This post is part of a learning series exploring good practice, examples, or new applications of MSD.
Thanks to Tim Stewart for this contribution.
As MSD practitioners, we are tasked with positively and sustainably impacting the poor, but what about the ultra-poor? Are criticisms valid that MSD doesn’t – or can’t – benefit the ultra-poor? Well, that somewhat depends on who’s defining them and how you reach them.
We examined 41 self-identified MSD programmes worth over USD 1.5 billion from 25 countries for their definitions of poverty and how they identified and impacted poor and ultra-poor target groups. Over half of these (22) were DCED-audited and a further ten were using the standard for monitoring and results measurement.
Firstly though, who are the ultra-poor (also termed ‘poorest of the poor’, ‘extreme poor’ and ‘hardcore poor’)? It turns out that there isn’t a commonly accepted definition. The UN and World Bank’s extreme poverty definitions are simple Purchasing Power Parity (PPP) figures of USD 1.25 and USD 1.90 respectively but these can be difficult to use for targeting and measurement purposes because incomes are not readily observable and rarely disclosed.
In response, multi-dimensional poverty indices have been developed that use criteria that are observable and more directly linked to wellbeing, such as nutrition, child mortality, schooling, access to cooking fuel, condition of housing and ownership of assets. Others such as the Chronic Poverty Advisory Network (CPAN) use similar criteria, and also define a ‘severe poverty’ PPP class of living on less than USD 0.7/day. Some MSD programmes use the Poverty Probability Index (PPI) which uses similar multi-dimensional poverty parameters to determine the probability that a given household is below a national or internationally defined poverty PPP threshold, but the PPI does not in itself constitute a definition.
The lack of a commonly accepted ultra-poor definition is a challenge for implementors and critics alike. Critics require a definition in order for their criticism to be valid, otherwise it’s just whatever they say it is. Implementors can find it challenging to satisfy the demands of critics and donors if there isn’t an agreed definition, and it is easier to enter into a spiral of demands to ‘do more’.
Leaving aside the ultra-poor, we examined how well MSD programmes defined their ‘poor’ target group. We found that MSD programmes were not very precise in defining their poor target groups: only around half (24) defined poverty using an economic (PPP), asset-based definition or both in their designs. The remainder used definitions such as ‘smallholder farmers’ without even defining what was meant by a smallholder in land area or other terms. However, around 20% of these went on to define them in more detail as the programme progressed (this is what inception phases are for!). Moreover, a substantial proportion did not set clear targets of how many poor they aimed to reach or by how much they aimed to increase their incomes: only a quarter did both. Clearly then, MSD programmes and their designers need to get better at analysing and defining poverty and using inception phases and adaptive management to act, learn, refine and respond to improve their targeting.
The main reason we identified for MSD implementers neither defining nor targeting an ultra-poor target sub-group is that they simply were not asked to by their donors in programme designs. Only one in our sample did, categorising their sectors for inclusiveness and growth, and identifying ones that had high potential in both, such as medicinal plants.
Most MSD programmes in our sample worked in agriculture: half were exclusively in agriculture and the remainder were in agriculture and other sectors. A simple ‘smallholder farmer’ definition is used – and is probably sufficient as far as the donor and country partner (e.g., host government) are concerned. It is assumed that the poor tend to be smallholders and smallholders tend to be among the poorest in society and that reaching the ultra-poor is a matter of reaching the poorest smallholders.
In rural areas, however, the ultra-poor may be landless and may be engaged in daily labour, collecting non-timber forest products, grazing livestock along road verges or on harvested fields. However, many ultra-poor may also be found in urban areas: although declining, around 29% of those living in urban areas live in slums. The poorest among these may also be engaged in daily labour – e.g., as porters, or work in other sectors such as solid waste management (picking and selling rubbish into the recycling supply-chain), construction, domestic work, or even begging. They may also migrate nationally or/and internationally. The ultra-poor may also be reached through other markets such as health, education, welfare, sanitation, and energy.
The overwhelming focus on agriculture, though appropriate for reaching large numbers of the poor, ignores opportunities to reach the ultra-poor. This is not a limitation of MSD per se but reveals perhaps the constraints of those who deploy it who may perceive it as an approach that is only applicable to agriculture or prefer to use more direct approaches to reach the ultra-poor. The ultra-poor deserve better: they are the victims of systemic failures, and systemic problems require systemic solutions. If donors want to impact the ultra-poor, they must be prepared to design and mandate programmes that are not in agriculture.
On the implementor side there may be perverse incentives at play. If development contractors and NGOs have to pursue large numerical impact targets upon which their commercial performance may rest, then it makes sense to target sectors that can support this achievement – the low-hanging fruit. Indeed, the donors who set programme objectives may also want to achieve large numbers, driven by value for money – VFM – (e.g., cost per beneficiary) and political reasons. From a utilitarian perspective, this may be the most good for the highest number of people, but it denies opportunities to create ultra-poor systemic impact.
Impacting the ultra-poor therefore requires working in non-traditional MSD sectors and using tactics that may look different to those commonly used in agriculture. Such tactics may need to be more direct, but this does not require a re-brand of MSD to a hybrid or reach down/lift-up approach. MSD tactics are, and have always been situational: same approach, different tactics. Anything is allowable provided the analysis has identified the entry point, that there is a clear vision of sustainable future performance, and there are legitimate market players that can be engaged with to bring it about – be they public, private, social or informal sector. This will require more flexibility too, and an acceptance of different VFM and impact metrics.
In summary, we recommend the following for improved MSD poor and ultra-poor targeting:
Define & Design: Be clear about the target group/s and their poverty characteristics, be sector agnostic to include sectors that engage the ultra-poor in rural and urban areas and consider the poor as consumers of services such as welfare and education. Also, donors need to consider different metrics and incentives for ultra-poor MSD programmes.
Decide: Funders and implementors need to determine what sustainability means to them, and decide what trade-offs are acceptable in reaching poorer people. This could be greater cost per beneficiary, longer timeframes or/and less outreach.
Data: Gathering evidence on poverty targeting, impact and outreach across MSD programmes is labour intensive and time consuming. It needn’t be. Most MSD programmes are using the DCED standard, and track Net Attributable Income Change and outreach. It would be relatively simple for donors and implementers to make these figures public and for agencies such as the BEAM Exchange to aggregate this data and allow for easier comparison of programme performance.
MSD, then, is absolutely fit for purpose to impact the ultra-poor in a sustainable and systemic way. It just needs to be deployed in the right way by committed donors and practitioners. However, whether it’s the poor or the ultra-poor, there is clearly room for improvement in the way that programmes define and target them.