On target? Why participant selection matters for development programmes

| April 13, 2015

icrisat.images_6119486174Many development programmes reach only a fraction of the people they aim to include. One reason for this is that attrition erodes target group participation at various stages between programme conception and completion. Programme targeting using selection criteria, eligibility assessment and participant registration is one of the ways this problem can be addressed. But how far does targeting address this issue?

3ie’s new systematic review on targeting and farmer field schools (FFS) throws up several insights about how targeting decisions can have a real impact on the ground. Evidence shows that many FFS programmes do not necessarily reach the participants they target. The key reason for this was that criteria for eligibility, the selection procedures used and the realities of potential participants were not always compatible.

The targeting debate

Discussions about targeting have typically focussed on the relative merits of universal versus targeted programmes and the problem of how large-scale programmes can reach the poorest in society. However, most development programmes focus limited resources on a specified population, with many targeting women, youth or other groups. The challenge of targeting is one that is of equal importance for these types of programmes.

Amartya Sen pointed out that the word ‘targeting’ turns beneficiaries into passive recipients instead of the agents of change. Targeting also focuses our attention on devising the most accurate way of identifying beneficiaries. However, people act and react to the programmes offered to them. Factors such as implementation, participants’ characteristics and programme context all have an important role to play in determining who participates – and by extension, who it is that stands to benefit.

Targeting analysis: the case of farmer field schools

FFSs have become one of the most common approaches to rural adult education and agricultural extension in the world. They have been implemented in over 90 countries, reaching an estimated 10 to 20 million people.

We found in our systematic review that some FFSs target the poorest or most disadvantaged groups in society on the grounds that they are most in need of the benefits that these programmes provide. However, many FFS programmes target farmers with more resources, more education and greater social agency, with the aim of maximising programme effectiveness.

Programmes targeting more experienced and educated farmers typically succeeded in reaching their target group, irrespective of the procedure they used to select them. The main driver behind this success was that the target groups’ socio-economic characteristics favoured their inclusion.

However, targeting was less successful when programmes were designed to be generally inclusive or to include a specified group such as female farmers. One example from the systematic review is illustrative of a wider trend of an FFS programme in Uganda. The programme intended to be inclusive. But community leaders’ prominent role in selection meant that many final participants had social connections to recruiters. In other cases, it was not selection criteria or procedures that precluded some peoples’ participation, but a lack of time and the resources needed to participate. In Liberia, an FFS programme targeting women failed to reach many female farmers because not enough consideration was given to their limited access to land and tools or their existing household and childcare commitments.

Programmes that were more successful in reaching their target groups were carefully designed, so that selection procedures were not prone to elite capture. They also incorporated additional components to ensure that the poorest or most marginalised had access to the resources needed to participate. Overall, our analysis highlighted the importance of a coherent and consistent logic underlying the targeting process. This encompasses not only the choice of who should participate and the identification procedures used to select them, but also consideration of the characteristics of potential participants. The goal of identifying target groups as accurately as possible should be balanced against the need to ensure that participants are able to participate.

Why targeting matters

In attempting to ensure that a programme reaches its intended audience, targeting aims to minimise resource wastage and increase cost-effectiveness. Of course, targeting is never without cost, and identifying the poor with real precision is likely to be costly. Targeting is also a zero-sum game. Money spent on it reduces the budget for addressing poverty and welfare goals. However, a decision not to spend resources on targeting is still a choice that can have important consequences for programme effectiveness and social equity.

The review of FFS programmes was based on a range of different evaluations that typically did not report on financial considerations or programme budgets. As a result, we were not able to examine how much time and resources were put into planning targeting and selection. However, there was some indication that implementers were sometimes caught between competing demands to deliver a programme that would have an effect on outcomes of interest, to be as inclusive as possible and to meet the practical necessity of reducing costs. Consequently, targeting and selection decisions were not always aligned with programme objectives, target group characteristics and the wider intervention context. Devoting some time and resources to targeting and selection considerations could go a long way to improving programme performance. By ensuring that the target group, selection procedures and target-group characteristics are as aligned as possible, programmes may stand a better chance of reaching those most in need.

Tags: ,

Comment on “On target? Why participant selection matters for development programmes

  1. Hugh WaddingtonHugh Waddington

    Thanks Daniel. I think one of the interesting points raised in your study is how much programme assignment ultimately DOES matter for external validity of impact evaluation findings. I am a big supporter of randomised assignment, as long as it provides policy-makers with a useful ‘impact quantity’. A really good example of randomised assignment is Rohini Pande’s study of pollution control in Gujurat (http://www.3ieimpact.org/publications/3ie-impact-evaluation-reports/3ie-impact-evaluations/3ie-impact-evaluation-report-10/) – a perfect experiment from an external validity perspective since random selection of factories should actually be a component of future roll-out. In most cases, randomisation will definitely NOT be part of any continuation or up-scale of the policy so the issue of targeting (or whichever term you feel comfortable using) becomes potentially crucial in programme effectiveness. More reason why we need systematic reviews such as yours.

Comments are closed.