Prevalence and Predictors of Affirmations of Intimate Partner Violence in Germany: A First Nationwide Study on Victimization in Women and Men

Abstract

Large-size studies on the prevalence of female intimate partner violence (IPV) victimization in Germany are rare and partly outdated; representative data on male IPV victimization are lacking altogether. The present survey addresses these gaps. For this study, the instrument of the WHO Multi-country study on women’s health and domestic violence has been translated to German and adapted to be used with females and males. A random route procedure in selecting household addresses has been used to gather data on IPV in combination with an omnibus survey on (mental) health issues. A total 2,503 respondents with a minimum age of 14 years have participated (response rate=44.1%). The resulting distribution of age and gender was representative for the German population above the age of 14 years. A total of 57.6% of female participants and 50.8% of male participants have reported victimization by intimate partners during their lifespan; gender distribution differs significantly (Chi²=43.43; p<0.001). Out of the different documented subtypes, psychological IPV was most prevalent (53.6% in females, 48.0% in males). Other forms ranged between 15.2% (physical IPV) and 18.6% (sexual IPV) for females, and 5.5% (sexual IPV) and 10.8% (physical IPV) for males. All forms of victimization regularly coincided, both in females and males. Experiencing any IPV was not only significantly associated with female gender, but also with older age, periods of unemployment, poverty, and IPV perpetration. The findings highlight the still much needed global efforts to prevent IPV against women – and in general. They further support previous research in underlining that fighting poverty might also be instrumental in reducing the likelihood of IPV. The discussion further addresses the issues of reciprocity in IPV.

Introduction

Protection of women’s rights, prevention of violence against women, and prosecution of perpetrators of domestic violence have—in many countries and societies—been widely debated issues in the last decades, at least since the introduction of the Convention on the Elimination of All Forms of Discrimination against Women (CEDAW) in 1979. The Council of Europe’s Convention on preventing and combating violence against women and domestic violence (better known as the “Istanbul Convention”), which has been ratified by 34 European countries since the opening of the treaty in 2011, has once again highlighted the importance of acknowledging violence against women as a violation of human rights. To better protect victims of intimate partner violence (IPV), however, it is important to understand the scope of the problem. As only a small percentage of IPV cases show up in official police or court statistics, population surveys are needed to reliably estimate the prevalence of IPV in the general population.

International reviews have established IPV against women as a large-sized problem (e.g., FRA – Fundamental Rights Agency of the European Union, 2014a; Jewkes et al., 2017; Stöckl et al., 2015). Most reviews were based on studies of self-reported incidents, an approach that is challenged by issues of reliability and validity, as studies have pointed to a possibility of under-reporting in both women’s and men’s self-reports (for an overview, see Chan, 2011). To acknowledge the challenges of self-reports, epidemiological findings are framed by using the term affirmations of IPV. Unlike violence against women, the size and the scope of IPV against men have hardly been established. In Germany, nationally representative studies on IPV against women are rare, and nationally representative data on IPV against men are lacking altogether. So far, the only EU-wide prevalence study including Germany was conducted by the European Union Agency for Fundamental Rights (FRA) in 2012. It was based on face-to-face interviews with around 42,000 women in the EU-28, including 1534 women in Germany (FRA – Fundamental Rights Agency of the European Union, 2014b). Participants between the ages of 18 and 74 were chosen representatively through random household sampling and interviewed orally (FRA – Fundamental Rights Agency of the European Union, 2014b). The women who took part in the study reported their experience of physical, sexual, and psychological violence, including intimate partner violence, as well as stalking, sexual harassment, and violence in their childhood. In congruence with the overall rate for the EU-28, the study found an average of 20% of women over the age of 15 years in Germany who affirmed having experienced physical and/or sexual violence by any partner (previous or current) (FRA – Fundamental Rights Agency of the European Union, 2014a). In Germany, 8% of female respondents reported sexual violence by a partner (EU-average: 7%). In addition, 50% of respondents reported psychological violence by any partner; a rate considerably higher than the EU average of 43%. The finding does not necessarily only indicate a higher prevalence of affirmed psychological violence in Germany than in other EU-states but might also partly originate in an increased readiness to perceive and disclose events as psychological violence. The scale on psychological violence also included two items on economic violence, with a rate of 11% of victimized women in Germany. Here again, the rates in economic powerhouse Germany were similar to the EU average of 12%.¹

Although the research method used was different (e.g., questionnaires), the first nationally representative study on violence against women in Germany showed partly similar findings on IPV prevalence (BMFSFJ - Federal Ministry for Family Affairs Senior Citizens Women and Youth, 2004). It was conducted nearly two decades ago in 2002–2004, and due to its scope and thoroughness, it is still often used as a reference for prevalence of violence against women in Germany. Data were collected from 10,264 women aged 16 to 85 years via oral interviews and questionnaires. The results showed that at least every fourth woman (25%) who ever had an intimate relationship affirmed physical or additionally sexual (7%) violence by a partner once or several times (BMFSFJ - Federal Ministry for Family Affairs Senior Citizens Women and Youth, 2004). When including women who were never in a relationship and women who skipped questionnaire items on violence, the rate of affirmed incidents dropped to 21.6%. The study showed a broad spectrum of violent acts of varying severity, which were differently pronounced and contextually embedded within the relationships (BMFSFJ - Federal Ministry for Family Affairs Senior Citizens Women and Youth, 2004). Of the women who affirmed having experienced physical and/or sexual violence by current or previous partners, just under a third (31%) stated that they had experienced a single incident of violence, whereas 36% named 2 to 10 incidents and a further third (33%) more than 10 to over 40 incidents. This indicates various degrees of severity of violence in intimate relationships.

So far, there has been little research on intimate partner violence against men in Germany (Mosser, 2016). To our knowledge, the only German study on violence against men was published in 2004 and collected data through 266 quantitative oral interviews, 32 qualitative interviews, and 190 additional questionnaires specifically on the topic of domestic violence (Forschungsverbund Gewalt gegen Männer, 2004). The paramount finding was that all types of intimate partner violence (previously known from the studies interviewing women) were also reported by men (Jungnitz et al., 2007). Even though the participants were sampled representatively, Jungnitz et al. (2007) labeled their findings as tendencies due to the small sample size. Twenty percent of interviewees reported having their social contacts controlled by their partners; 25% reported experiencing physical violence in their relationships at least once; several men reported experiencing coerced or unwanted sexual acts in their relationships (Jungnitz et al., 2007).

Predictors of Intimate Partner Violence Victimization

Recently, Yakubovich et al. (2018) offered a systematic review and meta-analyses of prospective-longitudinal studies of risk and protective factors for IPV against women. Among the 71 risk and protective factors in 64 prospective-longitudinal studies, several plausible proxies of socioeconomic status emerged as some of the most relevant risk factors. Further important risk factors were unplanned pregnancy, single parenthood, victimization in youth and childhood, traditional gender role attitudes, and substance abuse. Previous reviews highlighted similar risk factors (e.g., Capaldi et al., 2012; Stith & McMonigle, 2009). On the other hand, being older and being married were significantly associated with a lower risk of prospectively experiencing IPV (Yakubovich et al., 2018). With older age, perpetrating partners might become less physically violent—even in relationships with a long history of IPV. The finding on the lower risk of prospectively experiencing IPV if married is puzzling, but Yakubovich et al. (2018) offer no interpretation. In addition to risk factors at the individual level, the overview of prospective-longitudinal studies of IPV also included variables at the environmental level Surprisingly, Yakubovich et al. found evidence that living in more disadvantaged neighborhoods might be protective; however, that finding should be interpreted with caution, as it could be susceptible to residual confounding, selection bias, or framed perception of IPV.

Aims

Although self-report surveys are far from optimal when exploring the prevalence of IPV, self-report is a valuable approach as a complement to findings from administrative data from the police, the courts, shelters, and more. Administrative data are similarly challenged and do not adequately cover milder incidents of IPV that did not come to the attention of services or law enforcement. The most recent survey on the prevalence of affirmations of IPV against women in Germany was conducted almost a decade ago (FRA – Fundamental Rights Agency of the European Union, 2014a), and regularly cited findings from a larger-sized survey are almost two decades old (BMFSFJ - Federal Ministry for Family Affairs Senior Citizens Women and Youth, 2004). The present study therefore aims to provide a much-needed update on the prevalence of IPV against women in Germany and to introduce the first nationally representative data on IPV against men in Germany. The survey was conducted as part of the development of the online e-learning course, “Schutz und Hilfe bei häuslicher Gewalt [Protection and support against intimate partner violence],” a training course for professionals in the prevention, intervention, and prosecution of domestic violence.

Methods

Sample

This study was designed to gain a representative sample of the German population via systematic area sampling of geographic units and the random route procedure: Geographic units in the entire inhabited area of Germany were sampled based on the municipal classification of the Federal Republic of Germany, and households in every third building on a randomly chosen street were invited to participate in the study. To select participants in multi-person households a Kish selection grid was applied. Selected participants had to be at least 14 years of age and sufficiently master the German language. Of 5668 initially contacted households, 2503 persons (one per household) completed the survey (response rate = 44.1%). The main reasons for non-participation were refusal to identify a target person within the selected household (23.5%, in relation to a sample 5668 households), failure to contact anyone in the household after four attempts (13.4%), and refusal of the target person to participate (13.2%). The resulting distribution of age and gender was representative for the German population above the age of 14.

Variables

Victimization by intimate partner violence (IPV) was analyzed as the dependent variable. IPV was assessed by a translated and adapted questionnaire from the UN Multi-Country Study on Men and Violence (e.g., Jewkes et al., 2017). Items were adapted to include victimization of not only women but also men. They were grouped into psychological violence (5 items), economic violence (3 items), physical violence (3 items), sexual violence (4 items), and disclosure and help-seeking behavior (7 items). Responses to items on lifetime prevalence of violent incidents were Likert-scaled, with the response options never, rarely, occasionally, and regularly. Derived dependent variables included any IPV victimization, which covered all participants that checked the response at least rarely for at least one item. Any psychological IPV victimization, any economic IPV victimization, any physical IPV victimization, and any sexual IPV victimization repeated the “at least rarely for at least one item” rule for each subtype of IPV. Any regular IPV victimization covered all participants that chose the response regularly for at least one item. Multiple victimization counted if participants were victimized by one or several types of violent behaviors. The questionnaire screened for different incidents of victimization that do not necessarily have to co-occur. Accordingly, internal consistency as measured by Cronbach’s α was not expected to reach the same range as would be expected for scales on latent concepts. Cronbach’s α for subtypes ranged from α = 0.51 (economical IPV) to α = 0.88 (sexual IPV).

Demographic independent variables covered gender (male, female), age (birth month and birth year, grouped into 14–29 years, 30–59 years, and 60+ years for regression analyses), formal education (binary coded into college or university degree vs. any other), current residence in Eastern Germany, any period of unemployment. For household income below poverty level (derived from items on range of household income and number of persons in household) the EU definition for poverty thresholds was applied (Eurostat, n.d.). Particularly for psychological IPV, victimization may ignite (defensive) IPV perpetration (e.g., Stith et al., 2012). Any IPV perpetration (at least rarely for at least one item) was therefore introduced as a control variable. An interview conducted during the national lockdown due to the COVID-19 pandemic was introduced as a control variable. As dependent variables measured lifetime prevalence, potential increases in IPV during the lockdown could unfortunately not be analyzed. For data protection, only birth month and birth year and not birth day were available. Age at interview was therefore conservatively calculated by subtracting birth month and birth year from month and year of interview. Table 1 shows the distributions of the independent variables.

Table 1.

Distribution of Independent Variables.

	Total	Percent
Female gender	1256	50
Age group 30–59 years	1331	53
Age group 60+ years	765	31
College degree	598	24
Periods of unemployment	1192	48
Poverty	173	7
Any IPV perpetrated	1246	50
Residence in Eastern Germany	523	21
Right wing voting preference	150	6
Interview during lockdown	961	38

Procedure and Analyses

The interviews were conducted by the USUMA company (Berlin, Germany) from February 10, 2020 to April 25, 2020. A majority of interviews (n = 1542) were conducted before Germany went into a public lockdown on March 15, 2020, to curb the spread of COVID-19. To collect the remaining data, the original deadline was expanded. Interviews during the lockdown were collected following strict public health protection measures, including mandatory face masks.

Individuals who agreed to participate were given information about the study, and informed consent was obtained. For participants who were minors, informed consent was provided by their caregivers. Informed consent included transparent information on the content of the questionnaires and their focus on psychological health and well-being. First, socio-demographic information was collected face-to-face. Next, the researcher handed out a copy of the questionnaire and a sealable envelope. This questionnaire was self-completed due to the partly sensitive nature of the items. The researcher remained nearby in case the participants needed further information. The completed questionnaire was then linked to the respondent’s demographic data but did not contain name, address, or any other personal identifiers. The study was conducted in accordance with the Declaration of Helsinki, was approved by the Ethics Committee of the Medical Department of the University of Leipzig, and fulfilled the ethical guidelines of both the International Code of Marketing and Social Research Practice of the International Chamber of Commerce and the European Society of Opinion and Marketing Research.

All analyses were conducted with the software Stata version SE 16.0. Descriptive analyses were performed for prevalence rates; inferential statistics were modeled as either logistic regression on binary dependent variables of interpersonal violence, ordered logistic regressions for rank-ordered dependent variables (Likert-scaled variables), or Poisson regressions for counts (multiple types of interpersonal violence experienced).

Results

A total of 57.6% of female participants reported having been victimized by intimate partners during their lifespan; male participants affirmed having been victimized at a rate of 50.8%. Gender distribution differed significantly (Chi² = 43.43; p < 0.001); women affirmed experiencing IPV more regularly (Figure 1). Of the different subtypes, psychological IPV was the most prevalent (53.6% in women, 48.0% in men). Other forms ranged from 15.2% (physical IPV) to 18.6% (sexual IPV) for women, and from 5.5% (sexual IPV) to 10.8% (physical IPV) for men. Gender distribution varied significantly for all self-reported types of IPV (psychological IPV: Chi² = 27.59; p < 0.001; physical IPV: Chi² = 63.44; p < 0.001; economic IPV: Chi² = 26.00; p < 0.001; sexual IPV: Chi² = 100.76; p < 0.001), in both total frequency and more frequent regular victimization (Figure 1).

Figure 1.

Prevalence of affirmed intimate partner violence victimization by type of violence and chronicity.

All forms of affirmed victimization regularly coincided, in both women and men (Table 2). Particularly psychological victimization was almost universally present with other types of victimization. Overlap between the other types of affirmed victimization was also substantial for both women and men, albeit on a totally lower level for men (Table 2): For example, almost two thirds of women who reported having been physically victimized women also affirmed experiencing sexual violence; more than half of the women who self-reported having been sexually victimized also affirmed experiencing physical assault.

Table 2.

Prevalences (percentage) for Affirmed Single and Combined Victimization Types.

Gender	Victimization Type 1	Prevalence	Victimization Type 2	Prevalence	Combined Prevalence	Overlap
Female	Psychological	53,58	Economical	17,83	55,73	15,68
(n = 1,256)		53,58	Physical	10,83	54,06	14,73
		53,58	Sexual	18,55	55,41	16,72
	Economical	17,83	Physical	15,21	24,68	8,36
		17,83	Sexual	18,55	26,99	9,39
	Physical	15,21	Sexual	18,55	24,13	9,63
Male	Psychological	48,04	Economical	7,46	49,24	6,26
(n = 1,247)		48,04	Physical	15,21	49,08	9,78
		48,04	Sexual	5,53	48,84	4,73
	Economical	7,46	Physical	10,83	14,44	3,85
		7,46	Sexual	5,53	10,67	2,33
	Physical	10,83	Sexual	5,53	13,72	2,65

Note. All percentages are percentages per total female or male participants.

Victimization differed significantly across age groups in both women (Chi² = 53.10; p < 0.001) and men (Chi² = 44.31; p < 0.01). Victimization of women peaked in the age group 40–49 years, with one or several types of IPV (Figure 2(a)); multiple victimization was particularly pronounced in the age group 60–69 years, with 16.7% of women affirming experiencing two types and 18.8% of women reporting experiencing three or four types of IPV. For men, affirmed victimization peaked in the age group 60–69 years (Figure 2(b)) and decreased substantially for older age groups. IPV was also considerably lower in the youngest age group of men under age 30; however, they peaked in experiencing three or four types of IPV. Overall, multiple victimization was less prevalent in men than in women.

Figure 2.

a) Prevalence of female affirmed intimate partner violence victimization by age group and multiple victimization. b) Prevalence of male affirmed intimate partner violence victimization by age group and multiple victimization.

Tables 3 and 4 show predictors of affirmed IPV victimization: Experiencing any IPV was significantly associated with female gender, older age, periods of unemployment, poverty, and IPV perpetration. A similar pattern emerged for regular IPV victimization: The strength of association between poverty and IPV victimization increased; but with residence in Eastern Germany, the likelihood of experiencing regular IPV decreased. Multiple victimization was also associated with a lower likelihood in residents of Eastern Germany. In addition, it was less likely reported by participants who were interviewed during the COVID-19 lockdown.

Table 3.

(Ordered) Logistic Regressions for Any Affirmed Intimate Partner Violence Victimization and for Subtypes Separately.

	Any Victimization	Any Regular Victimization	Multiple Victimization
n	2456	2456	2456
LR Chi2	1225.36	160.25	1019.31
Pseudo-R	0.362	0.153	0.16
Female gender	2.03***	2.83***	0.41***
Age group 30–59 years	1.55**	2.47*	0.02
Age group 60+ years	1.45*	3.31**	0.08
College degree	1.24	0.71	0.02
Periods of unemployment	1.39**	2.12***	0.23***
Poverty	1.93**	3.96***	0.29***
Any IPV perpetrated	26.31***	3.66***	1.36***
Residence in Eastern Germany	0.92	0.23***	−0.24***
Interview during lockdown	1.04	0.94	−0.10*

Notes: Predictors are displayed as Odds Ratios or Coefficient for Ordered Logistic Regressions.

Table 4.

Logistic Regressions for Any Intimate Partner Violence Victimization and for Subtypes Separately.

	Any Victimization	Any Psychological Victimization	Any Economical Victimization	Any Physical Victimization	Any Sexual Victimization
n	2456	2456	2456	2456	2456
LR Chi2	1225.36	1167.89	255.36	242.21	259.03
Pseudo-R	0.362	0.343	0.137	0.128	0.143
Female gender	2.03***	1.79***	2.94***	1.52**	4.41***
Age group 30–59 years	1.55**	1.56**	0.73	0.81	0.94
Age group 60+ years	1.45*	1.30	1.44	0.85	0.97
College degree	1.24	1.17	0.74	1.26	0.97
Periods of unemployment	1.39**	1.38**	1.59**	1.87***	1.83***
Poverty	1.93**	1.60*	2.19***	1.75**	1.95**
Any IPV perpetrated	26.31***	22.86***	3.73***	4.83***	4.05***
Residence in Eastern Germany	0.92	0.97	0.35***	0.30***	0.81
Interview during lockdown	1.04	0.93	0.77	0.74*	0.74*

Notes: Predictors are displayed as Odds Ratios; *** p<0.001; ** p<0.01; * p<0.05.

The same pattern of significant predictors for any affirmed IPV victimization emerged for psychological IPV (Table 4). The pattern of significant predictors partly changed for other types of IPV: Age was no longer significantly associated with victimization, whereas female gender, periods of unemployment, poverty, and IPV perpetration still were. Residence in Eastern Germany was associated with a lower likelihood of economic and physical IPV victimization; an interview conducted during the first 2 weeks of the COVID-19 lockdown was associated with a lower likelihood of physical and sexual victimization.

Discussion

The first large-size study on affirmed IPV victimization in both women and men reveals a high prevalence both overall and for separate subtypes of victimization: More than half of the German population, both women and men, have experienced IPV victimization at least rarely: 8.1% of women and 3.1% of men affirmed having regularly experienced IPV. Psychological IPV victimization is widespread, as almost all victims of IPV also affirmed having experienced psychological violence by their intimate partners. Other types of victimization are also frequently reported and commonly overlap: From 15%-19% of women have experienced economic, physical, and/or sexual violence by their intimate partners, and the respective percentages for men range from 5% to 11%. Overlap between those subtypes ranges from around a third to more than half of victims. Thus, many affirm multiple victimization—in addition to psychological victimization. The findings highlight the burden of victimization of women: Women report more frequent victimization, more regular victimization, and more multiple victimization. Predictors of IPV victimization highlight the importance of economic stressors: Both periods of unemployment and current poverty were significantly associated with a higher risk of having experienced overall IPV or one of its subtypes.

Affirmed IPV victimization, particularly psychological IPV victimization, is widespread in both women and men in Germany. Around half of the sampled population reports experiences of psychological victimization at least rarely in their lifetime. An exploratory analysis of single items shows that particularly insults are ubiquitously reported by both women and men, whereas incidents perceived as more severe, such as threats of physical harm or destroying personal items, are still common but not as highly affirmed as insults. In addition, these more severe incidents are more frequently reported by women. Additional analyses will have to scrutinize the prevalence and regularity of different incidents of psychological IPV victimization, particularly as evidence suggests that the importance of (regular) psychological IPV victimization should not be underestimated, as it may lead to potentially severe psychosocial consequences (e.g., BMFSFJ - Federal Ministry for Family Affairs Senior Citizens Women and Youth, 2004; Potter et al., 2021).Whereas psychological IPV victimization is widespread, other subtypes of self-reported IPV victimization are also highly prevalent and commonly overlap. Between 1 in 7 and 1 in 5 women affirm having experienced economic, physical, and/or sexual violence. The respective ratios for men are between 1 in 20 and 1 in 9. Notably, economic IPV is also an important issue in Germany, which is one of the G7 countries. This underlines the need to conduct more research on economic IPV victimization in high-income countries. For women, the 24.12% combined prevalence of affirmed physical and/or sexual IPV victimization exceeds the corresponding rate of 21.6% physically and/or sexually victimized females in both the first nationwide population survey in 2003 (BMFSFJ - Federal Ministry for Family Affairs Senior Citizens Women and Youth, 2004) and the more recent Europe-wide FRA survey (FRA – Fundamental Rights Agency of the European Union, 2014a). The increase does not necessarily reflect only an actual increase in prevalence; widespread public sensitization might also have led to an increased readiness to self-perceive acts of IPV as such and, consequently, to disclose it in surveys (for an analysis of child sexual victimization rates in the context of the international 2017/2018 #MeToo debate, see Witt et al., 2020). Still, comparisons between the surveys should be interpreted cautiously due to variations in the instruments used and the contextualization of the items (see limitations below).

Although differences in prevalence rates for age groups are consistent with previous research—women in their teens and elderly women regularly report less violence than middle-aged women (e.g., Sanz-Barbero et al., 2019), they are partly difficult to interpret. As the lifetime prevalence of self-reported victimization has been measured, the increasing span of years in older people logically increases the likelihood of experiencing an incident of IPV victimization. However, the significantly higher likelihood for older age groups than for persons below the age of 30 does not hold up, if age is introduced as a predictor for multiple victimization and for separate subtypes. The overall prevalence of affirmed multiple IPV victimization decreases particularly substantially in older men. This finding of lower multiple victimization in older men does not necessarily reflect actual lower prevalence: In the context of changing societal values, older men might also be more reluctant to perceive and document incidents of violence, whereas younger men are more sensitive to these issues, as the peak of 3+ multiply victimized for men below the age of 30 might indicate. Besides a higher likelihood of selective memories in elderly participants, the higher mortality of regularly or multiply victimized participants might also have decreased the prevalence in older age groups, particularly for men, who have a generally lower life expectancy (e.g., The World Bank).

Although age and gender have already been discussed as predictors of the likelihood of being victimized, findings on other predictors also necessitate attention. The findings on potential proxies of socioeconomic status are congruent with previous research (cf. Yakubovich et al., 2018). Having experienced at least one period of unemployment and current household income below the poverty threshold are significantly associated with a higher likelihood of affirmed IPV victimization—not only for overall victimization but consistently so across subtypes and for regular and multiple victimization. As these economic predictors are controlled for in the multivariate logistic regressions, higher education as measured by at least college degree is not significantly associated with a risk for affirmed IPV victimization. Thus, higher education does not buffer against victimization if economic factors are accounted for.

This study finds an unusually large odds ratio for affirmed IPV perpetration as a predictor of overall and psychological IPV victimization. It seems that psychological IPV is largely reciprocal. However, these quantitative data do not provide information on the defensive or offensive use of psychological violence. Also, in the items regarding psychological violence the threshold was rather low compared to other studies. Two items asked about insults and humiliation without specifications (e.g., “in a way you felt hurt/threatened”). It is likely that almost all adults living in a relationship have felt insulted or humiliated by their partners once in their lifetime, and it is difficult to say if this always qualifies as IPV. Affirmed perpetration is also a significant predictor for IPV victimization in other subtypes, yet to a much smaller degree. Note that the predictors measure any type of IPV perpetration, so psychological violence by a female partner might have been responded to by a male partner or used as an excuse to perpetrate sexual violence. It is likely that the study found defensively or vindictively perpetrated violence.

To date, 22 years after the reunification of the German Democratic Republic (GDR) with the Federal Republic of Germany, several studies still find distinct differences in societal variables, many in the direction of socioeconomic disadvantages in the states of the former GDR (e.g., Wenau et al., 2019). We therefore introduced the East-West differentiation as control variable. Somewhat surprisingly, the likelihood of reports of regular and multiple IPV victimization and economic and physical IPV victimization is significantly lower for residents of Eastern Germany. The findings are puzzling and difficult to interpret, even though they might be partly congruent with the findings by Yakubovich et al. (2018) on the potentially protective effect of disadvantaged neighborhoods. An additional comparison between the economically strong southern states of Baden-Wurttemberg, Bavaria, and Hessian and the less economically thriving northern German states reveals no significant differences.

This study comes with several advantages: It is the first nationally representative study in Germany that includes both women and men as potential victims of intimate partner violence. In addition, it goes beyond measuring physical, psychological, and sexual IPV by also including economic IPV. There are some limitations, however. First, as previously addressed, the validity of self-report data on IPV is controversially debated, particularly in combination with the unresolved question of gender symmetry or asymmetry (for an overview, see Chan, 2011). A poor overlap in partners’ reports of IPV has been attributed to both underreporting of victimization in women (due to, for example, self-blaming or dependence) and under-reporting of perpetration in men, as many might perceive non-violent behavior as socially desirable (e.g., Dobash & Dobash, 2004; Chan, 2011; Hamby, 2014). Note that for this survey, only one member per household self-completed a questionnaire, which should have reduced a bias of adaptation to the response of the partner in the same household. We still observed a gap between the prevalence of self-reported incidents of victimization and the prevalence of perpetration. Even though the questionnaire did not cover motives and impact of the IPV incidence, which are both predictors of gender asymmetry (Chan, 2011), the findings still clearly indicate an asymmetry of higher female victimization than male victimization. Second, the dependent variables are measured as lifetime prevalence of IPV. It was therefore not possible to analyze a potential increase of affirmed IPV during the national lockdown during the COVID-19 pandemic in the spring of 2020. However, an interview conducted during the lockdown was introduced as a control variable. The decreased likelihood of physical, sexual, and multiple IPV for interviews conducted during the lockdown might be associated with selection bias in participation. However, this is speculative, as media attention to the topic of IPV during the lockdown might also have increased the likelihood to disclose victimization. Based on the same sample, Sachser et al., 2021 offer an interpretation on the perception of the family as a buffer against external threat. Third, the questionnaire on IPV victimization and perpetration was part of an omnibus survey on topics in (mental) health and was therefore limited in length. This comes with restrictions in variables: The dependent variables are limited to frequency of self-reported IPV victimization and do not include other relevant information such as (perceived) severity of IPV or onset. Moreover, the variety of topics included in the questionnaire might have influenced participants’ responses to some of the questions on IPV victimization as well as their willingness to share this information. In addition, the questionnaires did not (fully) cover potentially important predictors: Participants were not asked to provide detailed information on migration, particularly their country/region of origin. This is unfortunate, as we cannot differentiate between migration of mainly highly-qualified persons from culturally comparable high-income neighboring countries like Austria and Switzerland and migration of refugees from countries like Afghanistan and Syria that most likely have a history of experiencing multiple forms of violence. Overall, as the omnibus survey was of substantial length, it was potentially tiring and, in consequence, may have been linked to decreased reliability. In addition to the questionnaire on IPV victimization, other parts of the survey dealt with violence, such as items on incidents of child maltreatment or corporal punishment. Although the instructions and wording were checked for accessibility and clarity, some participants might still have had difficulty disentangling these items from the items on IPV. Finally, the items on IPV victimization were positioned after 73 questions that included other items on violence and (mental) health issues. Those might have triggered the topic and might have led to a higher readiness to disclose IPV victimization.

Conclusion

In the almost impossible struggle to determine the real prevalence of IPV (Chan, 2011), the present study provides a perspective on approximating the prevalence of IPV victimization in Germany by presenting updated self-reported prevalence data that partly reproduce findings of earlier studies. However, the design included men and women as well as questions regarding the perpetration of violence and forms of violence that have not been included widely in other prevalence research. In addition, the present survey further includes data on witnessing IPV in childhood and IPV perpetration. These data offer several further avenues to explore in additional outputs, particularly on the contribution of victimization in childhood and youth, which has been identified as a risk factor for IPV in adulthood (Yakubovich et al., 2018).

Epidemiological research on violence can and should inform strategies for intervention and prevention (e.g., World Health Organization, 2007). In this study, the findings on socioeconomic variables suggest that fighting poverty might also be instrumental in reducing the likelihood of IPV. Socioeconomic variables could be used to identify target groups for selective prevention that focuses on women’s economic independence. Too often, selective (and indicative) prevention are still niches in comparison to broadband, universal violence prevention, even though these approaches should complement each other. Selective and/or indicative prevention might also be needed to address IPV in the migrant population. Future studies should therefore include detailed information on migration—particularly for a country like Germany, where almost 20% of the population are first-generation immigrants (e.g., Bundeszentrale für politische Bildung, 2021).

In line with international evidence, women are the major victims of IPV. They are more frequently victimized, more regularly victimized, and more commonly multiply victimized. Ultimately, both in Germany and elsewhere, criminal statistics point to extensive rates of femicides (Bundeskriminalamt, 2020). Gender differences in severity of experienced IPV are another issue to analyze in future research: An exploratory analysis of single items in this study, for example, points to an increasing gap between men and women in perceivably more severe items of affirmed IPV victimization (e.g., being physically hurt by intimate partner). Overall, the findings of this study once more underline the importance of fostering change in policy and society. The (further) implementation of the Istanbul Convention and regular and secure funding for prevention and protection as well as research are paramount.

Author Biographies

Andreas Jud, PhD, has a chair on Child Maltreatment Epidemiology at University of Ulm’s department of Child and Adolescent Psychiatry/Psychotherapy, Germany, and is staff at Lucerne University of Applied Sciences and Arts, School of Social Work, Switzerland. He is an Associate Editor of Child Abuse and Neglect, the leading international journal in the area of child protection.

Bianca Grafe has a MA in Educational Science, Comparative Philology and Philosophy. She is currently writing her PhD thesis. Key aspects of her studies are culture, migration, gender and violence, as well as ethics and practical philosophy. She has worked in several international research projects with focus on violence and gender, particularly intervention against violence against women and children.

Ksenia Meshkova is a Gender Studies researcher and lecturer working in Berlin, Germany. She is currently based at the Humboldt University of Berlin, German Human Rights Institute and Alice Solomon University of Applied Sciences. The main areas of her expertise are domestic and intimate partner violence, femicide, workplace sexual harassment as well as connections between femininity, masculinity and violence.

Barbara Kavemann, Prof. Dr., is a senior researcher at the Social Science Research Institute on Gender Issues in Freiburg and member of the Independent Inquiry into Child Sexual Abuse in Germany.

Thomas Meysen, Dr., is managing director of SOCLES International Centre for Socio-Legal Studies. His research focusses on the interplay between law and society and vice versa in the fields of childhood, youth, family and gender. Main areas are violence against children, violence against women, inclusion, extremism, and ethics.

Ulrike Hoffmann, Dr. hum.biol., head of the working group “Knowledge transfer, Dissemination, E-Learning” at the University hospital Ulm’s department of child and adolescent psychiatry/psychotherapy. The working group develops online courses on child protection topics for professionals mainly from the medical, psychotherapeutic, educational and legal sectors.

Prof. Dr. Jörg M. Fegert, MD, is funder and Medical Director of the Department for Child and Adolescent Psychiatry/Psychotherapy at Ulm University Hospital as well as head of the Competence Area Mental Health Prevention in the Competence Network Preventive Medicine Baden-Württemberg. At the international level, he is President Elect of the European Society for Child and Adolescent Psychiatry (ESCAP).

ORCID iDs

Andreas Jud https://orcid.org/0000-0003-0135-4196

Thomas Meysen https://orcid.org/0000-0003-3904-5030