Methodological assumptions and limitations of life expectancy estimates for minoritised ethnic groups in the UK: implications for validity, practice, and policy

Experimental life expectancy estimates calculated by the Office for National Statistics (ONS) for the period 2011 to 2014 show significantly longer life expectancy for minoritised ethnic groups in England and Wales when compared with the white majority. These findings are in contrast to a large body of evidence of poorer health outcomes among certain minoritised ethnic groups (predominately Bangladeshi, Black Caribbean, Gypsy/Traveller and Pakistani groups), and have serious practice and policy implications if taken as definitive. We examine the data and methodology used by the ONS in producing these estimates, and consider the sources of error in that approach. We find that the estimates for minoritised ethnic groups exhibit high sensitivity to error that is not seen in the estimates for the White British population; although we note that even in our largest error scenario, many minoritised ethnic groups still have higher life expectancy than the White British group. Although the results are supported by evidence around the “healthy migrant” effect, and other global research on life expectancy by ethnic group, there is a risk that the ONS’ life expectancy estimates of minoritised ethnic groups may be being inflated due to the large amount of missing data among these groups, and the potential for those missing cohorts to be at higher risk of morbidity and mortality. The ONS’ estimates, while clearly labelled as experimental, have been used in academia, policy and the press without necessary caveats. We remind researchers of the experimental nature of the ONS’ life expectancy by ethnic group estimates, and advise caution in how they are used. • Experimental ONS ethnic group life expectancy estimates have been reported as fact in academic research & government policy. • This contrasts large body of evidence of poorer health outcomes among certain minoritised ethnic groups. • We examine the data and methodology used by the ONS in producing these estimates. • We find high sensitivity to missing data among minoritised ethnic groups. • We advise researchers to exercise caution when using the ONS estimates.