The Impacts of International Rape Laws Upon Official Rape Rates

Abstract

Relying on the official rape rates, previous cross-national researchers found a significant and positive relationship between gender equality and the rates of reported rapes, and they used such findings as supporting evidence of the backlash hypothesis that increased gender equality leads to augmented violence against females. However, such a conclusion may be faulty because of the contested validity and reliability of the official rape rate. This study examined how the variation in a comprehensive legal definition of rape is associated with the rates of reported rapes across 74 countries by using the rates of reported rapes from the United Nations Office of Drugs and Crime. The regression analysis suggests that the comprehensive legal definition of rape, more than a country’s gender equality and developmental level, is related to the official rape rate. A country records a high official rape rate when the country includes a broad range of sexual assaults in the category of rape crime, suggesting a reevaluation of the existing studies on this topic.

Keywords

rape law reform legal definition of rape validity and reliability of official rape rate cross-national study feminist theories of victimization

Cross-national studies by Austin and Kim (2000) and Chon (2013) using police rape rates from the Interpol (1995) and the United Nations Office on Drugs and Crime (UNODC, 2012), respectively, found a significant and positive relationship between gender equality and cross-national rates of reported rapes. These researchers used such findings as supporting evidence for the backlash hypothesis (Brownmiller, 1975; Russell, 1975), whereby males use sexual assault as a tool to keep females in their inferior positions when females achieve enhanced socio-economic status with improved gender equality in a country. However, their conclusion may be faulty because the validity and reliability of official rape statistics are questionable. Even in a highly developed country, the rates of reported rapes based on police records are not accurate, as up to 84% of rape incidents are not likely to be reported to the police (Yung, 2014). Further, official rape rates for some countries (mostly developing countries) are more likely than in others (mainly developed countries) to be underestimated.

Several factors, other than actual rape incidents, may contribute to the variation of official rape rates among countries globally. Rape law reforms have occurred in Western countries, such as the United States, Sweden, and Australia, resulting in significant changes to the context of rape. First, the broadening of the legal definition of rape led to the inclusion of an array of sexual assaults, such as marital rape and forced oral and anal sex, as well as traditional penile-vaginal intercourse. Second, national and state laws in some countries (e.g., Canada, England, Finland, the Republic of Ireland, the United States, and Wales) expanded the “classic” definition of rape (reflective of a female victim and male perpetrator) (Notonty, 2002) to being gender-neutral in recognizing a male as a rape victim and females as capable of committing the crime of rape (Rumney, 2007).

In contrast to most Western and developed countries, rape law reforms have not occurred in numerous developing countries globally, such as Egypt, Ghana, India, and Lebanon. Consequently, a narrow range of sexual assaults is included in the legal definition of rape in these countries. Also, rape victims are likely to be reluctant to report their victimization as social sensitivity toward sexual violence is relatively low, and the criminal justice system’s response to rape victimization is slow in these countries (Adinkrah, 2011; Htun & Jensenius, 2020; Randall & Venkatesh, 2015). Ahrens (2006) found that rape survivors commonly experienced adverse social reactions when disclosing their experience with key community systems (e.g., legal, medical, mental health, rape crisis centers, and religious community). These included being blamed for the assault, receiving insensitive responses (i.e., doubting, lack of sympathy), and ineffective victim assistance (i.e., lack of help from support providers). Even when rape victims reported, in some cases, the police refused to file a report or formally charge the assailant with rape. As such, “negative reactions may thereby serve a silencing function,” resulting in rape victims’ discontinuation of discussion about their experience to anyone (Ahrens, 2006, p. 265).

In short, it is possible that the rape law reform in Western countries has broadened the net and brought more offenders into the criminal justice system in comparison with the countries not experiencing rape law reform. Thus, the cross-national official rape rates may be an artifact of the comprehensiveness of the legal definition of rape across countries. In this study, the “comprehensive” legal definition of rape refers to the definition that includes broad acts of sexual assaults in the crime of rape. Consequently, a country that has a comprehensive legal definition of rape is likely to record a higher rape rate than the one that has a non-comprehensive legal definition of rape (Berger et al., 1994). The authors of this study used the comprehensiveness of rape law from the Organization for Economic and Co-operation and Development (OECD, 2020) to test its impact on official rape rates for 74 countries.

International Variation of Legal Definition of Rape and Official Rape Rates

Although the legal definition of all types of crimes is likely to vary from one nation to another, the variations in the legal definition of rape may be even greater than those of many other types of crimes. Although many types of sexual assault exist, rape is the most serious form. One of the factors contributing to the variations in the legal definition of rape is rape law reforms that have occurred in many Western countries over the last few decades (Dripps, 2008). Thus, it is essential to review the content of rape law reform.

Under English common law, rape includes only penile-vaginal intercourse, and the same law does not recognize a male rape victim. As a result, traditionally, deviant sexual assaults, such as forced oral and anal sex and the vaginal penetration by body parts and objects other than the penis, do not fall under rape crime. Rather, these deviant sexual assaults are classified as other less serious sex offenses. Also, the common law does not recognize rape that occurs within marriage, which is called a marital rape exemption. This exemption operates under the premise that a wife gives consent to have sex with her husband when she marries and that marital rape is a private matter (Bennice & Resick, 2003; Htun & Jensenius, 2020). Also, the cultural belief that marital rape is not “real rape” contributes to the marital rape exemption. Thus, “real rape,” the stereotypical context, is perceived as rape that occurs between strangers (Bennice & Resick, 2003).

Beginning in the 1970s, the feminist groups in several Western countries wanted to change traditional rape laws to include a broader range of sexual assaults in the legal definition of rape (Dripps, 2008). As a result, forced anal and oral penetration, as well as the traditional penile-vaginal penetration, has begun to be included in the definition of rape. Additionally, vaginal penetration by body parts and objects other than the penis falls under the category of rape (Lutz-Priefert, 2015). Even some jurisdictions in the United States reclassified rape into sexual assault and sexual battery by broadening the types of sexual assaults that should be included in rape crime¹ (Berger et al., 1994). Likewise, the Federal Bureau of Investigation (FBI) in the U.S. eliminated the requirement of the use of force and an act against a victim’s will in the definition of rape. Instead, a sexual assault without the consent of a victim can fall under rape. Thus, sexual assault can become rape when a victim is incapable of granting consent due to mental or physical incapacity (The U.S. Department of Justice, 2012).

Another essential element of rape law reform was that marital rape became criminally punishable in several countries including Argentina, Australia, Canada, France, New Zealand, Sweden, the United States, and most countries in Latin and Central America (Adinkrah, 2011; Bennice & Resick, 2003). When considering that a significant portion of rapes are perpetrated by current partners (Dartnall & Jewkes, 2013), whether a country includes marital rapes in rape statistics is likely to affect its rates of reported rapes. In the United Kingdom, 45% of all rapes were committed by current partners, and one-third of males in India said that they had forced sex with their wives (Htun & Jensenius, 2020). Von Hofer (2000) suggested that the rape rate in Sweden was 8.2 per 100,000, three times higher than that in the 35 other Organization for Economic Cooperation and Development (OECD) countries. According to Von Hofer (2000), this may be because marital rape has been recognized in Sweden since 1965.

A small number of non-Western and developing countries have been influenced by the rape law reform that has occurred in several Western countries. Fiji and Zimbabwe, for example, have changed and widened their legal definitions of rape and recognized marital rape (Adinkrah, 2011). However, many non-Western and developing countries still have not enacted such comprehensive rape law reforms. As a result, a narrow range of sexual assaults fall under rape crime in these countries. According to Touquet and Gorris (2016), in 79 countries, sexual violence against males is defined as sodomy, and still, many national laws do not recognize male victims in the definition of rape.

Sexual assaults committed against their spouses are either legal or not charged in many countries in Africa, the Middle East, and Asia, including Afghanistan, Sri Lanka, India, Kenya, and Malawi (Htun & Jensenius, 2020), and even in some OECD countries, such as the Czech Republic and Japan. For example, Lebanon entrenched a husband’s marital right to have sexual intercourse, and marital rape was not recognized as of 2014 (Randall & Venkatesh, 2015). National culture may influence the criminalization and non-criminalization of marital rape. Male and female college students in Ghana opposed to criminalizing marital rape, as they believed it is the cultural obligation of females to provide sex to their husbands whenever husbands want sex (Adinkrah, 2011). Additionally, perpetrators are not charged if they marry their victims in Angola, Bahrain, Bulgaria, Cameroon, Equatorial Guinea, Eritrea, Iraq, Jordan, Lebanon, Libya, the Philippines, Syrian Arab Republic, Tunisia, and the West Bank and Gaza (Htun & Jensenius, 2020; Randall & Venkatesh, 2015). Likewise, in India, males can be charged for marital rape only if their wives are younger than age 18 (Htun & Jensenius, 2020).

Such a comprehensive legal definition of rape, including legal recognition of marital rape, may be related to official rape rates in a country. In their study of 48 states in the United States, Berger et al. (1994) argued for the symbolic effects of rape law reforms because such reforms may influence cultural and public attitudes toward rape. Berger et al. (1994) concluded that a country is likely to record a high rape rate when it has a comprehensive legal definition of rape. The opposite may be true for countries with a narrow legal definition of rape.

Rape law reform also changed the way the criminal justice system perceives a rape case (Spohn & Horney, 1992). In some Western countries, to encourage reports of rape incidents by victims and others, legal protections for victims were established. Rape shield laws were enacted, prohibiting disclosure of rape victims’ identity and preventing a defense attorney from asking the victims questions about their sexual history during cross-examination at trial if it is not directly related to the case (Klein, 2008). Additionally, rape victims no longer must prove that they resisted sexual assault (Dripps, 2008; Von Hofer, 2000) and the requirement for corroboration was eliminated (Horney & Spohn, 1991). However, prior researchers have reported inconsistent results regarding the effect of such procedural rape law reform on victims’ police reporting and indictment. For example, Spohn and Horney (1992, 1993) and Horney and Spohn (1991) suggested that rape law reform occurring in the U.S. during the 1970s had a limited effect. Horney and Spohn (1991) found that only one of six U.S. jurisdictions in their study had a significant increase in reports of rape and indictment ratio during the period of post-reform. By contrast, Clay-Warner and Burt (2005) found that rape reporting had more broadly increased nationwide after the rape law reform.

Likewise, the comprehensiveness of the legal definition of rape may influence police officers’ reactions to sexual assault. Police officers are gatekeepers of the criminal justice system, and they exercise wide discretionary decision-making power when responding to an incident (Artz & Smythe, 2007). Legal factors (e.g., evidence of penetration, use of force, and a victim’s injury) as well as extralegal factors (e.g., an officer’s personal perception of rape, a victim’s alcohol/drug use, previous sexual relationship between the victim and the perpetrator, and a victim’s reactions) are likely to be related to a police officer’s decision to arrest a suspect (Venema, 2016). However, it may be difficult for police officers to ignore the legal definition in their jurisdictions as they must establish probable cause requirement for arrests, and department policies and procedures play a crucial role in their arrest decisions (Venema, 2016). For example, Marsh et al. (1982) reported that the rape law reform including legal definition change to include a broad range of sexual assaults as rape drastically increased arrest rates of rapes in Michigan in the United States. Thus, using their understanding of the legal definition of rape and information from witnesses and forensic evidence, police officers must decide if a reported crime meets the criteria for rape (Lea et al., 2003). Police officers are likely to arrest more suspects for rape when their country’s criminal law includes a broad range of sexual assaults under the category of rape.

Feminist Perspective on Rape Victimization

Pertaining to the impact of rape law reform on rape rates, two main feminist perspectives—amelioration and backlash hypothesis—may be useful in examining the relationship between gender equality and violence against females. The amelioration hypothesis posits that gender equality is conducive to a reduced level of violence against females. In societies where patriarchy is present, males have advantages over females and males dominate females in nearly every aspect of life. Within this context, males use violence against females to keep females in their inferior and subservient positions (Brownmiller, 1975; Whaley, 2001). For example, advocates of rape law reform view traditional rape laws as preserving male dominance and subjugating females as sexual objects (Brownmiller, 1975). Rape law reform would “symbolize rejection of this patriarchal view” and recognize that rape is a crime of violence (Spohn & Horney, 1992, p. 20). When females advance or gain more equal standing in society, instituting change in gender norms and societal culture, this over time leads to a reduction in violence against females. In theory, improved gender equality equips females with greater ability to defend themselves from male violence, resulting in the declining victimization of females (Chafetz, 1984; Schwendinger & Schwendinger, 1983).

In contrast to Chafetz (1984) and Schwendinger and Schwendinger (1983), Brownmiller (1975) and Russell (1975) suggested that gender equality leads to increased levels of violence against females as a significant amount of research contradicts the amelioration hypothesis. Russell (1975) called the unexpected findings the “backlash” hypothesis stating as females gain gender equality by advancing their social and economic status, their advancement challenges and threatens male dominance and leads to a “backlash” of violence against females. Thus, once males feel that their masculinity is threatened by females’ enhanced social position, they are likely to use violence against females to maintain their dominance over females (Whaley, 2001). As a result, improved gender equality in school education, occupational status, labor force participation, and political status increases the victimization of females, instead of reducing it (Austin & Kim, 2000).

By using official statistics, Austin and Kim (2000) and Chon (2013) suggested that improved female occupational status is related to increased rates of reported rape across countries, and they used such findings as supporting evidence for the backlash hypothesis. However, Austin and Kim (2000) and Chon’s (2013) conclusions may be faulty, as the official statistics from the Interpol (1995) and the UNODC (2012) may have validity and reliability issues. Using official rates of rape has a clear limitation because the majority of rapes and sexual assaults are not reported to the police despite changing legal definitions. Thus, the official rates of rape do not provide much information on the actual incidents of rapes.

The Present Study

As discussed earlier, the validity and reliability of official rape rates may be influenced by several factors, such as victims and others’ reporting behavior to the police, the manner in which the rape case is managed by the authorities, social sensitivity toward sexual assaults, and the manner in which the rape case is recorded. For example, given the high drop rate of rape cases by victims, the rape statistics are partially dependent upon the time of rape recording by the police. Hence, the rape rate for a country that records rape as soon as a victim reports is likely higher than that for the country that records after the conclusion of a criminal investigation (Htun & Jensenius, 2020). However, the availability of international information on these factors is limited. We could introduce only the comprehensiveness of legal definition rather than these factors as a predictor of the variation in official rape rates.

It is important to clarify that this study focused on the impact of rape laws upon official rape rates, as opposed to a victimization survey. We took advantage of the comprehensiveness of the legal definition of rape from the Organization for Economic Cooperation and Development’s (OECD, 2020) and examined how it is connected to rates of reported rapes from the UNODC (2020) while controlling other variables, such as a nation’s gender inequality index, the Gini-coefficient of income inequality, minority and female population size, and the Latin American region and Muslim country dummy variables. Given our earlier theoretical discussion, we hypothesized that the comprehensiveness of the legal definition of rape is positively related to the official rape rate. The more comprehensive a country’s definition of rape, the higher its rape rate.

Data and Methods

Dependent Variable

The dependent variable was the rape rate per 100,000 national populations (Table 1). The information on the rates of reported rapes of 74 countries was provided by UNODC (2020). The UNODC (2020) publishes rape counts and rates for the United Nations’ member countries, and the same organization defines rape as “sexual intercourse without valid consent” (UNODC, 2010, p. 24) and indicates that rape is mostly committed by males and the victims are females. The rates of reported rapes are based on police records circa 2014.

Table 1.

Descriptive Statistics.

	N	Min.	Max.	Mean	SD
Rape rate	74	0.12	58.14	10.80	11.37
Comprehensive legal definition of rape	74	0.00	0.80	0.30	0.26
Gender inequality index	74	0.04	0.58	0.27	0.16
HDI	74	0.42	0.95	0.78	0.12
Gini-coefficient of income inequality	74	25.60	51.70	37.03	7.75
Minority population (%)	74	1.50	90.50	29.22	21.71
Female population (%)	74	48.16	54.09	50.85	1.13
Age 15–29 group (%)	74	15.45	30.91	23.16	4.36

Independent Variables

The comprehensiveness of rape law

The OECD (2020) published information on the comprehensiveness of rape law in 2014. The OECD considers comprehensiveness of the legal definition of rape as an ordinal variable and correspondingly assigns a numerical value for each category. According to the OECD (2020), a value of 0 is designated for the country that provides a comprehensive legal definition of rape and includes “marital rape” in rape crime.² Perpetrators cannot escape prosecution, even if they marry the victims. Also, the implementation of such a comprehensive legal definition is effectively enforced. Next, a value of 0.25 indicates the country that provides a similar comprehensive legal definition as the country with a value of 0. However, such a comprehensive legal definition of rape is not effectively enforced. On the other hand, a value of 0.50 is given to the country that does not include marital rape in the legal definition of rape. Finally, theoretically, a value of 1.00 refers to the country that does not have a comprehensive legal definition of rape and thereby lacks its enforcement.

According to the OECD (2020), the following attributes are in the legal definition of rape:

0: There is specific legislation in place to address rape, marital rape is included, perpetrators cannot escape prosecution if they marry the victim and implementation is effectively enforced.

0.25: There is specific legislation in place to address rape, marital rape is included and perpetrators cannot escape prosecution if they marry the victim, although the implementation is not effectively enforced.

0.5: There is specific legislation in place to address rape, marital rape is not included and perpetrators cannot escape prosecution if they marry the victim.

1: There is no legislation in place to address rape.

Although the OECD (2020) provides a definition of each category and a specific numerical value, it appears that many countries were assigned a numerical value somewhere between these four values, and no country received a value of 1.00. To make it easier to understand, we reverse coded so that a low value indicated a low level of a comprehensive legal definition of rape law and its enforcement. We treated the legal definition of rape as an interval measure for further analysis.

Other variables

This study included other variables, such as the gender inequality index, the Gini-coefficient of income inequality, minority and female population sizes, and the Latin American region and Muslim country dummy variables. First, to examine feminist theories regarding rape victimization, the gender inequality index (hereafter GII), a comprehensive measure of gender inequality in a country, was introduced. The GII has been published by the United Nations Development Program (UNDP, 2015) since 2010. The GII measures three areas of gender inequality: reproductive health (maternal mortality and adolescent fertility), females’ empowerment (percentage of females’ seats in a national parliamentary body and females’ education for secondary school or above), and females’ labor market participation (labor force participation). The GII data for 2014 theoretically ranges from zero (very low gender inequality) to one (very high gender inequality).

We also included the Human Development Index (hereafter HDI) for the year 2014 (UNDP, 2015), which is more comprehensive than GDP per capita as an indicator of the overall development level of a country because the HDI includes life expectancy at birth and adult literacy and combined school enrollment as well as GDP per capita. The relationship between the level of development and the violence level can be understood within the frame of the civilizing theory (Elias, 1978). The criminal justice system in a civilized country is well developed and efficient, so its people use the courtrooms rather than violence to resolve interpersonal conflicts (Chon, 2013).

Second, although it is difficult to find a study exploring the relationship between income inequality and rates of reported rapes, Blau and Blau (1982) reported a positive association between income inequality and homicide rate. Pridemore (2002), utilizing the conflict theory (Quinney, 1970), contended that economically disadvantaged groups who feel being mistreated tend to show their frustration toward others through violence, resulting in a high homicide rate. Chon’s (2016) cross-national study found a significant and positive relationship between income inequality and the female homicide rate. Thus, this study used the Gini-coefficient, theoretically ranging from 0 to 100, as a measure of income inequality, and a high score indicates a high level of household income inequality. The Gini-coefficient of income inequality was obtained from the Human Development Reports (UNDP, 2015).

Third, the social disorganization theory (Shaw & McKay, 1942) posits that ethnic diversity in a nation may impede effective social control of violence by community members, as collective efforts to control social disorder are difficult when diverse ethnic groups exist in a community. For example, Stamatel (2018) found a significant and positive connection between national ethnic diversity and the female homicide rate. Thus, the minority population size (%) was introduced as a measure of ethnic diversity, which was based on the Time Almanac 2010 (Encyclopedia Britannica, 2009).

The Latin American region and Muslim countries were other predictors regarding domestic violence. Neapolitan (1994) proposed the existence of machismo, an exaggerated sense of masculinity emphasizing male dominance over females in many Latin American countries. According to Neapolitan (1994), machismo may be related to a high rape rate in Latin America. By contrast, several researchers suggested that Islam religion is related to the low violence rate in Muslim nations. It is possible that an exaggerated sense of honor and shame culture contributes to a low report rate by a rape victim as many Muslim societies stigmatize a rape victim and a raped female is not marriageable to other males except the rapist. Additionally, the patriarchal culture may adversely affect rape recording and conviction practices by police in Muslim countries. To convict a rapist, either the confession from a rapist or the statement from a witness is required, which does not commonly occur in rape cases (Haddad, 2017).

Thus, this study introduced dummy variables that indicate the Latin American region and Muslim countries. Given Neapolitan’s (1994) research, 16 South and Central American and Caribbean countries were classified in the Latin American region. On the other hand, a country was classified as Muslim country when the majority (50% or more) of its national population was affiliated with Islam, and seven countries were included as Muslim countries (Appendix A). The information on predominantly Muslim countries came from the Time Almanac 2010 (Encyclopedia Britannica, 2009).

Other variables that may be related to violence toward females are their gender and young age group population sizes. It may be necessary to include female population size, as most rape offenders are males and most victims are females. In addition to the female population size, the young age group population size may be a possible predictor of violence against females. For example, Yodanis (2004) found a significant and positive relationship between the young adult population size and sexual violence. Thus, we included female and teenage and young adult population sizes age 15 to 29 (%) from The UN Demographic Yearbook (The United Nations, 2015).

Diagnostic Tests and Analytical Strategy

Diagnostic tests were performed before conducting the regression analysis. No missing data problem existed, and all data for the dependent and independent variables were available for 74 countries. However, the rape rate was highly skewed and violated the normal distribution assumption of the regression analysis. Thus, the rape rate was transformed by using a natural logarithm. The zero-order correlations between the GII and the HDI (r = −0.87), between the HDI and age 15 to 29 (r = −0.80), and between the GII and age 15 to 29 (r = 0.84) were elevated, which may cause collinearity issues. Thus, these three variables were introduced separately in the regression models. Additionally, the regression models with and without the comprehensiveness of rape laws were developed to compare the effect of the introduction of the comprehensiveness of the rape law, which led to a total of six regression models.

The Results

Descriptive Statistics and Bivariate Correlation Analysis

The rates of reported rape significantly varied by country (Table 1). Egypt (0.12) and Sweden (58.14) had the lowest and the highest rate of reported rape per 100,000 population among 74 countries, respectively. Twenty-seven countries had the most comprehensive definition of rape law (0.80), including Sweden, the United Kingdom, and the United States. Conversely, six countries had the least comprehensive definition of rape law (0.00), including Algeria, Armenia, Bulgaria, Cameron, the Czech Republic, and Russia. Next, Denmark (0.04) had the lowest GII (the highest gender equality), and Cameroon (0.58) had the highest GII (the lowest gender equality). Finally, Burundi (0.42) and Norway (0.95) had the lowest and the highest HDI, respectively.

The zero-order correlation deserves a brief discussion (Table 2). The comprehensive legal definition of rape was positively associated with the rates of reported rape (r = 0.47, p < .01). The Latin American dummy variable was significantly and positively related to the rates of reported rape (r = 0.40, p < .01), and the Muslim country variable was significantly and negatively associated with the rates of reported rape (r = −0.42, p < .01). However, all other variables had no significant association with the rates of reported rape. The bivariate correlation was only a preliminary analysis, and clearer results must come from the multiple regression analysis.

Table 2.

Zero-Order Correlation Matrix.

	1	2	3	4	5	6	7	8	9
1. Rape rate	1
2. Comp. rape law	−0.47**	1
3. GII	−0.09	0.34**	1
4. HDI	0.19	−0.32**	−0.87**	1
5. Gini coefficient	0.19	−0.08	0.60**	−0.46**	1
6. Minority pop.	0.07	0.11	0.49**	−0.47**	0.36**	1
7. Female pop.	−0.16	0.14	−0.26*	0.18	−0.19	−0.30**	1
8. Age 15–29	−0.02	0.26*	0.84**	−0.80**	0.51**	0.38**	−0.31**	1
9. Latin America	0.40**	−0.18	0.42**	−0.20	0.67**	0.07	−0.16	0.42**	1
10. Muslim country	−0.42**	0.29*	0.26*	−0.20	−0.16	0.04	−0.27*	0.36**	−0.17

Note. Two-tailed test: *p < .05. **p < .01.

Regression Analysis

Table 3 presents the results of the rates of reported rapes. In the first regression model, the age 15 to 29 group was introduced with the Gini-coefficient of income inequality, minority and female population sizes, and the Latin America and Muslim country dummy variables. The Latin American dummy variable was positively related to the rape rate at a 0.001 level, and the Gini-coefficients of income inequality, female population, and Muslim country were negatively connected to the rates of reported rape at a 0.05 level, a 0.001 level, and a 0.001 level, respectively. However, the age 15 to 29 group and minority population size had no significant association with the rates of reported rape.

Table 3.

Regression Analysis of Rape Rate (N = 74).

	1	2	3	4	5	6
Comprehensive legal definition of rape	—	1.577** (0.502)	—	1.285** (0.502)	—	1.377** (0.534)
Gender inequality index	—	—	—	—	−1.878* (1.113)	−0.522 (1.192)
HDI	—	—	2.553* (1.237)	1.507 (1.257)	—	—
Gini-coefficient of income inequality	−0.043* (0.025)	−0.047* (0.023)	−0.027 (0.024)	−0.033 (0.024)	−0.028 (0.025)	−0.038 (0.025)
Minority population (%)	0.006 (0.007)	0.007 (0.006)	0.010 (0.007)	0.010 (0.006)	0.010 (0.007)	0.009 (0.007)
Female population (%)	−0.265* (0.118)	−0.157 (0.117)	−0.247* (0.115)	−0.168 (0.115)	−0.250* (0.116)	−0.168 (0.116)
Age 15–29 group (%)	−0.013 (0.040)	0.025 (0.039)	—	—	—	—
Latin America	1.451*** (0.428)	1.227** (0.408)	1.366*** (0.402)	1.288*** (0.388)	1.546*** (0.413)	1.347*** (0.404)
Muslim country	−1.873*** (0.512)	−1.621*** (0.487)	−1.671*** (0.447)	−1.381** (0.445)	−1.572*** (0.486)	−1.413** (0.471)
R-squared	0.378	0.459	0.414	0.467	0.403	0.457
Adjusted R-squared	0.322	0.402	0.362	0.411	0.349	0.400

Note. One-tailed test: *p < .05. **p < .01. ***p < .001.

The second regression model added the comprehensiveness of the legal definition of rape to the first model. The comprehensive legal definition of rape positively related to the rates of reported rape at a 0.01 level. The Latin America dummy variable positively linked to the rates of reported rape at a 0.01 level, and the Gini-coefficient and Muslim country dummy variable were negatively associated with the rates of reported rape at 0.05 and 0.001 levels, respectively. However, the female population size no longer had a significant relationship with the rates of reported rape, and the minority population and the age 15 to 29 group were not significantly related to the rates of reported rape. The adjusted R-squared improved from 0.322 in the first model to 0.402 in the second model.

In the third model, the human development index (HDI) was introduced, and the comprehensiveness of the legal definition of rape and the age 15 to 29 group were removed from the model. The HDI positively related to the rates of reported rape at a 0.05 level. The Latin America dummy variable positively correlated at a 0.001 level, and the female population size and Muslim country dummy variable negatively coupled with the rates of reported rape at 0.05 and 0.001 levels, respectively. However, the Gini-coefficient and minority populations maintained no significant relationship with the rates of reported rape.

In the fourth model, the comprehensiveness of the legal definition of rape was added, and it positively linked to the rape rate at a 0.001 level. By contrast, the HDI, the Gini-coefficient, and the female population size were no longer significant predictors of the rates of reported rape. The Latin American dummy variable maintained a positive and Muslim country had a negative relationship with the rates of reported rape at 0.001 level and 0.01 levels, respectively. The introduction of the comprehensiveness of the legal definition of rape improved the adjusted R-squared from 0.362 in the third model to 0.411 in the fourth model. These seven variables explained the highest percent (41.1%) of the variance in the rates of reported rapes across all the models.

In the fifth model, the gender inequality index (GII) was introduced, and the HDI was removed. The GII negatively related to the rates of reported rape at a 0.05 level, and the Latin American dummy variable positively associated with the rates of reported rape at a 0.001 level. The female population size and Muslim country dummy variable negatively connected to the rates of reported rape at 0.05 and 0.001 levels, respectively. Unlike these variables, the Gini-coefficient and minority population size had no significant association with the rates of reported rape.

In the sixth model, the comprehensiveness of the legal definition of rape was added to the fifth model. The comprehensiveness of the legal definition of rape positively related to the rates of reported rape at a 0.01 level. However, the GII was no longer a significant predictor of the rates of reported rape. The Latin America dummy variable positively correlated at a 0.001 level, and the Muslim country dummy variable was negatively associated with the rates of reported rape at a 0.01 level. However, the Gini-coefficient and minority and female population sizes maintained no significant relationship with the rates of reported rape. The adjusted R-squared improved from 0.349 in the fifth model to 0.400 in the sixth model.

Discussion and Conclusion

Interpretation of Results

Previous cross-national studies reported a positive relationship between gender equality and the official rape rate and used it as supporting evidence of the backlash hypothesis (Austin & Kim, 2000; Chon, 2013). The authors of this study hypothesized that such a conclusion may be faulty, as the validity and reliability of the official rape statistics are likely to vary by country. Among many other factors, the comprehensiveness of the legal definition of rape is expected to affect the official rape rates. Thus, this study tested the relationship between the comprehensiveness of the legal definition of rape and the official rape rate, while controlling for several other variables, including the human development index and gender inequality index.

The regression results showed that the comprehensive legal definition was related to an increased official rape rate. The country that includes a broad range of sexual assaults in rape crime is more likely to record a high level of rape rate than the country that includes a narrow range of sexual assaults in rape crime. Rape victims do not file a report with the police for several different reasons, such as perceiving the victimization as a personal matter, distrust of the police, fear of exposure of their victimization to others, fear of retaliation by their rapists, and accepting the blame for the victimization (Chon, 2014). Although these factors are crucial for rape victim’s reports of their victimization, it is possible that victims are likely to report sexual assaults and police officers are likely to record such reports as rapes when their countries maintain a comprehensive legal definition of rape. For example, Sweden showed one of the highest rates of reported rapes globally, and it may be because Sweden has a comprehensive legal definition of rape. By contrast, the opposite is possible in countries that have a narrow definition of rape. Many sexual assaults occur in these countries, but their rates of reported rapes may be low since they include a somewhat limited range of sexual assaults as rape crimes.

Another notable finding pertains to the human development index and gender inequality index. The human development index and gender equality maintained a positive relationship with the official rape rate. Overall, developed countries showed a high level of rates of reported rapes compared to developing countries, as reported by other studies (Soars, 2004). However, these variables were no longer significant when the comprehensiveness of the legal definition of rape was introduced in a regression model. Additionally, the introduction of the comprehensiveness of the legal definition of rape significantly improved the explained variation of the rates of reported rape. These findings may plausibly be interpreted as indicating the comprehensiveness of the legal definition of rape is more directly related to the official rape rate than a country’s development level and gender equality.

Another result that is consistent with the current hypothesis comes from the findings on a few other variables. Overall, the Gini-coefficient of income inequality, minority and female population sizes, and the age 15 to 29 group failed to show a significant association with the rates of reported rape. For example, the female population was significantly and inversely linked to the rates of reported rape. However, the same variable was no longer significantly associated with the rates of reported rape when the comprehensiveness of the legal definition of rape was added in the regression model. These variables showed a significant association with the international homicide rate in extant works (Chon, 2016). The international homicide rate may be subject to fewer validity and reliability issues than the rates of reported rape, as the homicide rate is usually based on an actual count of dead persons, and homicide sustains fewer legal definition issues across countries than rape. In sum, non-significant and non-consistent associations of these variables with the rape rate may imply validity and reliability issues with the official rape rate.

The findings on the two regional dummy variables, the Latin American region and Muslim country, are also interesting. Neapolitan (1994) argued the existence of machismo culture that emphasizes males’ dominance over females in Latin America. Thus, this study expected a positive relationship between the Latin American regional dummy variable and the rates of reported rape, and the relationship was supported by this regression analysis even after introducing the comprehensiveness of the legal definition of rape. Conversely, the Muslim country dummy variable was consistently and negatively linked to the rates of reported rape for all six regression models, even after controlling for the effect of the comprehensiveness of the legal definition of rape. Possibly, an exaggerated sense of honor and shame within cultures contribute to a low report rate by rape victims in Muslim countries. Additionally, patriarchal culture and a strict burden of proof requirement may adversely affect rape recording practice by the police in Muslim countries.

Research Implications

Researchers should exercise caution when they use official rape statistics, given the problem with the validity and reliability of this source. Using information from a survey of rape victimization is more effective than official rape statistics. However, an international survey of rape victimization, such as the International Crime Victimization Survey (ICVS International Working Group, 2005), covers only a small number of countries. If researchers want to use official international rape rates, they must consider ways to ameliorate the problem with the validity and reliability of official rape statistics. One such way is to introduce the comprehensiveness of the legal definition of rape from the OECD (2020) in the study to control its effect. However, researchers must be aware that introducing the comprehensiveness of the legal definition of rape does not fully address the validity and reliability of the official rape rate, as many other factors may affect the variation of official rape rates among countries. Given the findings of this study, it may not be appropriate to accept Austin and Kim (2000) and Chon’s (2013) findings as supporting evidence for the backlash hypothesis. Although this study used a cross-national approach, it has research implications for studies within a country. For example, states in the United States have different criminal codes, and the legal definition of rape varies for each state. Thus, researchers also must exercise caution when they compare the rates of reported rapes across various jurisdictions in a country.

Limitations

A few limitations of this study warrant attention. First, this study introduced only the comprehensiveness of the legal definition of rape. However, several other factors are likely to contribute to the validity and reliability issues with the official rape rate, such as reporting behavior by victims and others, recording practice of rape statistics by police agencies, and cultural and social sensitivity toward sexual violence against females. Thus, this study should be considered as a partial test of the validity and reliability issues with the official rape statistics. Additionally, the sample size of 74 countries in a cross-national study is not small in comparison with existing cross-national studies. However, larger sample size is desirable. The sample size could not be extended beyond this because of the limited availability of international information on the dependent and independent variables.

Conclusion

This study suggests that the comprehensiveness of the legal definition of rape is related to the international variation in the official rape rates, indicating the problem of validity and reliability of police rape rates. A country with a high rate of official rape has a comprehensive legal definition of rape, which includes many forms of sexual assault as rape crimes. This suggests the official rape rate does not necessarily indicate that a large number of rapes are committed in that country. Thus, existing studies, which have found a positive relationship between gender equality and the official rape rates and have used it to support the backlash hypothesis, should be reevaluated.

Footnotes

Appendix

Appendix A.

The List of Countries (N = 74).

Albania¹, Algeria¹, Argentina², Armenia, Austria, Azerbaijan¹, Belarus, Belgium, Bolivia², Bosnia, Brazil², Bulgaria, Burundi, Cameron, Canada, Chile², Costa Rica², Croatia, Czech Republic, Denmark, Dominican Republic², Ecuador², Egypt¹, El Salvador², Estonia, Finland, France, Georgia, Germany, Greece, Guatemala², Honduras², Hungary, India, Indonesia¹, Ireland, Israel, Japan, Kazakhstan, Kenya, Kyrgyzstan¹, Latvia, Lithuania, Luxembourg, Macedonia, Mexico², Mongolia, Morocco¹, Netherlands, New Zealand, Norway, Panama², Paraguay², Peru², Philippine, Poland, Portugal, Romania, Russia, Rwanda, Singapore, Slovakia, Slovenia, Spain, Sri Lanka, Sweden, Switzerland, Tanzania, Thailand, Trinidad and Tobago², Uganda, United Kingdom, Uruguay², the United States.

Muslim country.

Latin American country.

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.

ORCID iD

Don Soo Chon

Notes

References

Adinkrah

(2011). Criminalizing rape within marriage: Perspectives of Ghanaian university students. International Journal of Offender Therapy and Comparative Criminology, 55(6), 982–1010.

Ahrens

C. E.

(2006). Being silenced: The impact of negative social reactions on the disclosure of rape. American Journal of Community Psychology, 38(3–4), 263–274.

Artz

Smythe

(2007). Case attrition in rape cases: A comparative analysis. South African Journal of Criminal Justice, 20(2), 158–181.

Austin

Kim

Y. S.

(2000). A cross-national examination of the relationship between gender equality and official rape rate. International Journal of Offender Therapy and Comparative Criminology, 44(2), 204–221.

Bennice

J. A.

Resick

P. A.

(2003). Marital rape: History, research, and practice. Trauma, Violence, & Abuse, 4(3), 228–246.

Berger

R. J.

Newman

W. L.

Searles

(1994). The impact of rape law reform: An aggregate analysis of police reports and arrests. Criminal Justice Review, 19(1), 1–23.

Blau

J. R.

Blau

P. M.

(1982). The cost of inequality: Metropolitan structure and violent crime. American Sociological Review, 47(1), 114–129.

Brownmiller

(1975). Against our will: Men, women and rape. Simon & Schuster.

Chafetz

J. S.

(1984). Sex and advantage: A comparative macro-structural theory of sex stratification. Rowman and Allanheld.

10.

Chon

D. S.

(2013). Test of impact of gender equality and economic development on sexual violence. International Studies on Family and Sexual Violence, 28(6), 603–610.

11.

Chon

D. S.

(2014). Police reporting by sexual assault victims in western and in non-Western countries. Journal of Family Violence, 29(8), 859–868.

12.

Chon

D. S.

(2016). A spurious relationship of gender equality with female homicide victimization: A cross-national analysis. Crime & Delinquency, 62(3), 397–419.

13.

Clay-Warner

Burt

C. H.

(2005). Rape reporting after reforms: Have times really changed? Violence Against Women, 11(2), 150–176.

14.

Dartnall

Jewkes

(2013). Sexual violence against women: The scope of the problem. Best Practice & Research Clinical Obstetrics and Gynecology, 27(1), 3–13.

15.

Dripps

(2008). After rape law: Will the turn to consent normalize the prosecution of sexual assault? Akron Law Review, 41(4), 957–980.

16.

Elias

(1978). The civilization process: The history of manners. Pantheon.

17.

Encyclopedia Britannica. (2009). Time almanac 2010. Encyclopedia Britannica.

18.

Haddad

(2017, May 9). Victims of rape and law: How the laws of the Arab world protect rapists, not victims. Jurist. https://www.jurist.org/commentary/2017/05/mais-haddad-arab-world-laws-protect-the-rapist-not-the-victim/

19.

Horney

Spohn

(1991). Rape law reform and instrumental change in six urban jurisdictions. Law and Society Review, 25(1), 117–154.

20.

Htun

Jensenius

F. R.

(2020). Fighting violence against women: Laws, norms & challenges ahead. Daedalus, the Journal of the American Academy of Arts & Sciences, 149(1), 144–159.

21.

ICVS International Working Group. (2005). International crime victimization survey (ICVS) series. http://www.icpsr.umich.edu/icpsrweb/ICPSR/series/175

22.

Interpol. (1995). International crime statistics. Interpol.

23.

Klein

(2008). An analysis of thirty-five years of rape reform: A frustrating search for fundamental fairness. Akron Law Review, 41(4), 981–1057.

24.

Lea

S. J.

Lanvers

Shaw

(2003). Attrition in rape cases: Developing a profile and identifying relevant factors. British Journal of Criminology, 43(3), 583–599.

25.

Lutz-Priefert

(2015). A call for a more permanent international definition of rape. Creighton International and Comparative Law Journal, 6(85), 85–101.

26.

Marsh

J. C.

Geist

Caplan

(1982). Rape and the limits of law reform. Auburn House.

27.

Neapolitan

J. L.

(1994). Cross-national variation in homicide: The case of Latin America. International Criminal Justice Review, 4, 4–22.

28.

Notonty

(2002). Rape victims in the (gender) neutral zone: The assimilation of resistance? Seattle Journal for Social Justice, 1(3), 743–756.

29.

Pridemore

W. A.

(2002). Vodka and violence: Alcohol consumption and homicide rates in Russia. American Journal of Public Health, 92(12), 1921–1935.

30.

Quinney

(1970). The social reality of crime. Transaction Publisher.

31.

Randall

Venkatesh

(2015). The right to no: The crime of marital rape, women’s human rights, and international law. Brooklyn Journal of International Law, 41(1), 153–202.

32.

Rumney

P. N. S.

(2007). In deference to gender neutrality in rape. Seattle Journal for Social Justice, 6(1), 481–526.

33.

Russell

D. E. H.

(1975). The politics of rape: The victim’s perspective. Stein and Day.

34.

Schwendinger

J. H.

Schwendinger

(1983). Rape and inequality. Sage.

35.

Shaw

C. R.

McKay

H. D.

(1942). Juvenile delinquency and urban areas. The University of Chicago Press.

36.

Soars

R. R.

(2004). Development, crime and punishment: Accounting for the international differences in crime rates. Journal of Development Economics, 73(1), 155–184.

37.

Spohn

Horney

(1992). Rape law reform: A grassroots revolution and its impact. Plenum Press.

38.

Spohn

Horney

(1993). Rape law reform and the effect of victim characteristics on case processing. Journal of Quantitative Criminology, 9(4), 383–409.

39.

Stamatel

J. P.

(2018). Money matters: Dissecting the relationship between gender equality and female homicide victimization rates in the European Union. Feminist Criminology, 13(5), 435–455.

40.

The Organization for Economic Co-operation and Development (OECD). (2020). Country profile. https://www.genderindex.org/countries-territories-2014/

41.

The United Nations Development Program (UNDP). (2015). Human development report. http://www.undp.org

42.

The United Nations Office on Drugs and Crime (UNODC). (2010). International statistics on crime and justice. UNODC.

43.

The United Nations Office on Drugs and Crime (UNODC). (2012). Total sexual violence: A national level, number of police-recorded offenses. http://www.unodc.org/unodc/en/data-and-analysis/statistics/crime.html

44.

The United Nations Office on Drugs and Crime (UNODC). (2020). Statistic and data. https://dataunodc.un.org/crime/sexual_violence

45.

The United Nations. (2015). Demographic yearbook. The United Nations Statistical Division.

46.

The U.S. Department of Justice. (2012). An updated definition of rape. The Department of Justice Archive. https://www.justice.gov/archives/opa/blog/updated-definition-rape

47.

Touquet

Gorris

(2016). Out of the shadows? The inclusion of men and boys in conceptualisations of wartime sexual violence. Reproductive Health Matters, 24(47), 36–46.

48.

Venema

R. M.

(2016). Police officer schema of sexual assault reports: Real rape, ambiguous cases, and false reports. Journal of Interpersonal Violence, 31, 872–899.

49.

Von Hofer

. (2000). Crime statistics as constructs: The case of Swedish rape statistics. European Journal on Criminal Policy and Research, 8, 77–89.

50.

Whaley

R. B.

(2001). The paradoxical relationship between gender inequality and rape: Toward a refined theory. Gender & Society, 15(4), 531–555.

51.

Yodanis

C. L.

(2004). Gender inequality, violence against women, and fear: A cross-national test of feminist theory of violence against women. Journal of Interpersonal Violence, 19(6), 655–675.

52.

Yung

C. R.

(2014). How to lie with rape statistics: America’s hidden rape crisis. Iowa Law Review, 99(3), 1197–1256.