The challenge, of connecting micro with macro analysis raises questions, reflected in Doyle's and Bujra's articles about the efficacy of projects in significantly affecting gendered power relations. They both question women-only projects, and the effectiveness of interventions which use gender as an entry point for instrumentally tackling development problems without facilitating wider empowerment or equality. As Doyle reports, women at a workshop on AIDS awareness in Vietnam requested that men be similarly targeted. However, when men did participate, they changed their behaviour in sexual relationships but not the way they fundamentally thought about gendered relations of power. Bujra's research in Tanzania and Zambia also asks whether AIDS awareness campaigns significantly affect gender balances power and suggests that what needs changing is not the behaviour of individual men and women but the relations between them.

Changing ideas about men's roles, varying cultural conceptions of masculinity, and the need to challenge dominant definitions of 'what it is to be a man', are all strong themes in reported experience of dealing with men and masculinities. This is well illustrated in Thomson's description of Save the Children's work with boys in the UK who struggle to cope with changing roles (with the 'crisis' of masculinity), the discrepancy between publicly-sanctioned gender roles and what actually happens in families, and the dynamic nature of gender relations. Indeed this dynamism echoes throughout these articles, several of which link difficulties men may experience with responsible partnering and parenting with changing expectations of employment and wider societal change.

Montoya focuses on a campaign in Nicaragua aimed at preventing men's domestic violence, emphasising the need to understand the fears and insecurities that men experience in their relationships with women. Interestingly, Montoya links the increased tensions and conflicts in families to the environmental, economic and social devastation caused by Hurricane Mitch.

Smith's article on Oxfam projects supporting disadvantaged men in the UK highlights the problems associated with lack of employment and the stereotyping of alternative employment opportunities as 'women's work'. The changing shape of men's working lives and the ways in which policy interventions conceptualise gendered divisions of labour are key issues in studies of men and masculinities (see European Journal of Development Research, December 2000). For example, are different forms of work empowering or oppressing for men and women? How do gendered labour allocations impact upon health and wellbeing? Is it correct to assume that ‘women do all the work’ in developing countries? There is an urgent need to further investigate relations of power and domination in men’s working lives.

Kandirikirira's and Dolan's articles explicitly link individual state action (or inaction) with the development of damaging forms of masculinity, expressed in violence. They differ significantly however in their suggested policy implications. Kandirikirira attributes the sexually violent and abusive relationships between boys and girls in Namibia (sanctioned or ignored by elders) to the policies of the previous apartheid state which systematically distorted the image of black people and restricted their opportunities. Participatory approaches have begun to help overcome this legacy as individual stakeholders become aware of their own responsibility and capacity to tackle injustice and inequitable relations. Dolan, by contrast, in analysing the prevalence of gender- related violence in Uganda, attributes this to the weakness of the state, to its incapacity to maintain the rule of law and to the threat to masculine identities that this constitutes. The developmental challenge, Dolan argues, is to hold states rather than individuals to account and to focus more widely on the political context in which masculinities are formed.

Men and masculinities is a relatively new area in gender and development. Ideas concerning policy implications are in their infancy. How can research, policy, and training contribute to the debate and complete the shift from WID to GAD so that the situation of women and men is better understood? Suggestions include:

• investigating the changing roles, needs and identities of men over lifecourses

• researching men’s roles in families, the reproduction of gender inequities through work, and men’s specific health vulnerabilities

• tracking and monitoring changes in gender relationships over time, in different cultural contexts, in association with programmes and policies

• developing positive role models for men and boys by influencing mass media images, establishing activities in schools, NGOs, religious and youth groups

• ensuring that legal frameworks supports gender equity, through regulating working hours, parental leave provision, improved maintenance and inheritance laws, for example

• improving gender training within development organisations to focus on gender and not women alone: for example by increasing the number of male gender trainers and improving gender analysis frameworks.