Trends in the acceptability and prevalence of intimate partner violence: Evidence from Demographic and Health Surveys, 1999–2024

Irina Vartanova; Kimmo Eriksson; Pontus Strimling

doi:10.1371/journal.pgph.0006628

. 2026 Jun 24;6(6):e0006628. doi: 10.1371/journal.pgph.0006628

Trends in the acceptability and prevalence of intimate partner violence: Evidence from Demographic and Health Surveys, 1999–2024

Irina Vartanova ^1,², Kimmo Eriksson ^1,^3,^*, Pontus Strimling ^1,⁴

Editor: Zahra Zeinali⁵

PMCID: PMC13293389 PMID: 42340971

Abstract

Intimate partner violence against women (IPVAW) is a major global public health problem. Social norms that accept IPVAW are a key correlate of its prevalence, and recent research has documented a widespread, long-term global decline in such acceptability. What societal changes underlie this normative shift and whether it is associated with a reduction in violence are not yet well understood. We analyzed Demographic and Health Surveys (DHS) data from 1999–2024, including responses from 1,920,105 women across 69 countries and 539,470 men across 60 countries. We calculated the annualized rate of change between the first and last available survey waves for each country. We tested the association between these country-level trends and longitudinal changes in the Human Development Index (HDI) and Gender Development Index (GDI) at a 4-year lag. We also examined whether trends in acceptability were correlated with trends in the reported prevalence of physical IPVAW. Acceptability of IPVAW declined in 94% of countries for women and 89% for men, with a mean annual decline of approximately 1.3 percentage points. Countries with faster increases in HDI tended to show faster declines in acceptability at a 4-year lag (women: r = -0.33; men: r = -0.54). Overall societal development (HDI) showed somewhat stronger associations with declining acceptability than gender-specific equity gains (GDI). Finally, countries with faster declines in acceptability tended to show faster reductions in physical IPVAW prevalence (r = 0.35). This study provides the first cross-national evidence that improvements in human development are associated with a global decline in the acceptability of IPVAW, and that this normative shift corresponds with reductions in experienced violence. These findings are consistent with a role for comprehensive development investments in education, health, and economic prosperity, though the design cannot establish that such investments cause reductions in violence against women.

Background

Intimate partner violence against women (IPVAW) is a major global public health problem, affecting an estimated 641 million women worldwide [1,2]. It encompasses physical, sexual, and psychological violence, with severe and long-lasting health consequences including injury, depression, anxiety, post-traumatic stress disorder, and death [3]. Given its scale and impact, ending IPVAW is a central target of the United Nations Sustainable Development Goals (SDG 5.2).

The socio-ecological model [4] provides a framework for understanding how factors at individual, relationship, community, and societal (macrosystem) levels interact to influence violence risk. At the macrosystem level, social norms that justify IPVAW—such as the belief that a husband is entitled to physically discipline his wife—are consistently identified as a key correlate of its prevalence [5–7]. Both individual attitudes and community-level acceptability are strongly associated with IPVAW experience and perpetration [5,8–10]. Understanding how these norms are changing is therefore critical for prevention efforts.

Recent research has documented a widespread, long-term global decline in the acceptability of IPVAW. Throughout this paper, ‘acceptability of IPVAW’ refers to endorsement of the view that a husband is justified in hitting or beating his wife under at least one specified circumstance, as measured by items from the Demographic and Health Surveys (DHS), a long-running program of nationally representative household surveys conducted in low- and middle-income countries. Foundational work by Pierotti documented this decline during the first decade of the 21st century [11], and a comprehensive 2025 study by Bergenfeld and colleagues confirmed its durability: using DHS and Multiple Indicator Cluster Surveys (MICS) data from 83 countries until 2022, they found that acceptability declined significantly in most countries among both women and men [12]. However, two critical questions remain unresolved.

First, what societal changes underlie this normative shift? Modernization theory posits that economic development fosters value shifts toward greater gender equality and reduced tolerance of interpersonal violence [13,14]. This endogenous process of development-driven change is complementary to the role of international cultural diffusion of anti-violence norms emphasized by other scholars [11,15]. Cross-sectional evidence supports this relationship: countries with higher Human Development Index (HDI) scores exhibit lower acceptability of IPVAW [16], and associations exist between gross domestic product (GDP) per capita, urbanization, and lower domestic violence [17]. However, whether improvements in development over time correspond with declining acceptability remains unexplored.

Other scholars have emphasized exogenous explanations for normative change that operate independently of domestic economic development. Pierotti, analyzing DHS data from 26 countries, found that structural changes within countries—rising urbanization, education, and media access—could not account for the decline in IPVAW acceptability; she attributed the trend to the diffusion of a global cultural script condemning domestic violence, transmitted through international organizations, media, and education systems [11]. Swindle extended this by showing that mass media exposure is a primary conduit for normative diffusion, with media access predicting individual-level attitude change [15]. A parallel pathway operates through institutional mechanisms: True and Mintrom demonstrated that gender mainstreaming adoption across 157 countries was driven by transnational networks and participation in UN women’s conferences, not by domestic development levels [18]. Htun and Weldon, analyzing 70 countries over four decades, found that autonomous feminist mobilization—not national wealth or women’s political representation—was the strongest predictor of progressive policy on violence against women [19]. These endogenous and exogenous mechanisms are not mutually exclusive; countries experiencing development improvements may simultaneously face greater exposure to international human rights discourses. Development may facilitate normative diffusion by expanding education, media access, and civic capacity—making it difficult to disentangle whether norm change reflects endogenous value shifts or increased reception of exogenous global scripts. Our analysis documents associations between development trends and normative change but cannot adjudicate between these complementary explanations.

This raises an important question: is the observed normative shift associated with general societal progress (rising living standards for all), or specifically with improvements in women’s status relative to men? While HDI captures overall development, the Gender Development Index (GDI) measures gender gaps in health, education, and economic resources. Within Heise’s socio-ecological framework, gender inequality operates as a fundamental macrosystem-level driver of violence [4]. We therefore examine both overall development (HDI) and gender-equitable development (GDI) to understand which dimensions of societal change are most strongly associated with normative shifts.

Second, does normative change translate into reduced violence? As Bergenfeld et al. note, this “is far from straightforward” [12]. A 2023 study in The Lancet Global Health found an overall decline in physical and/or sexual IPVAW prevalence across 53 countries between 2000 and 2021, linking this decline to indicators of women’s empowerment [20]. Yet the relationship between attitudinal shifts and prevalence trends has not been directly examined. Understanding whether declining acceptability corresponds with declining violence is essential for assessing whether norm-change interventions may contribute to reducing IPVAW.

This study addresses both questions by analyzing data from the DHS, a series of nationally representative household surveys conducted across low- and middle-income countries, from 1999–2024, including responses from 1,920,105 women across 69 countries and 539,470 men across 60 countries. We test whether country-level trends in IPVAW acceptability are associated with earlier trends in HDI and GDI, examining temporal lags to identify when development improvements most strongly relate to norm change. We then examine whether trends in acceptability correlate with trends in the reported prevalence of physical IPVAW. (Fig 1, in the Methods section, illustrates the specific relationships tested in this study.) Within the broader socio-ecological framework of IPVAW, we focus on macrosystem-level pathways: whether improvements in human development correspond with declining acceptability of IPVAW, and whether declining acceptability corresponds with reduced violence prevalence. While correlational evidence cannot establish causality, correlations between trends provide a stronger test of hypothesized relationships than correlations in cross-sectional data. Cross-sectional correlations confound the hypothesized effect with all stable between-country differences (e.g., geography, colonial history, cultural heritage). Correlations between within-country changes over time control for these time-invariant factors, isolating the association between temporal co-movements [21].

Fig 1 — Arrows indicate the hypothesized relationships tested in this study: human development (HDI, GDI) is associated with trends in the acceptability of IPVAW, and trends in acceptability are associated with trends in prevalence.

By examining both overall development (HDI) and gender-equitable development (GDI), this analysis clarifies whether and how global progress in human development is translating into gender-equitable social norms and, ultimately, safer lives for women.

Methods

Data sources

This study is a secondary analysis of data from the DHS Program [22], publicly available at https://dhsprogram.com/. We used the rdhs R package to identify and download relevant datasets [23]. The DHS provides comprehensive and comparable population-level data across many countries and time points. While dedicated violence-against-women surveys using the World Health Organization (WHO) Multi-country Study methodology are considered the gold standard for measuring IPVAW, the DHS domestic violence module offers the broad geographic and temporal coverage that enables the large-scale trend analysis presented here.

We compiled all available DHS datasets containing the domestic violence module released between 1999 and 2024. To ensure comparability across surveys and time, our analysis is restricted to participants (both women and men) aged 15–49 years who are or have been married, and who provided information on IPVAW acceptability or prevalence. This comprises responses from 1,920,105 women in 69 countries and 539,470 men in 60 countries.

For the primary trend analysis, only countries with at least two surveys collecting data on IPVAW acceptability or prevalence were included. For cross-sectional analyses, such as correlating initial levels of acceptability with the Human Development Index, countries that participated only once were also included. A complete list of countries, survey years, and sample sizes is available in Table A in S1 Appendix.

Measures

Acceptability of IPVAW.

DHS measures attitudes toward IPVAW by asking respondents whether they agree that a husband is justified in hitting or beating his wife for each of several specific reasons: burning food, arguing with him, going out without telling him, neglecting the children, and refusing to have sexual intercourse with him. We focus on countries where all five contexts were included in the survey. Following established literature [6,12], a respondent was coded as accepting IPVAW if they did not reject that a husband is justified in hitting or beating his wife for at least one of these five reasons. We calculated the weighted percentage of respondents accepting IPVAW for each country-year.

It is important to note that the DHS does not include questions about justifying violence in response to suspected or actual infidelity. Research using the World Health Organization (WHO) Multi-country Study methodology has found that infidelity is consistently the most widely accepted justification for violence in many contexts [24]. Thus, our DHS-based estimates likely represent a conservative (lower-bound) measure of total IPVAW acceptability.

Prevalence of physical IPVAW.

The DHS Domestic Violence module is an optional component included in some but not all country surveys. It is administered to a randomly selected subsample of women in each household and measures various forms of violence, including intimate partner violence (for ever-partnered women) and non-partner violence.

We measured women’s experiences of physical IPVAW through questions directed to these randomly selected subsamples of women who ever lived with a partner. To align conceptually with the acceptability measure (whether a husband is justified in ‘hitting or beating his wife’), we focused on physical violence. A respondent was coded as having experienced physical IPVAW if she reported that a partner had pushed her, shook her, threw something at her, slapped her, twisted her arm or pulled her hair, punched her, or hit her with something in the last 12 months. We excluded two items about more extreme violence—being strangled or burnt, and being threatened with a weapon—because they are conceptually distinct from the ‘hitting or beating’ specified in the survey’s social norms questions.

The prevalence of physical IPVAW in a country-year is the weighted percentage of women reporting such violence in the past year.

Human development measures.

We obtained Human Development Index (HDI) data directly from the United Nations Development Programme (UNDP) Human Development Reports database [25]. The HDI is calculated as the geometric mean of normalized country indices for three key dimensions: health (life expectancy at birth), education (mean years of schooling for adults and expected years of schooling for children), and economic prosperity (logarithm of gross national income per capita). The UNDP publishes new HDI measures annually.

To distinguish between the roles of overall development versus gender-equitable development, we also analyzed the Gender Development Index (GDI). The GDI measures the ratio of female-to-male achievement in the same three HDI dimensions. A GDI value closer to 1 indicates greater gender parity, while values further from 1 indicate larger gender gaps. Improvements in GDI therefore reflect the narrowing of gender inequalities in health, education, and economic resources. GDI data were also obtained from UNDP.

This parallel analysis allows us to test whether normative change is associated primarily with rising living standards that benefit entire populations (captured by HDI) or specifically with women’s gains relative to men (captured by GDI). Fig 1 provides an overview of the relationships tested.

Covariates for robustness analyses.

For supplementary robustness analyses, we obtained four time-varying country-level covariates. Net official development assistance (ODA) received as a percentage of gross national income (GNI), the percentage of seats held by women in national parliaments, and trade openness (the sum of exports and imports as a percentage of gross domestic product) were obtained from the World Bank World Development Indicators [26]. Armed-conflict intensity was coded as the country-year maximum state-based intensity level from the UCDP/PRIO Armed Conflict Dataset version 25.1 (0 = no active conflict, 1 = minor, 2 = war) [27].

Statistical analysis

All country-level percentages for the prevalence and acceptability of IPVAW were calculated using the individual sample weights provided in the DHS datasets to ensure national representativeness. For acceptability data, we used standard DHS sample weights (v005). For prevalence data, we used the domestic violence-specific sampling weights (dv005), which adjust for both household selection probability and the within-household random selection of DV module respondents.

Estimating temporal trends.

Throughout this paper, a ‘trend’ refers to the estimated annual rate of change in a variable over the study period for a given country, measured in percentage points per year. To estimate the temporal trend for each country, we calculated the change in percentage points per year as the difference between the last and first observed values divided by the number of years between observations: (Y_last − Y_first)/ (t_last − t_first). This two-point estimator directly measures the observed change rate across the study period for each country. The standard error (SE) for this rate was computed by propagating the sampling uncertainty of the two proportion estimates: SE = √(SE_first² + SE_last²)/ (t_last − t_first), where SE_t = √[p_t(1 − p_t)/n_t]. In this formula, p_t represents the proportion (of acceptability or prevalence) and n_t the sample size at time t, while the denominator t_last − t_first accounts for the length of the observation period to produce a standardized annual estimate. This approach allows for the construction of 95% confidence intervals (Rate ±1.96 SE) that directly reflect the inherent sampling noise of the underlying DHS surveys. Importantly, any measurement error in these country-specific trend estimates acts to attenuate correlations toward zero in subsequent analyses (classical errors-in-variables bias); our correlational findings are therefore conservative.

While countries vary in the number of survey waves (range: 2–9, median: 4), this two-point approach focuses on the total observed change rather than fitting a linear trend to potentially non-linear intermediate data. This method provides unbiased estimates of the annualized rate of change between first and last observations, with confidence intervals that directly reflect sampling uncertainty. However, the results we present are virtually identical if we instead use ordinary least squares (OLS) estimates based on all points in time with available data (as shown in our robustness analysis).

This approach directly estimates additive changes in acceptability levels over time. It differs from the recent comprehensive study by Bergenfeld et al. [12], who used generalized linear models with a log link to model trends. While their figures refer to “annual percentage-point change,” log-link models actually estimate relative (multiplicative) changes. The two approaches are complementary and yield consistent results. Our percentage-point approach offers several advantages: (1) results are directly interpretable as absolute changes in acceptability, (2) trends can be directly compared to baseline prevalence levels, and (3) the relationship between acceptability trends and prevalence trends (both measured in percentage points) is more straightforward to interpret.

Association with human development.

To examine the relationship between trends and living standards, we used Pearson correlation coefficients. If increases in HDI are associated with subsequent decreases in IPVAW acceptability, this predicts that changes in HDI will be correlated with later changes in acceptability. However, a priori, the lag between development improvements and normative change is theoretically uncertain, and no prior empirical work specifies an expected duration. We therefore correlated country-level trends in acceptability with the corresponding change in the country’s HDI over equally long periods but occurring a few years earlier, using lags ranging from 1 to 10 years (for countries where the lagged start year preceded the earliest available HDI record, the period was truncated to begin at the first available data point, so the HDI change period may be shorter than the acceptability trend period). We similarly correlated initial baseline levels of acceptability (at the first point of measurement) with baseline HDI.

The same analysis was repeated using GDI in place of HDI to test whether gender-equitable development shows different associations with normative change than overall development. For HDI and GDI trends, we used OLS linear regression rather than the two-point estimator. With longer lags, the required start year for estimating development change may predate the earliest available HDI and GDI data in some countries; OLS regression accommodates this by fitting trends to all available data points within the specified period.

Relationship between declining acceptability and declining prevalence.

This same correlation method was used to assess the relationship between trends in acceptability and trends in prevalence. We reestimated time trends over the period for which data on both prevalence and acceptability were available. Where acceptability data were available for both men and women, their trends were averaged; in five countries that lacked trend data for men, only women’s acceptability trends were used.

Hierarchical longitudinal robustness analysis

As a robustness check addressing concerns about shared-trend confounding, we fit a hierarchical longitudinal (multilevel) model to the wave-level acceptability data. Each country-year DHS wave enters as an observation, inverse-variance weighted by its binomial sampling variance (n/ [p(1 − p)], normalised to mean 1 within sex). The model includes country-specific random intercepts and random linear time trends, plus a fixed global year term, so that each country’s idiosyncratic trajectory is absorbed and the within-country HDI coefficient is identified by deviations from that trajectory. Time-varying predictors — HDI (at the 4-year lag used in the main analysis) and four supplementary confounders — are decomposed into country-mean (between-country) and deviation-from-country-mean (within-country) components following Bell & Jones (2015). The outcome and all time-varying predictors are standardised per sex (pooled SD), so coefficients are standardised β. Three nested specifications are reported (Table B in S1 Appendix): HDI alone; plus plausibly-exogenous lagged controls (UCDP/PRIO armed-conflict intensity; ODA as % of GNI); plus partly-endogenous lagged controls (trade openness; women’s parliamentary representation). Models were fit in R with lme4 and lmerTest using REML and the bobyqa optimiser, with Satterthwaite degrees of freedom.

All data processing and analysis were conducted in R version 4.5.3 [28].

Ethical considerations

This study is a secondary analysis of publicly available, anonymized data from the DHS Program. All original DHS surveys are approved by the Institutional Review Board (IRB) of ICF and by a local IRB in the host country, and informed consent is obtained from all respondents before participation.

Results

Widespread decline in the acceptability and prevalence of IPVAW

Our analysis reveals a widespread and significant decline in the acceptability of IPVAW across the countries studied. For the acceptability of IPVAW among women, the estimated time trends were negative in 44 out of 47 countries (94%), with a mean decrease of 1.30 percentage points per year. This global decline was not confined to a specific period; for the 21 countries included in Pierotti’s original study [11], the decrease in the acceptability of IPVAW among women continued robustly: the mean decrease was 1.82 percentage points per year in 1999–2010 and 0.91 percentage points per year in 2011–2024. Country-level trend estimates for acceptability and prevalence are reported in Table C in S1 Appendix.

Results for time trends in the acceptability of IPVAW among men were similar. The trends were negative in 33 out of 37 countries (89%), with a mean decrease of 1.31 percentage points per year. These rates are comparable to those reported by Bergenfeld et al. [12], who found declining acceptability in 94% and 89% of countries for women and men respectively.

Fig 2 visually represents these downward trends in the acceptability of IPVAW among women and men. The prevalence analysis showed that 73% of countries (24 out of 33) experienced a reduction in reported physical IPVAW.

Consistency of trends in the acceptability of IPVAW across contexts

Within each country, the trend in acceptability of IPVAW was highly consistent across different contexts (burning food, arguing, neglecting children, going out without telling, refusing sex), as shown by the strong intercorrelations and similarity of average change rates reported in Table 1.

Table 1. Consistency across contexts in country-specific time trends in the acceptability of IPVAW among women.

	Argue	Burn food	Neglect children	Go out w/o telling	Refuse sex
Argue	-0.75
Burn food	0.50	-0.62
Neglect children	0.63	0.66	-1.10
Go out w/o telling	0.62	0.69	0.83	-1.04
Refuse sex	0.62	0.59	0.55	0.72	-0.71

Open in a new tab

Note. Entries are Pearson correlations between the estimated time trends for the acceptability of IPVAW among women in different contexts. The italicized numbers on the diagonal are the average trend in percentage points per year for that context. Based on n = 47 countries.

Consistency of trends in the acceptability of IPVAW across genders

The correlation between the rate of decline in acceptability among women and the rate of decline among men within the same country is r = 0.58, 95% CI [0.31, 0.76], n = 37 countries. This moderate-to-strong correlation indicates substantial consistency in how acceptability is changing across genders within countries, though not complete synchrony. While women and men are moving in similar directions at similar rates within each society, the imperfect correlation indicates that gender-specific factors also influence the pace of change.

Association with human development and gender equity

We first tested whether improvements in the Human Development Index (HDI) correspond with declining acceptability of IPVAW. We found a strong cross-sectional relationship: countries with higher HDI at baseline showed significantly lower acceptability among both women (r = -0.67, 95% CI [-0.78, -0.51], n = 68) and men (r = -0.35, 95% CI [-0.56, -0.11], n = 60).

We also found that the pace of HDI improvement predicted the pace of attitudinal change several years later. Testing temporal lags from 1 to 10 years, we observed a U-shaped pattern with negative correlations strongest at moderate lags (Fig 3). Using a four-year lag, countries with faster HDI growth showed faster declines in acceptability among both women (r = -0.33, 95% CI [-0.56, -0.05], n = 47) and men (r = -0.54, 95% CI [-0.73, -0.26], n = 37).

Fig 3 — Each line shows the correlation between change in development (HDI in green, GDI in orange) and change in acceptability of IPVAW, where the development change precedes the acceptability change by 1-10 years. Left panel: acceptability among women. Right panel: acceptability among men. Error bars represent 95% confidence intervals.

To test whether this relationship reflects general societal progress or specifically the closing of gender gaps, we repeated the analysis using the Gender Development Index (GDI). As Fig 3 illustrates, improvements in gender equity consistently showed weaker associations with declining acceptability than overall HDI improvements across all lag periods tested. While the difference was modest at each individual lag, the pattern was consistent: HDI improvements showed stronger associations with declining acceptability than GDI improvements.

Relationship between declining acceptability and declining prevalence

Acceptability trends (averaged across men and women except in five countries that lack trend data for men) show a positive correlation with trends in the prevalence of IPVAW (r = 0.35, 95% CI [-0.00, 0.62], n = 32).

Visual inspection of Fig 4 reveals Sierra Leone (SLE) as an outlier, the only country with a strong increase in the prevalence of IPVAW. If this outlier is excluded, the observed correlation strengthens to r = 0.51, 95% CI [0.20, 0.73], n = 31, suggesting that our estimate of the relationship between acceptability trends and prevalence trends is conservative. These correlations are unchanged if countries that lack trend data for men are excluded.

Robustness analyses

As a robustness check, we re-estimated all country-level trends using ordinary least squares (OLS) regression fitted to all available survey waves, rather than the two-point estimates used in the main analysis. OLS slopes were highly correlated with the two-point estimates across all measures (Fig A in S1 Appendix), and re-running the key analyses with OLS-derived trends produced virtually identical lag profiles (Fig B in S1 Appendix), acceptability–prevalence associations (Fig C in S1 Appendix), and summary statistics (Table D in S1 Appendix). This consistency indicates that our findings are not sensitive to the choice of trend estimation method.

To further examine the robustness of the HDI-acceptability association, we conducted two supporting analyses. First, a falsification test confirmed temporal ordering: lag correlations (HDI changes preceding acceptability changes) were negative, whereas lead correlations (HDI changes following acceptability changes) were near zero (Fig D in S1 Appendix). Second, we fit a multilevel longitudinal model to the wave-level data with a series of control variables (Table D in S1 Appendix). Each country-year proportion enters as an observation, inverse-variance weighted by its binomial sampling variance; country-specific random intercepts and random linear time trends absorb country-idiosyncratic trajectories, while a fixed global year term absorbs the common secular trend. Time-varying predictors are decomposed following Bell & Jones (2015) into country-mean (between) and deviation-from-country-mean (within) components. Outcome and all time-varying predictors are standardised, so coefficients are standardised β. The within-country HDI coefficient — which is identified by deviations from each country’s own trajectory — is the quantity of substantive interest. Three nested specifications were fit separately for women (W1–W3) and men (M1–M3). With HDI alone, the within-country standardised β is −0.49 (95% CI [−0.88, −0.10], p = .014) for women and −1.04 [−1.51, −0.58], p < .001 for men. Adding plausibly-exogenous lagged controls (armed-conflict intensity; ODA as % of GNI) moves the coefficients to −0.56 (W2; p = .006) and −0.91 (M2; p < .001); further adding partly-endogenous lagged controls (trade openness; women’s parliamentary representation) yields −0.67 (W3; p = .007) and −1.10 (M3; p < .001). The within-country HDI effect is thus consistent across specifications and, if anything, strengthens under fuller control; the between-country HDI effect is also robustly negative for women (β = −0.41 to −0.56, all p ≤ .011) but near zero and non-significant for men.

Discussion

This study investigated trends in the acceptability and prevalence of IPVAW in low- and middle-income countries over the past two decades. Consistent with recent comprehensive analyses [11,12], we confirm a widespread and enduring global decline in the social acceptability of this violence among both women and men. At the observed mean rate of decline (~1.3 percentage points per year), a country with 40% acceptability in 2000 would reach approximately 14% by 2020.

Four findings extend the existing literature. The decline is consistent across justifications, indicating a common underlying attitude rather than independently evolving context-specific beliefs. The pace of decline is correlated between women and men within countries, suggesting that normative change operates largely at the societal level. Faster HDI growth precedes faster declines in acceptability by several years, with somewhat stronger associations for HDI than for GDI; measurement caveats discussed below limit the substantive interpretation of this gap. Faster declines in acceptability also accompany faster declines in reported prevalence of physical IPVAW, providing the first cross-national link between the two trends and bearing on the puzzle of mixed progress on prevalence in earlier work [9,20].

Development and normative change

These findings are consistent with modernization theory [13,14] and extend prior cross-sectional work [16,17] by tracking the process over time. The endogenous modernization account complements the exogenous “global cultural script” diffusion model advanced by Pierotti [11] and Swindle [15]. The two are difficult to separate in observational data: countries experiencing development improvements are also exposed to international human rights discourse. Historical examples are illustrative rather than dispositive. Most US states renounced men’s right to physically discipline their wives by the late 1870s [29], and the UK followed in 1878 [30]. These cases suggest that internal development processes can contribute to normative change alongside international diffusion, but observational data of this kind cannot adjudicate between the two pathways.

The 3–5 year lag between HDI gains and attitudinal change carries two implications. First, normative change is not immediate: it unfolds as individuals experience the effects of development (education, economic security), as egalitarian ideas diffuse through expanding schools and media, and as community standards shift in response. Second, the lag is short. If normative change operated through cohort replacement, lags would run to decades and correlations might strengthen rather than decay. The 3–5 year window is consistent with intragenerational adaptation [31] — people updating their views during their lifetimes — rather than slower demographic turnover. Recent development improvements remain the proximate correlate of current attitudes; earlier gains are not erased but are superseded.

HDI tracked attitudinal change more closely than GDI across all lags and for both genders. The GDI is constructed as a ratio of female-to-male achievement, which compresses cross-country variation and likely attenuates correlations on measurement grounds. HDI and GDI are also highly correlated in our sample (r = 0.72), so their independent contributions are difficult to disentangle. Direct measures of gender-specific institutions (legal protections, female political representation) would provide a more rigorous test. With that caveat, the somewhat stronger HDI association is consistent with general development conditions being relevant to normative change among both women and men, with men’s own education, exposure, and economic circumstances likely shaping their attitudes. The stronger HDI-acceptability correlation among men than women (r = −0.54 vs. r = −0.33) is consistent with this reading. Given these measurement caveats, our results should not be read as evidence that broad development matters more than gender-equitable development for normative change. The pattern is consistent with both processes operating jointly.

From norms to behavior

The moderate correlation between acceptability and prevalence trends is consistent with a social-ecological understanding in which normative change is one component of a multi-level system and causal arrows may run in both directions.

Several factors can decouple the two. Strong legal frameworks can reduce violence where acceptability persists; weak enforcement can allow violence to continue where it does not. Economic stress can raise violence even as attitudes shift, while economic opportunities for women provide exit options. As violence becomes more stigmatized, survivors in some settings may become more willing to disclose it, inflating measured prevalence even as actual violence declines, while in other settings disclosure may fall. Post-conflict contexts (such as Sierra Leone in our sample) can show rising violence linked to social disruption despite declining acceptability. Despite these complexities, the positive correlation indicates that the two trends are linked at the population level.

Policy implications

Table 2 summarizes the key findings. Two readings follow, both proportionate to the correlational design. First, the cross-national pattern is consistent with HDI-aligned investments in education (SDG 4), health (SDG 3), and economic opportunity having relevance for both human welfare and norms about violence (SDG 5), with any normative correlate of development investments emerging over a medium-term horizon rather than immediately. Existing experimental and quasi-experimental evidence shows that gender-norms interventions can reduce IPVAW [32], particularly when individual-level empowerment is combined with community-level norm change [33]; the present trends underscore the value of integrating norm-change components into economic empowerment programs, school curricula, and health-system strengthening rather than relying on standalone campaigns. Second, the population-level association between norm trends and prevalence trends supports continued investment in norm change as one component of multi-level prevention frameworks such as WHO’s RESPECT [34], alongside legal protections, accessible services for survivors, and broader development investments.

Table 2. Summary of key findings and policy implications.

Finding	Evidence	Policy implication
Widespread decline in IPVAW acceptability	Women: 44/47 countries (94%). Men: 33/37 countries (89%).	Norm change initiatives have broad potential for success across diverse contexts
Consistent trends across contexts	Correlations 0.50–0.83 across violence scenarios	Universal messaging about the unacceptability of beating partners is supported
Increases in human development associated with decreases in acceptability 4 years later	Women: r = -0.33; Men: r = -0.54.	Pattern is consistent with development investments being relevant to both human welfare and normative change; causal inference about violence prevention requires interventional evidence.
Trends in acceptability associated with trends in prevalence	r = 0.35 (or 0.51 if outlier is excluded)	Norm change might translate into reduced violence

Open in a new tab

Limitations

Our design is correlational. Several factors connected to the HDI–acceptability association, including legal reforms, awareness campaigns, and women’s rights movements, may operate as mechanisms that transmit development effects, as confounders independent of development, or as both. Because we cannot empirically distinguish these roles, we do not control for these factors in the main analysis, and readers should weigh the evidence accordingly. The concerns below cover threats that operate largely independently of any specific causal pathway.

Survey research on IPVAW must consider social desirability bias [35–38]—the tendency for responses to systematically deviate from respondents’ genuine beliefs toward what is socially desirable. For our findings of declining acceptance of IPVAW, this concern shifts the interpretation from declining private acceptance to declining public acceptance: it has become less socially acceptable to endorse violence against wives. In practice, this distinction may not matter much. Public norms—what people believe they should say and what others in their community find acceptable—shape behavior through social sanctions, gossip, and community judgment, independent of private beliefs. Both private and public norms are sociologically meaningful, and both influence violence risk. Our data capture normative change in the broad sense, even if we cannot determine the precise mix of shifting private beliefs versus shifting public standards.

Organized backlash is a second concern. Anti-gender campaigns unite religious conservatives, right-wing populists, and opponents of “gender ideology” through legislation restricting reproductive rights, withdrawal from agreements such as the Istanbul Convention [39], and attacks on feminist organizations [40–42]. Our data (1999–2024) largely predate the most intensive phase of this mobilization, so the observed trends may not extrapolate to the current political landscape.

Several standard limitations apply. As a country-level analysis, we cannot identify the specific mechanisms (individual education, community norms, legal reforms) through which development affects attitudes, nor fully rule out unmeasured country-level confounders beyond those addressed in the robustness analyses. Our data are limited to low- and middle-income countries and do not include questions about infidelity, often the most widely accepted justification for violence [24], so our estimates of total acceptance are conservative. Countries with only two survey waves contribute less precise trend estimates, though such measurement error attenuates correlations toward zero rather than inflating them. Finally, our country-level averages conceal important variation within countries (urban vs. rural, rich vs. poor, different ethnic groups); future research examining these subgroups would reveal how normative change spreads through societies.

Conclusion

The social norms that have historically justified violence against women are eroding across the globe. Beyond confirming this trend, we provide the first longitudinal evidence that the decline is associated with improvements in human development and corresponds with reductions in experienced violence. Broad-based initiatives raising living standards, health, and education are associated with normative change toward rejection of domestic violence; whether such initiatives causally contribute to reductions in violence prevalence remains to be established by designs that can support causal inference. The cross-national pattern is consistent with development conditions in which both women and men come to hold more egalitarian attitudes, though the analysis cannot establish that any specific development strategy causes this shift, nor that broad development matters more than gender-equitable development. The moderate correlation between acceptability and prevalence trends is a reminder that normative change, while a correlate of progress, is not sufficient on its own. Achieving the vision of SDG 5.2 will require comprehensive, multi-level interventions that integrate broad development investments with targeted norm-change efforts, robust legal protections, and accessible support services for survivors.

Supporting information

S1 Appendix

Fig A. Comparison of two-point and OLS trend estimates for each measure. Each point represents one country. The dashed line indicates identity; the solid line is the linear fit. The near-perfect alignment confirms that the two estimation methods produce equivalent results. Fig B. Correlation between development trends and acceptability trends across different time lags, comparing two-point and OLS estimators. Lines show Pearson correlations at lags of 1–10 years for HDI (green) and GDI (orange), using either two-point (solid) or OLS (dashed) trend estimates. Left panel: women’s acceptability. Right panel: men’s acceptability. The near-identical profiles confirm that the lag structure is robust to the trend estimation method. Fig C. Scatterplot of OLS-estimated time trends for the prevalence of IPVAW (y-axis) against OLS-estimated trends in acceptability (x-axis), in percentage points per year. Labels are ISO country codes. Compare with Fig 4, which uses two-point estimates. Fig D. Lead-lag falsification test. Each point shows the Pearson correlation between HDI change and acceptability change at different temporal offsets, shown separately for women’s acceptability (left panel) and men’s acceptability (right panel). Negative offsets (lags) indicate HDI change preceding acceptability change; positive offsets (leads) indicate HDI change following acceptability change. The asymmetry—negative lag correlations and near-zero lead correlations—supports the hypothesized temporal ordering. Table A. Sample size per country-year. Only the acceptability of IPVAW was used in countries that participated once. The sample size for the prevalence of IPVAW is smaller because only random subsamples of women are asked questions from the domestic violence module. Note on data usage: All listed survey waves were utilized for the longitudinal analysis of the Human Development Index (HDI) and Gender Development Index (GDI) and for the associated lag analyses. For the calculation of the primary annualized rates of change in IPVAW acceptability and prevalence, only the first and last available survey waves (endpoints) for each country were used. Table B. Hierarchical longitudinal model of acceptability of IPVAW, estimated separately for women (W1–W3) and men (M1–M3). Each row is a country-year DHS wave, inverse-variance weighted by the binomial sampling variance of the country-year proportion. The outcome and all time-varying predictors are standardised (per-sex pooled SD), so coefficients are standardised β. All specifications include country-specific random intercepts and random linear time trends, plus a fixed global linear year trend. Time-varying predictors are decomposed into country-mean (between) and deviation-from-country-mean (within) components following Bell & Jones (2015). W2/M2 adds plausibly-exogenous lagged confounders (UCDP/PRIO armed-conflict intensity; ODA as % of GNI). W3/M3 further adds partly-endogenous lagged confounders (trade openness as % of GDP; women’s parliamentary share). Cells report β [95% CI], exact p (APA). Table C. Time trends in the prevalence and acceptability of IPVAW by country. Estimates represent annual change in percentage points per year, calculated as the difference between first and last observed values divided by the number of years. 95% confidence intervals are shown in brackets. Table D. Key correlations using two-point vs. OLS trend estimates (HDI and GDI at lag = 4; acceptability vs. prevalence). The table shows that results are virtually identical regardless of trend estimation method, confirming the robustness of the main findings.

(DOCX)

pgph.0006628.s001.docx^{(1.1MB, docx)}

Data Availability

Individual-level DHS data are available at https://dhsprogram.com/. The aggregated dataset analysed during the current study is available in the OSF repository, https://doi.org/10.17605/OSF.IO/YU3F7.

Funding Statement

This research was supported by the Knut and Alice Wallenberg Foundation (Grant No. 2022.0191 to PS). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

References

1.Devries KM, Mak JYT, García-Moreno C, Petzold M, Child JC, Falder G, et al. Global health. The global prevalence of intimate partner violence against women. Science. 2013;340(6140):1527–8. doi: 10.1126/science.1240937 [DOI] [PubMed] [Google Scholar]
2.World Health Organization. Violence against women prevalence estimates, 2018. Geneva: WHO; 2021. [Google Scholar]
3.Campbell JC. Health consequences of intimate partner violence. Lancet. 2002;359(9314):1331–6. doi: 10.1016/S0140-6736(02)08336-8 [DOI] [PubMed] [Google Scholar]
4.Heise LL. Violence against women: an integrated, ecological framework. Violence Against Women. 1998;4(3):262–90. doi: 10.1177/1077801298004003002 [DOI] [PubMed] [Google Scholar]
5.Heise LL, Kotsadam A. Cross-national and multilevel correlates of partner violence: an analysis of data from population-based surveys. Lancet Glob Health. 2015;3(6):e332-40. doi: 10.1016/S2214-109X(15)00013-3 [DOI] [PubMed] [Google Scholar]
6.Gunarathne L, Apputhurai P, Nedeljkovic M, Bhowmik J. Factors influencing attitudes toward intimate partner violence among married women: a multilevel analysis across 20 low- and middle-income countries. Violence Against Women. 2026;32(5):1039–55. doi: 10.1177/10778012251324890 [DOI] [PMC free article] [PubMed] [Google Scholar]
7.Jewkes R. Intimate partner violence: causes and prevention. Lancet. 2002;359(9315):1423–9. doi: 10.1016/S0140-6736(02)08357-5 [DOI] [PubMed] [Google Scholar]
8.Sardinha L, Nájera Catalán HE. Attitudes towards domestic violence in 49 low- and middle-income countries: a gendered analysis of prevalence and country-level correlates. PLoS One. 2018;13(10):e0206101. doi: 10.1371/journal.pone.0206101 [DOI] [PMC free article] [PubMed] [Google Scholar]
9.Uthman OA, Moradi T, Lawoko S. Are individual and community acceptance and witnessing of intimate partner violence related to its occurrence? Multilevel structural equation model. PLoS One. 2011;6(12):e27738. doi: 10.1371/journal.pone.0027738 [DOI] [PMC free article] [PubMed] [Google Scholar]
10.World Health Organization. Preventing intimate partner and sexual violence against women: taking action and generating evidence. Geneva: WHO; 2010. [DOI] [PubMed] [Google Scholar]
11.Pierotti RS. Increasing rejection of intimate partner violence: evidence of global cultural diffusion. Am Sociol Rev. 2013;78(2):240–65. [Google Scholar]
12.Bergenfeld I, Nguyen V, Wiederkehr K, Hadd AR, Molina EP, Clark CJ, et al. Global trends in men’s and women’s acceptance of intimate partner violence, 1999–2022: an analysis of population-based survey data from 83 countries. eClinicalMedicine. 2025;83:103199. [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Inglehart R, Norris P. Rising tide: Gender equality and cultural change around the world. Cambridge: Cambridge University Press; 2003. [Google Scholar]
14.Inglehart R, Welzel C. Modernization, cultural change, and democracy: the human development sequence. Cambridge: Cambridge University Press; 2005. [Google Scholar]
15.Swindle J. Pathways of global cultural diffusion: mass media and people’s moral declarations about men’s violence against women. Am Sociol Rev. 2023;88(4):742–79. doi: 10.1177/00031224231184576 [DOI] [Google Scholar]
16.Rodríguez Martínez P, Khalil H. Battery and development: exploring the link between intimate partner violence and modernization. Cross-Cult Res. 2013;47(3):231–67. [Google Scholar]
17.Arthur C, Clark R. Determinants of domestic violence: a cross-national study. Int J Sociol Fam. 2009;35(2):147–67. [Google Scholar]
18.True J, Mintrom M. Transnational networks and policy diffusion: the case of gender mainstreaming. Int Stud Q. 2001;45(1):27–57. doi: 10.1111/0020-8833.00181 [DOI] [Google Scholar]
19.Htun M, Weldon SL. The civic origins of progressive policy change: combating violence against women in global perspective, 1975-2005. Am Polit Sci Rev. 2012;106(3):548–69. [Google Scholar]
20.Ma N, Chen S, Kong Y, Chen Z, Geldsetzer P, Zeng H, et al. Prevalence and changes of intimate partner violence against women aged 15 to 49 years in 53 low-income and middle-income countries from 2000 to 2021: a secondary analysis of population-based surveys. Lancet Glob Health. 2023;11(12):e1863–73. doi: 10.1016/S2214-109X(23)00417-5 [DOI] [PubMed] [Google Scholar]
21.Wooldridge JM. Econometric analysis of cross section and panel data. 2nd ed. Cambridge (MA): MIT Press; 2010. [Google Scholar]
22.ICF. Demographic and Health Surveys (various) [Datasets], 2000-2025. Funded by USAID. Rockville (MD): ICF; [Distributor]. [Google Scholar]
23.Watson OJ, FitzJohn R, Eaton JW. rdhs: an R package to interact with The Demographic and Health Surveys (DHS) Program datasets. Wellcome Open Res. 2019;4:103. doi: 10.12688/wellcomeopenres.15311.1 [DOI] [Google Scholar]
24.García-Moreno C, Jansen HAFM, Stöckl H, et al. WHO multi-country study on women’s health and domestic violence against women: initial results on prevalence, health outcomes and women’s responses. Geneva: World Health Organization; 2005. [Google Scholar]
25.UNDP (United Nations Development Programme). Human Development Report 2025 - A matter of choice: people and possibilities in the age of AI. New York: UNDP; 2025. [Google Scholar]
26.World Bank. World Development Indicators [Internet]. Washington, DC: World Bank Group; 2025. Available from: https://data.worldbank.org/ [Google Scholar]
27.Davies S, Pettersson T, Sollenberg M, Öberg M. Organized violence 1989–2024, and the challenges of identifying civilian victims. J Peace Res. 2025;62(4):1223–40. doi: 10.1177/00223433251345636 [DOI] [Google Scholar]
28.R Core Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2025. Available from: https://www.R-project.org/ [Google Scholar]
29.Pleck E. Criminal approaches to family violence, 1640-1980. Crime Justice. 1989;11:19–57. [Google Scholar]
30.Dobash RP, Dobash RE. Community response to violence against wives: charivari, abstract justice and patriarchy. Soc Probl. 1981;28(5):563–81. [Google Scholar]
31.Tormos R. The rhythm of modernization: How values change over time. Leiden: Brill; 2020. [Google Scholar]
32.Semahegn A, Torpey K, Manu A, Assefa N, Tesfaye G, Ankomah A. Are interventions focused on gender-norms effective in preventing domestic violence against women in low and lower-middle income countries? A systematic review and meta-analysis. Reprod Health. 2019;16(1):93. doi: 10.1186/s12978-019-0726-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Clark CJ, Shrestha B, Ferguson G, Shrestha PN, Calvert C, Gupta J, et al. Impact of the change starts at home trial on women’s experience of intimate partner violence in Nepal. SSM Popul Health. 2019;10:100530. doi: 10.1016/j.ssmph.2019.100530 [DOI] [PMC free article] [PubMed] [Google Scholar]
34.World Health Organization. RESPECT women: preventing violence against women. 2nd ed. Geneva: WHO; 2025. ISBN: 978-92-4-011702-0. [Google Scholar]
35.Gibson MA, Gurmu E, Cobo B, Rueda MM, Scott IM. Measuring hidden support for physical intimate partner violence: a list randomization experiment in South-Central Ethiopia. J Interpers Violence. 2022;37(7–8):NP4238–57. doi: 10.1177/0886260520914546 [DOI] [PMC free article] [PubMed] [Google Scholar]
36.Cullen C. Method matters: the underreporting of intimate partner violence. World Bank Econ Rev. 2022;37(1):49–73. doi: 10.1093/wber/lhac022 [DOI] [Google Scholar]
37.Yount KM, Halim N, Schuler SR, Head S. A survey experiment of women’s attitudes about intimate partner violence against women in rural Bangladesh. Demography. 2013;50(1):333–57. doi: 10.1007/s13524-012-0143-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
38.Visschers J, Jaspaert E, Vervaeke G. Social desirability in intimate partner violence and relationship satisfaction reports: an assessment of the Inventory of Desirable Responding. J Interpers Violence. 2017;32(10):1528–46. [DOI] [PubMed] [Google Scholar]
39.Krizsán A, Roggeband C. Politicizing gender and democracy in the context of the Istanbul convention. Cham: Palgrave Pivot; 2021. [Google Scholar]
40.Kuhar R, Paternotte D. Anti-gender campaigns in Europe: mobilizing against equality. London: Rowman & Littlefield; 2017. [Google Scholar]
41.Corredor ES. Unpacking “gender ideology” and the global right’s antigender countermovement. Signs. 2019;44(3):613–38. doi: 10.1086/701171 [DOI] [Google Scholar]
42.Graff A, Korolczuk E. Anti-gender politics in the populist moment. London: Routledge; 2022. [Google Scholar]

PLOS Glob Public Health. doi: 10.1371/journal.pgph.0006628.r001

Decision Letter 0

Zahra Zeinali

26 Mar 2026

PGPH-D-26-00404

Human Development and the Acceptance and Prevalence of Intimate Partner Violence: A Global Analysis of Temporal Trends

PLOS Global Public Health

Dear Dr. Eriksson,

Thank you for submitting your manuscript to PLOS Global Public Health. After careful consideration, we feel that it has merit but does not fully meet PLOS Global Public Health’s publication criteria as it currently stands. Therefore, we invite you to submit a revised version of the manuscript that addresses the points raised during the review process.

Please submit your revised manuscript by 30 April 2026. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at globalpubhealth@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pgph/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A letter that responds to each point raised by the editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

We look forward to receiving your revised manuscript.

Kind regards,

Zahra Zeinali, MD MPH DrPH

Academic Editor

PLOS Global Public Health

Journal Requirements:

If the reviewer comments include a recommendation to cite specific previously published works, please review and evaluate these publications to determine whether they are relevant and should be cited. There is no requirement to cite these works unless the editor has indicated otherwise.

Additional Editor Comments (if provided):

Thank you for submitting your manuscript, “Human Development and the Acceptance and Prevalence of Intimate Partner Violence: A Global Analysis of Temporal Trends,” to PLOS Global Public Health.

Your manuscript addresses an important and policy-relevant question and draws on an extensive cross-national dataset. The reviewers and I agree that the paper has potential. In particular, the descriptive evidence suggesting declining acceptability of intimate partner violence against women across many settings is of interest. However, substantial revision is required before the manuscript can be reconsidered for publication.

After considering the reviewer reports and my own assessment, I am inviting you to submit a major revision.

The principal concerns relate to the alignment between the manuscript’s theoretical framing, analytic approach, and conclusions. At present, the paper relies heavily on modernization theory as the main explanatory framework, but the analysis does not adequately engage with alternative explanations, including global norm diffusion, institutional diffusion, or transnational advocacy processes. More importantly, the empirical strategy does not permit clear adjudication between these mechanisms. The revised manuscript should therefore either broaden and rebalance the theoretical framing or more clearly position the analysis as descriptive and correlational rather than explanatory.

A second major issue concerns the statistical and methodological approach. The current strategy of estimating country trends using first and last survey observations is simple and intuitive, but it discards intermediate observations and may obscure non-linear trends. In addition, the core analyses rely primarily on bivariate correlations of country-level trend estimates, which limits the strength of the inferences that can be drawn. The revised manuscript should strengthen the analytic approach where possible, provide clearer robustness analyses using all available time points, and temper the interpretation of the findings accordingly.

The manuscript also requires a more critical treatment of measurement. The operationalization of “acceptance” of IPVAW using the DHS item set should be discussed more carefully, including its conceptual limits, its binary construction, and the omission of certain commonly cited justifications such as infidelity. Relatedly, the discussion of social desirability bias and reporting bias should be expanded, particularly given the manuscript’s interpretation of temporal change in both attitudes and prevalence.

The interpretation of the comparative HDI and GDI findings also needs greater caution. As currently written, the manuscript risks overstating the conclusion that overall human development matters more than gender equality. Differences in observed associations may reflect measurement properties of these indices rather than substantive differences in explanatory importance.

Finally, the section linking change in attitudinal acceptance to change in violence prevalence should be interpreted more cautiously. The association is potentially important, but the direction of causality remains unclear and alternative explanations remain plausible.

In preparing your revision, please address the following:

strengthen and rebalance the theoretical framing, including engagement with alternative explanatory perspectives;

revise the manuscript to ensure that causal language does not exceed what can be supported by the design;

strengthen or more transparently justify the analytic strategy, including robustness analyses using all available survey waves;

expand the discussion of measurement limitations, including the construction and cross-context validity of the attitudinal measure;

address social desirability and reporting bias more fully;

interpret the HDI versus GDI findings more cautiously;

moderate the interpretation of the association between attitudes and prevalence; and

revise the manuscript for clarity, concision, consistency of terminology, table presentation, and standard English throughout.

Please also address the reviewers’ specific line-by-line and presentation-related comments in full, including comments on the title, abstract, acronyms, structure of the Results and Discussion, and table formatting. In addition, please ensure that the Data Availability Statement and underlying materials fully comply with journal policy.

When submitting your revision, please include a detailed response to the reviewers and a marked-up version of the manuscript showing all changes.

A decision of major revision does not guarantee acceptance, and the revised manuscript may be returned for further review. However, I would be willing to reconsider a substantially revised version that addresses the concerns above.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

-->Comments to the Author

1. Does this manuscript meet PLOS Global Public Health’s publication criteria? Is the manuscript technically sound, and do the data support the conclusions? The manuscript must describe methodologically and ethically rigorous research with conclusions that are appropriately drawn based on the data presented.-->

Reviewer #1: Partly

Reviewer #2: Yes

**********

-->2. Has the statistical analysis been performed appropriately and rigorously?-->

Reviewer #1: No

Reviewer #2: Yes

**********

-->3. Have the authors made all data underlying the findings in their manuscript fully available (please refer to the Data Availability Statement at the start of the manuscript PDF file)?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception. The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.-->

Reviewer #1: No

Reviewer #2: Yes

**********

-->4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS Global Public Health does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.-->

Reviewer #1: No

Reviewer #2: Yes

**********

-->5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)-->

Reviewer #1: The manuscript entitled “Human Development and the Acceptance and Prevalence of Intimate Partner Violence: A Global Analysis of Temporal Trends” addresses an important and timely question: whether improvements in human development are associated with declining social acceptance of intimate partner violence against women (IPVAW) and whether changes in norms correspond with changes in the prevalence of violence. My colleagues use Demographic and Health Survey data covering nearly 1.9 million women and more than half a million men across 69 countries between 1999 and 2023, which provides a remarkably large and geographically diverse empirical basis for analysis! Their manuscript therefore engages with a central issue in the global gender equality literature and has the potential to contribute to ongoing debates on the macro-level drivers of normative change regarding violence against women.

The authors’ central empirical claim is that acceptance of IPVAW has declined in most countries examined and that this decline is associated with improvements in the Human Development Index (HDI) and, to a lesser extent, the Gender Development Index (GDI). Furthermore, the authors report a moderate correlation between declining acceptance and declining prevalence of physical IPVAW. While these findings are potentially interesting, the study in its current form suffers from several theoretical, conceptual, and methodological weaknesses that significantly limit the strength of its conclusions.

In my opinion, the most fundamental limitation concerns theoretical framing. The authors rely heavily on modernization theory to interpret the observed associations between development and normative change. However, I find the engagement with the theory to be limited and insufficiently critical. The colleagues, on the one hand, present modernization as the primary explanatory framework, while, on the other hand, they only briefly mention and without substantive engagement, alternative perspectives, - such as global cultural diffusion, institutional diffusion through international organizations, transnational feminist advocacy networks, or the theory of the world polity. As a result, their theoretical discussion seems somewhat confirmatory rather than truly explanatory. Their analysis is structured in a way that implicitly assumes development-driven value change rather than rigorously testing competing explanations for the observed trends.

I wish to emphasize that this issue is compounded by the fact that the empirical strategy cannot convincingly distinguish between endogenous development processes and exogenous norm diffusion. In my view, both processes could plausibly produce the same empirical pattern: countries experiencing improvements in development indicators might simultaneously be more exposed to global human rights discourses, international development programs, and media narratives condemning domestic violence. I would like to draw the attention of colleagues to the fact that without incorporating variables capturing global institutional diffusion or policy adoption, the analysis risks attributing causal explanatory power to development indicators that may simply correlate with broader processes of globalization and norm institutionalization.

A second major concern relates to the measurement of the key attitudinal variable. If I understood correctly, the operationalization of “acceptance of intimate partner violence” relies on a composite indicator derived from five DHS questions asking whether a husband is justified in hitting or beating his wife in specific circumstances (e.g., burning food, arguing, going out without informing him). Respondents are coded as accepting violence if they endorse any of these scenarios. Here of course, although this operationalization follows established practice in DHS-based research, I find that it is conceptually blunt. It conflates very different normative positions into a single binary measure and does not capture gradations in moral justification or contextual reasoning. Furthermore, as the authors acknowledge, the DHS questions omit infidelity-related justifications, which are widely documented as among the most accepted triggers for partner violence in many societies. Consequently, do not colleagues believe that the indicator likely underestimates overall normative acceptance and may also vary in validity across cultural contexts?

The methodological strategy used to estimate temporal trends is also problematic. If again I have understood correctly, the authors calculate country-specific trends using only the first and last available survey observations, dividing the difference by the time interval between them. But, while this approach produces a simple measure of average change, it ignores all intermediate data points and assumes linear change over time. In my opinion, in countries with multiple DHS waves, this decision results in a substantial loss of information and prevents the detection of non-linear dynamics, reversals, or acceleration in normative change. The authors briefly state that results are “virtually identical” when using OLS estimates based on all available time points, but these alternative analyses are not presented in sufficient detail to allow readers to assess their robustness.

Although I am not a statistician and in general speaking, I find that the statistical design remains relatively weak for the type of causal interpretation suggested in the discussion. The core analyses rely on simple Pearson correlations between country-level trend estimates. While the authors introduce temporal lags between development indicators and attitudinal change, the analysis remains essentially correlational and does not control for potential confounders. I invite colleagues to reflect on how many factors (including legal reforms, public awareness campaigns, expansion of women’s rights movements, conflict dynamics, or shifts in reporting behavior), could simultaneously influence both development indicators and attitudes toward violence. Without multivariate panel models or hierarchical time-series analyses, the observed correlations cannot be interpreted as evidence supporting modernization-driven normative change.

Regarding the interpretation of the HDI versus GDI findings, I would add that warrants caution. The authors conclude that overall development appears to be more strongly associated with declining acceptance of IPVAW than improvements in gender equality. However, this conclusion may be partly an artifact of measurement. HDI and GDI are strongly correlated, and the GDI’s construction as a ratio between male and female outcomes compresses variation across countries. In my view, the weaker statistical relationship observed for GDI therefore does not necessarily imply that gender equality plays a smaller role in shaping norms around violence. I therefore believe that a more nuanced approach would incorporate direct measures of gender institutions, legal protections, or female political representation rather than relying solely on development indices.

As for the analysis linking attitudinal change to changes in the prevalence of violence, I believe it also requires more careful interpretation. The authors report a moderate correlation between trends in acceptance and trends in prevalence (approximately r = 0.35, rising to about 0.51 when an outlier country is excluded). However, the causal direction of this relationship remains unclear. Normative change may indeed influence behavior, but it is equally plausible that declining violence alters attitudes, that both are shaped by institutional reforms, or that measurement artifacts affect reported prevalence. In addition, changes in stigma surrounding domestic violence could simultaneously reduce expressed acceptance while increasing or decreasing reporting rates, thereby complicating interpretation of trend correlations.

Another issue concerns potential social desirability bias. As public discourse increasingly condemns domestic violence worldwide, respondents may feel greater pressure to provide socially acceptable answers in surveys. This possibility is acknowledged but not sufficiently examined. The colleagues assume that systematic increases in social desirability bias are unlikely, yet this assumption is not empirically tested and may be unrealistic given the global expansion of gender equality campaigns during the study period.

Despite these concerns of mine, I find that the manuscript has several strengths. The dataset is extensive, the research question is relevant, and the attempt to connect macro-development indicators with shifts in social norms is intellectually valuable. The descriptive evidence for a widespread decline in the acceptance of IPVAW across multiple regions is itself an important empirical contribution. However, the analytical framework and methodological design are not yet sufficiently developed to support the stronger causal claims advanced in the discussion.

In its current form, the manuscript reads more as a descriptive cross-national trend analysis than as a rigorous test of competing theories of normative change. To reach the standards expected of a serious public health journal, as it is PGPH, the paper would require substantial revision. In particular, the authors would need: a) to expand the theoretical framework, b) incorporate alternative explanatory mechanisms, c) employ more rigorous panel-based statistical models, and d) provide a deeper discussion of measurement limitations and potential reporting biases.

For these reasons, my recommendation would be major revision. The paper addresses an important topic and draws on a valuable dataset, but significant conceptual and methodological improvements are necessary before the conclusions can be considered robust.

Reviewer #2: Thank you for the opportunity to review this interesting manuscript on an important subject. I enjoyed reading this article, which is well-written and compellingly written. My primary feedback is the need to directly engage with recent patriarchal backlash and how the findings can be used to develop strategies to mitigate for anti-equity misinformation. I make some suggestions below to enhance clarity and would be interested to review a revision.

TITLE: It isn't immediately clear why "human development" is in the title without explictly mentioning the human development and gender development indices.; I wonder if the authors would consider amending the title to more clearly indicate the study methods (e.g. "global Demographic and Health Survey trends from 1999-2023")

ABSTRACT

Line 23 "Four years earlier" - than which time point? 2023?

Line 31 While interesting, I suggest that the point about greater reductions when excluding a post-conflict outlier is more detail than is needed in the abstract.

Line 33 "cross-country" - perhaps, global? Or international?

Line 38 "safer social norms" - perhaps "gender-equitable" social norms?

BACKGROUND

- Line 60 I suggest briefly summarising what the DHS is before reporting a previous finding from it.

- It would help to briefly define "acceptance" of IPVAW. Does it refer to "acceptability" (how acceptable people consider IPVAW to be) or people's personal acceptance/tolerance/lack of opposition to IPVAW. This is explained in the methods but authors would benefit of knowing this at the Background stage.

- Line 63-65 it is hard when reading this sentence not to wonder about current developments in the United States directly refuting the association between economic development and tolerance of (gender) equality. The authors might want to acknowledge this point or mention it in the Discussion.

- I suggest moving Figure 1 to the Methods section.

- Can the statement, "correlations between trends provide a stronger test of hypothesized relationships than correlations in cross-sectional data" be referenced and/or explained, briefly?

METHODS

Could the term "trend" be briefly statistically defined, to help less familiar readers understand its statistical meaning over and above the lay understanding of the word?

Line 240 I believe that the statement about use of AI (Claude) should be at the end of the manuscript rather than in the methods section.

RESULTS

Line 249-251 Since many readers will not be deeply familiar with the Pierotti study, the phrasing of "Pierotti period and "post-Pierotti period" would be better stated simply using the years being referred to.

Throught the manuscript, it would be clearer to say "the acceptability of IPV among women/men" rather than "women/men's acceptance of..." because "acceptance of" sounds like their readiness to personally accept IPV (for readers who may not have read the Methods section).

Line 289 "Consistent with modernization theory" - this is an interpretation that would be more helpful in the Discussion.

Line 299-302 repeats content in the Methods.

Line 305-308 "suggests that..." these interpretations would fit better in the Discussion.

Line 318-319 "Is consistent with..." would be better in the Discussion.

Line 321-324 the additional detail about Sierra Leone would be better in the Discussion.

Table 1 does not have a top line with labels for each column. The third column is also not information deduced from the data so this content could fit better in the Discussion.

DISCUSSION

This section is very interesting to read but rather long in relation to the rest of the manuscript. Some of the repetition of the results could be reduced.

Line 355-357 It does not necessarily follow that public health campagins will influence acceptability. This point is elaborated later so can be removed here.

Line 390 Please explain what the "exogenous "global cultural script" diffusion model" comprises.

Line 564 it would help to say something about what RESPECT is.

The authors only fleetingly acknowledge patriarchal backlashes, which by now are no longer localised matters but being implemented on a national scale in many countries. Can the authors acknowledge this and comment on how their proposal that investment in HDI components might serve to diminish support for populist misinformation as living conditions improve?

MINOR POINTS

- It would help to be consistent in the use of acronyms. IPVAW is mostly used but IPV appears on lines 55, 86, 94, 114 and then "intimate partner violence" is said in words on lines 129, 147, 246, 356, 372, 487, 534.

- Please define all acronyms at first use e.g. Line 60 MICS, Line 69 GDP, Line 74 HDI, line 200 OLS, line 237 ICF, IRB. It would also help not to redefine acronyms later, e.g. DHS on line 108, HDI on line 288, IPVAW on line 343-4, SDG on line 537.

- Line 60 "through 2022" is not a phrase all English speakers would understand. I suggest, "up to" or "until" 2022.

- Line 119 I suggest stating "years" after all ages.

- Line 214 "we don't know which lag to expect" is a little casual.

LInes 211-233 there is switching from active voice ("we correlated...") to passive voice ("analysis was repeated.." Consistency would help.

Line 280 and line 360 "not perfect synchrony" could be phrased more formally.

Line 354 I suggest "women" rather than "wives".

**********

-->6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

Do you want your identity to be public for this peer review? If you choose “no”, your identity will remain anonymous but your review may still be made public.

For information about this choice, including consent withdrawal, please see our Privacy Policy.-->

Reviewer #1: No

Reviewer #2: Yes: Roxanne Keynejad

**********

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

-->Figure Resubmissions:

-->-->While revising your submission, we strongly recommend that you use PLOS’s NAAS tool (https://ngplosjournals.pagemajik.ai/artanalysis) to test your figure files. NAAS can convert your figure files to the TIFF file type and meet basic requirements (such as print size, resolution), or provide you with a report on issues that do not meet our requirements and that NAAS cannot fix.-->-->

After uploading your figures to PLOS’s NAAS tool - https://ngplosjournals.pagemajik.ai/artanalysis, NAAS will process the files provided and display the results in the "Uploaded Files" section of the page as the processing is complete. If the uploaded figures meet our requirements (or NAAS is able to fix the files to meet our requirements), the figure will be marked as "fixed" above. If NAAS is unable to fix the files, a red "failed" label will appear above. When NAAS has confirmed that the figure files meet our requirements, please download the file via the download option, and include these NAAS processed figure files when submitting your revised manuscript.-->

PLOS Glob Public Health. 2026 Jun 24;6(6):e0006628. doi: 10.1371/journal.pgph.0006628.r002

Author response to Decision Letter 1

23 Apr 2026

Attachment

Submitted filename: renamed_2a1d4.docx

pgph.0006628.s003.docx^{(572.7KB, docx)}

PLOS Glob Public Health. doi: 10.1371/journal.pgph.0006628.r003

Decision Letter 1

Zahra Zeinali

1 May 2026

PGPH-D-26-00404R1

Trends in the Acceptability and Prevalence of Intimate Partner Violence: Evidence from Demographic and Health Surveys, 1999-2024

PLOS Global Public Health

Dear Dr. Eriksson,

Thank you for the substantial revisions made in response to the previous decision. The manuscript is notably improved. In particular, I appreciate the clearer framing of the attitudinal construct as acceptability, the strengthened engagement with alternative theoretical perspectives beyond modernization theory, the expanded discussion of measurement limitations and social desirability bias, the improved data availability and reproducibility statement, and the addition of robustness analyses addressing the earlier methodological concerns.

The paper is now much stronger and closer to being suitable for publication. However, a few issues remain that should be addressed before I can make a final decision. I am therefore inviting you to submit a minor revision.

The main remaining concern is the need for greater consistency in aligning the manuscript’s interpretation with the limits of the design. Although the revised manuscript is more careful than the prior version, some sections—particularly parts of the Abstract, Discussion, Policy Implications, and Conclusion—still at times imply stronger causal or policy-relevant inferences than can be supported by an observational, country-level correlational analysis. Please revise the text to ensure that the language remains consistently cautious and does not suggest causal effects where the evidence demonstrates association only.

Relatedly, the interpretation of the HDI versus GDI findings should be further tempered. The manuscript now appropriately notes that the weaker GDI associations may partly reflect measurement properties and compressed variation. However, some sections still lean toward the broader substantive conclusion that overall development matters more than gender-equitable development. Please ensure that this comparison is framed cautiously and consistently across the manuscript.

I also encourage you to slightly moderate several parts of the Discussion that read as more assertive than necessary. In particular, the historical examples used to argue against a diffusion-only explanation should be presented more modestly, as illustrative context rather than as evidence that resolves the theoretical question. Similarly, the Limitations section would benefit from a more neutral treatment of possible confounding versus mediating pathways.

Finally, please tighten the manuscript slightly for clarity and balance. The Discussion can be shortened somewhat by reducing repetition of the Results, and the Policy Implications section should be revised so that proposed implications remain proportionate to the strength of the evidence.

In summary, please address the following in your revision:

1. revise wording throughout to ensure that causal and policy language does not exceed what can be supported by the study design;

2. further temper the interpretation of the HDI versus GDI comparison;

3. moderate several more assertive passages in the Discussion, including the historical argument contrasting endogenous and exogenous explanations;

4. revise the Limitations section to adopt a more neutral tone regarding mechanisms versus confounders; and

5. tighten the Discussion and Policy Implications sections for concision and interpretive balance.

Please submit a clean revised manuscript together with a point-by-point response describing how each of these issues has been addressed.

Please submit your revised manuscript by 30 May 2026. If you will need more time than this to complete your revisions, please reply to this message or contact the journal office at globalpubhealth@plos.org. When you're ready to submit your revision, log on to https://www.editorialmanager.com/pgph/ and select the 'Submissions Needing Revision' folder to locate your manuscript file.

Please include the following items when submitting your revised manuscript:

A letter that responds to each point raised by the editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.
A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.
An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

As the corresponding author, your ORCID iD is verified in the submission system and will appear in the published article. PLOS supports the use of ORCID, and we encourage all coauthors to register for an ORCID iD and use it as well. Please encourage your coauthors to verify their ORCID iD within the submission system before final acceptance, as unverified ORCID iDs will not appear in the published article. Only the individual author can complete the verification step; PLOS staff cannot verify ORCID iDs on behalf of authors.

We look forward to receiving your revised manuscript.

Kind regards,

Zahra Zeinali, MD MPH DrPH

Academic Editor

PLOS Global Public Health

Journal Requirements:

Please review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript. If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Additional Editor Comments (if provided):

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

--> -->-->Figure Resubmissions:

PLOS Glob Public Health. 2026 Jun 24;6(6):e0006628. doi: 10.1371/journal.pgph.0006628.r004

Author response to Decision Letter 2

17 May 2026

Attachment

Submitted filename: Response_to_Editor_R2_2026-05-17.docx

pgph.0006628.s004.docx^{(17.7KB, docx)}

PLOS Glob Public Health. doi: 10.1371/journal.pgph.0006628.r005

Decision Letter 2

Zahra Zeinali

22 May 2026

Trends in the Acceptability and Prevalence of Intimate Partner Violence: Evidence from Demographic and Health Surveys, 1999-2024

PGPH-D-26-00404R2

Dear Dr. Eriksson,

We are pleased to inform you that your manuscript 'Trends in the Acceptability and Prevalence of Intimate Partner Violence: Evidence from Demographic and Health Surveys, 1999-2024' has been provisionally accepted for publication in PLOS Global Public Health.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. A member of our team will be in touch with a set of requests.

Please note that your manuscript will not be scheduled for publication until you have made the required changes, so a swift response is appreciated.

IMPORTANT: The editorial review process is now complete. PLOS will only permit corrections to spelling, formatting or significant scientific errors from this point onwards. Requests for major changes, or any which affect the scientific understanding of your work, will cause delays to the publication date of your manuscript.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they'll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact globalpubhealth@plos.org.

Thank you again for supporting Open Access publishing; we are looking forward to publishing your work in PLOS Global Public Health.

Best regards,

Zahra Zeinali, MD MPH DrPH

Academic Editor

PLOS Global Public Health

***********************************************************

Reviewer Comments (if any, and for reference):

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

S1 Appendix

(DOCX)

pgph.0006628.s001.docx^{(1.1MB, docx)}

Attachment

Submitted filename: renamed_2a1d4.docx

pgph.0006628.s003.docx^{(572.7KB, docx)}

Attachment

Submitted filename: Response_to_Editor_R2_2026-05-17.docx

pgph.0006628.s004.docx^{(17.7KB, docx)}

Data Availability Statement

Individual-level DHS data are available at https://dhsprogram.com/. The aggregated dataset analysed during the current study is available in the OSF repository, https://doi.org/10.17605/OSF.IO/YU3F7.

[pgph.0006628.ref001] 1.Devries KM, Mak JYT, García-Moreno C, Petzold M, Child JC, Falder G, et al. Global health. The global prevalence of intimate partner violence against women. Science. 2013;340(6140):1527–8. doi: 10.1126/science.1240937 [DOI] [PubMed] [Google Scholar]

[pgph.0006628.ref002] 2.World Health Organization. Violence against women prevalence estimates, 2018. Geneva: WHO; 2021. [Google Scholar]

[pgph.0006628.ref003] 3.Campbell JC. Health consequences of intimate partner violence. Lancet. 2002;359(9314):1331–6. doi: 10.1016/S0140-6736(02)08336-8 [DOI] [PubMed] [Google Scholar]

[pgph.0006628.ref004] 4.Heise LL. Violence against women: an integrated, ecological framework. Violence Against Women. 1998;4(3):262–90. doi: 10.1177/1077801298004003002 [DOI] [PubMed] [Google Scholar]

[pgph.0006628.ref005] 5.Heise LL, Kotsadam A. Cross-national and multilevel correlates of partner violence: an analysis of data from population-based surveys. Lancet Glob Health. 2015;3(6):e332-40. doi: 10.1016/S2214-109X(15)00013-3 [DOI] [PubMed] [Google Scholar]

[pgph.0006628.ref006] 6.Gunarathne L, Apputhurai P, Nedeljkovic M, Bhowmik J. Factors influencing attitudes toward intimate partner violence among married women: a multilevel analysis across 20 low- and middle-income countries. Violence Against Women. 2026;32(5):1039–55. doi: 10.1177/10778012251324890 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgph.0006628.ref007] 7.Jewkes R. Intimate partner violence: causes and prevention. Lancet. 2002;359(9315):1423–9. doi: 10.1016/S0140-6736(02)08357-5 [DOI] [PubMed] [Google Scholar]

[pgph.0006628.ref008] 8.Sardinha L, Nájera Catalán HE. Attitudes towards domestic violence in 49 low- and middle-income countries: a gendered analysis of prevalence and country-level correlates. PLoS One. 2018;13(10):e0206101. doi: 10.1371/journal.pone.0206101 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgph.0006628.ref009] 9.Uthman OA, Moradi T, Lawoko S. Are individual and community acceptance and witnessing of intimate partner violence related to its occurrence? Multilevel structural equation model. PLoS One. 2011;6(12):e27738. doi: 10.1371/journal.pone.0027738 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgph.0006628.ref010] 10.World Health Organization. Preventing intimate partner and sexual violence against women: taking action and generating evidence. Geneva: WHO; 2010. [DOI] [PubMed] [Google Scholar]

[pgph.0006628.ref011] 11.Pierotti RS. Increasing rejection of intimate partner violence: evidence of global cultural diffusion. Am Sociol Rev. 2013;78(2):240–65. [Google Scholar]

[pgph.0006628.ref012] 12.Bergenfeld I, Nguyen V, Wiederkehr K, Hadd AR, Molina EP, Clark CJ, et al. Global trends in men’s and women’s acceptance of intimate partner violence, 1999–2022: an analysis of population-based survey data from 83 countries. eClinicalMedicine. 2025;83:103199. [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgph.0006628.ref013] 13.Inglehart R, Norris P. Rising tide: Gender equality and cultural change around the world. Cambridge: Cambridge University Press; 2003. [Google Scholar]

[pgph.0006628.ref014] 14.Inglehart R, Welzel C. Modernization, cultural change, and democracy: the human development sequence. Cambridge: Cambridge University Press; 2005. [Google Scholar]

[pgph.0006628.ref015] 15.Swindle J. Pathways of global cultural diffusion: mass media and people’s moral declarations about men’s violence against women. Am Sociol Rev. 2023;88(4):742–79. doi: 10.1177/00031224231184576 [DOI] [Google Scholar]

[pgph.0006628.ref016] 16.Rodríguez Martínez P, Khalil H. Battery and development: exploring the link between intimate partner violence and modernization. Cross-Cult Res. 2013;47(3):231–67. [Google Scholar]

[pgph.0006628.ref017] 17.Arthur C, Clark R. Determinants of domestic violence: a cross-national study. Int J Sociol Fam. 2009;35(2):147–67. [Google Scholar]

[pgph.0006628.ref018] 18.True J, Mintrom M. Transnational networks and policy diffusion: the case of gender mainstreaming. Int Stud Q. 2001;45(1):27–57. doi: 10.1111/0020-8833.00181 [DOI] [Google Scholar]

[pgph.0006628.ref019] 19.Htun M, Weldon SL. The civic origins of progressive policy change: combating violence against women in global perspective, 1975-2005. Am Polit Sci Rev. 2012;106(3):548–69. [Google Scholar]

[pgph.0006628.ref020] 20.Ma N, Chen S, Kong Y, Chen Z, Geldsetzer P, Zeng H, et al. Prevalence and changes of intimate partner violence against women aged 15 to 49 years in 53 low-income and middle-income countries from 2000 to 2021: a secondary analysis of population-based surveys. Lancet Glob Health. 2023;11(12):e1863–73. doi: 10.1016/S2214-109X(23)00417-5 [DOI] [PubMed] [Google Scholar]

[pgph.0006628.ref021] 21.Wooldridge JM. Econometric analysis of cross section and panel data. 2nd ed. Cambridge (MA): MIT Press; 2010. [Google Scholar]

[pgph.0006628.ref022] 22.ICF. Demographic and Health Surveys (various) [Datasets], 2000-2025. Funded by USAID. Rockville (MD): ICF; [Distributor]. [Google Scholar]

[pgph.0006628.ref023] 23.Watson OJ, FitzJohn R, Eaton JW. rdhs: an R package to interact with The Demographic and Health Surveys (DHS) Program datasets. Wellcome Open Res. 2019;4:103. doi: 10.12688/wellcomeopenres.15311.1 [DOI] [Google Scholar]

[pgph.0006628.ref024] 24.García-Moreno C, Jansen HAFM, Stöckl H, et al. WHO multi-country study on women’s health and domestic violence against women: initial results on prevalence, health outcomes and women’s responses. Geneva: World Health Organization; 2005. [Google Scholar]

[pgph.0006628.ref025] 25.UNDP (United Nations Development Programme). Human Development Report 2025 - A matter of choice: people and possibilities in the age of AI. New York: UNDP; 2025. [Google Scholar]

[pgph.0006628.ref026] 26.World Bank. World Development Indicators [Internet]. Washington, DC: World Bank Group; 2025. Available from: https://data.worldbank.org/ [Google Scholar]

[pgph.0006628.ref027] 27.Davies S, Pettersson T, Sollenberg M, Öberg M. Organized violence 1989–2024, and the challenges of identifying civilian victims. J Peace Res. 2025;62(4):1223–40. doi: 10.1177/00223433251345636 [DOI] [Google Scholar]

[pgph.0006628.ref028] 28.R Core Team. R: a language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2025. Available from: https://www.R-project.org/ [Google Scholar]

[pgph.0006628.ref029] 29.Pleck E. Criminal approaches to family violence, 1640-1980. Crime Justice. 1989;11:19–57. [Google Scholar]

[pgph.0006628.ref030] 30.Dobash RP, Dobash RE. Community response to violence against wives: charivari, abstract justice and patriarchy. Soc Probl. 1981;28(5):563–81. [Google Scholar]

[pgph.0006628.ref031] 31.Tormos R. The rhythm of modernization: How values change over time. Leiden: Brill; 2020. [Google Scholar]

[pgph.0006628.ref032] 32.Semahegn A, Torpey K, Manu A, Assefa N, Tesfaye G, Ankomah A. Are interventions focused on gender-norms effective in preventing domestic violence against women in low and lower-middle income countries? A systematic review and meta-analysis. Reprod Health. 2019;16(1):93. doi: 10.1186/s12978-019-0726-5 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgph.0006628.ref033] 33.Clark CJ, Shrestha B, Ferguson G, Shrestha PN, Calvert C, Gupta J, et al. Impact of the change starts at home trial on women’s experience of intimate partner violence in Nepal. SSM Popul Health. 2019;10:100530. doi: 10.1016/j.ssmph.2019.100530 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgph.0006628.ref034] 34.World Health Organization. RESPECT women: preventing violence against women. 2nd ed. Geneva: WHO; 2025. ISBN: 978-92-4-011702-0. [Google Scholar]

[pgph.0006628.ref035] 35.Gibson MA, Gurmu E, Cobo B, Rueda MM, Scott IM. Measuring hidden support for physical intimate partner violence: a list randomization experiment in South-Central Ethiopia. J Interpers Violence. 2022;37(7–8):NP4238–57. doi: 10.1177/0886260520914546 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgph.0006628.ref036] 36.Cullen C. Method matters: the underreporting of intimate partner violence. World Bank Econ Rev. 2022;37(1):49–73. doi: 10.1093/wber/lhac022 [DOI] [Google Scholar]

[pgph.0006628.ref037] 37.Yount KM, Halim N, Schuler SR, Head S. A survey experiment of women’s attitudes about intimate partner violence against women in rural Bangladesh. Demography. 2013;50(1):333–57. doi: 10.1007/s13524-012-0143-7 [DOI] [PMC free article] [PubMed] [Google Scholar]

[pgph.0006628.ref038] 38.Visschers J, Jaspaert E, Vervaeke G. Social desirability in intimate partner violence and relationship satisfaction reports: an assessment of the Inventory of Desirable Responding. J Interpers Violence. 2017;32(10):1528–46. [DOI] [PubMed] [Google Scholar]

[pgph.0006628.ref039] 39.Krizsán A, Roggeband C. Politicizing gender and democracy in the context of the Istanbul convention. Cham: Palgrave Pivot; 2021. [Google Scholar]

[pgph.0006628.ref040] 40.Kuhar R, Paternotte D. Anti-gender campaigns in Europe: mobilizing against equality. London: Rowman & Littlefield; 2017. [Google Scholar]

[pgph.0006628.ref041] 41.Corredor ES. Unpacking “gender ideology” and the global right’s antigender countermovement. Signs. 2019;44(3):613–38. doi: 10.1086/701171 [DOI] [Google Scholar]

[pgph.0006628.ref042] 42.Graff A, Korolczuk E. Anti-gender politics in the populist moment. London: Routledge; 2022. [Google Scholar]

PERMALINK

Trends in the acceptability and prevalence of intimate partner violence: Evidence from Demographic and Health Surveys, 1999–2024

Irina Vartanova

Kimmo Eriksson

Pontus Strimling

Roles

Abstract

Background

Fig 1. Analytic framework.

Methods

Data sources

Measures

Acceptability of IPVAW.

Prevalence of physical IPVAW.

Human development measures.

Covariates for robustness analyses.

Statistical analysis

Estimating temporal trends.

Association with human development.

Relationship between declining acceptability and declining prevalence.

Hierarchical longitudinal robustness analysis

Ethical considerations

Results

Widespread decline in the acceptability and prevalence of IPVAW

Fig 2. Estimated time trends for the acceptability and prevalence of IPVAW.

Consistency of trends in the acceptability of IPVAW across contexts

Table 1. Consistency across contexts in country-specific time trends in the acceptability of IPVAW among women.

Consistency of trends in the acceptability of IPVAW across genders

Association with human development and gender equity

Fig 3. Correlation between development and acceptability trends across different time lags.

Relationship between declining acceptability and declining prevalence

Fig 4. Scatterplot of estimated time trends for the prevalence of IPVAW (y-axis) plotted against time trends in acceptability of IPVAW (averaged between men and women).

Robustness analyses

Discussion

Development and normative change

From norms to behavior

Policy implications

Table 2. Summary of key findings and policy implications.

Limitations

Conclusion

Supporting information

Data Availability

Funding Statement

References

Decision Letter 0

Zahra Zeinali

Roles

Author response to Decision Letter 1

Decision Letter 1

Zahra Zeinali

Roles

Author response to Decision Letter 2

Decision Letter 2

Zahra Zeinali

Roles

Associated Data

Supplementary Materials

Data Availability Statement

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases