Who Supports Honor-Based Violence in the Middle East? Findings From a National Survey of Kuwait

Abstract

This study reports findings from what is, to the authors’ knowledge, the first nationally representative survey of public attitudes toward honor-based violence (HBV) in a Middle East country affected by such violence, and one of very few if any such national surveys undertaken in any context. In a telephone survey of 1,050 adult citizens of the Arab Gulf state of Kuwait, we collect attitudinal and behavioral measures of support for HBV. Facilitated by vignettes, item order randomization, and other survey best practices designed to reduce social desirability bias, the survey measures both support for HBV in principle and support for legislation permitting it. We also assess the correlates of views toward honor-based crimes, including demographic attributes (age, sex, and education), social indicators (economic status, religiosity, and tribal attachment), and political indicators (political interest, engagement, and ideology). These unique data permit investigation into the individual-level predictors of support for HBV among citizens of a country that has experience with HBV and laws that afford reduced punishment for perpetrators of honor killing. Results reveal, first, considerable levels of public support in Kuwait for HBV. Half of citizens report agreement that physical violence is justified as punishment for female adultery, whereas a third of both men and women say they would support a law permitting violence against a female adulterer. The strongest positive predictors of individual support for HBV are tribal attachment, religiosity, and support for political Islam. Surprisingly perhaps, permissive views of HBV are most likely among the youngest generation of Kuwaitis. The findings have important implications for understanding the factors that help explain the existence and persistence of HBV in the Middle East, in Islamic contexts, and elsewhere where it occurs.

Keywords

honor-based violence honor killing Middle East and North Africa public opinion survey research

The Middle East and North Africa (MENA) is commonly identified as the geographical and cultural epicenter of honor killing, murder committed by a male family member for a woman’s perceived transgression of gender norms. Yet, the problem of honor killing and other forms of honor-based violence (HBV) in MENA societies, and elsewhere where it occurs, has been compounded by an acute lack of data with which to study the phenomenon. A paucity of official statistics and survey data continue to obscure its magnitude, prevalence over time, and demographic, attitudinal, and behavioral correlates. Indeed, a 2011 comprehensive review identifies a mere nine studies, or fewer than a quarter of publications on Middle East honor killing, that utilize primary data, and only 20 works that offer original analysis of any kind (Kulczycki & Windle, 2011).

Such data limitations owe both to the nature of HBV and to the contexts where it tends to occur. Weak administrative capacity in many Middle East states (Ayubi, 1996; United Nations Development Program, 2004) means that cases of honor-related violence reported to authorities may not be recorded or collected in a database, whether for incompetence, indifference, or lack of technical capability. The politically closed regimes of the MENA region also generally limit the availability of governance-related data, and they are particularly reluctant to report statistics that might be believed to harm their international reputation (Tannenberg, 2017), such as in the case of honor-related crime. Restricted press freedom in many MENA contexts similarly limits the viability of data collection through media sources (Thuroczy, 2010).

Meanwhile, at the individual level, the intimate and familial nature of honor-related violence has been blamed for chronic underreporting by victims (Rafferty, 2013). Lack of confidence in law enforcement and justice systems may also dampen incentives to report honor-related crimes in MENA societies with weak state institutions (United Nations Development Program, 2004). Finally, social sensitivity surrounding HBV complicates efforts by researchers to collect reliable independent data through fieldwork or surveys. This includes information on individual experience with as well as attitudes toward HBV.

The result is that extant empirical investigations of honor crime have had to rely on data with acknowledged methodological limitations with respect to sample size, representativeness, mode of collection, and the scope of information collected (Kulczycki & Windle, 2011). In the case of survey-based research in particular, such shortcomings reflect a basic tradeoff: smaller studies with limited target populations may permit more direct questions about HBV, but at the expense of generalizability; whereas large-N questionnaires can capture the views and experiences of a wider population, but for practical or political reasons may be unable to pose sensitive questions about HBV within the framework of a general public opinion survey.

The current study employs a survey intended to bridge this tradeoff between specificity and generalizability. In a nationally representative telephone survey of 1,050 adult citizens of the Arab Gulf state of Kuwait, we collect attitudinal and behavioral measures of support for HBV. Facilitated by vignettes, item order randomization, and other best practices designed to reduce social desirability bias, the survey captured both support for HBV in principle and support for legislation permitting it. We also assess the correlates of views toward HBV, including both social variables such as economic status, religiosity, and tribal attachment; and political indicators such as political interest, engagement, and ideology. These unique data permit investigation into the individual-level predictors of support for HBV among citizens of a country that has experience with HBV and laws that offer reduced punishment for perpetrators of honor crimes. To the authors’ knowledge, the study is the first national-level investigation into of the nature and correlates of attitudes toward HBV.

HBV in the Middle East

In societies where a family’s honor is perceived as important, failure to follow expected gender norms can result in shame to both male and female members of a family (Caffaro, Mulas, & Schmidt, 2016). Honor killing refers to a family’s attempt to restore honor by enlisting a male relative to kill a female who has brought shame to the family (Kulczycki & Windle, 2011). Such crimes are rooted in economic and legal systems that interpret women as the property of males through inheritance laws and by other means that encourage social policing of female behavior (Kulczycki & Windle, 2011). Here, we take a broad view of HBV that includes not only murder but any violent act understood by the perpetrator as “heroic” and/or “a fulfillment of a religious obligation,” rather than merely punitive (Chesler, 2009, p. 64). We thus follow Terman (2010) in making a conceptual distinction between HBV and domestic violence.

Although HBV is a global phenomenon, scholars have long noted its disproportionate prevalence in MENA and other Islamic societies. The United Nations estimates that around 5,000 women are killed annually in honor-related crimes, with some contending that the true number is several times higher (Chesler, 2010; Rafferty, 2013). In a review of 172 honor killings reported in English-language media between 1989 and 2009, Chesler (2010) finds that more than half occurred in Muslim-majority countries, and 91% of perpetrators were Muslims. While acknowledging the problem of honor killing in other traditions such as Hinduism and Sikhism, Chesler (2010) goes so far as to conclude on this basis that “honor killings, both worldwide and in the West, are mainly Muslim-on-Muslim crimes.” In the MENA, the greatest number of incidents are reported in Egypt, Jordan, Israel, Morocco, and Turkey (Rafferty, 2013). But what explains why some ordinary men and women in MENA countries condone violence, sometimes deadly, against female members of their own societies?

Terman (2010) identifies four characteristics of an honor killing. First, the murder is carried out to restore honor to a collective. Second, the perpetrator is encouraged by others to commit the murder. Third, no evidence is necessary to confirm or deny a suspicion. Finally, the murder is premeditated. Given its collective nature, explanations for the high incidence of HBV in MENA and other Muslim countries have focused on institutions and practices that encourage collective enforcement of social norms and underlie the notion of “honor” that perpetrators of HBV ostensibly aim to protect. In particular, previous studies have highlighted the role of two sociopolitical institutions with strong salience in most MENA countries: Islam and descent-based groupings such as religious denomination and tribe. An additional institution that bears consideration, but which has garnered relatively less attention, is an Arab criminal code inherited mainly from religious texts as well as Ottoman and European colonial powers, some of whose provisions have proven permissive of or even compatible with HBV.

The enduring societal salience of Islam and traditional social units such as tribe and sect in MENA society has been used to help explain prevalence of HBV, due to their role in the creation and collective enforcement of patriarchal social norms (Al-Nakib, 2015). Although most agree that HBV is “by no means exclusively Muslim” (Payton, 2014, p. 2866), scholars have documented how perpetrators and their families may use religion to justify honor-related violence (Al Gharaibeh, 2016; Kulczycki & Windle, 2011). Others have elucidated the key role of Islamist political movements in resisting legal punishment of HBV (Herb, 2004). And, on the side of religious leaders, Idriss’ (2020) study of Muslims in the United Kingdom concluded that some mosques shrink from addressing violence against women, including HBV, because it may threaten their patriarchic control over worshippers (p. 26).

Other research, however, makes clear that adherence to Islam is not a monolithic predictor of societal or individual attitudes toward HBV. In Jordan, when Islamists have justified honor killings through strict religious interpretations, the country’s religious establishment has issued contradictory statements condemning such acts (Kulczycki & Windle, 2011). A similar disagreement manifests in Pakistan, where some religious movements have led efforts to change adultery laws seen as permissive of HBV (Khan, 2003). Thus, Islam can either tolerate or condemn honor crimes depending on the interpreter.

Meanwhile, in most MENA states, descent-based groupings such as tribe and religious sect possess and enforce their own norms governing personal conduct (Al Gharaibeh, 2016; Charrad, 2009). These collectives represent additional layers of social scrutiny of female behavior as well as additional communal identities whose honor members may feel is at stake. Al Gharaibeh (2016, p. 123) suggests that tribalism “may be at the root of a culture” of HBV in the Arab world, by elevating the preservation of a woman’s virginity, institutionalizing the practice of arranged marriage, and facilitating collective enforcement of violations of family and tribal honor. Tribal and sectarian considerations may also influence retributions relating to honor (Carroll, 2011).

If religion and tribalism have important roles in the creation and promulgation of norms supporting HBV, legal structures in the Middle East serve as often to reinforce as to discourage honor crimes. The penal codes of many Middle Eastern nations continue to be influenced by religious texts enshrining the dominance of males (Caffaro et al., 2016) as well as 19th-century laws making allowances for crimes of passion inherited from colonial rule (Liebesny, 1956). The latter provide protection to the perpetrators of honor killings, despite being long abolished in their places of origin (Faqir, 2001). States whose penal codes retain these provisions include Lebanon, Jordan, Syria, Oman, and Kuwait (Moghaziel, 2000). Article 153 of Kuwait’s penal code, for example, provides relative impunity to a man who commits murder in an honor-related incident. The law, which stipulates punishments for a host of offenses ranging from crimes of passion to communication (Duffy & Alkazemi, 2017), was put into practice before the country’s independence from Great Britain in 1961—remnants of an old Arab legal system that developed over time (Liebesny, 1956). Its provisions state as follows:

Anyone who surprises his wife in a state of adultery, or surprises his daughter or mother or his sister in the presence of a man and kills her immediately or kills him or kills them together is punishable by imprisonment for a period not exceeding three years and a fine of not more than 3000 rupees [~45 USD] or one of these two punishments. (Advocacy and Lobbying, 2015)

Other aspects of MENA penal codes legitimize HBV in a different fashion, by setting punishments based on the degree of putative “honor” at stake for a victim. In Lebanon, for example, the law stipulates a harsher sentence if the victim is a virgin (Rebeiz & Harb, 2010).

In this way, the foregoing institutions—Islam, tribal and other kinship groups, and permissive laws—have been offered as key factors that help account for the disproportionate prevalence of HBV in the MENA. Yet, though this literature does help identify societal-level variables that may be correlated with HBV, it cannot directly answer the question of what makes a person more or less likely to engage in or support honor crimes. Such acts are committed by but a tiny minority of people, whereas the vast majority of a population, despite being Muslims or belonging to a tribe or being subject to the same legal incentives, does not attack their female relatives for violating gender norms. What explains this variation? To understand the factors associated with support for HBV, one must also observe negative outcomes in which HBV does not occur or is not considered acceptable. This is most easily achieved at the individual level of analysis.

The Case of Kuwait

Here, we study individual views toward HBV using the case of Kuwait. Kuwait, an oil-exporting Arab Gulf monarchy situated between Iraq and Saudi Arabia, is both an appropriate and innovative case in which to study the individual-level predictors of support for HBV. The country is representative of the societal contradictions and disagreements discussed in the foregoing section. Within the notably authoritarian Arab Gulf region, Kuwait stands out for the social and political freedoms its citizens enjoy. Kuwait was the first Arab Gulf state to establish a constitution and parliament, its National Assembly existing, with some interruptions, since independence from Britain in 1962 (Meyer, Rizzo, & Ali, 2007). Today, it remains the Gulf region’s only functional constitutional monarchy. The Kuwaiti media also enjoy more freedom than those in any other Arab Gulf state (Duffy & Alkazemi, 2017; Selvik, 2011), social media use being particularly high (Aladwani, 2015; Dashti, 2009). But Kuwait also is a society in which tribes and Islamist groups hold considerable social and political sway, both as allies of the government in parliament and more often in opposition (Ibrahim, 1993).

The history of female empowerment in Kuwait reflects this duality. When the parliament was reopened in 1999 following Kuwait’s liberation from Iraqi occupation during the First Gulf War, some political activists resisted the enfranchisement of women (Meyer et al., 2007). Islamists and blocs representing tribal constituencies were among the strongest opponents (Meyer et al., 2007; Tétreault, 2001), and ultimately proposals to grant women the right to vote in 1999 were defeated in parliament (Herb, 2002; Meyer et al., 2007). Only in 2005 did Kuwaiti women earn the right to vote and run for office, and four women became the country’s first parliamentarians (Al-Nakib, 2012). The next decade would witness similar jumps and starts. In 2009, Kuwaiti courts ruled as unconstitutional the requirement that a male guardian be present for a woman to obtain a passport (“Kuwaiti Women Win Right to Obtain Passport Without Husband’s Approval,” 2009). Yet, a December 2015 lawsuit challenging the constitutionality of gender segregation in universities failed to overturn legislation requiring the separation of genders (Ahmed, 2015). Thus, it is unclear precisely how much support exists for female empowerment in Kuwait, at either the elite or popular level.

Kuwait also has experience with HBV. The issue gained attention most recently in 2014, when a British newspaper published an article about a Kuwaiti man who stabbed his 24-year-old daughter to death (Pleasance, 2014). According to the report, the father claimed that his daughter was acting inappropriately within his culture and that her death restored his honor (Pleasance, 2014). A year later, a Bahraini newspaper detailed the case of a Kuwaiti woman who had been stabbed by her brother for her desire to live independently of her extended family (Toumi, 2015). The article recounts that the siblings’ hysterical father delivered his blood-covered daughter to a hospital and reported the incident to police (Toumi, 2015).

Method and Data

We investigate public attitudes toward HBV in Kuwait using an original, nationally representative opinion survey. The survey was conducted by telephone in July 2016 with a total of 1,050 adult citizens aged 21 years and above. Respondents were selected randomly via random digit dialing from a comprehensive national frame of cellular subscribers, with approximately 98% coverage of adult Kuwaitis. Kuwait, together with the other Arab Gulf countries, has one of the highest mobile phone penetrations in the world, making a cellular frame a common and valid means of probability sampling (Gengler, Le, & Howell, 2018). The response rate for the survey was 43.7%, with a sampling error of 3.2%. The survey was carried out by a professional and reputable polling firm with a history of administering social scientific surveys in Kuwait.

Survey Design

In view of the sensitive nature of honor violence, the survey employed several strategies aimed at mitigating social desirability bias (see Krumpal, 2013; Nederhof, 1985, for reviews)—that is, the tendency of survey respondents to report attitudes and behaviors that they perceive to be socially acceptable. In this case, it was expected that contested norms surrounding domestic abuse and women’s rights may make some Kuwaitis reluctant to express support for honor violence despite agreeing with it privately. The survey was designed to mitigate this, first, by introducing the questionnaire¹ much more broadly as an investigation into citizens’ opinions on “important social issues.” The interview schedule began with general questions about economic and social priorities for the country; it then asked about concerns facing Kuwaiti families specifically; and it concluded with questions about various sociopolitical opinions and behaviors, including respondents’ views about honor violence. In total, the average interview lasted less than 10 minutes, and we expect that the precise purpose of the survey—to assess public attitudes toward honor violence—would not have been obvious to most respondents.

The second strategy meant to reduce socially desirable reporting in the survey was to embed the most sensitive questions about gender attitudes and honor violence in multi-item batteries introduced by vignettes (Gengler & Mitchell, 2018; King, Murray, Salomon, & Tandon, 2004), in contrast to traditional, direct solicitations of opinion. This approach helps obscure the aim of a question while also depersonalizing the response by invoking a hypothetical scenario. Thus, for example, we gauge support for legislation permitting honor violence by asking,

Now please imagine that Kuwaiti lawmakers were debating different laws about the status of women in the country. I am going to list a series of laws. For each, please tell me whether you would strongly support, somewhat support, somewhat oppose, or strongly oppose the adoption of such a law in Kuwait.

A. A law barring women from running for parliament

B. A law allowing a woman to divorce her husband without his consent

C. A law giving male students priority over females in university admissions

D. A law permitting violence against a woman found in the act of adultery

A final and critical element of this technique is that the order of the individual items within the battery (here A through D) is randomized at the respondent level. This eliminates possible response-order effects that might result from a sensitive item appearing earlier or later in a battery (see, for example, Krosnick & Alwin, 1987; McFarland, 1981).

In conclusion, the nationally representative character of our survey is an important methodological advance over previous survey-based studies, allowing generalization about the views of a society in general, rather than about the views of narrower population groups such as, for example, adolescents (Eisner & Ghuneim, 2013), university students (Caffaro, Ferraris, & Schmidt, 2014; Caffaro et al., 2016), academics (Çalik, 2017), or victims of gender-based violence (Sheridan, Scott, & Roberts, 2015). However, another important innovation of our study is that it measures attitudes toward honor violence not just more broadly than before, but also in a manner that anticipates and takes steps to mitigate the response bias inherent in investigation of such a sensitive topic.

Model and Indicators

The analysis to follow focuses on two related but separate dependent variables measuring individual support for honor violence. The first, an attitudinal measure, captures agreement or disagreement in principle that violence against a woman who has committed adultery is justified. The survey asks,

The following questions are about your personal opinions about the principles that should determine the behavior and situation of women in our society. For each of the statements listed below, please indicate whether you strongly agree, agree, disagree, or strongly disagree:

A. “It is justified to use violence against a woman who has committed adultery.”

B. “A woman can become president or prime minister of a Muslim country.”

C. “A woman can reject a marriage partner that her family chose without her consent.”

D. “University education for males is more important than for females.”

The three other statements in this battery (B-D), the order of which is randomized, originate from the gender attitudes section of the widely used Arab Barometer survey (Tessler et al., 2016).

The second dependent variable, discussed already in the previous section, is a behavioral measure of approval of honor violence: support for a hypothetical law permitting violence against a woman found in the act of adultery. Both variables are coded on a 4-point categorical scale from “strongly disagree” to “strongly agree” and from “strongly oppose” to “strongly support,” respectively. The rationale for examining both types of variables is that individuals may condone honor violence personally but not wish to legislate it for society as a whole, and similarly that the correlates of support for honor violence in the abstract may be different than correlates of support for a law permitting it.

We predict individual support for HBV as a function of demographic, economic, social, and political indicators. Because our dependent variables are ordered choices, we estimate support using ordered logistic regression. The first category of explanatory variables includes gender, age, education, and marital status. Gender is a binary variable; years of age is coded as continuous; education level is coarsened into six categories (primary and below, secondary, vocational, 2-3 year diploma, bachelor’s degree, and graduate degree); and marital status is coded 1 for ever-married respondents and 0 otherwise. We also include in the model the square of age (age x age) to account for possible nonlinearity in the relationship between age and support for honor violence.

Economic status is measured as a self-assessment of the respondent’s household financial situation, on a 5-point scale from “very weak” to “very good.”

Two indicators reflect principal dimensions of social identity in Kuwait: tribal affiliation and religiosity. The former is coded as the response to the question, “For you personally, how important is tribe and tribal belonging?” Answers are categorical and range from “not at all important” to “very important.” Religiosity is a dichotomous variable coded 1 if a respondent cites “lack of prayer” as the primary obstacle to accepting a marriage proposal for her son, daughter, sister, or brother, and 0 otherwise. This is a common indirect method of measuring religiosity in MENA surveys (e.g., Tessler et al., 2016), used because it does not require respondents to self-identify as “religious” or “not religious” or to self-report observance of religious duties such as prayer. Such self-assessments may entail social desirability bias (see Gengler, 2015, pp. 120-121).

Political indicators include interest in politics on a 4-point categorical scale from "not at all interested" to “very interested,” and political participation as measured by self-reported frequency of voting in parliamentary elections from "never" to “always.” An additional indicator captures support for political Islam and is measured as the extent of agreement or disagreement with the statement, “The government and parliament should enact laws in accordance with the Islamic shari‘a.”

Descriptive statistics of the variables used in the analysis are given in the appendix.

Findings

Overall Distributions of Attitudes Toward Honor Violence

The results of the survey reveal a substantial level of approval of honor violence within the Kuwaiti population both in principle and in terms of legislation. More than half of Kuwaitis agree that physical violence is justified as punishment for female adultery, with 33% of males and 28% of females reporting “strong” agreement. Meanwhile, only 25% of men and 24% of women in Kuwait strongly disagree that violence against a woman is justified in such a case. Figure 1 depicts these results. As shown in Figure 2, a smaller but still considerable proportion of Kuwaitis support a law legalizing violence against a woman found in the act of adultery, with a combined 33% of men and 37% of women, respectively, saying they would support or strongly support such a law. However, the most common survey response to the hypothetical legislation is “strong” opposition, representing 39% of Kuwaiti male and 36% of female responses. Notably, nonresponses—“don’t know” and refusals—account for only around 5% of answers to each question. Combined with the significant number of citizens who voice either somewhat or very supportive views toward honor violence, this suggests that the data are not unduly affected by underreporting due to social desirability bias.

Figure 1.

Attitudinal support for honor-based violence in Kuwait, by gender.

Figure 2.

Behavioral support for honor-based violence in Kuwait, by gender.

It is also instructive to note that neither the attitudinal nor the behavior measure of support for honor violence is strongly related to more general orientations toward gender equality in Kuwait. The former is correlated at only r = .15 with a simple additive index of the three other survey items capturing overall gender attitudes (see page 10), and its highest correlation with any single item is r = .16. The behavioral measure of support for honor violence is correlated similarly at the r = .16 level with the three-item index of gender attitudes, with a maximum single-item correlation of r = .19. These findings suggest that in Kuwait and possibly elsewhere, broad orientations toward gender relations have minimal connection to views regarding honor violence specifically. This may have implications for studies that use general attitudinal proxies to draw conclusions about views toward HBV.

Predicting Support for Honor Violence: Multivariate Results

Perhaps surprisingly, the response distributions of the two measures of support for honor violence do not suggest a substantive disparity in attitudes according to respondent gender. One potential explanation is that gender-based differences in other important demographic, attitudinal, or behavioral predictors of support obfuscate the true relationship between gender and approval of honor violence. To test this possibility, and to assess more generally the individual-level correlates of support for honor violence in Kuwait, we now present the results of our multivariate analysis.

Given in Tables 1 and 2 are the estimated associations between support and the several explanatory variables discussed in previous sections. In the models reported in Table 1, support is operationalized in attitudinal terms as approval of honor violence in principle, whereas in the models of Table 2, support is measured behaviorally as support for a law that would authorize honor violence. Predictors are introduced into the model sequentially so as to show how coefficient and error estimates change with the introduction of additional regressors. Thus, the baseline Model 1 includes only the demographic variables of gender, marital status, age, and education level. Model 2 adds the indicator capturing economic status. Model 3 adds the identity-related factors of religiosity and tribal belonging. Model 4 introduces the three political explanators: interest in politics, participation in politics, and support for political Islam. Finally, Model 5 omits the insignificant political interest and participation variables to regain observations lost due to missing data (around 10% of respondents report being ineligible to vote) and thereby increase the statistical confidence of the remaining estimates. Note that variables are kept in their native metrics, and so the magnitude of coefficients cannot be compared directly. A comparison of substantive marginal effects of individual regressors follows later in Table 3.

Table 1.

Predictors of Attitudinal Support for Honor Violence in Kuwait.

	(1)	(2)	(3)	(4)	(5)
	Approval	Approval	Approval	Approval	Approval
Female	−0.104 (.377)	−0.0862 (.467)	−0.104 (.384)	−0.0667 (.610)	−0.117 (.332)
Married	0.301* (.026)	0.295* (.030)	0.262^† (.055)	0.241 (.104)	0.212 (.128)
Age	−0.0982** (.002)	−0.0978** (.003)	−0.0888** (.005)	−0.0810* (.015)	−0.0826** (.009)
Age²	0.00103** (.006)	0.00103** (.007)	0.000988** (.009)	0.000884* (.023)	0.000908* (.015)
Education	−0.0934* (.015)	−0.0916* (.020)	−0.0715^† (.092)	−0.0575 (.223)	−0.0552 (.201)
Economy		−0.0248(.633)	−0.0173(.746)	0.0168(.776)	−0.0214(.693)
Religiosity			0.329** (.008)	0.231^† (.088)	0.240^† (.060)
Tribalism			0.152** (.007)	0.135* (.026)	0.139* (.016)
Political interest				−0.0393(.592)
Political participation				0.00117(.984)
Political Islam				0.235** (.006)	0.227** (.005)
Cut 1	−3.351*** (.000)	−3.405*** (.000)	−2.600*** (.000)	−1.713* (.029)	−1.805* (.014)
Cut 2	−2.320*** (.000)	−2.382*** (.000)	−1.543* (.027)	−0.627(.425)	−0.748(.306)
Cut 3	−1.539* (.016)	−1.594* (.018)	−0.745(.285)	0.146(.853)	0.0587(.936)
N	967	958	942	829	923
pseudo R²	.007	.007	.013	.016	.016

Note. p values in parentheses; robust standard errors utilized.

†

p < .10. *p < .05. **p < .01. ***p < .001.

Table 2.

Predictors of Behavioral Support for Honor Violence in Kuwait.

	(1)	(2)	(3)	(4)	(5)
	Support	Support	Support	Support	Support
Female	0.0592(.619)	0.0780(.516)	0.0917(.448)	0.0239(.855)	0.0561(.646)
Married	0.262^† (.057)	0.260^† (.059)	0.230^† (.092)	0.307* (.037)	0.227(.104)
Age	−0.117*** (.000)	−0.120*** (.000)	−0.118*** (.001)	−0.122*** (.001)	−0.136*** (.000)
Age²	0.00112** (.005)	0.00115** (.004)	0.00119** (.005)	0.00129** (.003)	0.00142*** (.001)
Education	−0.131*** (.001)	−0.124** (.002)	−0.0999* (.021)	−0.0738(.118)	−0.0925* (.036)
Economy		−0.0617(.245)	−0.0528(.340)	−0.0545(.365)	−0.0579(.306)
Religiosity			0.574*** (.000)	0.442** (.001)	0.462*** (.000)
Tribalism			0.170** (.004)	0.112^† (.077)	0.138* (.021)
Political interest				−0.0317(.659)
Political participation				−0.0356(.534)
Political Islam				0.337*** (.000)	0.315*** (.000)
Cut 1	−3.356*** (.000)	−3.583*** (.000)	−2.729*** (.000)	−1.916* (.015)	−2.154** (.004)
Cut 2	−2.389*** (.000)	−2.626*** (.000)	−1.741* (.016)	−0.906(.251)	−1.156(.121)
Cut 3	−1.721** (.008)	−1.954** (.005)	−1.051(.144)	−0.161(.839)	−0.443(.554)
N	964	957	942	835	924
pseudo R²	.014	.014	.027	.031	.034

Note. p values in parentheses; robust standard errors utilized.

†

p < .10. *p < .05. **p < .01. ***p < .001.

Table 3.

Marginal Effects on Support for Honor Violence in Kuwait.

	Attitudinal			Behavioral
	Low	High	% Diff.	Low	High	% Diff.
Demographic variables
Female	−	−	−	−	−	−
Age	36%	28%	−20%*	26%	17%	−37%***
Economy	−	−	−	−	−	−
Education	−	−	−	21%	16%	−21%*
Married	−	−	−	16%	19%	+20%^†
Explanatory variables
Religiosity	27%	32%	+18%^†	16%	23%	+45%***
Tribalism	26%	32%	+25%*	16%	21%	+29%*
Political Islam	25%	32%	+27%**	15%	22%	+47%***
Combined explanatory
Religiosity, tribalism, and political Islam	21%	39%	+88%***	11%	30%	+172%***

Note. “Low” and “High” columns report the predicted probability of strong support for honor violence when an explanatory variable equals −1 and +1 standard deviation from the mean, respectively. Predicted probabilities are calculated at the mean values of all other regressors. The “% Diff.” column reports the relative percent change between the “Low” and “High” columns as well as the statistical significance of this change. Robust standard errors utilized.

†

p < .10. *p < .05. **p < .01. ***p < .001.

The two sets of estimations offer a quite clear and consistent picture of the factors related to greater support for honor violence among Kuwaitis. First, the strongest statistical predictor of support in the both attitudinal and in the behavioral model (using the final Model 5 results) is endorsement of political Islam. Meanwhile, other identity-related factors, namely personal piety and tribal identification, are independently associated with support for honor violence both in principle and in the legal domain. The other political variables, measuring interest and participation, respectively, are never linked to attitudes about honor violence, and neither is a person’s household economy. Among demographic characteristics, only age and, in the behavioral model, education level are robust predictors. Age is related to support for honor violence in a curvilinear fashion, with support being highest among younger Kuwaitis. More educated Kuwaitis are less likely to support legislation permitting honor violence, and the estimated coefficient on education is also negative in the attitudinal model but is not statistically significant.

Other demographic characteristics, such as marital status, appear as significant predictors in Models 1 and 2, but this statistical significance dissipates after additional indicators capturing social and political orientations are included. Thus, though it is true that support for honor violence is higher among married and more educated Kuwaitis, these associations exist due to links between these demographic categories and other explanatory variables, namely religiosity, tribalism, and support for political Islam. According to Models 1 and 2, for example, respondent education is a significant predictor of attitudinal support for honor violence at the standard p < .05 level. However, this is because education level is (negatively) correlated with tribalism (r = ‒.10) and political Islam (r = ‒.16). When the latter variables are included in the model, the results no longer evidence an independent effect of education on attitudinal support for honor violence. Similarly, married Kuwaitis are more likely than unmarried individuals to be religious and to ascribe to political Islam, so the significance of marriage in Models 1 and 2 gradually fades as other regressors are added. If the conceptual aim is to identify the basic demographic categories for which support for honor violence among Kuwaitis is highest, therefore, the results of Models 1 and 2 are valid and instructive. But if the purpose is to attempt to draw conclusions about why a person might support violence against a woman perceived to have transgressed gender norms, then it is more important to understand the relationships with underlying social and political explanators included in later models. For the latter type of investigation, it is the independent rather than absolute explanatory power of demographic variables that is of primary interest.

But what do the model estimates reported in Tables 1 and 2 mean in practical terms? And how can one assess the relative importance of the different independent variables? Summarized in Table 3 are the marginal effects of the respective demographic and explanatory variables, expressed in a manner that is substantively meaningful. Effects are evaluated as the difference in predicted support for honor violence attributable to an individual possessing a “low” value of a particular independent variable—say, low education—compared with having a “high” value of that variable. A low value is defined as 1 standard deviation below the mean value of that variable, and a high value is defined as 1 standard deviation above the mean value. Binary indicators—that is, female, married, and religiosity—are evaluated at 0 and 1, respectively. This represents a comparison that is both realistic and easily interpretable. Marginal effects for the attitudinal and behavioral models are given side-by-side. For each model, the column labeled “Low” gives the estimated likelihood of the highest category of support for honor violence when the independent variable in question has a “low” value (–1 standard deviation, or “no” for binary variables), and the “High” column gives the estimated likelihood when that variable has a “high” value (+1 standard deviation, or “yes” for binary variables). The disparity between these two predicted probabilities is given as a relative percentage change in the column labeled “% Diff.” This difference can be considered the practical marginal effect of an independent variable.

The percentages reported in Table 3 add substance to the statistical associations reviewed already. A religious individual is 18% more likely to express strong attitudinal support for honor violence, all else equal, as compared with a nonreligious person. A Kuwaiti with a strong affinity to his or her tribal identity is 25% more likely to support honor violence in principle, relative to a Kuwaiti with weak tribal identification. A respondent who advocates the mixture of religion and politics is 27% more likely to report attitudinal support for honor violence. And finally, a person who combines all three characteristics—high religiosity, high tribalism, and high support for political Islam—is a notable 88% more likely to express strong support for violence as punishment for a woman adulterer, compared with a person who possesses low values of these three variables.

Marginal effects on behavioral support for honor violence are notably exaggerated relative to the effects of independent variables in the attitudinal model, and moreover education and marital status are associated with higher support only when it is measured in behavioral terms. The percent change attributable to religiosity increases from 18% in the attitudinal model to 45% in the behavioral model, for example, whereas the impact of endorsement of political Islam grows from 27% to 47%. The effect of age also nearly doubles, from 20% to 37%. Finally, the combined effect of the three explanatory variables increases from 88% to a remarkable 172% in the predicted likelihood of support for honor violence. This amplification reflects the nature of the behavioral versus the attitudinal measure of support for honor violence, the former requiring a respondent to imagine engaging in an actual supportive behavior rather than simply report a supportive attitude. Such a result is consistent with prior research that has found behavioral survey indicators to possess greater validity by imposing a (hypothetical) cost to respondents (Hainmueller, Hangartner, & Yamamoto, 2015).

A final result that warrants additional treatment is the considerable and in many ways unexpected impact of respondent age on support for honor violence. Because we model age nonlinearly, the marginal effects of age on attitudinal and behavioral support calculated in Table 3 are misleading, assuming as they do a constant effect across all values of the variable—that is, across all ages. In fact, as depicted in Figure 3, the relationship between age and support for honor violence is highly curvilinear, with support greatest among the youngest adult Kuwaitis and then dropping precipitously before leveling out at around the age of 40 years. Recall that this effect represents the independent impact of age, holding constant all other variables in the model. The shaded area shows the 95% confidence interval of the estimated relationship, shown as the black line. An 18-year-old citizen is more than 40% likely to support a law permitting honor violence, but by 25 years of age this estimated likelihood already reduces to below 30%. At the age of 30 years, an individual is 23% likely to be supportive, and at age of 40 years, the likelihood is a mere 18%. As indicated by the shaded confidence interval, support does not increase significantly after the age of 45 years.

Figure 3.

Predicted probability of behavioral support for honor-based violence in Kuwait, by age.

Substantively speaking, the picture of Figure 3 utterly contradicts the notion of a more liberal youthful generation of Kuwaitis in conflict with the more conservative views of older citizens. Instead, the attitudes toward HBV seem to liberalize around the time that individuals have sons and daughters of their own who are approaching the age of marriage. One might speculate that the middle and older generations of Kuwaitis are more apt than younger individuals to consider how they would wish their own children to behave in domestic life.

Discussion

Our analysis indentifies numerous demographic, social, and political factors that help explain individual attitudinal and behavioral support for HBV in Kuwait. The negative relationship between educational achievement and support for HBV matches previous findings (e.g., Caffaro et al., 2014; Cihangir, 2013). The result that married Kuwaitis are more likely to support a law enabling HBV than unmarried individuals is also perhaps not surprising. Abdalla (2015) cites unsupportive husbands as a main obstacle for social and economic advancement of Arab women. Citizenship in Kuwait is passed patrilineally, meaning that women who might wish to marry nonnationals from less patriarchic foreign cultures can neither pass their citizenship to the spouse or children nor keep the extensive social benefits provided by the government such as housing (AlNakib, 2015). This begets an enormous disincentive for females to marry outside of the Kuwaiti population. Thus, married individuals may be more supportive of HBV due to male-dominated gender norms reinforced by within-group marriage.

The nonlinear relationship between age and support for HBV admits of several interpretations. A first explanation concerns what some see as the increasing influence of conservative religious factions over the Kuwaiti school curriculum. Dutton, Bakhiet, Essa, Blahmar, and Hakami (2017) note increased classroom time devoted to religious education “at the expense of scientific subjects,” pushed by Salafist sympathizers within the Ministry of Education. This conservative faction supports, among other things, the segregation of men and women in classrooms (Wahid, 2014), a policy which Kuwait University implemented in 2001 (Algharabali, 2010).

Such segregation, both at Kuwait University and in public K-12 schools, suggests a second explanation for the higher support for HBV observed among younger Kuwaitis: lack of integration of male and female students may lead to unrealistic perceptions of gender roles. AlNakib (2015) observes that segregated education in Kuwait allows men “to ignore the reality of women’s experiences.” Because this segregation begins in primary school and continues through students’ university experiences (AlNakib, 2015), it may not be until individuals have completed their education, started families, and perhaps given birth to the next generation of Kuwaiti women that support for HBV decreases. The curvilinear relationship between support for HBV and age reflected in the findings may be explained in this light.

A third piece of the explanation relates to generational differences in education. Kuwait University is the most established university in Kuwait, and its College of Social Sciences was established only in 1998 (Al-Qimlass, 2015). Circa 1995, an estimated 21% of the Kuwaiti population was illiterate (Bahgat, 1999). Hence, the tail end of the curvilinear relationship between support for HBV and age, which shows some evidence of support increasing after the age of 50 years, can be explained with reference to elderly individuals who lacked access to formal education about sociological concepts, including the changing roles of women in modern Kuwaiti life.

Along with age, the social and political indicators of tribal identification, religiosity, and support for political Islam are the most robust predictors of both tacit and behavioral support for HBV among Kuwaiti men and women. Shultziner and Tétreault (2011) argue that progress toward female empowerment in MENA countries is viewed by opponents as moral decay. In this way, HBV can be understood as a mechanism by which some seek to restore sociological balance among people organized by patriarchal lineage and/or adherence to a common ethnical code (Terman, 2010). This moral dimension of HBV likely underlies support among individuals with strong attachment to traditional institutions such as religion and descent group.

Separate from an individual’s personal religiosity, support for political Islam—for the mixing of religion and politics—also predicts support for HBV. Although there is diversity within political Islam, in Kuwait Salafism represents the largest current, receiving a quarter of the votes for Kuwaiti parliament in the 2016 elections (Dutton et al., 2017). Other strains include Shi‘a Islamists and supporters of the Muslim Brotherhood. The finding that support for political Islam is inversely related to female welfare in Kuwait runs counter to some existing survey-based results. Blaydes (2014), for example, found that women living in Muslim Brotherhood-controlled areas of Egypt enjoyed better reproductive health than women living in other regions, which had higher rates of high school drop-out and female circumcision. Arguably, the divergent result in the case of support for HBV in Kuwait reflects the diversity of Islamic parties, whose political programs may depend on very different interpretations of religious law. If our Kuwait data included more indicators about political orientations, one could test this theory by investigating potential differences in the effect of support for political Islam according to the specific faction or denomination with which an individual identifies.

Finally, the study’s negative findings about the significance of other aspects of political behavior—frequency of voting and interest in politics—should not be ignored. Supporters of HBV in Kuwait are not more likely to participate in parliamentary elections as compared with opponents, nor are they more interested in politics overall. Thus, the survey shows that two-thirds of Kuwaiti citizens are opposed to a law legalizing violence against a female adulterer, and that this same two-thirds of society is engaged in politics and votes in elections at rates equivalent to those of supporters of HBV-permissive legislation. Far from being overpowered by conservative forces with more active political consistencies, then, Kuwaitis who are against HBV represent a political majority that in principle should be equally potent.

No study is without its limitations. In the case of our survey, first, it does not distinguish between support for honor killing and support for HBV in general. Extensions to this research may seek to focus more specifically on the former, keeping in mind the challenges inherent in inquiring directly about such a topic. Furthermore, our study focuses on the extreme case of adultery and does not examine attitudes toward honor crimes committed for any number of lesser ostensible gender-related offenses. Finally, Kuwait may be unrepresentative in some respects of other Arab and Muslim nations, or at least of other Arab Gulf states, because of its relative social and political openness. Despite such limitations, the current study represents the first nationally representative survey of attitudes toward HBV in the Middle East, and it is one of very few such surveys undertaken of any type. In doing so, it also employs novel methodological tools that future work can leverage to continue the task of increasing the availability and reliability of data on honor killing and other forms of HBV.

Supplemental Material

Appendix_A_ – Supplemental material for Who Supports Honor-Based Violence in the Middle East? Findings From a National Survey of Kuwait

Supplemental material, Appendix_A_ for Who Supports Honor-Based Violence in the Middle East? Findings From a National Survey of Kuwait by Justin J. Gengler, Mariam F. Alkazemi and Alanoud Alsharekh in Journal of Interpersonal Violence

Footnotes

Appendix

Acknowledgements

The authors would like to acknowledge the members of the Abolish 153 Kuwait team, Lulu al Sabah, Sundus Hamza, Sheikha al Nafisi, and Amira Behbehani. We also thank the participants of the London School of Economics workshop on Legal Inequities for Women in the Middle East, who gave helpful feedback on an earlier version of the paper, as well as two anonymous reviewers.

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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Funding for data collection for this study was provided by a grant from the Middle East Partnership Initiative (MEPI).

Supplemental Material

Supplemental material for this article is available online.

Notes

Author Biographies

Justin J. Gengler is research assistant professor at the Social and Economic Survey Research Institute (SESRI) at Qatar University. He is a political scientist specialized in political behavior, public opinion, and survey methodology in the Arab Gulf states. His survey-based research has appeared most recently in Governance, Political Research Quarterly, and International Journal of Public Opinion Research.

Mariam F. Alkazemi is assistant professor of public relations at Virginia Commonwealth University. She has served as a fellow at the Middle East Center at the London School of Economics and co-authored a book on Kuwaiti media law, published by Wolters Kluwer. Some of her research appears in International Communication Gazette, Communication Law & Policy, and Newspaper Research Journal.

Alanoud Alsharekh is research associate at the London Middle East Institute at SOAS University of London. She conducts research on sociopolitical, cultural, and security issues in the Arab Gulf region. She is a board member of several academic and nongovernmental organizations, and director of the “Abolish 153” campaign to end honor killing legislation in Kuwait and the Arab Gulf states.

References

Abdalla

I. A.

(2015). Career facilitators and barriers of Arab women senior executives. International Journal of Business and Management, 10, 218-232.

Advocacy and Lobbying. (2015). Available from http://www.abolish153.org

Ahmed

(2015, December 17). Kuwait Law: No more segregation and animals now have rights. Retrieved from https://248am.com/fajer/law/kuwait-law-no-more-segregation-and-animals-have-rights/

Aladwani

A. M.

(2015). Facilitators, characteristics, and impacts of Twitter use. International Journal of Information Management, 35, 15-25.

Algharabali

(2010). Two cultures, one room: Investigating language and gender in Kuwait (Doctoral dissertation). London, United Kingdom: Queen Mary University of London.

Al Gharaibeh

F. M

. (2016). Debating the role of custom, religion and law in ‘honour’crimes: Implications for social work. Ethics and Social Welfare,10, 122-139.

Al-Nakib

(2012). Human rights, education for democratic citizenship and international organisations: Findings from a Kuwaiti UNESCO ASPnet school. Cambridge Journal of Education, 42, 97-112.

Al-Nakib

(2015). Education and democratic development in Kuwait: Citizens in waiting. Chatham House Research Paper, 7. London, United Kingdom.

Al-Qimlass

(2015). The Counseling Profession in Kuwait: Its Journey and Future. International Journal for the Advancement of Counselling, 37, 355-366.

10.

Ayubi

N. N.

(1996). Over-stating the Arab state: Politics and society in the Middle East. London, England: I.B. Tauris.

11.

Bahgat

(1999). Education in the Gulf monarchies: Retrospect and prospect. International Review of Education, 45, 127-136.

12.

Blaydes

(2014). How does Islamist local governance affect the lives of women? Governance, 27, 489-509.

13.

Caffaro

Ferraris

Schmidt

(2014). Gender differences in the perception of honour killing in individualist versus collectivistic cultures. Sex Roles, 71, 296-318.

14.

Caffaro

Mulas

Schmidt

(2016). The perception of honour-related violence in female and male university students from Morocco, Cameroon and Italy. Sex Roles, 75, 555-572.

15.

Çalik

K. Y.

(2017). Attitudes of Turkish academics regarding violence against women in the name of honor. Journal of Interpersonal Violence, 33, 3232-3254. doi:10.1177/0886260517739288

16.

Carroll

K. B.

(2011). Tribal Law and reconciliation in the new Iraq. The Middle East Journal, 65, 11-29.

17.

Charrad

M. M.

(2009). Kinship, Islam, or oil: Culprits of gender inequality? Politics & Gender, 5, 546-553.

18.

Chesler

(2009). Are honor killings simply domestic violence? Middle East Quarterly, 16, 61-69.

19.

Chesler

(2010). Worldwide trends in honor killings. Middle East Quarterly, 17, 3-11.

20.

Cihangir

(2013). Gender specific honor codes and cultural change. Group Processes & Intergroup Relations, 16, 319-333.

21.

Dashti

A. A.

(2009). The role of online journalism in political disputes in Kuwait. Journal of Arab & Muslim Media Research, 2, 91-112.

22.

Duffy

M. J.

Alkazemi

(2017). Arab Defamation Laws: A Comparative Analysis of Libel and Slander in the Middle East. Communication Law and Policy, 22, 189-211.

23.

Dutton

Bakhiet

S. F.

Essa

Y. A. S.

Blahmar

T. A.

Hakami

S. M. A.

(2017). A negative Flynn effect in Kuwait: The same effect as in Europe but with seemingly different causes. Personality and Individual Differences, 114, 69-72.

24.

Eisner

Ghuneim

(2013). Honor killing attitudes amongst adolescents in Amman, Jordan. Aggressive Behavior, 39, 405-417.

25.

Faqir

(2001). Intrafamily femicide in defence of honour: The case of Jordan. Third World Quarterly, 22, 65-82.

26.

Gengler

(2015). Group conflict and political mobilization in Bahrain and the Arab Gulf: Rethinking the Rentier state. Bloomington: Indiana University Press.

27.

Gengler

K. T.

Howell

(2018). Survey challenges and strategies in the Middle East and Arab Gulf regions. In Johnson

T. P.

Pennell

B.-E.

Stoop

I. A. L.

Dorer

(Eds.), Advances in comparative survey methods: Multinational, multiregional and multicultural Contexts (3MC) (pp. 555-568). New York, NY: John Wiley.

28.

Gengler

Mitchell

J. S.

(2018). A hard test of individual heterogeneity in response scale usage. International Journal of Public Opinion Research, 30, 103-124.

29.

Hainmueller

Hangartner

Yamamoto

(2015). Validating vignette and conjoint survey experiments against real-world behavior. Proceedings of the National Academy of Sciences of the United States of America, 112, 2395-2400.

30.

Herb

(2002). Emirs and parliaments in the Gulf. Journal of Democracy, 13(4), 41-47.

31.

Herb

(2004). Princes and parliaments in the Arab world. The Middle East Journal, 58, 367-384.

32.

Ibrahim

S. E.

(1993). Crises, elites, and democratization in the Arab world. The Middle East Journal, 47, 292-305.

33.

Idriss

M. M.

(2020). “The mosques are the biggest problem we’ve got right now”: Key agent and survivor accounts of engaging mosques with domestic and honor-based violence in the United Kingdom. Journal of Interpersonal Violence, 35(13-14), 2464-2491. doi:10.1177/0886260517703376

34.

Khan

(2003). Zina and the moral regulation of Pakistani women. Feminist Review, 75, 75-100.

35.

King

Murray

C. J.

Salomon

J. A.

Tandon

(2004). Enhancing the validity and cross-cultural comparability of measurement in survey research. American Political Science Review, 98, 191-207.

36.

Krosnick

J. A.

Alwin

D. F.

(1987). An evaluation of a cognitive theory of response-order effects in survey measurement. Public Opinion Quarterly, 51, 201-219.

37.

Krumpal

(2013). Determinants of social desirability bias in sensitive surveys: A literature review. Quality & Quantity, 47, 2025-2047.

38.

Kulczycki

Windle

(2011). Honor killings in the Middle East and North Africa: A systematic review of the literature. Violence Against Women, 17, 1442-1464.

39.

Kuwaiti women win right to obtain passport without husband’s approval. (2009, October 22). The Telegraph. Retrieved from http://www.telegraph.co.uk/news/worldnews/middleeast/kuwait/6402047/Kuwaiti-women-win-right-to-obtainpassport-without-husbands-approval.html

40.

Liebesny

H. J.

(1956). Administration and legal development in Arabia: The Persian Gulf principalities. The Middle East Journal, 10, 33-42.

41.

McFarland

S. G.

(1981). Effects of question order on survey responses. Public Opinion Quarterly, 45, 208-215.

42.

Meyer

Rizzo

Ali

(2007). Changed political attitudes in the Middle East: The case of Kuwait. International Sociology, 22, 289-324.

43.

Moghaizel

(2000). Crimes of Honor: Crimes of Horror. Al-Raida Journal, 89, 16-18.

44.

Nederhof

A. J.

(1985). Methods of coping with social desirability bias: A review. European Journal of Social Psychology, 15, 263-280.

45.

Payton

(2014). “Honor,” Collectivity, and Agnation: Emerging Risk Factors in “Honor”-Based Violence. Journal of Interpersonal Violence, 29, 2863-2883.

46.

Pleasance

(2014, October 10). Father who brought daughter, 24, to Britain to learn English stabbed her to death for using a mobile in honour killing. Daily Mail. Retrieved from http://www.dailymail.co.uk/news/article-2788439/

47.

Rafferty

(2013). International dimensions of discrimination and violence against girls: A human rights perspective. Journal of International Women’s Studies, 14, 1-23.

48.

Rebeiz

M. J.

Harb

(2010). Perceptions of rape and attitudes toward women in a sample of Lebanese students. Journal of Interpersonal Violence, 25, 735-752.

49.

Selvik

(2011). Elite rivalry in a semi-democracy: The Kuwaiti press scene. Middle Eastern Studies, 47, 477-496.

50.

Sheridan

Scott

A. J.

Roberts

(2015). Young women’s experiences of intrusive behavior in 12 countries. Aggressive Behavior, 42, 41-53.

51.

Shultziner

Tétreault

M. A.

(2011). Paradoxes of democratic progress in Kuwait: The case of the Kuwaiti women’s rights movement. Muslim World Journal of Human Rights, 7, 1-25.

52.

Tannenberg

(2017, June). The autocratic trust bias: Politically sensitive survey items and self-censorship (V-Dem Working Paper 49). Gothenburg, Sweden: University of Gothenburg.

53.

Terman

R. L.

(2010). To specify or single out: Should we use the term “honor killing”? Muslim World Journal of Human Rights, 7, 1-39.

54.

Tessler

Jamal

Bedaida

Abderebbi

Shikaki

Braizat, Gengler

Robbins

(2016). Arab barometer: Public opinion survey conducted in Algeria, Morocco, Jordan, Lebanon, Palestine, Yemen, and Bahrain 2006-2009. Ann Arbor, MI: Inter-university Consortium for Political and Social Research [distributor], 2016-02-26.

55.

Tétreault

M. A.

(2001). A state of two minds: State cultures, women, and politics in Kuwait. International Journal of Middle East Studies, 33, 203-220.

56.

Thuroczy

F. A.

(2010). Kuwait: Democratization in process: Five factors of democratization and their state in Kuwait. Lund, Sweden:

57.

Lund University. Toumi

. (2015, November 18). Kuwait man stabs sister for planning to live alone. Gulf News. Retrieved from http://gulfnews.com/news/gulf/kuwait/kuwait-man-stabs-sister-for-planning-to-live-alone-1.1621779

58.

United Nations Development Programme. (2004). Arab human development report 2004: Towards freedom in the Arab world. Retrieved from http://www.arab-hdr.org/Reports/2004/2004.aspx

59.

Wahid

(2014). Nurturing the Salafi Manhaj: A study of Salafi Pesantrens in contemporary Indonesia (Doctoral dissertation). Utrecht, Netherlands: Utrecht University.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.13 MB