In this study we examine whether gender bias in education depends on the extent of female decision-making power. Household headship is used as a measure of female autonomy, with different types of households theorized to reflect varying degrees of female autonomy. Most female-headed-households in Pakistan are formed either because women are widowed or because husbands migrate. Women in male-headed- households are hypothesized to have least autonomy followed by married women heads whose migrant husbands may retain some decision-making power. Widow heads are hypothesized to have the greatest degree of autonomy among women in different households. The econometric findings suggest that married women heads gender-discriminate as much as male heads but that widow-heads have significantly lower bias against girls in enrolment decisions than male heads. The results also suggest that educated female heads gender differentiate less than both uneducated female heads and than male heads. The evidence suggests that households having better educated women with more independent status discriminate against the education of their daughters less than other households.
Centre for Commonwealth Education, University of Cambridge, UK. WP07/03, 44 pp.