Parts Unknown: Risk Factors of Intimate Partner Violence in Azerbaijan,Kyrgyzstan,Tajikistan,and Moldova

Abstract

Extant research has argued that there are variations in predictors of intimate partner violence (IPV) across nations and it is necessary to examine country-specific correlates of IPV. Much remains unknown about factors that affect risk of IPV in transitional countries. Specifically, we explore risk factors for IPV among four former nations of the Soviet Union—Azerbaijan, Kyrgyzstan, Tajikistan, and Moldova—with data from the nationally representative Demographic and Health Surveys. Rates of physical abuse were similar within Tajikistan (18.96%) and Moldova (18.38%), while Kyrgyzstan (20.62%) had the highest prevalence and Azerbaijan (9.29%) had the lowest prevalence. In contrast, the prevalence of sexual abuse was very similar across the four countries and was much lower (between 1.00% and 3.50%) than cross-national estimates of physical abuse. Findings suggest that many of the same risk factors of IPV in developing and developed nations influence risk of IPV in transitional nations. Across the four former Soviet nations included in analyses, the experience of controlling behaviors, husband’s alcohol use, and witnessing IPV or experiencing abuse during childhood were all significantly associated with risk of physical and sexual abuse later in life. Indeed, the intergenerational transmission of violence may be a cross-national phenomenon. In addition, equality in decision making did not have an effect on risk of physical and sexual violence, except for risk of physical abuse in Moldova. This may be due to the historical context of these nations. Taken together, findings suggest that country context may influence individual risk factors of IPV.

Keywords

intimate partner violence transitional nations intergenerational transmission of violence

Individuals in both developing and developed nations retain permissive views of violence against women. Particularly, Russia decriminalized domestic violence in early 2017 and some countries, including Niger and Pakistan, do not have a law that specifically criminalizes domestic violence. Yet this cross-national variation in attitudes toward violence against women as well as prevalence rates of violence against women requires further exploration.

Intimate partner violence (IPV) can result in a range of physical, emotional, sexual, and health problems (Garcia-Moreno, Jansen, Ellsberg, Heise, & Watts, 2006; World Health Organization [WHO], 2014). Contemporary research on the incidence and prevalence of IPV in countries around the world reports lifetime national rates of IPV that range from about one in four women to one in two women (Alio, Daley, Nana, Duan, & Salihu, 2009; Black et al., 2011; Gómez & Speizer, 2010). Extant research has argued that there are variations in predictors of IPV across nations and it is necessary to examine country-specific correlates of IPV (Barrett, Habibov, & Chernyak, 2012). Much remains unknown about factors that affect risk of IPV in transitional countries (emphasis added; Yount, Halim, Hynes, & Hillman, 2011). Although many risk factors and theoretical propositions for the persistence of this form of violence remain a point of continued research and development, the present study provides a somewhat unique opportunity to explore, compare, and contrast individual-level correlates of IPV within four transitional countries.

Specifically, we explore risk factors for IPV among four former Soviet nations. All four of the nations included in analyses are transitional countries where little is known about the risk factors of IPV. Azerbaijan, Moldova, Kyrgyzstan, and Tajikistan are geographically dispersed and are arguably very different despite having a common history in their nations’ development. There is variation between these countries despite the fact they all once belonged to the Soviet Union. International comparative and exploratory research on IPV adds value by searching for commonality in risk factors to gain a better understanding of the common correlates of physical and sexual abuse. Exploratory work of this nature facilitates the understanding of the relative importance of cultural influences, versus individual experiences, on behavior and risk of victimization.

Transitional Countries of the Former Soviet Union

The former countries of the Soviet Union include Armenia, Azerbaijan, Belarus, Estonia, Georgia, Kazakhstan, Kyrgyzstan, Latvia, Lithuania, Moldova, Russia, Tajikistan, Turkmenistan, Ukraine, and Uzbekistan. Analyses focus on four of these former Soviet nations—Kyrgyzstan, Tajikistan, Azerbaijan, and Moldova. During the 1970s and while attending an international conference on women, a Soviet male delegate “declared that there was no discrimination against women in the Soviet Union” in addition to claiming “there was no such thing as sexism in his country because there was no such word in the Russian language” (Freedman, 2002, p. 108). Socialism at this time also attempted to provide alternatives to working mothers, including state-sponsored day cares and incentives for having children (Freedman, 2002).

Despite these nations’ collective socialist backgrounds, former Soviet nations each have independent identities and differ in their demographics and levels of gender equality. Kyrgyzstan, Tajikistan, and Azerbaijan are predominately Muslim countries, while citizens of Moldova primarily identify as Orthodox Christian. These four countries also vary in terms of gender inequality. According to the Social Institutions and Gender Index (2017), a measure of gender inequality (scale from 0 to 1) in regard to discrimination across social institutions (including laws, attitudes, and practices), Kyrgyzstan (0.1598) and Tajikistan (0.1393) were rated “medium.” In contrast, Azerbaijan (0.2403) was rated high in terms of gender inequality and Moldova (0.0513) was rated low in regard to gender inequality. Each country has a law that specifically criminalizes domestic violence and rape. Moldova is the only one of the four countries to include marital rape as a criminal offense. Although all four of these countries were part of the former Soviet Union, it is clear there are differences between each. Based on this overview, it is expected that Moldova differs from Kyrgyzstan, Tajikistan, and Azerbaijan.

Present Estimates of IPV

Estimates over time of IPV, or violent crimes committed against persons by a current or former spouse, partner, or significant other (boyfriend or girlfriend), indicate a number of important trends. First, estimates for the United States suggest that IPV primarily affects women, though it is not limited to female victims (Rennison & Welchans, 2000). Second, in the United States, rates of IPV were higher for low-income households and among those who were divorced/separated and never married. Finally, between 1976 and 1998, the rate of IPV homicide for White females remained stable (Rennison & Welchans, 2000). More recently, findings from the 2010 National Intimate Partner and Sexual Violence Survey report that in the United States 32.9% of women have experienced some form of physical violence by an intimate partner with 9.4% reporting being raped by an intimate partner in their lifetime (Black et al., 2011). These findings are important benchmarks for scholars of IPV to place global or regional findings in context.

Nevertheless, there are a number of cultural and structural differences between the United States and both similarly developed nations and their developing counterparts. And as the World Report on Violence and Health discussed, researchers “have suggested that structural inequalities between men and women, rigid gender roles, and notions of manhood linked to dominance, male honour and aggression increase the risk of IPV” (WHO, 2002, p. 100). Given the occurrence of these factors vary across the world, country context may shape the prevalence of abuse and risk factors of IPV. Recent reports suggest that globally 35.6% of women will experience some form of physical and/or sexual violence by an intimate partner (WHO, 2013). Further illustrating the global threat of IPV, Stöckl and colleagues (2013) report that globally 38.6% of femicides were the result of IPV.

More specific to the focal region in current analyses, prior research has indicated rates of physical abuse in Russia range from 27% to 40% (Cubbins & Vannoy, 2005; Gorshkova & Shurygina, 2003) and from 13% to 19% in Ukraine (Barrett et al., 2012; Centers for Disease Control and Prevention, 2003). According to extant research, the rate of IPV in Kyrgyzstan is 28% compared with 50% of Tajik women who have been physically abused by a husband and/or their in-laws (Human Rights Watch, 2015; WHO, 2000). It is difficult to make comparisons across these rates because they come from different data sources. The current study used a standardized measure to examine differences across four transitional former Soviet countries not only in rates of physical and sexual abuse but also predictors that may influence risk of IPV.

Risk Factors for Physical and Sexual Abuse

The WHO has identified a number of risk factors that have been associated with risk of experiencing IPV. These risk factors include lower levels of education, exposure to child maltreatment, witnessing family violence, harmful use of alcohol, and possessing attitudes that are accepting of violence and gender inequality. More specific risk factors for experiencing IPV (not including sexual violence) are a history of violence, marital dissatisfaction, and difficulty in communication between partners (WHO, 2012). Below, we review these risk factors in more specific detail, especially as they relate to the four transitional nations included in analyses.

Supportive Attitudes Toward IPV

There are a number of attitudinal characteristics that research has shown correlate with either increased likelihood of perpetration or acceptance of violence against women. Research on the acceptability of violence against women in the European Union suggests that men who know a perpetrator view violence against women as more acceptable and that higher levels of acceptability are associated with victim blaming (Gracia & Herrero, 2006). IPV against women was found to be widely accepted by both men and women in 17 Sub-Saharan African nations, only mitigated by higher levels of wealth, education, urbanization, access to media, and joint decision making (Uthman, Moradi, & Lawoko, 2009). Similar research conducted in Arab and Islamic countries reports that prevailing attitudes toward domestic violence are that in most instances the violence is a private matter and justifiable (Douki, Nacef, Belhadj, Bouasker, & Ghachem, 2003). Macro-level research suggests that in nations where positive or accepting attitudes toward violence against women are more prevalent, individuals within those nations hold more accepting attitudes toward IPV (Hayes & Boyd, 2017). Likewise, research has found that positive attitudes toward IPV are also an individual-level risk factor for perpetration (Fleming et al., 2015). For these reasons, we hypothesize as follows:

Hypothesis 1 (H1): Attitudes supportive of IPV will increase the risk of physical and sexual abuse.

Intergenerational Transmission of Violence

The intergenerational transmission of violence, rooted in social learning theory (Bandura, 1969, 1973), proposes children who witness or experience violence in their family of origin learn it is an appropriate response to conflict. Extant cross-national research on it, including studies that have relied on Demographic and Health Survey (DHS) data, has reported mixed effects. Dude (2007) found witnessing IPV as a child decreased risk of later IPV, whereas Barrett et al. (2012) found it increased risk of both physical and emotional abuse in Ukraine. O’Leary, Tintle, Bromet, and Gluzman (2008) found witnessing IPV as a child did not increase risk of IPV later in life. Barrett et al.’s (2012) findings are more consistent with cross-national research on the intergenerational transmission of violence and later risk of IPV, including prior research on Russian victims of IPV (Cubbins & Vannoy, 2005). Violence in one’s family of origin can also include experiencing abuse. Extant research had indicated that women who reported physical abuse during their childhood were more likely to experience IPV as an adult (Bensley, Van Eenwyk, & Simmons, 2003). This research suggests that within developing nations, experiencing or witnessing violence within one’s family of origin increases the risk of IPV later in life. It is important than to include predictors of the intergenerational transmission of violence when examining risk factors for physical and sexual abuse. Therefore, we hypothesize as follows:

Hypothesis 2 (H2): Experience of childhood violence will increase the risk of physical and sexual abuse.

Hypothesis 3 (H3): Witnessing violence against one’s mother will increase the risk of physical and sexual abuse.

Husband’s Alcohol Use

Generally, research suggests that alcohol use has an effect on the risk of both victimization and perpetration of IPV (see Devries et al., 2014; Foran & O’Leary, 2008). However, the nature of this relationship is unclear and meta-analytic review suggests that the relationship is small to moderate (Foran & O’Leary, 2008). Research has shown, though, that the relationship between alcohol and IPV persists even after controlling for a number of potentially mediating variables such as socioeconomic status and race (e.g., Pan, Neidig, & O’Leary, 1994). Alcohol consumption in excess has been reported in a number of former Soviet states (see Pomerleau et al., 2008; Treml, 1997). Indeed, Stickley and colleagues (2007) report that “acute alcohol poisoning has now reached unprecedent rates in parts of the” former Soviet Union (p. 444). Given the critical role of alcohol within this national context and previous findings that indicate alcohol use was associated with increased risk of physical and sexual abuse in Ukraine, a measure of husband’s alcohol use is included in analyses. For these reasons, we hypothesize as follows:

Hypothesis 4 (H4): Husband’s use of alcohol will increase risk of physical and sexual abuse.

Decision Making Within the Couple Dyad

Equality in decision making has been shown to affect risk of IPV (Coleman & Straus, 1986; Hindin & Adair, 2002; Mann & Takyi, 2009). However, the effect is not always in the expected direction. For instance, risk of IPV may increase when the husband makes all the decisions (Coleman & Straus, 1986; Hindin & Adair, 2002) or it may increase when the woman makes decisions because the husband feels powerless and uses violence to exert control (Mann & Takyi, 2009). What is unique about former Soviet nations is that communism influenced the woman’s role within the household (Freedman, 2002). Within communism, women were expected to work and contribute to the household. Child care was provided by the state. In these instances, equality in decision making may have a limited effect because the decisions a couple would have to make may already have been determined by the government. Prior research in the Ukrainian context found that decision making was not associated with risk of experiencing IPV (Barrett et al., 2012). Given prior research has not found an association between decision making and risk of IPV among former Soviet nations, we do not propose a hypothesis for this measure.

Experience of Controlling Behavior by an Intimate Partner

Within both developing and developed nations, controlling behaviors may be a motivating factor of physical and sexual abuse (Basile, 1999; Gage & Hutchinson, 2006; Tanha, Beck, Figueredo, & Raghavan, 2010). Physical and sexual abuse may occur then when the abusive partner is no longer able to maintain control through nonphysically violent behaviors. Prior research with the DHS has shown that controlling behaviors are associated with IPV in Nigeria (Antai, 2011), Mali (Hayes & van Baak, 2017), and Ukraine (Barrett et al., 2012). For the aforementioned reasons, we hypothesize as follows:

Hypothesis 5 (H5): As the number of reported controlling behaviors the respondent experienced increases, risk of physical and sexual violence will increase.

Respondent’s Use of Violence

Victims who engage in violence against an intimate partner may do so in an effort to resolve an argument and/or to resist the abusive partner (M. P. Johnson, 2005). Regardless of the victim’s intention, research has found that when IPV victims use violence, the likelihood of continued IPV increases (Bachman, Saltzman, Thompson, & Carmody, 2002; Fanslow, Gulliver, Dixon, & Ayallo, 2014). The current study examines the association of a respondent’s use of violence against her partner with risk of physical and sexual abuse and hypothesizes as follows:

Hypothesis 6 (H6): Respondent’s use of violence against an intimate partner will increase risk of physical and sexual abuse.

Theoretical Framework

According to I³ (I-cubed) theory (Finkel, 2007; Finkel et al., 2012), many of the known risk factors for IPV are elements of three core processes or pathways to violence. These three processes are referred to as instigation, impellance, and inhibition (Finkel et al., 2012). Although we do not specifically attempt to address this theory in the current analyses, our measures may capture elements of instigation, impellance (i.e., individual drives or dispositions toward violence), and inhibition (i.e., individual or contextual factors that may produce a breakdown in self-regulation). Notably, we include the respondent’s use of violence in our models, which may be a form of instigation. We measure attitudes toward IPV and controlling behaviors, which may capture the drives of impellance. Finally, we capture alcohol use by the respondent’s husband, which may be interpreted as a measure of inhibition. The I³ perspective suggests that these core elements are functional at an individual level, and that cultural effects or socialization toward IPV cannot provide a complete understanding of the nature of IPV (Finkel et al., 2012). More specifically, the perspective suggests that IPV is largely situational and the product of an interaction of instigation, impellance, and/or inhibition (Finkel et al., 2012).

Data and Method

Data came from four DHS surveys. The DHS, which have been conducted in more than 90 developing countries, are standardized to allow for comparisons both historically and cross-nationally. The goal of the DHS is to monitor demographic and health trends on a 5-year basis and to collect data on fertility, marriage, family planning methods, nutrition, and sex behaviors (Boyle & Carbone-Lopez, 2006). Since the DHS surveys were implemented, it has since been modified to include additional modules on key social issues, such as female genital cutting, out of pocket health care expenditures, maternal mortality, and IPV.

The current study used nationally representative data from the four former Soviet nations that conducted the IPV module and were not missing data on key predictors. The IPV module is limited to female respondents, which limited analyses to female respondents. Female respondents from Tajikistan (n = 3,093), Kyrgyzstan (n = 3,171), Azerbaijan (n = 2,909), and Moldova (n = 3,355) were included in analyses. As noted above, despite all these countries belonging to the former Soviet Union, each of these nations also have unique histories.

Dependent Variables

Physical abuse

The dependent measures captured the respondent’s experience of physical and sexual abuse.¹ Respondents across all four former Soviet nations reported whether their partner ever (a) pushed, shook, or threw something; (b) slapped; (c) punched with fist or hit by something harmful; (d) kicked or dragged; (e) strangled or burnt; (f) threatened with gun/knife or other weapon; and (g) twisted or pulled hair. Consistent with prior research (Lawoko, Dalal, Jiayou, & Jansson, 2007), responses were collapsed into a binary measure, where 1 = “experienced physical abuse” and 0 = “did not experience physical abuse.”

Sexual abuse

Sexual abuse was measured with the following questions that identified whether the respondent was (a) ever physically forced into unwanted sex by her husband/partner, (b) ever forced into other unwanted sexual acts by her husband/partner, and (c) ever physically forced to perform a sexual act the respondent did not want. Similar to the physical abuse measure, experience of sexual abuse was binary coded into 1 = “experienced sexual abuse” and 0 = “did not experience sexual abuse.” Descriptive statistics for the outcome measures and independent variables by country are presented in Table 1.

Table 1.

Descriptive Statistics.

	Tajikistan (n = 3,093)		Kyrgyzstan (n = 3,171)		Azerbaijan (n = 2,909)		Moldova (n = 3,355)
	M/%	SD	M/%	SD	M/%	SD	M/%	SD
Dependent variables
Physical abuse	18.96%		20.62%		9.29%		18.38%
Sexual abuse	3.54%		3.45%		1.46%		2.88%
Independent variables
IPV attitudes greater than country mean	45.01%		36.83%		41.58%		18.48%
Father abused mother	14.03%		14.73%		17.11%		35.16%
Respondent’s family violence history	3.74%		3.25%		3.37%		4.32%
Respondent’s use of violence	1.75%		3.51%		0.16%		5.98%
Husband’s alcohol use	18.30%		36.32%		53.88%		77.36%
Urban	27.07%		34.89%		56.09%		42.81%
Interview not in country’s native language	7.04%		39.71%		0.18%		25.33%
Controlling behaviors	1.66	1.48	1.69	1.28	1.54	1.30	1.54	1.73
Decision-making scale	2.35	1.68	3.51	1.02	2.64	1.43	2.91	0.36
Education	4.72	0.91	5.38	0.67	4.60	0.80	4.66	0.85
Age	32.74	8.15	33.10	8.41	34.21	8.19	34.49	8.65
Wealth	3.13	1.44	3.12	1.41	3.09	1.41	3.23	1.39
Total number of children	3.10	1.94	2.60	1.54	2.28	1.28	1.79	1.12

Note. IPV = intimate partner violence.

Independent Variables

IPV attitudes greater than country mean

Respondents were asked whether wife beating was ever justified in five scenarios: (a) “If the wife goes out without telling her husband,” (b) “If she neglects the children,” (c) “If she argues with her husband,” (d) “If she refuses to have sex,” and (e) “If she burns the food.” To control for variations across the four countries included in analyses, the average attitudes for each country were calculated for all respondents who were not missing data on these five measures (range = 1.00-5.00; Moldova M = 0.29; Tajikistan M = 2.19; Azerbaijan M = 1.39; Kyrgyzstan M = 0.80). A binary measure was then created to capture whether the respondent’s attitudes toward IPV were more supportive than her country’s average, where 1 = “respondent’s attitudes toward IPV were more supportive than the country average.”

Family history of violence

First, a binary measure that captured whether the respondent’s mother, father, or brother/sister ever physically abused her over her lifetime was included in analyses to represent one dimension of family history of violence. The second dimension includes witnessing violence in one’s family of origin, which can also affect risk of IPV later in life (Bensley et al., 2003; Coker, Smith, McKeown, & King, 2000; H. Johnson, Ollus, & Nevala, 2008). Whether the respondent’s father ever abused her mother was included as a dichotomous measure, where 1 = “respondent’s father ever abused her mother.”

Husband’s alcohol use

Whether the respondent’s husband consumed alcohol was captured with a binary measure, where 1 = “respondent’s husband drinks alcohol.”

Decision making

For three of the countries (Tajikistan, Azerbaijan, and Kyrgyzstan) responses across four questions were collapsed into a scale that measured whether the respondent had a say in health care, large purchases for the home, when to visit family, and how to spend the husband’s earnings (range = 0.00-4.00). Moldova was missing the measure that captured whether respondent had a say in how to spend her husband’s earnings. The decision-making scale for Moldova therefore ranged from 0.00 to 3.00. The higher the value on the decision-making scale, the greater the respondent’s say in household decisions.

Controlling behaviors

The respondent’s experience of controlling behavior was measured with a scale (range = 0.00-8.00) where higher values indicated the respondent experienced more controlling behaviors. Behaviors in this index included whether the respondent’s partner was ever jealous, whether her partner insisted on knowing where she was, whether her partner ever accused her of being unfaithful, and whether her partner ever humiliated, threatened, or insulted her.

Respondent’s use of violence

Whether the respondent ever initiated violence against her husband was measured with a dichotomous outcome, where 1 = “respondent initiated violence” and 0 = “respondent never initiated violence against her husband.”

Control variables

First, education was included as an ordinal measure (range = 1.00-6.00). On average, women in these former Soviet countries had some secondary education. Age and total number of children were included as continuous measures. If the respondent lived in an urban, compared with rural, location was included as a binary measure, where 1 = “urban.” Socioeconomic status was measured with the “Wealth Index,” which was a composite of the respondent’s standard of living, including ownership of goods, how one’s house was constructed, and access to water and sanitation (Rutstein & Johnson, 2004). The wealth index ranged from 1.00 (i.e., poorest) to 5.00 (i.e., richest). Finally, a binary measure captured whether the interview was conducted in the country’s dominant language. Civil wars have been fought in this region following the fall of the Soviet Union. The language the interview was conducted in served as a proxy for whether the respondent was part of a minority group.

Analyses

Bivariate statistics are presented in Table 2. A series of logistic regressions were used to examine the effect of the predictors on the odds of experiencing physical and sexual abuse across the four former Soviet nations. We also conducted coefficient comparisons to examine the differences across the coefficients of the independent samples (Paternoster, Brame, Mazerolle, & Piquero, 1998). We pay specific attention to the differences between Moldova compared with the other three nations included in analyses. Findings from the coefficient comparisons are based on an alpha value of .01. Missing data was listwise deleted. To correct for over- and under-sampled regions and to ensure the representativeness of the sample, proper weights assigned to respondents were used in the analyses (DHS, n.d.).

Table 2.

Bivariate Correlations Between Variables.

	Physical Abuse	Sexual Abuse	IPV Attitudes	Father Abused Mother	Family Violence History	Respondent’s Use of Violence	Husband Alcohol Use	Urban	Interview Not in Native Language	Controlling Behavior	Decision-Making Scale	Education	Age	Wealth	Total Children
Physical abuse	1
Sexual abuse	.29***	1
IPV attitudes	.12***	.04***	1
Father abused mother	.18***	.08***	.01	1
Respondent’s family violence history	.11***	.09***	.03***	.15***	1
Respondent’s use of violence	.21***	.15***	.00	.10***	.11***	1
Husband’s alcohol use	.17***	.07***	−.07***	.21***	.05***	.10***	1
Urban	−.04***	−.04***	−.18***	−.02**	−.03***	.02*	.06***	1
Interview not in native language	.00	.01	−.13***	−.03**	.02*	.09***	.00	.07***	1
Controlling behaviors	.40***	.24***	.17***	.14***	.09***	.17***	.12***	−.03***	−.01	1
Decision-making scale	−.02*	−.03**	−.17***	−.02**	−.04***	.00	.07***	.06***	.13***	−.11***	1
Education	−.01	.00	−.14***	−.09***	−.02	.00	−.04***	.15***	.15***	−.05***	.23***	1
Age	.04***	.03**	−.04***	.02*	−.04***	.00	.12***	.02*	.02*	−.10***	.21***	.14***	1
Wealth	−.06***	−.02*	−.23***	−.06***	−.02*	.03**	.00	.63***	.16***	−.04***	.08***	.23***	−.01	1
Total number of children	.09***	.05***	.14***	−.03***	−.01	−.03**	−.08***	−.20***	−.09***	−.02*	.08***	.01	.48***	−.22***	1

Note. IPV = intimate partner violence.

p < .05 level (2-tailed). **p < .01 level (2-tailed). ***p < .001 level (2-tailed).

Results

Physical abuse ranged from about one in 10 women in Azerbaijan to almost a fifth of women in Kyrgyzstan. Tajikistan (18.96%) and Moldova (18.38%) had very similar rates of physical abuse. In contrast, prevalence of sexual abuse across all four transitional countries was lower than the prevalence of physical abuse. Tajikistan had the highest prevalence of sexual abuse at 3.54%, followed by Kyrgyzstan (3.45%), Moldova (2.88%), and Azerbaijan (1.46%). As noted, controlling behaviors capture the experience of IPV that may not necessarily involve physically violent behavior. Across all four countries included in analyses, the average number of controlling behaviors reported was greater than one. The majority of women in these former Soviet nations had experienced at least one type of controlling behavior during their lifetime.

As reported in Table 3, if the respondent reported they had more supportive attitudes toward IPV (compared with the mean for her country), she was at increased of risk of physical abuse in all four countries and sexual abuse in Azerbaijan. These findings partially support H1. Across all four countries and supporting H2 and H3, a global experience of violence in one’s family of origin or witnessing IPV between parents significantly increased the risk of physical and sexual abuse. If the respondent witnessed her father abuse her mother, the risk for physical abuse was about 1.5 to 2 times greater in Tajikistan, Kyrgyzstan, Azerbaijan, and Moldova. Similarly, risk of sexual abuse was almost 4 times greater for Tajik women if they witnessed their father abuse their mother. The effect of witnessing one’s father abuse their mother on risk of sexual abuse significantly varied between Tajikistan and Moldova. Also, the effect of witnessing one’s father abuse their mother on risk of physical abuse significantly varied between Kyrgyzstan and Moldova. If the woman experienced violence from her mother, father, brother, or sister, her risk of sexual abuse was 4 times greater in Kyrgyzstan, over 6 times greater in Azerbaijan, and about 2.5 times greater in Moldova.

Table 3.

Logistic Regression for Odds of Physical and Sexual Abuse Across Four Transnational Nations.

	Tajikistan		Kyrgyzstan		Azerbaijan		Moldova
	Physical Abuse (n = 3,093)	Sexual Abuse (n = 3,094)	Physical Abuse (n = 3,171)	Sexual Abuse (n = 3,171)	Physical Abuse (n = 2,909)	Sexual Abuse (n = 2,909)	Physical Abuse (n = 3,355)	Sexual Abuse (n = 3,355)
IPV attitudes greater than country mean	1.35*	0.78	1.96***	1.20	1.98***	2.74***	1.32*	1.15
Controlling behaviors	1.49***	1.51***	2.10***	2.07***	1.87***	1.64***	1.96***	2.14***
Father abused mother	2.05***	3.77***	2.48***	1.19	1.47*	0.65	1.64***	0.77
Respondent’s family violence history	2.06***	0.77	1.61	4.17***	4.29***	6.73***	1.46	2.48*
Respondent’s use of violence	3.86***	2.96***	5.10***	3.49***	—	—	4.92***	1.74
Husband’s alcohol use	2.94***	2.45***	2.77***	2.56***	1.73***	0.98	2.38***	2.08
Decision-making scale	1.04	0.91	0.97	0.85	0.94	1.05	0.75*	0.93
Education	1.00	1.00	1.30***	1.00	1.04	0.77	0.93	1.21
Age	0.98*	1.03	1.03***	1.01	1.01	1.01	1.02***	1.01
Urban	1.83***	0.90	1.14	0.53	1.19	1.75	0.81	0.76
Wealth	0.94	0.96	1.12***	0.94	0.84*	1.20	0.82***	1.05
Total number of children	1.15***	1.10	1.14***	1.23*	1.11	1.15	1.22***	1.19
Interview not in country’s native language	0.52***	0.20**	0.76*	1.32	—	—	1.21	0.59
χ² (df)	433.53 (13)***	143.47 (13)***	893.43 (13)***	306.83 (13)***	381.65 (11)***	62.80 (11)***	1,031.04 (13)***	313.83 (13)***
Nagelkerke R²	.21	.17	.38	.36	.27	.15	.43	.39

Note. IPV = intimate partner violence. * p < 0.05; ** p < 0.01; *** p < 0.001

Husband’s use of alcohol increased the risk of physical abuse for women in all four countries. However, husband’s use of alcohol only increased the risk of sexual abuse for women in Tajikistan and Kyrgyzstan, lending partial support to H4. As expected, based on prior research, the respondent’s say in decision making was not significant across all countries and types of abuse except for the risk of physical abuse in Moldova. Across all four countries and supporting H5, as the number of controlling behaviors the respondent reported ever experiencing increased, the risk of both physical and sexual abuse increased. The effect of controlling behaviors on both physical and sexual abuse significantly varied between Tajikistan and Moldova. The respondent’s use of violence against her husband or partner increased the risk of physical abuse in Tajikistan, Kyrgyzstan, and Moldova and the risk of sexual abuse in Tajikistan and Kyrgyzstan. These findings lend partial support to H6.

Having more education increased risk of physical abuse in Kyrgyzstan. Older women were at increased risk of physical abuse in Kyrgyzstan and Moldova and decreased risk in Tajikistan. Tajik women living in urban areas were at increased risk of physical abuse. Wealth increased the risk of physical abuse in Kyrgyzstan but decreased the risk of physical abuse in Azerbaijan and Moldova. If the interview was not conducted in the country’s native language, risk of physical abuse decreased in Tajikistan and Kyrgyzstan and risk of sexual abuse was decreased in Tajikistan.

Discussion

Little is known about risk factors of IPV in these transitional nations. The current study examined the effect of predictors of IPV across four former nations of the Soviet Union. Overall, rates of physical abuse were similar within Tajikistan and Moldova, while Kyrgyzstan had the highest prevalence. In contrast, the prevalence of sexual abuse was very similar across countries and was much lower (between 1.00% and 3.50%) than cross-national estimates of physical abuse. Across the four former Soviet nations included in analyses, the experience of controlling behaviors, husband’s alcohol use, and witnessing IPV or experiencing abuse during childhood were all significantly associated with risk of physical and sexual abuse later in life.

Across four diverse former Soviet nations, controlling behaviors significantly increased the risk of physical and sexual abuse. This finding highlights the importance of coercive control cross-nationally, which has been found to increase the risk of IPV in Mali (Hayes & van Baak, 2017), Ukraine (Barrett et al., 2012), and in developed countries (Tanha et al., 2010). Similar to the conclusions of Barrett and colleagues (2012) analyses of the Ukrainian DHS data, the findings from the current study lend support to the continued use of the “Power and Control Wheel” (Domestic Abuse Intervention Services, 2008) in former countries of the Soviet Union. Prevention programs for violence against women should incorporate a focus on men’s use of power and control. In addition, scholars have noted that abusive men may actually view themselves as moral individuals, despite the behavior they engage in (Vecina, Marzana, & Paruzel-Czachura, 2015). Given moral socialization occurs early in life, family, teachers, and peers can help engage adolescents as they learn morality (Marzana, Vecina, & Alfieri, 2016) and may be a viable area of early intervention.

Across all nations, when respondents witnessed their father abuse their mother, they reported an increased likelihood of experiencing physical abuse. Social learning theory has been used to inform our understanding of the intergenerational transmission of violence. Witnessing one’s father abuse their mother may teach the child that love is paired with fear and that future intimate relationships have the potential to involve violence (Holt, Buckley, & Whelan, 2008). Similarly, experiencing violence in one’s family of origin increased the risk of either physical or sexual abuse across the four nations included in analyses. Triangulation of findings across these two measures strengthens the overall conclusion that the intergenerational transmission of violence is a cross-national phenomenon.

We find here that alcohol use and physical abuse are correlated in all four countries. This finding is consistent with research on IPV (Devries et al., 2014; Jewkes, 2002). The role of alcohol in violence, particularly in IPV, is multifaceted as consumption can interact with a variety of other factors to increase risk. Alcohol consumption has a number of physiological effects resulting in temporary conditions (e.g., reducing inhibitions), which may increase the likelihood that an individual may act or react violently. Given that our sample of former Soviet nations exhibited pervasive alcohol use and abuse (see Pomerleau et al., 2008), the nature and strength of the relationship could have serious implications.

With the exception of physical abuse in Moldova, decision making was not a significant predictor of IPV risk. One of the principal ideals of communism within the Soviet Union was that there should be equality between men and women and that women should also contribute financially to the household (Freedman, 2002). As evidenced by the descriptive statistics of the decision-making scale, the majority of women across these four nations had a significant say in household decisions. The underlying ideals of communism may have indeed carried into the 21st century. In comparison, Moldova is geographically distinct (i.e., located in the Europe continent) from the other nations (i.e., located in the Asia continent) included in analyses and has been in the process of developing a relation with the European Union. This may be why decision making was significant in Moldova. Decision making in Moldova may represent a form of empowerment that is distinct from the other former Soviet nations, suggesting a shift in gender norms and gender role expectations within the Moldavian context.

We find that many of the individual factors related to I³ are important in understanding IPV within these transitional countries. More specifically, controlling behaviors may suggest impellance, husband’s use of alcohol may be a form of inhibition and the respondent’s use of violence could be a form of instigation. Nevertheless, we cannot conclusively separate socialization toward IPV from immediate contextual and dispositional factors, nor do we evaluate interactions. Our findings include cross-national differences in effect sizes on individual-level items that may be the result of differential socialization. For example, we find that attitudes supportive of IPV that are greater than the nation’s mean are positively associated with IPV, which supports the notion of impellance. However, the size of this effect is significantly different across nations. Ultimately, IPV may be largely situational, though here differential social context appears to play a role. Future research should incorporate measures of I³ to better isolate cultural effects.

It is also important to note the low rates of sexual abuse across the four transitional countries given many studies, especially within developed nations, have found one in five women will be the victim of sexual assault at some point in their life (see Black et al., 2011). Barrett et al. (2012) also reported very low prevalence rates of sexual abuse in Ukraine. These low prevalence sexual abuse rates with DHS data may be the result of the operationalization of sexual abuse by the DHS rather than an accurate reflection of sexual assault rates. The sexual abuse measure included in the DHS focused on the use of physical force for unwanted sex acts by the respondent’s husband/partner. Given the patriarchal nature of these countries, women may not see the sexual act as forced but view it as part of the marital covenant (Barrett et al., 2012). In addition, many acts of sexual assault do not meet the criteria for rape. Measures included in analyses are more representative of rape than a broader definition of sexual violence. Future research should incorporate measures that capture both rape and sexual assault.

Nevertheless, there are limitations that merit mention. First, there were small sample sizes for respondent’s use of violence in Azerbaijan (n = 9). For the Azerbaijan analyses, respondent’s use of violence and whether the interview was not in the country’s native language were dropped from the models due to low variation within the measure. Second, the DHS were not conducted with male respondents in Tajikistan. Therefore, men’s attitudes toward IPV could not be compared across countries but remains an area of future research. Third, the DHS did not have recent data for all former Soviet nations. Even if data were available for all former Soviet nations, the sample size at the country level (i.e., Level 2) would still not be large enough to conduct a multilevel study that could assess country and individual-level predictors of IPV among former nations of the Soviet Union. This is because multilevel modeling has minimum sampling requirements at Level 2 to increase reliability of the model (Maas & Hox, 2005). Fourth, consistent with prior research utilizing DHS data (e.g., Hayes & van Baak, 2017; Lawoko et al., 2007), binary indicators were used to capture the experience of physical and sexual abuse because scalar measures were extremely skewed. Future research should examine how these risk factors impact the severity of IPV. Finally, controlling behaviors were treated as a predictor of physical and sexual abuse based on the theoretical propositions of coercive control (Stark, 2007). Nevertheless, there is no way to determine the time order of controlling behaviors and IPV. However, prior research with DHS data has treated controlling behaviors as a predictor of physical and sexual abuse (Barrett et al., 2012; Hayes & van Baak, 2017).

Conclusion

Findings from the current analyses suggest that many of the same risk factors of IPV in developing and developed nations influence risk of IPV in transitional nations. The experience of controlling behaviors, husband’s alcohol use, and witnessing IPV or experiencing abuse during childhood all increased the risk of experiencing physical and/or sexual abuse in four former nations of the Soviet Union. Therefore, the intergenerational transmission of violence may be a cross-national phenomenon. In addition, equality in decision making did not have an effect on risk of physical and sexual violence, except for risk of physical abuse in Moldova. This may do the historical context of these nations. Taken together, findings suggest that country context may influence individual risk factors of IPV.

Footnotes

Declaration of Conflicting Interests

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

Funding

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

Notes

Author Biographies

Brittany E. Hayes (PhD, John Jay College of Criminal Justice, CUNY) is an assistant professor in the Department of Criminal Justice and Criminology and the assistant director of the Crime Victims’ Institute at Sam Houston State University. Her research focuses on victimology and has been published in the American Sociological Review and the Journal of Research in Crime and Delinquency. She is the 2017 recipient of the Division of Victimology Researcher of the Year Award from the American Society of Criminology and 2018 recipient of the Outstanding Paper Award from the Academy of Criminal Justice Sciences.

Ryan Randa (PhD, University of Cincinnati) is an assistant professor in the Department of Criminal Justice and Criminology in the College of Criminal Justice at Sam Houston State University. Research interests include bullying behavior, victimization, and fear of crime. His recent work in this area has appeared in Crime & Delinquency, Journal of Criminal Justice, Youth Violence and Juvenile Justice, Victims & Offenders, and Journal of Interpersonal Violence.

References

Alio

A. P.

Daley

E. M.

Nana

P. N.

Duan

Salihu

H. M.

(2009). Intimate partner violence and contraception use among women in Sub-Saharan Africa. International Journal of Gynecology & Obstetrics, 107, 35-38.

Antai

(2011). Controlling behavior, power relations within intimate relationships and intimate partner physical and sexual violence against women in Nigeria. BMC Public Health, 11(1), Article 511.

Bachman

Saltzman

L. E.

Thompson

M. P.

Carmody

D. C.

(2002). Disentangling the effects of self-protective behaviors on the risk of injury in assaults against women. Journal of Quantitative Criminology, 18, 135-157.

Bandura

(1969). Principles of behavior modification. New York, NY: Holt, Rinehart and Winston.

Bandura

(1973). Aggression: A social learning analysis. Englewood Cliffs, NJ: Prentice Hall.

Barrett

B. J.

Habibov

Chernyak

(2012). Factors affecting prevalence and extent of intimate partner violence in Ukraine: Evidence from a nationally representative survey. Violence Against Women, 18, 1147-1176.

Basile

K. C.

(1999). Rape by acquiescence: The ways in which women “give in” to unwanted sex with their husbands. Violence Against Women, 5, 1036-1058.

Bensley

Van Eenwyk

Simmons

K. W.

(2003). Childhood family violence history and women’s risk for intimate partner violence and poor health. American Journal of Preventive Medicine, 25, 38-44.

Black

Basile

Breiding

Smith

Walters

Merrick

. . . Stevens

(2011). National Intimate Partner and Sexual Violence Survey: 2010 summary report. Atlanta, GA: National Center for Injury Prevention and Control, Centers for Disease Control and Prevention.

10.

Boyle

E. H.

Carbone-Lopez

K. C.

(2006). Movement frames and African women’s explanations for opposing female genital cutting. International Journal of Comparative Sociology, 47, 435-465.

11.

Centers for Disease Control and Prevention. (2003). Reproductive, maternal and child health in Eastern Europe and Eurasia: A comparative report. Retrieved from https://dhsprogram.com/pubs/pdf/OD28/00FrontMatter.pdf

12.

Coker

A. L.

Smith

P. H.

McKeown

R. E.

King

M. J.

(2000). Frequency and correlates of intimate partner violence by type: Physical, sexual, and psychological battering. American Journal of Public Health, 90, 553-559.

13.

Coleman

D. H.

Straus

M. A.

(1986). Marital power, conflict, and violence in a nationally representative sample of American couples. Violence & Victims, 1, 141-157.

14.

Cubbins

L. A.

Vannoy

(2005). Socioeconomic resources, gender traditionalism, and wife abuse in urban Russian couples. Journal of Marriage and Family, 67, 37-52.

15.

Demographic and Health Surveys. (n.d.). Using datasets for analysis. Retrieved from https://dhsprogram.com/data/Using-DataSets-for-Analysis.cfm

16.

Devries

K. M.

Child

J. C.

Bacchus

L. J.

Mak

Falder

Graham

. . . Heise

(2014). Intimate partner violence victimization and alcohol consumption in women: A systematic review and meta-analysis. Addiction, 109, 379-391.

17.

Domestic Abuse Intervention Services. (2008). Home of the Duluth model: Social change to end violence against women. Available from http://www.theduluthmodel.org/

18.

Douki

Nacef

Belhadj

Bouasker

Ghachem

(2003). Violence against women in Arab and Islamic countries. Archives of Women’s Mental Health, 6, 165-171.

19.

Dude

(2007). Intimate partner violence and increased lifetime risk of sexually transmitted infection among women in Ukraine. Studies in Family Planning, 38, 89-100.

20.

Fanslow

J. L.

Gulliver

Dixon

Ayallo

(2014). Hitting back: Women’s use of physical violence against violent male partners, in the context of a violent episode. Journal of Interpersonal Violence, 30, 2963-2979.

21.

Finkel

E. J.

(2007). Impelling and inhibiting forces in the perpetration of intimate partner violence. Review of General Psychology, 11, 193-207.

22.

Finkel

E. J.

DeWall

C. N.

Slotter

E. B.

McNulty

J. K.

Pond

R. S.

Jr. Atkins

D. C.

(2012). Using I³ theory to clarify when dispositional aggressiveness predicts intimate partner violence perpetration. Journal of Personality and Social Psychology, 102, 533-549.

23.

Fleming

P. J.

McCleary-Sills

Morton

Levtov

Heilman

Barker

(2015). Risk factors for men’s lifetime perpetration of physical violence against intimate partners: Results from the International Men and Gender Equality Survey (IMAGES) in eight countries. PLoS ONE, 10(3), Article e0118639.

24.

Foran

H. M.

O’Leary

K. D.

(2008). Alcohol and intimate partner violence: A meta-analytic review. Clinical Psychology Review, 28, 1222-1234.

25.

Freedman

E. B.

(2002). No turning back: The history of feminism and the future of women. New York, NY: Ballantine Books.

26.

Gage

A. J.

Hutchinson

P. L.

(2006). Power, control, and intimate partner sexual violence in Haiti. Archives of Sexual Behavior, 35, 11-24.

27.

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. The Lancet, 368, 1260-1269.

28.

Gómez

A. M.

Speizer

I. S.

(2010). Community-level intimate partner violence and the circumstances of first sex among young women from five African countries. Reproductive Health, 7, Article 11.

29.

Gorshkova

I. D.

Shurygina

I. I.

(2003). Nasilie nad zhenami v sovremennykh rossiiskikh sem’yakh [Violence against wives in contemporary Russian families]. Moscow, Russia: MAKS Press.

30.

Gracia

Herrero

(2006). Acceptability of domestic violence against women in the European union: A multilevel analysis. Journal of Epidemiology and Community Health, 60, 123-129.

31.

Hayes

B. E.

Boyd

K. A.

(2017). Influence of individual- and national-level factors on attitudes toward intimate partner violence. Sociological Perspectives, 60, 685-701.

32.

Hayes

B. E.

van Baak

(2017). Risk factors of physical and sexual abuse for women in Mali: Findings from a nationally representative sample. Violence Against Women, 23, 1361-1381.

33.

Hindin

M. J.

Adair

L. S.

(2002). Who’s at risk? Factors associated with intimate partner violence in the Philippines. Social Science & Medicine, 55, 1385-1399.

34.

Holt

Buckley

Whelan

(2008). The impact of exposure to domestic violence on children and young people: A review of the literature. Child Abuse & Neglect, 32, 797-810.

35.

Human Rights Watch. (2015, October 28). “Call me when he tried to kill you”: State response to domestic violence in Kyrgyzstan. Retrieved from https://www.hrw.org/report/2015/10/28/call-me-when-he-tries-kill-you/state-response-domestic-violence-kyrgyzstan

36.

Jewkes

(2002). Intimate partner violence: Causes and prevention. The Lancet, 359, 1423-1429.

37.

Johnson

Ollus

Nevala

(2008). Violence against women: An international perspective. New York, NY: Springer.

38.

Johnson

M. P.

(2005). Domestic violence: It’s not about gender—Or is it? Journal of Marriage and Family, 67, 1126-1130.

39.

Lawoko

Dalal

Jiayou

Jansson

(2007). Social inequalities in intimate partner violence: A study of women in Kenya. Violence & Victims, 22, 773-784.

40.

Maas

C. J.

Hox

J. J.

(2005). Sufficient sample sizes for multilevel modeling. Methodology: European Journal of Research Methods for the Behavioral and Social Sciences, 1(3), 86-92.

41.

Mann

J. R.

Takyi

B. K.

(2009). Autonomy, dependence or culture: Examining the impact of resources and socio-cultural processes on attitudes towards intimate partner violence in Ghana, Africa. Journal of Family Violence, 24, 323-335.

42.

Marzana

Vecina

M. L.

Alfieri

(2016). The morality of men convicted of domestic violence: How it supports the maintenance of the moral self-concept. Violence & Victims, 31, 1155-1170.

43.

O’Leary

K. D.

Tintle

Bromet

E. J.

Gluzman

S. F.

(2008). Descriptive epidemiology of intimate partner aggression in Ukraine. Social Psychiatry and Psychiatric Epidemiology, 43, 619-626.

44.

Pan

H. S.

Neidig

P. H.

O’Leary

K. D.

(1994). Predicting mild and severe husband-to-wife physical aggression. Journal of Consulting and Clinical Psychology, 62, 975-981.

45.

Paternoster

Brame

Mazerolle

Piquero

(1998). Using the correct statistical test for the equality of regression coefficients. Criminology, 36, 859-866.

46.

Pomerleau

McKee

Rose

Haerpfer

C. W.

Rotman

Tumanov

(2008). Hazardous alcohol drinking in the former Soviet union: A cross-sectional study of eight countries. Alcohol and Alcoholism, 43, 351-359.

47.

Rennison

C. M.

Welchans

(2000). Intimate partner violence. Retrieved from http://www.popcenter.org/problems/domestic_violence/PDFs/Rennison%26Welchans_2000.pdf

48.

Rutstein

S. O.

Johnson

(2004). The DHS Wealth Index (DHS Comparative Reports No. 6). Calverton, MD: ORC Macro. Retrieved from http://dhsprogram.com/pubs/pdf/CR6/CR6.pdf

49.

Social Institutions and Gender Index. (2017). 2014 results. Retrieved from http://www.genderindex.org/ranking/

50.

Stark

(2007). Coercive control: How men entrap women in personal life. New York, NY: Oxford University Press.

51.

Stickley

Leinsalu

Andreev

Razvodovsky

Vagerö

McKee

(2007). Alcohol poisoning in Russia and the countries in the European part of the former Soviet union, 1970–2002. European Journal of Public Health, 17, 444-449.

52.

Stöckl

Devries

Rotstein

Abrahams

Campbell

Watts

Moreno

C. G.

(2013). The global prevalence of intimate partner homicide: A systematic review. The Lancet, 382, 859-865.

53.

Tanha

Beck

C. J.

Figueredo

A. J.

Raghavan

(2010). Sex differences in intimate partner violence and the use of coercive control as a motivational factor for intimate partner violence. Journal of Interpersonal Violence, 25, 1836-1854.

54.

Treml

V. G.

(1997). Soviet and Russian statistics on alcohol consumption and abuse. In National Research Council (Ed.) Premature death in the new independent states (pp. 220-238) Washington, DC: National Academic Press.

55.

Uthman

O. A.

Moradi

Lawoko

(2009). The independent contribution of individual-, neighbourhood-, and country-level socioeconomic position on attitudes towards intimate partner violence against women in sub-Saharan Africa: A multilevel model of direct and moderating effects. Social Science & Medicine, 68, 1801-1809.

56.

Vecina

M. L.

Marzana

Paruzel-Czachura

(2015). Connections between moral psychology and intimate partner violence: Can IPV be read through moral psychology? Aggression and Violent Behavior, 22, 120-127.

57.

World Health Organization. (2000). Violence against women: Report on the 1999 WHO Pilot Survey in Tajikistan. Retrieved from http://www.who.int/violence_injury_prevention/media/en/150.pdf

58.

World Health Organization. (2002). World report on violence and health. Geneva, Switzerland: Author.

59.

World Health Organization. (2012). Understanding and addressing violence against women: Intimate partner violence. Retrieved from http://apps.who.int/iris/bitstream/10665/77432/1/WHO_RHR_12.36_eng.pdf

60.

World Health Organization. (2013). Global and regional estimates of violence against women: Prevalence and health effects of intimate partner violence and non-partner sexual violence. Retrieved from http://apps.who.int/iris/bitstream/10665/85239/1/9789241564625_eng.pdf

61.

World Health Organization. (2014). Violence against women: Intimate partner and sexual violence against women. Retrieved from http://www.who.int/mediacentre/factsheets/fs239/en/

62.

Yount

K. M.

Halim

Hynes

Hillman

E. R.

(2011). Response effects to attitudinal questions about domestic violence against women: A comparative perspective. Social Science Research, 40, 873-884.