Smart Subsidies for Ag Knowledge and Info Services, Online Forum Synthesis, FINAL Mar2011

1

MaFI—The Market Facilitation Initiative

Synthesis of the Online Discussion Smart Subsidies in Market Facilitation

Specifically for Agricultural Knowledge and Information Services (AKIS)

Facilitated by Andy Jeans, Ben Fowler and Lucho Osorio

Participants:

Abdur Rob, Practical Action (Bangladesh) Alex Mugova, Practical Action (Kenya)

Andy Jeans, APT Enterprise Development (UK) Alison Griffith, Practical Action (UK)

Ben Fowler, Consultant (Canada) Christian Pennotti, CARE (USA)

Ekanath Khatiwada, SNV (Zambia) James Tj, National Innovation Foundation (India) Jayantha Gunasekera, Practical Action (Sri Lanka)

Kamran Niazi, Consultant (Pakistan) Linda Jones, Aga Khan Foundation (Switzerland)

Luis E. (Lucho) Osorio-Cortes, Practical Action (UK) Marcus Jenal, Intercooperation (Bangladesh)

Mary Cockram, Consultant (USA) Rajiv Pradhan, IDE (Bangladesh)

Shawn Cunningham, Consultant (South Africa) Terence Isert, Consultant (USA)

Yogesh Kumar Dwivedi, Action for Social Advancement (ASA) (India)

March 2011

Editors: Gail E. Carter and Lucho Osorio

with the support of Practical Action

2

TABLE OF CONTENTS Introduction – 4

Defining Subsidies – 5

Are They Smart Subsidies, Interventions or Transactions? – 5

Moving the Discussion Forward – 6

The Issues: 1) Transparency – 8 2) Decreasing / Ending Subsidies – 9 3) Expectations and Preconceptions – 11 4) Public and Private Goods – 13 5) Coordinating Market Players and (or) Customizing Subsidies – 19 6) Financial Services – 22 7) Markets, Incentives and Investments – 23 8) Gender – 24

ANNEX A1

Key Principles for Subsidy Use – 26 Tips for Implementing Subsidies – 26 Justifiable Subsidies – 27 Difficult to Justify Subsidies – 28 Some Considerations When Subsidizing AKI Services – 28 Central Questions and Challenges – 28

1 Some ideas in this annex were inspired or adapted from the work done by the FIELD Facilitation Working Group: http://www.microlinks.org/ev_en.php?ID=26628_201&ID2=DO_TOPIC

3

₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪

Dear Readers: We hope you find this paper useful in your work. The discussions were rich and informative, and raised interesting questions for practitioners and policy-makers. We would like to hear about areas where you, your colleagues or partners experience knowledge gaps in applying subsidies in your work. A particularly important question raised during the discussions is: Can we use our combined experience to come up with factors that could convince farmers to pay for knowledge and information services? While we may know the principles, we sometimes struggle with how to apply them in practice. What areas do you think require more investigation? For example, do you find the principles presented in this paper to be universally applicable or does their usefulness and efficacy vary with context or target population? If the latter, where and with who? And, if there are other areas of the synthesis that you agree or disagree with, we hope you share your thoughts with us.

Contact Information:

Luis E (Lucho) Osorio-Cortes, MaFI Facilitator: [email protected]

To know more about MaFI: http://bit.ly/d9KsB8

₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪ ₪

4

Introduction

The Market Facilitation Initiative (MaFI) is a working group of the SEEP Network and the Livelihoods Network, with support from Practical Action. MaFI promotes learning and peer support amongst practitioners facilitating pro-poor market development programs that aim to achieve sustainability and impact at scale. It also assists practitioners move from design to implementation by advancing principles, techniques and tools. In 2008, MaFI created a learning agenda focused on Horizontal and Vertical Linkages, Capacity-Building and Subsidies. The central points of the Subsidies Learning Theme are the discussion, sharing and distilling of experiences; the advancement of practical principles, techniques and tools; and whether, when and how subsidies can be used to catalyze structural changes in market systems that improve the livelihoods of poor people. Given that subsidies is such a broad topic, MaFI members agreed to focus first on the use of subsidies to promote sustainable agricultural knowledge and information services (AKIS2) and their relevance to pro-poor market development in the agriculture sector. Traditionally, assistance to this sector has depended heavily on subsidies and finding “smarter” ways of using them to maximize the sustainability and scale of impacts is not only important but urgent. AKIS includes the implicit and explicit knowledge and information on the skills, techniques, technical assistance and market prices; the location and quality of suppliers; and the warning signals and prevention of pest infestations, diseases, etc.—that strengthen agricultural enterprises and the livelihoods of those who depend on them. On October 15th, 2009, MaFI circulated a discussion document3 to spark a learning conversation among its members. That paper presented a working definition of subsidies and several key principles of subsidy use, implementation tips, central questions and challenges and issues specific to AKIS. The discussions were held over a three-day period from March 23–25 20104. This document is a synthesis of those discussions. The goal of the document is to serve as a reference point for future discussions and to map out the current knowledge and the new questions, gaps and tensions related to the use of subsidies to promote sustainable access to AKIS. The document also includes case studies to illustrate how MaFI members and other development agents and researchers are applying subsidies in the field. We foresee the paper contributing to a more effective use of subsidies in pro-poor market development.

2 The initial idea was to use the term “agricultural extension” but some members thought it excluded other important types of knowledge and information services usually not supplied by extension agents. 3 See Annex A or go to http://www.slideshare.net/marketfacil/mafi-subsidies-discussion-paper 4 In reality the conversations took place during a longer period of time (March-April), but the online forum was convened during the three days mentioned.

5

Defining Subsidies

There are those who would argue that all development interventions are subsidies and effecting change in a market system has the potential to distort it. If this is the case, why are distortion and subsidy considered negative, but change and smart subsidies positive? Given the diversity of terms that practitioners use—incentives, buying down risk, investments, smart subsidies, risk capital, grants and start-up capital—defining a subsidy can be difficult. Subsidies are a way to reduce market actors’ costs and risk through cost-sharing with a development agency to achieve a specific development objective. Practitioners (and donors) consider subsidies to be investments intended to bring about a desired result—systemic change that permits their eventual discontinuation, increased food security or a developmental impact such as leveling the playing field or reducing exclusion, inequality and discrimination. Whatever the outcome, subsidies should not undermine the self-reliance of households or market systems. Boundaries define subsidies—some say that from an economic perspective subsidies do not exist. Others believe that subsidies exist, but their existence depends on the boundaries we use to define the social system we are trying to change (in our case, the markets). Any resource originating outside of the target system is a subsidy. If the system is the global economy, subsidies cease to exist. In a nutshell, subsidies are real but depend on constructed boundaries. If this is accepted, then the trick is in how we draw such boundaries. Some people understand public subsidies as a bad thing. Others, especially when it comes to crisis situations and extremely vulnerable people, accept that public subsidies are good and even necessary. Using the concept idea of system boundaries, two considerations come to mind: Do the boundaries include all the actors (public and private) who make the creation of wealth

sustainable in the long run? When we define the boundaries of a particular market system do we not leave out key players?

How feasible is it for actors within those boundaries (especially marginalized producers) to influence the way resources inside those boundaries are used?

Are They Smart Subsidies or Interventions?

Jayantha Gunasekera [Practical Action] wrote: “I am trying to understand what are subsidies, what are smart subsidies and what are not subsidies, but market development interventions.” He shared a case study that involved either subsidies or interventions in the dairy sector in Sri Lanka and said he wanted to learn if it met smart subsidy criteria. In brief, the project: Identified the need for improved breeds, feed and cattle management for better returns for

farmers Provided cattle management training for selected lead farmers, including training of trainers Introduced a new variety of grass for improved feed Organized and funded a one-off agricultural information clinic, and Facilitated interaction between the farmers, veterinarians and input providers.

If all development interventions are subsidies and effecting change in a market system has the potential to distort it, why are distortion and subsidy considered negative, but change and smart subsidies positive?

6

Andy Jeans [APT] responded to Jayantha’s request, noting that it was unclear how much subsidy was used and that the case was an intervention by an NGO as it appeared that the NGO met all of the costs. If that is so, then it was not really a subsidy. A subsidy would involve the intervener contributing part of the cost and one or more market actors covering the balance. A smart subsidy would lead to a systemic change that obviated the need for further subsidies—for example, in this case study, a subsidy that led to cattle management training, the provision of improved feed by local input suppliers and artifical insemination services becoming more widely available in the market with services purchased by farmers like those assisted. On the other hand, Linda Jones [Aga Khan Foundation] said that she appreciated the implication of Jayantha's question and, in a sense we can say that all interventions are subsidies and have the potential to distort the market system in which we are trying to intervene. In fact, when we try to effect change in a market system, are we not trying to distort it? So distortion and subsidy are terms we use negatively, but change and smart subsidies are meant to be the positive. To her, it all comes back to the context and outcome and she agreed with Andy's statement that subsidies should not undermine self-reliance of households or market systems. Rajiv Pradhan [IDE] proposed that it is useful to differentiate between project interventions and the transactions performed by market actors. Using Jayantha’s case as an example, Rajiv mentioned that training paravets so that they can demonstrate the benefits of their services is a legitimate project intervention that can be subsidized. The question then is who trains the paravets without subsidies in the long term? If the subsidy leads to sustainable transactions amongst market actors, then we can talk about smart subsidy. Moving the Discussion on Subsidies Forward

Discussions over the three-day period were facilitated by Lucho Osorio, Ben Fowler and Andy Jeans and focused on 1) Using Subsidies in Dysfunctional, but Stable Markets; 2) Using Subsidies in Dysfunctional, Unstable and Crisis-affected Markets; and 3) Lessons, Challenges, Policy Implications and a Future Research Agenda. Day One began by looking at the use of subsidies for AKI services in markets that, despite being relatively stable, are weak, thin or do not work well for the poorest. In addition, concepts and issues surrounding subsidies were introduced. Day Two focused on particularly vulnerable / disadvantaged producers, e.g., individuals / communities suffering from poor health, discrimination, conflict or natural disaster. Some may say that relief or welfare services (such as food aid, food for work, or distribution of seeds and tools) are the only practical response. But can the market for AKI services respond to their needs as well or better? And if so, in what ways can or should we use subsidies to bring this about? Examples from East Africa show the possibilities:

“…in a sense we can say that all interventions are subsidies and have the potential to distort the market system in which we are trying to intervene. In fact, when we try to effect change in a market system, are we not trying to distort it?” Linda Jones, Aga Khan Foundation

7

In the work of the National Union of Disabled Persons in Uganda (NUDIPU) with disabled people in conflict-affected Northern Uganda, partial subsidies are used to stimulate the hiring of disabled men and women as apprentices in ongoing small enterprises. The subsidy enables owners to adapt facilities and many apprentices go on to either start their own enterprise or to find employment elsewhere.

APT’s Rural Education and Economic Enhancement Program (REEP) in Western Kenya provides AKI subsidies (applied indirectly through community volunteers) to farmers who have HIV/AIDS. The subsidies enable them to establish horticultural production and marketing that leads to food security and social inclusion.

On Day Three we looked at areas that received less attention during the first two days: • Public and private goods: Are there agriculture knowledge and information services that can become sustainable only through permanent, public subsidies? Many practitioners have seen situations where the private provision of AKIS is nonexistent and public provision is extremely weak or absent—does a solution always exist in these areas? • Transparency: Should we be 100% transparent with all stakeholders about our intentions to use subsidies? • Power: Having resources to invest and deciding who to give them to makes us powerful compared to other market actors. How should we deal with this? • Expectations and preconceptions: How can we manage expectations strategically and minimize the damage preconceptions may cause? • Coordination: How can we coordinate our use of subsidies with other development agents? If we want extremely marginalized and vulnerable producers to engage with other public and private market actors, we need to expend extra effort during the capacity-building phase before including them in workshops with other stakeholders, negotiation rounds, drafting of joint action plans, etc. However, a lesson from APT shows that the opposite is also true—if the project is well facilitated, market engagement can actually build capacity to engage with other actors. And, it can also re-build self-esteem and break down cultural and social barriers, which is extremely important when talking about HIV/AIDS-affected populations and victims of conflict, disaster, etc. Within those broad parameters, the conversation ranged widely and raised a number of important and interesting questions. For instance: Can we predict market distortions and/or the scale of impact when we use subsidies? Can a good business plan tell us when to withdraw subsidies? Are there other indicators

we should use? What incentives do subsidy recipients have to tell us they no longer need subsidies? Kamran Niazi [Consultant] brought up the issue of who deserves assistance and how we think about those we provide with subsidized aid. Do we consider everyone so deserving that we want to give them everything? He considers it preferable to think of those we help with subsidies as potential market players and productive members of the economy who just need a helping hand to move further up the value chain. He provided a link to the Entrepreneur Bootcamp for Veterans

8

with Disabilities (EBV) 5, a training program at several U.S. universities for American war veterans, a group that, according to Kamran, certainly falls under the poor health / conflict-affected criteria that we saw in the NUDIPA example. He noted that the selection criteria for inclusion in EBV are very tough. Andy Jeans [APT] noted that in this forum we are considering the merits and value of subsidies to bring about systemic change in market processes that enable everyone, whatever their disadvantages, to work themselves out of poverty. In that context, no one is more deserving than any other and the question must be whether the targeted use of subsidies can enable AKI services to be accessed and made use of by even the most disadvantaged. Participants’ experiences in working with and familiarizing themselves with numerous projects enriched the conversation around these and other questions and helped refine our understanding of smart subsidies by identifying issues and suggesting different ways to use them. The Issues

In synthesizing the discussions, we organized the paper around the major themes that emerged during the discussions:

1) Transparency 2) Decreasing / ending subsidies 3) Public and private goods 4) Expectations and preconceptions 5) Coordinating Market Players and (or) Customizing Subsidies 6) Markets, Incentives and Investments 7) Power and relationships

We understand that there is necessarily a good deal of overlap between these themes and we encourage readers to comment on the content and its organization. 1) Transparency

Transparency is a multi-faceted issue. In some cases the question is, should development practitioners be 100% transparent with all stakeholders about their intentions to use subsidies? It depends... where loan guarantees (a financial subsidy) are concerned, transparency can lead to increased defaults—the loan is guaranteed and if borrowers do not repay, the guarantor will. Alex Mugova has had firsthand experience with more than ten loan guarantee programs that performed dismally once borrowers learned that the loans were guaranteed. It was clear to him that, “…in the case of loan subsidies, the overwhelming evidence from the field is that transparency will undo all the good intentions.”

5 http://www.whitman.syr.edu/ebv/

“For me, the main reason to be transparent about subsidies is that it will help manage expectations and it is easier to set up a process that is clear on why there is a subsidy, how long it will last, what the hoped for outcome is, etc.” Linda Jones, Aga Khan Foundation

9

There may not be a comparable market development example where transparency can help programs manage expectations by setting up a process that is clear on why there is a subsidy, how long it will last, what the expected outcome is, etc. In cases where subsidies consist of embedded inputs or services, transparency is paramount. No producer who receives a service such as training or advice wants to find out later that he is, in fact, paying for it when he receives less than expected for his crop or product. A lack of transparency can seriously undermine the confidence and mutual trust that are critically important to stimulating the emergence of inclusive and pro-poor markets. In Kenya, the case of a Community-Based Animal Health Workers (CBAHW) association that

works with the District Veterinary Office (DVO) on disease outbreaks and vaccination campaigns raises issues of transparency, participation in decision-making and empowering marginalized actors to decide how subsidies are used. The CBAHW was unable to get any information on the amount of donated inputs and supplies intended for the communities from the DVO, which received the supplies and distributed them. Not only could the district CBAHWs not find out anything about the supplies, they were denied a role in the decision-making and resource allocation processes. In this example, a public entity withheld valuable information from a vulnerable group by not informing members about the supplies it received (how much and for who) and not allowing them to help decide how to use the goods.

2) Decreasing / Ending Subsidies

The discussion paper states that capacity-building is a legitimate use of subsidies, but it is more difficult to justify physical assets and recurring operational costs (e.g., personnel, equipment, etc.). While some may argue that subsidies for machinery are a low-risk way of introducing a group of market actors to a new technology or technique, how do they distort the market for incumbent producers and what is the scale of their impact? Can we predict market distortions or impact? Can a business plan tell us when to withdraw subsidies? And, are there incentives that could convince recipients to voluntarily stop accepting subsidies? Gradually reducing subsidies may sound easy, but it can be very complex. Rajiv Pradhan finds that if development agencies understand business plans they can provide subsidies on a declining basis. This type of subsidy must be justified and based on: performance, the ability to tackle key problem(s) and an expanding market. However, will this work in rural areas? Will small farmer groups be able to identify others who can provide AKI services as subsidies decline? Another question is: Is it a good idea to link subsidy reductions to individual business

“…in the case of loan subsidies, the overwhelming evidence from the field is that transparency will undo all the good intentions.” Alex Mugova, Practical Action

Gradually reducing subsidies may sound easy, but it can be a very complex process. The subsidy must be justified and based on: performance, an ability to tackle key problems and an expanding market. Is it a good idea to link subsidy reductions to business plans? And, will it work in rural areas? Rajiv Pradhan, IDE

“…we organized ourselves into a microenterprise and got training from the national and local governments and NGOs. We now exhibit in shows and fairs and meet at the end of each month to coordinate our work. We are learning to produce other types of milk-based products.”

A Kamayok association member

10

plans and to avoid standard, one-size-fits-all subsidy packages? In Peru, 11 Kamayoq (community-based agriculture service providers) trained by Practical

Action established a producer association that received funds (a subsidy) from the national government to set up a yoghurt production plant. An association member says: “we organized ourselves into a microenterprise and got training from the national and local governments and NGOs. We now exhibit in shows and fairs and meet at the end of each month to coordinate our work. We are learning to produce other types of milk-based products.”

This story demonstrates that once-marginalized producers can learn how to access and combine funds and training services from government and NGOs and use them to build their businesses. Though the Kamayoq are now doing well, they continue using NGO subsidies and this raises the question of when subsidies for AKI services should be stopped. The subsidies were for training the Kamayok as AKI service providers and to purchase a yoghurt processing unit for them to become producers through their association. The first—promoting AKI service provision—is smart because it enables the poor to access information they need to improve production. However, the rationale for subsidizing a production plant is less clear because a) it may distort the market for competitors and b) resources needed to subsidize all producers would likely be beyond the scope of the project. Should the withdrawal of a subsidy for the Kamayok as service providers be based on the business plan they have for operating those services? In another example, this on Nepal’s essential oil industry from Rajiv Pradhan [IDE], the lack of

processing units near community forest users group (CFUG) member farms prevented the groups from producing essential oil products. The project began by first helping farmers improve their production and marketing skills and then facilitated their development of a business plan. Due to the ongoing conflict, businesses were reluctant to invest in processing machines so the project subsidized 50% of the cost of a unit to hasten the ROI for the groups. The bank financed the remaining 50%, without collateral, on a hire-purchase basis and the groups made regular installment payments. The lesson IDE learned was that a business plan was essential both to help the CFUGs understand the ROI and to convince the bank to finance the scheme. IDE found that the subsidy helped buyers access the product and did not distort the market. The upfront work the project did with the groups to improve production and marketing also gave them confidence in farmers’ ability to stick to the business plan.

When asked how farmers learned to maintain the machine and whether they could pay to learn and were able to make their own repairs, Rajiv said group members received on-the-job training and were involved in setting up the machine so they could operate and maintain it. (Other projects gave units, but did not provide training and groups had to return to them for maintenance and other assistance.) And, the community can pay / contribute when training or inputs benefit everyone. Another question concerned the bank and whether it had replicated the scheme on its own initiative? Rajiv’s response was that, yes, the bank had established a microenterprise unit and was considering supporting others.

11

Ekanath Khatiwada [SNV] said that introducing new technologies / innovations at the community level can be difficult, even if there is market demand for new (value-added) products. A demonstration of the tangible benefits that a new processing unit would provide may change attitudes and encourage acceptance. The Nepal cardamom value chain case shows how improved dryers introduced and commercialized locally benefited the community. (practicalaction.org/shelter//text/SED_19-1) In an example from Andy Jeans [APT], SITE, a Kenyan NGO, promoted milk quality services

to people already working in the warm milk chain—those who collect milk from farmers (such as coops), transport it (often by bicycle) and retail and/or process it at milk bars. SITE subsidized the initial training of these service providers whose eligibility for the training was based on their business plans and previous performance. Once they began delivering the milk quality services and clients paid their fees, SITE considered further subsidies to be unnecessary. A key factor in allowing discontinuation of subsidies was the role that SITE, in cooperation with Kenya Dairy Board, played in stimulating demand for the services by publicizing the importance of milk hygiene and quality and facilitating the accreditation of both the milk sector market players and the service providers.

Lucho brought up the importance of subsidizing the right actors, those who can provide AKI services to marginalized producers and buyers. Sometimes, when we aim for sustainability we need to subsidize non- or not-so-poor actors. This experience also stresses two essential selection criteria: a track record as a service provider and a business plan (which can be improved with the help of facilitators). Do you agree with these criteria? Are there others? 3) Expectations and Preconceptions

Ekanath Khatiwada contributed a case that highlighted the difficulty of creating market-based AKI services due to the long-standing tradition—and expectations—of government subsidies for agriculture. Much depends on changing attitudes. Very poor people cannot afford to buy information or inputs and a subsidy for training, supplies or machinery can enable them to participate in economic activities. A declining subsidy would allow them to begin sharing the cost of services and pay a larger share over time. DANIDA /MRMG supported Rural Income Generation Programme provided funds to the vegetable value chain in Nepal for awareness-raising, training, inputs and technology, demonstration plots and group formation. During the first year, these services helped the community to produce vegetables for selling in local markets. In year two, MRMG/RIGP reduced the inputs subsidy by 25% and in year three they cut it another 25%. In the final year, the project stopped all subsidies and facilitated linkages with a private agricultural center, which has worked out for everyone.

Lucho asked if he had seen structural transformations in the vegetable case that would help marginalized producers access sustainable AKI services? Did the farmers trained by the project go on to build their own links with fee-based or government AKI service providers?

12

Kamran Niazi worked with a project in India that supported grassroots NGOs in their efforts to help very small farmers form into groups that they provided with funds to set up a business venture. The groups were required to match the funds this modified venture philanthropic activity provided, but they often contributed more. They had good ideas for businesses that were market-led and quite promising. They would not make a lot of money, but the project subsidized the start-ups, providing from $700-$1200 for each group ($70-$120 per member) and they were happy with the returns they received (20 percent or more). Although there were some failures, there were more successes. In the end, the project introduced new technologies and inputs and made the markets work for the poor.

Kamran also said the discussion seemed to be focusing too much on paperwork at the expense of the idea or the people and wondered if we do not expect too much of those living at the grassroots level? Assuming that very poor and uneducated farmers who cannot read Urdu can put together a business plan or financial information in English is impractical and unrealistic. Even the NGO staff could not produce a well-written plan and the project had to train them. Lucho asked Kamran what he looks for when selecting marginalized producers to receive subsidies. Kamran cited former MIT entrepreneurship professor, Ken Morse, who said “I will fund an A Team with a B Idea, rather than funding a B Team with an A Idea” and “I do not look at the plan, I look at the people”. He then said that for him the critical selection factors are: An understanding that money is a privilege, not a right. A business approach or very clear idea about what they want to do and why. The ability to convince others and to know what they are talking about. He considers the ability

to write a plan in English to be far less important. An idea they conceived and want to develop; it should not something imposed on them from

outside, though they can adopt a good practice as long as it is adapted to the local context and conditions.

An appropriate concept, e.g., a processing unit that can take them from zero to 15, if that is what they can afford, rather than to 100. The unit should be built locally from local materials so others can copy the concept and afford to pay the full price. And, it should be energy efficient since most poor people rarely have access or the ability to buy expensive fuels.

Subsidy approval should take no more that 2-3 steps with the final decisions made by totally independent decision-makers. Lucho noted that Kamran’s comments brought the importance of flexibility, experience and intuition to the forum—and to donors who have strong incentives to push practitioners toward forms and checklists. He then referred to Yogesh Kumar Dwivedi’s [ASA] comments about farmers in very weak markets who have a small surplus, but are unaware of market and quality issues. There is a need for projects to work with government to create awareness, train farmers, form groups, provide agribusiness support, set up systems and procedures, build infrastructure such as warehouses, etc. Promoting low-input agriculture and low-cost, ICT-based technologies for

Subsidies should be used to strengthen the performance of the market and not individual firms, and government should provide public goods that include knowledge and information services. Shawn Cunningham, Consultant

13

higher productivity can help ensure increased net returns from agriculture and provide producers a fair price—something that rarely happens in India. Lucho said that the idea of collaborating with the government may be the key to success in such extremely marginal contexts. 4) Public and Private Goods

Shawn Cunningham [Consultant] endorses the view that subsidies should be used to strengthen the performance of the market, not of individual firms. Many believe that for markets to work, governments must get out of the way, but others think government should focus on providing public goods, including knowledge and information services that are prone to market failure because they are so difficult to value and compare. Shawn says that we should direct subsidies at creating public goods, such as education, technology demonstrations and the provision of information. Another market intervention would be to support the creation of positive externalities. For instance, when a farmer trains his staff, they may leave to work for another farmer who offers more money. This typically happens in areas where there is under-investment in education. Because farmers and businesses often under-invest in services such as training, subsidies for them can be very important. However, using subsidies to provide public goods, address externalities or overcome indivisibilities make sense only in environments where markets are supported, i.e., enabled and not prone to monopolies (by the state or other actors). You need some basic market infrastructure and institutions. In other words, subsidies should be directed at the causes of market failure and not at the symptoms. We must see the market, even if it appears to be simple, as a complex system. The agents that can affect the performance and configuration of this system may be government organizations or industry bodies. While we should not rule out the possibility that in some instances the cause might be addressed by providing a subsidy to a firm, we should be very skeptical about this. The topic of market failures is fascinating, but it is not often understood within the context of a complex adaptive market system.

Andy Jeans commented that, while many governments strive to deliver some public goods—such as health care, education and agricultural and veterinary extension, they find the cost difficult to meet. Shawn’s observation that subsidies should be directed at causes of market failure and not at symptoms sounds very sensible but what exactly are the causes? Some may argue that the factors that cause AKI services to not be available to all who could benefit from them may include insufficient economic viability (service delivery costs more than the economic return), mistrust (between service provider and recipient), or insufficient government will (or ability) to provide public goods except to a very few. There are probably many others. How can subsidies address those causes? In responding to Andy, Shawn said that he had asked similar questions and they had led him to embark on a seven-year study of market failures and its characteristics. Many donors consider market failure a pre-condition for interventions. However, they almost never talk about how to identify and measure market failure. For example, education and health are public goods in more than one sense. When they are not provided this is not called a market failure, but a government failure (a concept familiar to political scientists and philosophers). From an economics and socio-political perspective, market failures directly affect the allocation of resources in the marketplace.

14

And for economists, it is about efficiency of allocation. But, important public goods are more about the basic knowledge needed in an industry, information about quality standards, accreditation systems, etc. In many developing countries, knowledge is not yet valued, often because it is intangible and difficult to measure how such services affect industry performance. Furthermore, measures to support impersonal exchange (through markets or other means) are often lacking in societies where knowledge is not seen as a transactable good. This is why so many BDS interventions fail in these contexts. When people do not value advice (knowledge), they do not want to pay for it. For Alex Mugova, the classic distinction between public and private goods is at the heart of the subsidies debate. Traditionally, most argue that public goods, by their nature, always require subsidies. This is because it is difficult, if not impossible, to exclude those who are unwilling or unable to pay for public goods. On the other hand, those who are unwilling or unable to pay for private goods can easily be excluded and so they do not require any subsidies. This is as far as the theory goes and it is neat. But, when it comes to supporting poor and vulnerable groups in developing countries to move out of poverty, there is a wide divergence between this neat theory and the reality we find on the ground. Agricultural extension services are a public good that, according to the theory, must be subsidized for the good of the public. However, the reality is that most developing country governments do not have the resources to subsidize these services. And, because the poor are usually located in remote rural locations, they do not present a compelling business case for the private sector to provide embedded extension services. The question then is: are such poor and vulnerable people condemned to a life of hopelessness and poverty? He then provided examples of Practical Action’s work in this area. A number of Practical Action projects in remote rural locations demonstrate another way to

address this challenge—involving an NGO or other outside agency in organizing farmers and helping them understand their predicament and collaborate to build their bargaining capacity. The next step is to approach private firms and convince them that there is a compelling business case in working with poor and isolated farmers. In one of Practical Action's projects with small livestock farmers in the Guruve district of northern Zimbabwe, Practical Action convinced a private veterinary drug company that it would be possible to expand the market for its drugs and advisory services in this remote district. The approach involved identifying and training village based paravets. Once they had the requisite skills, they became the local foot soldiers for the company, selling drugs and providing advice and training in basic animal disease diagnosis, treatment of sick animals and improved animal husbandry techniques to farmers for a fee. The results of this intervention are dramatic and demonstrate what can be achieved when a private firm works with poor farmers in remote locations. The number of paravets in the area has risen from 200 to 600—the project trained the first 200 and Agritex, the state extension department, trained an additional 400 once it realized the efficacy of the approach. The government is now replicating the business model in five other districts of the country. The incidence of major livestock diseases such as CBPN, black leg and red water has decreased by nearly 90% in the last four years and livestock off-take rates in the participating districts have

15

increased from 5-15% over a four-year period with farmers’ incomes from livestock sales jumping almost 20% over the same period. This intervention demonstrates that even when government lacks the capacity to provide extension services and the private sector is unwilling to do so, development agency innovations can help break this impasse.

Lucho Osorio said that Alex's experience demonstrates the need to help market actors learn from successful business models in similar contexts and develop new models that can reach out to the most marginalized and vulnerable populations. He also commented on Alex’s observation that agricultural extension services are by nature a public good—which seemed more like an ideology than a solid theory—and he asked “what theory demonstrates this?” The very fact that the private sector is finding successful business models to reach out to marginalized and vulnerable producers with useful knowledge means that there is no inherent public nature in agricultural extension. He then highlighted the point that the business feasibility of knowledge services depends on the type of knowledge and the possibility of creating barriers to entry that can generate profits. If there are no barriers to entry, for example, when one farmer tells another how to protect his potatoes from a parasite—it cannot be the foundation of a business model, even though it can make the recipient’s business more lucrative. But, if there are barriers to entry, for instance, inoculating an animal with a vaccine, which requires investments in training and practice, there is a business case. Many of us have seen practitioners trying to create extension business models on the basis of easily transferable knowledge. But, Practical Action long-term evaluations of extension services suggest that agricultural extension services in the livestock sector, where there is a lot of knowledge with high barriers to entry due to the use of vaccines and the need for specialized skills, are more successful and sustainable than those in crop production where techniques are easily shared amongst farmers. Have others seen similar evidence? Does anyone have evidence that does not corroborate this observation? Marcus Jenal agrees with Lucho on the types of AKI services that can become a viable business model, i.e. those with relatively high entry barriers. Marcus provides an example from Bangladesh, where Katalyst used embedded services to tackle the issue of low barriers to entry. Katalyst supported the dissemination of easily transferable AKI (with low entry barriers) using ICTs. The business model is based on a central knowledge provider whose main task is to provide up-to-date AKI services to its clients. The content is not limited to agricultural knowledge, but also includes health and legal issues. The clients are mainly two large mobile phone companies, Grameenphone and Banglalink. The Grameenphone business model is to set up numerous Community Information Centers to sell a range of value-added services such as internet access, scanning, printing, mobile calling, etc., and farmers can also access AKI services. Banglalink, on the other hand, promotes access to content through a hotline. The farmer, who must be a Banglalink customer, can dial 7676 and ask any question relevant to his work. The call center agent then accesses the database to find a solution to the farmer's problem. While these examples raise some questions, they show that the private sector can provide information with a low entry barrier to farmers in remote areas when it is combined with other products, mobile phone services in this case, that already are a highly successful business.

16

Ekanath agreed with Alex’s experience, saying that it is a successful business model, especially in the livestock sector in rural areas, and SAPAP/UNDP Nepal adopted a similar approach in its program Nepal. Alex Mugova [Practical Action] then provided another example from Zimbabwe, highlighting

that subsidies can help to move poor and marginalized producers into meaningful players in commodity markets. “The challenge, as we all know, has always been how to ensure that subsidies are smart and enhance the attainment of intended objectives.” This example showed how a project used subsidies to help smallholder farmers become important players in a specific market. Practical Action and its partners—Agriseeds (a private company), Agritex and poor smallholder farmers—supported the production and use of guar bean for use in nickel production6. For two years farmers received subsidized seeds and fertilizers as an incentive to participate. After that time, farmers were able to produce and supply close to 30% (4,800 metric tons) of the local mining industry’s requirements, a significant achievement given that they had never before grown the crop. The project facilitated a link between the producers and Agriseeds (the buyer) from the beginning, which put them in a good position to negotiate with the buyer to access subsidies and learn about the penalties and the obligations on the part of the buyer in terms of prices and buying conditions.

Three key lessons emerged from this project. 1) Subsidies help poor and marginalized producers become meaningful actors in commodity

markets by increasing their productivity and output and making them consumers of other commodities.

2) Subsidies alone can do more harm than good. To be effective, they must be provided together with other key services and inputs that small farmers require—extension services and passable roads that enable them to get inputs to the farm and to deliver their products to markets.

3) The private sector can be an effective partner in providing subsidies that positively impact poor, smallholder farmers. Most important is to convince private firms that working with the poor makes a compelling business case.

In agreeing with Alex that private firms must see a compelling business case, Andy Jeans said it means the cost of the subsidy for seeds must be offset by a reduction in the crop price paid to the farmer to make the net subsidy to the farmer zero. This was a smart subsidy because the project did not have to continue facilitating the supply of seed. The transaction now is handled by the market actors—farmers and Agriseeds. Also important is Alex’s second lesson—that subsidies can do more harm than good. If applied inappropriately, subsidies can lead to reduced self-reliance and greater vulnerability of recipients when they are removed. Rajiv Pradhan’s response to the model Alex presented was a question—how can we have a private sector solution to what is normally understood to be a public service function—the 6 The gum extracted from guar beans is used as a lubricant for the perforation drills.

“That subsidies can help to move poor and marginalized producers into meaningful players in commodity markets is not in doubt. The challenge, as we all know, has always been how to ensure that subsidies are smart and enhance the attainment of intended objectives.”

Alex Mugova, Practical Action

17

provision of services to rural populations? Government services to the agriculture sector will exist for the foreseeable future. But, how can we make them more efficient and effective without additional resources and leverage the private sector to provide them? One way is for NGOs to do the initial modeling of how to make the business compelling for the private sector, then bring in government and, finally, scale it up with private sector companies that have an incentive to participate. When one can see the business rationale, e.g., that in the Katalyst ICT example Marcus provided or with contract farming, scaling up is relatively easy, but when large companies do not see the provision of information as a way to increase sales, it can get tricky. Another strategy would be to have farmer groups, local firms and government form into small, very local partnerships. When there is information that farmers must have (e.g., to prevent disease, discontinue using specific chemicals, etc.), these partnerships can get the word out—with government organizing a meeting and providing information and the companies and groups contributing funds and distributing printed materials to the farmers. We saw this happen with artificial insemination (AI) in poultry—government was holding critical information, but did not have resources to print the information or the funds to visit all farmers and so phoned a few large farmers. Their association copied the materials and members took it to their own groups to make farmers aware of AI and how to mitigate it in the industry. Though it is possible to use the media to do this, will the information reach all of the small farmers? The subsidy here would be used to form the partnerships and, if it can demonstrate that business could increase for participating companies—e.g., day-old chick sales would not decrease—then larger companies would be willing to participate in local-level partnerships. In Bangladesh, these companies have field officers who travel to different parts of the country. For now, they may still focus on large farmers, but partnerships could help them to reach out to smaller farmers. In this way the government mandate to serve all of the people is achieved and both the private sector and the farmers also gain. It would be a win-win-win situation as Kamran stated. Yogesh Kumar Dwivedi [Action for Social Advancement – ASA] offered a case from India

where he works with small farmers on both high potential production systems as well as those for resource-poor areas. In his work with 13 villages in one area, he found that farmers getting irrigation from a private lift irrigation system could easily pay the Rs 300/acre for irrigation and, in fact, were always ready to pay for all agriculture-related services (mostly inputs) because they were growing mainly cash crops and paying for services was not an issue for them. However, farmers did not want to pay for knowledge-based services since they felt government should provide them. ASA found local input suppliers to be the main source of agriculture information with only 2-3% of farmers approaching the regional agriculture research station (Krishi Vigyan Kendra), or state agriculture universities for information. In general, farmers who visited these places did so as participants in an NGO or government-funded project. In addition to providing technical support, the project supplied farmers with quality seeds and farm tools on credit—which they repaid in a timely manner. This also occurred in another area where ASA implemented a project with less-poor farmers who requested quality seeds and agriculture inputs and who paid the full costs, even travel expenses, without any problem.

ASA’s experience in resource-poor areas was very different. Farmers practicing agriculture in risk-prone zones do not have assured irrigation facilities and the frequent, severe droughts and

18

uncertainty of final production convince them not to invest in agriculture. They look for government subsidies and external support for extension services, support that Yogesh thinks they deserve until basic facilities such as irrigation, storage and grading infrastructure, information on drought coping strategies and short season, drought-resistant/tolerant varieties are made available. Subsidies could help these farmers deal with changing socio-economic and hydro-geological conditions. In India’s more developed states, Yogesh found that farmers were willing to pay for extension services if they were timely and high quality. Farmers elsewhere pay for inputs such as seeds, fertilizers, farm tools and pesticides but not for AKI services, mostly because the quality of the services is neither credible nor useful. Customizing, branding and making extension services work in the local socio-economic, agro-climatic and geo-hydro context is critically important when charging for them.

Lucho Osorio wondered if the ASA program was able to exit the direct provision of inputs (e.g. seeds and tools) and if farmers were requesting less from now than before? And why? This example also teaches us that farmers are willing to pay in some cases but not in others and that, in places where farmers are more reluctant to pay for knowledge, the accepted rule seems to be that government should subsidize it. The questions this example brings to mind include: are there political incentives for governments to subsidize access to knowledge (e.g., popularity and elections)? If this is unlikely to change, can those subsidies promote the transformation of market systems and their rules? Alison Griffith [Practical Action] spoke about those who argue that certain very fragile eco-systems are unsuitable for agriculture, even in a very extensive form, and subsidizing the provision of inputs and services in such areas artificially supports activities that are ultimately unsustainable. In these cases, they argue, it is preferable to do nothing and allow natural depopulation to a so-called sustainable level. This would be problematic because it would increase the vulnerability of the poorest. Or, we could promote alternatives, which also would be problematic since many have been tried and there are few examples of success, particularly at scale. A third option would be to promote pro-poor market development approaches to improve the efficiency of key subsectors. For example, the case of livestock in northern Kenya shows that, in spite of many alternative

income generation programs, the sector is fundamentally important for local livelihoods. The challenge is to develop an appropriate level of services and inputs and promote efficiency at various levels (e.g., management of natural resources, coordination of actors). Practical Action’s new work in the Mandera area is looking at the different roles public and private actors play in the milk and meat market systems and facilitating new models to promote efficiency of their activities. So, even in areas where viability is questionable, it may still be possible to use models discussed here. Attracting private input suppliers (e.g., veterinary drug suppliers) to such areas and understanding their risks and concerns, can be critical. We have choices when using subsidies in these kinds of situations and the use of Challenge Funds with the private sector is becoming more common. Perhaps the biggest advantage of this approach is that these funds devolve power from donors to actors who have shown a commitment to the future development of the business, as demonstrated by their application to the Challenge Fund. Does anyone have experience with either the Africa Enterprise Challenge Fund or others?

19

5) Coordinating Market Players and (or) Customizing Subsidies

Shawn Cunningham noted that another area of concern is coordination (horizontal linkages) in weaker markets—it is a very expensive and time-consuming endeavor and may justify a subsidy. For instance, coordinating several tea farmers to fill a container is a high-cost activity that could warrant subsidized support. Kamran Niazi worked with a community that had to sell its produce for a very low price in

local markets because they could not sell it as individuals in high-margin but distant metropolitan markets. When the project organized them into a group and brought them to a major market to sell their combined product, measured in tons, they received VIP treatment and sold out their entire inventory. The result was a 3million Rs (US$ 35,295) increase in revenue that year. The project provided a subsidy of 200,000 Rs (2,353 US$) for group organization and to assist them transport their produce to the market.

Targeted support should be extended and customized to the needs of those who really need it. At the same time, projects must be flexible, trust the participants and monitor implementation of subsidies. This is a time consuming (and costly) activity and though results are likely to be very good, the numbers and scale are going to be small. One way to increase scale is volunteers, for example, the Association for the Development of Pakistan, formed by American and Pakistani volunteers, raises money for grants and expenses, identifies potential opportunities, conducts due diligence and evaluates impact. The cost is very low and all projects are in conflict or post conflict areas. In other words, the decision about whether to support many people with a few generic products or only a few with customized products plays a big part in project impact. Andy Jeans commented on Kamran’s example about whether we are obliged to choose between supplying many people with a few generic subsidies or fewer people with customized subsidies and less impact. Some have observed that government-provided, low budget and widespread agricultural extension services are generic subsidies that often have a disappointingly low impact. Sometimes, though, there is room for a smarter choice. Rather than assisting only a fortunate few with customized services—and recognizing that many developing country governments cannot afford to provide effective, generic AKI services—we should strive to use subsidies as a time-bound catalyst to bring about systemic change that can, over time, enable all market actors to access the AKI services they need. In adding to this debate, Christian Pennotti [CARE] said that not having a standard subsidy package for all participants is one of the most important guidelines for subsidy use. CARE recognizes that those receiving subsidies are different in terms of their willingness or ability to take on risk and the sophistication of their enterprises. While taking a one-size-fits-all approach can get money into the hands of those who need it quickly, at a lower cost and with less effort, it may not conform to the underlying rationale for providing a subsidy in the first place. Tailoring subsidy packages based on the needs of individual enterprises, farmers or groups clearly is more expensive and time intensive, but it also can be more effective.

20

A challenge that customized services can present is, when providing different subsidies to people in the same community or value chain, what about transparency? Should each subsidy recipient know what others are getting? What if someone who receives less than another feels discriminated against? How should you handle those tensions if they arise? And, if there is no tension, do you know why? Alex Mugova brought up the critically important issue of how best to coordinate the use of subsidies with other development agents. He has found the coordination of subsidy provision to be a key to whether or not projects can achieve their intended objectives. In the example below, Alex demonstrates how all too often the efforts of one development project to develop pro-poor market systems are undermined by the uncoordinated and indiscriminate use of subsidies by another donor or practitioner. In Sudan, Practical Action assisted pastoralists in Kassala to organize, collaborate and develop

linkages with other actors in the livestock value chain to everyone’s benefit. The project identified and trained local paravets who were providing a valuable service to pastoralists, selling veterinary drugs and training them in improved livestock management skills. A viable pastoralist association was formed and helped farmers to meet regularly to discuss opportunities and constraints in the livestock value chain. The association was an effective vehicle for farmers to collectively bargain with buyers for better prices. It was evident that pastoralists were on the right path to improving their livelihoods. This promising development was dealt a severe blow when a new intervention by another development agent cropped up in the same region.

Instead of coordinating with Practical Action to discuss additional support required to enhance the growth of the livestock market system, the new intervention provided large subsidies to pastoralists in the form of free drugs and training. Promising veterinary drug stores and advisory services were undermined and, as expected, most of the paravets were pushed out of business. If there is one important lesson to be learned from this experience, it is that uncoordinated provision of subsidies by one development agent can completely undo the good work of another. It seems that, in spite of more than seven decades of development interventions, poverty remains intractable, especially in most of Sub-Saharan Africa, and a large share of the blame for this state of affairs should be placed on the indiscriminate and uncoordinated use of subsidies.

Kamran Niazi said that, in Pakistan, the Agribusiness Support Fund (ASF) worked with private

companies, who provided extension services for a fee, including a company that provided fruit fly traps for free, but made money by selling the insecticide and hormones for the traps. At up to 70% spoilage rate, anything that would limit losses would be good for farmers and this was a simple, cheap, low maintenance solution. ASF also worked with exporters’ cooperatives that together paid for the GlobalGAP certification of orchards owned by small farmers—40-50 farmers. The farmers were asked to sell the produce to the exporters and they also could sell the produce to others. The exporters could buy only from these farmers due to the certification requirements and they had to pay a premium price for the produce. But they in turn were able to export the produce and charge much more for the product. In the end, everyone was a winner.

21

Abdur Rob contributed an example of local learning groups of facilitators working in the same region or with the same subsector who can coordinate the use of subsidies when they share their experience to create synergies in their work. A particular example is DairyNet Bangladesh, a local learning group comprising CARE, Practical Action, Intercooperation and Plan International Bangladesh, all of which have been implementing dairy and livestock sector development projects in northern Bangladesh. The network realized that if it could develop a standard training manual for Animal Health Service Providers by using materials available to each organization, it would help them reduce the training subsidy. It also would enable them to start a dialogue with the Department of Livestock Services to underscore the training of animal health service providers that NGOs and private businesses had conducted and request acknowledgement. The network collected and reviewed materials and coordinated the participation of its members in the International Dairy Conference planned held at the Mymenshinge Agricultural University, Bangladesh in April 2010. All four organizations shared the cost of participation and promoted their work through DairyNet in an effort to influence institutions and policy makers.

Ben Fowler [Consultant] noted the increasing interest in applying subsidy principles when working with the very poor. He asked some interesting questions:

1) The majority of practitioners agree that a demand-driven approach relies on participants self-selecting into the project. Yet in many projects, the very poor or vulnerable do not end up participating due to their reduced capacity to take risks and minimal access to investment capital. For projects aiming to include the most vulnerable, what strategies have you used to encourage their participation? Are there ways you've been able to leverage the capacity and assets of better-off community members, for example, horizontal linkages (producer associations)? Have these approaches required subsidies specifically towards the vulnerable that are not available to others?

2) How are people defining the very poor? Comments so far indicate that the definitions differ quite a bit between cases. Are participants using formal tools to determine the most poor or vulnerable in a community (e.g., the poverty assessment tools), community self-assessments or more informal tools, or not measuring at all?

3) Some believe that below a certain point of insecurity, a pure market development approach is irrelevant and consumption stabilization, etc. are more important. CGAP's graduation model is built on such a concept. What do you think? If you agree, how has this belief impacted your programming?

In responding to Ben's point about project strategies that include the most vulnerable, Andy Jeans observed that most producer associations are formed by individuals with shared problems and concerns who tend to exclude those who have significantly fewer assets than they do, seeing them as potential competitors. However, they do make trading arrangements (for commodities or labor) with poorer farmers for large market orders. In the examples of farmers with HIV/AIDs in Kenya and disabled people in Uganda, the formation of exclusive producer groups proved to be effective due primarily to shared problems and concerns. Also responding to Ben, Kamran Niazi said that public subsidies will never be a permanent solution because governments simply do not have the money to cover all available farmland—population and size—especially in developing nations. Furthermore, expecting public subsidies to be

22

completely transparent is unrealistic—everyone is responsible to someone, whether shareholders, donors, government or other stakeholders. And, having resources does not mean you can dictate terms—at the end of the day, there needs to be collaboration and mutual agreement. Alison Griffith considers that a key aspect of practitioners’ role as market development facilitators is to provide spaces and opportunities for public and private actors to find solutions and ways of working together. When that happens, sometimes a subsidy is not needed because resources can be found within the system. Our Nepal program concluded that in the dairy sector it was crucial not to jump in with solutions to blockages, but to be patient and go through a (sometimes messy) process of letting actors determine the key issues and how they might tackle them. If extra resources are needed, a finance provider may step up, especially if they see that players are committed. Practical Action has found this to be the case in Nepal’s dairy sector and with hibiscus producers in Sudan. As Rajiv said, farmer collaboration is critically important to enabling other actors to work with them efficiently and effectively and to ensuring that, as marginalized and weaker players, they can benefit from new products, services and linkages. This is what makes the market development process pro-poor. We could argue that investing in this aspect can be as important as providing subsidies because when farmers are organized, confident and articulate, they are more attractive to AKI suppliers. In some cases, this could negate the need for subsidies. Has anyone else had similar experiences? 6) Financial Services

Ekanath brought up the issue of safety-net packages in areas where communities are impacted by conflict, natural disasters, etc. There is a need to develop products like community-based insurance for both enterprises and individual health and safety. Safety-net packages can increase the risk-taking capacity of the most vulnerable groups and, ultimately, protect them if their enterprises fail. The UNDP South Asian Poverty Alleviation Program in Nepal introduced and successfully piloted this type of package.7

When establishing a community-based insurance system there needs to be a strong savings and credit mechanism and social collateral within the community. In such a system, if a community group provides loans to very poor or vulnerable community members, they should embed the insurance premium .There also should be an insurance committee to monitor the scheme. Initially, communities need subsidized services from support agencies and facilitators to help link the insurance committee with professional insurance providers to develop an appropriate system, provide training on the system and help them develop a sustainable insurance scheme. Terrence Isert [Consultant] has been working with vulnerable populations (poorest of poor) who

are savings group members, encouraging them to become involved in program market access activities. The particular challenges have been the lack of social, physical and especially the human capital needed to achieve sustainable livelihoods. Whether a livelihood or a value chain approach is pursued, vulnerable populations, particularly youth, require additional services and support to increase their participation in a value chain to gain greater access to markets. They also require human capital support that includes life-skills development to help counter the effects of the conflicts to which they have been exposed and to build educational and coping skills that would enable them to manage a business.

7 http://www.undp.org/evaluation/documents/MicroMacro_SouthAsia.pdf

23

7) Markets, Incentives and Investments

Shawn Cunningham asked if we should make a distinction between enterprises competing in markets and people who are trying to survive in spite of markets? For the first group, improving market performance and structure can lead to increased income and the main risk of intervening in the system for practitioners is that someone captures a benefit that was supposed to target the system as a whole. When directing interventions (whether subsidies or subsidized expertise) at the entire market you are safe. But, market efficiency and the ability to access markets mean little for those who are just trying to survive. They do not care if markets select better prices, value or quality, or if they distribute goods from efficient producers into rural areas. We need to approach subsidies aimed at livelihoods. This is not about market failure, it is about something else—perhaps societal failure. It does not seem appropriate to use typical market failure logic to solve problems here. Subsidies to support survival need to be allocated using different criteria than those we use to select interventions aimed at better performing enterprises. Mary Cockram [Consultant] In Colombia, Aid to Artisans worked to integrate displaced people

into markets, particularly in cities where the overall market is stable, by providing incentives (usually subsidized equipment and cost-shared marketing services) for existing enterprises to hire them. In general, that worked pretty well; their sales increased and at least a month or two after the end of the project the displaced people were still employed. In cities hard-hit by conflict as well as serious poverty, existing enterprises were so tenuous that it was much harder to figure out the right incentive that still seemed market-based. A few entrepreneurial people started viable businesses, mostly selling in the more stable capital, but many groups could not overcome internal issues and mistrust.

Kamran Niazi presented three approaches to providing subsidies:

1) How do we distinguish between deserving communities and individuals within them? In every community there are people who have a good chance of success—with or without help. However, with support they have an even greater chance of succeeding. Should we consider the community a homogeneous group of people in need or a heterogeneous group comprising both relatively passive observers and potential change agents? If people are dying of hunger or disease, then everyone has to be supported. But if you are talking about market forces, you should consider who can make the greatest difference for the entire community and provide those groups or individuals with smart subsidies.

2) Are smart subsidies sufficient in themselves to achieve success if we do not select smart

people to use them? There are smart people in every walk of life. Poor people may not be formally educated, but many are sharp thinkers and their street knowledge can make them formidable. Andy’s observation about producer groups who exclude those having few assets because they are potential competitors is exactly what I am talking about here. In the end, many of these groups work with poorer farmers because they need their commodities or labor.

3) Is this about qualitative or quantitative impact? Providing a limited number of smart

subsidies can result in greater impact because people often copy successful enterprises. Is this better than the average or below average outcome that often results from low quality support to many? Another strategy is to create small niches for target groups (in different geographical locations), develop smart interventions and apply them in a single location. After observing the

24

results, group similar niches and replicate and adapt the support. This approach can also work by selecting a large group (in one locality), use a subsidy to identify need and help the group attain it. An example would be to arrange a visit to a leading cluster or market to see what a large group can achieve and then work with the group to identify activities they want to undertake. This option gives you the best of both worlds.

Marcus Jenal [Intercooperation] agreed with Shawn that practitioners need to differentiate between people who are able to interact with market forces and profit from them, assuming that the market system works appropriately, i.e., is pro poor, even for those whose primary concern is feeding their families every day. He provided an example:

The LEAF project in Bangladesh works with community-based organizations (CBOs) comprising a range of people—from the not so poor to the very poor. Improving their human and institutional development and organizational / business management capacities and establishing a service provision system for them has enabled the CBOs both to interact better with local government bodies and to profit from market opportunities. In general, this works very well. However, there is one category of people in the groups who are unable to profit from the opportunities opened by the CBOs, mainly those who have no assets at all. So far, the project has had little success in reaching those people and is working on different strategies:

1) Make the CBOs more socially inclusive. The project works with the CBOs to raise awareness of their responsibilities towards the whole community, particularly the extremely poor and marginalized, understanding that this is not the responsibility only of the CBO, but of the entire community. Due to their strengthened organizational capacity, the CBOs are in a position to take the lead in raising funds from the whole community for social activities or lobby the local government to better support the very poor through safety net programs or by giving them access to public resources like roadsides or public land and water.

2) The project encourages the CBOs to use their savings to finance economic activities for the very poor to start businesses. They can provide grants, soft loans or investments on a profit-sharing basis, a strategy shown to be quite successful. Some CBOs have bought livestock for the extremely poor and lobbied the local government to give them access to unused public land to feed the animals. The beneficiaries care for the animals and repay the loan by providing either a fixed amount of meat or offspring.

3) The project also encourages the CBOs to subsidize service provision for the businesses the very poor started with help from the CBO.

For us it is clear that there is a need for special efforts when targeting the extreme poor. Success to date has been very limited and it would be great to hear about other experiences and case studies on how we might make our efforts more successful. A major question is how to find systemic interventions that can change the status of the extreme poor. In contrast to market development activities, where it is clearer (although not always easy) how to target the system in order to make the market work better for the poor, when working with extremely poor people, the interventions are often direct and not very systemic. 8) Gender

James Tj [National Innovation Foundation] has found that subsidies given to women’s self-help groups (SHGs) have more impact than do those to individuals and he thinks this is an area that needs more

25

attention. Experience in southern India shows that women’s SHGs are very strong and both government and NGOs use this strength for development interventions. In India, banks do not hesitate to give loans to these groups as they are accountable, responsible and practical and their repayment rate is more than 90 percent. Support reaches the right person and is used for the right purpose and, since the entire group is accountable for the subsidy, all members are involved in monitoring loan use and repayment.

26

Annex A: Key Points of the Discussion Paper

Key Principles for Subsidy Use

The purpose of a subsidy is to mitigate risk and build capacity to support innovation and change and no mechanism works in all situations. Practitioners must use good judgment, be prudent, ensure transparency and facilitate negotiations between actors when identifying the need for a subsidy and implementing an agreement. While there are no universally applicable rules for using subsidies, some general principles do apply:

Avoid subsidies when possible—use them only with a clear vision for how they can achieve systemic change.

The better the facilitation, the less need there is for subsidies—create conditions that allow market actors to recognize business opportunities and pursue them without subsidies.

Consider the possible negative consequences of unnecessary subsidies and adopt a do-no-harm approach—do not destroy private businesses and competitiveness, create dependency, block indigenous resources or diminish creativity.

Use subsidies to reduce risk for market actors—particularly new technology / service adopters as long as they do not detract from sustainability or create dependency.

Use subsidies to strengthen demand / supply markets, not to subsidize transactions— link marginalized producers to market actors who can provide services and inputs (facilitators should not provide these services).

Develop / use strategies that promote ownership by market actors.

Determine who will pay for and undertake services when subsidies end.

Be as inconspicuous as possible; make your role as facilitator invisible. Base selection of the type of subsidy to use on the intended target group—development,

production, promotion, provision, consumption.

Use subsidies to improve market system structure and dynamics to benefit the most marginalized—for example, facilitate links between a credit provider and a local paravet to redress inequalities and exclusion and expand access to productive resources—financial, physical, human, etc.

Recognize that those who need subsidies are not always the poorest.

Encourage self-targeting strategies that help those who want to participate to step forward.

Tips for Implementing Subsidies

Identify public / private sector actors having an incentive to provide services / products and support development and testing of models.

27

Always co-invest with market actors—they should match your investment.

Find supporting market actors who can undertake specific educational activities.

Identify and clearly communicate subsidy purpose and target.

Set and communicate a timeframe for terminating a subsidy; transition it out over time.

Identify and address systemic constraints, i.e., non-financial reasons value chain actors do not invest.

Ensure market incentives, not subsidies motivate market actors.

Subsidize as small a portion of the total cost as possible.

Explore with market actors the possibility of embedding the cost of their services in an existing transaction.

Encourage self-selection to make subsidy support contingent upon a demonstrated financial or other commitment and improve the chances of success and impact sustainability. This may exclude those who are more disadvantaged.

Use subsidies for a limited time to get things going; this allows market actors to see and feel the impact of new ways of doing things and helps them build capacity and systems.

Identify and plan a long-term solution for meeting costs before introducing a subsidy, e.g., a new MFI loan product.

Have a clear exit strategy.

Justifiable Subsidies

Facilitation—a temporary, external intervention by a non-market actor

Capacity-building activities that allow new players enter the market system and explore new business opportunities

Training, demonstration and pilot activities—learning visits, trade fair attendance, etc.— that confirm the impact a new business model can have and provide opportunities to explore new ways of doing things—business / organizational models, technologies, techniques, etc.

Ice-breaking activities that link the marginalized to new markets, opportunities, etc.

Market research and feasibility studies to demonstrate that new ways of doing things can work.

Technical assistance for R&D to confirm the efficacy of new business models, technologies, etc.

Group formation.

28

Difficult to Justify Subsidies

Recurring operational and working capital costs, including personnel

Physical assets, e.g., buildings or machines, except when subsidies can stimulate communal infrastructure like collection centers that can catalyze change for many.

Operational knowledge needed to produce something or operate a business.

All subsidies that do not have an exit strategy.

Some Considerations When Subsidizing AKI Services

Good sources of AKIS: input suppliers (retailers, sales reps), outgrowers with a contractual selling agreement and (sometimes) marketing cooperatives

Ineffective government extension services: lack of performance incentives, information gaps between final users and extension agencies, researchers’ misconceptions about marginalized producers, farmers’ inability to influence government

Radio extension messaging often not commercially viable; usually needs ongoing subsidy or commercial sponsorship, e.g., from an input supply company.

Need for AKIS is continuous; not a one-time service. Projects must create a process that ensures ongoing communication, service delivery, access to and use of AKIS by orienting providers to help clients improve productivity and product quality over the long term, not just until the project ends.

Central Questions and Challenges

1) Blanket provision of subsidized services or inputs can reduce their effectiveness—provide subsidies to those who need or can make the most effective use of the service / input. How can we ensure a high rate of self-selection amongst market actors and what is the role of subsidies in achieving this?

2) Should we disclose to stakeholders that we have funds to subsidize services or inputs? Or, should we withhold this information until all non-subsidized alternatives are explored? How does this affect the ideal of self-targeting?

3) Subsidized services can prevent recipients from telling us which services they want. How can we make sure their voices are heard before using subsidies? How can we manage their expectation or assumption that we want to hear from them because we want to provide subsidized / free services?

4) If the targeted group is very disadvantaged, how do you meet the cost of training, group formation or other services? Should the market cover these additional social costs? What strategies work well when facilitating access to AKIS in areas lacking a moderately strong

29

private sector, e.g., input suppliers or purchasers? Is direct, short-term provision by a facilitator ever defensible?

5) Are there effective ways to increase the quality of government-provided AKIS? Is there evidence that this could evolve into a viable option for smallholders in remote areas?

6) Where should we focus our subsidies? Governments tend to play a more interventionist, subsidizing role in AKIS than they do in other sectors and are likely to continue doing so.

7) In developing a market for AKIS, how do we coordinate, complement or mitigate conflicts between public / private AKIS providers that have different subsidy practices?

8) How can ineffective public extension services be improved, particularly where inadequate infrastructure and adverse incentives limit their usefulness and efficacy?