twitter

Wednesday, 15 June 2016

Intimate partner violence against women and the Nordic paradox

Social Science & Medicine
Volume 157, May 2016, Pages 27–30
Short communication


doi:10.1016/j.socscimed.2016.03.040
Get rights and content

Highlights

•
Nordic countries have a high level of gender equality but high prevalence of IPV.
•
This Nordic paradox is a question rarely asked, and one that remains unanswered.
•
This paradox could express country differences in self-reported IPV.
•
The Nordic paradox may provide an avenue to guide new research on IPV.
•
Multilevel analyses of individual heterogeneity will help to understand the paradox.

Abstract

Nordic countries are the most gender equal countries in the world, but at the same time they have disproportionally high prevalence rates of intimate partner violence (IPV) against women. High prevalence of IPV against women, and high levels of gender equality would appear contradictory, but these apparently opposite statements appear to be true in Nordic countries, producing what could be called the ‘Nordic paradox’. Despite this paradox being one of the most puzzling issues in the field, this is a research question rarely asked, and one that remains unanswered. This paper explores a number of theoretical and methodological issues that may help to understand this paradox. Efforts to understand the Nordic paradox may provide an avenue to guide new research on IPV and to respond to this major public health problem in a more effective way.

Keywords

  • Intimate partner violence;
  • Gender equality;
  • Nordic countries;
  • Prevalence;
  • Violence against women;
  • Multilevel analysis;
  • Individual heterogeneity
The most common form of violence suffered by women is intimate partner violence (IPV) (Devries et al., 2013, European Union Agency for Fundamental Rights, 2014, Garcia-Moreno et al., 2006, Stöckl et al., 2013 and World Health Organization, 2013). With a global prevalence of 30%, and with a proportion of murdered women killed by a partner of 38.6% (Devries et al., 2013 and Stöckl et al., 2013), IPV against women remains a major public health problem worldwide (Campbell, 2002, Ellsberg et al., 2008 and World Health Organization, 2013). Gender inequality, has been considered a main factor explaining rates of IPV against women and, accordingly, increasing gender equality is a main target to reduce this major public health problem (García-Moreno et al., 2015, Heise, 2011, Jewkes, 2002 and Jewkes et al., 2015). High prevalence of IPV against women and high levels of gender equality would appear contradictory, but these apparently opposite statements appear to be true in Nordic countries, producing what could be called ‘the Nordic paradox’. Despite this paradox being one of the most puzzling issues in the field, interestingly, this is a research question rarely asked (Lundgren et al., 2001), and one that remains unanswered.

1. A Nordic paradox?

According to The Nordic Council of Ministers (2016), “gender equality refers to the equal rights, responsibilities and opportunities in every area of life of women and men and boys and girls. It means that every person - regardless of sex - have equal power and influence in society”. Equality between women and men is a fundamental value in the Nordic countries, which has contributed towards making the Nordic region the most gender equal region in the world today. A number of international indicators support the view that Nordic countries are the most gender equal countries in the world. For example, the three European Union (EU) Nordic countries (Sweden, Finland, and Denmark) have the highest Gender Equality Index, an index developed by the European Institute for Gender Equality (EIGE) based on six core domains (work, money, knowledge, time, power, and health), and allowing comparisons among all EU member States (EIGE, 2012). For example, the EU Nordic countries have indexes between 70.9 and 74.2, with a EU mean of 52.9 (the EU lowest index is 33.7). Also, Iceland, Norway, Finland, and Sweden are, according to the Global Gender Gap Index (Word Economic Forum, 2015), the countries with the smaller gap between men and women measured in four categories: economic participation and opportunity, educational attainment, health and survival, and political empowerment. Finally, according to the Gender Development Index (United Nations Human Development Reports, 2015) an index based on three dimensions of human development: health (life expectancy at birth), education (years of schooling), and command over economic resources (female and male estimated earned income), Nordic countries all achieve scores (ranging from 0.975 to 0.999) close to one (one representing absolute equality).
Despite these high levels of gender equality, however, Nordic countries have, at the same time, surprisingly high prevalence rates of IPV against women. A 2014 survey conducted by the European Union Agency for Fundamental Rights among the 28 EU Member States showed that the lifetime prevalence of physical and/or sexual violence against women by intimate partners in Nordic countries members of the EU was among the highest (FRA, 2014). With a EU average of 22%, and 13% being the lowest prevalence, Denmark with 32% (highest prevalence in the EU), Finland with 30%, and Sweden 28%, were clearly above average. This was also the case for psychological partner violence. In a separate survey conducted in Norway (a non-EU member), lifetime prevalence of IPVAW was 26.8 (Nerøien and Schei, 2008). In Sweden, another national survey showed that more than one in four women have at some point in their lives been victimized in a close relationship (Brå, 2014). When violence against women by non-partners is also considered, another study of women attending gynecology departments in the five Nordic countries (Denmark, Finland, Iceland, Norway, and Sweden) found that lifetime prevalence of violence against women ranged across countries between 38 and 66% for physical abuse, 19–37% for emotional abuse, and 17–33% for sexual abuse (Wijma et al., 2003). Also the EU survey showed that the lifetime prevalence of violence against women by non-partners was also the highest among Nordic countries (ranging between 46 and 52%). Official crime data for the period 2007–2011, showed that Scandinavian countries had among the highest rates of sexual assault compared to other European countries (Aebi et al., 2014). In general, available studies support the idea of a high prevalence of violence against women in Nordic countries (Aebi et al., 2014, Bra, 2014, European Union Agency for Fundamental Rights, 2014, Heiskanen and Piispa, 1998, Lundgren et al., 2001, Nerøien and Schei, 2008 and Wijma et al., 2003).
This data illustrating the Nordic paradox clearly posits a puzzle not only to researchers, but also to policy makers and program planners aiming to reduce violence against women. Why in countries, like the Nordic ones, where the gender gap in areas such as economic participation, educational attainment, health or political empowerment has been almost closed (Word Economic Form, 2015), prevalence of IPV against women has not been substantially reduced but rather remains remarkably high? So far, there is no clear answer to this question, and the challenge that posits the Nordic paradox for IPV researchers is to make sense of it.

2. Gender equality and IPV risk

Available research suggests that gender inequality is related to higher risk of IPV victimization for women, in particular in low and middle-income countries, and that women victimization is expected to decrease as gender equality increases, as it is the case in high-income countries (Archer, 2006). Data for Nordic countries appears to contradict this argument. As the Nordic paradox implies, it appears to be a link between gender equality and IPV prevalence but in the opposite direction than expected. When Nordic data is compared to other EU countries it emerges a positive relationship between country-level gender equality and prevalence of IPV against women. For example, countries like Portugal, Italy or Greece, with IPV prevalence rates of 19%, have all Gender Equality Indexes more than 30 points lower that Nordic countries, which in turn have substantially higher IPV rates (between 9 and 14 percentage points higher).
As to how country-level gender equality may modify the influence of other variables on the risk of IPV against women, available research suggests several possibilities. For example, studies in low and middle-income countries find that women's increased economic or educational status may either protect them or increase their risk of IPV victimization, depending on the cultural and interpersonal context (Abramsky et al., 2011, Heise, 2011, Jewkes, 2002 and Vyas and Watts, 2009). Women reaching a higher status in relation to their partners, and against a background of traditional and rigid gender norms, can be at higher risk of IPV victimization. But, what in the other end of the continuum of gender equality? How could high country-level gender equality provide a background against which the individual-risk for victimization or perpetration may be increased in particular contexts (e.g., interpersonal, group, community), resulting in higher overall prevalence levels than other countries with lower gender equality? We still do not know, but exploring a number of theoretical and methodological issues may help to better understand this puzzle.
Could there be unanticipated consequences of high gender equality for IPV risk? Could strong status discordance between partners (based on differential educational or economic attainment), or isolation from mainstream gender values and social norms against a background of high structural levels of gender equality generate excess risk for IPV against women victimization/perpetration? For example, some research in high-income countries suggests that women with higher economic status relative to their partners can be at greater IPV risk depending on whether their partners hold more traditional gender beliefs and expectations (Atkinson et al., 2005 and Macmillan and Gartner, 1999). Could this effect be exacerbated in highly equal countries like the Nordic ones? Do perceptions and expectations regarding womanhood and manhood at the individual level clash with strong normative gender equality? (Jewkes, 2002 and Jewkes et al., 2015). It has been suggested that in high-income countries with high levels of gender equality, women become more agentic, moving away from traditional gender stereotypes, and displaying more dominant traits like being directive and competitive (Archer, 2006 and Rudman and Glick, 2001). It has also been suggested that increased gender equality can create a backlash effect, with negative perceptions and responses (e.g., discrimination) against women in areas like managerial positions (Rudman and Glick, 2001). Could there also be a backlash effect in terms of higher IPV risk? Can a context of high gender equality, where women tend to be more agentic, reinforce victim-blaming attitudes, by which victimization can be perceived as deserved and perpetration excused or justified, contributing to a social climate where IPV against women is more tolerated, and thus more prevalent? (Gracia, 2014 and Gracia and Tomás, 2014). Within the Nordic countries, could risk of IPV against women be higher in some socio-demographic or ethnic groups where there may be a wider discordance between country structural norms and individual-level beliefs regarding gender? Is the effect of known individual level predictors of IPV modified by the country context as a whole? If this is the case, what is the role of contextual gender equality?

3. A multilevel perspective

IPV is a complex multilevel phenomenon (Heise, 1998, Heise, 2011, Heise and Kotsadam, 2015, Jewkes, 2002 and Jewkes et al., 2015), and its adequate investigation needs to rely on multilevel analytic approaches to take into account the intricacy of multilevel influences and across level interactions. After all, IPV occurs not only at the macro level but mostly at the interpersonal level, were the influence of risk factors, either for victims and perpetrators may be exacerbated or reduced by other factors working at different levels (from the most proximal to the most distal). We need to know the relative contribution of all relevant levels (e.g., interpersonal, peer group, community, and country) when it comes to understand individual IPV risk.
We need a better understanding of individual heterogeneity of responses rather than only rely on differences between country averages. Investigating between country differences in average IPV risk, as well as the association between contextual gender equity and IPV risk based only on country level aggregated information is a justifiable first analytical step for generating hypothesis. Relying only on this type of analyses is, however, inappropriate not only because the peril of the ecological fallacy (i.e., inferring that in the Nordic countries the most equalitarian partners are also the most violent against women) (Morgenstern, 1998), but also because this type of analyses may be misleading even if the inference is pretended to be at the country rather than at the individual level (Merlo et al., 2009). What matters in Public Health is not only to quantify differences between averages, but also to understand the individual heterogeneity around the averages by performing appropriate multilevel analyses (Merlo, 2014 and Merlo et al., 2009). That is, we need to know the share of the individual variance that can be found at the country level. The relevant question is if knowledge on women's country of residence helps us to discriminate with accuracy the women who suffer IPV from those who not. If this discriminatory accuracy is low, the average differences may provide inaccurate information for decision makers and even unnecessary stigmatization of individuals from certain countries of birth, a situation that has been denominated the “tyranny of the means” (Merlo, 2014).
Individual heterogeneity in IPV risk is certainly a complex phenomenon and for a better understanding we should decomposed this heterogeneity in different cross-classified and multiple membership levels including for instance the household, proximal social networks, the neighborhood, the work place as well as the region and the country where the individual is living. A simple multilevel analysis seeking for cross-level interactions by measures of association (i.e., fixed effects) would be conceptually insufficient (Merlo, 2014). The multilevel analysis of individual heterogeneity is therefore closer to complex system thinking and its need for computational modeling (Dammann et al., 2014, Diez Roux, 2011 and Huang et al., 2009). In this regard, future research should also consider complex systems approaches for analysis (e.g., system dynamics, network analysis, and agent-based modeling) to better understand the factors contributing to the gender equality-IPV association in Nordic and non-Nordic countries.
A better understanding of both between and within-country variations in IPV prevalence with complex analyses that take into account the relative influence of multiple factors working at different levels (from individual to macro-levels), but also the complexity of their cross-level interactions need to be taken into account for the design of more effective prevention and intervention strategies (Heise and Kotsadam, 2015).

4. Is the Nordic paradox just the expression of confounding or information bias?

Prevalence comparisons across settings or countries raise issues of comparability (Garcia-Moreno et al., 2006 and Hindin et al., 2008; WHO, 2013). However, the FRA survey from where we draw our data for Nordic countries prevalence comparisons is one of the few surveys using the same questions and methods, allowing comparisons among all 28 EU Member States. Although showing important variations between countries, the FRA survey tend to be in line with national surveys from EU Member States (FRA, 2014), and also shows less variations in rates of violence than other international surveys with comparable data like the WHO Multi-country study on Women' s Health and Domestic Violence against Women (Garcia-Moreno et al., 2006). On the other hand, it has been argued that different interpretation of survey questions across countries, the acceptability to talk to others about IPV victimization, or higher levels of disclosure in countries with higher gender equality, could explain differences across countries beyond real differences in prevalence (FRA, 2014). Regarding different interpretations of survey questions, while this circumstance could be a concern for subjective variables like self-reported health, it seems less relevant for measuring IPV. In the FRA survey, IPV is measured by objective (behavior oriented) questions, which prevent different interpretations across countries (e.g., the survey ask about whether one has been hit, burned, stabbed, cut or forced to have sexual intercourse).
It has also been argued that the higher prevalence of IPV in countries with high gender equality, like Nordic countries, may be just reflecting the fact that women in these countries feel freer to talk about their victimization leading to higher levels of disclosure (FRA, 2014). Therefore, this could be a potential information bias behind the observed Nordic paradox. Data would reflect not an actual higher prevalence but higher levels of disclosure than in less equalitarian countries. However, the same FRA survey provides data suggesting lower levels of disclosure of IPV to the police by women in Nordic countries as compared to other EU countries. For example the average percentage for the EU of women indicating that the most serious incident of IPV came to the attention of the police is 20%, whereas for Denmark and Finland is 10% and 17% for Sweden (FRA, 2014). In any case, the ‘higher disclosure’ explanation, however, would not solve the Nordic paradox, as these more ‘reliable’ levels of disclosure would rather reinforce the paradox posited by very high levels of IPV prevalence (prevalence rates around 30% is by all means disproportionate) in countries with high levels of gender equality. Other data regarding violence against women by non-partners, including rape, as well as data regarding levels of acceptability and victim-blaming attitudes in cases of violence against women in Nordic countries (Aebi et al., 2014, Gracia, 2014, Gracia and Lila, 2015 and Wijma et al., 2003), also support the view of high prevalence of IPV against women in Nordic countries. Finally, it could be possible that gender equality has nothing to do with the high prevalence of IPV in Nordic countries. It could be possible an explanation based on a third variable not taken into consideration. Nordic countries are similar in levels of gender equality, but they may also be equal or share other characteristics that increase IPV, like for example different drinking patterns (FRA, 2014). Clearly further research, including qualitative research, is needed to gain a better understanding of the social context in which IPV occurs. We cannot rule out the possibility that the Nordic paradox is just the expression of confounding or information bias, but at the same time we cannot rule out the possibility that the Nordic paradox reflects true differences in IPV prevalence.

5. In conclusion

IPV against women has been considered as a ‘global public health problem of epidemic proportions’ (WHO, 2013, p. 7), and researchers, governments and international bodies concerned about its prevention have made calls to urgent action (García-Moreno et al., 2015, Heise, 2011, Jewkes, 2014 and World Health Organization, 2013). Plans to action should be research-driven and based on suitable methodologies like quantitative multilevel analyses with information at different levels (e.g., individual, household, county, country). The Nordic paradox posits a challenging research question that should not be ignored. After excluding the possibility of confounding and information bias, this paradox needs to be urgently understood. By doing so we will advance our knowledge base on the determinants of individual IPV risk within and between countries and, thereby, provide better-targeted prevention initiatives. The Nordic paradox may provide an avenue to guide new research on IPV in order to appropriately respond to this social and public health problem in a more effective way.

References

    • Abramsky et al., 2011
    • T. Abramsky, C.H. Watts, C. Garcia-Moreno, K. Devries, L. Kiss, M. Ellsberg, L. Heise
    • What factors are associated with recent intimate partner violence? Findings from the WHO multi-country study on women's health and domestic violence
    • BMC Public Health, 11 (2011), p. 109
    • Full Text via CrossRef
    • Aebi et al., 2014
    • M.F. Aebi, G. Akdeniz, G. Barclay, C. Campistol, S. Caneppele, B. Gruszczynska, R. þórisdóttir
    • European Sourcebook of Crime and Criminal Justice Statistics
    • (5th Edition)Akateeminen Kirjakauppa, Helsinki (2014)
    • Bra, 2014
    • Brå [National Council for Crime Prevention]
    • Brott I Nära Relationer. En Nationell Kartläggning [Offences in close relationships: A national survey]
    • Brå, Stockholm (2014)
    • Gracia and Lila, 2015
    • E. Gracia, M. Lila
    • Attitudes towards Violence against Women in the EU
    • Publication Office of the European Union, Luxembourg (2015)
    • Heise, 2011
    • L. Heise
    • What Works to Prevent Partner Violence? an Evidence Overview
    • STRIVE, London School of Hygiene and Tropical Medicine, London (2011)

    • Heiskanen and Piispa, 1998
    • M. Heiskanen, M. Piispa
    • Faith, Hope and Battering: a Survey of Men's Violence against Women
    • Statistics Finland, Helsinki (1998)
    • Hindin et al., 2008
    • M.J. Hindin, S. Kishor, D.L. Ansara
    • Intimate Partner Violence Among Couples in 10 DHS Countries: Predictors and Health Outcomes
    • Macro International Inc, [DHS Analytical Studies No. 18]. Calverton, Maryland (2008)
    • Lundgren et al., 2001
    • E. Lundgren, G. Heimer, J. Westerstrand, A.M. Kalliokoski
    • Captured Queen. Men's Violence against Women in “equal” Sweden-a Prevalence Study
    • Fritzes Offentliga Publikationer, Stockholm (2001)
    • Morgenstern, 1998
    • H. Morgenstern
    • Ecological studies, in modern epidemiology
    • K.J. Rothman, S. Greenland (Eds.), Modern Epidemiology (second ed.), Lippincott-Raven, Philadelphia (1998), pp. 459–480
    • World Health Organization, 2013
    • World Health Organization
    • Global and Regional Estimates of Violence against Women: Prevalence and Health Effects of Intimate Partner Violence and Non-partner Sexual Violence
    • World Health Organization, Geneva (2013)
Corresponding author. Department of Social Psychology, University of Valencia, Avda. Blasco Ibáñez, 21, 46010 Valencia, Spain.
© 2016 Elsevier Ltd. All rights reserved.
Posted by at