The Scottish Government

« Previous | Contents | Next »

Listen

4. FACTORS ASSOCIATED WITH SUBJECTIVE WELL-BEING

4.1 It appears that subjective well-being in Scotland is high in absolute terms and also compares favourably with that of other countries in Europe. Satisfaction levels were lowest in relation to jobs, but even here the average score of 7 is comfortably above the mid-point of the scale. However, since the aim of Scottish Government policy is to improve the well-being of all people in Scotland, it is important to establish whether there are some groups of people who have lower levels of subjective well-being than others. If the latter is found to be true then policies can be targeted specifically at the groups of people in society who are most unhappy and dissatisfied with their lives.

4.2 In Galloway et. al. (2006), David Bell provides a comprehensive review of the literature on subjective well-being. In pulling together the key findings from previous research on the topic, he identifies a core set of factors that have been shown to be associated with subjective well-being. These are:

• Gender - with women tending to have higher levels of subjective well-being than men.
• Age - with subjective well-being found to be 'U shaped' in age - highest among the young and older age groups but low among the middle aged.
• Family status - with those who are married most satisfied and happy followed by those living together, those widowed or widowers, the divorced and finally the separated.
• Health status - with those reporting good self rated health more satisfied and happy than those reporting poor health.
• Employment status - with the unemployed having much lower levels of subjective well-being than those employed.
• Income - with the wealthy having higher levels of happiness and satisfaction than those less well off, despite the traditional argument that economic growth fails to bring higher levels of well-being.
• Social capital - with those who are most trusting of others having higher levels of subjective well-being than those who are least trusting.

4.3 This section explores the relationship between a range of demographic and subjective factors (collected in SSA 2007) and the five measures of subjective well-being included in the survey, to determine which factors are associated with higher or lower levels of well-being in the Scottish context. These include:

• Socio-demographic factors (sex, age, socio-economic classification, employment status, educational attainment, income, marital status, children living in the household)
• Presence of long-term illness, disability or health problem
• Self-perceived level of financial hardship
• Self-perceived health status
• Geography (urbanity/rurality, deprivation, living in Strathclyde or not)
• Social trust

4.4 As the aggregate scores of subjective well-being are so positively skewed, there is a limit to how much can be deduced from examining mean scores alone. ¹⁶ The real interest lies in revealing inequalities in well-being, or which groups of people in society exhibit lower than average subjective well-being. Therefore, for each of the five subjective well-being measures a new variable was derived which split responses to each question into one of three categories: lower than average, average (mean score) or above average. These variables were then used in the analysis to establish whether some groups of people living in Scotland are less happy and satisfied with their lives than others. ¹⁷

Gender

4.5 The only measure of subjective well-being that was significantly associated with gender was 'job satisfaction' (Figure 5). Of those in paid employment, men were significantly more likely than women to report a score that fell below the mean score of 7 (34% of men had a lower than average score, compared with 26% of women). While this appears to support the findings of Bell & Blanchflower (2007) and Blanchflower and Oswald (2004) who found that men are less satisfied than women with their lives as a whole, it is important to note that gender was not significantly associated with 'satisfaction with life as a whole' or the other 3 measures included in SSA 2007.

Figure 5 Scores on 'job satisfaction' by gender

Sample Size: All those in employment/waiting to take up paid work=829
Men=397, Women=432

Age

4.6 As has been noted, it is widely accepted that there is a relationship between age and subjective well-being. Much of the early analysis in this area concluded that happiness was either flat across age categories or increased slightly with age. More recent research suggests that the relationship between the two is curvilinear or 'U-shaped,' with both happiness and life satisfaction at their lowest among people in their mid-forties to mid-fifties ¹⁸. There are two possible explanations as to why this is the case. First, is that it is as a result of a cohort effect, that is, people born 45-55 years ago are, on the whole, less happy and satisfied with life than those born either before or after them and these people will continue to have low levels of subjective well-being as they move through their lives. Second is that an individual's happiness is not fixed but changes as they move through their life. It is the latter of these explanations that has received most support with Clark (2007) and Blanchflower and Oswald (2008) finding evidence of a 'U-shaped' life cycle effect in their research using longitudinal panel data.

4.7 Initial bivariate analysis suggests that age is indeed related to both 'satisfaction with life as a whole' and 'happiness.' Figure 6 indicates the proportion of people in each age group that had an above average score on these measures ¹⁹. There is evidence of a 'U-shaped' relationship with age on both measures although it is most pronounced on the happiness measure. Older and younger people were around twice as likely as people aged 35-54 to have a score of 9 or 10 on the happiness measure (48% for people aged 18-24, 56% for people aged 65+ and 28% of people aged 35-54). With regards to 'satisfaction with life as a whole' it was people aged 45-54 that were least likely to give a score of 9 or 10 and those aged 65+ that were most likely to do so (33% and 50% respectively).

Figure 6 Above average (9+) scores for 'life satisfaction' and 'happiness' by age

Sample size: All respondents
18-24=99, 25-34=197, 35-44=299, 45-54=284, 55-64=260, 65+=366

4.8 However, regression analysis showed that once a range of other factors is taken into account, age is not significantly related to any of the five subjective well-being measures in the survey. This contradicts many of the findings in the literature on the relationship between age and subjective well-being, and suggests that other factors that relate to age, such as employment status or health, may have a greater impact on both 'life satisfaction' and 'happiness' than age itself.

Education

4.9 Bivariate analysis suggests that people with no educational qualifications were around twice as likely as those educated to degree level to have lower than average job satisfaction (46% and 24% respectively). There was a similar, albeit slightly less pronounced, pattern between education and satisfaction with general standard of living. However, regression analysis showed that the relationship between both measures and education disappeared once other factors were controlled for. This supports much of the evidence on subjective well-being correlates, suggesting that education 'has only a small direct effect on happiness' (Layard, 2005).

Employment status and socio-economic classification

4.10 The relationship between well-being and employment status has been well documented and it is widely held that subjective well-being is lowest among the unemployed. ²⁰ Layard described unemployment as 'disastrous' because in addition to reducing a person's income it 'reduces happiness directly by destroying the self-respect and social relationships created by work' (Layard, 2005).

4.11 Employment status was found to be a key factor associated with four of the subjective well-being measures (excluding job satisfaction). Although the number of unemployed people that took part in SSA 07 was relatively small, it was still the case that they were significantly more likely than people in paid work to have lower than average levels of subjective well-being (Table 1). For example, half (48%) of the people unemployed had lower than average levels of 'life satisfaction' compared with a quarter (25%) of those in work or waiting to take up work. Subjective well-being was also particularly low for people who are permanently sick or disabled. Again there were relatively few people in the sample that fell into this category but around 6 in 10 had lower than average scores on the four measures. Subjective well-being was highest among retired people with only around 1 in 5 having lower than average well-being scores across the board. The exception was 'general standard of living' with 27% of those retired reporting lower than average satisfaction.

Table 1 Below average (7<=) subjective well-being scores, by employment status (cell %)

4.12 There was a significant association between socio-economic classification and both 'job satisfaction' and 'satisfaction with standard of living' (Figure 8). People with 'routine' or 'semi-routine' occupations were significantly more likely than those in any other occupation to have a below average score. The variation was most pronounced in relation to 'job satisfaction'. Half (48%) of those in 'routine' or 'semi-routine' occupations had a score below the mean score of 7 compared with only 2 in 10 of employers, managers and professionals.

Figure 8 Below average scores on job satisfaction and satisfaction with general standard by socio-economic classification

Sample sizes: Job satisfaction=All those with a job (829), Standard of living=All respondents (1508)
For base sizes for each category please see Tables A2 and A4 in Annex A

Marital status, children in the household

4.13 Regression analysis confirmed that a person's marital status was a key factor relating to how happy they were, how satisfied they were with their life as a whole and with their family or personal life. Married people, those living as married and those widowed or widowers were happier and more satisfied with their lives as a whole than people who were separated, divorced or not married (Table 2). For example, 46% of married people gave a score of 9 or 10 for 'life satisfaction' compared with 31% of unmarried people and 30% who were separated or divorced. Unsurprisingly, the relationship with marital status was most pronounced with regard to satisfaction with 'family or personal' life, with married people almost twice as likely as those who were separated or divorced to rate their satisfaction with family or personal life as 9 or 10 (60% and 35% respectively).

4.14 These findings support Bell and Blanchflower's conclusion that married people are happier than single people in Scotland (Bell and Blanchflower, 2007) as well as Layard's observation that 'family relationships (and our close private life) are more important than any other single factor affecting our happiness' (Layard 2005).

Table 2 Subjective well-being measures, by marital status (cell %)

4.15 Whilst Clark (2003) found that when children are born the parents' happiness levels increase then subside again within a couple of years, there was no link found between living in a household with children and overall life satisfaction in SSA 2007.

4.16 However, the presence of children in the household was significantly associated with levels of job satisfaction (Table 3). People living with children were more likely than those with none to have a lower than average score on job satisfaction (35% compared with 27%). This is perhaps unsurprising given that most people with children have to juggle work alongside the demands of having children and is particularly problematic for those who have employers who are unsympathetic to their needs.

Table 3 Job satisfaction, by children age 0-17 living in household 2007 (cell %)

Actual income and self-perceived hardship

4.17 The introduction to this report touched on the paradox that is central to the debate on subjective well-being - that western developed countries are becoming financially richer but are not becoming any happier. The picture is further complicated by evidence that within developed nations wealthy people are happier than those who are less financially well-off (Layard, 2006). Regression analysis confirmed that actual household income was significantly and independently associated with 'job satisfaction' and 'satisfaction with general standard of living'. The pattern was identical for both; people with the lowest household income were more likely than anyone else to have lower than average scores. On 'job satisfaction' for example, those in the lowest income quartile were more than three times as likely as those in the highest income band to have a score of 6 or less (60% and 18% respectively).

Figure 9 Percent below average score on 'job satisfaction' (6<=) and 'satisfaction with general standard of living' (<7=), by actual household income

Sample sizes: Job satisfaction=All those with a job (829), Standard of living=All respondents (1508)
For base sizes for each category please see Tables A2 and A4 in Annex A

4.18 A much stronger predictor of a person's subjective well-being than their actual household income was their own assessment of how they were coping financially. This was measured by asking people to evaluate how they were coping on their current household income with answer options ranging from 'living very comfortably on present income' to 'finding it very difficult on present income'. Although the size of the sample reporting that they were finding it difficult or very difficult was relatively small, regression analysis demonstrated that self-perceived financial hardship was a key factor relating to all five subjective well-being measures included in the survey. Happiness and satisfaction levels were highest among people who said they were living 'comfortably' or 'very comfortably' and lowest among those who were finding it 'difficult' or 'very difficult' to cope.

4.19 Figure 10 illustrates the strength of the relationship between self-perceived financial hardship and subjective well-being. Perhaps unsurprisingly, the relationship was most pronounced with regards satisfaction with general standard of living. People who were finding it 'difficult' or 'very difficult' to cope on their household income were four times as likely as those who were 'living comfortably' to have a score of 7 or below.

4.20 It is clear from Figure 10 that perceived financial hardship does not only affect a person's assessment of their standard of living but relates to how they evaluate other domains of their life too. Job satisfaction was lowest among those who were struggling to cope financially (55% compared with 22% of those who were living 'comfortably' or 'very comfortably'). People finding it 'difficult' or 'very difficult' to cope were around three times as likely as those living 'comfortably' to have lower than average scores for 'happiness,' 'satisfaction with family life' and 'satisfaction with life as a whole.'

Figure 10 Percent below average score for subjective well-being measures, by self-perceived financial hardship

Sample sizes: Job satisfaction = All those with a job (829), Else = All respondents (1508)
For base sizes for each category please see Tables A2 - A6 in Annex A

Long-term illness and self-perceived health status

4.21 Just as an individual's assessment of how they are coping financially was more strongly related to subjective well-being than their actual income, so too was perception more important than actuality with regards to the relationship between subjective well-being and health. People were asked whether they actually had any long-term illness, health problem or disability. The only measure of subjective well-being that this was significantly and independently related to was 'job satisfaction'. People who reported having an illness, disability or health problem were more likely than those with none, to have low levels of 'job satisfaction' (43% and 27% respectively) ²¹.

4.22 People were also asked how they thought their health was for someone of their age, with answer options ranging from 'very good' to 'very bad'. Figure 12 highlights the clear linear relationship between self-assessed health on four out of the five subjective well-being measures included in the survey ²². People who considered their health to be 'very good' were least likely to have a below average score for 'happiness', satisfaction with 'family or personal life', 'general standard of living' or with 'life as a whole.' On all of these measures it was the case that as self-assessed health status decreased the likelihood of having below average subjective well-being increased. For example, 16% of people who considered themselves to be in 'very good' health scored their satisfaction with 'life as a whole' as 7 or below, compared with 44% of those who described their health as 'fair' and 57% who described it as 'bad' or 'very bad'. People in 'bad' or 'very bad' health were more than twice as likely as those in 'very good' health to score their satisfaction with their family or personal life as 7 or below (39% compared with 16%).

Figure 12 Percent below average score for subjective well-being measures, by self-perceived health status

Sample sizes: All respondents=1508
For base sizes for each category please see Tables A3-A6 in Annex A.

Social trust

4.23 Recently, researchers have started to explore the relationship between social capital and well-being. ²³ Coined by Robert Putnam (1995), social capital is a term used to describe the extent to which a person is embedded into the community and society in which they live. There are several dimensions to social capital including social and civic participation, social networks, reciprocity and trust. SSA 2007 included a question to measure one of these dimensions - social trust. ²⁴ Layard (2005) argued that this kind of interpersonal trust affects happiness, with people who are more trusting of others more satisfied and happy with their lives.

4.24 Initial bivariate analysis suggests that there is a relationship between how trusting a person is of others in society and their subjective well-being. Those least trusting of others (i.e. who said you 'can't be too careful in dealing with people') were more likely than those who said that 'most people can be trusted' to have a below average score on all of the subjective well-being measures included in the survey (Figure 13). However, regression analysis demonstrated that, once other factors are controlled for, the relationship is no longer significant. So there is a link between social trust and subjective well-being but it is a relatively weak one and determining causality between the two is problematic.

Figure 13 Percent below average score for subjective well-being measures, by social trust

Sample sizes: Job satisfaction=All those with a job (829), Else=All respondents (1508)
For base sizes for each category please see Tables A2-A6 in Annex A.

Urbanity/Rurarlity/ Deprivation/ Strathclyde region

4.25 In addition to examining the relationship between individual characteristics and subjective well-being it was also possible to explore the correlation, if any, between where a person lives and their self-assessed happiness and life satisfaction. Three area based factors were considered: urbanity, deprivation and whether an individual lived in the Strathclyde region or not. The latter was included to see if there was any evidence to support Bell & Blanchflower's conclusion that 'residents of Strathclyde are particularly unhappy'(2007).

4.26 Initial bivariate analysis suggested that people living in remote small towns and accessible rural areas were less likely than those living in more urban areas to have lower than average scores on the 'life satisfaction' and 'happiness' measures. However regression analysis showed that these differences disappeared once other factors were taken into consideration indicating that the rurality of the area a person lives in is not a key factor relating to their perception of happy and satisfied they are with life.

4.27 Given the strong correlation between a person's income and the level of deprivation of the area they live in, we might expect to find some evidence of a relationship between well-being and deprivation. The 'National survey of public attitudes to mental health, mental well-being and mental health problems' found evidence of a relationship between the two, with people living in the most deprived areas of Scotland most likely to rate their health as poor and be more susceptible to mental ill-health than those living in less deprived areas ²⁵.

4.28 Using the Scottish Index of Multiple Deprivation ( SIMD) ²⁶ quintiles as the measure of deprivation for analysis, SSA data showed that it was not significantly and independently related to 'satisfaction with family and personal life' and 'satisfaction with general standard of the living'. It was however, one of the key factors related to 'happiness', 'life satisfaction' and 'job satisfaction'. People living in the most deprived areas in Scotland were almost twice as likely as those living in the least deprived areas to have a below average score on the happiness measure (46% compared with 21% in the 4 ^th quintile and 25% in the least deprived quintile). It was also the case the those living in the most deprived areas in Scotland were also significantly more likely than those in the least deprived areas to have low levels 'job satisfaction' ²⁷.

Figure 11 Percent below average score on 'happiness' and 'job satisfaction', by deprivation level ( SIMD)

Sample sizes: Job satisfaction=All those with a job (829), Happiness=All respondents (1508)
For base sizes for each category please see Tables A2 and A6 in Annex A

4.29 There was some evidence of a relationship between subjective well-being and living in Strathclyde or not ²⁸. At the bivariate level, people living in Strathclyde were more likely than those living elsewhere in Scotland to report lower than average (7 or less) scores on both self-assessed happiness and overall life satisfaction. Given that there was a significant relationship between deprivation and some of the self-assessment measures and as Bell & Blanchflower note that the majority of Scotland's most deprived areas are in and around the Strathclyde area this is perhaps unsurprising. (Bell & Blanchflower, 2007). Multi-variate analysis, which included all three area based measures, confirmed that it is deprivation that matters most.

Key points:

• Men were more likely than women to have a lower than average 'job satisfaction' score (34% compared with 26%).
• Age was not significantly and independently associated with subjective well-being.
• Employment status was highly correlated with 4 of the 5 subjective well-being measures (the exception was 'job satisfaction'). 48% of people who were unemployed had a 'life satisfaction' score lower than 8, compared with 25% of those in or waiting to take up paid work.
• People in routine or semi-routine occupations reported lower 'job satisfaction' and 'satisfaction with general standard of living' than those in any other occupation (48% had a below average 'job satisfaction' score compared with 20% of employers, managers and professionals).
• Married people, those living as married and widowed/widowers were happier and more satisfied with their lives as a whole than people who were separated, divorced or never married. (For example, 46% of married people had an above average 'life satisfaction' score, compared with 30% who were separated or divorced.)
• People living with children were more likely than those living with none to have a lower than average score on 'job satisfaction' (35% compared with 27%).
• Actual income was associated with 'job satisfaction' and 'general standard of living'. However, self-assessed hardship was a key factor relating to all 5 measures of subjective well-being. For example, people who were finding it 'difficult' or 'very difficult' to cope were four times as likely as those 'living comfortably' to have a below average score for 'satisfaction with standard of living.'
• People living in the most deprived areas in Scotland were almost twice as likely as those living in the least deprived areas to have a below average happiness score (46% compared with 25%).
• There was no relationship between subjective well-being and how urban or rural an area was.
• People who reported illness, disability or health problems were more likely than those with none to have low levels of 'job satisfaction' (43% compared with 27% respectively).
• Self-assessed health was more strongly related to the other measures of subjective well-being. For example, 16% of those who considered themselves to be in 'very good' health had a below average 'life satisfaction' score compared with 57% of those who described their health as 'bad' or 'very bad' .
• There was no significant relationship between how trusting a person was of others in society and their subjective well-being.

« Previous | Contents | Next »