Discrimination-related health inequality and role of social capital among marriage migrant women in South Korea

Table 4 Concentration indexes for inequality in self-rated poor health among marriage migrant women in South Korea

	n	Unadjusted model		Fully adjusted model
	n	CI (SE)	p	CI (SE)	p
Discrimination-related inequality
Total sample	40430	0.023 (0.001)	<.001	0.018 (0.002)	<.001
Subsample
No social capital (A)	12553	0.022 (0.003)	<.001	0.018 (0.005)	<.001
Bonding social capital (B)	14518	0.023 (0.002)	<.001	0.017 (0.004)	<.001
Bridging social capital (C)	12359	0.024 (0.002)	<.001	0.018 (0.004)	<.001
Non-linking social capital (A + B + C)	39430	0.023 (0.001)	<.001	0.018 (0.002)	<.001
Linking social capital (D)	1000	0.008 (0.008)	.332	0.007 (0.012)	.422
Income-related inequality
Total sample	33108	−0.026 (0.001)	<.001	−0.027 (0.002)	<.001
Subsample
No social capital (A)	10264	−0.030 (0.003)	<.001	−0.027 (0.005)	<.001
Bonding social capital (B)	12115	−0.023 (0.002)	<.001	−0.026 (0.004)	<.001
Bridging social capital (C)	9901	−0.024 (0.002)	<.001	−0.026 (0.004)	<.001
Non-linking social capital (A + B + C)	32280	−0.026 (0.002)	<.001	−0.027 (0.002)	<.001
Linking social capital (D)	828	−0.025 (0.008)	.003	−0.025 (0.014)	.014

Abbreviations: n sample size, CI concentration index, SE standard error
Source: 2009 National Survey of Multicultural Families, Republic of Korea
Note: CIs, SEs, and p values were estimated using the convenient regression: \( 2{\sigma}_r^2\left(\frac{h_i}{\mu}\right)=\upalpha +\upbeta {r}_i+{\sum}_j{\delta}_j{x}_{ji}+{\varepsilon}_i, \)
where σ ²_r and h _i/μ are the variance of the fractional rank (r _i) and share of the poor self-rated health, respectively. r _i is selected as the rank of discrimination experiences to estimate the discrimination-related CI; in contrast, r _i is selected as the rank of household income to estimate the income-related CI. x _j represents the confounding variables listed below to estimate indirectly standardized CI. For the discrimination-related inequality model, variables include age, length of residence, country of origin, household income, marital status, educational level, economic activity, language proficiency, change of subjective social position, and protest against discrimination. For the income-related inequality model, the variable of experiences of discrimination (dichotomous variable) was additionally inserted and the variable of household income was excluded from the model. The nonresponse weight provided from the data source was applied

ISSN: 1475-9276