NewThinking.org.uk

Do 'Best practices' for managing development projects help or hinder?

Many have raised concern about the effectiveness of development projects, both on a macro-level (e.g. Easterly 2006) and on a micro view (e.g. Dichter 2005). One often quoted problem is ineffective project implementation.

In response, funding agencies / donors use a range of procedures to increase project effectiveness. These include donor coordination, target group definition, tight project budgeting, logical frameworks, etc. These standard procedures have now become a requirement for many funding agencies, who believe all effective programs can comply.

This short paper shows how, in some cases, these procedures can distort project design and lead to good projects being rejected. It proposes a more flexible mind-set and more selective application of project management tools. This mind-set is more consistent with 'searching' for good development projects as proposed by Easterly (2006).

The paper uses the example of re-capitalisation of SANASA micro-finance cooperatives in Sri Lanka, following the Tsunami. SANASA is a network of approximately 8,000 thrift and savings cooperatives. 390 of these cooperatives, with a membership of 70,000, were in areas impacted by the Tsunami. Societies suffered from high levels of bad-debt and many were bankrupt. To survive, they needed a capital injection (a grant) equal to the losses incurred - a total of approximately $1.5m.

SANASA agreed a re-capitalisation program with the Canadian International Development Agency (CIDA) in August 2006. As of March 2007 the program had been implemented in the South of Sri Lanka and was underway in the East where conflict made implementation more complex. Overall, the program has been a success. It led to 380 MFIs being re-capitalised, ensuring the long-term delivery of micro-finance services to a population of 70,000. It delivered $1.5m of 'post-disaster' assistance to the people most affected by the Tsunami, money that will be used for long-term livelihood development. It reduced pressure to re-repay loans on those most impacted by the Tsunami. It delivered this with minimal overhead - 10 local staff worked on the program for a period of 4 months.

However, it took SANASA many months of negotiation with a number of funding agencies (both governmental and international NGOs) before support was secured. The delay was because the program did not comply with many of the procedures of donor funded projects.

Problem 1: Donor coordination

Donor coordination aims to make funding agencies' actions consistent and minimise overlapping programs. Almost every agency subscribes to the practice of donor coordination and there are countless conferences and workshops dedicated to it.

In practice, funding agencies meet to 'divide up responsibilities' for different sectors or geographies. At an early donor coordination meeting following the Tsunami (to which SANASA was not invited), responsibility for re-capitalisation of SANASA societies was allocated to a particular agency. This agency, however, did not realise that re-capitalisation required a grant, which it was not able to provide since it only provided loans to micro-finance institutions (MFIs). Even once this was realised, other agencies were reluctant to work with SANASA since this would go against the objectives of donor coordination.

'Carving up the market' eliminates performance incentive for agencies. Recipients of aid cannot choose to go to a different funding agency if one proves to be incompetent. A mechanism where agencies would compete for the right to work with recipients would increase their responsiveness, and ensure the type of confusion that SANASA faced is not repeated. Practically, this is what happens when there are a large number of donors and few local institutions, as was seen after the Tsunami in Sri Lanka. But this approach is fundamentally in conflict with the approach of donor coordination.

Problem 2: Identification of target segments

Many agencies require target segments or recipients to be exactly identified. They need to ensure their funds are focused on the poorest, need to report the gender mix of recipients, etc. The re-capitalisation program was not able to identify recipients, since funding was given to the micro-finance institutions, not to individuals. As a result, some donors rejected the program.

The program did, however, focus on the poorest. The people who benefited most from the program were those who were unable to pay their micro-loans - by definition poor. However, it was the nature of the intervention that targeted assistance to the poorest, not an explicit selection process. Agencies should be more willing to accept this mechanism.

Problem 3: Project budgets

Almost all funding agencies need an exact budget to fund a project. But a re-capitalisation program is not able to create an exact budget until it completes loan-by-loan analysis. Only an estimate is possible.

Exact budgets may not be the best solution in other programs also. Projects should be funded further if they are going well, or if additional needs are proven. Projects that are not going well should be stopped early.

Problem 4: Desire to ensure a 'proven sustainable impact upon livelihood' (consequence of logical frameworks)

Some funding agencies were initially interested in supporting the re-capitalisation program, but later wanted to add additional components to the project such as training or individual capacity building as they did not believe that 're-capitalisation alone would have a sufficient impact upon the livelihoods of the target population to create sustainable poverty reduction'.

However, adding additional components would have increased the overhead costs and delayed the project, making the re-capitalisation calculations more complex. Some funding agencies understood this, but said that they needed to add extra components to 'complete the logical framework' that had poverty alleviation as the overall goal. Rather than 'doing what they could', they wanted to 'do all they could' even if it was not likely to be effective. Funding agencies should focus on projects with more limited ambition that are likely to be effective, rather than making projects complex and difficult.

Problem 5: Desire to build institutional capacity / involve technical assistance

One of the reasons the re-capitalisation program has been a success is that it is very simple to implement. No foreign technical assistance was required to manage the program. No additional training of local staff was needed. The whole program was implemented with 10 existing local staff.

This simplicity ran against the expectations of some agencies. Some asked if a program of this size could be implemented without external assistance. Others stated that 'sustainable impact could not be achieved unless the program also built capacity within SANASA'. Agencies should accept that a simple project design is a better guarantee of success than extensive TA programs.

Problem 6: Need for donor visibility

Re-capitalisation did not create any new buildings, develop identifiable new businesses or provide any training courses. There were no photo opportunities. This turned off some agencies.

In summary, the procedures employed by development agencies, which have now become requirements, do not always help. While they are each useful in some circumstances, they should not be slavishly applied, otherwise good projects will be rejected or delayed. Because the SANASA re-capitalisation program did not comply with some of the standard procedures, there was a considerable delay in identifying funding, and the project nearly did not happen. The project now faces a number of implementation challenges caused by this delay.

This does not imply that the project management procedures should be rejected, many projects benefit from having a budget, or defined target segments. It does imply agencies should actively decide when to apply the tools. Agencies should first understand the level and nature of implementation risk in a project to determine if a tool needs to be applied. A project like the SANASA re-capitalisation, which has few implementation risks, will need few procedures.

To make this judgement, agencies need to make a mind-shift, from seeing themselves as project designers to 'providers of funding services'. Funding agencies should shift planning responsibility and performance accountability to the implementing agency by simplifying procedures and making future funding dependent upon the outcomes of previous projects, which would allow the funding agency to use a simplified set of controls and feed-back mechanisms.

The change in mentality is similar to the change that banks have gone through in retail lending. Previously, loans were 'granted', applicants had to fulfil a number of requirements and submit complex analysis. Now loans are 'sold', banks have found simpler ways to predict loan re-payment, based on lending record. This lowers cost and is more customer friendly, reflecting the fact that loans are products that need to be sold. Similarly, funding agencies should view funding as a 'product' and find simpler ways ensure funds are well used.

Easterly, W. 2006. The White Man's Burden: Why the West's Efforts to Aid the Rest Have Done So Much Ill and So Little Good. Penguin Press, New York.

Dichter, T. 2005. Time to Stop Fooling Ourselves about Foreign Aid: A Practitioner's View. In Cato Institute Foreign Policy Briefing no 86, September 2005