Beneath the colours: Unveiling Holi's impact on violence against women

While social norms are considered important to the issue of violence against women (VAW), there is little research on how norms impact VAW in public places. Analysing police data from Bihar, this article finds a stark increase of 170% in assaults against women during the festival of Holi as compared to otherwise similar days. The effect varies across districts with different perceptions among women and men regarding VAW.

One-third of all women and girls globally experience physical or sexual violence during their lifetime. Many of these incidents occur in public spaces such as public transportation, festivals, or carnivals – places where women should feel safe and free from harm. This lack of public safety perpetuates gender inequalities and has detrimental consequences for women’s educational choices and labour market outcomes (Borker 2021, Jayachandran 2021).

Deeply entrenched social norms of what constitutes acceptable behaviour towards women have been identified as a major contributing factor. However, little research has been conducted on how social norms impact incidents of violence against women (VAW) committed in public places, which are perpetrated by strangers and often in the presence of witnesses or bystanders.

To analyse how deeply entrenched gender norms affect VAW in public places, our research (Martínez and Poblete-Cazenave 2024) leverages the Indian festival of Holi where certain unwanted behaviours towards strangers are widely condoned, and uses this context as a natural social experiment.

During Holi – the festival of colours – the popular phrase ‘Bura na mano Holi hai’ (‘Don't feel offended, it's Holi’) has become a justification for inappropriate behaviour. This temporary decrease in the social consequences of harassing strangers enables some perpetrators (commonly men) to act by their underlying attitudes towards women. This is consistent with media reports of harassment during Holi which make headlines every year (Mandhani 2018, Dhar 2021). Our research quantifies the extent to which Holi promotes VAW during the festivities.

Sharp rise in violence against women during Holi

Using police data on incidents filed as First Information Reports (FIRs) from the Bihar police department, we examine whether VAW spiked during the Holi period. Based on Indian Penal Codes, we collect daily data on the number of incidents of assaults against women, sexual violence against women (including sexual harassment, attempts to disrobe women, voyeurism, and stalking) and an aggregated measure of VAW.

To estimate the causal impact of Holi on VAW we use the fact that Holi follows the lunar calendar, and therefore its dates vary from year to year – in 2018 it fell on 2 March, while in 2019 it was 21 March. This allows us to compare VAW during Holi against otherwise similar days, accounting for potential temporal factors.

Figure 1 below shows the impact of Holi on diverse measures of VAW in public spaces. We observe a stark increase of 170% in assault against women during Holi compared to otherwise similar days. Similarly, sexual violence and our aggregate variable of VAW increase by over 160% and 140% during Holi, respectively.

Figure 1. Impact of Holi on violence against women

Note: The impact is measured as the share of the mean (average).

Moreover, we observe a significant increase in VAW during days around Holi as shown in Figure 2. For example, panel (a) shows a large increase in assaults against women during Holi, particularly on the day of the festival (indicated by a dashed line). However, assaults increase slightly on the day before the festival. Also, we observe a significant increase in assaults even on the day after Holi – but not beyond. In panel (b) and (c) we observe similar dynamic impacts on sexual violence and our aggregate measure of VAW.

Figure 2. Event-study plots of the impact of Holi on violence against women

Note: This figure shows the prevalence of different offenses 30 days before and after Holi. Date 1 (indicated by the dashed line) is the date of Holi in that year, while date 0 is the day before Holi.

Is this increase in violence against women unique to Holi?

The rise in VAW during Holi could stem from the crowded streets, facilitating sexual harassment and public disturbances. Additionally, Bihar’s limited police capacity, particularly for mass gatherings and events, might contribute to the surge. To analyse this, we study whether the observed rise in VAW is exclusive to Holi, or if it extends to other celebrations, including Chhath Puja (the most significant festival of Bihar), Diwali, New Year's Day, Independence Day, and Christmas Day. Interestingly, we find that most festivals do not lead to an increase in VAW. However, during Diwali, there is a significant, though comparatively smaller, increase of 10% in our aggregate measure of VAW. This is in stark contrast to the surge of 140% during Holi. This suggests that factors behind the surge in VAW during Holi are less likely to be linked to common elements found in major events, like large crowds or limited police resources.

Furthermore, alcohol consumption has been identified as a significant contributor to sexual misconduct, particularly during social gatherings and parties. In Bihar, the production, sale, and consumption of alcohol have been legally prohibited since 2016. Our analysis of alcohol-related offenses across districts reveals no evidence that VAW during Holi is linked to alcohol consumption.

Role of social norms

Gender norms of both the perpetrator (usually men) and the victim (typically women) may play a crucial role in perpetuating this violence. In communities with deeply ingrained norms of male dominance, men might feel there are fewer consequences for their actions. On the flip side, assaulting women who support gender equality can be seen as more ‘rewarding’ for some men, as they might view empowered women as a threat to their traditional beliefs and norms.

We investigate whether the impact of Holi varied across districts with different norms regarding violence against women. Our primary analysis uses data from the National Family Health Survey (NFHS-4), focusing on attitudes about the justification of violence against women (measured by whether wife beating is justified or not). We study survey responses from women as well as men to examine norms from the perspectives of both potential perpetrators and victims. In Bihar, on average, 54% of women versus 37% men believe that this behaviour is justified. Note that the correlation between women’s and men’s perception is very low (-0.06).

Figure 3 shows the impact of Holi on assaults against women based on wife-beating perceptions of women and men. We find two different effects. The first is an intensification effect, with VAW being higher in districts where men believe it is justified to hit women. Holi seems to amplify these existing attitudes and beliefs. Second, we observed a male backlash effect, reflected by higher VAW in districts where women believe that wife-beating is inappropriate. This suggests that men may engage in more violence against women as a response to increasing female empowerment. Similar effects are found for sexual violence and for the aggregate measure of VAW.

Figure 3. Impact of Holi on assaults against women, by men’s and women’s attitudes towards violence against women

Role of underreporting and women's presence on the streets during Holi

In patriarchal societies, crimes against women often go underreported, especially where such behaviours are normalised. Similarly, in patriarchal societies, women might be less likely to participate in public festivities such as Holi; if so, the lower female presence in public festivities might reduce the number of potential victims. Both of these effects could explain the lower VAW in districts where women justify VAW.

To understand reporting behaviour, we analyse whether the delay between the incident date and the reporting date changes during Holi. In turn, to study changes in female mobility during Holi, we use data from the Google Mobility Index, which tracks individuals' presence in specific locations via mobile phone data.

We find that while offenses during Holi are reported with a higher delay than those occurring during other days, there are no significant differences in reporting delays across districts with varying attitudes towards VAW. This suggests that underreporting is not the primary driver of the higher VAW in districts where women justify wife-beating compared to those that do not.

Regarding women’s mobility, our analysis shows a large increase in time spent at home during Holi, indicating reduced mobility. However, as before, we find no clear impact of attitudes towards VAW on mobility during Holi, suggesting gender norms do not significantly affect participation in the festival.

Lessons for the future

We show a large increase in VAW in public spaces during Holi festivities in Bihar, which is primarily driven by entrenched gender norms. Our findings highlight, first, the importance of safety in public festivals or events for women, and second, the critical need for interventions to address harmful gender norms. These interventions (such as those put forth in Bandiera et al. (2020), Dhar et al. (2022), Edmonds et al. (2023), and Shah et al. (2023)) are essential for creating safer public spaces, which would particularly benefit women

Is this surge in VAW during festivities a problem only for India? While contexts and data availability vary across countries, making direct comparisons difficult, we argue that events that enable potential perpetrators to misbehave, for instance by lowering the social costs of harassing individuals, can exacerbate VAW. Indeed, using data from offenses during other carnivals, such as Mardi Gras in New Orleans, United States, and Carnival in Rio de Janeiro, Brazil, we also observe an increase in VAW.