In Zimbabwe, Substantial Minorities of Women Are Accepting of Wife-Beating

D. Hollander

First published online:

Half of Zimbabwean women believe that in some situations, a man is justified in beating his wife, according to an analysis of data from the 1999 Zimbabwe Demographic and Health Survey.1 Although the proportion who see some justification for this behavior varies among subgroups of women, it never is below one-third and typically exceeds one-half. Women are more accepting of wife-beating in some situations than in others; levels of acceptance are highest (33-39%) in cases in which a woman has argued with her spouse, gone out without telling him or neglected her children.

The 5,907 women in the nationally representative sample were predominantly urban, married and living with their spouse or partner in a monogamous union; virtually all of them belonged to one of the country's two main ethnic groups (Shona and Ndebele). On average, they were 28 years old and had been married for 12 years. Nearly half had at least a secondary education, and slightly more than half were employed (mainly as agricultural or manual workers); six in 10 lived in households whose level of wealth was categorized as low. Respondents' partners were generally older than they were and more likely to be employed. When asked how decisions about the household and the woman herself were made, half of respondents said that their partners did not have final say in any of five specified situations; one-third reported that the couple made at least two types of decisions jointly, and one-third said that they had final say in four or five areas.

In most subgroups, 52-62% of women considered wife-beating justifiable in some circumstances; even the lowest proportion (36%, among urban women) indicated substantial acceptance of this behavior. Comparisons based on chi-square tests revealed that acceptance differed significantly by most background and relationship characteristics: Urban residence, minority ethnicity, a high level of household wealth or educational attainment, older age and employment in a high-status occupation were associated with reduced levels of acceptance. Similarly, having a husband who was older than 36, had at least a secondary education or held a high-status job was related to lower acceptance of wife-beating. The level of acceptance was relatively low among married women, those who did not live with their partner and those married more than 10 years. Women who had final say or participated in household decision-making were less likely than others to consider wife-beating justifiable, while those whose partners were the main decision-makers were more likely to do so.

Fifty-three percent of respondents said that a man was justified in beating his wife in at least one of five hypothetical situations. Acceptance was highest for instances of a woman's arguing with her husband (39%), neglecting her children (37%) or going out without her husband's permission (33%). About three in 10 women considered a beating justified if a woman refused to have sex with her husband, one in 10 if she burned food.

Using logistic regression, the analyst examined the associations between women's background and relationship characteristics and their belief that wife-beating was justified in the specified situations. For each circumstance, rural women were more likely than urban dwellers to accept such behavior (odds ratios, 1.7-2.9), and the older a woman, the less likely she was to accept it (0.9-0.96). No other background characteristic was associated with acceptance of wife-beating in every situation specified, but several were associated with reduced odds of considering it justified in 2-4 situations: increased household wealth (0.96-0.98), having a primary education (0.7 for each of two reasons) or secondary schooling (0.4-0.5 for each of three reasons), and working in a high-status job (0.04-0.6 for each situation except refusal to have sex). Ndebele women had higher odds than Shona women of accepting wife-beating if a woman argued with her husband or neglected her children (1.3 and 1.6, respectively), but they, as well as women of other ethnicities, had lower odds than Shona women of accepting it in cases in which a woman refused to have sex (0.5-0.6).

Among relationship characteristics, duration of marriage was positively associated with acceptance of wife-beating overall and when food was burned or the couple had argued (odds ratios, 1.02-1.06); living with a partner was associated with elevated odds of acceptance overall, when food was burned and when children were neglected (1.3-1.4). Women whose partner had final say in household decision-making were more likely than others to accept wife-beating (1.1 for each reason), those who reported having final say had increased odds of accepting it if a woman burned food, argued with her spouse or neglected her children (1.1 for each); respondents who reported joint decision-making had reduced odds of justifying it for any reason except burned food (0.9 for each).

"Gender norms and expectations in Zimbabwe," the analyst remarks, "warrant the concern of public health researchers and practitioners." She notes that improving women's educational and occupational opportunities might help alter their perceptions of social norms, and promoting joint decision-making might help them achieve "equality in marriage." But she concludes that changing the acceptability of wife-beating in Zimbabwe will require widespread social and political reforms, including "improvement in national level regulations for women to . . . maintain control over their lives" and the promotion of "non-violent approaches to conflict resolution both within the household and outside."--D. Hollander


1. Hindin MJ, Understanding women's attitudes towards wife beating in Zimbabwe, Bulletin of the World Health Organization, 2003, 81(7):501-508.