The FRP began in 1995 as an innovative project to rehabilitate feeder roads in an area of Mozambique still suffering the consequences of long-term conflict. An important objective was to increase local interest within the public and private sectors and in civil society for sustainably improving physical access for people in rural Zambezia. Measures to achieve this included using provincial rehabilitation priorities which opted for labour-based methods of road building rather than the more usual mechanical building methods; training and employing a workforce comprising provincially-based contractors and indigenous workers.

By 1998 it was apparent that these aspects of sustainability were not just dependent upon factors affecting the availability of physical infrastructure, but on broader issues affecting people's ability to use and maintain roads. For example:

• roadside communities faced the threat of land grabbing by logging and other companies;
• people recently displaced by the war faced insecurity of land tenure;
• women were not sufficiently represented in road gangs although the project aimed for an equitable distribution of earnings;
• a question over the need for roads passable by motor vehicles above those passable year round by foot or bicycle;
• the threat of HIV infection to members of road gangs and road camp supporters

The approach of the project was sufficiently process-orientated to enable a closer look at these factors and complementary activities to be approved following an Output-to-Purpose Review in 1998 (DFID (1998): Report of an Output to Purpose Review of the DNEP/DEP Feeder Roads Project. A Mission by A Wray et al. Mozambique, 12 - 23 July). These innovations amounted to:

• 1998 - adding a component to address land insecurity among roadside communities;
• 1999 - introducing measures to redress gender imbalance in recruitment to road gangs (men are now recruited for work gangs only after a target figure for women recruits has been reached);
• 1999 - piloting of women-only maintenance gangs; commissioning a study of other barriers to employment for women;
• 1999 - raising awareness of HIV/AIDS and setting up mitigation activities (e.g. introducing theatre groups, condom distribution and health workers in camps);
• 1999 - strengthening the socio-economic impact analysis and the selection of roads for rehabilitation;
• 1999 - commissioning an environmental impact study;
• 1999 - providing emergent contractors with assistance in business development;
• 1999 - commissioning a study into 'Barriers to Access' in Zambezia and holding a stakeholder workshop to discuss the findings;
• 2000 - adopting a revised logframe (this refers to added ORAM component and to previous modifications of OVIs, especially the adding of qualitative indicators);
• 2000 - reviewing the findings of these initiatives and the scope for further support to enhance livelihoods through access improvements.

In shifting from a focus on the product - the roads - to how roads and road building affect the people who live nearby and who might eventually use them as part of their livelihoods strategies, the project exemplified a key characteristic of sustainable livelihoods approaches - the fact that they are people centred.