coenders, m. and scheepers, p. - research.manchester.ac.uk · web viewhowever, the evidence...
TRANSCRIPT
Is ethnic prejudice declining in Britain? Change in social distance
among ethnic majority and minority Britons
Ingrid Storm, Maria Sobolewska and Robert Ford (University of
Manchester)
Authors' pre-peer review version. Revised version accepted for
publication after peer review in British Journal of Sociology on
25.04.2016.
Abstract: Most literature on racial prejudice deals with the racial
attitudes of the ethnic majority and ethnic minorities separately. This
paper breaks this tradition. We examine the social distance attitudes of
white and non-white British residents, to test if these attitudes follow the
same trends over time, whether they are driven by the same social
processes and whether they are inter-related. We have three main
findings. Firstly, we show that social distance from other ethnic groups
declines over time for both white and ethnic minority Britons. For the
white majority there are both period and cohort elements to this decline.
Secondly, we see some evidence that social distance between the
majority and minority groups is reciprocal. Specifically, minorities who
experience rejection by the white British feel a greater sense of distance
from them. Thirdly, we find that all groups share the perception of the
same ethnic hierarchy. We see evidence of particularly widespread
hostility towards Muslim Britons from all ethnic groups, suggesting that
Muslims have become a ‘pariah’ group for many British residents of all
1
backgrounds, including a large number who do not express hostility to
other groups.
Word length: 8455 (including title, abstract, references, notes, tables
and figures)
Date of submission: 20.07.2015
Introduction
Ethnic diversity is rising rapidly across Europe, a consequence of both
high immigration and the growth of native born ethnic minority
communities. This historically unprecedented demographic shift has
triggered intense political and academic debate. Concerns have been
raised about the negative impact of rising diversity on social trust and
support for the welfare state (Putnam 2007; Banting and Kymlicka 2006).
The perceived threat to traditional national and ethnic identities has
fuelled the emergence of a new family of radical right parties in Europe
(Mudde, 2007; Ford and Goodwin, 2014) and also generated anxiety
among Europe’s mainstream political elites, with politicians of all
ideological persuasions speaking out about the alleged ‘failure’ of
multiculturalism (Banting and Kymlicka 2013).
Against this negative narrative, there is growing evidence that majority
group citizens of many diverse countries are slowly becoming more
tolerant towards ethnically distinct groups (Ford 2008, Coenders and
Scheepers 2008; Bobo 2001). Research into observable behaviours shows
2
increased social contact between different groups, including growing
rates of interethnic marriage (Muttarak and Heath 2010). One
explanation for this apparent disjuncture is that the debates around
immigration tend to adopt a relatively short time horizon. The values and
identity attachments of individuals are difficult to change in the short
run, being shaped most strongly by conditions in youth and early
adulthood (Mannheim 1970), but over the long run new generations rise
with different values, producing steady change in aggregate attitudes
through the process of cohort replacement (Inglehart 1997; Tilley 2005).
Earlier research on Britain (Ford 2008), the US (Bobo 2001) and Europe
(Coenders and Scheepers 2003) has provided evidence that this
generational process of racial attitude change and minority integration is
ongoing in countries with long established migrant minority groups.
However, this research leaves several important questions unanswered.
New data collected in 2013 by the British Social Attitudes (BSA) survey
enable us to take this argument forward in several respects.
Firstly, the existing research on Britain ends in the mid-1990s. Since this
time, Britain has experienced a second wave of large scale immigration,
and an increase in popular anxiety about identity conflict and threats
from minorities. It is not clear whether the period and cohort shifts
towards more tolerant racial attitudes observed in the 1980s and 1990s
will persist in the more polarized contemporary social environment. On
the other hand, the majority of white adults in the Britain of 2013 have
3
grown up with ethnic and racial diversity as an accepted fact of British
social life, and so racial and cultural distinctions may not be as relevant
in drawing group boundaries as they were 20 years ago. We can use this
data to test whether race still functions as a ‘bright boundary’ (Alba
2005) in British society, as earlier research (Ford 2008; Song 2004: 864)
suggests it did in the 1980s and 1990s.
Secondly, we examine the variation in majority reactions to different
minority groups. Researchers have found evidence of persistent ‘ethnic
hierarchies’ (Hagendoorn 1995) in many social contexts. Ethnic groups
who are socio-economically disadvantaged and racially or culturally
different from the majority group are consistently less valued, by
different ethnic groups within the same society (Hagendoorn 1995; Ford
2011; Sidanius and Pratto 1999). Other work has suggested that many of
the distinctions in such hierarchies have become ‘blurred’ over time as
groups are accepted as part of the mainstream culture. However, some
remaining ‘bright boundaries’ of differences in race, religious practice
and cultural background can leave some groups outside this mainstream,
and stigmatized by all within it (Alba 2005; Modood 2003; Song 2004).
The new data measures attitudes about a wider range of groups: Black
Africans, East Europeans and Muslims, alongside the Asian and
Caribbean groups asked about in earlier British data. These additional
categories provide a greater range of variation on racial, cultural and
religious difference. Specifically, it allows us to test whether Muslims
have become singled out as a particularly stigmatized or problematic
4
‘pariah minority’ (Saggar 2008; Modood 2003), attracting more intense
and persistent hostility from the majority group.
Research on the integration of immigrant minorities in the US has long
observed that it is a two way process of mutual change and acceptance
(Alba and Nee 1997). Integration requires not only that the white
majority population be fully comfortable having intimate social relations
with ethnic minorities, but that ethnic minorities feel similarly
comfortable about contact with the white majority. This other side of the
social integration issue has not been subject to systematic quantitative
research over time in a British context. Using ethnic minority survey
data, collected in 1993-4 and 2010, we can compare the racial attitudes
and attitudinal change of ethnic minorities with the majority. This offers
us a unique opportunity to extend existing explanatory theories to
include the attitudes of non-white minorities (Song 2004: 874).
Looking at increasing rates of intermarriage among the younger cohorts
of ethnic minorities, Muttarak and Heath (2010) found ethnic differences
even after controlling for generational differences and the ethnic
composition of residential areas, with people of Indian, Pakistani and
Bangladeshi background being less likely to marry outside their own
group. Since marriage in the vast majority of cases will reflect the
preferences of both parties involved, it is hard to know whether the
trends are driven by greater white resistance to marrying into South
Asian minority groups, or vice versa. It is also possible that these
5
processes are related: with the attitudes of the minority reciprocating
those of the majority. We will test this possibility, observing whether
experience of rejection by the white majority, through racial
discrimination, increases minorities’ hostility to the majority group.
Theory and hypotheses
1. Why should social distance decline over time and between
generations?
Interethnic prejudice and ingroup favouritism are cognitive shortcuts for
quickly judging the risks and potential rewards of associating with
someone one has never met. Research in social psychology shows that
prejudice can be reduced by intergroup contact, but also that this
requires some initial willingness and trust (Allport 1954; Pettigrew and
Tropp 2009). Attitudes to social contact with outgroup members should
thus be a good indicator of prejudice more generally. In this paper we
focus exclusively on interpersonal social distance, and more specifically
on acceptance of an in-law from an ethnic out-group. This question is
strongly associated with other social distance measures (Pettigrew and
Meertens 1995), and shows that the same patterns and trends in Britain
as accepting an ethnic minority as one’s boss (Ford 2008).
We expect four main mechanisms to reduce interethnic social distance in
Britain since the 1980s due to four main mechanisms. Firstly, the non-
white ethnic minority share of the population increased from 3.9 per cent
in 1983 to 11.9 per cent in 2013i (OPCS 1985; ONS 2014) both through
6
immigration and a steady increase in the size of the British born ethnic
minority population. Interpersonal contact between ethnic groups is
hypothesized to reduce anxiety about ethnic and cultural difference and
increase empathy for people from different ethnic groups to oneself.
Secondly, even those who live in places that are largely unaffected by
immigration will watch television broadcasts, support sports teams, and
read magazines that now more frequently reflect ethnic diversity as an
established part of British cultural and social life. Thirdly, participation in
higher education has expanded rapidly in the UK since the 1960s, and
education is associated with a reduction of prejudice (Coenders and
Scheepers 2003; Ford 2008; Strabac and Listhaug 2008: 271). Fourthly,
there has been a general shift in norms and values over this period, and
blatant racism has become less socially acceptable. Compared to the mid-
20th century there is almost no cultural, intellectual or elite political
support, for discrimination on the basis of ethnicity (Ford 2008). This
previous research shows change to have continued over time as a result
of both period and cohort effects. Thus, our first hypotheses are that:
(H1a) white social distance to ethnic minorties has continued to decline
over time from the 1980s until the 2010s
(H1b) social distance between the white majority and ethnic minorties is
lower for cohorts born in the 1980s and 1990s than for previous
generations.
7
2. Why might Muslims be singled out?
Islamophobia, or majority group hostility directed particularly at Muslim
minorities, has become a growing concern in developed democracies
(Strabac and Listhaug 2008). Despite their ethnic, national and religious
diversity, Muslims may be singled out for particular hostile attention due
to both cultural and socio-demographic differences which set them apart.
The first cultural mechanism is the association between Islam, religious
fundamentalism and violence. Terrorist acts by Islamist extremists, such
as the 11 September attacks and the Madrid and London train bombings,
have increased the salience of Islam in Europe during the 2000s, and
inspired narratives portraying Muslim religion and culture as violent and
threatening, both physically and symbolically, to Western civilisation and
values. Muslims in Britain are also on average more religious, and have
more conservative social norms on gender and sexuality than the
majority (Field 2011). Negative views to Muslims can therefore also stem
from a perception that Muslims are opposed to liberal secular Western
values and lifestyles (Werbner 2005: 8).
A second issue is the distinct socio-economic status of Muslim minorities.
Britain's Muslim population tends to concentrate in relatively deprived
areas, and like other ethnic minorities, have lower social mobility than
average (CoDE 2014). The Muslim population also tends to be more
residentially segregated than other ethnic minority groups. The
combination of high segregation and concentration in poor urban areas
8
has inspire a narrative of Muslims as unwilling or unable to integrate into
mainstream society, despite their mobility patterns being driven by
similar social and economic factors as those of other ethnic and religious
groups in Britain (Finney and Simpson 2009).
From this we would expect that:
(H2a) members of the white majority will express more social distance to
Muslims than to ethnic minority groups
(H2b) the difference between the social distance to Muslims and social
distance to other ethnic minorities is greater among the cohorts coming
of age around 2001.
3. Are minority perceptions of social distance shaped by the same
mechanisms as those of the majority?
As outlined in the introduction, we look at the social distance felt by the
ethnic minority groups when faced with a possibility of acquiring an in-
law from a different ethnic minority, or from the white majority group,
with the aim to test four hypotheses.
The first expectation is that minority groups share an ethnic hierarchy of
preference with the majority (Hagendoorn 1995). All groups have a
preference for social contact with members of their own group. This is
followed by the white majority, whose status is the highest in the ethnic
hierarchy; the willingness to have an in-law from other ethnic minority
groups then follows the same status ordering for all groups, including the
9
majority. Such consensual ethnic hierarchies have been found in a
number of countries including the Netherlands and former Soviet
societies (Hagendoorn 1995), Sweden (Snellman and Ekehammar 2005),
Canada (Kalin and Berry 1996) and the US (Sidanius and Pratto 1999).
Hagendoorn (1995) proposed an inter-group differentiation mechanism
to account for this pattern. If a minority group perceives another group
to be lower in the status hierarchy than they are, their resistance to
social contact with members of that group should increase, as such
contact could be a risk to their own status. In contrast, contact with
groups who are perceived to be higher in the hierarchy is seen as status
enhancing, and will be less resisted.
(H3a) Ethnic hierarchies: all minority groups will prefer intra-marriage,
followed by in-laws from the white majority and with the Muslim in-laws
as least desirable.
In addition to status maximisation, another mechanism that can drive the
formation of ethnic hierarchies is cultural distance. It has been shown
that some immigrants and minority group members are more welcomed
(Ford 2011) and assimilated faster into their new societies (Maxwell
2012) than others. The literature on New Racism argues that although in
the past race was a more salient differentiator between groups, cultural
difference has become the more prevalent justification for prejudice
(Pettigrew and Meertens 1995: 59). South Asians, being culturally distant
from the white British due to greater differences in religion, language
10
and social norms, may have become more likely to be rejected by the
majority (Maxwell 2012). Cultural distance may also be a relevant
mechanism for how ethnic minority attitudes to the majority are shaped.
As a result, we expect that
(H3b) Cultural distance: South Asian groups, compared to other ethnic
groups, will feel a greater sense of social distance from the white
majority and other ethnic groups.
The mechanisms of status maximisation and cultural distance that
underpin the consensual ethnic hierarchy both make assumptions about
how the attitudes between majority and minority perceptions of social
distance are related. There are two possibilities here. Firstly, as white
prejudice towards ethnic minority groups declines, minorities may be
more willing to engage in social contact with the majority in a reciprocal
manner (Kalin and Berry 1996). This would narrow the distance from the
majority group, but not necessarily between different minority groups. A
second possibility is that the same underlying social trends drive declines
in prejudice for all groups regardless of their position in the ethnic
hierarchy. Education and experience of diversity should diminish younger
whites’ sense of distance from other groups; we can expect the same
effect for ethnic minorities. As minorities experience educational upward
mobility, and the British born make up an increasingly larger share of
young ethnic minorities, we can expect that they too will become more
tolerant, cosmopolitan and open to other cultures.
11
(H4a) Reciprocity: ethnic minorities’ social distance from the white
majority will decrease with diminishing negative attitudes of the white
majority, and increase with experience of discrimination
(H4b) Parallel changes: generational change and education will reduce
social distance between all ethnic groups: young, British born, and better
educated minority group members will be more open to out-group
contact.
Data and methods
We employ two sets of data to analyse majority attitudes, and two
parallel sets to analyse minority attitudes. In eight British Social
Attitudes (BSA) survey waves conducted between 1983 and 1996, white
respondents were asked if they would mind a close relative marrying an
Asian or a Caribbean (West Indian)ii person. In 2013, these items were
repeated, and additional items measuring views of intermarriage with
Muslims, Eastern Europeans and Black Africans were added.
Our two ethnic minority datasets use comparable questions. The Fourth
national survey of Ethnic minorities (FNSEM) from 1993-1994 and the
Ethnic Minority British Election Survey (EMBES) from 2010 both asked
questions about intermarriage with other ethnic groups, including other
minorities and the white majority group. (For full details of all the
variables used in the analysis see Appendix Table AI.)
Analysis
12
1. Trends over time
Figure I charts the trend in British attitudes to a relative marrying an
Asian or Caribbean person over the thirty year period covered by the
data. Attitudes were relatively stable during the 1980s, with around 50
per cent expressing opposition to intermarriage with either group in each
of the first four time points. Aggregate opposition then declined
considerably in the 1990s, falling in each of the three time points, and
declined again in 2013. Roughly half of the opposition to intermarriage
recorded in 1983 was gone by 2013. However, the evidence suggests
that the pace of decline has slowed. Opposition fell from c.55 per cent to
c.35 per cent in 13 years between 1983 and 1996 (about 1.5 percentage
points per year), and then to c.25 per cent in the 17 years from 1996 to
2013 (about 0.6 points a year).
There is very little difference in the aggregate trends in attitudes towards
the two ethnic minority groups. Further analysis (Appendix Figure AI)
indicates that the change in social distance to Asians and Caribbeans can
be attributed to both period and cohort effects. Each younger cohort is
more tolerant than the last, but there is an additional decline in social
distance over time within each cohort, supporting hypotheses H1a and
H1b.
<Figure I>
<Figure II>
13
2. Cohort structure of majority group attitudes
Figure II shows the distribution of opposition to intermarriage with five
groups across birth cohorts in the 2013 BSA data. The overall
generational shift towards acceptance identified in earlier analysis (Ford,
2008) is seen here for all five groups. In every case, opposition is highest
amongst the oldest, pre-migration, cohorts and declines in a more or less
linear fashion as we move to younger cohorts. While more than a third of
the oldest cohorts express opposition to intermarriage with each ethnic
minority group, less than one in ten of the youngest cohort (those born in
the 1990s) express opposition to any group except Muslims.
Reactions to the five groups asked about do not suggest a clear or
consistent ethnic hierarchy: Social distance from four of the groups is
rather similar, particularly amongst respondents born in the 1950s or
later. Only the oldest cohorts show greater opposition to the non-white
groups than to Eastern Europeans, suggesting racial difference is salient
for them, but is not so for younger cohorts. However, there is clear
evidence of much greater social distance from Muslims than any of the
other groups, supporting the idea that Muslim religious and cultural
distance functions as a ‘bright boundary’. This distinct reaction to
Muslims also shows a generational dynamic - it is very intense amongst
the oldest cohorts: over two thirds of those born in the 1940s or earlier
oppose a Muslim in-law. Opposition then falls steeply amongst younger
cohorts, with evidence of convergence towards other ethnic groups.
14
Opposition to intermarriage with other groups runs at around 20 per cent
for those born in the 1950s, but 55 per cent for Muslims. Among those
born in the 1980s, the figures are 10-20 per cent for non-Muslim groups
and 28 per cent for Muslims. There is also evidence of somewhat greater
opposition among the very youngest cohort, the first to grow up in the
post 9/11 and 7/7 environment, but as this group is still relatively small in
our data (as those born after 1995 were under 18 at the time of the
survey) this cannot be taken as conclusive.
To determine which factors seem to be driving the decline in social
distance from ethnic minorities, we run two ordered logistic regression
models with the dependent variables being whether the respondent
would mind a lot, a little or not at all if a relative marries an Asian, or
Caribbean person respectively (See Table AI for details). The results are
shown in Table I.
The first model includes dummy variables for each year since 1991, and
compares them with the 1980s (when attitudes were stable). We also
include gender and a centred variable of birth year, as well as birth year
squared to account for the possibility that the relationship with
age/cohort may be curvilinear.
We test the impact of education on reactions to minority groups
(Coenders and Scheepers 2003; Ford 2008; Strabac and Listhaug 2008:
271) using details of the formal qualifications respondents have received,
15
with no qualification as the reference category. As prejudice could be
associated with economic insecurity (Bobo 1983), three other
socioeconomic indicators were also included: the Registrar General’s
measure of social class (the only measure available in all survey years),
unemployment, and housing tenure.
Being married and having children have been associated with more
socially conservative attitudes (Tilley 2005). Religious people may be
more concerned about religious differences between spouses (Voas and
Crockett 2005). As previous research (Ford 2008; Storm 2011) found that
Christian affiliation is generally associated with negative attitudes to
immigration and minorities, while regular church attendance is
associated with positive attitudes, we control for both in the model.
The second model includes interactions between birthyear and
educational qualifications and gender respectively, as Ford (2008)
showed a much steeper decline by birthyear in social distance to ethnic
minorities among women and people with higher education, and
including these interactions improved the model fit.
<Table I>iii
A number of key results emerge from this table. First, social distance to
both groups is significantly lower in the 1994, 1996 and 2013 surveys
than the earlier surveys, confirming that there is a sustained overall
16
decline in social distance this period even after controlling for
generational shifts and other social factors (Hypothesis 1a).
The period effects suggest an overall liberal shift across all respondents,
but in support of hypothesis 1b, we also find significant cohort structure
in the attitudes to both groups. The oldest respondents are the most
sceptical of new ethnic minority in-laws, and younger cohorts are
consistently more liberal. However, for attitudes to black Caribbeans,
this cohort trend levels off - those born in the 1980s or 1990s are no
more liberal than those born in the 1960s and 1970s, after controlling for
period (see Figure AI).
We also find significant gender differences - women are less opposed to
intermarriage than men - and educational differences: those with A-levels
are more liberal than those with GCSEs or less, and those with university
degrees are more liberal still. The interaction effects show that the
generational change in social distance to Asians has occurred to a
greater extent for women than men, as was also found in earlier analysis
(Ford 2008). Education appears to have had most influence on the
tolerance of cohorts born between 1930 and 1970. University graduates
led the trend towards social acceptance of interethnic marriage, but for
cohorts of graduates born in the 1980s or later such acceptance is
already near-universal. In these youngest cohorts, acceptance of
intermarriage is a social norm only rejected by those with no educational
qualifications at all and by a shrinking minority even in this group. (See
17
Figure AII for the predicted probabilities by birth-cohort and highest
qualification.)
3. Are Muslims a ‘pariah’ minority?
Muslims stand out in the 2013 data as the only group which still faces
widespread hostility from across the white population (see Figure II).
This supports hypothesis (H2a) that opposition to intimate social contact
with Muslims would be stronger than similar contact with other out-
group members. We also hypothesized that this phenomenon would have
a generational aspect. Young people born in the 1980s and 90s, growing
up amid increased media salience of Muslim terrorism are expected to
have more divergent attitudes towards Muslims and other minority
groups (H2b). To test this hypothesis, and see what other predictors can
explain Muslims’ position at the bottom of the hierarchy, we show the
results of a logistic regression analysis restricted to the data from 2013
(when people were asked about Muslims, East Europeans and Black
Africans for the first time) in Table II. The model is otherwise similar to
the previous analysis (Table I), and shows that the generational
differences in social distance from Muslims are as large, or possibly
larger, than those seen for other groups. There is no evidence that those
born in the 1980s and 1990s express greater opposition to inter-marriage
with Muslims. Instead, the opposite is true.
18
Table II also shows that in addition to age or year of birth, attitudes to
Muslims are largely determined by education, which also predicts social
distance towards other minority groups.
<Table II>
Social distance from Muslims is much higher than from any other ethnic
group. This suggests some people single Muslims out, even while
accepting other minorities. We can analyse the drivers of this 'Muslim
penalty' (Heath and Martin 2013), as all those who were asked about
their attitudes to Muslim intermarriage were also asked about either
black Caribbeans or Black Africans. This enables us to identify those who
say they oppose inter-marriage with Muslims, but do not oppose inter-
marriage with one of the black groups. As Table III shows, at least half of
those who said they mind if a relative marries a Muslim, would not mind
if they married the other ethnic minority group they were asked about
(Black African or Caribbean), while only a small proportion of
respondents express opposition to inter-marriage with the black group,
but not the Muslim group.
<Table III>
The logistic regression model in Table IV shows that those who single out
Muslims in this way are primarily people over 30, married and with
some, but not higher, education. In comparison, those who mind both
19
Muslims and Blacks, are older, less educated, and more likely to have an
Anglican or non-Christian religious affiliation. This is consistent with the
earlier analysis and indicates that there is a sizable group of relatively
younger and more educated people, who do not have general high levels
of ethnic prejudice, but who still show considerable social distance to
Muslims. This could reflect perceived conflicts between Islam and liberal
social values (Werbner 2005) related to the on average higher levels of
religiosity and material deprivation of Muslim minorities.
<Table IV>
Figure III shows the predicted probabilities of minding a Muslim in-law
by birth cohort when controlling for all the other variables in the model.
Overall opposition to Muslim intermarriage, shown by the darkest line,
falls sharply between cohorts. However, distinguishing between those
who only mind Muslims, and those who mind both Muslims and black
ethnic minorities, reveals a more differentiated pattern. The oldest
cohorts tend to oppose intermarriage with both groups, suggesting a
generalized opposition to outgroup marriage. Among those in middle age,
opposition to the black group declines, but opposition to the Muslim
group remains. These middle cohorts - born between the 1940s and the
1970s - are most likely to single out Muslims. In the youngest age groups
however, the difference is considerably smaller. Despite the increased
salience of Muslims as an outgroup after 2001, the cohorts coming of age
in this period (born in the 1980s and 90s) are in fact less likely to single
20
Muslims out than their parents. In other words, hypothesis H2b was not
supported - the experience of growing up in the polarised post 9/11
environment has not made the youngest cohort of Britons more likely to
single out Muslims as a "pariah" group.
<Figure III>
4. Ethnic hierarchy, parallel trends or reciprocity? Social distance
attitudes among ethnic minorities.
Figure IV shows that the period decline in white social distance from all
minority groups is also found in ethnic minority attitudes towards whites:
in almost all cases opposition to a white in-law declines between 1994
and 2010. The exception is the black Caribbean group, whose social
distance from whites was already low in 1994, when 11.6 per cent
expressed opposition to intermarriage with whites, notably lower than
the 38.5 per cent of whites who opposed marrying a Caribbean in data
from the same year. This figure remained unchanged in 2010. Although
resistance to intermarriage declines in all three Asian groups between
the two time points, they all express higher opposition than either the
African, Caribbean or Chinese groups, a pattern which does not change
over time, and is also reflected in marriage statistics (Muttarak and
Heath 2010). The particularly negative views about intermarriage with
Muslims found among whites also appear to be mirrored in the views of
Muslim minorities themselves - the highest opposition to intermarriage
21
with whites is found in the predominantly Muslim Pakistani and
Bangladeshi groups.
<Figure IV>
Table V based on EMBES 2010, suggests that a three tier ethnic
hierarchy operates in Britain, with the white majority at the top, Muslims
at the bottom and other ethnic minorities in-between. Ethnic minorities
and the white majority broadly agree on the relative acceptability of
different groups, supporting Hypothesis H3a. As predicted, white in-laws
are the least opposed by all minority groups, while the group which
consistently evokes the most opposition is Muslims. Non-Muslim Indians,
Black Caribbeans and non-Muslim Black Africans all showed much
greater social distance to Muslims on this question than they did to other
ethnic groups, suggesting that the pariah status of Muslims is recognised
by non-Muslim ethnic minorities. As with the BSA, the question about
Muslims was unfortunately only asked once and there is no comparative
question for other religious groups.
There does not seem to be a consistent pattern according to group status
outside of this three-tiered division. As shown in Table V, Black and Asian
non-Muslims are equally accepted by whites, yet non-Muslim Asians are
by far the least comfortable with an out-group in-law. Table V also shows
that while African Muslims are one of the groups most keen to
intermarry, Asian Muslims reject white in-laws the most. Hypothesis
22
(H3b) that singled out South Asians as most culturally distant from other
groups and thus least favourable to intermarriage is supported.
<Table V>
To further examine the cultural distance (H3b), reciprocity (H4a) and
parallel change (H4b) hypotheses, we conducted multivariate analysis. In
Table VI we analyse ethnic differences in social distance found in the
EMBES 2010 data in a logistic regression model, controlling for sex,
birth cohort, country of birth and education as well as the ethnic and
religious groups used in Table V. To test the reciprocity hypothesis (H4a)
we also include a dichotomous variable of whether the respondent has
experienced discrimination based on their race, ethnicity or religion in
the past five years. To control for the salience of religiosity to the
respondent’s identity (Perry 2014), we also include the frequency of
religious service attendance.
<Table VI>
Even after controlling for these variables, some ethnic differences in
social distance persist. Most notably, the analysis offers further support
for hypothesis H3b, as Asians are more likely than Africans and
Caribbeans to mind their relatives marrying a white person, and non-
Muslim Asians are more likely to mind intermarriage with a Muslim. As
for the differences between Muslim and non-Muslim respondents the
23
results are mixed: Asian Muslims are somewhat more likely to be
bothered by the prospect of White in-laws, but non-Muslims are more
bothered by Black ones. There is no significant difference between
Muslims and non-Muslims among the Africans.
In contrast to the white majority, there is no significant cohort effect, nor
do we see a clear association between gender or education and social
distance among the ethnic minorities. This leads us to reject hypothesis
H4b that parallel social changes are driving declines in social distance in
the majority and minority populations. However, what does seem to make
a difference is how long the respondent has been in the UK. Those who
are born in the UK are more likely than first generation immigrants to
accept white in-lawsiv suggesting that social distance is in part based on
experience of and contact with the white majority. However, contact with
the majority can cut both ways: those who report having experienced
ethnic discrimination are generally more opposed to marriage with the
white majority, in line with the reciprocity hypothesis (H4a). We also find
evidence that religiosity encourages minority social distance from all
other groups, with a particularly large effect for the question about
Muslims. This could reflect several factors - more religious minority
respondents may be more embedded within their own group (Perry,
2014), may have more conservative social values (Altemeyer, 1981) or
may simply place greater weight on cultural and religious difference
when judging those from other groups.
24
These findings suggest that cultural difference, but not group status,
shape minority attitudes towards other ethnic groups. There is evidence
also of a strong reciprocity mechanism with regards the white majority -
ethnic minorities who have experienced discrimination become more
opposed to inter-marriage with the white majority group. The decline in
ethnic minority social distance from the white majority does not seem to
be driven by the same social processes as drive the parallel decline in
white social distance. Minority attitude change is not the product of
cohort value shifts, or educational expansion, but instead the result of
being born and raised in Britain, and thereby having extended social
contact with the white majority.
Conclusions
This study showed a continued decline in white social distance toward
ethnic minorities measured as expressed opposition to interethnic
marriage in one's close family. The decline is clear whether measured by
survey year or by birth cohort, and applies to all ethnic groups where
there is data available, thus supporting Hypothesis 1a and 1b. The trend
seems to be caused in part by the growth in education, but the effect of
education also appears to have diminished in recent years. A large
majority of British respondents born in the 1990s accept inter-marriage
with all minority groups except Muslims.
The finding that all minority groups express smaller or similar levels of
social distance to whites compared to the other ethnic out-groups seem
25
to offer partial support for the ethnic hierarchy hypothesis (H3a), that the
same hierarchy of social distance can be observed among different ethnic
groups. Nonetheless, the ‘hierarchy’ we observe is a very rudimentary
one. The majority white group is consistently ranked at the top, one
pariah group (Muslims) is consistently ranked at the bottom, while the
other groups fall in between in no particular order. Race also matters as
a distinguishing factor for older whites, and in earlier surveys, but
younger whites in the 2013 survey showed similar responses to all
minority groups except Muslims.
The consistency of the large social distance towards Muslims expressed
among both the majority and minority non-Muslim respondents, does
offer support for the description of Muslims as a ‘pariah’ minority (H2a)
attracting uniquely intense hostility from all other groups. The
significance of religious affiliation and attendance in some of the models
also seem to underscore the importance of cultural and religious identity
as determinants of social distance, at least when measured by attitudes
to interethnic marriage in the family.
While the social distance from Muslims may in part be the result of
recent events and public discourses, the results do not support the
hypothesis (H2b) that this has disproportionately affected cohorts coming
of age in the 2000s. The relative social distance to Muslims as compared
to other groups is no higher for younger cohorts than for middle-aged
ones. While the results suggest that there is a specific anti-Muslim bias in
26
social distance, there is no indication that this is a growing generational
trend. However, measures for more than one year would be needed to
establish with certainty whether the specific anti-Muslim bias is declining
over time.
The decline in white social distance from ethnic minorities is mirrored in
minority attitudes to the majority - social distance from whites declined
between 1994 and 2010 for the three groups where it started highest,
and in the fourth group, black Caribbeans, it started low and did not rise.
The mechanisms driving this change appear to be distinct, however:
attitude change in minorities is associated with British birth, secularism
and positive social contact with whites rather than educational expansion
and cohort value change. Direct social experience clearly matters for
ethnic minorities - those born in Britain, who thus have more extensive
contact with the majority group, report much lower social distance, but
those who report experience of discrimination report greater social
distance. However, we do not find evidence that Muslim minorities, who
are on the receiving end of more hostile attitudes, react by becoming
more hostile to other groups: the views of Asian and African Muslims are
no different to those of non-Muslims with the same ethnic origin.
Taken together, the results suggest that racial boundaries between
whites, blacks and Asians have blurred in recent decades, and continue
to do so. Far fewer people express opposition to inter-marriage with
other ethnic groups. However, some boundaries remain bright, or may
27
even have brightened over time. Muslims are singled out for unique
hostility from both the white majority and other minorities, including
from many who express inclusive attitudes towards other groups.
Notes
28
Bibliography
Alba, R. 2005 ‘Bright vs. Blurred Boundaries: Second-Generation
Assimilation and Exclusion in France, Germany, and the United
States’, Ethnic and Racial Studies 28(1): 20–49.
Alba, R, and Nee, V. 1997 ‘Rethinking Assimilation Theory for a New
Era of Immigration’, The International Migration Review 31(4): 826–
74.
Allport, G. W. 1954 The Nature of Prejudice, Reading, MA: Addison-
Wesley.
Altemeyer, B. 1981 Right-wing authoritarianism, Winnipeg: University
of Manitoba Press.
Banting, K. and Kymlicka, W. 2013 ‘Is There Really a Retreat from
Multiculturalism Policies? New Evidence from the Multiculturalism
Policy Index.’, Comparative European Politics 11(5): 577–98
Banting, K. and Kymlicka, W. (eds) 2006 Multiculturalism and the
Welfare State: Recognition and Redistribution in Contemporary
Democracies, Oxford: Oxford University Press.
Bobo, L. 1983 ‘Whites’ opposition to busing: Symbolic racism or realistic
group conflict?’ Journal of Personality and Social Psychology 45(6):
1196–210.
Bobo, L. 2001 ‘Racial Attitudes and Relations at the Close of the
Twentieth Century’ in N.J. Smelser, W.J. Wilson, and F. Mitchell (eds)
America Becoming: Racial Trends and Their Consequences, National
Academy Press.
29
CoDE (Centre on Dynamics of Ethnicity). 2014 ‘Addressing Ethnic
Inequalities in Social Mobility: Research Findings from the CoDE and
Cumberland Lodge Policy Workshop.’ (Accessed 29.06.2015:
http://www.ethnicity.ac.uk/medialibrary/briefings/policy/code-social-
mobility-briefing-Jun2014.pdf.)
Coenders, M. and Scheepers, P. 2008 'Changes in Resistance to the
Social Integration of Foreigners in Germany 1980–2000: Individual and
Contextual Determinants', Journal of Ethnic and Migration Studies
34(1): 1-26
Coenders, M. and Scheepers, P. 2003 ‘The Effect of Education on
Nationalism and Ethnic Exclusionism: An International Comparison’,
Political Psychology 24(2): 313–43.
Field, C. D. 2011 ‘Young British Muslims since 9/11: A Composite
Attitudinal Profile’, Religion, State and Society 39(2-3): 159–75.
Finney, N. and Simpson, L. 2009 Sleepwalking to Segregation?
Challenging Myths About Race and Migration, Bristol: Policy Press.
Ford, R. 2008 ‘Is racial prejudice declining in Britain?’, The British
Journal of Sociology 59(4): 609‐36.
Ford, R. 2011 'Acceptable and unacceptable immigrants: the ethnic
hierarchy in British immigration preferences', Journal of Ethnic and
Migration Studies 37(7): 1017-1037
Ford, R. and Goodwin, M. J. 2014 Revolt on the Right: Explaining
Support for the Radical Right in Britain. Abingdon: Routledge.
30
Hagendoorn, L. 1995 ‘Intergroup biases in multiple group systems: The
perception of ethnic hierarchies’, European Review of Social
Psychology 6(1): 199–228.
Heath, A. and Martin, J. 2013 ‘Can religious affiliation explain “ethnic”
inequalities in the labour market?’, Ethnic and Racial Studies 36(6):
1005–27.
Inglehart, R. 1997 Modernization and Postmodernization: Cultural,
Economic, and Political Change in 43 Societies, Princeton, NJ:
Princeton University Press.
Kalin, R. and Berry, J. W. 1996 ‘Interethnic Attitudes in Canada:
Ethnocentrism, Consensual Hierarchy and Reciprocity’, Canadian
Journal of Behavioural Science 28(4): 253-61
Mannheim, K. 1970 ‘The Problem of Generations’, Psychoanalytical
Review 57: 378–404.
Maxwell, R. 2012 Ethnic minority migrants in Britain and France:
integration trade-offs, Cambridge: Cambridge University Press.
Mudde, C. 2007 Populist Radical Right Parties in Europe, Cambridge:
Cambridge University Press.
Muttarak, R., and Heath, A. 2010 ‘Who intermarries in Britain?
Explaining ethnic diversity in intermarriage patterns’, The British
Journal of Sociology 61(2): 275–305.
ONS (Office for National Statistics). 2014 ‘Quarterly Labour Force
Survey, October - December, 2013’ [computer file]. 2nd Edition.
Colchester, Essex: UK Data Archive [distributor], SN:
7468: 10.5255/UKDA-SN-7468-2
31
OPCS (Office of Population Censuses and Surveys). 1985 ‘Labour
Force Survey, 1983’ [computer file]. Colchester, Essex: UK Data
Archive [distributor]. SN: 2029: 10.5255/UKDA-SN-2029-1
Perry, S. L. 2014 ‘More Like Us: How Religious Service Attendance
Hinders Interracial Romance’, Sociology of Religion 75(3): 442–62.
Pettigrew, T. F. and Meertens, R. W. 1995 ‘Subtle and Blatant
Prejudice In Western Europe’, European Journal of Social Psychology
25(1): 57–75.
Pettigrew, T. F. and Tropp, L. R. 2008 ‘How Does Intergroup Contact
Reduce Prejudice? Meta-Analytic Tests of Three Mediators’,
European Journal of Social Psychology 38(6): 922–934.
Putnam, R. 2007 ‘E Pluribus Unum: Diversity and Community in the
Twenty-First Century’, Scandinavian Political Studies 30(2): 137–74.
Saggar, S. 2008 Pariah Politics : Understanding Western Radical
Islamism and What Should Be Done, Oxford: Oxford University Press.
Sidanius, J. and Pratto, F. 1999 Social Dominance: An Intergroup
Theory of Social Hierarchy and Oppression, Cambridge: Cambridge
University Press.
Snellman, A. and Ekehammar, B. 2005 ‘Ethnic Hierarchies, Ethnic
Prejudice, and Social Dominance Orientation’, Journal of Community
and Applied Social Psychology 15(2): 83–94.
Song, M. 2004 ‘Introduction: Who’s at the Bottom? Examining Claims
About Racial Hierarchy’, Ethnic and Racial Studies 27(6): 859–77.
32
Storm, I. 2011 ‘”Christian Nations”? Ethnic Christianity and Anti-
Immigration Attitudes in Four Western European Countries’, Nordic
Journal of Religion and Society 24(1), 75–96.
Strabac, Z. and Listhaug, O. 2008 ‘Anti-Muslim prejudice in Europe: A
multilevel analysis of survey-data from 30 countries’. Social Science
Research 37: 268–286.
Tilley, J. R. 2005 ‘Research Note: Libertarian-Authoritarian Value
Change in Britain, 1974-2001’, Political Studies 53(2): 442–453.
Voas, D. and Crockett, A. 2005 ‘Religion in Britain: Neither Believing
nor Belonging’, Sociology, 39(1): 11–28.
Werbner, P. 2005 ‘Islamophobia: Incitement to Religious Hatred -
Legislating for a New Fear?’, Anthropology Today 21(1): 5–9.
33
Table I: Opposition to a minority in-law, ordered logistic regression
Asian in-law Caribbean in-law Model 1 Model 2 Model 1 Model 2 Coeff SE Coeff SE Coeff SE Coeff SE
Period (ref: 1980s)1991 -0.215* 0.091 -0.215* 0.091 -0.232* 0.097 -0.243* 0.0971994 -0.354*** 0.084 -0.353*** 0.084 -0.488*** 0.085 -0.486*** 0.0851996 -0.524*** 0.086 -0.523*** 0.087 -0.493*** 0.084 -0.495*** 0.0842013 -0.796*** 0.115 -0.786*** 0.116 -0.653*** 0.115 -0.642*** 0.115Sex (Female) -0.234*** 0.063 27.172*** 6.773 -0.135* 0.064 18.716** 6.733Birthyear centred 1.308*** 0.360 0.812* 0.381 0.378 0.346 -0.179 0.369Birthyear squared /1000 -0.341*** 0.093 -0.206* 0.098 -0.104 0.089 0.046 0.095Highest qualification (ref: no qualification)GCSE -0.159 0.082 25.257** 9.000 0.032 0.080 24.426** 8.747A levels -0.258** 0.091 33.994*** 9.417 -0.118 0.093 32.102** 9.935Degree -1.023*** 0.149 45.883** 16.084 -0.709*** 0.140 61.849*** 15.939Social class (ref: Unskilled/partly skilled)Skilled manual/nonmanual -0.064 0.076 -0.048 0.076 0.053 0.077 0.064 0.077Professional/managerial -0.216* 0.097 -0.217* 0.098 -0.170 0.098 -0.162 0.098Unemployed -0.186 0.143 -0.237 0.144 0.211 0.139 0.159 0.140Tenure (ref: Own) Rent council -0.221** 0.081 -0.228** 0.082 -0.165 0.084 -0.162 0.084Rent other -0.305** 0.110 -0.306** 0.110 -0.146 0.109 -0.142 0.110Marital status (ref: married)Separated/divorced -0.070 0.108 -0.091 0.109 -0.201 0.110 -0.218* 0.110Widowed -0.068 0.104 -0.121 0.106 -0.047 0.106 -0.087 0.108Never married -0.266* 0.106 -0.300** 0.107 -0.209 0.107 -0.243* 0.108Children in household -0.083 0.073 -0.082 0.074 0.111 0.072 0.102 0.072Religion (ref: No religion) CofE/Anglican 0.394*** 0.074 0.411*** 0.075 0.227** 0.074 0.237** 0.074Roman Catholic 0.107 0.115 0.131 0.115 0.193 0.118 0.214 0.118Other Christian 0.163 0.097 0.180 0.097 0.128 0.096 0.133 0.097Non-Christian 0.641* 0.316 0.723* 0.318 0.784** 0.287 0.833** 0.288Religious service attendance -0.157 0.083 -0.192* 0.084 -0.192* 0.085 -0.209* 0.086Highest qualification x birthyearGCSE*birthyear -0.013** 0.005 -0.013** 0.005A levels*birthyear -0.018*** 0.005 -0.017** 0.005Degree*birthyear -0.024** 0.008 -0.032*** 0.008Gender*birthyear -0.014*** 0.003 -0.010** 0.003Intercept 1 -1293.717 351.341 -780.210 372.020 -394.157 337.114 173.192 360.495Intercept 2 -1292.634 351.340 -779.123 372.020 -393.031 337.115 174.321 360.496N 4826 4826 4717 4717Log likelihood -4282.03 -4266.49 -4234.4 -4220.6Pseudo R2 0.066 0.0696 0.064 0.067BSA 1983-2013 (only white respondents), * p<0.05, ** p<0.01, *** p<0.001
34
Table II: Opposition to minority in-law (2013), ordered logistic regression
Muslim Asian Black African Caribbean East European
Coeff SE Coeff SE Coeff SE Coeff SE Coeff SESex (Female) -0.049 0.115 -0.404* 0.204 -0.303 0.224 -0.370 0.205 -0.240 0.213
Birth cohort (ref: born before 1940)
1940s-50s -0.247 0.179 -0.058 0.286 -0.720* 0.292 -0.128 0.286 0.205 0.302
1960s-70s -0.941*** 0.219 0.506 0.380 -1.628*** 0.403 -0.409 0.377 0.517 0.398
1980s-90s -1.509*** 0.272 -0.321 0.537 -1.782** 0.562 -0.411 0.455 -0.703 0.615
Highest qualification (ref: no qualification)
GCSE -0.102 0.167 -0.673* 0.277 -0.186 0.296 -1.051*** 0.271 -0.726* 0.292
A levels -0.445* 0.181 -0.747* 0.314 -0.662* 0.325 -1.199*** 0.308 -0.540 0.319
Degree -0.958*** 0.207 -1.491*** 0.410 -1.066** 0.395 -1.785*** 0.368 -1.688*** 0.453
Social class (ref: Unskilled/partly skilled)
Skilled manual/nonmanual 0.078 0.155 -0.074 0.250 0.064 0.274 -0.050 0.263 -0.326 0.257
Professional/managerial -0.239 0.179 -0.420 0.308 -0.447 0.324 0.042 0.315 -0.635* 0.312
Unemployed 0.023 0.308 -1.037 0.667 -1.012 0.790 0.222 0.488 -2.327* 1.060
Tenure (ref: Own)
Rent council -0.053 0.169 0.510 0.274 -0.563 0.311 0.464 0.273 0.604 0.293
Rent other 0.377* 0.183 -0.546 0.358 -0.116 0.415 0.528 0.300 -0.203 0.367
Marital status (ref: married)
Separated/divorced 0.040 0.166 0.183 0.302 0.029 0.320 -0.246 0.316 -0.079 0.320
Widowed -0.139 0.191 0.076 0.303 0.032 0.310 0.776* 0.300 -0.172 0.326
Never married -0.326 0.183 -0.195 0.316 -0.047 0.390 0.221 0.315 -0.338 0.339
Children in household -0.085 0.142 -0.712* 0.277 -0.006 0.293 0.210 0.256 -0.737* 0.290
Religion (ref: No religion)
CofE/Anglican 0.249 0.146 0.633* 0.256 0.549* 0.263 0.208 0.250 0.452 0.263
Roman Catholic -0.334 0.199 -0.005 0.393 -0.146 0.390 -1.162** 0.431 -0.438 0.445
Other Christian 0.019 0.155 0.670* 0.273 0.208 0.298 -0.589* 0.282 0.471 0.281
Non-Christian 0.099 0.410 1.424 0.925 1.911** 0.664 0.300 0.565 0.512 1.166
Religious service attendance 0.282 0.158 -0.162 0.272 -0.320 0.296 0.548 0.281 -0.427 0.303
Intercept 1 -1.121*** 0.288 0.026 0.466 -0.586 0.500 -0.319 0.479 0.116 0.489
Intercept 2 0.037 0.286 1.119 0.471 0.624 0.503 0.819 0.482 1.442*** 0.499
N 1334 674 683 680 678Log likelihood -1277.454 -427.3963 -390.148 -442.325 -386.548
Pseudo R2 0.077 0.097 0.123 0.111 0.110
BSA 2013 * p<0.05, ** p<0.01, *** p<0.001
35
Table III: Percentage who would mind if a close relative marries a […]:
Would mind…(%) Mind both Mind (1) only Mind (2) only Mind neither % N(1) Muslim and (2) Black African 20.3 28.0 1.8 49.9 100 932(1) Muslim and (2) Caribbean 23.9 25.9 1.3 48.9 100 919(1) Asian and (2) East European 16.3 7.5 3.4 72.9 100 910N 557 567 60 1577BSA 2013
36
Table IV: Would mind if a close relative marries a […]:
Only mind Muslim in-law
Mind Muslim and Black in-law
Coeff SE Coeff SE
Sex (Female) 0.218 0.135 -0.343 0.155
Birth cohort (ref: born before 1940)
1940s-50s 0.145 0.214 -0.302 0.213
1960s-70s -0.361 0.258 -0.807** 0.282
1980s-90s -1.070** 0.329 -0.830* 0.360
Highest qualification (ref: no qualification)
GCSE 0.670** 0.199 -0.557** 0.203
A levels 0.581** 0.215 -0.933*** 0.230
Degree 0.250 0.244 -1.405*** 0.281
Marital status (ref: married)
Separated/divorced -0.139 0.190 0.076 0.227
Widowed -0.801** 0.234 0.548* 0.226
Never married -0.526* 0.217 -0.033 0.254
Religion (ref: No religion)
CofE/Anglican -0.039 0.172 0.403* 0.188
Roman Catholic 0.237 0.227 -0.614* 0.303
Other Christian 0.265 0.179 0.027 0.209
Non-Christian -1.142 0.642 1.138* 0.466
Constant -1.257*** 0.335 0.167 0.356
N 1321 1321
Log likelihood -743.060 -603.726
Pseudo R2 0.049 0.124
BSA 2013 * p<0.05, ** p<0.01, *** p<0.001 (Controlled for social class, unemployment, tenure, marital status, children and religious service attendance)
37
Table V: Percentage bothered by a […] in-law by ethnoreligious group in 2010 (2013)
Black Caribbean African non-Muslim
African Muslim
Asian non-Muslim Asian Muslim
White (BSA 2013)
White in-law 14.24 22.11 18.12 28.24 37.57
Black in-law 41.62 42.14 23.12
Asian in-law 22.61 29.15 18.25 24.02
Muslim in-law 37.30 42.93 58.69 49.52
EMBES 2010 (BSA 2013)
38
Table VI: Ordered logistic regression: Bothered by […] in-law
White in-law Black in-law Asian in-law Muslim in-law Coeff SE Coeff SE Coeff SE Coeff SE
Sex (Female) 0.229 0.136 0.451** 0.159 0.147 0.184 0.191 0.150Birth cohort (ref: Before 1960)1960s-1970s 0.260 0.203 0.057 0.233 0.114 0.267 -0.057 0.2051980s-1990s 0.145 0.222 -0.199 0.267 -0.389 0.282 -0.351 0.229Born in UK -0.434** 0.158 -0.190 0.171 0.277 0.229 0.028 0.173Highest qualification (ref: No qualification)GCSE -0.210 0.202 -0.265 0.236 -0.231 0.285 -0.099 0.245A-levels -0.193 0.204 -0.448 0.242 -0.417 0.295 -0.273 0.234Degree -0.179 0.201 -0.240 0.234 -0.573 0.314 -0.047 0.234Ethnic/religious group (ref: Caribbean) (ref: Asian non-Muslim) (ref: Caribbean) (ref: Caribbean)African non-Muslim 0.116 0.244 0.430 0.226 0.010 0.197African Muslim -0.168 0.338 -0.061 0.339Asian non-Muslim 0.922*** 0.201 1.326*** 0.166Asian Muslim 1.100*** 0.189 -0.409* 0.172Discrimination 0.307* 0.140 0.342* 0.171 0.459* 0.183 0.381* 0.154Religious attendance 0.233*** 0.049 0.179** 0.057 0.293*** 0.072 0.372*** 0.058Intercept 1 2.133*** 0.326 0.468 0.335 2.587*** 0.440 1.578*** 0.340Intercept 2 2.908*** 0.339 1.183*** 0.339 3.237*** 0.449 2.420*** 0.346N 2117 1090 984 1277Log Likelihood -1482.723 -1209.188 -481.281 -1202.445Pseudo R2 0.048 0.026 0.037 0.081EMBES 2010 * p<0.05, ** p<0.01, *** p<0.001
39
Figure I: Percentage who minds if a close relative marries Asian or Caribbean
840
10
20
30
40
50
60
70AsianCaribbean
BSA 1983-2013 (only white respondents)
40
Figure II: Percentage who minds ethnic minority in-law by birth cohort
1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s
0
10
20
30
40
50
60
70
80
MuslimBlack AfricanAsiansCaribbeanEast European
BSA 2013 (only white respondents)
41
Figure III: Predicted probability of minding if relative marries a Muslim by birth cohort
1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
Mind MuslimMind Muslim onlyMind Muslim and Black
BSA 2013 (Controlling for all variables in Table IV)
42
Figure IV: Percentage who mind if a close relative marries a white person […] by year and ethnic group
year 1993-4
year 2010
year 1993-4
year 2010
year 1993-4
year 2010
year 1993-4
year 2010
year 1993-4
year 2010
Pakistani Bangladeshi Indian Black Caribbean Chinese Black African
0
10
20
30
40
50
60
4th National Survey 1994/ EMBES 2010. Not all ethnic groups were included in both surveys.
43
Appendix
Table AI: List of variables used in the analysis
British Social Attitudes Survey 1983-2013
Variable Question Original Categories Recoded Categories Percent / Mean(SD)
N
Asian in-lawvvi Do you think that most white people in Britain would mind or not mind if one of their close relatives were to marry a person of [Asian / West Indian] origin? b) And you personally? Would you mind or not mind? (If mind) A lot or a little?
Not mind 1) Not mind [Ref] 57.4 6961Mind a little 2) Mind a little 20.2Mind a lot 3) Mind a lot 22.4
West Indian in-law [Seeabove] Not mind 1) Not mind [Ref] 56.3 6751Mind a little 2) Mind a little 21.4Mind a lot 3) Mind a lot 22.3
Year Survey year 1983, 1984, 1986, 1989 0) 1980s 56.2 302641991 1) 1991 9.61994 2) 1994 11.51996 3) 1996 12.02013 5) 2013 10.7
Sex Sex of respondent 1) Male [Ref] 1) Male [Ref] 45.1 302642) Female 2) Female 54.9
Birthyear centred Recoded Year of birth Range [1889,1996] Range [-57.2,49.8] 0.0(19.1) 30165Birthyear squared / 1000
Recoded Year of birth Range[1889,1996] Range [3568.3,3984.0] 3782.2 (74.4) 30165
Highest qualification Highest educational qualification obtained No qualification 0) No qualification [Ref] 38.0 26487CSE or equiv. / O-level or equiv. / Foreign or other
1) GCSE 27.9
A-level or equiv. / Higher education below degree 2) A-level 24.1Degree / Postgraduate degree 3) Degree 10.0
Social class Registrar General's social class Class V / Class IV / Never had a job 0) Unskilled / partly skilled [Ref] 28.2 29637Class III manual / Class III non-manual 1) Skilled manual / non-manual 44.1Class II / Class I 2) Professional / managerial 27.7
Unemployed Current economic position 10 other categories of employment and economic inactivity 0) No [Ref] 94.0 30264Unemployed 1) Yes 6.0
Tenure Accommodation tenure Own outright / Own on mortgage 1) Own [Ref] 68.2 29986Rent local authority / Rent new town 2) Rent council 22.2Housing association / Property company / Rent from employer / Other organisation / Rent from
3) Rent other 9.6
44
relative / Rent from other individualMarital status Marital status (summary) Married 1) Married [Ref] 63.5 30217
Separated/divorced 2) Separated/divorced 8.3Widowed 3) Widowed 10.6Never married 4) Never married 17.8
Children in household What is (name)'s relationship to you? [Recoded from household grid ]
Other relationship of household members 0)No son / daugther in household [Ref] 59.0 30264Son / daugher [incl step/adopted] 1) Son/ daughter of respondent lives in
the household41.0
Religion Religion summary No religion 0) No religion [Ref.] 37.6 26718C of E / Anglican 1) C of E / Anglican 33.9Roman Catholic 2) Roman Catholic 9.8Other Christian 3) Other Christian 15.7Non-Christian 4) Non-Christian 3.0
Religious service attendance
Frequency of attendance No religion / Twice a year / Once a year / Less often / Never 0) Less often or never [Ref] 79.2 29448Once a week / Once in two weeks / Once a month 1) At least once a month 20.8
British Social Attitudes Survey 2013
Variable Question Original Categories Recoded Categories Percent NBirth cohort Year of birth Range [1916,1939] 1) before 1940s 14.11 3239
Range [1940,1959] 2) 1940s-50s 32.66Range [1960,1979] 3) 1960s-70s 33.59Range [1980,1996] 4) 1980s-90s 19.64
Asian in-law Do you think that most white people in Britain would mind or not mind if one of their close relatives were to marry a person of [Asian / West Indian/ Black African / Muslim / East Asian] origin? b) And you personally? Would you mind or not mind? (If mind) A lot or a little?
Not mind 1) Not mind [Ref] 76.0 920Mind a little 2) Mind a little 13.0Mind a lot 3) Mind a lot 11.0
West Indian in-law [See above] Not mind 1) Not mind [Ref] 75.4 951Mind a little 2) Mind a little 13.8Mind a lot 3) Mind a lot 10.8
Black African in-law [See above] Not mind 1) Not mind [Ref] 78.4 957Mind a little 2) Mind a little 13.0Mind a lot 3) Mind a lot 8.7
Muslim in-law [See above] Not mind 1) Not mind [Ref] 50.5 1874Mind a little 2) Mind a little 22.5Mind a lot 3) Mind a lot 27.1
45
East Asian in-la [See above] Not mind 1) Not mind [Ref] 80.4 921Mind a little 2) Mind a little 12.3Mind a lot 3) Mind a lot 7.4
EMBES 2010
Variable Question Original Categories Recoded Categories Percent NFor each of the following statements please indicate how much it would bother you…
White in-lawvii If one of your close relatives were to marry a White person [if not White]
Not at all / Not very much 1) Not mind [Ref] 73.54 2,604A little 2) Mind a little 10.64Rather a lot / A great deal 3) Mind a lot 15.82
Black in-law If one of your close relatives were to marry a Black person [if not Black]
Not at all / Not very much 1) Not mind [Ref] 60.61 1,508A little 2) Mind a little 14.66Rather a lot / A great deal 3) Mind a lot 24.73
Asian in-law If one of your close relatives were to marry an Asian person [if not Asian]
Not at all / Not very much 1) Not mind [Ref] 75.75 1,171A little 2) Mind a little 9.82Rather a lot / A great deal 3) Mind a lot 14.43
Muslim in-law If one of your close relatives were to marry a Muslim person [if not Muslim]
Not at all / Not very much 1) Not mind [Ref] 54.05 1,543A little 2) Mind a little 16.27Rather a lot / A great deal 3) Mind a lot 29.68
Sex Gender of respondent Male 0) Male [Ref] 48.04 2,787Female 1) Female 51.96
Birth cohort Year of birth Range [1913,1959] 1) Before 1960 [Ref] 18.98 2,787Range [1960,1979] 2) 1960s-1970s 47.76Range [1980,1992] 3) 1980s-1990s 33.26
Born in UK Country of birth [Any other country] 1) Born outside UK [Ref] 63.83 2,759England/ Scotland / Wales / Britain / UK 2) Born in UK 36.17
Highest qualification Do you have any British or overseas educational or work-related qualifications? [If British:] Which is the highest British qualification you have? [If overseas:] Which phrase on this card comes closest to describing what sort of overseas qualification this is?
0) No qualification [Ref] 23.75 2,3541) GCSE or eq 28.592) A levels or eq 21.843) Degree 25.83
Ethnic/religious group Screen 5 Ethnicities + Do you regard yourself as belonging to any particular religion? If yes, which one?
Black Caribbean 1) Black Caribbean [Ref] 23.97 2,787Black African + Not Muslim 2) African non-Muslim 15.29Black African + Muslim 3) African Muslim 5.49Indian / Pakistani / Bangladeshi + Not Muslim 4) Asian non-Muslim 20.02Indian / Pakistani / Bangladeshi + Muslim 5) Asian Muslim 35.24
Discrimination In the past 5 years, do you feel that you have No / Language or accent / Age / Gender / Sexuality / 0) No 67.32 2,659
46
experienced discrimination or been treated unfairly by others in the UK because of your ethnicity, race, skin colour, language, accent, religion, age, gender, sexuality or disability?/ If yes, do you feel you have been discriminated against for any of these reasons?
DisabilityEthnicity, race or skin colour / Religion 1) Yes 32.68
Religious attendance Do you regard yourself as belonging to any particular religion? / Frequency of religious worship with others
No / Not at all 1) Not at all 22.31 2,761Yes + Occasionally / only on festivals 2) Occasionally / only on festivals 19.81Yes + At least once a month 3) At least once a month 10.83Yes + At least once a week 4) At least once a week 33.65Yes + At least once a day 5) At least once a day 13.4
Fourth National Survey of Ethnic Minorities 1993 / 1994Variable Question Original Categories Recoded Categories Percent N
Ethnic groupviii Ethnic origin White 1) White 35.56 8,012Caribbean 2) Caribbean 14.93Indian 3) Indian 24.08Pakistani 5) Pakistani 15.3Bangladeshi 6) Bangladeshi 7.46Chinese 7) Chinese 2.67
In which country were you born? / Family origins
African Asian Classified as 3/5/6 according to country of birth or family origins
White in-law Would you personally mind if a close relative were to marry a white person?
No 0) No 62.83 2,381Yes 1) Yes 37.17
47
Non-response and “Don’t know” has been coded as missing throughout.
48
Figure AI: Percentage who mind (a little or a lot) if their relative marries an Asian or Caribbean, by birth cohort and survey year
BSA (1983-2013)
49
Figure AII: Predicted probabilities of opposition to Asian or Caribbean in-law
1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8Mind Asian in-law No qual
GCSE
A-level
Degree
1900s 1910s 1920s 1930s 1940s 1950s 1960s 1970s 1980s 1990s0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8Mind Caribbean in-law No qual
GCSE
A-level
Degree
BSA 1983-2013 (only white respondents) – Average marginal effects based on logistic regression models, controlling for all variables shown in Table 1.
50
i This figure is based on the Labour Force Surveys and includes all non-white
minorities.
ii The question asked used the term ‘West Indian’ in all years of the BSA. However, to
be consistent with the other survey data we will refer to this category as ‘Caribbean’
in the remainder of this article.
iii The model does not fit the proportional-odds assumption or the parallel regression
assumption for either attitudes to Asian in-laws or Caribbean in-laws, due to the
relationship with education. However, as all the other variables meet the
assumptions, and have the same relationship in the multinomial, logistic and ordered
logistic regressions, we only show the ordered logistic models here. Multinomial
regressions (available on request) show that all levels of education are negatively
associated with ‘mind a lot’, whereas only degree level is negatively associated with
‘mind a little’. GCSE level education is positively associated with minding ‘a little’ if a
relative marries a Caribbean.
iv In alternative models, age of migration to, and length of residence in the UK were
both significant in the same direction, but due to multicollinearity these variables
were excluded from the model.
v For some of the analysis the three category ordinal outcome variables (how much the respondent minds if their relatives marry someone of a different ethnic background) in the BSA and EMBES has been recoded into a binary variable where Don't mind (1) = 0 is contrasted with Mind (2/3) = 1
vi In the BSA surveys before 2013 one half of the sample was asked about Asian origin and the other about black or West Indian origin. In 2013 there were three versions of the questionnaire and each respondent were only asked about two of the five ethnic minority groups: 1. Asian and East European; 2. Black African and Muslim; 3. West Indian and Muslim. The exact question wording was: a) Do you think that most white people in Britain would mind or not mind if one of their close relatives were to marry a person of [Asian] origin? b) And you personally? Would you mind or not mind? (If mind) A lot or a little? The variable (b) was recoded into a dichotomous variable with “mind a little” and “mind a lot” coded as 1 and “not mind at all” coded as 0. We refer to “West Indian” as “Caribbean” throughout the paper
vii In EMBES, respondents were asked about all except their own ethnic or religious group. In order to test the reciprocity hypothesis (H2a), they have been divided into categories that were as close as possible to the categories they were asked about, (and that white respondents were asked about in the BSA 2013). The original ethnic categories were Black Caribbean, Black African, Indian, Pakistani and Bangladeshi, and the respondents were also asked their religious affiliation which enabled us to distinguish between Muslims and non-Muslims.
viii The FNAS ethnicity variable had a category “African Asians”, which have been redistributed according to their country of origin as either Indian, Pakistani or Bangladeshi to make it comparable to the EMBES categorisation.