute only the minimum required to receive the employer contribution for the first half of working life. When earnings increase and/or children leave home, they should then markedly increase their saving rate for retirement.

  • This profile of the appropriate saving rate over working life in the presence of contingent employer pension contributions – flat at the minimum required employee contribution, and then increasing markedly when children leave home – is robust to a range of plausible assumptions about the rate of return on saving.

  • Uncertainty over the future path of earnings deters individuals from leaving all their retirement saving to a short period of time at the end of working life. In the presence of uncertainty about earnings and employment, individuals should save more at younger ages, in particular in years when earnings are high. However, the general pattern remains – that is, many would be expected to save the minimum amount early in working life, and then increase their saving rate substantially when children leave home.