Contextual Factors Associated With Women's Attitudes Toward Intimate Partner Violence in Tajikistan: Findings From the 2012 and 2017 Demographic and Health Surveys

Abstract

The study employed data from the 2012 and 2017 Tajikistan Demographic and Health Surveys to examine two time periods on media access, interview setting, and sociodemographic predictors of intimate partner violence (IPV) justification attitudes among representative samples of women aged 15–49 years old (9,656 and 10,718 women, respectively) in Tajikistan. The odds of justifying IPV were higher for women who had access to radio and lower for those who had access to newspaper and television. The presence of the husband and other women during the interview was associated with lower justification of IPV. The findings emphasize the importance of contextual factors in developing effective IPV intervention policies.

Keywords

intimate partner violence attitudes women Tajikistan media access interview context

Introduction

Intimate partner violence (IPV) against women, the most prevalent form of domestic violence, is an acute problem with negative physical and mental health implications (Campbell, 2002; Golding, 1999) that persists pervasively in many countries around the world (Garcia-Moreno et al., 2006; Uthman et al., 2009; Waltermaurer, 2012). Tajikistan, a post-Soviet Central Asian country, is no exception in this respect. A limited number of comparative studies on the subject place Tajikistan among the most IPV-tolerant countries in the post-Soviet arena (Ismayilova, 2015; Joshi & Childress, 2017). For example, in 2005, in Tajikistan, the prevalence of IPV justification (74.5%) for at least one of five reasons (if a wife goes out without telling her husband, neglects the children, argues with him, refuses to have sex with him, and burns food) was over six-fold and one and half times higher than in Kazakhstan (12.3%) and Kyrgyzstan (45.3%), respectively (Joshi & Childress, 2017). The numbers are comparable with some poverty-stricken countries of sub-Saharan Africa where rates of supportive attitudes toward IPV are among the highest in the world. A study of DHS data conducted between 2003 and 2007 in 17 sub-Saharan countries found that the percentage of women who justified IPV ranged from the lowest in Madagascar (28%) to the highest in Ethiopia (74%) (Uthman et al., 2009). Despite being a widespread and serious public health problem, pro-violence attitudes toward IPV justification against women in Tajikistan has received little scholarly attention. A small number of studies examined women‘s attitudes toward IPV as potential correlates of physical IPV in Tajikistan (Chernyak, 2018, 2020). Although both studies that focused solely on Tajikistan (Chernyak, 2018) and placed Tajikistan in comparative perspective with four other post-Soviet countries (Chernyak, 2020), listed women's pro-violence attitudes among risk factors for physical IPV, they did not analyze predictors of IPV justification attitudes among women. Haarr’s (2007) descriptive study that surveyed 400 women in three districts of the Khatlon region was one of the earliest attempts to research wife abuse in Tajikistan systematically. Looking at the perceptions of currently and formerly married Tajik women toward wife abuse, including justifiable circumstances, and the prevalence of wife abuse by husbands and mothers-in-law, the study found that most women approved of wife abuse by one‘s husband and/or mother-in-law. Women approved of a husband beating his wife in 12 out of the 13 analyzed scenarios that ranged from the lowest 55% to the highest 86% (Haarr, 2007). Since the focus of the study was on three districts of the southern region of the country, the generalizability of the findings may raise questions. Two other studies by Ismayilova (2015) and Joshi and Childress (2017) included Tajikistan in a comparative analysis of post-Soviet countries. Drawing on the 2012 DHS and the 2005 MICS (Multiple Indicator Cluster Survey) data, respectively, these studies found high prevalence of IPV acceptance in Tajikistan. However, both studies focused only on a subpopulation of currently married women. To avoid a selection bias and get a better understanding of IPV, a large number of studies suggested moving beyond the wife–husband relationship restriction and considering unmarried and widowed/divorced/separated categories as well (Waltermaurer, 2012). Previous research on pro-violence attitudes among women in Tajikistan has also paid little attention to women's access to media and some important social aspects of the research environment (e.g., lack of privacy during the interview) that have been strongly associated with IPV justification (Antai & Antai, 2009; Lawoko, 2006; Trott et al., 2017). Given the recent social and political changes concerning the problem of domestic violence in the country, it is necessary to employ the most recent survey to determine any notable shifts in women's attitudes across years. Additionally, since IPV justification rates among women in Tajikistan have been among the highest in the former Soviet countries and in the world, it is advisable to shed light on some understudied key causes of women's tolerant attitudes toward IPV. For this purpose, grounded in cognitive dissonance theory (Antai & Antai, 2009; Festinger, 1957) and social norms perspective (Paluck & Ball, 2010; Trott et al., 2017), the study examines to what extent women's access to media and the interview context (presence of husband, other men, and other women) influence women's attitudes toward IPV justification.

Country Profile

Following the Soviet legacy, Tajikistan remains one of the smallest and the most impoverished countries of the Soviet bloc, populated mainly by Sunni Muslims (Zainiddinov, 2013). According to the World Bank, about one third of Tajik citizens (29.5%) live in poverty (World Bank Group, 2019). The independence of Tajikistan in 1991 following the breakdown of the Soviet Union, a devastating civil war of 1992–1997, and the transition from a planned to a market economy resulted in the collapse of infrastructure, deterioration of health and social welfare services, rise of unemployment, poverty growth, and mass labor migration that further disempowered women and diminished the Soviet-era gained gender equalities (Joshi & Childress, 2017). Traditional patriarchal Tajik society, characterized by the highest level of religiosity in the post-Soviet region (Zainiddinov, 2018), widely tolerates violence against women, confining it to the private sphere that perpetuates women's inferior and submissive role (Haarr, 2007; Harris, 2004; Sharipova & Fabian, 2010).

A long-term absence of legal framework to protect women from domestic violence served as an indication of acceptance of IPV in Tajik patriarchal society. Although the Tajik government acknowledged violence against women as a serious issue in the late 1990s (Joshi & Childress, 2017), it was one of the last states among the post-Soviet countries to adopt a law on domestic violence. The Law on the Prevention of Violence in the Family came in 2013, as a result of a 10-year joint advocacy effort by local and international NGOs, civil society groups, and the women's movement (Human Rights Watch, 2019). The law recognizes the rights of victims to legal, medical, and psychosocial assistance, yet fails to criminalize domestic violence and lacks the rigorous steps needed to be taken toward its full implementation (Human Rights Watch, 2019).

Interview Context and Media Access as Key Factors of Women's Pro-violence Attitudes

Given the high prevalence of IPV justification among women in Tajikistan, it is important to examine how victims of violence themselves view such violence. Among different root causes of IPV justification that have been identified by previous research, some recent studies argued that a victim's attitudes of tolerance may have a stronger predictive power of IPV compared to socioeconomic factors that were previously thought to be powerful predictors, including education, work status, and poverty (Faramarzi, Esmailzadeh, & Mosavi, 2005). Since women's pro-violence attitudes are shaped by social context, it is necessary to explore key potential avenues intensifying women's acceptance of violence. Previous research has identified interview setting (Trott et al., 2017) and access to information (Antai & Antai, 2009; Lawoko, 2006) as such potential avenues. Examining the influence of interview context on women's attitudes toward IPV in Ethiopia, Trott et al. (2017) found that the odds of justifying IPV were nearly twice as high in the presence of other women, and nearly half in the presence of men. Access to media, in particular to newspapers in Zambia (Lawoko, 2006) and television in the Niger Delta (Antai & Antai, 2009), reduced the likelihood of justifying IPV among women. Previous research on other health problems in Central Asia, such as modern contraception utilization in Tajikistan and Kyrgyzstan, has also emphasized the importance of access to information in raising women‘s awareness and changing their attitudes about the issue (Habibov & Zainiddinov, 2017).

The importance of the interview context could be explained by a social norms perspective, whereas cognitive dissonance theory could serve as a possible theoretical explanation for access to information. The social norms perspective suggests that women's favorable responses correspond to the perceived attitudes of other women and men present during the interview (Paluck & Ball, 2010; Trott et al., 2017). Festinger's theory of cognitive dissonance (1957) postulates that a person experiences dissonance, a negative drive state, whenever the person simultaneously holds two psychologically conflicting cognitions (beliefs, ideas, opinions). Since the occurrence of dissonance is unpleasant, the person tries to reduce it by adding consonant cognition or by altering one or both cognitions to make them more consonant with one another (Festinger, 1957). Resorting to information from media might help women reduce their cognitive dissonance and explain women's justification of IPV.

Drawing on the social norms perspective (Paluck & Ball, 2010) and cognitive dissonance theory (Festinger, 1957), as well as findings from previous studies (Antai & Antai, 2009; Lawoko, 2006; Trott et al., 2017), two sets of hypotheses are constructed. First, it is hypothesized that the interview context will lead to more supportive attitudes among women toward IPV to resemble the perceived attitudes of the husband, other men, and other women present during the interview. Second, media access will either decrease women's pro-violence attitudes through raising their awareness and changing their distorted attitudes toward IPV, or increase women's acceptance of wife beating through reducing cognitive dissonance that serves as a coping mechanism with IPV justification.

Other Factors Associated with Women's Pro-violence Attitudes

A number of demographic and socioeconomic characteristics have been identified as potential determinants of IPV-tolerant attitudes. A review of quantitative studies on IPV justification representing 61 countries (30%) worldwide show that among sociodemographic characteristics, age, education, marital status, household wealth, employment status, and area of residence have been generally shown to be associated with legitimizing attitudes toward IPV justification against women (Waltermaurer, 2012). Although correlates associated with the acceptance of IPV might differ across various countries, these studies demonstrated that in general respondents who tended to agree more with IPV justification were younger, with little or no education, currently married, poor, unemployed, and from rural areas (see Waltermaurer, 2012, for a review). Similarly, a comparative multicountry analysis of 15 sub-Saharan countries identified that the likelihood of IPV justification decreased with increasing wealth status, educational attainment, urbanization, and access to media (Uthman et al., 2009). Previous research on Tajikistan and other former-Soviet countries align with findings from other countries around the world that sociodemographic factors such as younger age, rural residence, lower educational attainment, and household poverty were associated with women's tolerance toward IPV against women (Haarr, 2007; Ismayilova, 2015; Joshi & Childress, 2017). Based on previous research, it is hypothesized that women who are young; married; from a region) (rather than a city); and with disadvantaged socioeconomic status manifested through lower household wealth, lower educational attainment, and unemployment, will be more likely to believe that wife beating is justified.

Method

Data

The data for the study came from the 2012 and 2017 Tajikistan Demographic and Health Surveys (TjDHS), nationally representative sample surveys that provide current and reliable information on women's sociodemographic and household characteristics, maternal and reproductive health measures, education and employment, and decision making, as well as women's knowledge, attitudes, and behaviors about health issues and domestic violence. The surveys were implemented by the Statistical Agency of the Republic of Tajikistan in collaboration with the Ministry of Health of the Republic of Tajikistan with technical support from ICF International, and financial support from the United States Agency for International Development (USAID), the United Nations Children‘s Fund (UNICEF), and the United Nations Population Fund (SA, MOH, & ICF, 2013; SA, MOHSP, & ICF, 2018). The 2012 and 2017 TjDHS employed a stratified two-stage sampling strategy to select the samples of women. Initially, sample clusters (or PSUs [primary sampling units]) with a probability proportional to their size within each sampling stratum were selected. A total of 356 and 366 clusters were selected in the 2012 and 2017 TjDHS, respectively. In the second stage, households were randomly selected within each selected cluster. The standard DHS women's questionnaire was administered to women (15–49 years of age) who were either residents or visitors in the sampled households at the time of the surveys. In the 2012 and 2017 TjDHS, interviews were completed with 9,656 and 10,718 women, respectively. The detailed descriptions of sampling procedures are presented in the DHS 2013 and 2018 country reports (SA, MOH, & ICF, 2013; SA, MOHSP, & ICF, 2018). Sample cases were weighted to obtain nationally representative estimates. The weighted samples, after excluding cases due to missing data on one or more variables, included 9,599 and 10,525 respondents in the 2012 and 2017 surveys, respectively. The final weighted sample of both surveys included in multivariate analyses was comprised of 20,124 cases. The STATA vif and collin commands were used to test for collinearity among the key predictors and sociodemographic characteristics used in the study. The VIF (variance inflation factor) values for all independent variables ranged from the lowest 1.02 to the highest 2.34, which are below the cutoff value of 10.

Dependent Variables

To assess women's attitudes toward IPV, the respondents were asked, “Sometimes a husband is annoyed or angered by things which his wife does. In your opinion, is a husband justified in beating his wife in the following situations?” The five situations were: (a) if she goes out without telling him; (b) if she neglects the children; (c) if she argues with him; (d) if she refuses to have sex with him; and (e) if she burns the food.

Based on the five situations, two sets of dependent variables were created. First, for each of the situations, a separate dichotomous variable was created where responses of “yes” were coded 1 and responses of “no” were coded 0. Second, all five answers were combined to create a single dichotomous variable, a score or a count measure, where respondents who agreed with at least one situation were identified as having supportive attitudes toward IPV and given a value of 1, and respondents who disagreed with any of the situations were coded 0.

Independent Variables

Key predictor variables in the study include women‘s access to media and interview context. Women‘s access to media was assessed with three indicators, namely, frequency of listening to radio, reading newspapers/magazines, and watching television. Responses were dichotomized by combining “less than once a week” and “at least once a week” into one group with a value of 1, and “not at all” in the other group with a value of 0. The interview context was assessed with four sub-questions that followed the wife beating justified questions and captured the presence of husbands, other males, other females, and children under 10 years of age during the interview. Possible options “yes—listening” and “yes—not listening” formed one group of the dichotomy and were coded 1, whereas the option “no” formed the other group of the dichotomy and was coded 0. Since the focus is on the importance of men and women's presence during the interview, the presence of children variable was dropped.

Given that findings of previous studies are inconsistent concerning the significance and direction of the effects of various sociodemographic factors on IPV justification, the decision was made to select a few demographic and socioeconomic status variables that have shown strong association with women's attitudes toward IPV. Demographic variables included age in years (ranging from 15 to 49 years old), marital status [never married, (reference category), married/living with partner, and widowed/divorced/separated], and region of residence [a series of five dummy variables signifying five regions of Tajikistan: Dushanbe (reference category), Districts of Republican Subordination (DRS), Sughd, Khatlon, and Gorno-Badakhshan Autonomous Oblast (GBAO)]. Socioeconomic status measures included education in years (ranging from 0 to 21), household wealth index [originally grouped into five quintiles was recoded as tertiles: poor (reference category), middle, and rich], and current employment (yes/no). To compare respondents who participated in the 2017 survey with those who participated in the 2012 survey, two datasets were appended for the purpose of multivariate analysis and a dichotomous variable “year 2017” was created. It should be noted that respondents in the combined dataset are not tracked over time but rather are unique samples across the two survey administrations.

Analytic Plan

A unified methodology used by the DHS in administering standard DHS women's questionnaires in both years allows for comparison over time. To show changes across two survey years and provide descriptive statistics for the independent and control variables, univariate analyses were used. Next, bivariate analyses were conducted using chi-square and t-tests to assess whether the independent and control variables differ significantly on six outcome measures. The goal is to see whether significant attitudinal differences were observed on key predictors and sociodemographic characteristics among women in both survey years. Finally, binary logistic regressions were run with the combination of the 2012 and 2017 surveys to identify key predictors and sociodemographic determinants of pro-violence attitudes among women aged 15–49 years old in Tajikistan. All statistical analyses were performed in Stata 16 using the survey command (-svy-), to account for DHS's stratified multistage cluster sampling design and weighting.

Results

Figure 1 shows the percentage of women aged 15–49 years old who agreed that a husband is justified in beating his wife for specific reasons in 2012 and 2017.

Figure 1.

Prevalence of acceptance of wife beating by women aged 15–49 years old, DHS 2012 and DHS 2017. Note: DHS 2012 N = 9,599 DHS 2017 N = 10,525. Asterisks indicate statistically significant differences across two years *p ≤ .05, **p ≤ .01, ***p ≤ .001.

Overall, there was an increase in pro-violence attitudes among women in 2017 compared with 2012 for all situation-based questions, except for refusing to have sex with husband. Acceptance of wife beating does not only remain prevalent in Tajikistan but also is growing. The proportion of women who agree that wife beating is justified is higher in 2017 for four out of the five specified reasons: if she goes out without telling him (increased from 50% in 2012 to 51% in 2017); if she neglects the children (from 44% in 2012 to 50% in 2017); if she argues with him (from 40% in 2012 to 52% in 2017); and if she burns the food (from 28% in 2012 to 29% in 2017). A negative trend was also observed in agreement with at least one reason for wife beating. The proportion of women who agreed with one or more justification for wife beating increased by almost 4% between 2012 and 2017 (59.58% vs. 63.51%).

Chi-square tests show that the across-time differences are statistically significant for three outcome measures, namely, when the wife neglects the children (χ² = 58.59, p ≤ .001), argues with her husband (χ² = 300.70, p ≤ .001), and agrees with at least one specified reason (χ² = 32.81, p ≤ .01). As for the remaining outcomes, including when the wife goes out without telling husband, refuses to have sex with her husband, and burns the food, differences are not significant across the years of 2012 and 2017.

Table 1 shows descriptive statistics for the independent variables and sociodemographic controls in the 212 and 2017 DHS surveys.

Table 1.

Descriptive Statistics for the Independent Variables in the Analyses, DHS 2012, 2017.

Variables	DHS 2012		DHS 2017
	M or %	SD	M or %	SD
Media access
Reads print media	49.78		45.45
Listens to radio	41.96		34.7
Watches TV	95.21		95.58
Interview context
Husband presence	1.21		1.84
Other male presence	1.10		1.81
Female presence	7.17		8.57
Demographic characteristics
Age in years (range: 15-49)	28.89	9.83	29.49	9.64
Age square	931.31	615.03	962.71	606.71
Marital status
Never married	27.35		22.24
Currently married	67.40		72.36
Formerly married	5.25		5.40
Region
Dushanbe	9.10		9.02
GBAO	2.26		1.98
Sughd	29.80		31.15
DRS	23.21		21.71
Khatlon	35.63		36.14
Socioeconomic status
Education in years (0–21)	9.71	2.79	10.22	2.59
Wealth index
Poor	39.25		39.00
Middle	19.72		19.55
Rich	41.03		41.45
Work status
Employed	27.37		24.69
N	9,599		10,525

Notes: Means and standard deviations are presented for continuous variables.

Proportions (%) are presented for categorical variables.

Proportions are given for answer “yes.”

The majority of respondents are currently married, watch TV, live in rural areas, and belong to wealthy households. Compared with respondents in 2012, respondents in 2017 read newspapers or magazines less, listened to the radio less, and had lower employment. Women in the 2017 survey are more likely to be married and had slightly higher levels of education than women in the 2012 survey. In both surveys, there was an unequal distribution among the five regions of the country with respondents from the capital city and GBAO Autonomy Region making up the smallest proportions.

Table 2 presents proportions of women aged 15–49 years old reporting each type of pro-violence attitude by key predictors (access to media and interview context), as well as demographic (age, marital status, region of residence) and socioeconomic characteristics (education, wealth index, and employment status).

Table 2.

Proportion of Women Reporting Each Type of Proviolence Attitudes by Key Predictors and Sociodemographic Characteristics, DHS 2012, 2017.

	Go out without telling		Neglect the children		Argue with husband		Refuse sex		Burn the food		Agree with at least one reason
	2012	2017	2012	2017	2012	2017	2012	2017	2012	2017	2012	2017
Media access
Reads print media
Yes	46.75***	45.72***	39.49***	46.13***	36.12***	49.49 ***	22.64***	20.19 ***	24.04***	24.50***	56.48***	61.07**
Not at all	54.19***	55.55***	49.19***	52.77***	43.83***	54.45***	31.88***	31.48***	32.65***	33.32***	62.66***	65.54**
Listens to radio
Yes	52.42*	49.70	44.31	50.33	40.91	54.85**	26.51	22.07 ***	26.12*	25.84***	60.48	65.02
Not at all	49.09*	51.82	44.40	49.45	39.33	50.78**	27.84	28.63***	29.98*	31.16***	58.94	62.71
Watches TV
Yes	50.41	50.87	44.11	49.40**	39.67*	52.02	26.74***	25.70***	28.06*	28.84***	59.52	63.40
Not at all	52.12	55.74	49.30	57.37**	46.33*	55.86	38.14***	40.44***	34.34*	39.58***	60.78	65.88
Interview context
Husband presence
Yes	46.59	33.11**	32.99*	31.52**	37.33	44.43	22.76	19.10	24.21	20.42*	61.33	49.63**
No	50.54	51.41**	44.50*	50.09**	40.02	52.34	27.34	26.49	28.41	29.48*	59.56	63.77**
Other male presence
Yes	45.02	39.61*	34.84	31.78***	34.98	52.02	23.37	17.28 *	28.14	19.73 *	57.21	58.30
No	50.55	51.30*	44.47	50.09***	40.05	52.20	27.33	26.52*	28.38	29.49*	59.61	63.61
Female presence
Yes	37.02***	41.20***	35.15**	37.32***	43.76	47.36	23.35	18.46***	19.36***	23.21**	60.58	54.75 ***
No	51.53***	52.01***	45.07**	50.92***	39.70	52.65	27.59	27.09***	29.06***	29.88**	59.51	64.33***
Demographics
Age
Age in years (15–49)	29.51*** 9.25	30.38*** 8.81	29.68*** 9.34	30.36*** 9.12	29.42*** 9.20	30.45*** 8.83	30.11*** 8.99	31.04*** 8.38	29.48*** 8.97	30.27*** 8.90	29.57*** 9.29	30.25*** 9.11
Age square	960.13*** 587.51	1005.5*** 566.66	971.37*** 593.76	1006.5*** 586.07	955.81** 583.57	1010.4*** 567.82	993.78*** 576.87	1038.7*** 549.47	957.53* 567.80	998.05** 573.13	965.02*** 588.67	1000.6*** 582.62
Marital status
Never married	35.78***	29.58***	31.15***	31.64***	28.19***	30.97***	14.36***	7.98 ***	20.72***	17.46***	43.24***	43.69***
Currently married	57.50***	58.06***	50.38***	55.43***	45.36***	59.11***	32.97***	31.90***	31.92***	33.08***	67.20***	69.91***
Formerly married	37.16***	46.20*	35.99***	48.34	32.52**	46.89*	21.61*	27.72	22.49**	27.61	46.92***	59.37*
Region
Dushanbe	38.74***	29.92***	29.91***	36.82***	24.18***	37.43***	14.11***	15.86***	11.77***	19.16***	46.18***	45.20***
GBAO	34.05***	32.96***	35.38***	48.10	23.34***	32.16***	16.30***	11.96***	21.96*	23.87*	45.03***	57.60*
Sughd	43.62***	41.91***	32.27***	38.12***	37.29	53.85	24.65	20.10***	24.71*	19.13***	57.50	59.69**
DRS	58.32***	53.94	54.99***	49.84	47.09**	49.40	34.59***	31.53***	26.36	32.91*	62.32	64.09
Khatlon	55.17**	63.56***	51.81***	63.05***	42.73	57.23***	28.78	32.04***	37.36***	38.76***	63.89**	71.35***
Socioeconomic characteristics
Education
Education in years (0–21)	9.39*** 2.74	9.93*** 2.38	9.38*** 2.74	9.96*** 2.47	9.37*** 2.74	10.05*** 2.37	9.34*** 2.64	9.89*** 2.38	9.19*** 2.71	9.89*** 2.46	9.49*** 2.71	10.04*** 2.45
Wealth index
Poor	55.83***	55.19***	51.82***	53.02**	45.37***	54.56*	33.08***	26.38	37.26***	31.30*	65.74***	66.48*
Middle	52.52	57.99***	45.85	55.94***	39.74	58.36***	27.64	32.34***	27.92	33.80***	60.93	70.33***
Rich	44.40***	43.97***	36.51***	43.76***	34.96***	47.05***	21.57***	23.50***	20.06***	25.32***	53.04***	57.50***
Work status
Employed	48.97	46.62***	43.60	45.35***	38.29	47.53***	25.85	22.49***	28.80	24.65***	59.55	59.31***
Unemployed	51.06	52.55***	44.65	51.20***	40.63	53.72***	27.82	27.62***	28.20	30.84***	59.60	64.89***
Total	50.49	51.09	44.36	49.75	39.99	52.19	27.28	26.35	28.36	29.31	59.58	63.51

Notes: DHS 2012 N = 9,599; DHS 2017 N = 10,525.

Means and standard deviations are presented for continuous variables; proportions (%) are presented for categorical variables.

T-tests are used to assess significant differences between means for continuous variables; χ² are used for categorical variables.

*p ≤ .05, **p ≤ .01, ***p ≤ .001.

With a slight difference of 4%, about two-thirds of women aged 15–49 years old in Tajikistan in both survey years agreed that a husband is justified in beating his wife for at least one of the specified reasons. The most widely accepted justification for wife beating among respondents in 2012 was going out without telling her husband (50.49%), followed by neglecting the children (44.36%), and arguing with her husband (39.99%). Arguing with husband (52.19%) becomes the most widely accepted reason in 2017, whereas going out without telling a husband (51.09%) remains nearly at the same level as in 2012, followed by an over 5% increase in neglecting the children (49.75%). A smaller proportion of women (over one-quarter) across 2 years agree that refusing to have sex with the husband and burning the food are acceptable justifications for wife beating.

In both 2012 and 2017, significantly higher proportions of women who read newspapers or magazines (56% and 61%, respectively) were more likely to agree with at least one reason for wife beating than those who do not read newspapers or magazines (63% and 66%, respectively). Compared to women who do not listen to radio, significantly higher proportions of women who listen to radio agreed that wife beating is justified if a wife goes out without telling her husband (52% vs. 49% in 2012) or argues with him (55% vs. 51% in 2017). Nearly half of respondents in the 2017 survey who read newspapers or magazines (49.5%) and watch television (49.4%) agreed that wife beating is justified if the wife argues with her husband and neglects the children, although the proportions are significantly higher among women who do not read newspapers or magazines and watch television (54.5% and 57.4%, respectively). In 2017, agreement with at least one reason for wife beating was more likely among women who were interviewed in the absence of the husband (64%) and in the absence of other females (64%), compared to women who were interviewed in the presence of the husband (50%) and the presence of other females (55%). Compared to 2012, in 2017, the presence of males was a strong predictor of women‘s supportive attitudes toward wife beating on all outcome measures, except for arguing with the husband.

Significant attitudinal differences among women were also observed on all demographic and socioeconomic characteristics in both survey years, except for employment status in the 2012 survey. In 2012 and 2017, currently married women were more likely (67% and 70%, respectively) than never married women (43% and 44%, respectively) and formerly married women (47% and 60%, respectively) to agree with at least one justification for wife beating. There were varied significant differences across regions on all outcome measures. Agreement with at least one reason for wife beating in 2012 and 2017 is lowest among women in the capital city of Dushanbe (46% and 45%) and highest among women in Khatlon (64% and 71%). Agreement with at least one reason for wife beating declines with increasing wealth in 2012; however, it does not decline consistently in 2017. Employment status was a strong predictor in the 2017 survey across all outcomes with a higher percentage of employed persons reporting less acceptance of wife beating than those who are unemployed.

Table 3 presents results from six sets of binary logistic regression models predicting women's pro-violence attitudes for the 2012/2017 combined dataset.

Table 3.

Binary Logistic Regression Predicting Proviolence Attitudes Among Women Aged 15–49 Years Old in Tajikistan, DHS 2012, 2017.

	Go out without telling	Neglect the children	Argue with husband	Refuse sex	Burn the food	Agree with at least one reason
Media access
Reads print media (1 = yes)	0.92 (0.84–1.01)	0.92 (0.84–1.01)	0.92* (0.84–1.00)	0.76*** (0.68–0.85)	0.89* (0.81–0.98)	1.03 (0.94–1.13)
Listens to radio (1 = yes)	1.36*** (1.22–1.52)	1.38*** (1.23–1.53)	1.41*** (1.28–1.56)	1.09 (0.98–1.22)	1.05 (0.94–1.17)	1.36*** (1.23–1.51)
Watches TV (1 = yes)	1.08 (0.89–1.33)	0.92 (0.75–1.13)	0.92 (0.77–1.11)	0.68*** (0.55–0.82)	0.87 (0.73–1.03)	1.10 (0.90–1.36)
Interview context
Presence of husband (1 = yes)	0.65* (0.45–0.95)	0.64* (0.43–0.95)	0.64* (0.44–0.91)	0.79 (0.55–1.14)	0.81 (0.56–1.19)	0.62* (0.42–0.92)
Presence of males (1 = yes)	1.18 (0.80–1.74)	0.89 (0.62–1.26)	1.12 (0.79–1.57)	0.91 (0.62–1.33)	1.08 (0.75–1.55)	1.15 (0.78–1.72)
Presence of females (1 = yes)	0.68*** (0.56–0.83)	0.77* (0.63–0.94)	1.02 (0.85–1.22)	0.74** (0.60–0.91)	0.77** (0.64–0.94)	0.91 (0.75–1.10)
Demographics
Age (range:15–49)	1.07*** (1.03–1.10)	1.06*** (1.03–1.09)	1.05** (1.02–1.08)	1.05** (1.01–1.09)	1.06**** (1.02–1.09)	1.08*** (1.05–1.11)
Age square	0.998*** (.998–.999)	0.999*** (.998–.999)	0.999*** (.998–.999)	0.999** (.998–.999)	0.999*** (.998–.999)	0.998*** (.998–.999)
Currently married^a	2.92*** (2.56–3.32)	2.49*** (2.19–2.84)	2.62*** (2.32–2.97)	3.50*** (2.94–4.17)	1.96*** (1.70–2.28)	2.81*** (2.48–3.18)
Formerly married^a	1.57*** (1.30–1.89)	1.61*** (1.35–1.92)	1.70*** (1.42–2.02)	2.55*** (2.00–3.25)	1.42*** (1.15–1.77)	1.54*** (1.29–1.84)
GBAO^b	0.98 (0.76–1.27)	1.42** (1.13–1.79)	0.84 (0.65–1.08)	0.93 (0.70–1.23)	1.43** (1.11–1.93)	1.19 (0.96–1.49)
Sughd^b	1.21* (1.02–1.44)	0.91 (0.77–1.08)	1.62*** (1.36–1.92)	1.41*** (1.15–1.74)	1.28* (1.02–1.62)	1.37*** (1.17–1.61)
DRS^b	1.95*** (1.66–2.28)	1.75*** (1.49–2.06)	1.64*** (1.34–2.02)	2.28*** (1.81–2.87)	1.76*** (1.39–2.22)	1.55*** (1.32–1.81)
Khatlon^b	2.31*** (1.96–2.72)	2.21*** (1.89–2.63)	1.88*** (1.57–2.26)	2.02*** (1.65–2.49)	2.56*** (2.03–3.23)	2.00*** (1.70–2.36)
Socioeconomic status
Education (range: 0–21)	0.94*** (0.92–0.95)	0.95*** (0.94–0.97)	0.95*** (0.94–0.96)	0.97** (0.96–0.99)	0.96*** (0.94–0.97)	0.93*** (0.92–0.95)
Middle^c	1.00 (0.87–1.15)	0.99 (0.86–1.13)	0.93 (0.81–1.07)	1.04 (0.91–1.19)	0.94 (0.82–1.08)	0.97 (0.84–1.11)
Rich^c	0.79*** (0.70–0.89)	0.78** (0.68–0.88)	0.78*** (0.69–0.89)	0.87* (0.77–0.99)	0.78*** (0.68–0.89)	0.73*** (0.64–0.83)
Employed (1 = yes)	0.89* (0.80–0.99)	0.88* (0.79–0.98)	0.84*** (0.77–0.93)	0.86** (0.78–0.96)	0.86** (0.78–0.96)	0.89* (0.80–0.99)
Year 2017	1.03 (0.90–1.17)	1.28*** (1.12–1.46)	1.68*** (1.49–1.90)	0.90 (0.79–1.04)	1.03 (0.90–1.19)	1.19** (1.06–1.35)
Constant	0.24*** (0.14–0.39)	0.24*** (0.14–0.39)	0.20*** (0.12–0.34)	0.09*** (0.05–0.17)	0.13*** (0.07–0.23)	0.30*** (0.18–0.50)

Notes: N = 20,124.

Effect estimates are presented as odds ratios. Confidence intervals are given in parentheses.

Never married.

Dushanbe.

Poor.

*p ≤ .05, **p ≤ .01, ***p ≤ .001.

Among the key independent variables, reading newspapers or magazines, listening to radio, presence of husband during the interview, and presence of other females during the interview are strong predictors of women's attitudes toward wife beating. Significant associations between these variables and women's pro-violence attitudes are found on nearly all outcome measures. Women who read newspapers or magazines and were interviewed in the presence of the husband and other females were less likely to accept wife beating than women who do not read newspapers or magazines and were interviewed in the absence of the husband and other females. In contrast, listening to radio increases women's pro-violence attitudes. On the first (going out without telling the husband), second (neglecting the children), third (arguing with husband), and sixth (agreeing with at least one specified reason) outcomes, the odds of accepting wife beating are 1.36, 1.38. 1.41, and 1.36 times (p ≤ .001) as high for women who listen to radio as they are for those who do not listen to radio. Watching television is negatively associated with intermate partner violence on one scenario. Women who watch television are 32% (OR = 0.68, p ≤ .001) less likely to accept wife beating for refusing to have sex with the husband than women who do not watch television.

Except for a middle category of wealth index, all sociodemographic variables are significant determinants of women's pro-violence attitudes. Interestingly, demographic characteristics, such as age, marital status, and region of residence were positively associated, whereas socioeconomic status characteristics, such as education, rich category of the wealth index, and employment status were negatively associated with women's pro-violence attitudes. With each additional year of increase in age, women's acceptance of wife beating increased by 5%–8%. Yet, this effect continues until a turning point, as the negative effect of age squared shows that as women get older the effect of age on pro-violence attitudes declines. The STATA margins command was used to determine the turning point. The turning point for most outcomes was 25 years of age. Categorization of age into seven subgroups (not presented here) showed similar results. Proportions of respondents who agreed with at least one reason for wife beating, as well as justified wife beating if she argued with the husband and neglected the children were lowest among women 15–19 and 45–49 years of age, compared to the remaining age groups (available from the author upon request). Currently and formerly married women, as well as women residing in different regions of the country were more likely to report pro-violence attitudes than unmarried respondents and those who were from the capital city of Dushanbe. Across all outcome measures, with each additional year of increase in education, women's pro-violence attitudes decreased by a minimum 3% and a maximum 7%. Similarly, women from rich households and who were currently employed were less likely to accept wife beating for nearly all specified reasons than those from poor households and who were unemployed. The variable year 2017 confirms the results from Figure 1 showing significant differences between 2 years on three outcome measures (neglecting the children, arguing with husband, and agreeing with at least one reason for wife beating). Women in the 2017 survey are 28%, 68%, and 19% more likely to show pro-violence attitudes on these three outcomes, respectively, compared to women in the 2012 survey.

Discussion

The study examined changes in attitudes and potential contextual factors (media access and interview context) associated with IPV justification among women aged 15–49 years old in Tajikistan between 2012 and 2017. In the five-year period between the two DHS surveys, the justification of IPV among women not only remained common but also increased. In 2017, two-thirds of women in Tajikistan (63%) believed that wife beating was justified in at least one or more scenarios, showing a 4% increase (59%) since 2012. The two most common reasons for justifying wife beating, also highly significant across 2 years, included scenarios in which a wife neglected the children (44% in 2012 vs. 50% in 2017) and argued with her husband (40% in 2012 vs. 52% in 2017). The justifications of wife beating for burning the food and refusing to have sex were less common across the 2 years, yet the differences were insignificant. Although attitudes toward wife beating among women in Tajikistan have barely changed since 2012 for some of the questions, they raise alarm given the fact that the trend has not only stagnated but reversed. A comparison of the 2012 TjDHS survey with the 2005 Multiple Indicator Cluster Survey (2005 MICS) indicates that over seven previous years, there was a substantial reduction (ranging from 11% up to 28%) in pro-violence attitudes for each of the five specified reasons (SA, 2013; SCS, 2007). It is counterintuitive that despite the adoption of the Law on the Prevention of Violence in the Family and socioeconomic development of the country [between 2000 and 2017, the poverty rate reduced from 80% to 29% and average incomes increased from $162 to $800 (World Bank Group, 2019)], women's justification of IPV is increasing and remains one of the highest in the world. One may link it to the strong influence of pervasive patriarchal norms and cultural and religious restrictions imposed on women that might guide their attitudes toward IPV. The patriarchal structure of society enforced by religious, economic, and political institutions, as suggested by feminist theory (see Chernyak, 2018, for a review), perpetuates the dominant position of men and submissive status of women leading to the rise of women's justification of IPV. The lack of women's awareness and perpetrators’ impunity are among other reasons that lead to high acceptance of pro-violence attitudes among women. Out of 1,296 complaints of domestic abuse made to police between 2013 and 2017, only 65 criminal prosecutions were initiated under various articles of the Criminal Code (Human Rights Watch, 2019). The law does not criminalize domestic violence and its implementation in regions of the country remains unsatisfactory (Human Rights Watch, 2019).

Insignificant and low rates of justification of wife beating for burning the food and refusing to have sex across 2 years (Figure 1) merit attention. One possible explanation could be linked to the differences in interpretation and understanding of concepts of violence that exist in Tajik and Western societies. Sexual abuse by husbands is rarely viewed as violence by women in Tajik society; rather, it is considered a normal act in a spousal relationship (Sharipova & Fabian, 2010).

Another objective of the study was to investigate whether and to what extent media access and lack of privacy during the interview (presence of other women, husband, and men in the interview setting) affected women's pro-violence attitudes. Reading newspapers or magazines and watching television had a negative relationship with IPV tolerance, whereas listening to radio had a positive relationship with IPV tolerance. The findings partially support the proposed hypothesis on cognitive dissonance theory and corroborate previous research suggesting that women having access to newspapers (Lawoko, 2006; Marshall & Furr, 2010; Uthman et al., 2009) and television (Antai & Antai, 2009) were less likely to justify IPV, whereas women having access to radio (Antai & Antai, 2009) were more likely to justify IPV. Why does listening to radio increase the likelihood of acceptance of IPV among women? Perhaps radio broadcastss outnumber television broadcasts and newspapers or magazines in covering IPV incidents. Listening to IPV stories that are told with high frequency on radio may lead women to associate with other victims and accept wife beating as a norm. Association with other IPV victims serves as a consonant cognition that helps women reduce dissonance, leading toward more justification of wife beating. On the contrary, television and print media might be more selective in choosing their IPV coverage. There are no daily printed newspapers in Tajikistan. Due to financial constraints that print media faces, newspapers are printed once a week. Except for very limited occasional programs devoted to violence in the family, the coverage of IPV on television comes as a part of crime news reporting. Stories told in newspapers and on television are usually accompanied by horrific images of victims and arrests of IPV perpetrators. Obviously, women's exposure to such stories and images does not trigger IPV justifying responses, but rather leads to a decrease in acceptance of violence among women.

The effect of interview setting on women's attitudes toward IPV was strong. Women who were interviewed in the presence of the husband and other women reported significantly lower odds of agreeing with wife beating. The finding is inconsistent with previous research (Trott et al., 2017) and does not support the social norms perspective. In contrast to the proposed hypothesis, the interview context did not lead to more supportive attitudes among women toward IPV. It rather reduced women's pro-violence attitudes, contradicting the perceived attitudes of husbands, other men, and other women present during the interview. One explanation is that the survey respondents possibly responded in socially desirable ways. Previous studies conducted in Tajikistan show that women view violence as a family issue and talking about it is considered taboo (Sharipova & Fabian, 2010), accompanied with a fear of being labelled a “bad wife” (Haarr, 2007). Another explanation could be due to the pressure that exists with the presence of the husband and other women during the interview. Women are impeded from sharing sensitive information or are not honest in their responses. Qualitative research on domestic violence in Tajikistan has demonstrated that women were uncomfortable responding to violence-related questions even in neutral locations, such as health centers (Sharipova & Fabian, 2010).

Another interesting observation related to the effect of interview context is that the presence of women was a significant predictor on all outcome measures, except for arguing with the husband, whereas the presence of men was significantly associated with going out without telling the husband, neglecting the children, and arguing with the husband. It is possible that women intend to send a message to husband observers that these behaviors should not be considered a breach of socially accepted gender roles. Previous research suggests that IPV against women in Tajikistan is widely accepted as a justifiable punishment when a woman transgresses society's social norms (Haarr, 2007).

With regard to socioeconomic measures, full support was found for the proposed hypothesis on the positive association of disadvantaged socioeconomic status characteristics with justification of IPV. The findings corroborate previous research demonstrating that low education, household poverty, and unemployment are associated with higher justification of wife beating among women (Trott et al., 2017; Uthman et al., 2009; Waltermaurer, 2012), including in Tajikistan (Haarr, 2007; Ismayilova, 2015; Joshi & Childress, 2017). This consistent finding on socially and economically disadvantaged women's tendency to report more pro-violence attitudes can be linked to women's limited socio-economic resources.

Not all findings on demographic factors align with the proposed hypothesis. A curvilinear relationship was found for age. Initial linear relationship of age with women's justification of IPV continued until a turning point of 25 years and then declined as women get older. Explanation for younger women's low tolerance toward IPV against women can be sought in their high levels of awareness of women's rights and problems. A declining acceptance of IPV after 25 years of age could be linked to increasing social status of women in both family and community, which contradicts previous research that found no significant differences across various age groups irrespective of women's growing status (Haarr, 2007). Women who have never been married report less tolerance toward IPV than their married and formerly married counterparts. This finding corroborates the hypothesis and previous studies (Waltermaurer, 2012). It is counterintuitive that married and formerly married women who are or were more likely to experience IPV endorse pro-violence attitudes. Perhaps these women did not provide truthful responses of support for IPV out of the fear of stigma. On the contrary, unmarried women's autonomy might foster their progressive thinking about women's rights and problems. Finally, in agreement with findings from another dataset on Tajikistan (Joshi & Childress, 2017), living in the capital city is associated with lower justification of IPV. Regional residency might expose women to more conservative and nonegalitarian gender relationships. Patriarchal traditions and gender norms perpetuating the submissive status of women and acceptance of violence are better preserved in regions (Haarr, 2007; Sharipova & Fabian, 2010). Additionally, access to various anti-violence programs and support centers run by both local and international NGOs is limited in regions.

Limitations and Future Directions

The present study is subject to a few limitations worth noting. First, given that the dataset that was not solely designed to examine IPV against women, the study was not able to include all potential factors associated with women's attitudes toward IPV. Research suggests that IPV is a learned behavior that transmits from generation to generation (see Stith et al., 2000, for a review). The intergenerational transmission of violence was considered a cross-national phenomenon also observed in former-Soviet countries, including Tajikistan (Hayes & Randa, 2021). Questions regarding women's exposure to IPV and experience of abuse during their childhood, which may not only lead women to see violence as an accepted social norm but also transform them into either future docile victims or aggressive perpetrators, were administered only to one subpopulation of women. Excluding never married and widowed/divorced/separated women results in a selection bias and impedes our better understanding of women's attitudes toward IPV. Second, related to this, the validity of attitudinal questions on IPV justification has raised concern due to their heavy focus on women's normative roles in daily life and failure to capture nondomestic features, such as women's financial status and husband's drunkenness (Lawoko, 2006; Uthman et al., 2009). Future studies should consider broader factors of attitudes associated with IPV. Third, the presence of related and unrelated women and men in the interview context could affect interviewees’ honest disclosure of private information, increase their socially desirable responding, and lead to their underreporting of IPV. DHS trained interviewers should account for differential effects of the presence of related and unrelated observers in the interview setting that could either mitigate or intensify socially acceptable responding. Finally, DHS surveys do not provide information about the content and quality of IPV stories presented in the media. This might partially explain mixed and conflicting results of studies conducted in different countries on the impact of media outlets (Uthman et al., 2009). Future qualitative studies might take survey questions on the media access to the next level to explore the content and quality of disseminated IPV stories via media and shed light on the differences that exist among various media outlets in increasing or decreasing the likelihood of IPV acceptance among women.

Conclusion

IPV remains a serious and prevalent problem in Tajikistan. Given that women's IPV-supportive attitudes evolve across time and serve as a marker for social acceptability, tracing changes over time and understanding previously understudied contextual factors (media access and interview setting) are essential for developing effective interventions to alleviate the situation. This study draws attention to the alarming fact that following the previous sharp decline of IPV acceptance, there was an unacceptable increase in women's pro-violence attitudes between 2012 and 2017. Since efforts, including the adoption of the Law on the Prevention of Violence in the Family, have shown small results in reducing women's pro-violence attitudes, there is a need for additional studies to obtain a complete picture of the extent to which women continue demonstrating high levels of IPV acceptance. The study further expands our understanding of the role of such important, yet under investigated contextual factors associated with pro-violence attitudes as media access (Antai & Antai, 2009; Lawoko, 2006) and interview setting (Trott et al., 2017). The findings of the study emphasize the importance of the interview environment and media access that need to be considered for the development of IPV prevention policies in the former-Soviet Central Asian countries, given the patriarchal nature of these countries, and in other countries around the world with a similar socioeconomic background. They highlight the importance of bringing the issue out from the private realm that considers violence an integral part of the family, to the public arena that understands it as a complex problem whose solution requires close cooperation among program developers, policy makers, stake holders, and professionals. Since high IPV acceptance levels among women are associated with older age, being currently or formerly married, region residency, low educational attainment, household poverty, and unemployment, many of the same demographic and socioeconomic status correlates identified by pervious research, intervention programs and anti-violence information should be tailored to these disadvantaged groups.

Footnotes

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Hakim Zainiddinov

Author Biography

Hakim Zainiddinov (diploma, Tajik State University; MSSW, Columbia University; MA, Rutgers University; PhD, Rutgers University) is a visiting assistant professor in the Department of Sociology at Bowdoin College, Brunswick, Maine, USA. His research interests include health policies and disparities, race and ethnicity, Islam in the United States, state–religion relationship, terrorism, and globalization. His publications have appeared in Ethnic and Racial Studies, Journal of Historical Sociology, Journal of Church and State, Behavioral Sciences of Terrorism and Political Aggression, Sexual and Reproductive Healthcare, European Journal of Public Health, Central Asian Journal of Global Health, International Journal of Health Planning and Management, and Early Child Development and Care.

References

Antai

(2009). Collective violence and attitudes of women toward intimate partner violence: Evidence from the Niger Delta. BMC International Health and Human Rights, 9(12), 1–10. https://doi.org/10.1186/1472-698X-9-12

Campbell

J. C

. (2002). Health consequences of intimate partner violence. Lancet, 359(9314), 1331–1336. https://doi.org/10.1016/S0140-6736(02)08336-8

Chernyak

(2018). Intimate partner violence in Tajikistan: Risk and protective factors. Violence and Victims, 33(6), 1124–1147. https://doi.org/10.1891/0886-6708.33.6.1124

Chernyak

(2020). A comparative study of intimate partner violence in post-Soviet countries: Evidence from national surveys. Journal of Family Issues, 41(12), 2297–2327. https://doi.org/10.1177/0192513X20942818

Faramarzi

Esmailzadeh

Mosavi

. (2005). A comparison of abused and non-abused women’s definitions of domestic violence and attitudes to acceptance of male dominance. European Journal of Obstetrics Gynecology and Reproductive Biology, 122(2), 225–231. https://doi.org/10.1016/j.ejogrb.2004.11.047

Festinger

(1957). A theory of cognitive dissonance. Stanford University.

Garcia-Moreno

Jansen

H. A.

Ellsberg

Heise

Watts

C. H.

(2006). Prevalence of intimate partner violence: Findings from the WHO multi-country study on women’s health and domestic violence. Lancet, 368(9543), 1260–1269. https://doi.org/10.1016/S0140-6736(06)69523-8

Golding

J. M.

(1999). Intimate partner violence as a risk factor for mental disorders: A meta-analysis. Journal of Family Violence, 14(2), 99–132. https://doi.org/10.1023/A:1022079418229

Haarr

R. N.

(2007). Wife abuse in Tajikistan. Feminist Criminology, 2(3), 245–270. https://doi.org/10.1177/1557085107302869

10.

Habibov

Zainiddinov

(2017). Effect of TV and radio family planning messages on the probability of modern contraception utilization in post-Soviet Central Asia. International Journal of Health Planning and Management, 32(1), E17–E38. https://doi.org/10.1002/hpm.2318

11.

Harris

(2004). Control and subversion: Gender relations in Tajikistan. Pluto Press.

12.

Hayes

B. E.

Randa

(2021). Parts unknown: Risk factors of intimate partner violence in Azerbaijan, Kyrgyzstan, Tajikistan, and Moldova. Journal of Interpersonal Violence, 36(5-6), NP3346–NP3368. https://doi.org/10.1177/0886260518772105

13.

Human Rights Watch (2019). Violence with every step: Weak state response to domestic violence in Tajikistan. https://www.hrw.org/report/2019/09/19/violence-every-step/weak-state-response-domestic-violence-tajikistan

14.

Ismayilova

(2015). Spousal violence in 5 transitional countries: A population-based multilevel analysis of individual and contextual factors. American Journal of Public Health, 105(11), e12–e22. https://doi.org/10.2105/AJPH.2015.302779

15.

Joshi

Childress

(2017). A national survey of attitudes toward intimate partner violence among married women in Kazakhstan, Kyrgyzstan, and Tajikistan: Implications for health prevention and intervention. Social Work in Health Care, 56(4), 294–319. https://doi.org/10.1080/00981389.2016.1268660

16.

Lawoko

(2006). Factors associated with attitudes toward intimate partner violence: A study of women in Zambia. Violence and Victims, 21(5), 645–656. https://doi.org/10.1891/0886-6708.21.5.645

17.

Marshall

G. A.

Furr

L. A.

(2010). Factors that affect women’s attitudes toward domestic violence in Turkey. Violence and Victims, 25(2), 265–277. https://doi.org/10.1891/0886-6708.25.2.265

18.

Paluck

E. L.

Ball

(2010). Social norms marketing aimed at gender based violence: A literature review and critical assessment. International Rescue Committee.

19.

Sharipova

Fabian

(2010). From soviet liberation to post-soviet segregation: women and violence in Tajikistan. In K. Fabian (Ed.), Domestic violence in postcommunist states: Local activism, national policies, and global forces (pp. 133–170). Indiana University Press.

20.

State Committee on Statistics (SCS) of the Republic of Tajikistan (2007). Tajikistan Multiple Indicator Cluster Survey 2005. Final Report. Dushanbe, Tajikistan: SCS.

21.

Statistical Agency under the President of the Republic of Tajikistan, Ministry of Health and Social Protection of Population of the Republic of Tajikistan, and ICF (2018). Tajikistan Demographic and Health Survey 2017. Dushanbe, Republic of Tajikistan, and Rockville. Statistical Agency under the President of the Republic of Tajikistan (SA), Ministry of Health and Social Protection of Population of the Republic of Tajikistan (MOHSP), and ICF.

22.

Statistical Agency under the President of the Republic of Tajikistan (SA), Ministry of Health [Tajikistan], and ICF International (2013). Tajikistan Demographic and Health Survey 2012. SA, MOH, and ICF International.

23.

Stith

S. M.

Rosen

K. H.

Middleton

K. A.

Busch

A. L.

Lundeberg

Carlton

R. P.

(2000). The intergenerational transmission of spouse abuse: A meta-analysis. Journal of Marriage and Family, 62(3), 640–654. https://doi.org/10.1111/j.1741-3737.2000.00640.x

24.

Trott

C. D.

Harman

J. J.

Kaufman

(2017). Women’s attitudes toward intimate partner violence in Ethiopia: The role of social norms in the interview context. Violence Against Women, 23(8), 1016–1036. https://doi.org/10.1177/1077801216654018

25.

Uthman

O. A.

Lawoko

Moradi

(2009). Factors associated with attitudes towards intimate partner violence against women: A comparative analysis of 17 sub-Saharan countries. BMC International Health and Human Rights, 9(14), 1–15. https://doi.org/10.1186/1472-698X-9-14

26.

Waltermaurer

(2012). Public justification of intimate partner violence: A review of the literature. Trauma, Violence & Abuse, 13(3), 167–175. https://doi.org/10.1177/1524838012447699

27.

World Bank Group (2019). Tajikistan - Country economic memorandum: Nurturing Tajikistan's growth potential (English). http://documents.worldbank.org/curated/en/283081560956198220/Tajikistan-Country-Economic-Memorandum-Nurturing-Tajikistan-s-Growth-Potential

28.

Zainiddinov

(2013). The changing relationship of the secularized state to religion in Tajikistan. Journal of Church and State, 55(3), 456–477. https://doi.org/10.1093/jcs/css051

29.

Zainiddinov

(2018). Institutional isomorphism in religious entities of post-Soviet Tajikistan. Journal of Historical Sociology, 31(3), 346–362. https://doi.org/10.1111/johs.12181