Racial Discrimination

Occupational Medicine 2007;57:18–24
Published online 23 August 2006 doi:10.1093/occmed/kql088

Racial discrimination, ethnicity and work stress
Emma Wadsworth1, Kamaldeep Dhillon2, Christine Shaw1, Kamaldeep Bhui2,
Stephen Stansfeld2 and Andrew Smith1
...................................................................................................................................................................................

Background Previous research has suggested higher work stress among minority ethnic workers.
...................................................................................................................................................................................

Aims

To determine levels of work stress in three ethnic groups, consider the contribution of racial discrimination to the groups’ profiles of occupational and demographic associations with stress, and assess the association between work stress and well-being.

...................................................................................................................................................................................

Methods

A household quota sample design was used, and 204 black African–Caribbean, 206 Bangladeshi and
216 white (UK born) working people took part in structured interviews.

...................................................................................................................................................................................

Results

...................................................................................................................................................................................

Conclusions Perceived work stress may be underpinned by exposure to racial discrimination at work among black
African–Caribbean women, and this may affect their psychological well-being.
...................................................................................................................................................................................

Key words

Ethnicity; racial discrimination; work stress.

...................................................................................................................................................................................

Introduction
Work stress can be defined in several ways. First, it is often viewed as a characteristic of the work environment similar to other environmental hazards, such as noise. In this case, it is measured by considering the relationship between exposure and health. Second, it is seen as a physiological response to a threatening or difficult aspect of work, and may be measured directly (e.g. an adverse reaction to work stress can be measured using cortisol).
Third, it may be seen in terms of an interactional framework, as in the effort–reward imbalance model [1], where the imbalance between effort and reward at work interacts to influence health. Finally, transactional theories [2] focus on the cognitive processes and emotional reactions individuals have with their work environment. In this case, perceptions of stress are primarily used. In the work described here, the impact of work characteristics, including both environmental stressors (such as noise and working hours) and organizational stressors (such as
1
Centre for Occupational and Health Psychology, Cardiff University,
63 Park Place, Cardiff CF10 3AS, UK.
2
Centre for Psychiatry, Queen Mary, University of London, Charterhouse
Square, London E1M 6BQ, UK.

Correspondence to: E. Wadsworth, Centre for Occupational and Health
Psychology, Cardiff University, 63 Park Place, Cardiff CF10 3AS, UK.
Tel: [phone]; fax: [phone]; e-mail: [e-mail]

effort–reward imbalance and job demand), on perceived work stress was considered.
In earlier work [3], 30% of non-white respondents reported very or extremely high stress compared to
18% of white workers. However, no further investigation was possible because of the small proportion of minority ethnic respondents.
Ethnic minority groups make up 8% of the UK population. This represents an increase over the last four decades. Minority ethnic groups also have a younger age structure than the white (UK born) population, reflecting past immigration and fertility patterns. They will therefore continue to rise as a proportion of the working population well into the 21st century [4].
There is considerable evidence that perceived stress at work is widespread [3] and is associated with ill-health
[5,6]. Recent figures showed that .2 million people in the UK reported suffering from work-related ill-health
[7]. It has also been suggested that ethnic minorities experience particular negative work environments which can lead to stress [8], and that the pattern of long-term illness among those of working age varies with ethnicity
[9,10]. Furthermore, there is a large literature on ethnicity and mental health [11]. However, there is very little
UK information about work stress and health in different ethnic groups. A recent review of ethnic minorities’ occupational health and safety identified the lack of

Ó The Author 2006. Published by Oxford University Press on behalf of the Society of Occupational Medicine.
All rights reserved. For Permissions, please email: [e-mail]

Downloaded from http://occmed.oxfordjournals.org/ by guest on January 13, 2013

More black African–Caribbean respondents reported high work stress than either Bangladeshi or white respondents. Reported racial discrimination among black African–Caribbean female respondents was strongly associated with perceived work stress. Among the black African–Caribbean respondents, women who reported experiencing racial discrimination at work had higher levels of psychological distress.

E. WADSWORTH ET AL.: RACIAL DISCRIMINATION AND WORK STRESS 19

Methods
The quota sampling method was based on that used in the EMPIRIC (Ethnic Minority Psychiatric Illness Rates in the Community) Survey [18] and aimed to achieve interviews with 200 respondents from each of the three ethnic groups. Altogether, 3181 households were selected on a 1:15 ratio from a database drawn from the Electoral
Registers and Post Office Address Files for five council wards in Hackney and Tower Hamlets, East London. The total population of the five wards was 47 722, of which
4708 (10%) were black African–Caribbean, 8998 (19%)
Bangladeshi and 26 456 (55%) white [9]. Interviews were attempted at 3176 households. Up to four recruitment attempts were made at each household. If the interview was refused, or the resident did not fit the research criteria, interviewers tried up to six households on either side of the identified one until a participant was recruited.
This occurred 2769 times. Inclusion criteria were being in paid work, aged between 18 and 65 years, self-reported ethnicity of black African–Caribbean, Bangladeshi or white (UK born). The team of 25 interviewers comprised men and women from each of the ethnic groups being investigated. Interviewer and interviewee ethnicity and gender were matched as often as possible [19].
The structured interview included demographics, health (psychological well-being), lifestyle, occupational factors and discrimination. The demographic, health and lifestyle factors are summarized in Table 1 (available as
Supplementary data at Occupational Medicine Online), and the nine occupational and two discrimination factors are described in more detail below.
1. Work stress: measured using a single item asking
‘In general, how do you find your job?’ with responses

2.

3.

4.

5.

6.

7.

8.

9.

on a five-point Likert scale (not at all stressful, mildly stressful, moderately stressful, very stressful, extremely stressful). Those responding very or extremely stressful were compared with those responding not at all, mildly or moderately stressful. This measure and cut-point have been validated [3] and used elsewhere [3,22].
Working hours: respondents were asked how many hours a week they worked on average, and those working $30 h were compared with those working less.
Contract: respondents were asked whether their job was permanent, temporary/casual or fixed contract, and those with permanent jobs were compared with the others.
Position: respondents were asked which of seven items best described their current position at work
[self-employed (251 employees in company), selfemployed (,25 employees), self-employed (no employees), manager (251 employees), manager (,25 employees), supervisor, employee]. Employees were compared with those who were self-employed, managers or supervisors.
Type: socio-economic class was determined from job title using Computer Assisted Standard Occupational
Coding [23]. Those with non-manual jobs were compared with those with manual jobs.
Treatment: respondents were asked ‘Do you feel that you have been treated unfairly at your present place of work, e.g. when applying for promotion?’. Those who answered yes were compared with those who said no.
Ethnic mix: respondents were asked ‘In general, what is the predominant ethnic or cultural origin of the people you work with?’, and responses recorded on a five-point scale. Comparisons were made between those who answered all or mostly different from you, and those who answered all or mostly the same as you or mixed—half and half.
Effort–reward imbalance [24]: two dimensions were computed—effort–reward imbalance [ratio of effort
(situational factors which make work more demanding) and reward (pay, status and opportunities for advancement)] and intrinsic effort (personal factors, such as motivation and commitment to work). For each dimension, respondents above and below the median were compared.
Job strain [25]: three dimensions were computed—job demand (pace and intensity of work), control (amount of control the worker has over work and the skill and variety involved) and support (support from colleagues and superiors). For each dimension, respondents above and below the median were compared.

Respondents were asked ‘Have you had any experience of discrimination at work (e.g. being treated unfairly) in your present job on the basis of the following: age, caste, class, culture, disability, ethnicity, gender, language, race,

Downloaded from http://occmed.oxfordjournals.org/ by guest on January 13, 2013

evidence on ethnicity and work-related health issues as a research priority [4].
It has been suggested that social and economic inequalities, underpinned by racial discrimination, are fundamental causes of ethnic inequalities in health [12].
Several studies have shown associations between racial discrimination and both physical illness [13,14] and psychological distress [15,16], and discrimination has recently been recognized as a psychological stressor and possible risk factor for physical illness [17]. The dearth of UK research on ethnicity, occupational stress and health is, therefore, further compounded by failure to consider the potentially powerful influential role of discrimination in these relationships.
This study aimed to establish rates of perceived work stress in three ethnic groups. In addition, it focused on the contribution of the reported experience of racial discrimination to the groups’ profiles of occupational and demographic associations with perceived stress, and assessed the association between perceived work stress and well-being.

20 OCCUPATIONAL MEDICINE

religion, sexuality, village kinship?’. Responses were yes or no. Two discrimination variables were derived:
1. Racial discrimination: those who answered yes to discrimination on the basis of race, culture or ethnicity were compared with those who said no.
2. Other discrimination: those who answered yes to discrimination on the basis of age, class, disability, gender or sexuality were compared with those who said no.

Results
In total, 626 people took part in the survey. After exclusions (those who were ineligible, non-contacts, etc.), the response rate was 55% (see Appendix 1, available as
Supplementary data at Occupational Medicine Online).
Two hundred and four (33%) respondents were black
African–Caribbean, 206 (33%) Bangladeshi and 216
(34%) white (UK born).
There were some significant differences between the groups in terms of gender [more black African–Caribbean respondents were female: 119 (58%) compared to 96
(44%) among white and 61 (30%) among Bangladeshi respondents, P , 0.001], marital status [more Bangladeshi respondents were married: 145 (71%) compared to 102
(47%) and 91 (45%) white and black African–Caribbean respondents, P , 0.001], education [more whites had a further education qualification: 132 (61%) compared to
105 (52%) and 73 (36%) among black African–Caribbean and Bangladeshi respondents, P , 0.001], income [more whites earned $£20 000: 113 (54%) compared to 64
(33%) and 36 (19%) among black African–Caribbean and Bangladeshi respondents, P , 0.001], work hours
[more Bangladeshi respondents worked part-time: 67
(33%) compared to 49 (24%) and 40 (19%) of black

Downloaded from http://occmed.oxfordjournals.org/ by guest on January 13, 2013

The study was approved by the East London and City
Health Authority Local Research Ethics Committee.
Associations between the dependent variable (work stress) and each of the independent variables were first assessed using chi-square tests and analysis of variance.
These analyses were carried out for each ethnic group separately, and for the sample as a whole. Any variables significantly associated with work stress (for any ethnic group or the whole sample) were included in backward stepwise logistic regression models used to determine associations with work stress for each ethnic group separately and for the whole sample. In these models, P ,
0.05 was used as the entry criterion and P . 0.10 as the removal criterion. The Hosmer and Lemeshow
Goodness-of-Fit statistic was fixed at P . 0.05 [26]. The association between racial discrimination and psychological distress was considered using analysis of variance.
The analyses were carried out using the Statistical Package for Social Sciences (SPSS for Windows, Release 12.0.2).

African–Caribbean and white respondents, P 5 0.0004], positions [more whites were managers, supervisors or self-employed: 89 (41%) compared to 50 (25%) and 47
(23%) among Bangladeshi and black African–Caribbean respondents], contract [fewer Bangladeshi respondents had permanent contracts: 148 (73%) compared to 186
(87%) and 167 (83%) of white and black African–
Caribbean respondents, P 5 0.002] and manual work
[more whites had non-manual jobs: 148 (69%) compared to 116 (57%) and 105 (51%) of black African–Caribbean and Bangladeshi respondents, P 5 0.001].
Overall, 13% of respondents reported experiencing high work stress: 18% (n 5 37) of black African–
Caribbean, 11% (n 5 24) of white and 8% (n 5 17) of
Bangladeshi respondents (P 5 0.01).
Eight per cent reported experiencing discrimination at work: 12% (n 5 24) of black African–Caribbean, 7%
(n 5 14) of Bangladeshi and 6% (n 5 14) of white respondents (P 5 0.09). The overall level is similar to the 7% who reported experiencing discrimination in an earlier study [3].
Initial analyses between perceived work stress and occupational and demographic characteristics were carried out for the sample as a whole, and for each ethnic group separately (Table 2). These suggested an association between work stress and ethnicity: more black African–
Caribbean respondents reported high work stress. Both gender and reported racial discrimination were associated with perceived work stress for the whole sample.
For gender, this reflected an association among black
African–Caribbean respondents only, while for racial discrimination, the association was apparent among black African–Caribbean and white (UK born) but not
Bangladeshi respondents.
Logistic regression analyses were then carried out for the whole sample and each ethnic group separately.
Reporting experiencing racial discrimination at work, gender, negative affect, contract, background noise and the work characteristics, effort–reward imbalance and job demand, were all associated with perceived work stress for the whole sample (Table 3). Among the white respondents, effort–reward imbalance and unfair treatment at work were associated with perceived work stress. Position and intrinsic effort were associated with perceived work stress among the Bangladeshi respondents. For the black African–Caribbean group, both reported exposure to racial discrimination and gender were associated with perceived work stress, together with negative affect and background noise.
Among the black African–Caribbean group, 69% of the 13 women who reported experiencing racial discrimination at work reported high work stress. This is considerably higher than the 20% among the 105 women who did not report experiencing racial discrimination.
A similar effect was suggested among the men (30% of the 10 who reported experiencing racial discrimination

Table 2. Associations between work stress and each of the independent variables by ethnicity
All respondents

White

191
184
165
309
229
508
32
34.86
292
248
328
181
86
111
75
263
7.31
493
47
389
142
425
109
379
158
180

(89)
(92)
(82)
(91)
(84)
(90)
(63)
(11.36)
(87)
(88)
(88)
(86)
(90)
(87)
(92)
(86)
(4.88)
(89)
(73)
(86)
(92)
(86)
(93)
(88)
(85)
(93)

24
17
37
32
45
59
19
36.86
42
35
45
30
10
16
7
44
10.65
61
17
65
13
69
8
50
28
13

(11)
(8)
(18)
(9)
(16)
(10)
(37)
(11.03)
(13)
(12)
(12)
(14)
(10)
(13)
(8)
(14)
(4.68)
(11)
(27)
(14)
(8)
(14)
(7)
(12)
(15)
(7)

222
136
494
41
319
221
9.88
0.10

(84)
(85)
(90)
(65)
(86)
(89)
(3.70)
(0.15)

41
24
55
22
50
28
12.91
0.26

(16)
(15)
(10)
(35)
(14)
(11)
(4.15)
(0.28)

10.42
31.36
12.89
388
152
406
134
347

(2.55)
(6.99)
(2.45)
(88)
(85)
(89)
(84)
(91)

12.14
30.53
11.80
52
26
52
26
36

(2.45)
(6.68)
(2.70)
(12)
(15)
(11)
(16)
(9)

193 (82)
439 (90)
100 (79)

42 (18)
51 (10)
27 (21)

Low work stress P

High work stress 106
85
182
9
38.18
91
100
87
100
34
23
17
116
7.09
169
22
153
38
163
27
112
79
90

(88)
(90)
(91)
(64)
(11.71)
(90)
(88)
(90)
(89)
(87)
(92)
(90)
(89)
(4.84)
(90)
(79)
(87)
(97)
(88)
(93)
(89)
(89)
(93)

14
10
19
5
36.71
10
14
10
13
5
2
2
15
11.04
18
6
23
1
22
2
14
10
7

(12)
(10)
(9)
(36)
(10.09)
(10)
(12)
(10)
(11)
(13)
(8)
(10)
(11)
(4.63)
(10)
(21)
(13)
(3)
(12)
(7)
(11)
(11)
(7)

75
25
177
13
132
59
9.84
0.09

(87)
(81)
(92)
(59)
(89)
(88)
(3.82)
(0.10)

11
6
15
9
16
8
13.75
0.36

(13)
(19)
(8)
(41)
(11)
(12)
(4.58)
(0.39)

10.61
33.08
12.85
143
48
162
29
132

(2.72)
(6.74)
(2.57)
(91)
(84)
(91)
(81)
(91)

13.00
30.29
10.71
15
9
17
7
13

(2.19)
(7.87)
(3.17)
(9)
(16)
(9)
(19)
(9)

Low work stress P

High work stress P

Low work stress High work stress 126
56
172
12
28.46
129
55
137
32
25
44
48
65
7.38
172
12
120
64
129
52
140
42
70

(92)
(92)
(92)
(86)
(7.22)
(91)
(95)
(93)
(89)
(96)
(88)
(96)
(90)
(5.03)
(91)
(100)
(90)
(96)
(90)
(98)
(94)
(84)
(95)

11 (8)
5 (8)
15 (8)
2 (14)
30.69 (5.81)
13 (9)
3 (5)
11 (7)
4 (11)
1 (4)
6 (12)
2 (4)
7 (10)
10.00 (5.29)
17 (9)
0 (0)
14 (10)
3 (4)
15 (10)
1 (2)
9 (6)
8 (16)
4 (5)

1.00

77
88
154
11
38.15
72
93
104
49
27
44
10
82
7.48
152
13
125
40
133
30
127
37
20

(92)
(75)
(86)
(48)
(11.73)
(79)
(84)
(81)
(79)
(87)
(85)
(77)
(79)
(4.77)
(85)
(54)
(82)
(82)
(81)
(86)
(83)
(79)
(91)

7
30
25
12
39.69
19
18
24
13
4
8
3
22
10.69
26
11
28
9
32
5
27
10
2

(8)
(25)
(14)
(52)
(12.40)
(21)
(16)
(19)
(21)
(13)
(15)
(23)
(21)
(4.52)
(15)
(46)
(18)
(18)
(19)
(14)
(17)
(21)
(9)

63
50
169
12
94
90
10.45
0.10

(90)
(89)
(92)
(80)
(90)
(94)
(3.30)
(0.14)

7
6
14
3
11
6
13.18
0.20

(10)
(11)
(8)
(20)
(10)
(6)
(3.70)
(0.18)

84
61
148
16
93
72
9.28
0.10

(79)
(84)
(85)
(62)
(80)
(84)
(3.89)
(0.19)

23
12
26
10
23
14
12.24
0.23

(21)
(16)
(15)
(38)
(20)
(16)
(4.05)
(0.21)

10.33
29.56
13.09
130
54
124
60
120

(2.28)
(6.80)
(2.22)
(92)
(92)
(91)
(92)
(95)

11.41
31.38
12.13
12
5
12
5
6

(2.35)
(6.38)
(2.13)
(8)
(8)
(9)
(8)
(5)

10.29
31.31
12.71
115
50
120
45
95

(2.62)
(7.02)
(2.54)
(82)
(81)
(84)
(76)
(85)

11.92
30.32
12.40
25
12
23
14
17

(2.55)
(6.12)
(2.39)
(18)
(19)
(16)
(24)
(15)

P

0.009

0.01
,0.0001
0.15
1.00
0.52
0.48

,0.0001
0.001
0.07
0.04
0.29
0.01

,0.0001
0.46
,0.0001
,0.0001
,0.0001
0.33
,0.0001
0.35
0.13
0.003

0.002

59 (84)
155 (90)
36 (84)

11 (16)
17 (10)
7 (16)

0.83
0.01
0.56
0.67
0.83
0.95

,0.0001
0.10
0.09
0.75
1.00
0.15

,0.0001
0.82
,0.0001
,0.0001
,0.0001
0.06
,0.0001
0.22
0.14
0.17

0.28

64 (85)
150 (94)
33 (83)

Downloaded from http://occmed.oxfordjournals.org/ by guest on January 13, 2013

White
Bangladeshi
Black African–Caribbean
Male
Female
No racial discrimination
Racial discrimination
Age: mean (SD)
Married or cohabiting
Other
,£20 000
$£20 000
No educational qualifications
O/GCSE level or equivalent
A level or equivalent
Higher
Negative affect: mean (SD)
No other discrimination
Other discrimination
Full-time job
Part-time job
Permanent contract
Other contract
Employee
Other
Colleagues ethnicity all/ most same
Mixed
All/most different
No unfair treatment
Unfair treatment
Non-manual work
Manual work
Intrinsic effort: mean (SD)
Effort–reward imbalance: mean (SD)
Job demand: mean (SD)
Control: mean (SD)
Support: mean (SD)
Night work seldom/never
Often/sometimes
Shift work seldom/never
Often/sometimes
Long or unsociable hours seldom/never Often/sometimes
Background noise seldom/never
Often/sometimes

High work stress Black African–Caribbean

11 (15)
10 (6)
7 (17)

0.34
0.23
0.41
0.50
0.39

0.04
0.60
0.19
0.07
0.04
0.48

0.13
0.32
0.002
0.009
0.07
0.30
0.11
1.00
1.00
0.02

0.05

70 (78)
134 (85)
31 (71)

20 (22)
24 (15)
13 (29)

0.002
,0.0001
0.48
0.47
0.70
0.65

,0.0001
0.001
1.00
0.63
0.67
0.34

0.01
0.58
,0.0001
0.001
0.001
0.43
0.51
0.84
0.23
0.21

0.05

E. WADSWORTH ET AL.: RACIAL DISCRIMINATION AND WORK STRESS 21

Low work stress Bangladeshi

22 OCCUPATIONAL MEDICINE

Table 3. Multivariable associations with work stress (only the last step of the analyses is shown)
OR

White
Unfair treatment
No
Yes
Effort–reward imbalance
Low
High
Bangladeshi
Position
Employee
Manager/supervisor/self-employed
Intrinsic effort
Low
High
Job demand
Low
High
Black African–Caribbean
Racial discrimination
No
Yes
Gender
Male
Female
Negative affect
Low
High
Background noise
Seldom/never
Often/sometimes

1.00
2.71

1.25–5.90

1.00
1.92

1.09–3.41

1.00
2.58

1.39–4.82

1.00
2.75

1.40–5.40

1.00
2.47

1.25–4.88

1.00
0.35

0.13–0.93

1.00
1.81

0.99–3.32

1.00
5.74

1.88–17.53

1.00
10.81

2.42–48.33

1.00
4.25

1.43–12.68

1.00
3.95

1.04–15.00

1.00
3.23

0.93–11.23

1.00
8.82

2.94–26.46

1.00
4.38

1.49–12.82

1.00
2.79

1.12–6.94

1.00
2.69

1.10–6.60

Discussion

reported high stress compared to 5% of the 74 who did not). These differences were also apparent when mean work stress scores were considered (negative affect and background noise were included as covariates). Women

A higher proportion of black African–Caribbean respondents reported high work stress compared to Bangladeshi and white respondents. Reported racial discrimination among black African–Caribbean females was strongly associated with high perceived work stress. These findings were strongly supported by data from a follow-up interview phase [19]. Previous work has also identified racial discrimination as a contributory factor to stress among minority ethnic workers [27].
The analyses also showed that, among the black
African–Caribbean respondents, women who reported experiencing racial discrimination at work had higher
GHQ scores. This suggests an association between the reported experience of racial discrimination and psychological distress among these respondents. It was not clear why this was not reflected in overall differences in psychological distress between the ethnic groups. However, the groups perceived work, work stress and racial discrimination in different ways [19], suggesting different expectations as well as possible differences in attributional and response styles.

Downloaded from http://occmed.oxfordjournals.org/ by guest on January 13, 2013

Whole sample
Racial discrimination
No
Yes
Gender
Male
Female
Negative affect
Low
High
Effort–reward imbalance
Low
High
Job demand
Low
High
Contract
Permanent
Temporary
Background noise
Seldom/never
Often/sometimes

CI

who reported experiencing discrimination had the highest perceived stress level [3.48 (SE 5 0.26)]. This was significantly higher than women who did not report experiencing discrimination [2.65 (0.09), P 5 0.003].
Similarly, men who reported experiencing racial discrimination at work had higher perceived stress levels than men who did not [2.91 (0.30) compared to 2.23 (0.11),
P 5 0.03]. Among those who did not report experiencing racial discrimination, women reported higher perceived stress than men (P 5 0.006).
Psychological distress was determined using the
GHQ-28 [20] and defined at the clinical cut-point of $5. Overall, 23% (n 5 139) met this criterion: 24%
(n 5 47) of Bangladeshi, 23% (n 5 48) of white and 22%
(n 5 44) of black African–Caribbean respondents.
Analyses suggested no association between psychological distress and ethnicity, but a clear association with work stress and racial discrimination. Fifty per cent
(n 5 36) of those who reported high work stress were also experiencing psychological distress compared to
19% who reported low work stress (n 5 102) (P ,
0.001). Similarly, 43% (n 5 22) of those who reported experiencing racial discrimination at work also had psychological distress compared to 21% of those who did not report experiencing racial discrimination (n 5
117) (P 5 0.001).
Focusing on the black African–Caribbean group, which had the highest proportion of respondents with high perceived work stress, showed that women who reported experiencing racial discrimination had the highest mean GHQ scores (Table 4).

E. WADSWORTH ET AL.: RACIAL DISCRIMINATION AND WORK STRESS 23

Table 4. Mean GHQ score by gender and racial discrimination among black African–Caribbean respondents
1. Male no racial discrimination Mean
SE
Overall F, P
P compared
P compared
P compared
P compared

2. Male racial discrimination 3. Female no racial discrimination 1.26
0.46

1.56
1.31

2.99
0.39

–
0.83
0.004
,0.001

0.83
–
0.30
0.002

0.004
0.30
–
0.001

4. Female racial discrimination 6.86
1.05

8.83, ,0.001 to to to to

1
2
3
4

up on this basis), so the study has no data about the nature of this discrimination. However, this does suggest the possibility of wider race relations’ problems.
The cross-sectional design was a limitation, and the geographically localized nature of the sample means that the generalizability of the findings is unclear. However, it is likely to be typical of these ethnic groups in UK inner cities. Replicating the analyses as far as possible in existing datasets from elsewhere in the UK [3,22] supported the main findings (data not shown). This emphasizes the importance of ethnicity in work practices and policy suggested by the findings reported here and the follow-up data [19].
Both non-response and response bias may also have affected the study. It is possible that some groups were less willing to take part, or that some respondents were less comfortable with certain aspects of the interview.
The study was also limited to three ethnic groups.
Further research is necessary to establish whether these associations are apparent in a wider range of ethnic groups and in more geographically dispersed populations.
Nevertheless, the study showed that more black
African–Caribbean respondents reported high work stress than either Bangladeshi or white respondents. Among black African–Caribbean females, the reported experience of racial discrimination at work was strongly associated with both perceived work stress and psychological distress.
This suggests that perceived work stress may be underpinned by reported exposure to racial discrimination at work among black African–Caribbean women, and that this may affect their psychological well-being.

Key point
• The experience of racial discrimination reported by black African–Caribbean female respondents is associated with perceived work stress.

Acknowledgements
This work was funded by the Health & Safety Executive.

Downloaded from http://occmed.oxfordjournals.org/ by guest on January 13, 2013

Several studies have recognized the important role of racial discrimination in health [11,13–16,27]. Recent UK research suggests that minority ethnic teachers experienced discrimination on a weekly or even daily basis, and almost half of them described their workplace as displaying signs of institutional racism [27]. This social and economic inequality, which is one of the fundamental causes of ethnic inequalities in health, may well be underpinned by racism [12]. Findings from the US also indicate that racial bias, in terms of both institutional and interpersonal prejudice at work, was a more important predictor of job satisfaction among African–American women than other occupational stressors, such as low decision authority and heavy workload [28], and that racial discrimination affects job stress among black women [29].
The implication that black African–Caribbean women may be particularly at risk of work stress and its consequences because of the potential for gender and/or racial bias has, therefore, been apparent for sometime. However, there has been very little UK research in this area; occupational stress research has almost always been carried out among predominantly white groups.
It was not possible to establish causation from the study’s cross-sectional data. However, it is possible that these complex relationships are at least in part what lie behind previous work indicating lower levels of psychological well-being among black women compared with white women [30].
Compared with black African–Caribbean respondents, fewer Bangladeshi respondents in this study reported experiencing racial discrimination at work. This was also supported by the follow-up data, and may in part reflect differences in their experience and perception of discrimination [19]. It may also be explained by a response bias (Bangladeshi respondents may have been less willing to discuss racial discrimination with an interviewer), an attributional bias (such as denial) or a difference between the groups, such as that more Bangladeshis work with others of the same or similar ethnic group.
Some white (UK born) respondents also reported racial discrimination at work. None reported discrimination at the second interview (despite being selected for follow-

,0.001
0.002
0.001
–

24 OCCUPATIONAL MEDICINE

Conflicts of interest
None declared.

References

Downloaded from http://occmed.oxfordjournals.org/ by guest on January 13, 2013

1. Siegrist J. The effort-reward imbalance model. J Occup
Health Psychol 1996;1:27–41.
2. Lazarus RS, Folkman S. Stress, Appraisal and Coping. New
York: Springer, 1984.
3. Smith A, Johal S, Wadsworth E et al. The Scale and Impact of Occupational Stress: the Bristol Stress and Health at Work
Study. Health & Safety Executive Research Report No.
265. London, UK: HSE Books, 2000.
4. Szczepura A, Gumber A, Clay D et al. Review of the Occupational Health and Safety of Britain’s Ethnic Minorities.
Health & Safety Executive Research Report No. 221.
London, UK: HSE Books, 2004.
5. Jones JR, Hodgson JT, Clegg TA et al. Self-Reported
Work-Related Illness in 1995: Results of a Household Survey.
Sudbury, UK: HSE Books, 1998.
6. Kearns J. Stress at Work: The Challenge of Change. London:
BUPA, 1986.
7. Jones JR, Huxtable CS, Hodgson JT et al. Self-Reported
Work-Related Illness in 2001/02: Results from a Household
Survey. Sudbury: HSE, 2003.
8. Walcott-McQuigg J. Worksite stress: gender and cultural diversity issues. Am Assoc Occup Health Nurses J 1994;
42:528–533.
9. Census. Office for National Statistics 1991.
10. Census. Office for National Statistics 2001.
11. Nazroo JY. Ethnicity and Mental Health: Findings from a
National Community Survey. London, UK: Policy Studies
Institute, 1997.
12. Nazroo JY. The structuring of ethnic inequalities in health:
Economic position, racial discrimination and racism.
Am J Public Health 2003;93:277–284.
13. Klonoff EA, Landrine H, Ullman JB. Racial discrimination and psychiatric symptoms among Blacks. Cultur Divers
Ethnic Minor Psychol 1999;5:329–339.
14. Troxel WM, Matthews KA, Bromberger JT et al. Chronic stress burden, discrimination and sub-clinical carotid artery disease in Black American and Caucasian women.
Health Psychol 2003;22:300–309.
15. Roberts RK, Swanson NG, Murphy LR. Discrimination and occupational mental health. J Ment Health 2004;13:
129–142.

16. Bhui K, Stansfeld S, McKenzie K, Karlsen S, Nazroo J,
Weich S. Racial/ethnic discrimination and common mental disorders among workers: findings from the EMPIRIC
Study of Ethnic Minority Groups in the United Kingdom.
Am J Public Health 2005;95:496–501.
17. Contrada RJ, Ashmore RD, Gary ML et al. Ethnicityrelated sources of stress and their effects on well-being.
Curr Dir Psychol Sci 2000;9:136–139.
18. Weich S, Nazroo J, Sproston K et al. Common mental disorders and ethnicity in England: the EMPIRIC Study.
Psychol Med 2004;34:1543–1551.
19. Smith AP, Wadsworth EJK, Shaw C et al. Ethnicity, Work
Characteristics, Stress and Health. Research Report No. 308.
Sudbury: HSE Books, 2005.
20. Goldberg DP, Hillier VE. A scaled version of the General
Health Questionnaire. Psychol Med 1979;9:139–145.
21. Eysenck H. Eysenck personality inventory/questionnaires.
In: Hersen M, Bellack A, eds. Dictionary of Behavioural
Assessment Techniques. Oxford, UK: Pergamon Press, 1988;
206–209.
22. Smith A, Wadsworth E, Moss S et al. The Scale and Impact of Drug Use by Workers. Health & Safety Executive Research
Report No. 192. London, UK: HSE Books, 2004.
23. Elias P, Halstead K, Prandy K. Computer Assisted Standard
Occupational Coding (CASOC). London: HMSO, 1993.
24. Siegrist J, Peter R. The effort-reward imbalance model.
Occup Med State Art Rev 2000;15:83–87.
25. Karasek R, Brisson C, Kawakami N et al. The Job Context
Questionnaire (JCQ): an instrument for internationally comparative assessments of psychosocial job characteristics. J Occup Health Psychol 1998;3:322–333.
26. Lemeshow S, Hosmer DW Jr. A review of the goodness of fit statistics for use in the development of logistic regression models. Am J Epidemiol 1982;115:92–106.
27. Miller GVF, Travers CJ. Ethnicity and the experience of work: job stress and satisfaction of minority ethnic teachers in the UK. Int Rev Psychiatry 2005;17:317–327.
28. Hughes D, Dodge MA. African American women in the workplace: relationships between job conditions, racial bias at work and perceived job quality. Am J Community Psychol
1997;25:581–599.
29. Mays VM, Coleman LM, Jackson JS. Perceived race-based discrimination, employment status and job stress in a national sample of Black women: implications for health outcomes. J Occup Health Psychol 1996;1:319–329.
30. Allen L, Britt DW. Black women in American society: a resource development perspective. Issues Ment Health
Nurs 1983;5:61–79.

References: Downloaded from http://occmed.oxfordjournals.org/ by guest on January 13, 2013 1 2. Lazarus RS, Folkman S. Stress, Appraisal and Coping. New York: Springer, 1984. 265. London, UK: HSE Books, 2000. London, UK: HSE Books, 2004. Sudbury, UK: HSE Books, 1998. 6. Kearns J. Stress at Work: The Challenge of Change. London: BUPA, 1986. Survey. Sudbury: HSE, 2003. Institute, 1997. Sudbury: HSE Books, 2005. Report No. 192. London, UK: HSE Books, 2004.