This short story looks at people’s assessment of their own well-being by different ethnic groups using the April 2011 to March 2012 Annual Population Survey. As well as providing estimates for ethnic groups and ethnic group by sex, it provides further information that may explain some of the differences observed.
Asking people to provide their own assessment of their own well-being is part of ONS’s approach to measuring the well-being of the nation. Showing differences for different population groups is important as it allows for debate about why these differences may exist and provides useful information for policy making. This short story considers the differences in well-being between different ethnic groups and offers some potential reasons for these observed differences. It is published as part of the ONS’s Measuring National Well-being programme which aims to monitor and understand the well-being of the UK.
As well as a range of demographic and labour market questions the Annual Population Survey asks a cross section of 165,000 adults aged 16 and over to make their own assessment of how they feel about their lives.
Figure 1 shows that when asked how satisfied people are with their lives, respondents from the White ethnic group gave an average rating of 7.4 out of 10 in 2011-12.
This was in contrast with people from the Black ethnic group who gave the lowest ratings with an average of 6.7 out of 10. This was also lower than the Bangladeshi (7.0 out of 10) and Mixed/multiple ethnic groups (7.1 out of 10) who also gave significantly lower ratings on average than the White ethnic group. The Black ethnic group is made up of Black, African, Caribbean and Black British people.
In contrast, people who identified themselves as Indian (7.5 out of 10), Chinese (7.4 out of 10) and Other Asian backgrounds (7.4 out of 10) gave similar ratings on average to the White population.
As well as asking about how satisfied people are with their lives, the Annual Population Survey also asks to what extent people feel the things that they do in their lives are worthwhile.
The White ethnic group, with an average rating of 7.7 out of 10, was higher than all the other ethnic groups. The lowest average rating for this question was given by Bangladeshi respondents who reported an average of 7.3 out of 10.
Pakistani, Chinese, Black, and the Other ethnic group all reported a rating of 7.4 out of 10.
As well as asking people to make an assessment of how their lives are going overall, it is also important to understand how people’s day-to-day emotions differ when making an assessment of people’s well-being. For example, the Annual Population Survey asks about how happy and anxious people felt in the previous day.
The ethnic group that provided the lowest average rating for the ‘happy yesterday’ question was the Black ethnic group (6.9 out of 10), significantly lower than the average for the White group (7.3 out of 10). The Indian and Other Asian background groups both reported the highest ratings of 7.4 out of 10.
Figure 2 shows that for ‘anxiety yesterday’, all other ethnic groups reported on average higher levels of anxiety than the White ethnic group.
Indian respondents reported significantly higher levels of anxiety (3.4 out of 10) compared with the White ethnic group, which is in contrast to happiness yesterday, life satisfaction and worthwhile things in life, where their responses were similar. Bangladeshi (3.6 out of 10) and Pakistani (3.5 out of 10) groups also reported significantly higher ratings compared with the White group (3.1 out of 10) meaning that on average people who identified with these ethnic groups reported being more anxious in the day before interview.
Overall, regardless of ethnic group, according to the 2011-12 Annual Population Survey women in the UK reported on average slightly higher levels of life satisfaction, worthwhile things in life, the same levels of happiness yesterday, but higher levels of anxiety yesterday than men (ONS, 2012).
Figure 3 shows that when looking at the differences between men and women broken down by ethnic group the differences were also small for each group. Given for some ethnic groups the sample sizes are small, caution should be adopted when interpreting the differences between men and women as they will often not be statistically significant.
Some of the largest differences between men and women for different ethnic groups were for the average ratings for the question that asked about the extent to which people feel the things they do in their life are worthwhile. The averages for White men and women for this question were 7.6 and 7.8 out of 10 respectively.
The largest difference between men and women were for the:
Pakistani ethnic group (7.2 for men vs. 7.6 for women), the
Bangladeshi ethnic group (7.1 for men vs. 7.5 for women), and the
Other ethnic group (7.3 for men vs. 7.7 for women).
However, sample sizes are small, particularly for the Bangladeshi group which is not a statistically significant difference. The Chinese group is the only ethnic group where men reported on average higher levels of feeling things are worthwhile than women, however again the sample is small and this is not a statistically significant difference.
There are distinct differences in the average ratings between ethnic groups, and these are unlikely to be explained by different ratings given by men and women, as these differences are also small.
So what could be the main reasons for the different average ratings given by people in different ethnic groups?
The differences may in part be due to what can be described as ‘cultural bias’. This may be because people from different cultures may interpret the question scales differently or be more likely to give more extreme or moderate ratings when asked to make an assessment of their life in this way. However, although the research literature suggests there are some cultural differences in the patterns of observed responses, it is difficult to say to what extent this represents error in the data rather than genuine differences in how people feel, or how they assess their lives (OECD, 2013). This presents a challenging research agenda for the future.
An alternative to the cultural bias theory is that these differences exist because they reflect people’s responses to the circumstances that they find themselves in.
As Figure 4 shows, unemployment rates for Pakistani, Bangladeshi and the Black ethnic groups are amongst the highest. In October to December 2012 they stood at 18.4%, 20.0% and 15.7% respectively. Given the large association between unemployment and subjective well-being this may go some way to explaining some of the reason for the lower average life satisfaction, worthwhile things, happiness yesterday and higher average anxiety levels for these ethnic groups (ONS, 2013).
Pakistani and Bangladeshi groups also on average report the highest rates of poor health and limiting and long-standing illness or disability (ONS, 2008), other factors highly associated with subjective well-being.
Within the broader ethnic groupings, pupils of any Black background have historically been least likely on average to achieve 5 or more GCSEs at grade A* to C or equivalent including English and mathematics GCSEs or iGCSEs (DfE, 2013).
Further, more complex analysis is required to look in more detail to better determine how much these differences in circumstances can explain the observed differences in average ratings of subjective well-being between ethnic groups. The Annual Population Survey with its large sample and wide variety of different variables of interest is a useful source for further research in this area.
Many of the specific measures that are related to an individual's well-being are also related to each other such as age, employment status, marital status and health. ONS is currently analysing1 the relationship between each measure and subjective well-being. A methodological report providing more detailed results will be published in May 2013.
The questions developed and included in the Annual Population Survey were developed with expert academic advice as well as benefiting from discussions between members of the National Statistician’s Measuring National Well-being Advisory Forum and Technical Advisory Group.
The questions that ONS has asked on the Annual Population Survey include:
Overall, how satisfied are you with your life nowadays? (Evaluative approach),
Overall, to what extent do you feel the things you do in your life are worthwhile? (Eudemonic approach),
Overall, how happy did you feel yesterday? (Experience approach),
Overall, how anxious did you feel yesterday? (Experience approach).
All answered using a 0 to 10 scale where nought is ‘not at all’ and 10 is ‘completely’
Further information on the ONS approach to measuring subjective well-being can be found in the paper ‘Measuring Subjective Well-being’ (240.8 Kb Pdf) published by ONS in July 2011.
When comparing differences between mean ratings for groups or areas it is important to realise that these comparisons are made using aggregate statistics based on samples with those characteristics. Just because the average of the sample has a certain rating of subjective well-being does not necessarily mean that all people with that certain characteristic have that particular outcome.
For example, even though women on average have higher life satisfaction than men, it is important not to infer that all women are more satisfied with their lives than men.
It is also important to note that although subjective well-being estimates in this report have been analysed by people with different characteristics we are not inferring causation, even though some groups are more likely to give higher life satisfaction ratings on average, it may not be the particular characteristic that is causing them to have higher subjective well-being.
There are other factors that could also be influencing people’s ratings which would need to be controlled for in a regression model, and even then causation is often difficult to infer.
The size of differences between mean ratings of subjective well-being for groups of people with certain characteristics or in different circumstances can appear fairly small. These differences must be interpreted in light of the fact that subjective well-being is affected by many different life circumstances with a variety of factors having a bearing on a person’s subjective well-being.
For example, if someone is in poor health, that is only one aspect of their lives so when making a cognitive judgment about how satisfied they are with their lives overall, although their poor health may have a bearing, other factors may mitigate against that and mean they do not give such a low score as we may perhaps first expect.
The Annual Population Survey, first conducted in 2004, combines results from waves one and five of the Labour Force Survey and the English, Welsh and Scottish LFS boosts. The survey asks 155,000 households and 360,000 people per dataset about their own circumstances and experiences regarding a range of subjects including housing, employment and education.
The survey provides enhanced data on key social and socio-economic variables. For the subjective well-being questions responses by proxy are not collected. That is only people who respond directly to the survey are included rather than including responses made on behalf of other people. For this reason the sample size for the subjective well-being data is around 165,000.
One of the benefits of collecting subjective well-being data on a large scale survey such as Annual Population Survey is that large sample sizes, all else being equal, produce smaller standard errors. The standard error is an indication of the accuracy of an estimate and gives users an indication of how close the sample estimator (in this case the mean subjective well-being rating) is to the population value: the larger the standard error, the less precise the estimator. Large sample sizes also allow for comparisons across population groups and areas. The Annual Population Survey is a mixed mode survey and uses both face-to-face and telephone interviews. Different collection modes can affect responses and subjective well-being estimates are no exception.
This article is published as part of the ONS Measuring National Well-being Programme.
The programme aims to produce accepted and trusted measures of the well-being of the nation - how the UK as a whole is doing.
Measuring National Well-being is about looking at 'GDP and beyond'. It includes headline indicators in areas such as health, relationships, job satisfaction, economic security, education, environmental conditions and measures of 'subjective well-being' (individuals' assessment of their own well-being).
Find out more on the Measuring National Well-being website pages.
Details of the policy governing the release of new data are available by visiting www.statisticsauthority.gov.uk/assessment/code-of-practice/index.html or from the Media Relations Office email: firstname.lastname@example.org