Safe at home? Examining the extension of criminal penalties for marital rape in cross-national context, 1979–2013

Abstract

While sociologists have focused on the national adoption of public-sphere women’s rights such as the right to vote in elections or participate fully in economic matters, less work has examined the diffusion of private-sphere women’s rights, rights of women in the home. We address this gap by examining the cross-national adoption of laws that criminalize marital rape. Building on prior research that finds that women’s rights organizations and women’s rights focused treaties, we explore the cross-national determinants of the criminalization of marital rape. Using an event history analysis covering 131 countries from 1979 to 2013, we find support for the global institutionalist framework that contends that socialization into the global system and direct advocacy efforts of global organizations contribute to faster rates of criminalization of marital rape. Further, we suggest that these global institutionalist processes become amplified when they are focused by events that set the agenda for international organizations. Implications for world-society scholarship on the global adoption of women’s rights are further discussed.

Despite a recent wave of countries enacting laws criminalizing marital rape, rape within marriage remains legal in many countries. In fact, the most recent United Nations (UN) Progress of the World’s Women report (2019) declared that only four in ten countries had criminalized the act. While legislation in some nations simply has not explicitly addressed the legality of the act with regard to existing sexual assault laws, other nations provide explicit legal exemptions for nonconsensual sex that occurs within marriage (UN WOMEN 2011). While legal guarantees for various economic, political-civil and legal rights for women have diffused across the world throughout the twentieth century (Ramirez et al. 1997), private-sphere rights have been slower to adopt even in countries that were early adopters of public-sphere rights. For instance, the United Kingdom did not criminalize marital rape until 1991, while North Carolina, the last U.S. state to do so, did not criminalize marital rape until 1993.

What explains this relatively recent trend toward the criminalization of rape within marriage and why do some countries enact legislation more quickly while others lag behind? While scholars have not explored this phenomenon on a cross-national empirical basis, we argue that sociological theory may lend insight into the criminalization of marital rape across nations. First, we test the explanation that world-society dynamics may be at play because international women’s rights organizations may contribute to the diffusion of criminal penalties for marital rape through the spread of international norms and their ability to advocate for such policies (Frank et al. 2010). While scholars have investigated the effects of active women’s rights organizations on the enactment of women’s suffrage rights (Ramirez et al. 1997) and the structure of rape-law broadly writ (Frank et al. 2009), research has not examined their impact on the less visible issue of marital rape. Second, we explore the effects of women’s rights-related treaties – arguing that when countries become parties to treaties that promote protection from discrimination for women, they are more likely to develop criminal penalties for marital rape. By contextualizing these explanations, we demonstrate how the criminalization of marital rape transformed from a desirable progressive policy to a world society norm and construct a foundation for examining the primary factors that account for trends in cross-national reform.

Scholars of law and society have made interventions into the global institutional paradigm by arguing that there is a complex interplay between human rights institutions, such as treaties, and courts and other legal apparatuses that can provide opportunities to make social change practically possible (Merry 2006; Van der Vet and McIntosh Sundstrom 2023). Scholars argue that activist mobilization around social issues is reflective of a legal opportunity structure that makes social change possible (Comstock 2023; Kahraman 2023; Van der Vet and McIntosh Sundstrom 2023; Vanhala 2012). This opportunity may arise even given the tension between transnational human rights efforts and local social movement actors, as global reform efforts lack the issue-and-place-specific vernacular that is of focus to local movement actors (Merry 2006). This focus on how actors encounter legal opportunities made available by shifts in macro-institutional change can be harmonized with world society scholarship that focuses in part on non-governmental organizations (NGOs) and NGO-affiliated actors who can serve as the agents of social change. Equally crucial as movement actors in the process of transforming legal commitments to on-the-ground legal change is the focus on the ways that international norms are shaped and subsequently implicated in the diffusion of law (Wolf 2020).

This research addresses the global spread of criminal penalties for marital rape by conducting a cross-national analysis of 131 nations from 1979 to 2013 using a unique dataset merged from several available sources and original data. Estimated as an event history analysis, this analysis specifically focuses on timing to criminalization across nations. Our findings suggest that the increased presence of women’s rights organizations is strongly related to faster criminalization of marital rape – here, criminalization referring to both legal change with regard to changes in the text of laws but also implementation through enforcement in practice (Rubin 2019). Similarly, lengthy national membership in the Convention on the Elimination of All Forms of Discrimination against Women (CEDAW), a human rights treaty focused on eliminating discrimination against women, is also strongly related to faster criminalization of marital rape. We find that these dynamics become amplified after the UN General Assembly adopted the Declaration on the Elimination of Violence against Women (DEVAW) in 1993 which transformed criminalizing marital rape from a desirable liberal policy to a world society norm by explicitly codifying it in the text of the declaration. This paper makes an innovative theoretical contribution to world society scholarship by exploring the ways that global institutionalist dynamics, especially gendered dynamics, influence the provision of women’s private-sphere rights which are often hidden from the public arena.

Private-sphere women’s rights

Despite the diffusion of a wide range of gender-based legal protections in many countries, women’s rights advocates argue that family law has been slower to change than other forms of discrimination against women (UN WOMEN 2019). Scholars interested in cross-national analyses of women’s rights have also focused on changes in the public sphere such as measuring changes in women’s political representation (Hughes et al. 2019). It appears as though there is a gap between the diffusion of what might be called women’s public-sphere rights – that is, legal protections intended to incorporate women into the political system through citizenship, access to courts and rights to property ownership – and women’s private-sphere rights – which represent a distinct set of legal protections for women within the familial context. While private-sphere women’s rights such as protections from sexual violence in marriage have lagged behind the adoption of women’s public-sphere rights, some nations have adopted reforms that offer such protections.

Broadly speaking, we define the criminalization of marital rape as the enactment of new legislation that criminalizes nonconsensual sex within marriage or the removal of marital exemptions from existing sexual violence laws. It appears as though countries have adopted criminal penalties for marital rape in several distinct waves. Early adopters such as Poland, Czechoslovakia and the Soviet Union removed marital exemptions for rape as early as 1932 (Poland), though these early laws were never used in criminal proceedings until late 1969 and early 1970 (Warias-Michalska 2016). Other early adopters include nations like Sweden (1965) and Italy (1976), which make up the initial wave of enactments of criminal penalties for marital rape. The end of the Cold War saw a new wave of criminalization diffusing across Western Europe, North America and East Asia in the 1980s and 1990s. Finally, a more recent wave has occurred in the Global South – particularly in Latin America. There is evidence that prominent International Organizations such as the UN and their various arms are applying international pressure in the diffusion process by highlighting the issue of marital rape as a problem of central concern (Randall and Venkatesh 2015: 171).

Scholars suggest that the gap between women’s public- and private-sphere rights exists because public-sphere rights became world society norms much earlier than private-sphere women’s rights (Ramirez et al. 1997). In contrast, norms related to private-sphere women’s rights like the criminalization of marital rape have emerged from “penumbras” or areas of law that may be inferred but are not explicitly codified. Wang and Schofer (2018) introduced this concept after conducting a cross-national analysis of variation in divorce rates. They found that treaties and organizations that promoted individual rights and gender equality contributed to variation in divorce rates, a private-sphere norm, even though they did not explicitly address this right. Examining the historical factors that brought the criminalization of marital rape into the international conversation will provide valuable insights into how it transformed from a desirable liberal policy to a world society norm.

Historical background: criminalization of marital rape

On December 18th, 1979, the UN General Assembly adopted the CEDAW. It is a legally binding international treaty which states that, “States Parties shall take all appropriate measures to eliminate discrimination against women in all matters relating to marriage and family relations and in particular shall ensure, on a basis of equality of men and women.” It includes several specifications related to freely entering into marriage, equal rights during marriage and its dissolution, and equal rights and duties to their children. However, there is no explicit mention of marital rape or sexual violence in marriage.

On June 25th, 1993, the World Conference on Human Rights adopted the Vienna Declaration and Programme of Action which pointed to violence against women as a violation of human rights. It reads, “the World Conference on Human Rights stresses the importance of working towards the elimination of violence against women in public and private life … calls upon the General Assembly to adopt the draft declaration on violence against women and urges States to combat violence against women in accordance with its provisions” (13). This was one of the largest human rights assembly in history with 171 states and approximately 800 NGOs in attendance (Boyle 1995). This call for the elimination of violence against women in public and private life, supported by a substantial number of nations and NGOs, was a critical juncture in solidifying private-sphere women’s rights as a world society norm through international resolution.

On December 20th, 1993, the UN General Assembly adopted the DEVAW. This resolution was designed to complement and strengthen the CEDAW. It speaks directly to “violence against women in the family and in society” and put private-sphere women’s rights at center stage. Article 2 states:

Violence against women shall be understood to encompass, but not be limited to, the following: (a) Physical, sexual and psychological violence occurring in the family, including battering, sexual abuse of female children in the household, dowry-related violence, marital rape, female genital mutilation and other traditional practices harmful to women, non-spousal violence and violence related to exploitation.

This is the first explicit mention of marital rape in a UN resolution, which appeared 14 years after the adoption of the CEDAW. While the DEVAW is not legally binding, it transformed the private-sphere women’s right of protection from violence in the home from a desirable liberal policy that may be inferred from the CEDAW to an explicitly codified world society norm. In addition to stating that gender violence is a human rights violation, it asserts that it is the state’s duty to combat this type of violence (Randall and Venkatesh 2015). In the following sections, we will outline the two possible explanations for why some countries enact legislation that criminalizes marital rape more quickly than others.

Explanations for the diffusion of private-sphere women’s rights

Global norms, international organizations and women’s rights organizations

World Society approaches argue that states become socialized with regard to the norms of the international system through interaction with a global civil society – or a wide range of NGOs (Boli and Thomas 1997; 1999; Simmons 2009). These organizations, often with unique and privileged access to the halls of power within states, have been thought of as advocacy groups or social movements – pressuring, shaming and otherwise manipulating state behavior to institutionalize their norms into the system (Davis and Zhang 2019; Keck and Sikkink 1998). Beyond direct activism, other mechanisms at work in this process are that pathways to legitimate behavior for states in the international system become institutionalized through imitation and coercion (DiMaggio and Powell 1983; Meyer and Rowan 1977) – as well as through the direction of politically powerful actors guiding the international organization’s development toward particular “scripts” or templates for action (Kentikelenis and Seabrooke 2017).

World society theory suggests that countries’ embeddedness in the global community may contribute to the enactment of women’s rights legislation through exposure to Western social norms related to bodily integrity and the influence of international organizations who promote gender equality (Boli and Thomas 1997; Stroup and Murdie 2012). From this perspective, countries are becoming increasingly integrated into a system of international actors – notably international non-governmental organizations (INGOs). INGOs are viewed as important actors in world society diffusion processes due to their ability to advocate and at times shame countries into enacting policy. They contributed to reforms, including the expansion of women’s rights, by developing a universal framework to distribute legal templates as experts, promoting legal reform through the transmission of technical advice and information, lobbying governments, holding international conferences and/or making the push for legal reform in exchange for membership and positive press (Fallon 2003; Frank et al. 2010). For instance, Frank et al. (2009) found that the higher the density of a country’s linkages to world society, the stronger the relationship is between rape-law reforms and police reporting due to the fact that INGOs provide material resources and “receptor sites for unscrambling global signals for local constituencies” (Frank et al. 2009: 277).

More broadly, these actors make up a dense influence network through which global scripts, norms or standards are created and diffused (Boli and Thomas 1997; Meyer et al. 1997a). The more deeply embedded a country is in this network, the more likely it is to adopt these social norms and implement corresponding legislation (Hughes et al. 2015; Schofer and Longhofer 2011). This global-institutionalist framework has become one of the more influential schools of thought in studies of global and comparative social processes, finding broad support for the notion that more dense integration into the global system is associated with a variety of important outcomes of interest.

For instance, international embeddedness influenced the diffusion of universal suffrage during the twentieth century, during which time, women gained the right to vote across most of the globe (Boli and Thomas 1997). After World War II, a greater focus on the rights of the individual and gender equality began to spread as a dominant script (Frank and Mceneaney 1999; Pandian 2019) and laws shifted from protecting collective entities, like family and state, to protecting individualized persons and their bodies (Frank et al. 2010; Frank and Phillips 2013). While the criminalization of marital rape is one issue that has not fully diffused as a universal right across nations, previous research suggests that international embeddedness may play a role in the ongoing diffusion process among countries that have criminalized marital rape to date (Frank et al. 2010; Frank and Moss 2017).

A variant of this theory suggests that more national connections to specific issue-focused INGOs, rather than INGOs broadly writ, should be particularly implicated in global diffusion processes. For example, a well-developed literature has suggested that human rights focused INGOs such as Amnesty International play a role in affecting countries’ human rights performance by publicizing government abuses against populations that can potentially threaten a country’s standing in the international community, and by proxy, their access to goods such as foreign capital investment (Barry et al. 2013; Murdie and Davis 2012). Scholars have also explored the role of INGOs that specifically focus on advocacy for lesbian, gay, bisexual and transgender rights, demonstrating links between these specific issue-focused INGOs and the enactment of progressive policy toward sexual minorities (Velasco 2018).

Women’s rights focused INGOs (WINGOs) can positively influence the passage of legislation related to specific political and social rights for women (Cherif 2010; Cole 2013; Htun and Weldon 2012; Murdie and Peksen 2015; Paxton et al. 2006; Simmons 2009). Since the 1970s, WINGOs have grown dramatically and now represent a significant advocacy force in the international arena (Berkovitch 1999). While scholars suggest that there is a strong positive relationship between nations’ international organizational memberships and the number of women’s rights organizations (Frank et al. 2009), WINGOs may have a unique effect on the passage of legislation related to women’s rights. For instance, WINGOs may contribute to women’s focused policy implementation by forming dense networks of activists that can leverage organizational capacity into social movements and civic activity that can, in turn, assert direct pressure upon governments to push for policy change (Paxton et al. 2006). WINGOs provide an outlet for political participation for a variety of actors, including women who have been historically marginalized in the political process.

Women’s participation in civil society organizations may contribute to the enactment of women-centered legislation because participants are able to translate their civic participation into political capital by interacting with the government bodies that enact and enforce legislation. Feminist activists and members of women’s organizations advocate for legal reform by lobbying the government (Gornick and Meyer 1998), holding demonstrations (Fallon 2003) and writing newspaper and magazine articles (Rinaldo 2008). In fact, scholars have found that feminist activists’ lobbying of the government played an important role in the reform of rape laws that took place during the 1970s (Gornick and Meyer 1998). After legislation is enacted, organizations continue to contribute to an increased level of accountability by acting as watchdog groups (Frank et al. 2009). For instance, Neumann (2017) finds that feminist organizations in Nicaragua provide female victims of domestic abuse with legitimacy when navigating through legal institutions.

Women’s participation in WINGOs also empowers women as individuals in their local communities. Research shows that participating in women’s organizations may help to educate women about their rights (Kunovich and Paxton 2005; McVeigh et al. 2003) and gender equality (Rinaldo 2008), provide support services (Neumann 2017) and activist training (Rinaldo 2008) and give women the ability to pool their resources (Fallon 2003). Local groups are especially important in nondemocratic states where women are excluded from government and rely on such groups as protected spaces where they can voice their concerns and conceptualize the issues that they want to address (Fallon 2003). The theory that WINGOs positively influence the passage of legislation related to women’s rights leads us to our first major hypothesis:

H1: Nations with more WINGOs should criminalize marital rape faster than nations with fewer WINGOs.

Human rights treaties and INGO enforcement

A related line of research suggests that analyzing commitments made by countries in terms of treaties and international agreements may have a direct impact on states’ behavior. While countries rarely face harsh penalties for not complying with the terms of human rights treaties, leading to a variety of potential responses post treaty adoption, treaties are sometimes effective tools for the adoption of treaty-compliant behaviors across the world through two key mechanisms.

One mechanism is that socialization contributes to nations’ decisions to sign human rights treaties and comply with the terms of those agreements (Cole 2013; Goodman and Jinks 2004). This perspective suggests that countries adopt new behaviors as they gradually recognize norms that are encoded in the text of treaties. This mechanism aligns with Wang and Schofer’s (2018) observation that “international treaties and discourses enshrine general norms” (680).

The second mechanism is that INGOs promote compliance with human rights treaties because they monitor human rights practices and prompt increased mobilization against non-compliant regimes (Keck and Sikkink 1998; Risse et al. 1999; Ron et al. 2005). In this case, nations comply with the agreements set forth in human rights treaties because INGOs promote human rights norms and mobilize citizens when nations fail to adopt these norms. Cole (2013) finds that countries that had adopted the CEDAW for longer periods of time had stronger women’s political rights and the presence of WINGOs strengthened this effect. The theory that countries adopt the norms related to treaties they are party to allow us to suggest a second set of hypotheses:

H2a: Nations that are parties to the CEDAW treaty should criminalize marital rape faster than countries that are not parties of the CEDAW treaty.

H2b: Nations that have been under the CEDAW treaty for more years should criminalize marital rape faster than countries that have been under CEDAW for fewer years.

Research design

We make use of a merged dataset from several publicly available sources and novel measures based on our original data collection and coding. The timeframe of the analysis ranges from 1979 to 2013, a period in which the vast majority of laws criminalizing marital rape were passed, and covers 131 countries. The data are structured to constitute a risk-set, that is, a country-year set is included in years in which countries might potentially adopt criminalization. We right-censor observations once a country is no longer at risk of adoption. While it is theoretically plausible that a country may reinstate impunity measures once they previously criminalized marital rape, we do not observe such an instance in our measure of marital rape criminalization. Because we begin our analysis in 1979 to coincide with the UN General Assembly’s adoption of the CEDAW treaty (1979), a small number of cases are left censored (USSR, Poland, Italy, Norway and Sweden).¹ Table 1 below provides descriptive statistics for all variables in our analysis.

Table 1. Descriptive statistics for variables in analysis

Note: Statistics calculated based on country years at risk.

We use event history analysis to model the effects of our variables on a country’s likelihood of criminalizing marital rape in a given year or their hazard rate. Event history techniques estimate the relationship between our various predictors and the rate at which country i experiences an event j given that the country has not yet experienced event j (Allison 1984; Tuma and Hannan 1984). We estimate this rate as an exponential “constant rate” model which holds the hazard rate as a constant function of time but allows it to vary according to changes in predictors (Cole 2005). This modeling choice follows standard procedures in the world society tradition (Cole 2005; Frank and Hironaka 2000; Hironaka 2002; Koo and Ramirez 2009; Meyer et al. 1997b; Ramirez et al. 1997; Schofer 2003; Swiss 2009; 2012) and is appropriate given that there is no a priori justification to assume that time has an effect on the hazard rate that is independent of the covariates in the model (Cole 2005; Swiss 2009). Finally, an inspection of the baseline hazard rates strongly suggest empirical support for the exponential “constant rate” estimation procedure.

The full exponential rate model procedure treats time as continuous and takes the form:

\begin{equation*}\it{log}\left[ \it{r\left( t \right)} \right]\, = \,\it{B'X}\end{equation*}

where r is the rate at which countries adopt criminalization policies, X is a vector of theoretically relevant predictor and control variables and B represents the coefficients for these covariates (Swiss 2012). An exponential term on both sides of the equation yields the rate of criminalization and the effect of each covariate on this rate (Ramirez et al. 1997; Swiss 2012).

We begin by presenting three models testing for the effects of each of our focal predictor variables as well as theoretically relevant control variables. Model 1 includes only the (logged) number of Women’s Rights INGOs as well as all control variables, Model 2 introduces a dichotomous indicator for whether or not the CEDAW was ratified in a country and finally Model 3 includes the count of years since the CEDAW was ratified in a country to examine the long-term impact of ratification. Because CEDAW ratification and the length of time since ratification are highly collinear (r > 0.8), we do not include these in the same model. These models include jackknife estimation procedure as a robustness check against the threat of outliers, as well as robust standard errors clustered on country-code. Since the UN transformed criminalizing marital rape from a desirable liberal policy to a world society norm by explicitly codifying it in the text of the DEVAW in 1993, we estimate a final set of models first accounting for the “pre-DEVAW” time period (1979–1992) and the “post-DEVAW” time period (1993–2013). All models include heteroskedastic robust standard errors clustered on a numerical country code to correct for autocorrelation.

Dependent variable

Our focal dependent variable is a binary indicator for whether or not a nation has criminalized marital rape, instituting criminal legal penalties for the act that are effectively enforced. More specifically, we define criminalization of marital rape based on a standard that excluded policy in name only. As such, criminalization includes either (a) explicit policy adopted through the enactment of new laws that are effectively enforced, (b) the removal of marital exemptions from existing sexual violence laws that are effectively enforced, or (c) in rare cases, prominent courts will offer reinterpretations of existing sexual violence laws removing marital exemptions. While there is some variation in this process, many cases in which marital rape has not been criminalized are clear-cut as nations often retain explicit exemptions for marital rape with legal language codifying this status with wording suggesting that rape is impossible within marriage as the institution implies consent.

We coded this variable using a variety of sources including annual editions of the UN Progress of Women report (UN WOMEN 2011–2011; UN WOMEN 2018–2019) that often include information as to whether such a law existed or not, though these often include no information on timing of legislation. We supplement this initial source with a variety of publicly available official documents and media sources to collect full information on timing of criminalization of marital rape across the 131 countries in our study. Inconsistencies across these reports were meticulously investigated and ultimately resolved. In sum, we identified 73 cases of full criminalization and effective enforcement across the 131 countries included in this study.

We made several coding decisions that we describe below. We developed a relatively high threshold for considering a nation to have fully criminalized marital rape. For instance, several nations have in recent decades removed explicit marital exemptions from existing sexual assault laws, yet reputable reports have found that the crime was never prosecuted and treated with skepticism from legal actors. For federalist political systems such as Australia and the United States, we include the year when the last state or subnational division criminalized marital rape. We coded countries as having criminalized marital rape in cases when laws were enacted that unambiguously criminalized marital rape either by treating it as an aggravating circumstance, special category or the inclusion of explicit language outlawing the act. In several cases, criminalization was achieved through a high court’s ruling that resulted in effective prosecutions for the crime going forward. This procedure represents an advance for Law and Society scholarship in that our coding procedure is sensitive to the variability in the types of avenues that criminalization of marital rape may take, a technique that may be repeated in other studies of implementation.

We plot the overall cumulative survival and hazard rates for the countries in our study in Figures 1 and 2. The plot of the survival rate visualizes the cumulative proportion of countries that retain legal impunity for marital rape – demonstrating a precipitous decline in recent decades. The plot of the hazard rate reveals a moving estimation of risk of criminalization of marital rape that increases in recent years.

Figure 1. Overall survival estimates for impunity for marital rape (1979–2013).

Figure 2. Overall hazard estimates for the criminalization of marital rape.

Independent variables

To address the expectations in our first hypothesis, we test the extent to which the presence of WINGOs is related to the rate of criminalization of marital rape, following previous literature on this topic (Cole 2013; Hughes et al. 2017b; Hughes et al. 2018; Murdie and Peksen 2015). Our primary data source is the imputed WINGO count provided by Hughes and colleagues (2017b) although we supplement this file to include information from original coding based on the Yearbook of International Organizations (Union of International Associations (UIA) Various years). Because the WINGO data are publicly available from Hughes and colleagues (2017b) at non-annual intervals, we follow convention and use linear interpolation to estimate values for missing years (Nyseth Brehm and Heger Boyle 2018). As inspection of this variable reveals positive skew, we use the natural log (ln) of this measure in all estimations.²

We explore our second hypotheses using two measures related to the CEDAW. First, we include a measure of whether or not a country had ratified the CEDAW treaty (1) or not (0), with subsequent years also coded as “1.” Second, we include a cumulative measure of the running total of years since ratification of the CEDAW treaty. Both of these measures are inspired by Cole’s (2013) work on treaty effects and are coded based on the list of signatories and ratifiers provided by the UN’s online Treaty Collection (treaties.un.org). Following convention, we include countries as parties to the treaty once they have ratified or otherwise ascended to treaty membership.

Controls

In addition to our independent variables that we theorize should affect the timing to criminalization of marital rape, we include several control variables that we expect may constitute threats of spuriousness for our focal associations. Save for the regional controls we will discuss shortly, these variables are included in our final full model, a model that presents the net effects of our focal predictors given the effects of these control variables. Each of these measures can potentially covary with our outcome of interest and our predictors based on associations found in related research and thus should be included in estimations.

As we seek to account for the possibility that broad gender relations may affect the rate of criminalization of marital rape. To capture this dynamic, we first include the female labor force participation rate (World Bank 2020). This measure is interpolated because it is collected at non-annual intervals for many years. In addition, we include infant mortality rate aggregated by the Varieties of Democracy (V-Dem) project in the V-Dem + others dataset (V 10) but originally sourced from CLIO-INFRA’s historical database (Zijdeman and de Silva 2014; Aboubarb and Anessa 2007). This measure presents a potential threat to spuriousness as it is a measure of development commonly associated with women’s rights and has clear implications regarding reproductive health that may also be reflective of attitudes towards women’s bodily integrity.

Beyond these broad measures of gender relations, scholarship on women’s political leadership shows that increasing women’s representation in public office may potentially contribute to widespread political and societal change (Franceschet et al. 2012), including the enactment of women’s rights legislation. At the global level, research has shown that institutional arrangements regarding women’s representation may promote progressive policy (Fallon et al. 2012; Hughes et al. 2015; 2017a; 2019). Quota reforms, in particular, or policies that ensure that women hold a certain percentage of the legislative seats, have proliferated in recent years ensuring improvements in official representation for women in politics (Hughes et al. 2017a). Cross-national research has demonstrated that quotas have become more effective in increasing women’s representation over time (Paxton and Hughes 2015). Access to powerful lawmaking positions may put women in a better position to advocate for and enact private-sphere women’s rights.

To control for the potential impacts of women’s representation in government, we include two indicators drawn from separate data sources. First, we include a dichotomous measure of whether or not the head of state is female (1) or male (0) aggregated in the V-Dem 10 dataset (Coppedge et al. 2020) but originally sourced from Melander (2005) and Paxton and Hughes (2008). This measure captures whether or not women are represented at the highest levels of government. Second, we include a dichotomous measure of whether (1) or not (0) a nation had a gender quota for their legislature, also gathered from the V-Dem 10 dataset in the indicator “v2lgqugen” Lower chamber gender quota. This measure assesses the presence and type of gender quota that exists in each country. Originally, this measure is presented as an ordinal scale from “0” (no gender quota) to “4” (reserved seats for women), with intermediate values representing whether or not the existing gender quota provided sanctions for non-compliance. We recode this variable as a dichotomous variable because the vast majority of country-years in our study have no form of gender quota whatsoever. As a result, the meaningful comparison here should be between those with any form of gender quota and those with no gender quota at all.

We include several alternative factors that may account for a nation’s cultural and social position in the global system. First, we include a categorical variable for the largest religion among a nation’s population. Following Swiss (2012), we use four religious categories (Protestant, Catholic, Islam and Other). Data for this variable are gathered from the most recent available year from the World Religion Project dataset (Maoz and Henderson 2013). In addition, we include the widely used alternative indicator for globalization through the KOF social globalization index (Gygli et al. 2019).³ Next, as several early adopters were communist systems, we include a dichotomous variable that captures whether or not the country had a communist system.

We also include an indicator of development in the logged gross domestic product per capita (GDP per capita). These measures are also included in the V-Dem + others dataset but are originally sourced from the Maddison Historical Statistics database (Bolt and van Zanden 2014; Coppedge et al. 2020). These measures are potentially relevant as previous research on the adoption of sex-related criminal laws (Frank and Moss 2017), and laws protecting sexual minorities (Velasco 2018) have found economic development to be at least marginally related to the various outcomes of interest. Another global factor that is accounted for in our models is whether or not the United Kingdom/Great Britain was the primary imperial power that colonized the country in question – here, scholarship has described how many Commonwealth countries maintained formally codified colonial era marital rape exemptions that remain as an enduring effect of colonial legal structures (Anderson 2016; Elvy 2015).

An additional variable that may potentially confound the relationship between our focal variables of interest is a country’s level of democratization. World society scholarship routinely employs indicators for the level of democratic quality in countries as adhering to bureaucratic templates in such a way that the form may confound the effects of global-institutional factors of interest. For our particular study, we also consider the potential that more democratic countries will have stronger protections with regard to bodily integrity rights – an issue clearly implicated in legal considerations of sexual assault. Additionally, previous research has shown democratic quality to be related to stronger penalties for rape (Frank and Moss 2017). To account for democratic quality, we make use of democracy scores that provide information on regime characteristics provided by the V-Dem dataset, specifically with regard to their variable “v2x_polyarchy” which corresponds to the level of electoral democratic institutional strength in a country – here, we follow tradition in cross-national scholarship and code countries scoring above “0.5” on the polyarchy scale as “democracies” (Davis et al. 2022; Davis and Zhang 2019; Lührmann et al. 2018). Because previous research has found a negative association between countries’ total population and national adoption of liberal laws regarding same-sex relations (Frank and Mceneaney 1999), we have included a measure of the logged population sourced from the World Bank (2019).

Finally, we include dichotomous variables for regional effects in all of our models. An examination of hazard plots by region suggests that there are notable regional effects that need to be parsed out to assess the associations of our focal predictors and the rate of criminalization of marital rape. These regional “dummies” follow the logic of a fixed-effects estimator by accounting for within-unit unobserved heterogeneity (Allison 2009; Hill et al. 2020). In this case, our models control for threats associated with unobserved heterogeneity at the regional level.

Results

In the section that follows we present the results of our event history analysis that has estimated the rate of criminalization of marital rape across 131 countries from 1979 to 2013. All results are summarized in Tables 2 and 3 in the text that follow.

Table 2. Event history models predicting the effects of women’s rights organizations and CEDAW treaty ratification on the criminalization of marital rape

Note: Jackknife S.E. in parentheses. S.E. clustered on numerical country identifier. Regional effects included in all models.

*** p < 0.001,

** p < 0.01,

* p < 0.05,

Table 3. Comparison of event history estimates marital rape (pre- and post-DEVAW)

Note: Jackknife S.E. in parentheses. S.E. clustered on numerical country identifier. Regional effects included in all models.

***p < 0.001,

** p < 0.01,

* p < 0.05,

+ p < 0.1 (two-tailed tests).

Results from our full model provide evidence that country-level connections to WINGOs are related to a faster rate of criminalization of marital rape. We find that the (logged) number of WINGOs that have memberships in countries relates to an increased hazard of criminalization of marital rape in Model 1 (β = 5.3, p < 0.001) but also in Models 2 (β = 2.38, p < 0.05) and 3 (β = 3.49, p < 0.001) that account for CEDAW ratification and years since CEDAW ratification. Findings, here, show a large and statistically significant effect of the number of WINGOs in a country on the country’s timing to criminalization of marital rape. In addition, we find that CEDAW treaty membership is similarly related to faster criminalization of marital rape. CEDAW ratification is related to the criminalization of marital rape in Model 2 (β = 5.43, p < 0.001), indicating that countries that ratify the CEDAW criminalize marital rape faster than those that do not. Second, referring to Model 3, we find that the number of years since CEDAW ratification carries a positive and significant association with the hazard of criminalization (β = 0.13, p < 0.001).⁴ In Figures 3 and 4, we plot the cumulative hazard rate across values of logged WINGO counts and CEDAW ratification or not. Here, we dichotomize WINGOs at the mean with values falling below coded as “low” WINGOs and those above as “high.”

Figure 3. Cumulative hazard estimates for logged WINGOs.

Figure 4. Cumulative hazard estimates for CEDAW ratification.

Models 4 through 7 reveal that the main effects illustrated in our full models in Table 2 are driven in part by dynamics in the post-DEVAW period. We begin by encouraging cautious interpretation of estimates based on the pre DEVAW period as the total number of criminalization events (16) is quite small yielding large effect sizes. Here, in the pre-DEVAW period, only CEDAW ratification yields significant effects on the speed of criminalization. Yet a focus on the post-DEVAW period reveals that the logged number of WINGOs (Model 6: β = 3.33, p < 0.01, Model 7: β = 2.78, p < 0.01), CEDAW ratification (β = 11.98, p < 0.05) and the number of years since CEDAW ratification (β = 0.12, p < 0.05) are positively related to the rate of criminalization of marital rape in the post-DEVAW period. This finding allows us to suggest that global institutionalist forces related to international organizations as well as treaty-effects became more important once the issue of marital rape was brought to the forefront of the global stage through the UN’s adoption of the DEVAW. Likewise, these findings are reflective of the general silence of the global community on the issue prior to DEVAW.

Discussion and conclusions

Our results strongly support our first hypothesis reflecting the institutionalist tradition that countries with more women’s rights organizations (WINGOs) criminalize marital rape faster than countries with fewer WINGOs. The effect of WINGOs on the hazard of criminalization is present in our reduced model and persists in our full model. These findings suggest that WINGOs are effective actors in promoting private-sphere women’s rights, such as the criminalization of marital rape, in addition to the political rights documented by previous research (Cole 2013; Murdie and Peksen 2015). In addition to viewing INGOs more generally as carriers of global norms that can diffuse through the international system, WINGOs have been shown to engage in advocacy efforts as well as naming and shaming behavior to improve women’s rights across the globe (Murdie and Peksen 2015). Linking this institutionalization process to the criminalization of marital rape suggests that this issue has risen to the forefront of international organizations’ information and advocacy efforts.

We also find consistent support for treaty effects in this study. Both the dichotomous indicator for CEDAW treaty membership and the cumulative years of membership have a robust association with the hazard rate of the criminalization of marital rape. CEDAW members and countries that have been members of the CEDAW treaty for longer tend to criminalize marital rape faster than countries that have been members for fewer years or are not members at all. This finding provides support for the socialization perspective which argues that over time the norms that are encoded in the text of human rights treaties, such as CEDAW with respect to women’s rights, become increasingly associated with countries’ actual behavior (Cole 2013; Goodman and Jinks 2004; Merry 2006; Swiss 2012).

Taken together, these findings offer evidence that global institutional processes influence the enactment of sex-related laws (Frank et al. 2010; 2009; Frank and Moss 2017; Velasco 2018) and women’s rights more broadly conceived (Cole 2013; Murdie and Peksen 2015; Swiss 2009; 2012; Velasco 2018). WINGOs and women’s rights-related treaties such as the CEDAW serve as vehicles for the global institutionalization of ideas and norms related to bodily integrity that extend to at-home life. This process departs from accounts of institutional decoupling where leaders make commitments to international human rights law, yet fail to abide by the obligations set forth in those venues (Cole and Ramirez 2013; Hathaway 2002; Swiss 2009). It also demonstrates the efficacy of the CEDAW and related international agreements, such as the DEVAW declaration, in eliminating certain forms of violence against women. Despite this perhaps optimistic finding, the effects of treaty membership are not universal – and there are many cases of CEDAW members failing to abide by their international commitments.

Our final models offer some evidence that global institutionalist forces become activated when the international community bring an issue to light. The overall rate of criminalization increases significantly after the UN General Assembly adopted the DEVAW declaration-addendum to the CEDAW treaty. Additionally, global forces such as the number of WINGOs in a country and connections to the CEDAW treaty became more important during the post-DEVAW period. We suggest that issue-raising events such as the adoption of the DEVAW may become important agenda-setting moments that can work to assist in the diffusion of private-sphere rights that have lagged behind women’s political empowerment. In addition to raising the issue of marital rape to the UN General Assembly, it explicitly codified marital rape as a type of domestic violence covered under the CEDAW treaty, demonstrating the reach of international institutions into private-sphere women’s rights (Wang and Schofer 2018).⁵ In line with legal scholarship on legal change (Comstock 2023; Kahraman 2023; Van der Vet and McIntosh Sundstrom 2023; Vanhala 2012), this context shaped the legal opportunity structure and made activists and world-society institutions more effective in advocating for penalties for marital rape.

While we find strong support for the global institutionalist framework, we find less support for national level factors. Upon an examination of cases, this result may not be particularly surprising as countries in Western Europe and North America that lead in the enactment of criminalization very rarely had gender quotas or female heads of state in the 1980s and 90s when this wave of criminalization occurred (e.g. Canada in 1983, Ireland in 1990, the United Kingdom in 1991, Spain via court order in 1992, the United States in 1993, France in 1994 and Germany in 1997). While many of these countries have had female heads of state (Canada, the UK, France, Ireland, Spain and Germany at different periods), none of their terms coincided with the criminalization of marital rape and many served their terms long after criminalization occurred. Likewise, many early adopters did not have any form of national legislative gender quota at the time of or before criminalization, though recent trends have shown that more recent cases have had national gender quotas, such as Mexico, Honduras and Venezuela.

This article contributes to several ongoing lines of research exploring the links between international organizations, especially women’s rights organizations, and increasing women’s rights practices around the world (Cole 2013; Murdie and Peksen 2015; Swiss 2009; 2012). We move beyond work that centered on public-sphere political, economic and social rights to focus on women’s rights in the private sphere. Specifically, we focus on the enactment of criminal penalties for rape within marriage, asking what global and national characteristics are related to faster criminalization. This contribution offers a call for scholarship in Law and Society to consider the global dimensions of domestic legal change. Here, our findings reveal how domestic laws can be derived from global processes and interactions with global institutions, rather than being seen as solely the result of domestic pressures and interests.