We all know the story. Fishermen use mobile phones to call ahead to different markets to find the one with best prices, so they can sell their catch for the largest profit. We’ve heard this line a thousand times, and probably have even told it ourselves.
The claim that mobile phones are used to increase fishermen income was first made in academic literature by Robert Jensen in his 2007 paper, The Digital Provide: Information (Technology), Market Performance, And Welfare In The South Indian Fisheries Sector. Google says over 1,077 other research reports cite this paper. It may be the most cited work in ICT4D academic literature. Sadly, it’s certainly wrong.
Subscribe to ICTworks now and get more insights like this delivered to you 3x a week.
The Case Against Fishermen Using Mobile Phones for Market Prices
In a stinging critique, Dr. Jacques Steyn, Head of the School of Information Technology at Monash South Africa, demolishes Jensen’s claim that mobile phones improved the economic welfare of Kerala fishermen using ideological, paradigmatic, methodology, logical, statistical, and semantic arguments.
A Critique of the Claims About Mobile Phones and Kerala Fisherman shows that the fisherman using mobile phones for market prices myth must be buried and forgotten – except to serve as an example of how research in complex social systems should not be conducted.
Mobile Phones Are Not Used for Market Prices
Other researchers contradict Jensen’s first claim, that fishermen use mobile phones for market prices. They found that smaller fishermen who typically fish within 2-5km offshore use mobile phones to communicate with their financiers and family about when they would return to shore. They do not communicate about market prices.
In fact, fishermen are more concerned with ham radio, which has a much greater range than mobile networks, and even then for non-market price usage. Fishermen have petitioned the Indian government for free ham radio and GPS equipment to improve weather and location awareness. They have also set up VHF networks for weather alerts, not market prices.
Fishermen Cannot Choose Their Markets
Central to Jensen’s argument is that fishermen can chose which market to sell their catch at any time. However, fishing boat movement is regulated by law and by their lenders. The Kerala Marine Fishing Regulation Act (1980) requires boats to be licensed by port and restricts fishermen to sell their catch only at their registered port. Within that port, auction agents, who finance fishermen, require the fishermen to only sell their catch to them, not any other auction agent.
Other researchers also report that fishermen would not go to another port because they would not know the people at the other sites, and therefore would not trust them. Fishermen prefer their assigned port where they have trusted people networks.
This can be a reason why fishermen, when they do use mobile phones, use them to call about when they will arrive in port, not about market prices or which port to go to. They have no choice in port locations or the market price they’ll get there.
Confusing Correlation as Causation
Jensen’s next error is a classic one. He conflates an increase in mobile sales with increases in fish market prices, and an increase in fisherman’s profits. First, while mobile phone sales did increase in India, at the time of Jensen’s research, only fishermen with large, mechanized fishing boats had adopted mobile phones. These trawlers usually fish 20Km or more offshore, well out of range of land-based mobile networks.
Next, while fish market prices increased, there is no evidence it was from mobile phone usage vs. general population or market changes (more people, more wealthy people, more demand for fish, etc). Finally, there is no evidence that fishermen have seen increased profits from increased fish prices. Fishing costs could have risen, driving fish prices up, without any change in underlying profitability.
Using an Anecdote as Evidence
Jensen’s final error is one we’ve all done too. From 5 ports in one particular sub-region he overgeneralizes firstly to all fisheries in Kerala, and makes a giant leap to everywhere in the developing world. Just read his conclusion:
“By improving access to information, ICTs may help poorly functioning markets work better and thereby increase incomes and/or lower consumer prices. In fact, it has become increasingly common to find farmers, fishermen, and other producers throughout the developing world using mobile phones, text messaging, pagers, and the Internet for marketing output.”
Now isn’t that the same conclusion you’ve given to your ICT4D experiments and pilots? That one idea tried in one context could change the entire world? Could it even be immoral for us to announce that our latest ICT4D innovation is going to turn poor communities into vibrant first world economies? Its certainly not good research.
Interesting article, and worthwhile caution to not ‘exaggerate from anecdote’. However, SMS based ‘price-checking’ is used in a number of small agro-fisheries sectors — going back to the original e-Purji system from many years ago – sugar cane farmers checking prices faster, reducing crop loss from rotting. Here’s one quick reference:
http://kuisc.blogspot.com/2014/04/sms-for-sugarcane-farmers-purji.html
http://www.e-agriculture.org/content/impact-sms-based-agricultural-information-indian-farmers
https://www.usaid.gov/div/portfolio/text-tips-increasing
Dear Alain, Thanks for responding . Through our Wireless Reach program we have invested in several programs providing information services to Fishermen in 4 countries in 3 continents. Our experience of working with fishermen in India confirms that Fishermen find market prices of fish useful as they can negotiate better with the traders. Will be responding in detail soon
Hey Wayan,
I think the issue that people below seem to be responding to is the way the title is worded – as if to say the entire premise is wrong and invalid. I think the point you were making was a good one – to avoid hyperbole and over-exaggeration of anecdotal findings, but there has been, since this manuscript, a much larger body of trial and discovery in support of market-price access by agro-fisheries. Do you agree with this ?
-Alain
This is a very disappointing article. The title is only meant for clickbait and suggests the Jensen purposely misled people. This is far from the case. There is no “lie” here. There is one highly cited, peer reviewed academic publication, making claims based upon detailed analysis of data using economic theory (Jensen’s paper). Then there is a new working paper making claims that there were misunderstandings of the political and economic context in that paper (Steyn’s paper). We cannot assume that Steyn is correct because his paper is new.
A couple questions:
1. Did you try to contact Jensen for comment?
2. Has Steyn’s paper received critical review?
A highly cited paper that on closer inspection is highly flawed, warrants a highly contentious title to counteract the original paper’s highly suspect acclaim.
I disagree. Titles like this do dis-service to good and interesting writing that is usually shared through this site. And you are are still assuming that Steyn is correct. I am not convinced by his analysis or methods and his paper does not seem to have been subject to critical review.
Then we will have to disagree. Titles like this get people to click, which is the hardest aspect of information dissemination today. As to validity of opinions, Steyn makes a convincing argument, better than Jensen, which I had not read till now. Not everyone needs to wait for peer reviews to change their mind.
The is no good in disseminating information if it is not based upon particularly good analysis. In fact, this exactly the problem with reporting in science. From my brief reading of Steyn, I can’t see how you find his critique convincing. For example, Steyn says “Jensen does not state how the data on market prices was collected” (2). If fact, Jensen very clearly states that they interviewed fishing units “regarding that morning’s market, each fishing unit was asked for the amount of fish caught, market of sale, quantity sold, sale price, time of sale, costs, and whether they used a mobile phone” (890-891). This is from a very quick scan of the papers. I’d be very surprised if there weren’t other examples like this
Again, I think that comments from Jensen could have made this a valuable piece of writing. This is an important question and issue. The research on prices shared through ICT is not clear. There are numerous papers on these questions that illustrate the complexity of the question. I’ve added a list of papers below . Impacts of agriculture prices shared through ICT are diverse and underscore how complicated agricultural markets are. Debate in the practitioner and researcher community on this topic is very important and valuable to developing new programs. But misleading articles don’t help create valuable discussions.
Aker, J. C. (2010). Information from markets near and far: Mobile phones and agricultural markets in Niger. American Economic Journal: Applied Economics, 2(3), 46-59.
Fafchamps, M., Hill, R. V., & Minten, B., (2008). Quality control in nonstaple food markets: evidence from India. Agricultural Economics, 38(3), 251-266.
Goyal, A. (2010). Information, direct access to farmers, and rural market performance in central India. World Bank Policy Research Working Paper 5315.
Mitra, S., Mookherjee, D., Torero, M., & Visaria, S. (2015). Asymmetric Information and Middleman Margins: An Experiment withIndian Potato Farmers. Working Paper.
Steyn tears apart the very question Jensen uses, which is “did you used a mobile phone” – not what did you use a mobile phone for. Yes, they used a mobile, to say when they would arrive, not to get market prices.
I would argue that this is not a particularly close inspection, and certainly not one that would warrant calling the original paper highly suspect. Steyn’s main attacks on Jensen’s evidence are cherry-picking and not even doing that well.
For example, Steyn claims that Jensen bases his entire premise of ICT use for output markets on newspaper articles. Jensen himself calls that evidence anecdotal and cites it as the motivation to investigate micro-level effects of mobile phones in context.
Steyn claims that Jensen’s micro-level surveys and primary research are worthless by citing his own fieldwork which consisted of a small set of “informal conversations” as he puts it. There is very little in the Steyn paper to indicate that a rigorous evaluation of Jensen’s data or underlying assumptions has actually taken place.
A newer paper does not necessarily mean better evidence. I would hope you were more careful with what you are putting forth as a debunking. It’s disappointing to see this on what is normally a very high quality blog.
Jensen’s central claim is that fishermen adjust ports of call based on calls about fish market prices. Once that claim was proven false, what does it matter about the rest? I only included it to show just how flawed the original paper was.
There’s long been a need to take a good hard critical look at Jensen’s article in part because of its inordinate influence on both academic and practitioner.
Consider a critique of Jensen’s paper written by myself and Janaki Srinivasan… published in the ACM ICTD conference 2013 — http://dl.acm.org/citation.cfm?id=2516618 — free download here: https://markets.ischool.berkeley.edu/files/2014/09/p56-srinivasan.pdf
Our paper was peer reviewed in a double blind process, had three reviewers (2 reviewers and 1 meta-reviewer) and, in particular, was reviewed by an economist. Our larger point is that parsimonious models (including Jensen’s) are extremely problematic when used to draw conclusions about policy and technology design. I can’t vouch for Steyns paper but you will see that he cited our paper extensively. For many fishermen in Kerala, market prices are irrelevant.
Jensen’s account of the mechanisms underlying market efficiency gains in Kerala are slightly better than guesswork. We attempted to shore that up with qualitative empirical research (systematically interviewing all roles in the supply chain about mobile phone use broadly). And at any rate, economists know that these studies say little in terms of generalizability (talk to Angus Deacon).
Also am currently co-authoring a paper for Agricultural Economics with development economist Jenny Aker to point to many of the blindspots in economics in terms of thinking about technology’s “impact” on poverty in an agricultural contexts.
I am always surprised at who pushes back when closely held beliefs are questioned.
And that push back is the strength of good quality research. What bothers me here is not the questioning of Jensen’s conclusions, but the disregard for research rigor and the relative quality of the two papers.
Kyle and others – this is the stuff which “you know who” thrives on. You’ve swallowed his clickbait and wasted a precious half hour deigning this with a thoughtful response. This post is no less dramatic than many others and one person’s lack of interest in rigor and process, overshadowed by obnoxious opinion is pushed on readers through this bully pulpit paid for by …. ?
Ah, I always love the weak who hide behind anonymity instead the brave, like everyone else here who is willing to stand behind their comments. I may not agree with them, but at least they garner respect for being genuine.
I had not read Jensen’s paper before I saw Steyn’s. Going back and reading both, I found Jensen’s to be lacking and Steyn’s to be insightful. Subsequently, work like this has proven my conclusions valid.
While this critique makes interesting and important arguments about the explanation offered for the statistics observed, I was extremely surprised that the article didn’t actually address the central observation of Jensen’s paper–that price variance was reduced following the introduction of cell phones. The title of this article suggests that the statistics about reduced variance were overturned, EITHER because they were shown to be driven by something else OR because there was some error in data collection (as BJaques has pointed out, data collection was well documented). But the central statistical observation of the Jensen paper, about reduced variance in prices, was never addressed, so the point of this critique is kind of unclear to me.
Are you trying to say that the stats hold up but you don’t believe the mechanism story being told about them, or that the stats themselves are flawed? The former would be a great conversation to have, and could be the subject of further study to try to understand why the observed results were obtained. This kind of productive dialogue is what moves science forward! The latter critique would require a much deeper engagement with the data collection and statistical methods, none of which are mentioned here so I will assume that this is not your intention.
My concern, as the author themselves noted in the comments, is that when things are phrased bombastically people simply don’t want to consider the critique, even when it might have merit. That just pushes people into separate camps who won’t engage with one another. If the goal is actually to come to a better understanding of ICT usage, then it would be helpful to be clear what is the critique you are making, in a way that is commensurate to the new facts brought to the table (i.e. not implying the statistics were overturned when you don’t support that claim).
It’s pretty clear that fishermen didn’t use mobile phones to find which market to sell their fish at, which is the main point if Jensen’s paper. After that, it doesn’t matter if price variance increased or decreased, neither was due to mobile usage to find better ports of call.
Thanks to everyone who took the trouble to write a few notes and comments on this topic. I respond to comments in this single post.
Most of us active in the domain of ICT4D do so because we have a passion to enrich the lives of the so-called have-nots. Being activists and evangelists for change, we are sometimes blinded by hype. Personally, being involved in this arena for about 20 years, it seems to me most initiatives have very short-lived results, and that the actual impact is often less than claimed. If we wish to act responsibly in this domain, we need to critically scrutinize claims. Those of us who are scientists are obliged to bring critical scientific tools to the debates. For us it is not about who says what, but what is said — thus, not the people, but the ideas. We can fight about ideas, and then afterwards have a friendly drink. It is not about persons. My “attack” is certainly not aimed at Jensen, my fellow human being, but about the ideas he wrote. Unfortunately those ideas are widely cited in support of ICT4D projects; more unfortunately, those ideas achieved legendary status.
I welcome the discussions and debates. If we are serious about making a difference, we should investigate proposed ideas seriously. I also welcome criticism, which is valuable as it helps to sharpen our own positions.
Allow me to enter this debate, by responding to some remarks and questions. Don’t believe anything I say, I tell me students. But counter-criticism must be logical, provide counter-evidence, and be based on solid scientific argumentation. That evidence should stand the scrutiny of critical analysis.
To start off the discussion, also refer to discussions about this topic on the ICTWorks Facebook page. I draw your attention particularly to Jenna Burrell’s post. I hope she does not mind that I copy one of her posts here:
[start quote] Yes, I’ve been saying this for years. See journal pub “The Myth of Market Price Information” (https://markets.ischool.berkeley.edu/…/the-myth-of…/), see open access ICTD conference paper “Revisiting the Fishers of Kerala, India” (https://markets.ischool.berkeley.edu/…/revisiting…), see our group blog where we tackle this fallacy constantly – Beyond Market Prices (http://markets.ischool.berkeley.edu/) … but yeah, glad y’all are catching up on this [end quote]
***
Here, on the ICTWorks website, Anirban says regarding the Wireless Reach program that “…Fishermen find market prices of fish useful as they can negotiate better with the traders.” It would be helpful if Anirban could offer more details, such as what type of “fishermen” you refer to? Do you use the word generically, or for the chaps who draw the nets? And if for the chaps who draw the nest, do you mean the small boat fishermen (vallams) or trawlers? In which areas is this program implemented?
To Ben Jaques-Leslie: I certainly do not suggest any deliberate misleading on the part of Jensen. Oldness, newness has nothing to do with the debate. Some old literature suggest that the earth is flat, stands on pillars, etc. The merit of any academic paper is in the argumentation, methods, evidence, and so forth. Critical analysis of any proposed notions is what academia is about, including my own paper – and that will be welcomed, so that we can freshen up our ideas about ICT4D.
On my statement in the paper “Jensen does not state how the data on market prices was collected” – and the counter-statement by Ben Jaques-Leslie that Jensen does indeed state his method on pages 890-891: See Table 1 in Jensen’s paper (p882) – the note reads: “Data from the Kerala Fisherman Survey conducted by the author. The first column contains the average 7:45–8:00 A.M. price of sardines in each market on Tuesday, January 14, 1997, in rupees per kilogram.” This data is collected early in the morning. On page 890-891 there is reference to “in the afternoon”, and fishermen were surveyed. He does not state how he collected the early morning data.
How should we interpret this? Do the fishermen in the afternoon give the data of the prices of earlier in the morning? – if they remember the price and time… Were fishermen also surveyed early in the morning? If yes, why two sets of surveys? If not, what is the source of the early data? In Kerala there are government surveyors who visit markets randomly and record the prices fetched for each auctioned lot. Did Jensen perhaps use this data? He does not say.
Keep in mind that academic research papers are public, and responses are made in public. I would love to ask Robert to clarify this matter – but we would do so through the academic communication channels, i.e. journal papers. I can only respond to what he wrote.
Kyle Murphy says “Jensen himself calls that evidence anecdotal…” Indeed! But it is certainly not a particular scientific approach to use anecdotal claims as evidence for research, especially if those claims are the foundations of a paper. This is exactly the problem with Jensen’s paper: using anecdotal “evidence”, and then build his arguments on them.
About “…fieldwork which consisted of a small set of ‘informal conversations’….” Note that the wording is (I emphasise here in italics) “Informal conversational interview and emergent interview approaches..” These methods used are based on well-established practices in ethnographic and similar domains research – see Patton (2002), Glaser (2001), Zhang and Wildemuth (2009), DeWalt and DeWalt (2010). The words in italics are technical terms used in research methodologies, and should not be interpreted as informal, casual words. This research approach has been around for close to a hundred years, and used by anthropologists, psychologists and sociologists. The findings from such methods are as valid as the results of any formal survey. For the more technically minded: formal statistics is a positivist approach to data, while informal interview techniques are interpretivist. The problem with positivist approaches used to interpret complex human behavior is that they reduce complexities to singularities. If you take apart a car engine, and analyse only the carburettor, you will have a poor understanding of how an engine works without considering all the other aspects. Particular statistical analyses typically suffer from reductionism; while human behaviour is much more complex. In the case of fish markets in Kerala, they are much, much more complex than some literature suggests.
The critique is not only about interviews. A long list of counter-arguments are presented in the critique, each in its own right causing damage to Jensen’s claim.
Caitlin raises the point of “price variance was reduced following the introduction of cell phones”. This is the crux of confusing correlation with causation. That statement implies it is the use of cell phones that caused a reduction in price variation, which is one of the core problems with Jensen’s claim. There are several other possible causes for stabilisation. A more paradigmatic problem is the reduction of changes to the classic theory of monocausation. See Mario Bunge’s “Causality and modern science” for a critique of the oversimplification of a theory of causality. Had Jensen not linked price stabilization to mobile phones, but to a range of other possibilities, the critique would most likely not have been necessary.
On the statistics: Yes, the statistics is not the issue of the critique – any statistical analysis depends on data collected, the statistical instruments chosen, and most importantly, the assumptions made, and conclusions based on the analysis. The critique was not against the statistical method as much as against the data used as input, and the resulting conclusions.
Phew! that was a long response.
Finally, it is through respectful, open and critical debate that we might be able to make progress. Personally, I wish that mobiles would make an economic difference to the lives of the poor. But as far as I can determine, the value is personal – keeping in contact with friends and family, and for entertainment. Mobiles are indeed used to make business contact – so did paper letters, telexes and telegrams. Mobiles just make it easier. But mobiles causing economic growth? That is a very dubious claim, and I have yet to see the evidence….
Keep on chatting!
‘I have the best theories, I really do.’
In a British kind of way I have to start with an apology, so sorry, but I am going to add another comment criticising the sensationalist style of the headline and the article. Jensen’s article was a very early, serious attempt to apply some rigour to what was then, as now, an overhyped ICT4D sector (and I write with a history of hyping). Times move on, more research becomes available: thoughtful analyses like Steyn’s advance our understanding. It would have been good to use this well-read site to invite Jensen to comment on the Steyn article, have the two debate the issues. And maybe some recognition of the negative impacts that follow from shouty headlines would help get to that point.
Pete,
While you may be disappointed that I didn’t spend several hours tracking down Jensen and Steyn, convincing them to spend hours developing a debate around this paper, and then writing a sensible headline that no one would’ve clicked (Two Views on Market Pricing or Jensen vs. Steyn: An Academic Debate), I did achieve something greater with less effort.
You found this erstwhile blog and were so moved by the overall point, that you felt the need to comment. As was Ben Jaques-Leslie, Kyle Murphy, and many others here, who like you, had never commented before either. In fact, we’re now up to 21 comments, including a long detailed response by Steyn, creating that very debate you want to see.
If there is a downside, I dare say its inconsequential versus the years of hype and failed projects that relied on the flaws of the original paper.
Wayan
I responded because I found the language offensive and distracting. Calling something a lie equates to calling the person who wrote it a liar, accusing him of a deliberate act of deception. Using that kind of language is typical of the kind of online debate I don’t think you’d like to be associated with. Isn’t the acid test to consider if you would say the same thing face to face in a public setting, like, ‘Robert Jensen, you are a liar” during a Q&A in a conference, rather than asking him to comment on the critiques of his 2007 article.
And it’s not just niceties: using that kind of language in a well-read blog normalises it. At what point then would you censor offensive comments, making personal attacks on you?
For your headline there could have been any number of alternative phrasings between the bland and the sensationalised which could have attracted attention to the issue.
If I were to debate this in person, I would’ve said, “that’s bullshit” so my alternate title would be, “Calling Bullshit on Fishermen Benefitting From Market Prices.” Not sure you’d think that’s better language usage..
The greater point is that sometimes we need to use strong language to get a point across. The fisherman market price theory is so deeply entrenched in ICT4D that any less of a dramatic counterpoint would’ve been lost in the usual chaos of life.
This is an important debate and the article’s headline an effective way to launch it. Quite apart from the challenge to the integrity of Jensen’s research methods, there is much fishy business here to explore.
Our experience in Trinidad is that market prices has been a highly sought feature for users of a mobile application suite for fishers. It has given them a sense of confidence and introduced new pathways to negotiation with vendors with whom they have traditionally had no sway. The use of this simple service has been an entry point to basic instrumental as well as informational services. These are important outcomes of themselves. However, we have as yet found no evidence of economic impact and the market Prices facility has not enabled them to choose markets to sell their catch. Instead what we found was that the dynamics of the markets in the small island states in which the application suite has been used are characterized by interdependencies between agents that present a practical barrier to market choice.
We have found in some islands such as Trinidad and Tobago that the most pressing concern of fishers is safety at sea. While the mobile phone is limited by coverage range, these fishers do not generally use vhf radios despite receiving them as donations through various initiatives. Ham radio use, which requires a license based on the passing of written examinations, has virtually no subscription amongst local fishers. Perhaps vhf and ham can work in India but even if that is the case, they may be entirely inappropriate for other jurisdictions.
Small scale fisheries occupies a very complicated space and I would exert much caution to anyone, outside of embedded context, who prescribes interventions based on anecdotal or legitimate claims for solutions.