2

Religion and Culture Make a Difference

Religious inequality

In this book, ‘religious inequality’ (or ‘inequality of religion or belief’) is used

as a descriptive term to denote any variation in circumstances between
members of groups defined by religion or belief. We focus especially on
inequalities that exert major effects on the life chances of individuals. An
unequal situation does not necessarily imply unequal or unfair treatment,
since the latter judgement will depend on what has brought about the
unequal situation. Religious inequalities may, nevertheless, point up areas of
potential concern regarding injustices of religion or belief such as prejudice,
hate crime or discrimination, which are covered in more detail in the next
chapter. We argue in this chapter that it is necessary to achieve a balanced
view of the extent and causes of religious inequality in order to provide the
foundation for appropri- ate policy initiatives.

The focus In this chapter lies mainly on the minority (non-Christian) religions,
for three main reasons. First, concerns about inequality and unequal
treatment – have arisen mainly in relation to these religions, especially Islam.
Second, as argued in the previous chapter, it is some- times difficult to know
what meaning to attach to findings based on ‘Christian’ self-identifications,
especially in relation to Census data. Third, where Christian self-identifiers
provide a large proportion of the respondents (as in the Census data once
again), the average figures for any variable of interest are heavily influenced
by the ‘Christian’ totals for technical reasons. As a result, the statistics are
unlikely to reveal dis- tinctive inequalities in the situation of the Christian
population.

One of the main themes of this book is the emergence of religion as a

dimension of the equalities agenda in its own right, as distinct from
27

51 28 Ethnic, Racial and Religious Inequalities

Race on the one hand and ethnicity on the other. This point is reflected in the
fact that much of the information about religious inequality has been inferred
in the past from data drawn up in categories relating to race and/or ethnicity.

The extent of these inequalities related to race (and to some aspects of

ethnicity) had already been well established by the 1980s, either in general
terms (Brown and Gay, 1985) or in specific fields. This literat- ure, which
extends to the present, has been covered in detail elsewhere (Modood et al.,
1997; Macey, 2009), but some of the main findings are summarised here as
significant features of the context within which the issue of religious
inequality has arisen.

The current disadvantaged position of BME groups can be traced back to

their initial (post-WWII) settlement in Britain in response to recruitment
drives in the Caribbean and the Indian subcontinent. For not only was no
provision made for their arrival in terms of, for example, housing, but they
were confronted by deeply entrenched (institu- tional) racism on all
dimensions of ‘social wealth’ – health (Gordon and Newnham, 1986),
including mental illness (Burke, 1988); housing conditions (Association of
Metropolitan Authorities, 1985; Brown and Gay, 1985; Greater London Action
for Race Equality, 1987); education (All Faiths for One Race, 1982;
Department of Education and Science [DES] [the Swann Report], 1983; All
London Teachers against Racism and Fascism, 1984; Troyna and Williams,
1986; Troyna, 1993); employment (Allen et al., 1977; Association of
Metropolitan Authorities, 1985; London Association of Community Relations
Councils, 1985; Newnham, 1986; Brennan and McGeevor, 1987; Gifford et
al., 1989); social services and social security (Gordon and Newnham, 1985;
Connelly, 1987; Rooney, 1988); and the criminal justice system (Home
Office, 1981; Gilroy, 1982; Commission for Racial Equality [CRE], 1987;
Pilger, 1988; Tompson, 1989).
Nor was this position subject to rapid improvement. Brown and Gay’s
national research in 1985 showed that there had been no overall change in
the geographical or economic position of black Britons since they first arrived
as immigrant workers in the 1950s and 1960s. And in 1990, Allen and Macey
observed that the position of many of their children had actually worsened,
partly due to the economic situation, and despite the implementation of
three Race Relations Acts (in 1965, 1968 and 1976).

More recent work has confirmed that these inequalities are stub- bornly
persistent, and members of minority ethnic groups remain, on the whole, in
more deprived positions than their white counterparts on many indices of
social wealth. They continue to experience worse employment prospects, for
example, and are paid less. Bangladeshi men have the highest
unemployment rate, at 20%-four times that of white men (ONS, 2002) – and
Pell (2007) has emphasised the sixteen-point gap in economic participation
rates between the BME and white popu- lations, which are 60% and 76%
respectively. However, this headline figure glosses considerable variation in
the rates for different groups, which are discussed further below. These
findings are supported by a careful analysis of the Census data for 1991 and
2001 from Clark and Drinkwater, who conclude that ‘in spite of increasing
employment rates for most ethnic minorities, large employment deficits
remained in 2001, even for those ethnic groups whose employment rates
had risen' (2007: 45). The results, nevertheless, varied between groups: ‘for
Indian and Chinese men, there is very little “problem” as far as employment
pen- alties are concerned, while for Pakistani, Bangladeshi and Black men,
these are severe’ (Clark and Drinkwater, 2007: 47).

In relation to earnings, Berthoud (2000) refers to an ‘ethnic deficit/ penalty’

in male weekly pay, showing that when such factors as educa- tion and age
are controlled for, there remains an ethnic penalty of 7% for Indians, 24% for
Caribbeans, 36% for Bangladeshis and Pakistanis and 40% for Africans. Clark
and Drinkwater concur that ‘earnings penalties are a fact of life for ethnic
minorities in the UK labour mar- ket’ (2007: 47). Platt (2007b) points to the
gendered nature of pay dif- ferentials, demonstrating substantial pay gaps
not only between men and women, and between majority and minority
ethnicities, but also between minority men and women, with an average
weekly pay gap for minority women of 20% over three times as high as that
suffered by minority men, at 6%. The EOC’s (2007) statutory investigation
into Bangladeshi, Pakistani and black-Caribbean women and work in Britain
identifies five ‘employment gaps’: low participation rates, unemploy- ment,
lack of progression to senior levels, the pay gap and occupational
segregation.

On other dimensions, too, gaps remain between members of majorities and

minorities, though clearly some of these are linked to employment and
earnings or social class status and the differ- ences in provision that are
correlated with this, as in the ‘postcode lottery’ (Janes and Mooney, 2002).
These include poverty, with over half of Pakistani, Bangladeshi and black
African children growing up in poverty (Platt, 2002; 2007a,b,c). In health,
there are disparities in health status, lifestyle and access to services on the
basis of ethnicity/ religion (Heath Education Authority, 2000), as well as
differential treat- ment (CRE, 2004; Scottish Health Survey [Bromley et al.,
2005]; Social

5230 Ethnic, Racial and Religious Inequalities

Exclusion Unit, 2005). And both the provision and use of health and social
services vary by ethnic religious and racial group (Chambra et al., 1998;
Ahmad et al., 2000; Atkin et al., 2002; Katbamna et al., 2002; Chahal and
Ullah, 2004). Minorities are over-represented in the mental health system
(Nazroo, 1997; Sproston and Nazroo, 2002; Sproston and Mindell, 2006;
Healthcare Commission, 2007). In housing there is rela- tive overcrowding,
poor conditions and poor neighbourhoods (Sellick, 2004; Beckford et al.,
2006), as covered in greater detail below. In educa- tion, while gaps remain
between majority and minority ethnic groups, these vary by both gender and
ethnicity/religion, with some minori- ties (such as Indians and the Chinese)
doing as well as, or better than, their white counterparts, while Pakistanis are
underachieving, though girls are beginning to catch up (Abbas, 2002; 2003;
Equal Opportunities Commission, 2007; T. Phillips, 2007).

In the criminal justice system, there are a number of areas of concern: on the
one hand, black and Asian people are over-represented as far as being
stopped and searched by the police and receiving custodial sentences when
charged with offences (Jones and Singer, 2008); they are also more likely
than white people to be the victims of racially motivated violence, include-
ing homicide (Jones and Singer, 2008; Institute of Race Relations, 2008;
2009). On the other hand, they remain under-represented in employment in
the criminal justice system, particularly at the higher levels.

Overall, then, there is clear evidence of inequalities related to race and

ethnicity, covering a broad range of social indicators and types of service
provision. The evidence also reveals the differentiation between race and
ethnicity, where the situations of members of the same racially identified
groups vary according to their ethnicity.

Since religion is correlated with both race and ethnicity, it has always been
possible to infer the existence of religious inequality from the data available
on race and ethnicity. But it is only in the past fifteen years that studies have
begun to make explicit reference to religion. And it is even more recently
that larger-scale or more comprehensive studies have appeared, using
quantitative or qualitative methods that are more sophisticated and/or more
reliable in their conclusions. These ‘pioneer studies’ help to provide a clearer
picture of the distinctive profile of reli- gious inequality, and form the main
focus of attention below.

Religious disadvantage

Following Weller et al. (2001), the term ‘religious disadvantage’ (or ‘dis-
advantage of religion or belief’) will be used to denote the position of
Religion and Culture Make a Difference 31

Particular groups of religion (or belief) that are treated less favourably than
other groups through lack of equal recognition in official or insti- tutional
contexts, including the law. There are a number of clear-cut examples of
religious disadvantage in Britain.

The most significant examples are the constitutional ones. The Church of
England is the established church in England, and in Scotland the
(Presbyterian) Church of Scotland is the national church. The Church in
Wales, which belongs to the Anglican Communion, is, however, dis-
established. Given its position as the established church, the law of suc-
cession to the Crown privileges Anglicanism, and the House of Lords includes
only Anglican representatives of religion by right. In these respects, all non-
Anglican religions suffer disadvantage in Britain (and non-Presbyterian
groups similarly in the Scottish context).

There are a number of ways in which non-religious believers experi- ence

disadvantage. As a result of the 1996 Education Act, members of the
alternative spectrum of belief (who make up the majority of the population)
do not have their views reflected in the school curriculum, since the teaching
of non-Christian beliefs tends to comprise ‘world faiths’ (i.e. religions), rather
than agnostic or secular belief systems. Secularists tend not to be
represented on such decision-making bodies as School Boards of Governors,
and in some publicly funded initiatives such as the Inter-faith Network (which
if truly inclusive would become a ‘Network of Faith or Belief’). Humanist
weddings are not recognised in England and Wales. The Charities Act 2006
(which applies only to England and Wales) defines a charity in part by
reference to purposes that include the advancement of religion if it is for the
public benefit (Jahangir, 2008), but it excludes the similar advancement of
secular belief systems. The decision by the BBC to exclude secular speakers
from its radio programme ‘Thought for the Day’ offers a minor example of
this kind of disadvantage.

More broadly, provision is less frequently made for non-Christian

denominations (and for non-religious belief systems) than for their Christian
counterparts in connection with hospital and prison chap- laincies and
facilities for the practice of faith within these institutions. A national survey
of multi-faith chaplaincy arrangements in hospitals in England and Wales
showed that the majority of full- and part-time chaplains were Christian
(93.3% and 91.4% respectively); 6.7% of full- time chaplains were Muslim,
and of the remaining 8.6% part-time chap- lains few were from non-Christian
faith groups (Sheikh et al., 2004). A similar situation applies in Scotland,
where most chaplains have been appointed and employed historically by the
Church of Scotland’s
53 32 Ethnic, Racial and Religious Inequalities

Board of National Mission (Levison, 2005). Nevertheless, it should be

emphasised that the Muslim share of hospital chaplains is higher than their
overall share of the population, so that any verdict of proven dis- advantage
needs to be approached with caution. The NHS has issued guidance on
meeting the religious and spiritual needs of patients and staff in hospitals
which requires the provision of suitable ’spaces’, access- ible 24 hours a day
(Department of Health [DoH], 2003). Spalek and Wilson (2002) looked at
religious provision in prison with respect to access to imams. Because the
penal system is basically oriented towards Christianity, particularly the
Church of England, imams are placed in a relationship of dependency with
Anglican chaplains and are marginal- ised from decision-making
mechanisms.

Since religious disadvantage (in the sense used here) by definition involves
unequal treatment on the grounds of religion or belief, it cre- ates a prima
facie case for the existence of religious injustice. The remedy for this
injustice is also clear-cut, at least in principle: give equal recog- nition in
official contexts to all varieties of both religion and belief. What this might
imply in practical terms is discussed in Chapters 3 and 6. It is, however,
worth noting at this stage that the principle of equal advantage leaves open
a wide range of measures. It is consistent, for example, with secularism,
which gives no recognition to any religion or belief system. This chapter
returns to the discussion of the social and economic dimensions of religious
inequality.

Religion, economic activity and social class

Brown’s (2000b) investigation into the economic activity of South Asians in

Britain is the most prominent early example of a study in which reli- gion is
made an explicit variable, and which, significantly, predates the 2001
Census. The study is based on the 1994 National Survey of Ethnic Minorities,
with a large sample (N = 3,795). This is, however, a 'pioneer study’ not only
in the chronological sense, but because Brown’s findings set the agenda for
subsequent research, in three respects especially.
First, Brown’s findings show that ‘neither ethnic nor religious group
boundaries alone capture the labour market participation among South
Asians. Thus, among Indian men, Sikhs experience more than dou- ble the
unemployment of Hindus, while among Muslims, Pakistanis and Bangladeshis
record more than double the unemployment level of Indians’ (2000b: 1047).
This finding has been borne out by more recent work, with unemployment
rates among Muslim men of working age (between 16 and 64) ranging from
nearly 30% for black African Muslims

Religion and Culture Make a Difference

33

To between 15% and 20% for Bangladeshi, Pakistani and white British
Muslims, down to just over 10% for Indian Muslims. Unemployment rates for
Muslim women show a similar range and pattern by ethnicity (ONS, 2006).

These results give credence to the idea that ‘hyphen-groups’ of joint racial,
ethnic and religious composition might form the most appro- priate units of
socio-economic analysis, since it is ‘Indian Muslims’, say, rather than
‘Indians’ or ‘Muslims’ who have a distinctive profile of activity. There is an
interesting pattern here, reminiscent of the con- sequences of the earlier
trend to distinguish ‘ethnicity’ more explicitly from ‘race’, which was noted
above. When ‘religion’ is treated as dis- tinct from both ethnicity and race, it
appears that religious background differentiates the social outcomes for
those of the same ethnic/racial groups. Conversely, ethnicity and/or race
differentiate the social out- comes for those of the same religious
background. This reinforces the case for treating communal (RER) identity as
composed of race and ethnicity and religion, since it seems that all three
elements must be addressed in order to generate an adequate picture of
social inequality: in particular, religion does make a difference.

Second, Brown comments: ‘whilst it is relatively straightforward to

demonstrate differences between religious groups in economic activity, it is
harder to establish causality’ (2000b: 1058) and ‘proof of religious
discrimination is difficult to establish in the absence of hard statistics’
(2000b: 1037). This may either be because religion acts as a proxy for other
factors relating to culture of origin, or because it has operated historically by
contributing ‘indirectly to differences in factor endow- ments of migrants
entering Britain’ (Brown 2000b: 1059). This under- lines the point made at
the beginning of the chapter, that the issue of religious inequality is distinct
from the issue of religious injustice.

Brown’s conclusion on the issue of discrimination is worth quoting in full

because of its broad relevance:

A more direct effect [of religious affiliation] may be at work in the form of
religious discrimination. The relatively disadvantaged posi- tion of Muslims, in
particular, may tie in with the view (discussed earlier) that Muslims (and
Islam) are increasingly experiencing a neg- ative profile within the British
media and society at large. However, the revealing of a relatively
advantaged Indian Muslim minority would seem to challenge this thesis
(though there may be differences between Indian, Pakistani and Bangladeshi
Muslims in the expres- sion of their faith that might have a bearing on their
identification as

34

Ethnic, Racial and Religious Inequalities

Muslim and exposure to discrimination). Finally, a direct effect may operate

through attitudinal and aspirational differences between religious groups that
impinge on, and may affect the ‘success’ of, labour market participation.
(2000b: 1059)

This quotation makes the third point that can be gleaned from Brown’s work.
In general, religious inequalities are likely to be caused by a com- bination of
(a) factors internal to a group and/or its members (include- ing cultural
values and economic endowments) and (b) factors in the external social
environment of the group and/or its members (including religious, racial
and/or ethnic discrimination). The relative contribu- tions of these different
factors will vary from case to case.

Lindley (2002) conducted a similar analysis on the same data set, but using
slightly different statistical techniques. Her conclusions, perhaps
unsurprisingly, were in broad agreement with Brown’s. She did, how- ever,
undertake a specific analysis on the position of Muslims, which concluded
that ‘Muslims do experience some unexplainable employ- ment penalty,
relative to other non-white religions, over and above all other characteristics
(including ethnic differences and language flu- ency)’ (2002: 438). This is
consistent with the operation of direct or indirect discrimination in the
employment relationship, but the results also suggest that ‘Muslims appear
to be less assimilated and/or have less transferable human capital than other
non-whites’ (2002: 439). This bears out the general conclusion reported in
the previous paragraph on the need to combine internal with external
factors.

Social class or ‘socio-economic status’ as it has become known more recently

– is important because of the ways in which class positions pro- vide
resources that affect not only employment opportunities but also a wide
range of other life chances. Census data shows clearly how class
differentiation not only varies between religious groups but depends also on
ethnicity and gender. Thus, both men and women from Jewish and Buddhist
groups have substantially higher proportions of middle- class (managerial
and professional) occupations than the population at large. Jewish men and
women have very low representation among working-class occupations
(semi-routine and routine). There are more Buddhist men and women in such
working-class occupations, but still fewer proportionately than the national
averages.

By contrast, Muslim and Sikh men and women are less likely to have middle-
class occupations, and more likely to have working-class ones. There is,
however, some variation by ethnicity within religion, with Indian Muslim men
having a somewhat more favourable employment Religion and Culture Make
a Difference 35
Profile than Pakistani Muslim men (in terms of the conventional rank- ing of
social class positions), who in turn have a more favourable profile than
Bangladeshi Muslim men. Interestingly, Muslim women’s employ- ment
profiles show less ethnic variation than Muslim men’s, which sug- gests that
ethnicity makes less difference to women’s employment than to men’s
employment (ONS, 2006). Detailed information on these class differences by
religion is given in the Statistical Appendix (Table 3). Importantly, the results
given for Jewish and Buddhist groups demon- strate that minority religious
status per se is not necessarily associated with social deprivation in Britain.

Khattab’s (2009) study sheds further light on the relationship between

religion and social class by introducing educational achieve- ment as an
intervening variable. The study is based on a sub-sample of the Census data
for England and Wales, but it still has a huge sample size (N = 250,000). The
study compares educational achievement with relative occupational status
for a series of groups, taking the (majority) Christian White British (CWB)
group as the reference group. The find- ings create an interesting and
complex picture of the situations of dif- ferent groups, as follows:

(a) Above-average educational qualifications and higher occupational

status than the level of qualifications would predict: Jewish White
British; Non-Religious White British and Other White British.

(b)Above-average educational qualifications and the same оссира- tional

status as education would predict: Hindu Indians.

© Above-average educational qualifications and lower occupational status

than education would predict: Christian Black Caribbeans; Christian Black
Africans.

(c) Below-average educational qualifications and the same occира- tional

status as education would predict: Muslim Indians.
€ Below-average educational qualifications and lower occupational status
than education would predict: Muslim Bangladeshis; Muslim Pakistanis;
Muslim Whites; Sikh Indians.

It can be seen that the groups are differentiated sharply in terms of edu-
cational qualifications, in ways that mirror some, but not all, of the gen- eral
correlations between religion and social class. The White Jewish, No Religion,
Hindu Indian and Black Christian groups have higher quali- fications than
average, whilst the Muslim and Sikh groups have lower than average. The
picture changes, however, when it comes to converting these qualifications
into an employment status corresponding with the 36 Ethnic, Racial and
Religious Inequalities

Level of qualifications. Here the White Jewish and White No Religion groups
achieve a higher occupational status than their qualifications would lead one
to expect (relative to the majority reference group). These groups are under-
qualified for their class positions. By contrast, the Black Christian, Sikh and
Muslim groups (apart from the Indian Muslims) are over-qualified for their
class positions, and have not converted their edu- cation into occupational
status as one might anticipate. The two Indian groups (of Hindus and
Muslims respectively) convert qualifications into occupational position in a
manner that is very similar to the reference group, but at two different
educational levels.

This pattern of findings certainly reinforces the ‘RER’ perspective once again,
since it shows the contributory effects of race and ethnicity and religion on
the fate of group members. But these findings also high- light the difficulties
in inferring causal linkages from statistical correla- tions. It is tempting to
think that both Sikhs and (non-Indian) Muslims suffer from religious and/or
racial discrimination in employment, which helps to explain why they are
unable to gain the jobs to which they are suited by their educational
qualifications. A similar inference would suggest that black Christians face
racial discrimination in the labour market. But the same argument would also
imply that Indians do not face any discrimination in access to employment
(whether they are Muslims or Hindus) and that, for example, there is positive
discrim- ination in favour of white Atheists (those of ‘No Religion’) in Britain.
Since the latter inferences appear unlikely, there is reason to doubt the
previous inferences as well, because they are derived from analogous
reasoning. The very different social positions of those who are not eas- ily
distinguishable in terms of appearance, such as Indians (Christians, Hindus
and Sikhs) relative to Bangladeshis and Pakistanis (Muslims), may give pause
for thought about the causal influence of ethnic or reli- gious discrimination.
Or consider the possible causes of the position of white Muslims. Does
(something about) their religious identity lead to below-average
qualifications and/or lower occupational achievement for this group?
Alternatively, are those whites who become Muslims more likely to do so
because of these deficits? Or is there some independent factor that is
correlated with the identity and which causes the deficits? The data by itself
cannot decide between these possibilities. As we will argue in the next
chapter, determining the incidence of discrimination requires methods that
are directed specifically at the justice issue, as opposed to the inequality
issue.

The furthest one can go at this stage is to follow the findings of both Brown
and Lindley and conclude that it is a combination of external Religion and
Culture Make a Difference 37

Factors (including the possibility of racial, ethnic and religious discrim-

ination and the effects of local labour market conditions) and internal factors
(including Brown’s ‘attitudinal and aspirational differences’ and a range of
cultural differences) that are responsible for the observed pat- terns. Clark
and Drinkwater (2007) make a similar point whilst discuss- ing the
‘explained’ and ‘unexplained’ effects in their models, where the residual
(unexplained) terms include the possible effects of discrimina- tion. On the
other hand, Khattab says that ‘it is probably not the colour of skin, although
it is still important, but the meaning and values that the dominant group
attaches to the skills and who possesses them [that makes the difference to
occupational achievement)’ (2009: 12). This comment seems to exclude the
influence of the skills differentials them- selves, as opposed to the perception
of these differentials, in achieving higher occupational status. It thus
downplays the possible effects of the internal factors (skills as indicated by
qualifications) in favour of exter- nal factors (quasi-racist perceptions by ‘the
dominant group’).
It should be emphasised that these results concerning average pro- files are
compatible with a full range of variation of class positions within each
religious group. Minority religious communities have all been established
long enough in Britain for class differentiation to have developed within
them. And some of these differences may be founded on the differential
possession of social and/or economic capital at the time of their settlement
in Britain, acquired in varied countries and (urban or rural) regions of origin.
The position of the Ugandan Asians has already been mentioned in this
respect, and the point is underlined in the quotation from Brown above:
migration can exert social effects long after the process of physical relocation
has taken place.

It follows that there are issues of economic inequality within each ethno-
religious community as well as between a given community and the rest of
society. It would be very helpful to be able to pursue this analysis into the
highest reaches of the class system, for example, and to invest- igate the
ownership of property and businesses as well as employment status. Sadly,
there is relatively little research evidence on this issue. It is reported,
however, that fewer than 3% of FTSE 100 directors were from minority ethnic
backgrounds in 2004 (Clark and Drinkwater, 2007). This figure can be read in
two ways. On the one hand, this proportion is considerably less than the
minority ethnic representation in the popu- lation at large (8%); on the other
hand, there clearly are a number of high net worth individuals in Britain from
these backgrounds. Ansari (2002) reports that there are over 5,000 Muslim
millionaires in Britain, with combined liquid assets above £3.6 billion.

56 38 Ethnic, Racial and Religious Inequalities

The other difficulty with estimating the economic position of members of

some ethno-religious communities lies in the effect of established communal
or kin-related practices for the redistributions of resources. Bolognani
(2007a) and Ballard (2004) discuss the trans- national dispersion of funds to
relatives in the region of origin, for example and the importance of a cash
economy among Pakistani Muslims, both of which must be taken into
account to estimate their overall economic position. It is clear, nevertheless,
that there is a full range of variations of wealth or social status within each of
the reli- gious groups, so that there are many wealthy Muslims and numerous
socially disadvantaged Jews.

We emerge from this section with three points that are essential to a
balanced view of the relationship between religion and equality:

1. Hyphen-groups of combined racial, ethnic and religious identity are

often the salient entities of social action or empirical comparison;

2. Injustice is distinct from inequality;

3. Religious inequalities are caused in general by a combination of fac-

tors external to and factors internal to religious communities.

In the next section, we consider two examples of what can go wrong in policy
terms if these essential points are not taken to heart.

60/76 and Moving on Up? A cautionary tale

In 2007, Gordon Pell, who was then the Chief Executive of Retail Markets for
the Royal Bank of Scotland, presented a report into minority ethnic
employment from a prestigious body called the Business Commission on
Race Equality in the Workplace. He had been selected by the Prime Minister,
Gordon Brown, to chair this Commission on behalf of the UK’s National
Employment Panel and to ‘advise Government on poli- cies and practical
measures to increase the recruitment, retention and progression of ethnic
minorities in the private sector’ (Pell, 2007: 7).

In the Foreword to the report, Pell explained that:

We have entitled this report ‘60/76’ because it is important to get one
striking fact into as many minds as possible. 76% of white people in the
working age group have a job; only 60% of working age people from ethnic
minorities are in employment. This gap has remained largely unchanged for
a generation. It is unjust, socially divisive, and bad for the economy. (2007:
3) Religion and Culture Make a Difference

39

The report goes on:

Britain takes itself to and among thest tolerant and open-minded countries in
the wave up to this ideal most economically efficended we will never
persists. It is unfair, it flust the ethnic minority Buployment gap persists from
the workplacements social strife; it emploes productive talent from the
workplace; and it is not going to go away on its own. (2007: 5)

The report accordingly recommends that the government establish the long-
term goal of ‘eradicating the ethnic minority employment gap [of Joxteen
percentage points] within twenty-five years’ (Pell, 2007: 21). A number of
quite forceful legislative and administrative measures are then
recommended as ways of closing this gap, which enlist the powers of state to
improve the performance of both the public and the private sectors. An
action plan is established, with detailed quarterly targets running from the
year of the report up to 2015, under the ultimate supervision of the
Chancellor of the Exchequer.

The premises of this strategy are evidently (a) that the sixteen-point gap is
caused largely by factors in the social environment of ethnic minorities,
especially labour market discrimination (‘it is unjust’; ‘it is unfair’) and (b)
that the state and private employers acting together in a concerted fashion
have the power between them to close the gap.
Each of these assumptions is highly questionable. In the first place, it
transpires that there is not ‘one striking fact’ about the employment gap that
needs to be lodged in ‘as many minds as possible’, but two strik- ing facts.
This is because the employment gap is very different for men and for
women. For men, it is not sixteen points, but eleven points; for women, it is
no less than twice that figure, at twenty-two points (figures estimated from
the data given at Pell, 2007: 56). The headline figure of six- teen points
emerges as the average of these two very different percentages. And the
reason for the difference between the genders is that the employ- ment rates
for Pakistani and Bangladeshi women are both 24%, which represents an
employment gap of no less than 49% compared to the cor- responding figure
for white women (which is 73%), and 38% compared with Indian women
(whose employment rate is 62%). Since it is reasonable to assume that
Pakistani and Bangladeshi women have Muslim religious backgrounds, it is
clear that there is a significant component of religious

Inequality within the ethnic minority employment gap for women. The
purpose here is not to challenge the goal of eliminating the employment gap
for women far from it but rather to appeal for

157 40

Ethnic, Racial and Religious Inequalities

A higher level of sociological awareness regarding its achievement. In

particular, the elimination of the gap implies a three-fold increase in the
employment rates of the Pakistani and Bangladeshi women con cerned, from
about one-quarter to about three-quarters. This would require a substantial
shift in a range of attitudes and practices within reeir communities. Clark and
Drinkwater’s conclusion on this issue of women’s employment, under the
heading ‘religion matters’, is worth citing in full:

Our regression models suggested that religion is an additional source of

variation in labour market behaviour. In particular, there is some evidence
that, controlling for other factors, Muslims have lower employment rates
than individuals with another, or no, religion. Quantifying this is
problematical for some of Britain’s ethnic groups simply because ethnicity
and religion are highly correlated. Cultural attitudes and norms may underlie
some of the low employment rates, especially for Pakistani and Bangladeshi
women, but separating the influences of ethnicity and religion is extremely
difficult, both con- ceptually and empirically. It is also true that it may be
tradition, rather than religious belief per se, that influences attitudes to
female labour force participation and childcare. (2007: 48)

Given that the causes for the lower employment rates involve a com-
bination of (a) external circumstances (including RER and gender dis-
crimination) with (b) individual choices conditioned by religion and/or culture,
any policy aimed at increasing participation rates would need to address
both sides of the question. Insofar as the causes are internal to the ethno-
religious communities concerned, it is beyond the power of external
employers to bring such changes into being, however formid- able they may
be as economic organisations. Moreover, the campaign to close the gap
would at some point encounter religious opposition (or at least opposition
phrased in religious terms), and it might be inconsist- ent in some respects
with other aspects of government policy such as multiculturalism or
community cohesion (or so the argument would go, at any rate). It is thus
unrealistic to announce a government programme to eliminate inequality
that rides roughshod over all these social factors and political considerations.

As a second point, it is interesting to note that Pell proposes to com- mission

direct research into the extent of employment discrimination – ‘Benchmark
Discrimination Testing’. This is indeed essential to cre- ate an accurate
picture of the minority ethnic employment position, Religion and Culture
Make a Difference 41

But Pell’s recommendations overlook the fact that the appropriate tar- bet for
the gap-closing campaign depends on the results of this research, getich has
yet to be undertaken. The report therefore jumps the gun by announcing the
target as 60/76 simeste gap caused by exter- bal labour market
discrimination is almost certainly much smaller nal lahe sixteen-point
discrepancy up to a third of the gap may be accounted for by the situation of
Pakistani and Bangladeshi women afone. In short, the Pell Commission has
misidentified the target for policy (since it has been chosen in advance of the
facts) and mistaken the means of reaching it (since closing the gap is beyond
the power of the measures it proposes).

The Important conclusion for the purposes of this book is that these failings
in policy can be traced back directly to faults in the underlying model implicit
in the work of the Commission: a one-sided focus on the external
(environmental) factors at the expense of internal (communal) factors,
coupled with the habit of conflating inequality with injustice. Both these
faults are characteristic of what may be termed an ‘ideologi- cal’ stance in
this field, as opposed to an analytical stance.

The same points are illustrated by the statutory investigation into

Bangladeshi, Pakistani and black-Caribbean women and work in England,
Scotland and Wales conducted by the EOC, and published under the title
Moving on Up? The researchers actually stated at the outset that they were
determined to ‘move away’ from cultural, eth- nic or religious influences to
focus on workplace discrimination (EOC, 2007: 4). This is to endorse
explicitly the focus on external factors at the expense of internal factors that
is characteristic of the Pell Report. Subsequently, however, the authors list
under ‘work based discrimina- tion’, such factors as lack of language skills,
overseas qualifications and lack of support for family and childcare
responsibilities. These factors are all related (in various ways and to varying
extents) to circumstances internal to community life, including migration
histories and family roles. This research adds other elements to the inventory
of problematic assumptions, by relying entirely on perceptions, rather than
evidence, of discrimination, as well as including many ‘barriers to
employment’ which apply to both majority and minority women, and so may
be the consequences of gender discrimination rather than ethno-religious
dis- crimination.

These two studies from official bodies display what we will call in this book
‘conceptual inflation’. Conceptual inflation stretches the application of value-
laden concepts such as injustice or discrimination beyond their proper
spheres of reference. This typically has the effect 42. Ethnic, Racial and
Religious Inequalities
Of misrepresenting the incidence of social harms or wrongs, by exag-
gerating some and masking others. We will see in Chapter 3 that the
examples of conceptual inflation in these two sources are not isolated
instances – the problem recurs throughout the contemporary literature on
religion and equality.

The localisation of minority religions

One of the main findings from the 2001 Census is the extent to which the
minority religious (non-Christian) populations are localised in par- ticular
areas of the country. These populations tend to live in particular towns or
cities in which members of their own group are represented much more
frequently than their average throughout the country, and to live in
neighbourhoods within these cities in which members of their own group are
represented much more frequently than their aver- age throughout the city
as a whole. A similar point even applies at a national level, to the constituent
nations of the UK.

These facts of social separation do not in themselves constitute an issue of

inequality, but they are connected with the concerns of this book in several
ways. First, the social and economic prospects for mem- bers of minority
religions will depend in part on the areas in which they live. Second, the
levels of social separation condition (though they do not determine) the
prospects for social interaction and cultural exchange between different
religious traditions, and the corresponding dynamics for the formation of
ethno-religious communities (Carling, 2008). Third, the social, political and
administrative implications of religious diversity are experienced much more
directly in some parts of the country than in others.

In terms of Britain as a whole, the overwhelming majority of the minority

religious population over 95% is resident in England (2,940,000), compared
with much smaller numbers in Scotland (95,000), Wales (44,000) and
Northern Ireland (5,000). In terms of urban areas, London is unique in the
range and size of its minority religious popu- lation, with a total of 1,250,000-
no less than 40% of Britain’s entire minority religious population. This
population is otherwise heavily concentrated in a few English urban areas
outside London – and by no means always those with the largest populations
(ONS, 2006). These pro- vincial centres include Birmingham (196,000),
Bradford and Leicester (both 86,000), Luton (35,000), Slough (33,000) and
Blackburn (28,000). Minority religious proportions range in these cases from
17% of the local population (London) up to 31% (Leicester). By contrast,
there are no other local authority areas where the proportion rises above
15%. The detailed figures are set out in the Statistical Appendix (Table 4), but
it is only a slight exaggeration to say that the issues raised by minority
religions are (in numerical terms at least) issues primarily for London and a
handful of provincial English cities. This point has potentially far-reaching
implications for policy in the field of religion. To give just one example: Coles
and Bonney (2003) consider the pressures imposed on the education system
as a result of geographical concentration.

Analysing the residential distribution of specific religions shows, more- over,

that some minority religions tend to be concentrated even more heavily in
particular places. (See Statistical Appendix, Figure 3.) The non- Christian
religious populations of Blackburn and Bradford are largely Muslim, for
example (over 85%). In both these cases, it is known that the Muslim
population comes not only from one country – Pakistan – but also from one
particular area of that country, with migrants linked to each other by close
ties of kinship. For example, at least 70% of the Pakistani Muslim population
of Bradford originates from the Mirpur region of Azad Kashmir, and some
sources put this closer to 90% (Moss, 2006). (For a description of this region,
see Ballard, 2002.)

Scotland and Wales, London, Leicester and Slough are, by contrast, much
more heterogeneous in their minority religious composition, although
Muslims form the largest grouping in every case except Leicester. Luton and
Birmingham fall between these two poles. This difference in compo- sition
underlines the geographical concentrations of particular minority religions:
there are more Muslims in Bradford (75,000) than in the whole of Scotland,
Wales and Northern Ireland combined (66,000). And 56% of the Jewish
population of the UK resides in London (ONS, 2006).
There are, of course, understandable reasons why members of ethnic and
religious minorities tend to be found in some areas of the country more than
others. Ballard (1994) has discussed the process of ‘chain migration’ by
which immigrants follow in the path of relatives or other contacts who have
already settled in particular locations. Simpson (2005) has written similarly
of the process by which ethno-religious communities are formed, offering
support to established migrants and hospitality to newcomers. The presence
of existing communities and facilities for worship adds a specifically religious
ingredient to this mixture of factors (Lewis, 2002), and transcontinental kin
marriage is another powerful force promoting consolidation in specific
places.

In terms of the effects of residential separation on economic oppor- tunities,

it is noteworthy that some minority religious populations are located in areas
with relatively poor economic performance, such as

Economies. A shat are experiencing the trans those pom industrial

economies. A similar point applies at the local level of members of relies
Alam and lustek employment within local own.ethnic communities (les and
Husband, 2006), are constr their by lack of local opportunik or education
(radition of not travelined outside the locality for work or education (Bolliver,
2006). Clark ang Duiskwater have emphasised the importance of the local
area’ for eco- nomic outcomes and identified a corresponding range of
‘neighbour- hood effects. They note that ‘individuals who express a
preference for living in areas with a majority of people of the same ethnic
group have unemployment rates of up to 20 percentage points higher’
(2007: 50). And they conclude that:

The tendency of some ethnic groups in the UK to exhibit ‘opposi- tional

identities’ has also been noted by Battu et al. (2003) and such attitudes may
reduce interactions with other groups and hence limit labour market
opportunities. Of course, a desire to isolate oneself from other ethnic groups
may be borne of experiences of discrimina- tion or victimisation. (2007: 50)

The concerns raised here about the effects of residential separation within
towns and cities are related to the issues of ‘parallel lives’ and ‘community
cohesion’, which received much attention in the wake of the disturbances
that took place in the cities of Burnley, Oldham and Bradford in 2001 the so-
called Northern Riots. (These are covered in more detail in Chapter 4.)

There has been a corresponding, and often lively, academic debate about the
extent of residential separation (sometimes called ‘segrega- tion’) in the city
of Bradford, and its trend over time (Johnston et al., 2002; 2005; Simpson,
2002; 2004; 2005; Poulsen, 2005; Carling, 2008). It has now been
established that the degree of minority ethnic residential separation in
Bradford, and some similar places in Britain, is high by international
comparison, and approaches the level found for some of the most racially
segregated cities in the US. Moreover, there is good evidence that the
degree of separation in Bradford increased during the 1990s (Carling, 2008),
contrary to the claim made by Simpson (2004).

Residential separation is matched, or sometimes exceeded, by edu- cational

separation, that is, a tendency for children and young people from different
ethnic communities to be educated at different schools Religion and Culture
Make a Difference 45

(Burgess and Wilson, 2003; Miller 2004; Burgess et tal., 2005; Johnston (Bals
2006a,b; Osler, 2007). The concern here is that young people from different
ethnic and religious backgrounds can grow up lacking regular difftacts with
members of other groups, even when living in close prox. Imity to them.
Special measures have consequently been put in place to facilitate such
contacts, such as Bradford’s Linking Schools project, Which was evaluated
very positively by Raw (2006), and was included as an example of good
practice in What Works in Community Cohesion? (DCLG, 2007b).

The academic debate about community separation has been con- ducted
largely in categories of race and/or ethnicity rather than reli- gion per se,
because of the constraints imposed by the available data. D. Phillips (2006)
and her colleagues (Phillips et al., 2002; 2005) have, however, provided
some direct evidence of religious self-separation in the West Yorkshire region
– that is, families or individuals choosing to live in certain areas in order to
enjoy proximity to co-religionists. The general conclusion is that patterns of
separation are caused by a com- bination of factors that are internal to
communities (such as relative preferences for closeness with those of shared
culture and/or religion) and factors that are external to communities (such as
the experience or apprehension of hostility or discrimination). These factors
may even be combined, as suggested by Clark and Drinkwater in the quo-
tation above. It should be emphasised, however, that preferences for co-
association may not be absolute and do not necessarily exclude a
simultaneous desire for inter-cultural (or inter-religious) contacts under
appropriate circumstances. Self-separation need not, therefore, imply self-
seclusion (Carling, 2008).

The first large-scale study of residential patterns to make the religion

variable explicit was conducted by Peach (2006) in relation to London, based
on Census data. Taking the main religious groups as a whole, it found that
Sikhs and Jews are highly separated residentially. Hindus are moderately
separated; Muslims have low separation and Christians have very low
separation.

However, the low Muslim result, which suggests that Muslims are dispersed
relatively evenly throughout London, turns out to be a con- sequence of the
fact that the various Muslim populations of different eth- nicity are highly
separated from each other (in different parts of London), as well as from the
general population and from other ethno-religious groups. ‘Despite some
strong linkages, such as [that between] the Indian and Pakistani Muslims…,’
Peach (2006: 364) reports that ‘[the] residential C

46

Ethnic, Racial and Religious Inequalities

Pattern of Muslim groups in London is largely fragmented along ethnic lines’

and that ‘the Ummah exists spiritually, but is not manifested in residential
terms’. He concludes that ‘there is not a single Muslim.com- munity but a
community of communities’ (2006: 368)18
Muns finding reinforces the social units once again, it suggests that the
opeth religion units regarding choice of residince it subine elements of both
religion and ethnicity. It should be empha sised, however, that social class
remains the strongest determinapha. Sisidential location, because of the way
that class factors affect access to Housing markets, and the Peach analysis
does not take these factors into hccount. Since class is known to be
associated with religion, part of the findings for religious separation in
London will reflect social class con- straints rather than ethno-religious
preferences (but see Carling, 2008, for the demonstration that ethnicity has
effects on residential separa- tion in Bradford that are independent of social
class).

The provision of religious buildings

It was mentioned above that the existence of facilities for worship is one of
the factors attracting migrants to particular locales. Religious build- ings are
perhaps the most striking symbols of the religious character of a
neighbourhood, and one aspect of Britain’s Christian heritage is its
dominance of the British skyline. The issue of discrimination against minority
faiths in relation to planning permission for religious build- ings has attracted
considerable attention in recent years (Nye, 2001; Weller et al., 2001; Edge,
2002; Gale and Naylor, 2002; Gale, 2004; 2005). Beckford et al. summarise
the position by saying that planning processes surrounding the religious
buildings of Hindus, Muslims and Sikhs ‘have frequently been problematic’
(2006: 54). If such discrimina- tion were proven, it would count as an
example of religious disadvant- age in the sense defined above.

However, according to Vertovek and Peach (1997) there were at least 839
mosques and a further 950 Muslim organisations in Britain by the mid-1990s.
McLoughlin refers to ‘the pervasive Islamisation of the inner city’ in Bradford
(2005: 1046), noting that the majority of Bradford’s Muslims live within the
five square miles of the city centre, an area that contains 44 mosques. There
are currently 143 Hindu mandirs in Britain: 121 in England, 3 in Northern
Ireland, 4 in Scotland and 15 in Wales (National Council of Hindu Temples,
UK, 2008). Singh says that the modern gurdwara movement in Britain now
embraces 250 Sikh institu- tions, including the £17 million building opened in
Southall in 2003.
Religion and Culture Make a Difference

47

He calls this ‘the premier symbolulticultural Britain Britain, one of the He

calls erging, “cathedrals” of my in England and Wa Singh, 2006: 147 new,
there were 202 gurdwareland (Weller, 200gales (5 in Wales), 11 in Scotland
and 1ly positive situation in Brit 2003), Notwithstand, 11 apparently situated
britain with respect to rel thous buildings may have been affected by the
politics around inter giotional and national terrorism, as events in Dudley in
the Midlands seem to suggest. There, in 2007, planning permission for a new
mosque was refused following a sustained public campaign largely
orchestralud by the UK Independence Party (UKIP) and the British National
Party (BNP), both of which play on public ignorance and fears in relation to
migration, asylum seekers and refugees (Reeves et al., 2009).

Whether or not planning processes in Britain have been ‘problematic’ in the

past and, indeed, continue to be so in some cases, the significant number of
religious buildings stands in stark contrast to the situation in France, for
example, where it was not until the 1990s that any real change in attitudes
towards minority religious buildings took place (de Galembert, 2005).
Opposition to religious buildings in various parts of Western Europe is
frequently couched in terms of fears of ‘Islamisation’, as in Switzerland,
which has recently banned the building of minarets (BBC News, 2009), and in
Germany, despite its strong record of approv- ing the building of mosques
(Spiegel International, 2008). Dzi (2009) comments that religious buildings
signify the message 'We are part of this society and we will stay here’, but
religious buildings can also be said to symbolise the assertion by minority
faiths of their right to rep- resentation in the public domain. This has wider
implications than the creation of spiritual and social spaces, and some
authors see it as a chal- lenge to the Enlightenment distinction between the
public and private domains, or, indeed, to the prevailing definitions of
modernity itself (Eisenstadt, 2000), as discussed in Chapter 5.

Religion and housing

Poor housing conditions form one aspect of inequality which has often been
associated with BME status, as noted above. In the light of this long-standing
finding, and the results reported above, it can be inferred that there will be
associations between minority religious status and housing deprivation. As in
other fields, however, it is only relatively recently that research on housing
has taken explicit account of religion, driven partly by government concerns
with managing diversity and neighbourhood renewal/regeneration (Sellick,
2004).

48 Ethnic, Racial and Religious Inequalities

The Census provides information on housing tenure by religion, set out in the
Statistical Appendix (Figure 4). Sikhs, Jews and Hindus were most likely to
own their own home, followed by Buddhists and Muslims (ONS, 2004;
Graham et al., 2007). Muslims were the most likely to live in social rented
accommodation: in 2001, 28% of Muslim households were in social rented
accommodation, compared to less than 10% of Hindu, Sikh or Jewish
households. Buddhists were the most likely to be living in private rented
accommodation (24%). A very small percentage of all households live rent-
free (2%), but Muslim households were twice as likely as others to do so
(4%). This data as a whole runs counter to the stereotypes of Muslim owner-
occupation.

There is a correlation between ethno-religious groups and the afflu- ence of

their residential locations that, as expected, mirrors the findings for social
class. Hindus tend to live in more affluent locations; Sikhs are more likely to
live in middle-class areas than Muslims, who tend to occupy poorer housing
(Beckford et al., 2006). Muslims form 10% of the population resident in the
most deprived decile of residential neighbour- hoods (an over-representation
of more than three times), and there is a less pronounced tendency for
Hindus and Sikhs to share this residential deprivation (ONS, 2003). A similar
variation was found in a study of BME housing in Wales conducted by the
Welsh Assembly (2005): the least disadvantaged were Indian and Chinese
residents and the most dis- advantaged were black-Caribbeans, Pakistanis
and Bangladeshis (who in the latter two cases can be assumed to be mostly
Muslim).
The number of rooms In a house relative to the number of occupants is
commonly used to determine housing deprivation. Of all households, 40%
Muslim, 26% Hindu and 22% Sikh experience housing deprivation of this type
(Beckford et al., 2006). In 2001, just 6% of Christian house- holds
experienced overcrowding. The high proportions for Muslim, Sikh and Hindu
households in these totals are to some extent a reflection of their larger
family sizes their average numbers being 3.8, 3.6 and 3.2 people
respectively, compared with 2.3 people among Christian, Buddhist and
Jewish households (ONS, 2006).

However, overcrowding can be seen as a Eurocentric measure of depriva-

tion, which does not take into account lifestyle preferences and the finan-
cial and familial advantages of living as extended families. The need for
larger housing to accommodate extended and inter-generational families was
raised by Hindu and Muslim respondents to the consultation for the Weller
Report (Weller et al., 2001). Beckford et al. (2006) also note the issue of
shar’ia compliance in relation to the acquisition and manage- ment of
properties as a directly religious factor in Muslim housing needs. Religion and
Culture Make a Difference

49

At the same time, a careful study conducted by Sellick for the Housing At
theration brought to light another dimension of the issue, that the 15% of
Muslim one person household preflect not only the independent status of
high achieving independent professionals, but also the margin ality of ex-
offenders and the mentally ill (2004: 5). Sellick’s considered jew on the
housing situation of Muslims could stand as a conclusion on the housing
situation of the minority religions more generally:

[this] report has… highlighted the diversity within the… population and the
need for sensitivity to the differences borne not only of reli- gion but also of
socio-economic position, age, gender, locality and transnational
commitments. (2004: 4-5).
Once again, it is not religious identity by itself that is the key to policy in this
field, but religion read in conjunction with a variety of other aspects of
personal identity and social situation.

Religion, health and disability

The Census provided systematic evidence for the first time on certain
aspects of the relationship between health, religious identity and gender.

The broad findings, whichh are given in greater detail in the Statistical
Appendix (Figure 5), show that Jewish men and women have the best health
of any group. The Christian, Buddhist and No Religion popula- tions fare
slightly worse, with women’s health in each category being slightly better
than men’s. On the other hand, the health of women with Muslim, Sikh,
Hindu or ‘Other Religious’ identities is worse than that of their male co-
religionists, and with the exception of Hindus – worse for both genders than
for the general population. The differences reported by the Census are
striking: Muslims are almost twice as likely to suffer from long-term illness or
disability as Jews (about 23% com- pared to about 12%).

It is interesting that these results follow a similar pattern to the social class
differences by religion, so that it is not at all clear to what extent the health
statistics are expressing class differences rather than religious differences.
And it is also noteworthy that there is a deficit in women’s long-term health
or disability compared to men’s for all the minority religions, including
Judaism, with Buddhism being the lone exception.

As in other examples, the religious inequalities are clear, but the

explanations for them are harder to determine, and a review of the pub-
lished literature in this area does not take the discussion much further
forward. On the one hand, and/or arch highlights the numerous ways for
which religious beliefs and/or practices can impact on either the incidence on
the treatment of health or disability, and on the well-being incthose with
health conditions. This impact can be either positive negative, which implies
a corresponding ‘religious’ effect on inequal ities of health or disability.

On the other hand, this field of research is characterised by its frag-

mentation. There are a considerable number of studies, but they tend to
involve small sample sizes, often confined to one city or health authority,
and to focus on very specific topics within the overall field: an example is the
study by Heim et al. (2004) of the alcohol consump- tion of 174 Indian,
Chinese and Pakistani men in Glasgow aged 16-25. These small-scale
individual studies are not generally placed within a systematic framework of
theory and evidence. A flavour at most is given here of the connections
between religion, health and disability revealed by this literature; fuller
information is contained in the review by Macey et al. (2010).

Religious individuals often demonstrate their faith via charitable acts or

social works, and religious orders continue to be involved in health care
provision in Britain (Farnell et al., 2003). There is some research on the
relationship between religiosity and psychological well-being, with modestly
positive conclusions (Koenig, 1998; Maltby et al., 1999; McCullough and
Larson, 1999; Koenig et al., 2001; Coleman et al., 2004; Francis et al., 2008).
The relationship between ethnicity and mental health has been a long-
standing concern, prompted by the over-representation of BME individuals in
this sphere. This has been blamed on racism in psychiatry, though
alternative models are now being considered (Singh and Burns, 2008).
However, a survey by King et al. (2006) of 4,000 BME adults found no
difference in the preva- lence of common mental disorders (CMDs) between
religious and non-religious people.

One of the main ways in which religion affects health and disability is where
either a religious prescription or a cultural stigma attaches to a particular
social activity. If the activity has negative consequences for health, then the
religious or cultural prohibition will have posit- ive consequences for health,
so long as the prohibition is observed. If, however, the prohibition is not
observed and health deteriorates as a consequence, then the stigma
attached to the condition may lead to denial and make health care more
difficult. There are a number of studies tracing the incidence of such effects
in relation to activ- ities such as drinking, drug-taking or smoking (Khan,
1997; Bush et al., Religion and Culture Make a Difference 51

2003; Douds et al., 2003; Helm, 2007, 2004; Ross et al., 2004; Bradby and
Williams, 2006; Mirza et al., and relationship of religion to the incidence of
HIV/AIDs also falls

This category, where the the practice of to monogamous martialle in the

Christian tradition, or the practice of circumcision in the Musage in thewish
traditions, should help to prevent the spread of the condi and, and yet these
traditions tend to stigmatise those who fall vid to it (Anderson and Doval,
there are reland Sheikh, 2004; Adogame, 2007). On the other hand, there are
religious practices that increase the risk of infection, including Catholic
teaching on birth control ase practices associated with Islamic culture such
as polygamy, the shar ing of razor blades at the conclusion of Hajj and
certain conventions of childbirth and breastfeeding.

Religious beliefs themselves can also have implications for the explana- tion
of health conditions, and for their stigmatisation or denial (Rozario, 2005).
Pote and Orrell (2002) found, for example, that some ethnic groups viewed
schizophrenia as spirit possession or witchcraft. And a series of studies have
traced the effects of religion on the social recep- tion and psychiatric
treatment of mental illness in a variety of different ethno-religious groups
(Hussain and Cochrane, 2003; Dogra et al., 2005; MacKenzie, 2006; Leavey
et al., 2007).

Similar themes apply more generally to disability, where religious beliefs – in

reincarnation, for example – can lead to disability being seen as a
punishment, sin or curse (Confederation of Indian Organizations, 1987;
Greater London Association for Disabled People, 1987; Shah, 1995; Atkin et
al., 2002). In some cases, cultural beliefs suggest that disabilities can be
cured (Channabasavanna et al., 1985; O’Hara, 2003), and the reli- gious
issue of the marriage prospects of disabled children can loom large for their
families (O’Hara and Martin, 2003; Hussain, 2005). On the other hand,
people of faith can be more ready and/or able than those without a religious
belief to accept disabled children (Ali et al., 2001; McGrother et al., 2002;
O’Hara, 2003; Rozario, 2005). And some of the difficulties may be ascribed to
lack of knowledge or understanding, or to language barri- ers, rather than
the effects of religion per se (Begum, 1992; Smith, 1994; Nadirshaw, 1997;
Chambra et al., 1998; Shah, 1999; Katbamna et al., 2002; McGrother et al.,
2002). The evidence on service provision, and service acceptance, for
disabled young people from minority religious backgrounds is similarly
mixed, with some studies finding that cultural barriers exist, including
concern about lack of sensitivity to religious dif- ferences (Shah, 1999; Atkin
et al., 2002; Gilliat-Ray, 2007), whereas other studies go against this
conclusion (Ahmad et al., 2000).

52. Ethnic, Racial and Religious Inequalities

Perhaps the most difficult issue in this area concerns the relative Indence of
learning die south Asian can reach three times the incional average in soman
South Asian communities, with 19 the familles (Mir et al., 2001; and see
Emerson et al., 1997, for a similar finding). Mietic disorders are also
significnities higher in South Asian – partic

Bradford, for Darly Pakistani Mus Christianson and Modein the general
population Overall et al., 2002; Modell, 2004). In Brop Instance, 152 different
autosomal recessive conditions have been iden tified, relative to the
estimated 25 in ‘typical’ districts (Sulelden 2009). Genetic disorders ate
dometimes connected to consanguineons 200riage, which is estimated to
account for 35% to 60% of all mar- mages among Pakistanis (Shaw, 2001;
Hussain and Bittles, 2004). The Hink between consanguineous marriage and
the prevalence of genetic disorders is both complex and contested – the
latter particularly by non-medical researchers such as Ahmad (1995), Darr
(1997) and Stacey (1997). These concerns also relate to general issues of
women’s health and family size (Phillipson et al., 2003). Rozario has noted
the con- sequences of religious revivalism, in which contraception, prenatal
testing, termination and sterilisation may be seen as un-Islamic (2005: 190,
192). All these factors are linked to material and social depriva- tion, which
are linked, in turn, to inequalities in access to, or use of, maternal health care
services and to environmental and genetic risk factors (Baxter, 1998).
Secrecy surrounds the question of genetic disorders, partly because of
threats to family marriage prospects, and there are issues about the use of
interpreters who are not trusted to maintain confidentiality (Rozario, 2005;
and see Gilliat-Ray, 2007). On a positive note, however, Walji (2008) states
that the introduction of a screening programme at his local mosque in
Birmingham, which includes advising against mar- riage between carriers,
has led to a marked reduction in the incidence of birth defects.

The demonstration of all these varied connections between religion and

health or disability supports the case for ‘religious literacy’ in the provision of
services, or perhaps just as importantly ‘religious and cul- tural literacy’,
given the relevance in many cases of distinguishing the impact of religion
from the impact of culture and ethnic background. There is, nevertheless, a
real requirement in this field for more system- atic and comprehensive
research in order to provide a secure founda- tion for efforts to ameliorate
the associated inequalities of health or disability.

Religion and Culture Make a Difference

53

Towards a balanced perspective on religion and social inequality

The two main conclusions of this chapter pull in different directions. On the
one hand, it has bes of social wealthearly that ‘religion mations.

Difference’ to many key areas social attainth and personal well-beinkes a

differeligious identity affects social attainment independently of othin that re.
This point applies to economic activity, educational and occer factonal
achievement, residential that some mining, health and dis ability. Yet it is
highly signifist groups, are platinority religious groups, ablecially Jewish and
Buddhist gecially Hindu grobove the average on espre indicators, and others,
especially Hindu groups, are placed at or around the average level. There is
at least one finding which reaches the latter conclusion about a Muslim
group, regarding the occupational attainment of Indian Muslims noted above
(Khattab, 2009). This sug- gests that minority religious status is not in itself a
barrier to social equality in Britain. As far as we are aware, there is no
comparable find- ing in relation to minority racial status. And it is important
to note, in comparing the occupational attainment of Jewish and Buddhist
groups with that of black Christian groups (both African and Caribbean), that
race may indeed be the influential factor, as well as (or even instead of)
minority religious status (Khattab, 2009).

This helps to make the general point that it is not usually religious identity on
its own that makes a difference. A series of findings in the same key areas of
social wealth and well-being suggest that a hyphen- group perspective is
necessary, because the social position of religious groups is differentiated
according to both race and ethnicity. This double-edged conclusion is
summed up in the chapter title, ‘religion and culture make a difference’. An
important consequence of this per- spective is that there is always likely to
be uncertainty about whether a given characteristic of religious groups
should be ascribed to the influ- ence of religion in the strict sense or to the
influence of its associated ethnic culture. But the empirical connection
between religious iden- tity and other aspects of social identity goes well
beyond the relation- ship between religion and culture. The combination of
religion and culture with gender is central to the economic position of
Pakistani and Bangladeshi Muslim women, for example. And it is very striking
how often the pattern of inequality for religious groups matches the
inequalities of social class. This occurs to varying extents with educa- tional
achievement, residential location, housing conditions and long- term health
and disability.

54 Ethnic, Racial and Religious Inequalities

These observations have potentially far-reaching implications for policy in the

field of religion and equality. The problem of inequality is excluty sikhs and
most (but by specific groups. But it does not follow that the remedy for
religious inequality is to rely on meand Drinkwater the religious aspects of
identity. Example, Clark and De statistical models phasised the central in. For
exam of education: “the statistical models show that education har positive,
increasing and significat affect on employment for virtually all ethnic groups,
with education boat for higher (level 4/5) quality cations’ (2007: 15) and
‘education boosted the employment chances cation minorities by more than
for the White group’ (2007: 45; and see also Phillips, 2007). Suppose, then,
that the educational achievement of disadvantaged Sikh and Muslim groups
were to be improved, and that other measures were taken to enhance their
class positions. Suppose also that cultural (as opposed to religious) barriers
were addressed, and the levels of gender equality improved within some
ethno-religious communities. All these measures would impact favourably on
religious inequality without any attention having to be paid to the wider
social reception of religious identity.

Take as an example the case of material inequality as it affects the poorest

ethno-religious groups. Berthoud comments: ‘Name any group whose
poverty causes national concern – pensioners, disabled people, one-parent
families, the unemployed Pakistanis and Bangladeshis were poorer’ (1997:
180). The data summarised above shows that these groups are characterised
by larger than average family size; extended family living; a high incidence
of children with disabilities; women who do not work outside the home;
unemployment or low-paid male employment that is often related to lack of
educational qualifications and skills; and the maintenance of enduring links
with countries of origin, which have financial implications through, for
example, trans- national marriage, transcontinental travel and overseas
remittances.

The policy Initiatives that might help to alleviate the material pov- erty
experienced by a family in situations such as these will obviously depend on
the reasons for their situation. Thus, if the reason for there being only one
wage earner is labour market discrimination, then an appropriate response is
through existing equalities legislation. If, how- ever, it is due to the belief
that a woman’s main role is centred on child rearing and the home, then
such legislation will have no effect. Similar points apply to this hypothetical
family’s need for specialist disabil- ity services. If the wife’s inability to work
outside the home is due to

Religion and Culture Make a Difference 55

Lack of provision of such services, then a logical response is to improve
provision, or to remove any element of discrimination surrounding it. If,
however, such services are available but are rejected for religious or

As et al. (2002) suggest, it is more difficult to see how this family can be
some communities lie within the scope of community action as well, fois not
clear that measures take these canal to the community will have an effect on
the inequalities in these cases.

Our appeal, then, is for aed, evidence-based, the question of religious

inequality, and for a balanced, evid inequalities assessment of all the face.
Tors that contribute to the observed inequalities. Such an approach relles to
the following four points. The hyphen-group orientation is the first of these.
The second is that there is an important distinction betwest religious
inequality and religious injustice, so that examples of religious inequality
cannot be ascribed to the effects of religious injustice with out further
evidence of these effects. The third point is that a balanced view of the
causation of religious inequality will take account of factors that are internal
to ethno-religious communities as well as factors that are external to them,
derived from their social environments. The final point is perhaps the most
far-reaching, and acknowledges the difficult policy choices to be made in this
area. In particular, if multicultural- ism acts to shield cultural and religious
values or practices that repro- duce inequalities within communities, the
choice may have to be faced between multiculturalism and equality as policy
options.

In the course of reaching these conclusions about a balanced perspec- tive,

we have begun to trace out some of the consequences of depart- ing from
the balance provided by these points. The problems with the
recommendations of the Pell Commission arise from their moving too swiftly
from inequality to injustice, ignoring the importance of gen- der and
ascribing ethnic inequality solely to external causes. A similar problem
applied to the EOC report on minority women’s employment, Moving on Up?
We have touched on two other examples of a similar diffi- culty. In his
analysis of the effects of skills differentials, Khattab (2009) has downplayed
the internal factor (the skills differentials themselves) in favour of the
external perceptions of these skills by those outside the ethno-religious
community. And a similar problem affects some writ- ers on the issue of
residential segregation, who seem very reluctant to accept that the observed
patterns of residential separation are caused in part by the culturally
conditioned choices made by members of ethnic or religious minorities
themselves. These writers include both Simpson (2002; 2004; 2005) and
Phillips (D. Phillips, 2006; Phillips et al., 2002; 2005; see Carling, 2008, for
further analysis). In all these cases, we are beginning to see the effects of an
approach

That is one-sided and therefore ‘ideological’ rather than analytical, as

indicated by the departure from the balanced perspective outlined above. It
will be shown in the course of the next chapter that an ideo- logical approach
is not confined to one or two solecisms in the text, or to a small number of
sources, but appears to be characteristic of a wide range of influential work
on religion and equality. In particular, conceptual inflation – such as the
treatment of inequality as evidence in itself of injustice – seems to have
become an occupational hazard for writers in this field.