Numeracy Among England''s Over-Fifties: The Links With Retirement Saving And Well-Being
01 Nov 2010
More numerate people among the over-50 age group in England tend to have substantially more liquid financial wealth than less numerate people. That is one of the central findings of research by Professor James Banks, Cormac O''Dea and Zoe Oldfield, published in the November 2010 Economic Journal.
Their study also finds that the most numerate individuals have higher rates of saving (excluding pension saving) in the years leading up to retirement – and they have higher rates of ''dissaving'' after retirement, that is, they show a greater inclination (or perhaps ability) to run down their assets on retirement.
Both of these results emerge from analysis of data from the English Longitudinal Study of Ageing (ELSA), which collects detailed measures of both cognitive function and financial variables on a large sample of adults aged 50 and over in England. Both results remain true after controlling for education, literacy, other dimensions of cognitive ability (such as memory and executive function) and other socio-economic factors.
How do levels of numeracy relate to people''s ability and willingness to save for retirement? This research examines the relationship between numerical ability and retirement saving among older adults in England.
As well as showing that more numerate individuals have higher wealth than the less numerate, even when taking account of their greater education and incomes, it shows that they accumulate liquid private wealth faster than less numerate individuals before they retire and that they run it down faster after retirement.
But the research also finds that although numeracy is correlated with the private financial saving behaviour of those approaching and going through retirement, there is no strong evidence that the more numerate groups are systematically different to the less numerate when it comes to the changes to income, consumption or subjective wellbeing that occur around the time of retirement.
Understanding the relationship between numerical ability and retirement saving is becoming increasingly important, both from a theoretical and an empirical perspective. Declines in the generosity of state pension provision and other welfare benefits have meant that many of today''s households need to provide substantial proportions of the resources needed for retirement.
At the same time, the typical pension and savings products held by households have been getting more complicated. It is important to evaluate, therefore, the extent to which individuals are equipped to deal with these increasingly complex decisions and assess what role there is for government policy in this area.
To investigate these issues, the study analyses ELSA data collected every two years from 2002 to 2008. The individuals in the sample are divided into four groups on the basis of their response to a series of short mathematical questions applied to simple everyday situations, such as interest on a bank account or the price of a sofa in a sale.
Those in the most numerate group (around one in nine people) could correctly answer a question involving an understanding of compound interest while those in the least numerate group (around one in six people) were those whose ability does not extend beyond being able to do simple division (such as dividing 300 by 2).
While the study''s findings on the correlation between numeracy and levels of wealth and saving are instructive, it is important to analyse not only whether numeracy is correlated with retirement saving behaviour but also whether it is correlated with broader retirement outcomes.
To assess this, the research looks at the correlation of numeracy with other measures of welfare in retirement and also the way these change around the time that individuals leave the labour market. These measures include replacement rates (the ratio of household pre-retirement income to post-retirement income), expectations of having inadequate resources at some stage in the future and self-reported measures of happiness.
Among the findings, the research shows that retirees in the lowest numeracy group do have the lowest replacement rates – that is, they tend to experience the largest proportionate falls in income on retirement. But there is no statistically significant evidence of a link between numeracy and replacement rates after adjusting for other potentially confounding factors in the analysis.
Looking at the other measures of welfare, there is no evidence of a systematic relationship between numerical ability and well-being as individuals move towards, into and through retirement.
James Banks comments:
''Our research clearly identifies that those with different levels of numerical ability behave very differently when it comes to private forms of saving for retirement. It is not necessarily the case, however, that the different behaviour we observe is evidence of poor planning on the part of the less numerate.
''Given the increasing complexity of decisions that need to be taken by households when planning for retirement, developing an understanding of the link between numeracy and well-being in retirement is an important area for future research.''
''Cognitive Function, Numeracy and Retirement Saving Trajectories'' by James Banks, Cormac O''Dea and Zoe Oldfield is published in the November 2010 issue of the Economic Journal.
The research was funded by the Economic and Social Research Council and the US National Institute of Ageing.
professor of economics | University of Manchester | email@example.com
senior research economist | IFS | firstname.lastname@example.org
senior research economist | IFS | email@example.com