Violence against women by their intimate partners is a major public health problem globally and remains even more so in western societies. Gender inequality has been considered a factor in explaining violence against women. However, although high levels of gender equality are expected to be linked to a lower prevalence of intimate partner violence against women, this is not always the case. The high prevalence of violence against women by their partners in countries with the highest levels of gender equality, such as the Nordic ones, was labeled as the “Nordic paradox” in a 2016 paper published in Social Science & Medicine.
A survey conducted by the European Union Agency for Fundamental Rights, providing comparable data on IPVAW across the 28 European Union member states for the first time, showed clearly higher levels of lifetime physical and/or sexual partner violence against women prevalent in the European Union Nordic countries (Denmark, Finland, and Sweden). The prevalence of physical and/or sexual partner violence in these countries ranged between 32% and 28%, prevalence rates substantially higher than the European Union average (22%), or the lowest prevalence in these countries (13%).
Although this survey had the advantage of asking the same set of questions regarding physical and sexual violence by intimate partners across all European Union countries, some methodological limitations in this survey could cast doubts about whether cross-country differences reflected real prevalence differences or were the result of a potential measurement bias (e.g. people of different countries understanding or interpreting questions in a different way).
To make appropriate cross-country comparisons, measurements need to be not only reliable and valid but most importantly should demonstrate cross-cultural measurement invariance. However, the psychometric properties (i.e., reliability, validity, and cross-cultural measurement invariance) of the set of questions used in the European Union Agency for Fundamental Rights survey have not been tested and are unknown.
In our study, we set out to determine whether the differences in lifetime prevalence of physical and sexual partner violence against women between Spain and Sweden, two European Union member States that illustrate the “Nordic paradox,” reflected real differences or were the result of measurement bias. These two countries provide a clear example of the “Nordic paradox.” Sweden is ranked number 1 in the European Index of Gender Equality, whereas Spain is ranked 11 in the same index. However, the estimated prevalence of intimate partner violence against women in Sweden is 28% and 13% in Spain. In other words, the prevalence of violence against women by their intimate partners in Sweden is 15 percentage points higher than in Spain, even though the country-level gender equality is substantially higher in Sweden than in Spain.
To make appropriate and valid comparisons between these two countries, it is key to ensure that the measures of physical and sexual partner violence are reliable, valid, and comparable across Sweden and Spain. To do so, in our study, we conducted state-of-the-art psychometric analyses. First, to test the internal construct validity of the set of questions used in the survey, we conducted confirmatory factor analysis and assessed the reliability of the resulted factors, as well as their construct validity (i.e., their relationship with variables with expected links to partner violence, such as self-perceived health). And second, to ensure the comparability of partner violence scores across these countries two measurement invariance techniques were used: differential item functioning analyses were performed to assess whether respondents from Sweden and Spain respond to the items differently; and multigroup confirmatory factor analysis was conducted to assess three increasingly demanding levels of measurement invariance across Swedish and Spanish samples: configural, metric, and scalar.
Our analyses allowed us to establish an equivalent measurement model which ensure the comparability of physical and sexual partner violence scores (prevalence and latent factor scores) between Sweden and Spain. Once the comparability of partner violence survey data across these two countries was ensured, the results of our study showed clearly higher prevalence of intimate partner violence, both physical and sexual, in Sweden than in Spain. In terms of the effect size of the differences between Sweden and Spain, our results regarding physical violence showed that 89.1% of the Swedish sample had higher values in this factor than the Spanish average. These differences were even more remarkable for sexual violence, with 99.4% of the Swedish women scoring higher in the sexual partner violence factor than the Spanish average.
These results not only confirmed higher levels of intimate partner violence against women, both physical and sexual, in Sweden, as compared to Spain, but also that this prevalence differences across these two countries was not the result of methodological bias. In this regard, the results of this psychometric study provide clear support for the Nordic paradox.
So far, scholars do not yet understand why countries like Sweden, with levels of country-level gender equality ranking among the highest in the world, have at the same time higher levels of intimate partner violence against women. Thus, the “Nordic paradox” remains an important and challenging research question for those aiming to better understand and respond to this important social and public health problem.
These findings are described in the article entitled Prevalence of intimate partner violence against women in Sweden and Spain: A psychometric study of the ‘Nordic paradox’, recently published in the journal PLOS One.