project home

On Jensen and his sources

More information and interpretation of the Jensen (2007) article is provided here.

Landing Centres

Jensen collected data from 15 landing centres. Cities and large towns have several landing centres, but Jensen only mentions the city name, so it is unknown which particular landing cites were visited.

Kasaragod District

Kannur District

Kozhikode District

Newspaper articles on ICT4D

Here is more about the sources Jensen (2007) used for his claim that mobile phones are used in the developing world for economic purposes. It is extremely unfortunate that Jensen obtained ICT4D information from mass media sources, and not from academic research in the field. This mistake lead him to accept weak, and false data, as the basis for his argument. Because the data cannot be trusted, Jensen's conclusions cannot be trusted.

Alam (2005)

Alam's article is about how mobile phones makes communication easier for the poor. With such an angle there is no problem. But then he continues with conclusions and implications for which no evidence is offered. Alam claims “mobile phone technology is revolutionizing Hussein's life for the better, enabling him to cut out cheating middlemen and deal directly with buyers from district markets.” without offering any evidence for this claim, while Jensen piggybacked on this claim. It is the kind of claim created by mobile marketers, as it is commonly found in such literature, and also promoted from public podia where mobile technology evangelists have are in the spotlight.

Properly researched evidence from the field that mobiles are used for such purposes is scant. That is, for research conducted following scientific methodologies and independent from those of the marketers of mobile technologies. Despite firsthand experience in the field of ICT4D over the past more than 15 years, attending too many conferences I can remember, and reading scientific literature in the field, I have never come across any solid evidence for such claims. In fact, the lack of solid scientific research is shocking.

Alam's is a newspaper article, so he can be forgiven for not offering proper references or sources, and for twisting realities for the sake of sensation, on which mass media thrive. But that is not science. A critical reader can easily spot the weaknesses in the article. For example, Alam writes: “In China, too, poor farmers have benefited from the spread of mobile phones through text messages giving them information about prices, weather forecasts and pest control. Ministry of Information figures released in January showed the number of Chinese subscribers topped 334 million in 2004, up 65 million on the previous year.”

Common sense indeed allows us to conclude that mobile phones benefit, among others, farmers, and also that they might use text messages. But to jump from this to farmers using text messages to obtain prices is too large a pill to swallow. How does it follow? Where is the evidence? An example? Worse though, is that the conjunction of those two sentences is unfortunate as the reference to the Ministry of Information subtly implies that the growth and farmer usage are conceptually linked, which they are not.

No source is given for the claim that Chinese farmers use their phones for the purposes Alam states. The logic of the two sentences quoted is: farmers use phones; mobile phone sales grew by 65 million, therefore farmers must use mobiles to determine market prices. It is common in media to make such illogical links between statements. Jensen too fell into that trap. He probably got the logic from Alam. Jensen notices increasing sales in mobiles in Kerala. During that period markets stabilized. Hence, mobiles cause the stabilization: cum hoc ergo propter hoc.

Alam's other example points to the non-economic use of mobile phones. The story of the South African taxi driver, with the name Mazantsana, shows that the poor use phones for personal communications, not for business. Whereas in the past one had to write a paper letter, taking weeks to get a reply, the phone is a much quicker method. Landlines would offer the same convenience, except that landline infrastructure in South Africa, and other poor countries, is primitive, unreliable and costly. A mobile phone has exactly the same function as a landline phone except that mobile costs are relatively lower, thus more affordable, and also much more convenient. Of course, fancy mobile phones can do a lot more than handle calls, but our own observations and experience with the world of the poor is that by far the most important function a mobile phone is used for is for personal communications, not for economic activity. This is even true in highly developed regions, such as the USA (see Steyn and Das 2014 - names withheld during review process).

Arnold (2001)

Arnold refers to mobile network operators who claim: “In Thailand, one operator found shrimp farmers calling Bangkok to get prices -- and the upper hand on middlemen. Similarly, Smart Communications, a Philippine cellular operator, learned that vegetable farmers who had never had regular fixed-line phone service were using mobile phones to market their produce in Manila.”

The Arnold article is based on media releases of mobile phone operators. These evangelists must keep shareholders happy, and will twist facts to get what they want.

Arnold says: “In theory, the laws of supply and demand should drive the wireless industry to develop so that it can deliver some version of its product to anyone willing to pay for it.” Theoretically yes, but logistically, no. About 60% of the population of Africa are subsistence farmers. Their markets are in survival mode, not driven by theoretical constructs such as supply and demand. The same is true of many poor in other regions on this planet.

The poor, and rural, also suffer from logistics nightmares. The cost to get their produce or goods anywhere outside their neighborhood or village is prohibitive. Very few products make it out of their area, and no matter what communication technology, or the widest possible broadband, it is not going to change the business model for them. Economies of scale apply not only to the economic activity itself, but also to logistics.

LaFranier (2005)

LaFraniere's article is about the growth of the mobile phone market. Some examples are given about mobiles used for commerce.

However, the example of a fish vendor along the Congo River comes from a Vodacom (a mobile phone company) spokesperson, not from scientific research. Mobile phone companies in particular are known for inflating numbers to keep shareholders happy. For example, the nonsense of huge market penetration, even more than 100% per population in some countries is an indication of what such hype does to the realities. Mobile networks report sales based on SIM card sales. Even in poor countries many citizens have more than one SIM card. But there are still many individuals who do not have phones in the poor regions. Such inflated figures make a mockery of common sense, yet too many people, including academics, fall into the trap of believing the glitter of promise offered by marketers.

Rai (2001)

Rai, writing in prosaic style, claims that mobile phone ring tones interrupt the drone of motorized fishing boats in southern India. She has very obviously never visited a beach market, where the noise of the engines drowns the ring of mobile phones. And having done some fieldwork along the Kerala coastline in 2013, we have not come across constantly ringing phones. Of course phones sometimes do ring, as they do anywhere in the world where mobile phones are used.

But to jump from these phones interrupting engine noise to “...rural India has discovered the convenience of doing business on mobile phones.“ is an exceptionally huge jump. This is idealized journalism, not scientific research. Rai mentions 5000 fishermen who use mobile phones. One wonders how this number was obtained, as no source is given. Rai wrote in 2001, and as reported by Marine Fisheries Census 2010. Kerala, by 2010 a total of 36,906 fishermen in Kerala used mobile phones. This means that if Rai'' figure of 5000 for 2001 was correct, the annual growth from 2001 to 2010 was a measly 3545 new phones sold to fishermen every year. Not a particularly spectacular growth, if one considers that there are 145,396 active fishermen in Kerala.

Moreover, even if 5000 fishermen indeed used mobile phones, where is the evidence that they used their phones for business purposes?

Rai concludes with this: ''Life without a mobile phone,'' said Mr. Karthikeyan, the Cochin fisherman, ''is unthinkable.'' Indeed! But it is not clear what purpose Mr. Karthikeyan had in mind when uttering this sentence. It is quite possible that prominent in his mind was personal communication. It is a huge logical jump from this sentence to conclude that mobiles are commonly used for business purposes,and a magically huge jump to lead to determining market prices.

* * *

What does “doing business” mean? A business process consists of many components, communication perhaps the most crucial. Obtaining information is not yet completing a transaction. Market price is indeed one component of business communication, and it goes without saying that mobile phones might be used to source such information. But in 2001 all this was hopeful speculation. At that stage no research had been done to determine what happens on the ground. Although there are publications that claim that mobiles are used to determine market prices, there are plenty projects that suggest that the dominant use of mobiles is for personal communication, not for business purposes. When scrutinized, some claims about mobiles used to determine market prices can easily be refuted. It seems that a myth, originally created by mobile phone marketing hype mislead even serious academics into blindly assuming this must be a fact.

Correlation and causation

A correlation between two data sets certainly does not imply causation. However, Jensen made this mistake. Based on a correlation between the data sets of mobile phone growth, and market price increase and stabilization, he concluded that the use of mobiles was the cause.

The website of Tyler Vigen, Spurious Correlations jokingly lists many examples of unrelated matters that indicate a strong correlation (http://www.tylervigen.com/).

Here is just one tongue in the cheek example indicating that a correlation in no ways implies causation:

Per capita consumption of cheese (US) correlates with Number of people who died by becoming tangled in their bedsheets

Spending on science, space and technology also correlates with the number of suicides by hanging, strangulation and suffocation. Science is thus not good for you!

References

Alam 2005 Alam, Shafiq, “Mobile Rings Changes for World’s Poor,” Sydney Morning Herald, April 22, 2005.

Arnold 2001 “Cell Phones for the World’s Poor,” The New York Times, January 19, 2001, C1.

Jensen R (2007) The Digital Provide: Information (Technology), Market Performance, And Welfare In The South Indian Fisheries Sector. Quarterly Journal Of Economics, CXXII(3): 879-924.

LaFranier 2005 LaFraniere, Sharon, “Cellphones Catapult Rural Africa to 21st Century,” The New York Times, August 25, 2005, A1.

Marine Fisheries Census 2010. Kerala. Central Marine Fisheries Research Institute, Kochi. Part II: 6.

Rai 2001 Rai, Saritha, “In Rural India, a Passage to Wirelessness,” The New York Times, August 4, 2001, C1.

© Jacques Steyn 2014