Download BibTex file | 

As the degree to which individuals are expected to provide their own resources for retirement increases, there is a correspondingly increasing importance of individuals being able to understand the financial choices they face and to choose savings products, portfolios and contribution rates accordingly. In this paper we look at numerical ability and other dimensions of cognitive function in a sample of older adults in England and examine the extent to which these abilities are correlated with various measures of wealth and retirement saving outcomes. The key findings are:

a) Relatively large fractions of the population can be seen to have relatively low levels of financial numeracy and these numeracy levels decline systematically with age.

b) Numeracy levels are correlated with measures of retirement saving and investment portfolios, even when we control for other cognitive ability and education.

c) Numeracy is correlated with knowledge and understanding of pension arrangements, and with perceived financial security, even when we control for other cognitive ability, education and the level of overall retirement saving.

The lessons of our analysis are threefold, even though at this point we have only crosssectional data available on which to base our analysis. Firstly, it shows yet another dimension in which inequalities amongst older individuals are apparent. Second, the analysis suggests that in the short run there may be a role for targeting simple retirement planning information at low numeracy, low wealth, low education groups. Third, it suggests that a longer run policy goal might want to target numeracy levels more generally in order to reduce the fraction of the population with low basic skills. Whether such a policy would have knock on effects on to retirement planning arrangements, however, is a more difficult question to answer on the basis of the conditional correlations we present here. On this topic in particular, there is much further work to be done.

Download full version (PDF 522 KB)