Cross-Cultural Patterns of Interracial Marriage
Автор работы: Пользователь скрыл имя, 07 Сентября 2012 в 18:12, контрольная работа
Описание
This paper compares patterns of interracial marriage in seven different cultural
contexts. Four aspects of intermarriage are considered. First, log-linear models are as
estimated to gage the extent of overall homogamy and race specific homogamy in each
setting. Second, residuals from these models are used to assess gender differences in
intermarriage.
Работа состоит из 1 файл
1
Cross-Cultural Patterns of Interracial Marriage
By
Tim B. Heaton
And
Cardell K. Jacobson
Department of Sociology
Brigham Young University
September 2004
2
Cross-Cultural Patterns of Interracial Marriage
Abstract
This paper compares patterns of interracial marriage in seven different cultural
contexts. Four aspects of intermarriage are considered. First, log-linear models are as
estimated to gage the extent of overall homogamy and race specific homogamy in each
setting. Second, residuals from these models are used to assess gender differences in
intermarriage. Multinomial logistic regression is then used to evaluate age and
educational differences. Age is included as a surrogate for trends over time, and
education is included as a measure of social status. We hypothesize that intermarriage
will increase over time-but not necessarily at the same rate for each racial group or racial
category. We also hypothesize that higher education is associated with higher rates of
outmarriage from low-status groups, but lower rates of outmarriage from high-status
groups. In other words, the average difference in status between two groups will be
associated with the degree of influence education has on intermarriage.
Census data were obtained for each of the cultural contexts. We begin with the
United States because a substantial body of research focuses on the United States. Canada
shares some culture, history, and language with the United States, but does not have a
legacy of slavery and has different patterns of immigration. New Zealand, like the U.S.
and Canada, has as history of British and European settlement, but the indigenous
population of New Zealand is a much larger share of the total population. In contrast to
these three immigrant societies, South Africa experienced colonization, but the
indigenous population remained the numerical majority. South Africa also had a period
when contact between groups was legally and socially restricted. The next two data sets
are from provinces in China. Beijing is selected because it is the political and educational
center of the society. We also include a province where a minority group is the numerical
majority. Finally, we include Hawaii because of its reputation as a melting pot for a wide
variety of different groups.
3
Ideological, social, and political forces have been working to reduce racial
discrimination and to create more equal relationships among racial and ethnic groups.
Marriage is often viewed as the last barrier to racial integration, making intermarriage a
critical indicator of inter-group relationships. While numerous researchers have examined
inter-group marriage between racial and ethnic groups in the United States, fewer have
examined factors associated with inter-group marriage in an international perspective.
Using recent census data, we examine the factors predicting inter-group marriage in the
United States, Canada, New Zealand, South Africa, and parts of the Peoples Republic of
China. We also include Hawaii as a special case since that state has a long history of
racial tolerance and intermarriage. The six societies are geographically diverse, have
different racial and ethnic histories, and compositions.
Increased international migration in recent decades has increased the racial and
ethnic diversity in a number of countries, resulting in greater social proximity between
individuals from different groups and contact in the workplace and other social
institutions. This international migration is, at least in part, a major factor leading to
interracial unions (Sung, 1990; Tzeng, 2000). At the same time, cultural norms against
inter-group marriage have decreased resulting in less pressure to marry homogamously.
The United States, Canada, and New Zealand are all immigrant societies in which Euro-
Americans or whites are the dominant group.
South Africa provides a contrast to these immigrant societies because while it was
a colonized country, the indigenous peoples remained a numerical majority. The whites,
of course, had the political, social, and economic power.
China provides an additional contrast. While many people consider current-day
China to be ethnically homogamous, China actually is composed of 56 groups, 55 of
which China recognizes officially as minority groups. Altogether, the minority groups
comprise four percent of the population. The groups are not evenly distributed across
China, however. We examine two areas in China, Beijing, its capital city and Xinjiang
Province in Northwest China. Xinjiang Province is composed of Uyghers (percent), Han
(percent), as well as several other smaller groups. Though the Han are the dominant
group in China, they number less than half the population in Xinjiang Province and are
mostly immigrants or the children of immigrants in the province.
In this paper we examine three factors that have been shown to be related to inter-
group in previous research and compare the factors across the six contexts. More
specifically; gender, age and education have been used as indicators of cultural and social
indicators of racial and ethnic status.
Gender Imbalances in Interracial Marriage
Gender imbalances in interracial marriages are common in the United States and
Canada (Qian, 1997; Jacobson and Heaton, 2000), and Canada (Ram, 1990; Tzeng,
2000). In both countries white male-Asian female marriages are much more common
than the reverse. Black-white marriages in the United States, however, show an inverse
pattern with Black male-white female marriage numbering roughly twice the number of
white male-black female marriages. The Asian female-white male marriages in Canada
also tend to occur when the wife has high educational levels or works more often than her
husband (Tzeng, 2000).
4
Studies conducted in the United States also show that interethnic marriage is more
common for foreign females than it is for foreign males (Lee and Yamanaka, 1990). This
is especially true for those from male-dominated cultures such as Japan (Tinker, 1982).
Some of the gender differences in interracial may reflect decreasing traditional
norms. In most Asian cultures the family name is continued through the sons, not
daughters, and historically sons have been valued more than daughters. While this gives
Asian sons more latitude in choosing a mate, the daughters may also look outside their
traditional community to seek a mate (Koo 1985; Watson, Jashcok and Miers, 1994;
Cheung, 1997; Hays, 2001).
A similar phenomenon may be occurring in Black-White marriages. As
traditional norms about marriage decline white women may seek partners outside of their
own group.
Generation or Age
The literature on inter-group marriage suggests that younger generations are
consistently more likely than older generations both to approve of inter-group marriage
(Schuman et al., 1997) and to marry exogamously. Research in the United States has
consistently shown age to be inversely associated with inter-group marriage (Gurak and
Fitzpatrick, 1982; Tinker, 1982; Wong, 1989; Sung, 1990; Lee and Yamanaka, 1990;
Qian, 1997; Jacobson and Heaton, 2000). Lee and Yamanaka’s (1990) study, for
example, found that native-born Asian Americans were three times more likely to
intermarry than foreign-born women. Researchers have suggested that younger
generations have shed discriminatory beliefs and practices, resulting in a more open
attitude towards other races (Gurak and Fitzpatrick, 1982). Some researchers also believe
that younger generations are better able to adapt to a new culture and society (Tinker,
1982; Gurak and Fitzpatrick, 1982). Further, Younger generations, or second or
subsequent generations, generally are more assimilated and thus less likely to have strong
ties to their native land and cultures (Lee and Yamanaka 1990; Sung, 1990; Wong, 1989).
Generational status and intermarriage are also important in Canada where younger
generations are significantly more likely than younger generations to marry exogamously
(Ram, 1990; Tzeng, 2000). For white ethnic groups the reverse is true (Kalbach, 2000).
For white ethnic groups, couples often are married before their arrival in Canada
(Kalbach, 2003). A second explanation for the lower rates of outmarriage among the
younger generations is that they often do not speak either of the official Canadian
languages. Tzeng (2000), for example, estimates that 98 percent of those who are
outmarried are capable of speaking one or both of the official languages.
Cultural factors, no doubt, also affect the rates of intermarriage. Older members
of society are more simply more set in their ways, adhere to older social norms, and thus
are less to intermarry. Furthermore the norms and acceptance of Intergroup marriage
have changed dramatically over the past 40 or so years (see Schuman et al., 1997).
Education
A number of researchers have found that individuals from higher socio-economic
status groups are more likely than others to exhibit both higher acceptance of others and
are more likely to marry exogamously (e.g. Sandefur and McKinnel, 1986; Tinker, 1982;
Schoen, Wooldredge, and Thomas, 1989). While inter-group contact within educational
institutions often remains less than ideal for the improvement of racial attitudes
(Schaefer, 1996), educational institutions provide opportunities for increased inter-group
5
contact and acceptance that may not exist in other parts of society. The increased
approval may result from at least two aspects of education. First, attendance at college
may facilitate normative acceptance of others. Indeed, the research indicates that those
who attend college are the most likely of all educational levels to approve of inter-group
marriage (see Glick, 1970; Tinker, 1973; Fitzpatrick and Gurak, 1979, Kitano, 1984;
Mayer, 1985; Sung, 1990; Qian and Preston, 1993; Tzeng, 2000; Fu and Heaton 199??;
Tzeng, 2000).
Other researchers have found that education is positively associated with actual
rates of intermarriage (Sandefur and McKinnel, 1986; Tucker and Mitchell-Kernan,
1990: Kalmijn, 1993; Qian, 1997; Heaton and Jacobson, 2000; Jacobs and Labov, 2002;
Jacobson and Heaton, 2003; Jacobson, Yaw, and Heaton, 2004). Some research,
however, has not shown this positive association (Hwang, Saenz, and Aguirre 1995), and
some shows no association at all (Kitano et al., 1984).
The use of education in the analysis presents some potential problems, and the
actual effects of education may be underestimated. For example, the mere co-presence of
other groups does not necessarily lead to contact, the co-presence of large numbers of
members of other groups may actually present a racial threat to some whites (see Bobo,
1983; Bobo and Kluegel, 1993; Quillian, 1996; Taylor; 1998).
Additionally, how education is measured varies across nations. And some who
have attended college may have received their education years or decades earlier. The
effects of education may thus be underestimated, and the relationship between education
and inter-group marriage may not be as strong for older generations as it is for younger
generations.
Some research has also found that educational homogamy is related to the level of
economic development. Smits, Ultee, and Lammers (1998) used educational
homogamy rates as a measure of openness. They found that as the level of development
increases, educational homogamy increases. For the highest levels of economic
development, however, educational homogamy decreases again. While the United States,
Canada, and New Zealand are economically developed countries, most parts of China are
not, and parts of South Africa are not. These differences may account for some of the
differences we find across the six countries.
Studies generally show educational homogamy, that people tend to marry within
their own level of education (Qian, 1999). Studies conducted with Chinese Americans
also show that Chinese Americans who intermarry have more education, regardless of
gender (Sung, 1990; Wong, 1989). But, education sometimes interacts with gender or
generation. For example, Tzeng (2000) found that the effect of the wife’s educational
attainment has twice the effect that men’s education did in predicting interracial
marriage. And females in the United States are more likely to marry someone with a
higher degree of education (Qian, 1999). At the same time, minority males are more
likely to marry a female with a lower educational status in exchange for her higher ethnic
status (Kalmijn, 1993; Lee and Yamanaka, 1990).
Data sources
Data for the analysis are taken from national census. Although census data are
limited by lack of information about conditions when the marriage occurred, they are the
most comparable source of data for cross-cultural analysis. Public use samples we
6
available for the US in 2000, South Africa in 2001, and China in 1990. Special
tabulations were obtained for Canada in 1996 and New Zealand in 1996.
In each census, we selected currently married partners of the opposite sex
matched on household id numbers. A few households with multiple married couples were
eliminated to insure correct matching of partners.
Loglinear models are used to examine patterns of intermarriage and sex differences.
Multinomial logistic regression is employed to identify patterns by age and education.
Results
Table 1 shows the racial/ethnic distribution for groups used here. Some smaller
groups did not include sufficient cases for detailed analysis. Heterogeneous groups are
also lumped together. Still these numbers do allow us to compare general trends, with
roughly similar numbers of groups in each context.
Many contexts have a single large majority group. In the U.S., Canada, and New
Zealand Whites constitute at least 80 percent of the total population. Han are also a vast
majority in Beijing. Blacks are less of a majority in South Africa, but still constitute two-
thirds of the population. Hawaii is more divided with Asians totaling slightly over half of
the population. Finally, Xinjiang Province has to groups that are roughly equal in size,
along with two other smaller groups.
In most cases, the distribution of husbands and wives is nearly equal. Hawaii is
the only exception. Selective patterns of marriage and migration could each account for
the differences in Hawaii.
Results of model fitting are shown in Table 2. As expected, the model of simple
independence does not fit in any of the settings. Homogamy is common in each of these
contexts. In each case, adding one parameter for the universal tendency toward
homogamy accounts for the large majority of the association between husband’s and
wife’s race/ethnicity. In most cases, the homogamy effect explains over 90 percent of the
association. The exceptions are Beijing and New Zealand, when homogamy still accounts
for over 80 percent of the association.
The third model allows each race/ethnic groups to have its unique endogamy
parameter. In each country we see an additional improvement in model fit, indicating that
the degree of homogamy varies across groups.
Parameters for universal homogamy from model 2 and group specific homogamy
from model 3 are shown in Table 3. In-marriage is comparative low in Hawaii, New
Zealand and Beijing. In these context, and for the groups included here, groups are about
ten time more likely to marry endogamously that would be expected by chance. The U.S.
and Canada have values over three times greater than in these three contexts. Still, these
values are small compared to South Africa and Xinjiang Province where parameters
exceed 200. Clearly, there is substantial variation in tendencies toward in-marriage.
Individual group parameters show even greater variation. Where Whites are a
majority, they tend to have low rates of in-marriage. Likewise, Hans in Beijing have very
low rates of endogamy compared with all other groups considered here. Presumably,
when minority groups marry someone from a different group, they tend to marry into the
majority. This could be because of contact and tendencies to assimilate into the dominant
society. In contrast, Whites in South have the second to highest parameter of any group
considered. South African Whites are over 2000 time more likely to marry endogamously
7
that would be expected on the basis of change. The only other group in this range is the
Uighur of Xinjiang Province.
Blacks are the most isolated group in the United States and Hawaii. However, in
Canada where there is not a legacy of slavery, Blacks are less isolated in the marriage
market than Arabs or Hispanics. Indigenous populations In the U.S., Canada, Hawaii, and
New Zealand have relatively low rates of endogamy. Asians generally have higher
endogamy rates than Native Americans. Asian in-marriage is quite high in New Zeland
and especially high in South Africa.
Hispanic rates of inmarriage vary by context. The ratio for Hispanics is five times
higher in the U.S. than in Hawaii, but 12 time higher in Canada than in the U.S.
Not surprisingly, given its history of race relations, South Africa has high rates of
in-marriage for most groups. Only the Coloureds, who have a long history of
intermarriage, have moderate rates of inmarriage.
Gender
The final model fits an additional parameter for each combination of
race/ethnicity on the off-diagonal. The remaining association occurs because males and
females have different patterns of intermarriage. In the two Chinese samples and in South
Africa, this remaining association is small and statistically nonsignificant. Each other
country shows some gender difference. In the U.S. there is a larger than expected number
of marriages between White men and Asian women, and between Black men and White
women. Also, Asian men are more likely to marry Hispanic women, but Hispanic men do
not marry Asian women.
In Hawaii, there is also a tendency for White men to marry Asian women. But
Asian men are inclined toward Pacific Island or Hispanic wives. There is also a tendency
for White women to marry Hispanic and Pacificic Island husbands.
The tendency for White men to marry Asian women is also evident in Canada.
There are also more marriages bewteen White men and Black or Hispanic women than
would be expected. In contrast, Asian men appear to prefer Native Canadians and
Hispanic wives. White women show a tendency to marry Native and Arabic men.
In New Zealand, European men apparently prefer to marry women from Asia or
other pacific Islands rather than Maoris. Maori men are more likely to marry European
women. There are also higher than expected marriages between men from other Pacific
Islands and Maori women, and between Asian men and women from the Pacific Islands
including Maoris.
Age
Age is included as a surrogate for trends. Age is not a perfect indicator because
intermarriage could be associated with older age at marriage, higher rates of marital
dissolution, and differential mortality. But given the lack of longitudinal data for many
societies, age is about the best we can do.
Table 4 shows coefficients for age. In the U.S., most effects are negative,
implying an increase in intermarriage. The greatest changes are for marriages between
Blacks and Whites, and Blacks and Hispanics. Native Americans have a longer history of
inter-marriage such that coefficients tend to be less negative for them.
In Hawaii, age coefficients are mixed. Marriages between white men and Pacific
Island or Asian women apparently have been occurring for a long enough period such
8
that no age pattern is evident. There is more evidence that marriage with Hispanics is
becoming more common, but this is not universal. Still the overall tendency is for
coefficients to be negative, suggesting that the tradition of intermarriage in Hawaii has
accelerated somewhat.
Most of the age coefficients in Canada are negative and statistically significant,
implying increases in intermarriage. But there are some interesting exceptions to the
pattern. Some coefficients for marriages between whites and natives are positive.
The New Zealand data does not include husband’s age. People providing the data
did not want to give too much detail because of confidentiality of data. Thus, we used
wife’s age in each set of equations. The coefficients are generally negative, implying
increases in inter-group marriage. The exceptions are for marriages between European
men and Maori or Pacific Island women, and between European women and Maori men.
Perhaps these marriages have been occurring for a long enough period that recent
increases are not evident.
In South Africa, most age coefficients are negative, the most interesting exception
being marriages between Whites and Blacks. Patterns suggest that race barriers between
Blacks, Coloureds and Asians are eroding. Likewise, barriers between Coloureds, Asians
and Whites are also eroding. But the gap between Whites and Blacks is apparently as
wide as ever.
All of the coefficients for Beijing are negative supporting the hypothesis that
intermarriage is on the rise. There were not enough intermarriages in Xinjiang to justify
more detailed.
Education
Education coefficients are reported in Table 5. In the U.S., twenty-five of the
thirty-four coefficients are positive, indicating that education facilitates out-marriage
across many groups. The coefficients are particularly large for intermarriages involving
marriages to Asians. But a few coefficients are negative. White men who marry Native
Americans are less educated than White men who marry endogamously. White women
who marry Native Americans or Hispanics have lower educational attainment. Asian
women who marry black men also have comparatively low educational attainment.
Educational attainment is higher for whites and Asians than for Blacks, Native
Americans and Hispanics. Averaging coefficients across groups shows that people who
marry from higher education groups into lower education groups tend to have lower
status, and those from less educated groups who marry into more educated groups have
higher status. Thus, there is support for both hypotheses that more education facilitates
intermarriage, and that this is especially true for people from groups with lower overall
education. There is not a clear tendency for status exchange to be gender specific. For
example, coefficients for Black, Native American or Hispanic men who marry White
women are comparable to coefficients for Black, Native American or Hispanic women
who marry White men.
Since patterns are complex, and additional step was taken to summarize education
effects. An aggregate data file was created with an observation for each possible pair of
race combinations. The dependent variable is the education coefficient from the
multinomial regression reported in Table 5. Average education for the race group in each
pair was included as well as the gender of the person whose age and education were
9
included in the multinomial regression are also included. In essence, this is a multi-level
analysis. Because of the small number of pairs in some countries, the large sample sizes,
and the complexity of multinomial regression, multi-level statistical programs were not
used in the analysis.
Regression models are reported in Table 6. The dependent variable is the
education coefficient. Independent variables include the difference in education for the
two groups being compared, gender, and a gender-education gap interaction. Coefficents
from these regressions are used to test three hypotheses. First, the constant term test the
hypothesis that education facilitates intermarriage. Positive values are consistent with this
hypothesis. Second, the education gap variable tests the hypothesis that the importance of
education depends on the difference in the average education levels for the groups being
compared. Finally, the interaction term tests the exchange hypothesis that education
matters more for men than for women.
In the United States, only the first hypothesis is supported. On average,
coefficients are positive implying that higher education is associated with higher
probabilities of out-marriage. The education gap variable and the interaction are not
statistically significant and the explained variance is small. In the U.S., it appears that the
importance of education is not contingent on the average education of ethnic groups.
In Hawaii, Whites who marry out tend to have lower education that those who
marry endogamously. In contrast, other groups who marry whites have above average
educational status. With some exceptions that are not statistically significant, coefficients
for intermarriage between Pacific Islanders, Asians and Hispanics are positive. Thus,
higher education appears to facilitate out-marriage, except for whites. Higher order
analysis reported in Table 6 indicates that, on average, education is not a strong predictor
of out-marriage. However, the education difference between groups is a good predictor of
the importance of education, explaining over 60 percent of the variation in the magnitude
of education coefficients. The gender interaction indicates that higher education is less
important in facilitating female out-marriage into higher status groups or deterring female
out-marriage to lower status groups.
In New Zealand, White and Asian men who marry Maori women have lower
education, as is the case for White or Asian women who marry Maori or Pacific Island
men. People who marry exagamously and have White or Asian partners tend to have
higher education, especially if they come from the lower status Maori or Pacific Island
groups. Thus, New Zealand is consistent with the hypothesis that people who marry into
higher status groups have more education, but the reverse is true for people who marry
into lower status groups. Regression results for New Zealand support each of the three
hypotheses regarding education. The overall effect of education is to facilitate out-
marriage. This is especially the case for out-marriage to higher status groups. Indeed,
education deters out-marriage to lower status groups. Finally, the magnitude of
educational exchange is more salient for males than females.
Patterns of educational exchange are also quite consistent. Higher education is
associated with greater chances for marriage into higher status groups and lower
education is associated with greater chances for marriage into lower status groups. The
average education effect is slightly positive and the gender interaction is small. So that
the educational exchange described above is the dominant patters in the data.
10
Education coefficients for Beijing indicate that Hui and Manchu men who marry
Han women have above average education. Likewise, Han men who marry Hui women
have above average education. Coefficients for marriage of Han men and Manchu
women are negative. There are not enough pairs to run a regression predicting magnitude
of education coefficients, but the correlation between the education gap between pairs of
ethnic groups and the magnitude of the education coefficient is a very small -.057. Thus
Beijing data do not support the education exchange hypothesis.
There were not enough intermarriages in Xinjiang to justify further analysis.
Conclusions
Racial and ethnic homogamy are evident in each of the societies we consider here.
But rates of in-group marriage vary dramatically across societies, and across groups
within societies.
There is a common pattern of marriage between white men and Asian women.
Other patterns (black men and white men) are country specific.
Age differences suggest a general pattern of increase in intermarriage, but not for
some groups with a longer history of intermarriage, or for groups with persisting gaps
such as whites and blacks in South Africa.
There is strong support for the three education hypothesis in some countries
(Canada and New Zealand), partial support in others (USA, South Africa and Hawaii).
11
References
Cheung, Fanny M. 1997. “Engendering Hong Kong Society: A Gender Perspective of
Women’s Status.” Hong Kong: Chinese University Press.
Gladney, Dru C. 1996. “Muslim Chinese: Ethnic Nationalism in the People’s Republic.”
2
nd
Ed. Cambridge, Mass.: Council on East Asian Studies, Harvard University: Harvard
University Press.
Heaton, Tim B., and Cardell K. Jacobson. 2000. “Intergroup Marriage: An Examination
of Opportunity Structures.” Sociological Inquiry 70:30-41.
Hwang, Sean-Shong, Rogelio Saenz, and Aguirre E. Benigno. “The SES Selectivity of
Interethnicly Married Asians.” International Migration Review 29: 469-491. 1995.
Kitano, Harry H. L., Wai-Tsang Yeung, Lynn Chai, and Herbert Tatanaka. 1984. “Asian
American Interethnic Marriage.” Journal of Marriage and the Family. 46: 170-190.
Kalmijn, Matthijs. 1993. “Trends in Black/White Intermarriage.” Social Forces, 72:119-
146.
Kitano, Harry H.L. and L.K. Chai. 1982. “Korean Interracial Marriage.” Marriage and
Family Review 5:75-89.
Lee, Sharon M. and Keiko Yamanaka. “Patterns of Asian American Intermarriage and
Marital Assimilation.” Journal of Comparative Family Studies 1990 Sum; Volume 21
(2): 287-305. Canada: University of Calgary.
Mackerras, Colin.1995. “China’s Minority Cultures: Identities and Integration Since
1912.” New York: St. Martin’s Press.
Mamet, Rizvan. 2002. “Ethnic Intermarriage in Beijing and Xinjiang China, 1990.”
Master of Science, Thesis Brigham Young University.
Qian, Zhenchao, and Samuel H. Preston, 1993. “Changes in American Marrriage, 1972
to 1987: Availability and Forces of Attraction by Age and Education.” American
Sociological Review, 58: 482-495.
Qian, Zhenchao. “Who Intermarries? Education, Nativity, Region, and Interethnic
Marriage, 1980 and 1990.” Journal of Comparative Family Studies 1999 Fal;Volume
30(4): 579-597. Canady: University of Calgary.
Qian, 1997
12
Sung, Betty L. 1990. “Chinese American Intermarriage.” Journal of Comparative Family
Studies 21: 337-352.
Tzeng, Jessie, M. 2000? “Ethnically Heterogamous Marriages: The Case of Asian
Canadians.” Journal of Comparative Family Studies, Vol?;321-337.
Wong, Morrison G. “A Look at Intermarriage among the Chinese in the United States in
1980.” Sociological Perspectives 32: 87-107. 1989.
13
Table 1 - Race/Ethic Composition of Cultural Settings (percent)
USA
Hawaii
Husband Wife
Husband Wife
White
80.2
79.7
White
32.8
28.6
Black
6.2
5.9
Black
2.4
1.5
Asian
3.1
3.6
Asian
51.6
56.9
Native American
0.6
0.6
Native American
8.3
7.8
Hispanic
9.9
10.2
Hispanic
4.9
5.2
(n)
(559,983)
(n)
(8,839)
Canada
New Zeland
Husband Wife
Husband Wife
White
87.1
86.9
European
82.7
82
Asian
8.7
9
Mauri
9.6
9.7
Black
1.5
1.4
Other Pacific Islander
3.7
3.6
Hispanic
0.6
0.7
Ascan
4.1
1.7
Native
1.0
1.1
(n)
(46,344,704)
Arah
1.1
0.9
(n)
(10,192,270)
South Africa
Beijing
Husband Wife
Husband Wife
Black
67.9
67.7
Black
95.5
96.4
Colured
11.3
11.6
Colured
2.3
2.3
Asian
4.2
4.2
Asian
2.2
1.3
Whole
16.6
16.5
(n)
(22,119)
(n)
(305,758)
Xinjiang
Husband Wife
Uighur
46.9
47
Han
44.9
44.8
Hui
4.8
4.7
Kazakh
3.5
3.4
(n)
(27,048)
14
Table 2 - Log Linear Models for Cross-tabulations
of Husbands and Wives
Race/Ethnicity
Model
USA
Hawaii
χ
2
d.f. % of 1
χ
2
d.f. % of 1
1. Independence
571077.7
16
---
7960.9
16
---
2. Universal homogany
12684.2
15
2.22
373.7
15
3.Group Specific homogany
827.3
11
0.14
23
11
0.29
4. Gander Symetry
128.6
6
0.02
19.3
6
0.24
New Zealand
South Africa
χ
2
d.f. % of 1
χ
2
d.f. % of 1
1. Independence
12005398.3
9
---
726554.2 9
---
2. Universal homogany
1658883
8
13.8
3062.6
8
0.42
3.Group Specific homogany
114800
5
0.96
361.8
5
0.05
4. Gander Symetry
29622.2
3
0.25
3.2
3
<.01
Canada
Beijing
χ
2
d.f. % of 1
χ
2
d.f. % of 1
1. Independence
7,255,190.3 25
---
3417.1
---
2. Universal homogany
219618
24
3.02
673
4
19.7
3.Group Specific homogany
12122
19
0.17
0.4
3
0.01
4. Gander Symetry
396.3
0.05
Xinjiang
χ
2
d.f. % of 1
1. Independence
50437.04
9
---
2. Universal homogany
301.5
8
0.59
3.Group Specific homogany
128.09
5
0.25
4. Gander Symetry
5.46
3
0.01
15
Table 3 - Odds Ratios for Homogamous Marriage
Overall Homogamy
Group Specific Homogeny
USA
35.88
White
10.07
Black
692.29
Native American
33.78
Asian
164.02
Hispanic
39.65
Hawaii
9.68
White
3.86
Black
232.76
Asian
14.01
Pacific Island
20.29
Hispanic
8.33
Canada
37.11
White
5.64
Asian
287.15
Black
217.02
Hispanic
473.43
Native American
16.95
Arab
796.32
New Zealand
9.631
European
6.11
Maori
3.53
Other Pacific Islander
53.52
Asian
210.61
South Africa
237.70
African
376.15
Colored
25.28
Asian
685.4
White
2,164.62
Beijing
11.38
Han
1.48
Hui
869.40
Manchu
6.65
Xinjang
287.15
Uighur
2,344.90
Han
330.30
Hui
23.1
16
Kazakh
473.43
17
18
T
a
b
l
e
4
-
E
f
f
e
c
t
s
o
f
A
g
e
o
n
I
n
t
e
r
m
a
r
r
i
a
g
e
U
S
A
W
i
f
e
H
us
ba
nds
:
A
ge
of
W
hi
t
e
B
l
a
c
k
N
a
t
i
ve
A
m
e
r
i
c
a
n
A
s
i
a
n
H
i
s
pa
ni
c
W
hi
t
e
H
-
-
-
.044*
-
.024*
-
.018*
.034
W
-
-
-
-
.041*
.003
-
.001
.008*
B
l
a
c
k
H
-
.032*
-
-
-
-
.025*
-
.013
-
.040*
W
-
.053
-
-
-
-
.020
-
.020
-
.010*
N
a
t
i
ve
A
m
e
r
i
c
a
n
H
.002
.031*
-
-
-
-
-
-
-
.015*
W
-
.029*
.012
-
-
-
-
-
-
.000
A
s
i
a
n
H
-
.018*
-
-
-
-
-
-
-
-
-
-
.028*
W
-
.033
-
-
-
-
-
-
-
-
-
-
.002
H
i
s
pa
ni
c
H
-
.005*
-
.011*
-
.004
-
.012*
-
-
-
W
-
.045*
-
.039
-
.023*
-
.039*
-
-
-
H
aw
ai
i
W
i
f
e
H
us
ba
nd:
A
ge
of
W
hi
t
e
B
l
a
c
k
A
s
i
a
n
H
i
s
pa
ni
c
W
hi
t
e
H
-
-
-
.005
-
.007
-
.026*
W
-
-
-
-
.032*
-
.031*
.014
P
a
c
i
f
i
c
I
s
l
a
nde
r
H
-
.022*
-
-
-
-
.022*
-
.043*
W
-
.004
-
-
-
-
.016
-
.011*
19
A
s
i
a
n
H
.022*
.021*
-
-
-
-
-
-
W
-
.002*
.022*
-
-
-
-
-
-
H
i
s
pa
ni
c
H
-
.005
.013
-
-
-
-
-
-
W
-
.040*
-
.062
-
-
-
-
-
-
C
an
ad
a
W
i
f
e
H
us
ba
nd:
A
ge
of
W
hi
t
e
A
s
i
a
n
B
l
a
c
k
H
i
s
pa
ni
c
N
a
t
i
ve
A
m
e
r
i
c
a
n
A
r
a
b
W
hi
t
e
H
-
-
-
.321*
-
.378*
-
.538*
-
.488*
-
.409*
W
-
-
-
-
.310*
-
.196*
-
.252*
.153*
-
.040*
A
s
i
a
n
H
-
.425*
-
-
-
-
.248*
-
.414*
-
.773*
.137*
W
-
.294*
-
-
-
-
.026
-
.114*
-
.013
-
.251
B
l
a
c
k
H
-
.373
.267*
-
-
-
-
.699*
-
.499*
-
.650*
W
-
.425*
.388
-
-
-
-
.301*
.012
-
.415*
H
i
s
pa
ni
c
H
-
.457*
-
.459*
-
.499*
-
-
-
-
.906*
-
1.015
W
-
.623*
-
.553*
-
.834*
-
-
-
-
.368*
-
.268*
N
a
t
i
ve
A
m
e
r
i
c
a
ns
H
-
.040*
-
.119*
-
.066
-
.967*
-
-
-
-
.045*
W
-
.540*
-
.788*
-
.617*
-
.860
-
-
-
-
.663
A
r
a
b
H
.336*
-
.507*
-
.769*
-
.999
-
1.306*
-
-
-
W
-
.399*
-
.367*
-
.580*
-
.438*
-
.535
-
-
-
N
e
w
Z
e
al
an
d
W
i
f
e
H
us
ba
nd:
E
ur
ope
a
n
M
a
ur
i
O
t
he
r
P
a
c
i
f
i
c
I
s
l
a
nd
A
s
i
a
n
20
E
ur
ope
a
n
-
-
-
-
.348*
-
.348*
-
.408*
-
-
-
-
.122*
.122*
-
.049*
M
a
ur
i
-
.015*
-
-
-
-
-
-
-
.190*
-
.445*
-
-
-
-
-
-
-
.402*
O
t
he
r
P
a
c
i
f
i
c
I
s
l
a
nde
r
-
.129*
-
.428*
-
.428*
-
.473*
-
.520*
-
328*
-
.328*
-
.687*
A
s
i
a
n
-
.074*
-
.411*
-
.411*
-
-
-
-
.379
-
.189*
-
.189*
-
-
-
S
ou
t
h
A
f
r
i
c
a
W
i
f
e
H
us
ba
nds
:
B
l
a
c
k
C
ol
our
e
d
A
s
i
a
n
W
hi
t
e
B
l
a
c
k
-
-
-
-
.035*
-
.043*
.029
-
-
-
-
.050*
-
.051*
-
.027
C
ol
our
e
d
-
.022*
-
-
-
-
.021*
-
.021*
-
.018*
-
-
-
-
.043
-
.062*
A
s
i
a
n
-
.060*
-
.022*
-
-
-
-
.069*
-
.135*
-
.013*
-
-
-
-
.099*
W
hi
t
e
-
.003*
-
.024*
.039*
-
-
-
-
.002*
-
.017*
-
.033
-
-
-
B
e
i
j
i
n
g
W
i
f
e
H
us
ba
nds
:
H
a
n
H
ui
M
a
nc
hu
H
a
n
-
-
-
-
.023*
-
.032*
21
-
-
-
-
.040*
-
.025*
H
ui
-
.037*
-
-
-
-
-
-
-
.015
-
-
-
-
-
-
M
a
nc
hu
-
.008
-
-
-
-
-
-
-
.023
-
-
-
-
-
-
22
T
a
bl
e
5
E
f
f
e
c
t
s
of
E
duc
a
t
i
on on
I
nt
e
r
m
a
r
r
i
a
ge
U
S
A
W
i
f
e
E
duc
a
t
i
on of
W
hi
t
e
B
l
a
c
k
N
a
t
i
ve
A
m
e
r
i
c
a
n
A
s
i
a
n
H
i
s
pa
ni
c
W
hi
t
e
H
.078*
-
.080*
.199*
.048*
W
.152*
.164*
.038*
.277*
B
l
a
c
k
H
.112*
.096
.184*
.068*
W
-
.017
.219*
-
.066*
.216*
N
a
t
i
ve
A
m
e
r
i
c
a
n
H
.163*
.170*
.118*
W
-
.083*
-
.040
.142*
A
s
i
a
n
H
.124*
.002
W
.254*
.229
H
i
s
pa
ni
c
H
.224*
.154*
.044*
.286*
W
.026*
-
.017
-
.040
-
.016
H
a
w
a
i
i
W
i
f
e
W
hi
t
e
P
a
c
i
f
i
c
I
s
l
a
nde
r
A
s
i
a
n
H
i
s
pa
ni
c
W
hi
t
e
-
.177*
-
.222*
.069*
-
.147*
.272*
P
a
c
i
f
i
c
I
s
l
a
nde
r
.159*
-
.028
-
.007
-
.034
.100
.171*
A
s
i
a
n
.178*
.102*
-
.026
.044
23
H
i
s
pa
ni
c
.169*
.126*
-
.134*
-
.024
T
a
bl
e
5 C
ont
i
nue
d
C
an
ad
a
W
i
f
e
W
hi
t
e
A
s
i
a
n
B
l
a
c
k
H
i
s
pa
ni
c
N
a
t
i
ve
A
m
e
r
i
c
a
n
A
s
i
a
n
W
hi
t
e
.333*
.142*
.211*
-
.231*
.426*
.361*
.192*
.418*
.250*
.346*
A
s
i
a
n
.282*
.119*
.184*
-
.081*
.382*
.347*
.180*
.464*
.416*
.116*
B
l
a
c
k
.120*
.175*
.147*
-
.305*
.261*
.164*
.211*
.217*
.336*
-
.066
H
i
s
pa
ni
c
.296*
.234*
-
.092
-
.266*
.534*
.271*
.280*
.067
.516*
.386*
N
a
t
i
ve
A
m
e
r
i
c
a
n
.196*
.387*
.395*
.028
.822*
-
.121*
.053*
.093*
-
.058
.406*
A
r
a
b
.180*
.186*
-
.165*
.023
.138*
.399*
.393*
-
.091*
.516*
.696*
N
e
w
Z
e
al
an
d
24
W
i
f
e
E
ur
ope
a
n
-
.280
-
.127
.152*
.319*
.445*
.239*
M
a
ur
i
.427*
.155*
.588*
-
.256*
.178*
.049*
O
t
he
r
P
a
c
i
f
i
c
I
s
l
a
nde
r
.526*
.062*
.988*
-
.147*
.058*
-
.022
A
s
i
a
n
.146*
-
.365*
-
.341*
.297*
.255*
.288*
T
a
bl
e
5 C
ont
i
nue
d
S
ou
t
h
A
f
r
i
c
a
W
i
f
e
B
l
a
c
k
-
.027*
.123*
.214
-
.143*
-
.040*
-
.112*
C
ol
or
e
d
-
.147*
-
.142*
.276*
-
.050*
-
.053*
-
.078*
A
s
i
a
n
-
.161*
.007
.206*
.062
.162*
-
.040
W
hi
t
e
-
.078*
-
.071*
.023*
0.196
.208*
.132
B
e
i
j
i
n
g
W
i
f
e
25
H
a
n
.348*
-
.154
.137
-
.272
H
ui
.333*
.596*
M
a
nc
hu
.148*
.497*
26
27
Table 6
Regression of Education Coefficients on Gender Average Education of Groups
Constant
Education diff. Gender Gender Interaction R2
1. USA
.117*
-.009
-.040
-.017
.139
2. Hawaii
.029
-.086
.030
.020
.692
3. Canada
.178*
-.370*
.081
.145
.419
4. New
Zealand
.161*
-.791*
-.027
.449*
.821
5. South
Africa
.042
-.029*
-.022
.004
.706
Информация о работе Cross-Cultural Patterns of Interracial Marriage