On December 15-16, 2007, the second International Conference on Information and Communication Technology and Development (ICTD) was held in Bangalore, India. This explicitly multi-disciplinary conference aimed to bring together academic researchers from fields including computer science, electrical engineering, information systems, political science, urban planning, sociology, anthropology, economics, industrial design, and the private sector to discuss original research on ICT applied to development. More than 200 individuals from countries including India, the Philippines, Canada, the United States, Pakistan, Chile, the Netherlands, and the United Kingdom attended the two-day event.  

In this conference report I consider five themes that emerged from the conference[i]. First, two broad development themes:  

A)   The effect of occupation changes on poverty and the role for market-based solutions to development goals.

B)   Second, three ICTD-specific themes:

1. The relationship between micro and macro-level analyses in ICT for development project design,

2. The meaning of ‘development’ in ICTD, and

3. The need for meso-level ICTD research.

The conference was divided into eight panel sessions and a poster session, in addition to two keynote speeches. The keynotes played an important role in framing the ICT-focused topics of the individual papers in broader perspectives of development. Anirudh Krishna of Duke University began the proceedings with a discussion of his work on the flows of poverty in both developed and developing countries. Based on an interview and survey data on poverty within eight communities in India, Kenya, Peru, Uganda, the United States, he showed that despite overall statistics, which highlight small gains or losses in the fight against poverty, many of these communities have experienced major shifts in poverty. Through the analysis he found that poverty in these communities was largely not constant and instead many families were moving into or out of poverty, such that in one case there might be twenty percent of families moving out of family, but eighteen percent of families, who were previously above the poverty line, have moving into poverty, resulting in a net gain of only two percent. In addition, he found that matters related to health care are often major causes of poverty, with low-income families unable to cope with large and unexpected medical costs. Finally he found that small improvements in family members’ occupations, rather than major income shifts, were the most common mode for exiting poverty.

The closing keynote, offered by long-time development activist Paul Polak, emphasized a market-based approach to micro-level development. He highlighted the work of the development organization he has run for more than 25 years, which develops products in local communities and then markets them broadly at reasonably low costs. Examples included treadle pumps and low-cost drip systems for irrigation. Members of the audience questioned how this market-based approach fit into broader development models and whether he was advocating a development strategy that excluded government participation. In response Polak noted that this was not an anti-government model, but one that attempted to make improvements at the local level that could complement any broader initiatives.

In each of these cases, prescriptions derived from the keynotes focused largely on micro-level opportunities for poverty reduction. From the perspective of an ICTD practitioner, this might imply an emphasis on local initiatives working with individual groups to find ways in which ICTs might enable small improvements in development, rather than the major jumps often promoted in the early days of ICTD enthusiasm. This attention to micro-level opportunities was clearly present, and probably dominant, in the papers presented at the conference. At the same time, another recurrent theme in the papers and, more frequently, the question-answer sessions was the potential disconnect between micro-level initiatives and broader perspectives on society and political economy. How do, or can, initiatives focused on the micro-level maintain an awareness of the broader macro-issues that may play an important role in whether a project fails or succeeds? A village wireless network may have many local benefits, but if it unsustainable because of connectivity costs or legal restrictions on wireless transmissions, then what is the long-term value of the project? I return to this issue in discussion of individual papers below.

Another important theme that emerged from the discussions concerned conference participants’ and practitioners’ understanding of ‘development’ and how differing expectations of donors, beneficiaries, and analysts can lead to unintended consequences in ICT for development initiatives. This is linked to the question of micro- versus macro-perspectives, but draws greater attention to the specific motivations of individuals and organizations involved in a given project and how these motivations may conflict even when they all fall within the broader ICT for development agenda. This may be a question of beneficiaries wanting better access to water, rather than access to computers, or it may be differing views among donors and users regarding the specific purposes for which computers should be used. In either case, these differences, if not addressed explicitly within a project and with any relevant stakeholders, could lead to failure in what might otherwise have been a productive initiative.

These broad themes, in addition to other panel-specific topics, are understood best in the context of the individual papers. The first set of papers shared a focus on the design of applications for use in environments outside the standard Western model. This included discussion of an electronic mechanism for helping to teach blind students how to write using a low cost Braille writing tutor tool. A second paper showcased an initiative to develop a robust sensor network for predicting evacuation conditions in flood-prone areas of Honduras. The final paper considered a video-based agricultural extension intervention used to improve uptake of improved farming practices in rural India. To a certain degree, each of these cases involved an important interaction between the technology designers and their target audience as a part of the design process. In the agricultural extension case, farmers became an integral part of the technology innovation itself, as the project team discovered that members of the public viewing the intervention videos were more likely to take up the recommended practice if the person in the video was one of their peers. While the concept of incorporating users into the design process is not a new one, these projects highlight to varying degrees the way in which this model has been incorporated into ICTD initiatives.

These three initiatives differed, however, in their attention to broader political issues of the project environment. While the flood warning project incorporated government users into their processes, at least in part because government officials have decision-making power over whether or not call for evacuation, the farmer education project chose to see its initiative as a “digital pipe” that was not in itself political. This technological agnosticism belies the fact that the project, like many ICT for development initiatives, was implemented in partnership with an NGO, which has a particular philosophical and political approach to agricultural intervention. While their sustainable agriculture model may have many merits, the choice to pursue this technology initiative with an NGO that sees itself has having a significantly different approach to agriculture than the Indian government is a choice with political ramifications. Efforts to expand this initiative in other parts of India could then face political obstacles that were not accounted for in the pilot.

The second panel had only one paper due to last minute changes, but the topic, international flows of e-waste, was quite an important one from environmental and political economic perspectives. Indeed, the most recent issue of National Geographic has a major story on the detrimental effects of e-waste in many developing countries. The paper highlighted the importance of regulatory regimes and enforcement of regulations to support more sustainable practices in disposal and recycling the millions of “obsolete” computers and electronics produced each year. 

The majority of the first afternoon was spent in a poster session where twenty poster presenters were given an opportunity to discuss their work with participants. Given the intimacy of the conference and the small number of posters this was a nice opportunity for presenters to receive thoughtful comments and questions from a reasonably large number of people. The poster topics ranged from local wireless networks and ICT in agricultural production chains to the political economic characteristics of telecenter models and the effects of regional politics on telecenter distribution.  

The final set of papers on the first day considered examples of the telecenter model as a tool of development. Drawing on examples of projects from Zambia, Chile and the state of Kerala in India, the presenters drew a distinction between the expectations of project planners and eventual project outcomes. In comparing telecenter initiatives in Chile and Kerala, the first presenter argued that these projects, which were initially targeted to poor communities, are actually serving what the authors call an “emerging middle class” of women. While these women are benefiting from the centers, this is a different social outcome than was intended by the projects’ initial goals. Another analysis of the Kerala initiative showed that despite attempts to build “e-literacy” with subsidized training programs, many individuals did not take or finish the training course, either because they did not see the relevance of the training in their lives or they were uncomfortable with the location of the telecenter. The third paper presented a connectivity project in a rural Zambian village, in which the goals of the project were broader than implementing telecenters. In this case the introduction of an Internet café was complemented by other technology applications intended to support health, education and community development. While the initial pilot was deemed successful, planned expansion to other villages is likely to be constrained by the continued high cost of bandwidth in many Sub-Saharan African countries. Each of these cases highlighted the broad theme of stakeholder conflict and show that while donors may have specific goals for ICTD initiatives, in many ways the power of ICTs is in their potential to be used in user-defined ways. If these goals are in conflict, then the outcome may be unexpected negative effects on the project itself.  

Day one concluded with a broad discussion of ICTD research by a panel of representatives from academia, NGOs, government, and the private sector. The goal of the panel was to address a difficult aspect of ICTD research, which is the development of theory in a largely practical domain. The panel chair, Professor Annalee Saxenian, noted that most ICTD research has been at the individual level of single projects or cases, and little work has been done to compare across projects and attempt to develop middle- or meso-level theories. The first speaker, founding member of the IT for Change NGO Anita Gurumurthy, built on these initial comments by emphasizing the need to problemitize ‘development’ in the ICTD space in order to be more aware of the assumptions behind research in this area. A more self-conscious perspective by researchers should help to improve the rigor of research in this field in addition to increasing practitioners’ understanding that ICTD is in many ways a study of power, where power is located and the ways in which it is dislocated by the introduction of new technologies. Richard Heeks, a professor from the University of Manchester, echoed Saxenian’s call for mid-level theory and argued that ICTD researchers should build on the decades of previous research on technology and society, rather than starting from scratch with each research project. Karishma Kiri of Microsoft, who gave an industry perspective, noted that research must take a holistic approach, which considers all of the technical, economic, and human resources issues that go into creating a successful ICT initiative in developing countries. The last speaker was Pankaj Agrawala, an Indian Administrative Services officer from the Indian state of Uttar Pradesh, who has been a key proponent of ICTs in India over the last decade. He gave the most explicitly practical recommendation, which was that researchers should focus on language applications, in order to extend the reach of new information technologies beyond the English-speaking public.  

The second day kicked off with a set of papers considering “alternatives to real-time Internet.” These efforts responded to a general emphasis in the ICTD field on improving connectivity by providing interesting perspectives on the potential for connectivity options other than always-on broadband connections. Applications ranged from voice recorders that wirelessly transmit women’s comments and contributions to a Kenyan community radio station to the use of mobile phones for farmers to collect and submit agricultural data in rural India. In each case the projects took into consideration the specific resources and requirements of the local area in order to develop practical and useful tools for rural environments.  

The next panel took a much broader view of ICT and development, focusing on some of the meta-level issues that were given less consideration in other discussions during the weekend. The first paper developed a revised model of economic growth and IT investments, showing that when IT infrastructure is included in cross-national analyses there is a statistically significant and positive relationship between increased diffusion of IT and increased economic growth. This model contrasts with earlier analyses that only found a significant positive relationship in developed countries and not in developing countries. The next paper problemitize the underlying educational model of the One Laptop per Child initiative, showing that the design guidelines of the OLPC product to date seem based on the assumption of a single learning model, rather than a model which assumes multiple possible modes of learning across children. The final paper brought out a thought-provoking distinction between goals of welfare and agency in ICTD initiatives. Through a formal model and analysis of two telecenter initiatives, the authors showed explicitly the potential conflict between the goals of donors and users of ICTD projects. While donors may want to improve the welfare of aid recipients, the recipients themselves may have alternative desires for using telecenters than those promoted by donors. The authors provided examples such as users who want to view entertainment websites rather than collecting agricultural information, which may conflict with donor goals. But they ask if this entertainment increases users’ perceptions of overall welfare, then does it not actually serve donors’ overall goals? This conflict between the agency of aid recipients and the perception of welfare goals by donors can then have important outcomes for the success or failure of telecenter projects, which may not serve the needs of local users if this potential conflict cannot be resolved in the design or redesign process. This discussion highlights the broader issue of donor goals in development projects and how to design initiatives that are not constrained by stakeholders’ differing objectives.  

The final panel provided two perspectives on classroom-based computing in India. The first paper offered a broad perspective on usage models, from single-user, single-computer to multiple-user, shared computing environments. The authors argue that these models should be evaluated in terms of their overall costs, socio-cultural suitability, and educational effectiveness. As was noted in the earlier analysis of the OLPC initiative, no single model works best in all situations. The second paper presented findings from interviews with parents in four Indian states regarding their opinions on the use of computers in schools. The results highlighted a broad range of opinions, often linked to region and profession, with some parents seeing computers as their children’s ticket out of the village while others felt that the investment in computers could be better spent elsewhere.  

Overall the conference was well organized, enjoyable, and offered a wealth of information on individual initiatives combined with broader perspectives that should help the ICTD community continue to improve efforts on the ground and the analysis of these efforts in broader socio-economic and political contexts. At the next ICTD I expect that we will see even greater awareness by practitioners of their own perspectives on development and how this feeds into the specific goals and outcomes of projects, in addition to the perspectives of other stakeholders who in many cases play an important role in the ultimate success or failure of these initiatives.