The area around Makola market is crowded with traders trying to carve out a small space to display their goods and attract the attention of potential customers. Rentable shops and market stalls are in short supply. It is fairly common for a property owner to ask for 10 years up front to rent such a space. Informal arrangements are common. For example, a shop owner might permit sellers to use the space just outside her door in exchange for keeping an eye on things. The video below depicts a market clearing exercise by the Accra Municipal Authority (AMA) from a few years ago:
There are now quite a number of experimental designs and RCTs (randomized controlled trials) of market information systems measuring revenue gained by rural agriculturalists (farmers, fishermen) and traders. While some show improvements to the prices farmers get for their crops, others disappointingly show a lack of impact. There are also studies of agricultural extension services delivered via mobile phone which suggest that crop productivity is boosted by such services. How might these studies guide what actions are taken (i.e. to provide such services) in the real world?
What is ‘ecological validity?’ It is the relationship between controlled, or lab-based experiments and what unfolds in a non-lab, non-controlled environment (aka the ‘real world’). Out in the world multiple variables interact in different ways. So what can we take away as actionable from an RCT that shows a market information system has resulted in boosted revenues for farmers who used it? How do we reconcile the successes with similarly designed studies that showed a failure of impact?
Recent RCTs of MIS have been designed to provide the ideal circumstances for their success. They do so in the following ways:
1) providing the service for free [see Cole and Fernando 2014, Fafchamps and Minten 2012, Nakasone 2013, Hildebrandt et al. 2013]
2) automatically enrolling farmers and pre-customizing information delivery [Nakasone 2013]
3) providing extensive training on the use of the system to overcome literacy limitations [Hildebrandt 2011]
Why do they do this? Because these studies intend to show what impact information (in isolation from other issues) has on markets, farmer profits, and revenues. That does NOT quite make these studies true evaluations of a particular system in real-world use.
So what does this mean for the real-world use of MIS systems such as Esoko, Reuters Market Light, etc?
For item (1) it means that a successful business model will most likely require that it be provided for free, (at least in the short to medium term) if it is to yield the effects measured by recent studies.
For (2) and (3) it means that service adoption and literacy issues remain significant and substantial barriers. Artificially eliminating them for the purpose of an RCT is problematic because it obscures this fact. Fortunately, many programs in this space are finally becoming more attuned to this reality and offering such services through more user-friendly and accessible IVR or call center models (as Esoko now does). For more literacy-demanding services, training (or infomediaries) will be needed. Where training is needed, the scalability of such services is an issue.
Susan Wyche and Chip Steinfield have a paper that’s just come out evaluating the user interface issues around MIS. They look at the affordances of such services and the challenges they pose for farmers. By contrast, RCTs championed by economists fail to address usability and literacy issues when they treat ‘information’ as the only important variable to isolate. It will likely require more interest and support for researchers with different kinds of training (HCI, social science) to make sure this important issue is not trivialized.
Cole and Fernando (2014) ‘The Value of Advice: Evidence from the Adoption of Agricultural Practices‘ Harvard Business School Working Paper.
Fafchamps and Minten (2012) Impact of SMS-Based Agricultural Information on Indian Farmers. The World Bank Economic Review.
Nakasone, Eduardo (2013) The Role of Price Information in Agricultural Markets: Experimental Evidence from Rural Peru. Selected Paper prepared for presentation at the Agricultural & Applied Economics Association’s 2013 AAEA & CAES Joint Annual Meeting, Washington, DC, August 4-6, 2013.
Hildebrandt, N. (2011). Teach a Man to Text. Esoko, Insights from the Field.
Hildebrandt, N., Y. Nyarko, Romagnoli, and Soldani (2013). Market Information Systems for Rural Farmers: Evaluation of ESOKO MIS – Year 1 Results. New York University: Africa House Blog.
Wyche, S. and C. Steinfield (2015). Why Don’t Farmers Use Cell Phones to Access Market Prices? Technology Affordances and Barriers to Market Information Services Adoption in Rural Kenya. Information Technology for Development
A warm welcome to visitors who’ve found their way to this page from the Catholic Relief Services ICT4D conference.
To know more about the objectives of this 5+ year research project visit our about page.
If you are involved in an ICT4D project in the domain of agricultural, markets, and trade please visit our projects page to see if it is listed.
If your project isn’t listed let us know (or post a comment below) and we’ll add it to our listings!
A very small selection of hundreds of interviews with farmers, fishermen, traders, and others have been written up as profiles. Read these to get a richer sense of how ‘the poor’ and the ‘non-affluent’ manage, survive, and thrive with mobile phones in hand.
Reposting this entry from The Connectivity, Inclusion, and Inequality Group blog at the Oxford Internet Institute. (see the original post here and download the full report on connectivity and the tea sector in Rwanda)
by Chris Foster
Market information systems are growing in popularity as an intervention by governments, NGOs and private firms alike. New information provision for farmers, often price infromation over mobile, have been touted as a way of helping farmers improve the price they get for their produce, and reduce their dependency on middlemen. In our research we explored the impacts of connectivity in the tea sector in Rwanda, and we spent much time mapping the actors, information flows and relations that gave us critical insight into the potentials and limitations of market information systems.
What determines price?
Before we come onto talk about market information systems, I want to talk about ‘price’ and I want to problematize the assumption that the low prices that smallholders receive are the result of a ‘lack of information’. Let’s look at the tea sector in Rwanda. In Rwanda it is only recently that smallholder farmers have been paid based upon a market price (as opposed to a flat government-set rate). So, the tea sector provides us with an interesting natural experiment to observe what happens when smallholders are exposed to market prices – do they seek out and use information and knowledge to improve prices?
Not really! In fact we found little evidence in our research that new price-based trading changed farmer’s behaviour. Quality (and hence price of tea) is not really determined by elements that the farmer could influence. Price is more linked to international market fluctuations and to ‘natural’ endowments (location, soil quality, climate). For farmers who looked to increase quantity, the nature of land sub-division in Rwanda means that it is difficult for smallholders to greatly increase yields and new techniques, fertilisers and irrigation are rarely designed for smallholders. Finally, there is little room for smallholder to choose or skip intermediaries – tea leaf from smallholders needs to quickly be processed, so farmers have little choice about which processor they sell to. It has to be the nearest one.
Even with the above reservations, we explored if smallholders might have some ability to improve their prices through information and knowledge. What types of market information are available and how are they used in the tea sector in Rwanda?
There have been some sectoral efforts to disseminate price information in tea, and there is a price information system available. This market information system was supposedly accessible for tea co-operative associations (which all smallholders are a member of), but when we interviewed co-operatives there was an inconsistency in knowledge of the system: some co-ops had access, but some co-ops didn’t know the system existed, still others used other relationships to find out about price. In short, our feeling was that it was difficult for co-ops and smallholders to trust price information, and how they might use this to improve was unclear.
There were also a number of other social factors that meant that efforts to improve tea through market information had been less than successful. The demographics of tea smallholders in Rwanda does not favour innovation. Agriculture in general, and specifically tea is not seen positively, and the backbreaking work of sowing, weeding and harvesting tea is not one to attract dynamic entrepreneurs – the majority of tea producers are male and older, whilst youth tend to prefer to look for work in the informal sector in a local city. In Rwanda policy also works against tea, where certain regions are exclusively assigned to tea growing, and so there is less ability for diverse portfolios or rich inter-cropping approaches for smallholders.
Market information and the bigger picture
In recent years there has been a growth in discussion of ‘pro-poor’ value chains, of poor farmer becoming involved in more commercially-orientated agriculture. Across the board, NGOs, donors and commercial firms have subscribed to this perspective and I think that market information systems is one element in this new paradigm in agriculture. Bringing development through profit, new clever finance, insurance, ICTs, market information and prices. However, pro-poor value chain approaches are often narrowly focussed on the immediate economics and neglect the structural (politics, colonial histories, land ownership), social (diversity of livelihoods, risk and gender) and complexities of production.
Market information does have a place though. From our work in tea we see a number of ways that implementers and developers of market information systems might improve activity. As I hope this outline shows, the value of qualitative research is invaluable to designers. My recommendation would be that all market system designers should spend time seeking to understand the supply chains, the relationships and small holder goals before starting to develop any system. How does price break down in the supply chain? Where might price (or other) be useful and when might it not? In hand with this, our own evidence suggests that more integrated approaches are important for success – in terms information, such as price, farmer information (pests, fertiliser, skills), contacts, market intelligence, and in integrating the diverse communication ecologies of farmers.
There is also value is designers thinking of information not only as a top-down exercise. In the tea sector, we saw sporadic online sharing between co-ops relating to tea management and production (horizontal rather than top-down information flows). Indeed, intermediaries such as cooperative were a key node in these information flows. Thus market information systems which consider bottom up information flows and intermediary players more overtly would be useful in improving the ability for market information systems to have impact.
The daily financial lives of the poor are, without doubt, immensely complex. Irregular, uneven income streams are frequently unable to meet recurring and emergency expenses. Savings deposits and insurance schemes are almost non-existent, and reliance on credit is high. These credit instruments in turn are multifaceted; low-income populations diversify their credit sources, from the more formal bank loans (wherever they are actually eligible for these) to semi-formal microfinance loans, to the more common, often reciprocal, informal loans that they give to and borrow from their immediate social networks. The mobile phone is slowly and steadily transforming the culture of lending to and borrowing from one’s immediate friends, family, and neighbors. Moreover, the mobile phone can be a window into the social networks of people, and more specifically, their financial relationships as well as conversations. Jackson is a boda-boda driver from Kampala, Uganda, and his call logs reveal both, his financial relationships, as well as his conversations about money. Read his full profile here.
In the absence of agricultural extension officers (paid by the government to disseminate information and expertise on farming), many farmers rely on the local seller of agricultural inputs for tips and advice on handling crop diseases and pests and dealing with other farming problems. Many new mobile phone apps propose to deliver agricultural extension services via SMS or, in a few cases, through interactive voice response (IVR). An unanswered question is whether farmers are willing or able to pay ‘purely’ for information provided through these services. The agricultural product seller makes his money not directly from his good advice, but from the farming inputs he sells. His good advice may, however, draw farmers to his shop and motivate them to buy the products he sells.
Read the full profile here.
We are now in the process of analyzing data from last summer’s fieldwork on mobile phone use and money practices (conducted by Ishita Ghosh and Aisha Kigongo – with remote support thanks to Skype and Dropbox from Jenna Burrell). A standard schema of categorization when trying to understand the personal finances of low-income populations is to identify the repertoire of financial instruments used (bank savings accounts, bank loans, microfinance loans, savings groups/ROSCAs, remittances, etc). These instruments are, in turn, often grouped into two broad categories ‘formal’ and ‘informal’ [see Portfolios of the Poor (Collins et al) for a good example]. In development terms, mobile phones are defined within this schema as a gateway, an access point into the ‘formal.’ This is what is communicated by the notion of “banking the unbanked.” We are finding through our more all-encompassing approach to examining the multifaceted ways low-income populations incorporate the mobile phone into money practices that this division (and the presumption of a transition from informal to formal) is too limiting. Rather when considering the consequentiality of the mobile phone (and mobile money) more material concerns – liquidity, security, concealment/privacy, instanteneity, and obligation fulfillment at a distance — are perhaps better terms to reflect the concerns and advantages as they are framed by low-income Ugandans themselves.
Below we have financial instrument usage for our research participants in Kampala, Uganda (the small subset that were interviewed weekly throughout the summer). Color coding offers a sharp all-at-once visualization of this breakdown. Red is for financial instruments the research participant does not use (at all and not ever). Yellow is for financial instruments used, but that are not an important part of the participants routines (such as a bank account opened, but almost never deposited into). Green is for financial instruments used regularly and part of money management strategies.
Clearly the bulk of financial instrument usage remains in the ‘informal’ category rather than the ‘formal’ among these research participants.
formal financial instrument usage (Uganda)
informal financial instrument usage (Uganda)
The mobile phone
In Kampala, mobile money services have taken off (it’s another story entirely in our Delhi, India site). As much as mobile phone-based financial services might in some sense ‘formalize’ financial practices (for example, by creating a persistent record of money transfers and payments and by channeling formerly cash-based money movement through the networks of telecom companies and banks) they also resolve some recognized problems and limitations on the informal side. For example, phone calls and mobile money are a routine part of making requests for small interest-free loans from friends, family, and neighbors. They improve the ease of making such requests and also of seeking repayment. A question rarely asked is whether the mobile phone might make more legitimate and attractive the wide range of ‘informal’ financial practices. Might the mobile phone preserve informality? While the value of money transfer services for remittances (generally seen as ‘informal’) has been widely recognized, the mobile phone appears to have a role to play in small loans channeled through interpersonal relationships, and in the organization of savings groups (ROSCAs or chit funds).
This month Janaki Srinivasan presented at the ICTD conference on her project revisiting the site of a canonical study on mobile phone use by fishermen in north Kerala. The original study (by economist Rob Jensen) showed compelling evidence that fishermen used mobile phones to check prices at different beach landing sites leading to improved efficiency in the market as a whole and welfare gains to small and large-scale fishermen. In our revisit we sought to explore the use of mobile phones in this industry more broadly and to consider the history of the region, the groundwork that made these efficiency and welfare gains possible. In the process we raised issues about the potential to scale ICTD results across politically/historically/economically heterogeneous regions. Our findings also question how policy-makers and technology designers might go about drawing ‘implications’ from such studies.
[reposted from the IMTFI blog.]
Field site and findings
We first encountered our field site through a study that has become canonical in ICTD: economist Robert Jensen’s study of mobile phone use in fishing markets in north Kerala. Very simply put, Jensen found that the use of mobile phones to share market price information made fish markets more efficient while also improving producer and consumer welfare. Our goal with this project was to understand the geographic and political economic conditions in which Jensen’s findings hold and to examine questions of generalizability. In addition, we wanted to expand definitions of welfare to encompass more than an increase in income or savings.
We used a different methodological approach to frame and tackle these questions, constructing an ethnographic case study comparing two sites, one in north Kerala where Jensen’s study was conducted and the other in south Kerala. We conducted three-and-a-half months of participant observation and interviews in 2012 at fish landing centers in Kozhikode district in north Kerala and Thiruvananthapuram district in south Kerala. We spoke with fishers (owners of small and large boats, workers), fish purchasers (wholesale traders, small-scale vendors, including male and female), auctioneers who conducted the ‘auctions’ by which most fish was transacted between fishers and buyers, activists and others involved the fish supply chain. Based on our research, we found that
- fish trade in north Kerala is a special case. The region’s geography as well as prevalent investment and credit relationships allowed fishers the flexibility to sell at different markets. In south Kerala, we found that all-season landing centers were farther apart than in the north. Further, boats were bound by a credit relationship to a single investor rather than to several investors as in the north. For these reasons, fishers in the south had less flexibility to sell in different markets.
- Moreover, in both regions, only specific categories of actors within a landing center found price information critical in making trading decisions and regularly used phones to ascertain it.
- A majority of those at the fish market were using mobile phones in a much wider range of activities related to their livelihoods.
- While a majority of these individuals perceived mobile phones as having enhanced their livelihoods and well-being, their implicit definitions of ‘welfare’ were rarely focused on improved incomes alone, emphasizing instead how they used their phones to maintain relations within and outside the market, and protected themselves during times of risk, vulnerability, or emergency.
Describing our first and second conclusions, and the list of differences between fishing in north and south Kerala, is beyond the scope of a short blog post such as this one. Therefore, we focus here on our third and fourth conclusions. We discuss the varied uses of phones and the many definitions of welfare that were emphasized by different actors in the fishing industry. Based on our observations, we argue that the seeking of market price information via mobile phones should not be given an over-privileged role.
Mobile phone use and definitions of welfare in fish landing centers of Kerala
At the time Jensen conducted his research (1997-2002), phones cost Rs. 5000 (around $106) on average, and there was a clear division between those who possessed phones and those who did not. By the time we conducted our study (2012), phones could be purchased for as little as Rs. 700 (around $15) and many owned multiple phones. No boat went out to sea without a phone and most typically had multiple handsets onboard. Nor was this restricted to fishers: an auctioneer told us “There’s no business here without mobiles” and we heard this from almost all categories of actors operating in both north and south Kerala sites. Jensen’s focus on who possessed phones and who did not is therefore less significant today than it was during his study. The more interesting question today is how phones are being used by different categories of users. In addition, we saw that other technologies such as GPS and echo sounders have become popular since Jensen’s study, making it worth asking how phones are being used in conjunction with these technologies.
Among the broader uses of the mobile phone, co-ordination work between the different actors in the fish economy constituted an important category of uses. This was important in fish marketing activities, as well as in fish preservation. Boat owners and fishing crew described, and we saw, how they would call their auctioneers a few minutes before they arrived at the shore to ensure someone was on hand to perform the auctions. Fishers in the south also mentioned discussing the timing (rather than site) of landing to optimize pricing. The ice-seller on the shore called the ice company to order ice based on how much fish was being transacted on a given day. Wholesale merchants and export agents also mentioned using the phone to communicate details of the trucks on which they were sending fish to agents at the destination. The perishability of fish, of course, was part of what made this coordination work critical.
Phones were also mentioned in the context of coordinating or balancing work and home concerns, most often by women vendors who operated in south Kerala. With the growth of fish exports, the presence of export company agents on the beach and the entry of cheaper fish from neighboring Tamilnadu, small-scale vendors are increasingly being marginalized in this region. Many small-scale vendors in this region have started traveling to markets in Tamilnadu to buy cheaper fish. Women comprise a part of this population that travels long distances everyday to purchase fish. A woman vendor’s work day, which includes traveling by public or hired transport because she doesn’t own a vehicle, attending an auction, purchasing fish and selling it at a market or at individual houses, can last longer than 12 hours. Since women are also seen as the caretakers of this family in the prevalently patriarchal structure of the region, they worry about their families and their children throughout the day that they spend away from home. Many of them mentioned that having a phone helped them inform their family of their schedules and delays, know what was going on at home, and relieved them of constant worry.
Just as frequently as coordination work, people on the ground mentioned fish-finding as a prime reason for using mobile phones. Fishers used phones at sea and on shore to gauge fishing grounds on a given day. We found, in addition, that phones were often used in conjunction with the Garmin GPS units that all fishing units carried. The GPS was used to mark and specify the exact location where fish had been found. Fishers both used these markers themselves at a later date to look for fish, and also shared them with friends and relatives, a practice also noted by Abraham and Sreekumar. The widespread use of GPS and echo sounder technologies to pinpoint the location of fish and the use of GPS coordinates to share such prime fishing locations with other fishers post-dates Jensen’s study and is another element of the changing industry. While the sharing of fish locations is limited to fishers’ close social networks, it is worth noting that Jensen dismisses the likelihood of such a practice existing at all and as being against fishers’ self interest.
Finally, mobile phones (along with other communication devices) were perceived to be important in times of emergency as others have also noted. Fishers used both phones and wireless sets (the latter were typically installed only on large, mechanized boats) to contact the shore or other fishers in case of emergencies (such as running out of fuel, a damaged engine). Fishers frequently mentioned the dangers of fishing. A fisherman in north Kerala relayed a story of being out at sea when the fuel finished and his eventual rescue following a phone call placed on a satellite phone to a coastguard office, adding “I have great respect for this device because it saved our life.”
Using examples from north and south Kerala, we outlined six primary uses of the mobile phone – (1) price information gathering in combination with (2) arbitrage work (as considered by Jensen), as well as (3) coordination work, (4) balancing work and family (5) fish-finding, and (6) emergency response. We did this to question the often singular attention placed on the first two, and the pithy statement that commonly circulates in the aid sector and the mass media that ‘farmers/fishermen use mobile phones to get a better prices for their goods.’ What we heard from fishing industry actors in the field in both north and south Kerala is that there is no single practice that prevails as the most significant or universally valued use of the phone. It is important here, we argue, not to mistake the focus and priorities of disciplines (such as the concern in economics for how information asymmetries affect market functioning) for the interests and priorities of target populations. There are opportunities in the ICTD space (perhaps underexplored) to support the underlying needs that these alternate practices reflect.
The varied uses of the phone among these actors are matched by almost as many understandings of ‘welfare’ in their lives. People did not define their well-being or welfare primarily in terms of their income, or in terms of optimizing it. Many of them, especially the owners and crew of vallam and gillnet boats, and small-scale vendors, spoke instead in terms of managing or coping. They spoke of their physical and mental well-being, sometimes prioritizing that over an increased income (such as fishers who spoke of wanting to sell quickly and move on to rest, rather than wait to get the best price). The survival of a fishing unit lost at sea or caught in a storm is, of course, critical to fishers’ own long-term welfare and that of their families. Fishers and others in the fishing supply chain spoke also in terms of maintaining relationships, with fellow fishers, their auctioneers, or regular buyers, rather than solely in terms of optimizing their incomes (as reflected in practices of sharing fish-finding locations). These practices may very well eventually lead to improved incomes, but in a longer term and less easily measurable way. They also lead us to ask if Jensen’s definition of the fisherman’s problem as “maximizing profits by choosing where to sell their fish” or concluding with income increases as “welfare benefits” doesn’t narrow our understanding of the reality of the fish market.
The broader implication of our work, which we are only able to flag in this post (but is available in more detail here), concerns the importance of situating analyses of technology use in the history and political economy of a region and a sector. Through our conversations and examination of archival material, we found that fish markets in Kerala have been shaped by the regulatory influences of both fishers’ collectives and the government, in particular the creation of fishers’ co-operative societies, the systematization of an auctioning system and the loosening of traditionally coercive credit relations between fishers and those who provided them capital to purchase and operate their fishing craft and gear. We suggest that when studies (such as Jensen’s) omit these features in order to create parsimonious models and explanations, we end up representing particular markets as more “free” than is warranted. In turn, this potentially blinds us to the power dynamics that shape the daily working of such a market, including who uses technology and towards what end.
Ms Sun* is a university graduate in her mid 20s, originally from a village in the Shaanxi province. She graduated in geography from a university in Beijing, and on her last year as a student she took the exam to become a “university graduate village official” (大学生 村官). This is a country-wide program that started in 2006 with the dual goal of providing job opportunities to university graduates, and of supporting village officials by sending them educated aides to help with administrative tasks, especially using computers. … Read the profile …