Life Expectancy and Ethnicity
July 2008
This article aims to investigate the age makeup of the Lancashire population by ethnicity; and estimate whether there are any differences in life expectancy between white and black and ethnic minority residents in Lancashire.
Proportions of People by Ethnicity and Age
The Office for National Statistics provided figures for the ethnic profile of different age groups from the 2001 census. This allows the comparison of the age makeup of both the white and black and ethnic minority (BME) populations of Lancashire. Overall from the census, 5.3% of the population were from a non-white group in 2001 (note that these figures are now seven years old), but this is skewed heavily towards the younger age groups, as shown in Figure 1. While one in ten of the BME population in Lancashire is aged between 20 and 24 years, only one in twenty of the white population is in the same age group. In contrast while almost one in twenty-five white people were aged between 75 and 79 years (3.7%), the figure for those from a BME background is only 1.1%. The average ages of the white and BME population in Lancashire in 2001 were 39.8 and just 26.1 years respectively.
Looking at these figures as the proportion from each age group in each ethnic groups gives Figure 2. The proportion of the population from an ethnic minority falls as age increases, as would be expected from Figure 1. This is important for older people's services since only a small proportion of Lancashire's older people come from an ethnic minority. For example, only 1.4% of Lancashire's population aged 65 years and over is from a BME background. This means that while an important group, they may be difficult to find and provide services for, simply from the small numbers (aside from any cultural issues). Geographically, the older BME population will also follow that for the overall BME population in Lancashire, with the vast majority in the districts of Pendle, Preston, Burnley and Hyndburn.
Implications for Future Service Provision
Figure 2 is also useful as a guide to use for future planning in older people's services. This is because the number of older people from a BME background in Lancashire will increase in future, as the higher proportions of younger BME people age. So while only 1.4% of those aged 65 years and over are from an ethnic minority at the moment, as this is 2.3% of those aged 45 years and over, the proportion will have increased significantly in 20 years for example. This may well be worth investigating further in future.
| Age greater than: | Proportion from a BME background |
|---|---|
| 90+ | 0.7% |
| 85+ | 0.5% |
| 80+ | 0.7% |
| 75+ | 0.8% |
| 70+ | 1.0% |
| 65+ | 1.4% |
| 60+ | 1.6% |
| 55+ | 1.7% |
| 50+ | 1.9% |
| 45+ | 2.3% |
| 40+ | 2.6% |
| Source Office of National Statistics, Census 2001 | |
Life Expectancy by Ethnicity
There is some anecdotal evidence for differences in life expectancy by ethnicity in Adult and Community Services, and particularly in healthy life expectancy. Also in the 2001 Census, the highest self-reported proportions of both poor health and long term illness by ethnic group nationally were from Pakistani and Bangladeshi men and women (see the ONS story 'Manual workers die earlier than others'; also note that the majority of the BME community in East Lancashire have Pakistani heritage). Both of these measures are linked to life expectancy.
There are currently no available figures for life expectancy by ethnicity for the Lancashire area from the Office of National Statistics. Experimental life expectancy estimates are available by ward, as are figures for the ethnic distribution of wards. Average life expectancy is linked to the area where people live, by such factors as levels of deprivation and the quality of local housing. Therefore the aim of estimating life expectancy by ethnicity is to investigate whether there is any support from the ONS figures for any differences by ethnic group.
Looking at figures at a district level, the lowest life expectancies are for the districts of Burnley, Hyndburn and Preston, which have three of the highest proportions of BME residents. The highest life expectancies are in the districts of Ribble Valley, South Ribble and Wyre, all of which have relatively low proportions of BME residents.
Methodology
By comparing the ONS predicted life expectancies for each of the 257 wards in the Lancashire County Council area with the ethnic population of that ward, a mean average life expectancy at birth was calculated for both white and BME residents in Lancashire. This was done by summing the number of people in each group multiplied by their predicted ages for each ward, and dividing the total by the number of people from that ethnic group in the county.
Limitations
It is important to remember though that these are merely estimates to provide starting figures, in the absence of stronger data.
In particular the following assumptions and approximations are made:
- that ONS estimates are accurate. These are experimental statistics based on the death rate of the ward and the proportion of people rating their health as poor in the 2001 census. They are also based assuming that the people living in an area do so throughout their lives while for many people this will not be true; and
- that life expectancy in a ward is equal between white and BME residents, (ie that the life expectancy depends on location rather than ethnicity). Whether this is a fair assumption or not is unclear: while many aspects will be similar to all residents of a ward (eg deprivation levels), it is unclear whether there will also be any cultural differences in life expectancy. (For example this could be any differences in rates of specific illness between different groups, or there could be a difference due to different diets, or rates of alcohol consumption).
Life Expectancy
Life expectancy at birth for a particular area and for a particular time period is an estimate of the average number of years a newborn baby would survive if he or she experienced that area's age-specific mortality rates for that time period throughout his or her life. The figure reflects mortality among those living in the area in the particular time period, rather than mortality among those born in the area. It is not therefore the number of years a baby born in the area in the time period could actually expect to live, both because the death rates of the area are likely to change in the future and because many of those born in the area will live elsewhere for at least some part of their lives. Bearing in mind the limitations of both the data and the method used, the resulting averages for life expectancy at birth are shown in the table below.
| Ethnic group | Life expectancy at birth (years) | Life expectancy at 65 (years) |
|---|---|---|
| BME | 75.5 years | approx. 80.5 years |
| White | 77.7 years | approx. 82.7 years |
| Difference | 2.2 years | |
| Source Calculated using data from the Office of National Statistics: Population by ethnicity by ward (Census 2001) and life expectancy experimental estimates (based on 2001-3 data). | ||
There is a difference of over two years between the two groups. This is quite a large difference in life expectancy at birth, equal to that between the overall figures for the relatively less deprived Fylde district and relatively high deprived Hyndburn district.
The life expectancy at birth is the expectancy across all ages in an area. The life expectancy for someone aged 65 will be higher than the figure above since this depends on the mortality rate of those aged over 65, rather than the mortality rate across all ages, (ie someone at 65 has survived childhood illnesses etc and so will expect to live longer than the overall average). While more relevant to older people's services, the life expectancy at 65 is not available at ward level, but at district level in Lancashire, the average difference between life expectancy at 65 and life expectancy at birth is 3.8 years longer for men and 6.0 years longer for women. Therefore we would expect that the life expectancy at 65 to be around five years older than the figures above, and therefore around 80.5 years for BME residents and around 82.7 years for white residents, (as above).
In the case of both of these life expectancies, it is the difference between ethnicities of over two years that should be noted, more than the absolute age due to the uncertainty of the methods used.
Healthy Life Expectancy and Disability-Free Life Expectancy
There is also some anecdotal evidence from Older People's services of lower healthy life expectancy in BME groups in Lancashire. Healthy life expectancy and disability-free life expectancy are important since it is as people live beyond these that they are more likely to require access to relevant health and social care services. These depend on the population's self rated perceptions of health, and whether they have a limiting long term illness.
Healthy life expectancy is defined as the average number of years a person aged "x" would live in good/fairly good health if he or she experienced the particular area's age-specific mortality and health rates for 2001 throughout their life. Disability-free life expectancy is defined as the average number of years a person aged "x" would live disability-free (no limiting long-term illness) if he or she experienced the particular area's age-specific mortality and health rates for 2001 throughout their life. Ward data for healthy life expectancy , and disability-free life expectancy. from birth are both available at ward level, based on 2001 data. Therefore it is possible to calculate figures for these as above, subject the same caveats. The estimated figures for healthy and disability-free life expectancy are shown below, alongside upper and lower confidence limits, which are taken from the 95% confidence limits of each age from the ONS.
| Ethnic group | Disability-free life expectancy at birth (years) | Healthy life expectancy (years) | ||||
|---|---|---|---|---|---|---|
| 95% lower limit | Estimated average | 95% upper limit | 95% lower limit | Estimated average | 95% upper limit | |
| BME | 55.8 | 57.1 | 58.4 | 63.7 | 65.2 | 66.7 |
| White | 59.9 | 61.2 | 62.6 | 67.7 | 69.3 | 70.9 |
| Difference | 4.1 | 4.2 | 4.2 | 4.1 | 4.1 | 4.1 |
| Source Calculated using data from the Office of National Statistics: Population by ethnicity by ward (Census 2001) and disability free and healthy life expectancy experimental estimates (based on 2001-3 data). | ||||||
Again, there is a difference between white and BME residents for both types of life expectancy. It is also worth noting that the upper limit of the BME score for each is less than the lower score of the corresponding white expectancy. This shows that there is a significant difference between the two scores, (assuming that the estimates assumptions are fair). The 4 year differences between the groups are similar to the differences between the overall expectancies for South Ribble and Hyndburn districts for example.
Also, as for the overall life expectancy, the above ONS estimates are also available at district level from age 65. The difference between these estimates at birth and at 65 years is greater than for the overall life expectancy. In Lancashire at district level, the average difference in disability-free expectancy is 10.9 years for women and 12.4 years for men. Therefore the expectancies at 65 are much higher, but the differences by ethnic group would remain under the model.
Conclusions
Proportions of People by Ethnicity and Age
The proportion of the population of Lancashire from an ethnic minority is not linear and is much younger, on average than the white population. For older people's services, this means that only a small proportion of the potential service users come from a BME background. This means they may be more difficult to reach to provide services to.
The proportion of the older population who are from a BME population is going to increase though as the larger numbers of younger BME residents age.
Life Expectancy by Ethnicity
There does appear to be some evidence to show that overall life expectancy as well as that for disability-free and healthy life expectancy at birth is lower for those from BME backgrounds in Lancashire, (by around 2 years, 4 years and 4 years respectively). However, the methods used to calculate this are experimental and the assumptions made in the model are sizeable. Therefore these results should be seen as indicative of a difference between the white and BME population, but it should be noted that the true difference could be different.
There is some national support for the finding that disability-free and healthy life expectancy at birth is four years lower for those from a black and ethnic minority background. This could be partly explained by the highest rates of self reported long-term illness and poor health on Census 2001 coming from those from a Pakistani and Bangladeshi background. The majority of the BME population in east Lancashire have Pakistani heritage.
For further details, please contact:
Steven Knuckey
Tel 01772 531811
Email Steven.Knuckey@lancashire.gov.uk