X Close

The Bartlett Development Planning Unit


Collective reflections about development practice and cities


Why is it so tempting for livelihood projects to ignore poor people?

By Julian H Walker, on 9 February 2012

PHOTO: J. Walker

Approaches to development which prioritize economic growth have been consistently criticized on the basis of their trickle-down assumptions, and for losing sight of equality as an objective. An ongoing theme in international development, therefore, has been attempts to develop frameworks which ensure that the needs of poor women and men are understood, and catered for, in economic development and livelihood programming.  Yet all too often there seems to be a sort of slippage between the intention of such frameworks and their application, whereby, in practice, they are perversely used to justify the exclusion of poor people.

One anti-poverty framework which is currently in the ascendancy is the ‘Making Markets Work for the Poor’ (M4P) approach[1]. Championed by major donors such as the UK DfID and the Swiss Development Cooperation (SDC), M4P was developed in part as a corrective to previous anti-poverty approaches, such as the Sustainable Livelihoods Approach, which,  while it had a strong emphasis on the strategies and assets of poor households, failed to robustly address the structural and institutional constraints that exclude poor women and men from accessing markets and employment.

In this light, the M4P approach explicitly aims to combine growth with active measures to ensure the access of the poor to markets. Its intention is to foster systemic change focused on the systems of entitlement and the (formal and informal) ‘rules’ or institutions that support or impede poor people’s access to and control over markets.

Over the last six months I, along with DPU associate Nadia Taher,  have been working with the Swiss Development Cooperation (SDC) in their South Caucasus Progamme[2] (which covers their development assistance activities in Armenia, Azerbaijan and Georgia). This work builds on a long history of collaboration between the DPU and the SDC working on gender equality issues. In this case we are working with the SDC’s Gender and Governance advisor to bring a stronger gender equality focus to their work in the region.

One of the issues that we are addressing with the SDC team is how to make sure that the eight Economic Development and Employment (EDE) projects that they support in the South Caucasus (which have all been structured using the M4P approach) actively cater to the needs of poor women and men. These eight projects, reflecting the prevalence of poverty in remote rural areas in the South Caucasus, are all in the field of horticulture and animal husbandry, and attempt in different ways to connect village households dependent on farming to urban markets for their produce.

A critical issues here is that, not infrequently, the NGOs implementing these EDE  projects appear to be working in ways which sideline the inclusion and interests of the poor. Many of the project teams feel that they should primarily focus on support to existing private sector enterprises (including, in some cases, established international businesses), but are hesitant to work with the poorest women and men farmers or agricultural labourers.  Ironically, the justification for this approach is typically that to work directly with poor women and men would contravene the principles of the M4P framework, which proposes a ‘light touch’ facilitative approach, working with existing actors and processes, rather than interventions which create new processes and institutions, which are dependent on the project and may therefore be unsustainable. Whatever the justification, the outcome is odd for poverty focused projects when,  for example, it is seen as in keeping with the M4P framework  to purchase lorries for an established dairy processing business, while direct interventions such as support to the creation of cooperatives or famers associations, or start up grants or loans to poor households are disallowed on the basis that they are ‘unsustainable’.

In other cases, project interventions prioritize the interests of growth, but do not attempt to promote equitable access to the wealth created. For example one project focuses on supporting established businesses to develop fruit processing in high quality fruit value chains, arguing that this will create wage labour (casual agricultural labour and work in processing facilities), but envisages no interventions to support the rights and labour conditions of a casual agricultural labour force, despite the fact that this is a labour force which is notoriously vulnerable to poorly paid and exploitative working conditions.  Thus their interpretation of the systemic change envisaged by M4P appears to be about changing the systems of market access for medium sized business, while leaving the systems whereby agricultural labours are exploited untouched.

The SDC are aware of these issues, and, in response, have been stressing that there is space within the M4P framework for a more active, rights based interpretation. For example, they point out that the M4P approach advocates working with a full range of ‘market players’ – and, while it is important to work with the private sector in the interest of economic growth, rights based and pro-equality interventions also require working with other market players, specifically supporting civil society and government bodies working on issues related to labour rights, governance, and market regulation, or producers associations which protect the rights and negotiating capacity or women and men engaged in farming.

So why, in this case, is it that the application of pro-poor frameworks such as M4P often lead, in practice, to pro-business interventions which sideline the poor?  Is it that models of growth-led development are so embedded in our minds that we can’t take alternative forms of enterprise, such as cooperatives, or state regulated markets, seriously? Or that the ways in which the performance of economic development projects are measured (for example economic return on investment) mean that a truly pro-poor orientation will always score badly in the short term? Or that poor people are difficult to reach, because they don’t fit into neat organised associations which are easy to work with, and conform to the requirements of our framework? Or that dealing the institutions that underpin poverty requires confronting vested interests, and sensitive political structures that project teams feel are ‘out of reach’? Whatever the reason, it seem very clear that however sound frameworks  such as M4P are on paper, we need to apply constant critical scrutiny to what they deliver in practice, as they have a tendency to create a new logic all of their own when they hit the real world.

[1] http://www.m4phub.org/

[2] http://www.sdc.admin.ch/en/Home/Countries/Commonwealth_of_Independent_States_CIS/Southern_Caucasus_Georgia_Armenia_Azerbaijan

3 Responses to “Why is it so tempting for livelihood projects to ignore poor people?”

  • 1
    David Elliott wrote on 18 September 2012:

    The blog makes reference to the limitations of “remote reformers” (growth focused) at one end of the spectrum, and “impulsive interveners” (poverty focused) at the other. M4P doesn’t see these growth and poverty objectives as mutually exclusive, but as mutually reinforcing. It believes this not because of some academic notion of conviction or ideology, but because in the real world evidence is clear that growth without poverty impact is destabilising, and poverty can’t reduce sustainably without growth.

    Empirically, developing countries with similar rates of economic growth have experienced quite different levels of poverty reduction. There is no linear rate at which economic growth translates into reduced poverty levels. Take Armenia by way of example. From 1999 to 2003, GDP grew from USD 572 to 874, and the population living below the poverty line declined from 55% to 43% over the period. However, this growth and poverty reduction was not linear. Poverty rates in Yerevan declined by around 37%. In Suynik, the region of focus of the SDC programme this blog refers to, poverty rates increased by almost 6%.

    In light of such dynamics, at the time the dominant approach of SDC (and most other donors) in Armenia (and the South Caucasus region) was to respond directly to these problem pockets of poverty. They worked with local authorities to be more responsive to their citizens. They worked to organise selected poor to be grateful recipients of handouts (cows, bee hives, rosehip collection buckets etc). An independent and high level agency review of the SDC portfolio in the South Caucasus at the time (around 2003) confirmed these efforts were scattered, diverse with no sense of strategy, cohesion, limited (if any) tangible impact, and zero likelihood of sustainability. The poor groups were formed by projects, and entirely dependent on them. They were supported to work around, rather than with and through main market channels. The poor were being increasingly isolated and dependent – hardly a vision of equality or empowerment. Government was happy to take the new cars, computers and office equipment, but did little to be more accountable or responsive to its citizens.

    The conscious shift in approach from SDC at time was driven by a clear sense of failure of existing approaches to form project groups (in isolation of wider economic processes), and to support local authorities on governance matters (in isolation of wider political economy processes). The shift towards more of an M4P approach was based on the belief that development could do better, and doing so required much greater focus on poverty – understanding the potential (assets, skills, resources, connections etc) of the poor in regions such as Suynik, Racha and Agjabedi – and growth – how the poor can better connect and participate in the wider growth processes in Armenia, Georgia, and Azerbaijan.

    The question posed by the author is whether this is all just unfounded rhetoric; some duping grand conspiracy to bolster the profits of the rich and further indenture and marginalise the poor. A conspiracy overseen by SDC as the funder, by all the implementing partners (Oxfam, CARE, Mercy Corps, Inter-Cooperation, HEKS-EPER to name a few) and lubricated by the technical backstopping efforts of The Springfield Centre. Well, let’s explore this in response to the myriad of rambling, partially informed, ideologically fuelled comments and criticisms:

    Does sustainability and scale matter?
    Sustainability is at the very heart of M4P, as is the promise of impacting at scale through tackling the causes of underdevelopment rather than sticking plasters on its many symptoms. The blog criticises M4P for preferring to work through, and shape, existing actors and processes, rather than creating new ones which are project dependent and unsustainable. It is criticised for not, as a point of principle, creating new cooperatives and farmer associations, and dishing out directly dollops of grants to poor households. Well, the answer is simple: decades of development is littered with moribund project driven cooperatives and associations. They are not sustainable and don’t endure because they are formed by projects, for the benefit of projects, and fail when projects end. And all development projects end. Similarly, dishing out the goodies to the lucky few without changing the risk profile or appetite for risk of financial services providers or investors will undermine scale, and cannot be sustainable.

    Does this mean that M4P, on principle, rules of working to support cooperatives and associations, or using grants or loans to pursue development aims? No, it does not. But it will only do so if such actions respond to a clearly identified problem, and can be part of a sustainable and scalable solution.

    Does working for the poor mean only working with the poor?
    In a paper written some years ago by SDC, analysis showed that dividing all the official development assistance by the number of chronically poor people in the world would mean an annual development budget of around £0.87 per poor person. With such limited resources, working only with the poor to directly meet their needs clearly wont’ get us very far at all. Leverage is important. Private investment within, and into developing countries dwarfs development budgets. Working with private enterprise to recognise better the value of the poor – as employees, as business partners, as consumers – and to invest more in the poor is as legitimate as it is imperative.
    Bizarrely, the blog seems to differentiate the poor from private business. The poor that the blog refers to, mainly, are agricultural producers. They produce and market surpluses. They might be farmers, but they are private farmers – and farm as a business. Seeing the poor as, well, the poor – not as businesses, not as economic actors, without aspiration and needing of charity – isn’t we would argue particularly empowering.

    Rights or realities?
    The author seeks to take a principled position of rights based approaches, and when projects fail to live up to whatever utopian ideal the author has of rights, then projects are seen to have failed. Access to water is a human right. Yet, more than 1.2 billion people in the world still lack access (i.e. quality, volume, and proximity) to water. This number of hasn’t really changed for more than 50 years, in spite of 98% of all fixed water infrastructure in the world being in public hands. Some people hide behind rights and notions of what should be. Those interested in M4P respect such rights and principles governing them, but work in the real world and try to deal with and improve people’s realities.

    If the blogger lived in the real world, and took an interest in history he’d learn not only from failed development experience in the Caucasus region as referred to above; but also from the long history of failed cooperatives and state dominated markets that he seems so keen on. Should he care to ask poor farmers in Georgia, Armenia or Azerbaijan, he’d find few have any interest or appetite for re-joining the state led kolkhoz (cooperative)…!

    Objectives and performance measurement
    Poverty is – at its core – about incomes, but it is also about the wide set of characteristics often associated with lack of income. It is related to assets and income-earning opportunities; to consumption; to nutritional and health status and access to appropriate services; to learning and educational opportunities; to wider political freedoms and rights; to people’s ability to deal with shocks and insecurity; and to their status and sense of dignity.

    This diversity of poverty’s symptoms means that it does not lend itself to neat definition. Yet, from a programming perspective projects must define what poverty means for them, to set objectives and a clear framework for performance assessment and results measurement. This means making choices, narrowing down and tackling a particular aspect of “poverty”.
    For the SDC portfolio in the South Caucasus this has meant a focus on income poverty – consistent with the first of the MDG’s, and perfectly legitimate in its own right rather than as some intermediate proxy for any and all other aspects of poverty. The performance of the portfolio in the South Caucasus should be measured against what it’s trying – legitimately – to achieve, which is improved incomes for poor men and women in the areas and sectors defined for each project in the portfolio. So, how is it doing in this respect…? Well, a recent independent strategic assessment of SDC’s work in the region confirms:

    a) The high degree of relevance of the M4P approach in responding to the multifaceted challenges of generating higher returns to rural enterprise in the South Caucasus region;
    b) The programme has become more consistent, and has enhanced its credibility since following a unifying conceptual approach, as provided by M4P and
    c) The approach is delivering real and sustainable impact, and can be scaled up.

    What does this mean in more tangible terms? Well, for example, for thousands of farming families in the Armenian region of Goris, income poverty among farmers was reduced by over 10% during a time when poverty rates were increasing in Armenia. Poverty amongst the very poorest (living on less than US$1.60 a day) reduced by 9%.

    Not all projects in the portfolio are performing equally, but the majority are starting to deliver sustainable impact, at scale. They should be applauded and supported for delivering what they’ve been tasked to deliver; not vilified for falling short of things they’ve not been tasked to deliver…!

  • 2
    Tim Stewart wrote on 18 September 2012:

    An important question is raised by this blog entry: can approaches to poverty reduction that do not work exclusively or even directly with the poor (such as M4P) achieve their objective of inclusive growth. The author thinks not. As a former manager of an M4P program in the South Caucasus, I would like to correct this assertion and some of the inaccuracies presented as evidence, and call for the author to base their critique on a more balanced review of the body of work which exists on M4P, and a presentation of sound evidence that backs their claim.

    The SDC portfolio in the South Caucasus is used as the basis of the argument presented in the blog, and makes for a useful case for the counterargument based on my experience. Historically, projects in the region (not just SDC ones) used an approach that the author of the blog would on the one hand seem to advocate: a direct focus on delivering services and products to poor women and men such as micro-finance, agricultural training, seeds, tractors, milking equipment and artificial insemination; then on the other hand reject when they assert that approaches such as the Sustainable Livelihoods Approach failed to address structural and institutional constraints to pro-poor growth.

    Traditionally development projects in the South Caucasus often achieved their results through a process of direct farmer group formation (either cooperatives or “non-profit legal entities”). In assisting the formation of groups the implementer was at best assisting farmers in achieving economies of scale, and at worst improving their own economies of scale by organizing farmers so they could deliver project activities and social agendas more efficiently. Commonly though, these projects were disparate in nature, suffering from a lack of strategy, scale and sustainability. Thousands of farmers sat through trainings and attended demonstrations yet productivity remains resolutely low: millions was spent on tractors, milk processing equipment, farming journals and market price systems but very little remains in commission.

    The reasons for this litany of near-universal failure are many and varied. Firstly the implementer was often attempting to meet multiple and mixed objectives: gender equity, good governance, food security, environmental as well as livelihood and income objectives. In their endeavour to meet these objectives, these groups had no focus, did not have clear ownership, and invariably lacked any coherent economic incentive to sustain themselves. The numerous unused and by now unserviceable milk collection centres, cheese factories, vegetable washing stations, tractors and implements across the region bears grotesque witness to the missed opportunity for rural women and men to improve their incomes these failures represent.

    SDC’s adoption of the M4P approach was a brave acknowledgement of these failures and a commitment to do better. Firstly, it recognized that poverty is about not having money and prioritised economic development objectives in reducing income poverty. Secondly, it moved beyond vacuous rhetoric to invest in analysing and tackling the systemic constraints that limit the inclusion of poor women and men in the market. And thirdly it took the ambition of sustainability and scale seriously by working with those market actors (private, public & civil society) with both the capacity and incentives to improve their legitimate roles in creating economic opportunities for the poor.

    One of the examples cited in the blog as being intrinsically anti-poor is the purchase of lorries for dairy processing businesses and bares further examination as an example of M4P at work. A milk collection centre that one implementer worked with had been given their equipment by an INGO and was essentially moribund due to a lack of strategic thinking that would have generated ownership of the women’s group that had been established to run it. Once the project exited, the group was left with over $10,000 in assets, no market for the milk and collapsed. The one active member of the group ended up managing the centre and found a market for milk without any support but was limited in how far farmers –mainly women – would walk with their milk. In the absence of a formal source of capital, the project intervened to co-invest 65% of the cost of a truck to allow collection from three nearby villages. This created market access for an additional 800 small farmers of which over 190 were women, creating additional income of $33,000 over two seasons over and above what they would have earned from more laborious home cheese production. These benefits have now been sustained for three years. Nearby, there are at least five similar milk collection centres that still lie idle.

    This is not to say that M4P projects should not and do not help farmers to form groups, or never work directly with the poor: nothing is ‘off the table’ as a tactic. Rather, that in doing so, projects should have a clear justification and strategic aim to produce long-term inclusive market access for the poor at scale. For farmer groups this means understanding and addressing the underlying reasons that prevent farmers coalescing (e.g. lack of trust, tax disincentives, capacity gaps, weak market incentives) rather than organising farmers to make our ‘direct delivery’ lives easier.

    To suggest that M4P is a “trickle-down” approach is erroneous: M4P is explicitly pro-poor, focussing on achieving increased incomes for poor women and men. If the means of doing this, which often involves working with non-poor market actors, does not result in direct, tangible and verifiable impact on the poor, then this is a failed M4P project; and there are examples of these. To ensure that benefits are realised by the poor, M4P projects implement a robust impact measurement system based on an impact logic that must explicitly show how working with non-poor actors improves the incomes of the poor.

    The blog also states that M4P projects are hesitant to work with the poor which is certainly not true. What is true is that many interventions focus on non poor actors in the market but with an absolutely explicit and clear strategy of impacting on the poor.

    Proponents of M4P make no apology for working with sectors that include non-poor, but a fundamental criterion for an M4P project is that it works in areas that have a substantial proportion of poor (or potential for the poor) to participate directly. What is unacceptable for an M4P project is for the non-poor to be capturing the majority of the benefits: such a project would be a failure in my view. This is all explicit within the M4P approach.

    Another issue the blog raises is the use of economic indicators to measure M4P project success. This is curious because their overriding objective is to reduce income poverty: much as a health project measures health indicators to measure its success. Surely for any development project to be deemed successful it must create tangible benefits over and above its cost? SDC’s projects in the South Caucasus are going much further than firm-level economic returns (although this is important for sustainability) and measuring social return on investment, accounting for income gains for the poor generated through project partners as well as the benefits of time saved by poor women and men (e.g. women home cheese makers).
    In Georgia, robust results measurement audited to international standards from two of SDC’s programs implemented by Mercy Corps has resulted in over 2,100 women and 17,637 men increasing their incomes by more than $693,000 (with one project only a year into implementation). Moreover the programs were used as examples of good practice in gender mainstreaming and served as the basis for the development of guidelines for the wider community M4P community (DFID, 2012. M4P & Women’s Economic Empowerment: Guidelines for incorporating WEE into M4P programs).

    What the blog author is advocating is a return to nebulous interventions delivering project-defined direct benefits to a project-selected group. This is little more than inefficient, wasteful charity rather than development and it would be far preferable to find an efficient means of directly transferring money to the poor rather than establishing large inefficient institutions to administer it. Unfortunately, whilst charity may make us feel good, it will not tackle the vast problems that befall poor women and men today: adapting to climate change, peak oil, and population increase to name but three.
    Where I perhaps differ most from the author, is on the issue of sustainability. This is perhaps partly to do with the fact that I would differentiate humanitarian aid (keeping people alive) and welfare support (direct benefits delivered to the poor) from sustainable pro-poor livelihood development (increasing the income of the poor). The effectiveness of development needs to be judged more on what it leaves behind after it ceases than on what it does during the life of its intervention. However, on a positive note, I do have some common ground with the author in agreeing that projects around such issues as labour rights, producer associations and the care economy could be greatly augmented by an integration of the core principals of M4P and in doing so would lead to a much higher likelihood of delivering tangible benefits to poor women and men engaged in agriculture.

  • 3
    Julian H Walker wrote on 25 September 2012:

    Thank you for some interesting, and challenging, responses to this post from the Springfield Centre. I have a number of thoughts in relation to your posts, which I outline below.

    A point which you all raise relates to the focus of the M4P approach on sustainable, structural interventions, as opposed to charitable hand-out type projects with limited long term impact. I absolutely agree with this, and the blog post was not intended to suggest that sustainability and scale do not matter – indeed one of the important contributions that the M4P approach makes is this focus on sustainability. However, the concern that I express in the blog is that while the M4P approach advocates flexibility on sustainability principles, allowing for ‘transformative’ direct transfers where this could be seen as catalytic, to contribute to longer term structural change of markets, this flexibility in practice appears to be applied more in favour of businesses and service providers, than in favour of poor women and men in farm households. I would also argue that some trade offs between sustainability and direct intervention need to be made in practice if interventions are to be solidly pro-poor, as the most easy routes to sustainable impacts may not always be the most desirable (ie presumably the types of intervention that are most likely to be sustained after the withdrawal of project support would be those that serve the interests of market players in positions of power, and with greater control over market mechanisms, who are unlikely to be the poorest).

    Another important point that you raise is whether or not, in their poverty reduction efforts, M4P interventions should primarily focus on access to income for the poor. Tim Stewart notes that “Another issue the blog raises is the use of economic indicators to measure M4P project success. This is curious because their overriding objective is to reduce income poverty: much as a health project measures health indicators to measure its success.”

    As you note, income is an important measure of poverty. However an exclusive focus on income as a measure of success in poverty alleviation can be misleading. Many organisations working on poverty advocate an understanding of poverty which relates to disempowerment and social exclusion – ie the underlying factors which would mean that poor women and men are unable to command entitlements to income, services or decent employment. In fact SDC’s definition of poverty is rather broader than inadequate income, seeing poverty as discrimination, obstacles and exclusion in, for example, “participating in the formulation and decision-making stages of the social, political and economic transformation processes”. In this light short term gains in access to income may need to be set against longer term control over systems of entitlement that determine access to resources and public services – see for example the discussions on access to decent employment in the UNRISD 2010 report on Combating Poverty and Inequality, or the discussions on the trade offs between women’s employment, care work and time poverty in the 2012 World Development Report on Gender Equality and Development. I would argue that the challenge therefore is to explore the linkages between changes in access to income and changes in other parameters of poverty relating to well-being and social inclusion.

    A couple of other points, specifically in response to the post from David Elliot – I would like to point out that many of the arguments that you are positioning yourself against are not arguments that were made in the original blog post. For example, when you address ‘the question posed by the author’ of whether the M4P approach is in fact a pro- rich ‘conspiracy’. The blog post does not suggest that the M4P approach is a conspiracy supported by the SDC/ implementing partners and the Springfield Centre to bolster the profits of the rich and marginalise the poor. Perhaps this should have been expressed more clearly, but the concern discussed in the blog is that, despite the fact that all of these organisations are engaging in good faith in efforts to address rural poverty, project interventions not infrequently sideline the inclusion and interests of the poor. None of the suggested possible reasons for this, briefly touched on in the final paragraph of the post, include the idea that this is a deliberate ‘conspiracy’. In fact since I wrote this blog, back in February, many of the projects in the SDC South Caucasus portfolio have gone on to do very interesting work which attempts to address some of the power relations and market barriers which exclude the rural poor. The blog however is trying to explore why this, and other frameworks which attempt to tackle poverty, in many instances appear to end up focusing on market development more generally without explicitly examining how the benefits of this growth will be distributed in a way which reaches the poor (which is not ever an easy task).

    You also challenge the question “Does working for the poor mean only working with the poor?”. Again this is not a suggestion that was made by the original blog posting. Clearly if we understand poverty as a relationship of exclusion (see for example the work on relational understandings of poverty by researchers such as David Mosse or Frances Cleaver) then confronting poverty must involve working with actors on both sides of this equation. What the post does argue, therefore, is that it is not possible to work on poverty issues without explicitly addressing the ‘inclusion and interests of the poor’. As you rightly point out, the interests of scale and sustainability might mean that the best way to do this is not to work directly with poor women and men in implementing project activities (using a welfare/ charity approach that has been widely critiqued), but surely it must mean working with the poor in research to understand the obstacles that they face in accessing markets and employment, and also in working with organisations (often from civil society and/ or the public sector) which explicitly advocate for the interests of the poor.

Leave a Reply