Understanding men’s true perspectives on gender-based violence in Mali

Despite policy efforts to combat gender-based violence (GBV), its persistence and justification are widely documented, causing profound and lasting harm to human development (UN 2015, 2020). While the majority of research on GBV focuses on women, understanding men's justifications for and perpetration of GBV is crucial to explain the persistence of such harmful practices and design successful public policies to prevent them (Jewkes et al. 2015, Vaillant et al. 2020).

Mali, which ranks 172nd out of 181 countries on the Gender Inequality Index (UNDP 2024), is a particularly important context to explore men’s behaviours and attitudes towards GBV. According to the 2018 Demographic and Health Survey (DHS), around half of all women have been victims of GBV in their lifetime, 92% have undergone female genital mutilation (FGM), and 60% were married before 18 (against 2% of Malian men). Despite making progress to combat GBV in recent decades, strong resistance remains against banning child marriage and FGM. The current Malian Family Code also continues to uphold the principle that a wife is obligated to obey her husband.

What do we know about Malian men’s attitudes related to gender-based violence?

The DHS questionnaires ask men about their attitudes with respect to two dimensions of GBV: intimate partner violence (IPV) and FGM. Regarding IPV, men are asked whether they believe it is justified for a man to beat his wife in various situations, such as if she burns food. The share of men who justify IPV in at least one of these situations decreased from 61% to 46% between 2001 and 2018. In Bamako, the capital city where our research takes place, this dropped from 52% to 26%. On the other hand, support for FGM remained stable over this period and remains widespread, with 75% of Malian men in favour of it.

These figures may reflect genuine attitudes and/or the internalisation of perceived prevailing social norms—whether those norms oppose IPV or support FGM. The direct-question survey technique used by the DHS may suffer from misreporting and social desirability bias, especially given the potential sensitivity of GBV-related questions (Kadam et al. 2025).[1]1. The sensitivity of GBVs makes their measurement a challenge (Kadam et al. 2025, Park et al. 2023).  As a matter of fact, the diverging discourse on GBV, FGM, and gender-equality issues promoted by various actors—such as international organisations, government, religious authorities, and members of the civil society—may push men to misreport their support for GBV (Bertelli et al. 2024).

Using list experiments to estimate men’s true perspective

We estimate the importance of such declarative biases and elicit arguably unbiased prevalence rates by administering a set of list experiments among a representative sample of 1,200 men in Bamako (Bertelli et al. 2025). The main advantage of list experiments is to provide stronger confidentiality than standard direct questions by keeping respondents’ answers unknown to the interviewer and in fine to the researchers.

In a list experiment, respondents are randomly allocated to either a control or treatment group. They are given a list of statements and asked how many of these statements they experienced or agree with, without specifying which ones. In the control group, our lists are made of three ‘baseline items’, while in the treatment group they are made of four–the additional item being the sensitive one. Given the random allocation of respondents into the two groups, the prevalence rates of the baseline items should not differ across groups. The difference between the two groups’ average responses thus provides an estimate of the sensitive item prevalence.

We conduct seven list experiments, each with a different GBV-related sensitive item:[2]2. The survey contained eight lists but one of them, which sensitive item was the perpetration of verbal violence, did not pass the statistical tests for list experiments' validity.  attitudes towards husbands’ economic control over their wives, female child marriage, FGM, IPV, and the perpetration of physical and sexual violence (Column 1, Table 1).

In addition to the list experiments, we directly ask the control group about the sensitive items with direct questions, which are reported in Column 2, Table 1. We interpret the difference between the prevalence rates elicited from the direct questions and list experiments as a measure of the declarative bias in respondents’ direct answers.

Table 1: Attitudes and behaviours of interest

Men misreport their support for GBV when directly asked, but not their practices

Focusing on the list experiments, which arguably help elicit prevalence rates adjusting for response bias, we find that about half of our respondents are in favour of FGM: 44% believe that a ‘good wife’ is cut and 54% would have their daughter cut if they had one. Moreover, 43% of respondents justify IPV, and nearly one third support female child marriage (31.6%) and perceive as normal for a husband to control his wife’s money (32%). Last, the practice of GBV is not rare, with 26% and 1% of men having already hit and raped a woman, respectively.

These prevalence rates are fairly homogenous across respondents’ age and ethnicity. On the contrary, men with a secondary level of education are those who practice and justify GBV the least.

Comparing prevalence rates between survey methods reveals several declarative biases, as reported in Figure 1. The most striking difference between the two survey techniques concerns respondent support for FGM, which is much lower under the list experiment technique relative to what men declare in direct questions. The size of the bias is considerable: men overreport their support for wife cutting by 24.9 percentage points, and daughter cutting by 21.6 percentage points.

Figure 1: Attitudes towards GBV and behaviours: Prevalence rates and response biases

Conversely, when asked directly, men underreport their justification of IPV by more than 10 percentage points. The opposite biases concerning FGM and IPV are homogenous across respondent characteristics.

One possible reason for the existence of these declarative biases is that the mounting social pressure against IPV pushes respondents to conceal their true (favourable) opinion, when in contrast FGM is likely still perceived as an important cultural and/or religious trait of the Malian society, urging them to comply with the norm.

Last, the two survey techniques yield comparable prevalence rates of the perpetration of physical violence (26%) and rape (1%). Overall, we find that wife-beating is, on average, more justified (44%) than practiced (26%) by men, suggesting that justification of violence is not (simply) an ex-post rationalisation of violent behaviours.

Implications for tackling gender-based violence

Effectively combating GBV remains a critical policy question. The contribution of our results to this debate is twofold. First, unsurprisingly, we observe that male formal education is a potentially important tool to reduce GBV perpetration and justification. Second, when assessing the effectiveness of policies tackling GBV, we reveal the importance of accounting for the perceived social norms and, consequently, behaviours and attitudes perceived as deviant. The desire to comply with prevalent social norms may push respondents to conceal their true opinions, making it hard to assess behavioural and attitude changes through direct questions. Our study shows that, while IPV seems to be disparaged by two-thirds of Malian men, a large share of men actually conceals their true support for IPV. On the other hand, men are much less in favour of FGM than what they openly state, which could signal room for policy change—calling for further research on the determinants of social norm perception.

References

Bertelli, O, T Calvo, M Coulibaly, E Lavallée, M Mercier, S Mesplé-Somps, and O Z Traoré (2024), “Justification des violences conjugales par les hommes : résultats d'une enquête expérimentale au Mali,” Revue d'économie du développement, 32(3–4).

Bertelli, O, T Calvo, E Lavallée, M Mercier, and S Mesplé-Somps (2025), “What one thinks, what one says and what one does: Male justifications and practices of gender-based violence in Mali,” Journal of Development Economics, 176.

Jewkes, R, M Flood, and J Lang (2015), “From work with men and boys to changes of social norms and reduction of inequities in gender relations: A conceptual shift in prevention of violence against women and girls,” The Lancet, 385(9977).

Kadam, A, E B McCullough, T McGavock, and N P Magnan (2025), “Are sensitive questions sensitive to survey method?” VoxDev.

Park, D S, S Aggarwal, D Jeong, N Kumar, J Robinson, and A Spearot (2023), “The challenge of measuring intimate partner violence via self-interviewing,” VoxDev.

UNDP (2024), "Human development report 2023–24: Breaking the gridlock: Reimagining cooperation in a polarized world."

UN (2015), "World’s women 2015: Trends and statistics."

UN (2020), "World’s women 2020: Trends and statistics."

Vaillant, J, E Koussoubé, D Roth, R Pierotti, M Hossain, and K L Falb (2020), “Engaging men to transform inequitable gender attitudes and prevent intimate partner violence: A cluster randomised controlled trial in North and South Kivu, Democratic Republic of Congo,” BMJ Global Health, 5(5).