Lori Heise: Cross-national and multilevel correlates of partner violence

Levels of intimate partner violence vary hugely between countries, within countries and even across neighbourhoods and regions. In this studyDr Lori Heise and Andreas Kotsadam uncover crucial macro-level explanations for these disparities.

Published in The Lancet Global Health, this paper assesses how women’s status, gender norms and Gross Domestic Product (GDP) seem to affect levels of partner violence in different settings. The prevalence of previous-year intimate partner violence (IPV) ranges between less than 4% in many high-income countries and at least 40% in some low-income settings. Previous research into IPV has largely ignored the macro-level factors that affect a woman’s risk of violence and the geographical distribution of abuse.

To fill this important gap, Heise et al analyse data from population-based surveys in the WHO Multi-country Study of Domestic Violence and Women’s Health and from Demographic and Health Surveys (DHS). They ask four questions:

  1. Do macro-level gender variables correlate with the geographical distribution of partner violence in the directions feminist-informed theory would suggest?
  2. What best accounts for the apparent association between a country’s level of socioeconomic development and its overall prevalence of partner violence?
  3. Which factors remain important at the macro level when analysed in the presence of other macro-level and individual-level predictors of violence?
  4. Do important cross-level interactions exist between macro-level and individual level factors that affect a woman’s personal risk of partner violence?


Data were compiled from:

  • 54 separate DHS from January 2000 to April 2013
  • 15 population-based surveys representing 10c ountries obtained between 2000 and 2004 as part of the WHO Multi-country Study of Domestic Violence and Women’s Health
  • 2 national replication studies of the WHO study (Turkey and New Zealand)
  • 1 national-level survey of partner violence from Germany that used similar measures and methods

Prevalence surveys were selected for their similarity in terms of violence questions, methods and ethical controls. Overall, the surveys represent 481,205 women between 2000 and 2013.


At the population level:
  • Macro-level factors related to women’s status, gender-related norms, and gender inequality predict the population prevalence of physical and sexual partner violence within the past 12 months.
  • This is the first study to demonstrate empirically that gender-related factors are central to defining overall levels of partner violence.
  • Norms that justify wife beating and male authority over female behaviour are especially predictive of partner violence levels by country.
  • Gender-related discrimination in family law, including differential rights to child custody, to inherit land and money, and to marry and divorces, also predict levels of partner violence across settings.
  • Current partner violence decreases as GDP per person increases, but the association becomes non-significant in the presence of norm-related measures. This suggests that GDP per capita is a marker for other social transformations that accompany economic growth.
In terms of individual level risk:
  • Living in a country that discriminates against women in access to land and other property is also a strong driver of abuse-related risk.
  • Education is more protective in countries or regions where justification of wife beating is greater.
  • Working for cash increases a woman’s risk of partner violence substantially more in settings where few women work than in settings where many women work.
Norms related to male authority over female behaviour, norms justifying wife beating and ownership laws and practices that privilege men over women are robustly and significantly associated with higher levels of violence. These factors drive the geographical distribution of partner violence.

These findings are significant for programming to address IPV and the many health outcomes associated with it. Further research and analysis are needed to understand:

  • how macro-level socioeconomic development and gender norms affect these associations at the individual level
  • potential pathways of change
  • how programming effectiveness might vary across settings because interventions might be enhanced or hindered by macro-level indicators of gender equality


No comments yet.

Leave a comment