Department for International Development: Operating in insecure environments

Executive Summary

National Audit Office Value for Money Report

1. Many of the poorest countries in the world are affected by insecurity and violence (Appendix 1). Over 50 countries have experienced violent conflict within the last decade. Where a government cannot assure the security of its citizens it is rarely able to tackle poverty effectively and insecure countries are lagging behind other developing countries in their progress against poverty reduction goals. Insecurity has human and economic impacts, both for affected countries and their neighbours. Many insecure countries also receive lower levels of aid per capita than stable countries. These are good reasons to provide assistance in insecure countries but there are also difficulties and risks.
2. This report examines how the Department for International Development (DFID) works in insecure environments, ranging from some of the world’s most insecure countries where armed conflict is still present and stabilisation is required, to less insecure contexts where donors may have more scope to engage in long term development projects. It examines what DFID is achieving and how it designs and manages its programmes, including dealing with risks to its staff. We define insecurity by reference to the incidence of political violence and the level of threat to aid workers. Our work included four country case studies, literature and documentation review, data analysis and surveys (Appendix 2).
   
   

   Our findings:

   
   On DFID’s increasing interest in insecure environments
   
3. DFID has more than doubled its support to insecure countries within the past five years and plans further increases. In 2007-08 it spent over £1 billion, or 46 per cent of its bilateral expenditure, in 19 countries with significant insecurity. DFID has also diversified its expenditure in insecure environments so that long term development projects now outweigh its humanitarian expenditure there. DFID spends around five per cent of its country programme budgets on administration costs. DFID’s increases in expenditure in insecure environments followed its decision to increase support to ‘fragile states’ and conflict prevention. Post war recovery in countries emerging from conflict in the 1990s, such as Rwanda and Mozambique, also encouraged DFID to increase aid in currently insecure countries. It has been ahead of many other donors in recognising the importance of assisting insecure and often previously under-aided countries. As well as increasing its expenditure, DFID has encouraged other donors to work in some previously neglected countries and actively promoted coordination there.
   
   On what DFID expenditure has achieved in insecure environments
   
4. The overall context is that income poverty has not yet reduced in most insecure environments, although there has been progress against some important indicators, such as those for health. Progress in reducing insecurity has been mixed and its damaging effects hinder poverty reduction in affected countries. DFID has invested in security stabilisation in some previously insecure countries, helping to disarm warring factions and establish the conditions to address poverty. At project level DFID is achieving results in difficult circumstances, working closely with governments and other donors in priority sectors such as infrastructure, governance and conflict prevention. Despite the high risk, complex situations and experimental nature of many of the projects it has funded in insecure environments, we saw examples of projects which had achieved tangible benefits for poor people in all the countries we visited.
5. DFID rated around two thirds of its development expenditure in insecure countries as achieving all or most of its objectives, compared with around three quarters in secure countries. In the most insecure countries, around half of development expenditure achieved these ratings. Around 87 per cent of humanitarian expenditure was rated as achieving all or most of its objectives in highly insecure countries, not far behind success rates elsewhere. In highly insecure countries DFID has increased the proportion of assistance provided through development projects from 41 per cent of its country programmes in 2002-03 to 59 per cent in 2006-07. Much of the increase has been in governance and economic sectors, both of which are priorities in many insecure countries. But projects in those sectors performed less well in the most insecure countries than elsewhere. DFID has engaged in riskier or experimental projects in high risk contexts – many of which are unstable and operationally challenging – with a view to securing long-term benefits. Its average project success scores there declined as it did so.
   
   On the design and management of country programmes
   
6. DFID’s rationale for working in insecure countries is clear but its operational guidance is less developed. DFID country teams in most insecure countries assess the extent and nature of conflict, but these assessments rarely make explicit links to programme choices and management. There is guidance on the merits of different types of aid in different environments. But DFID has not so far made enough use of its staff’s experience in the field to feed into more practical advice; on adapting preferred approaches to highly insecure situations, collating security information to inform risk assessments and programme management. There is also little guidance about ensuring that individual projects and whole country programmes have no negative side-effects for insecurity or conflict. Although some projects clearly addressed sensitivities, such as targeting of aid on particular groups, country teams are not yet consistent in how they assess and respond to problems of insecurity. We found that other important elements of good project design in insecure environments included setting realistic objectives and timeframes, retaining flexibility and explicit consideration of long-term viability.
7. DFID uses a range of partners, including developing country governments, multilateral agencies and non-governmental organisations, to deliver its projects. Research on the relative effectiveness of different partners in insecure environments is limited. At country level DFID does not have a consistent and thorough approach to assessment of the relative strengths and weaknesses of different potential partners. Two thirds of the projects we reviewed had problems arising from weak partner capacity. In managing and monitoring its aid DFID considers risks at various levels although the quality of risk assessment in project design varied widely. DFID has to balance the need to monitor project progress with the risks of travel in insecure areas. It has however encountered poor performance and corruption by its partners and in some cases was slow to identify such problems due to limited project monitoring.  
   
   On managing staff resources and costs 
   
8. DFID staff are committed to delivering assistance in challenging circumstances. While staff express satisfaction with the skills of senior managers in country teams in insecure environments, there are still difficulties in building skilled and experienced teams. DFID has nevertheless been able to staff local offices in a range of insecure environments and in some cases it has more people on the ground than other donors. Issues such as high staff turnover, limited experience and staffing gaps remain challenges for DFID teams as they manage increasing budgets in insecure environments. DFID had not given security management sufficient priority when we began our study. Its approach to date in managing security and risk has varied between country teams, which led some staff to feel inadequately protected. But fully implementing the recommendations of a recent internal review should help DFID to tighten its security arrangements. DFID does not systematically collate or analyse the extra costs of running its business in insecure environments.
   
   Value for money:
   
9. DFID’s staff have worked hard in adverse circumstances and have delivered benefits for the poor. Achieving development in insecure environments is more difficult and requires different approaches compared with work in secure countries. DFID needs to ensure all its programmes in insecure environments are sufficiently adapted to the difficult context, in terms of both design and management. Learning has been hindered by incomplete project reporting. Quicker and fuller dissemination of the lessons from front-line experience will help DFID to improve the effectiveness of its aid there. The theoretical guidance which underpins DFID’s policy needs to be complemented by practical guidance on security issues, drawing on lessons learned from different countries. DFID is now starting to improve its approach to security management and to attracting staff to work in difficult environments. Its management information on the human and financial costs of operating in insecure environments has been limited, and it needs to develop stronger comparative analysis and management of its administrative and security costs.
10. We make the following recommendations:
    
    • There is limited research and experience on delivering effective aid in insecure environments, so the information on which DFID is able to base its decisions is weak. DFID should, with other development partners, continue to promote further research and evaluation of different ways of delivering aid in insecure environments. The emphasis of this work should be on practical approaches which work well in insecure situations, such as effective use of risk assessment and management to support staff to deliver projects safely in insecure areas. All annual and final project reports need to be completed and lessons on working in insecure environments should be well disseminated across its network.
    • DFID staff do not have enough practical guidance on working in insecure environments. DFID should use the results of research and experience of its own projects to provide more practical guidance to its teams. This guidance should help teams to:
      • make better use of information on the actual security situation to inform their operational decisions, such as when insecurity makes the financial and practical risks of an approach unacceptably high;
      • inform their country programmes through conflict assessments which analyse the risk that aid, by benefiting some more than others, could increase inter-communal tensions and therefore insecurity; or that insecurity could impair the effectiveness of different aid mechanisms;
      • ensure the design and day to day management of projects and overall programmes have no negative implications for security.
    • Weak partner capacity has undermined effectiveness in some insecure countries. DFID should establish a consistent approach to assessing potential partners’ capacity to operate in insecure places when considering their suitability to deliver services there. Where partners have less capacity than needed, DFID should provide capacity-building assistance before scaling up its funding. It should increase its use of leverage from central funding contributions to encourage multilateral partners to increase their capacity in key countries.
    • Successful projects have good design features that should be applied more consistently. DFID should ensure that the design of all bilateral projects in insecure environments includes:
      
      • realistic objectives capable of being monitored and a realistic timescale;
      • a thorough risk assessment; in particular how all projects and programmes could affect insecurity or be affected by it, including monitoring the impact of aid on different groups;
      • consideration of the prospects of longer term success and how the project will be reviewed after completion to learn lessons arising and to assess sustainability.
    • Monitoring in insecure environments is difficult, but DFID needs to identify and respond to problems as early as possible. Improvements should include more consistently:
      
      • requiring implementing partners to provide details of how their oversight will be sufficient to identify and report irregularities before agreeing funding; periodically checking this oversight by conducting visits; and sharing the results of visits with other donors where possible;
      • requiring multilateral organisations to provide DFID with a similar level of detail in progress reports as non-governmental organisations;
      • disseminating innovative approaches to monitoring in insecure environments, such as using video footage or sharing transport for planning and monitoring visits with other donors.
    • DFID needs better management information on its costs to inform its decisions and achieve value for money. To supplement improved information collection on results and experience of DFID’s work in insecure environments DFID should collect better information on its costs to allow it to assess the relative risks and costs of operating in different environments. It should also seek to get the most out of its administrative inputs by:
      • adopting a consistent approach to recording administrative and security costs;
      • obtaining data on the costs for other donors and UK government departments in each country and seeking to identify opportunities for cost savings and sharing, with consistent and transparent apportionment of costs;
      • promoting staff retention and productivity through better facilities and support for individuals who work in insecure environments.