Notes on Sustainability

Kenneth Keniston

            The conjunction of the “information age” with the idea of “development” has brought an apparently irresistible world-wide push toward “closing the digital divide”, or, as it is now known, “ICT4D.”  Involved here is the union of the newer electronic information and communication technologies, chiefly computers, email and internet-web, on the one hand, with the idea of development through technological progress and innovation on the other hand.  Joined together, these two ideas have proved persuasive to virtually every international and global agency, to almost every national and state government, to small and large nongovernmental organizations and to new organizations and affluent individuals.  The idea of ICT4D has been made more compelling by the recent argument in prestigious journals like the Harvard Business Review that properly-designed projects for the “poorest of the poor” cannot only help the poor but can yield a profit for the entrepreneurs or capitalists who are designing them.  Thus, ICT4D combines three of the most compelling ideas (or ideologies) of the 21st century – that of the information age, that of development, and that of triumphant capitalism[1].


Translated into concrete projects, ICT4D has meant, over the last five or six years, two major kinds of commitments.  For one, virtually every nation wishes to replicate the success of India in developing a vibrant software industry, an enterprise that adds to the Indian balance of payments and--perhaps more important symbolically--demonstrates that Indian scientists, engineers, and entrepreneurs are as good as the best in the world.

            More relevant to my discussion here is a second meaning of “ICT4D.”  This is the attempt to use modern information and communication technologies to improve the condition of ordinary people in developing nations.  The core image is the infokiosk[2] at the center of an impoverished village, providing information services and utilities for local citizens, and resulting in poverty alleviation, improved health, diminished corruption, speedier access to government, appropriate and useful information about agricultural inputs, accurate pricing of the sale of local produce, and so on.  The infokiosk model, of course, has many variations just as it has many names:  above all, and especially in India, it has many many incarnations.  In my own research, often based on a number of months of field research, I have been able to visit more than 25 sites that use ICTs for the benefit of ordinary people, ranging from the sugar refinery at Warananagar to the municipal centers of E-Seva in Hyderabad, from the Drishtree infokiosks in Haryana to the SARI entrepreneurial info centers in the Madurai-Melur area. 

            So attractive is this concept that literally billions of dollars have been committed to its implementation.  Perhaps the most dramatic single commitment was that of the G8 in Okinawa several years ago, which committed more than one billion dollars to ICT4D development.  But since then, every agency of the United Nations, plus the World Bank, the International Monetary Fund and other international agencies, plus the development aid governmental groups of countries like Canada, the Netherlands, and the US, plus national, state and local governments in many countries (not least of all India), and finally, plus major international non-governmental organizations and new organizations designed specifically for the purpose of producing “digital dividends”—all of these have committed large resources to “ICT4D.”

            The initial ICT4D projects were essentially “demonstration” or “proof of concept” projects: they were heavily funded by outside agencies (Government of India, Development Alternatives, IDRC, etc.) and their initial goals were primarily to test, explore, and define the idea that information technologies could be of constructive use to local people, and especially to people in the weaker sections of society.  A few projects were organized around existing commercial operations—e.g., the Warananagar Project financed chiefly by the Governments of India and Maharashtra and organized around the Warananagar sugar cooperative in Maharashtra.  Others were taken as proof of concept projects by externally-funded NGOs like the M.S. Swaminathan villages outside of  Pondicherry.  And some others were initiated by enterprising civil servants to provide government services, records, and licenses to citizens within highly disadvantaged and impoverished sectors like Dhar, where the Gyandoot project for a time flourished.  Each project had some unique feature.  For example, the HoneyBee-Drishtree Project in Ahmedabad aimed above all at using ICTs for recognizing and disseminating local Gujarati wisdom, technology, and knowledge to a wider international group, all the while providing adequate rewards for local village inventors.

            In these early stages of exploration and demonstration, issues of cost and sustainability were often secondary.  Donors, governments, and NGOs were above all interested in ascertaining whether or not these projects “made a difference.”  Precisely what “making a difference” meant varied from project to project, and indeed, was often not clearly defined.  Above all, the central question of the entire ICT4D movement was often glossed over--namely, the question of whether funds spent on modern information technologies in developing countries yielded better results than the same funds spent for other, perhaps more traditional, purposes.  But in all fairness, without a “proof of concept” or set of demonstrations, there was no way of answering this question.  That is, unless, or until, there were a number of trials which demonstrated whether or not ICTs, properly deployed, could indeed alleviate poverty, improve health, provide work, increase agricultural productivity, diminish corruption, increase the responsiveness of government, etc., there was no way of comparing these ICT expenditures with other alternatives. 

            Thus, it was perhaps reasonable, in the first few years of what we are calling the ICT4D movement, that questions of sustainability and scalability were either not mentioned or were deferred to a later point.  To be sure, some projects like the Gyandoot Project in Dhar were from the beginning built on a revenue model which enabled local operators to charge for their services, and which, at best, produced for them a reasonable income--including even repaying the loans they had taken out in order to purchase their basic equipment.  But most projects did not attempt even this rather limited concept of payback, but were instead focused on how to achieve maximum effectiveness with the local population.  Be it noted, however, that even the judgment of “effectiveness” was most often dealt with more by anecdotes and casual visitors who wrote glowing newspaper accounts than through attempts at systematic evaluation. 

            Today, India has at least 50 ICT4D projects aimed at improving the lot of ordinary citizens.  There is no sign of any diminution in the enthusiasm for these projects or in the commitment to funding them.  On the contrary, journalistic and other accounts continue to be almost uniformly positive and almost uncritically enthusiastic.  New and old organizations are planning extensions of what were once pilot projects to larger and larger areas, to more communities, to additional regions and states.

My guess, based on following the worldwide literature on this topic, is that in India alone, there are more ICT4D projects than in all the rest of the developing world together.  Be that as it may, India certainly holds the national record for innovative and expansive projects in this area.  It is thus an extraordinarily useful test bed to evaluate both the effectiveness and the sustainability or scalability of these projects.

            In the comments to follow I will address primarily the issue of sustainability and scalability, and in rather narrow economic terms.  As preface, however, I should note that “sustainability” is not simply a matter of economics:  it also means rootedness in local communities; it means cultural and political acceptance; it means value to individuals.  All of these are as important, and perhaps more important, than “economic scalability.”  Moreover, I am well aware that, if the advantages of ICT4D turn out to be as enormous as sometimes hoped, ICT4D may well be defined as a public utility rather like the school system, such that issues of payback, scalability, etc. defined in narrow financial terms become irrelevant.  If so, it is responsibility of public authorities, presumably using revenues from taxes, to provide each community its own appropriate ICT facility.  That day, however, is still far off in India, especially when the primary school system itself urgently needs additional resources.

            In reading the meager literature (apart from journalistic and public relations accounts) on ICT projects for the weaker sections of society, one is struck with the almost complete absence of any adequate or realistic definition of sustainability.  Indeed the term is rarely even mentioned.  A few projects do mention initial capital costs:  e.g., the Warananagar project involved stated an initial investment of 2.5 crores, mostly from the Government of India and the Government of Maharashtra.  Although the precise figures for the Tarahaat project are not, to my knowledge, published, a careful reading of the official reports suggests that the value of donations, labor, and installation of the VSATs involved in this project must have run, in dollar terms, to several hundred thousand dollars.  In every project, as noted below, there are background costs of programming, installation of equipment in sometimes climatically and geographically hostile regions, maintenance, and the like, which usually go unmentioned. 

            The importance of sustainability is nonetheless obvious.  In the absence of continuing, guaranteed and massive support from the private or public purse, ICT projects for ordinary people must somehow find the means to recover their initial costs, maintain themselves, and finally to replicate themselves in new settings.  If we imagine that ICT4D is a potential benefit to all Indians, and if we wish to extend that benefit to the entire population, (e.g., to every one of India’s 600,000 villages), then initial costs, cost recovery, and “scalability” are obviously essential.  NGO money may provide the first few demonstration projects; but soon after that, as we know, NGOs tend to withdraw their support and expect self-financing. 

In what follows, I will outline some of the factors that should be considered in evaluating the sustainability and scalability of ICT4D projects.  I do this in the hope that, increasingly, projects will become more explicit about their expenses and revenues, so that informed judgments can be made about the steps necessary to ensure the scalability of projects whose benefits have been demonstrated. Incidentally, however, I note the need for many projects in India and elsewhere also to demonstrate more clearly their “benefits”:  isolated anecdotes about individuals who benefit are only a beginning of a careful evaluation of the benefits of an entire project, as it relates to its cost.  But in what follows, I focus almost entirely on some of the components of “costs” and “revenues.”

I.  The Costs of ICT4D Projects

In the comments that follow, I will summarize nine distinct but related areas that bear on the issue of sustainability.  My viewpoint will be that of a self-interested investor who is asked to provide funds for a series of new infokiosks to be located in a rural area in India.  As an investor, I have many choices with what to do with my money, and I must decide whether investment in the ICT4D project will, on the whole, be profitable.  (I would also, of course, be considering whether other investments might be even more profitable, but that consideration will be omitted in the discussion that follows.)  Put bluntly: my concern is financial sustainability defined as, in the long run, profitability.  It follows, by definition, that if the project is not profitable, it will lose money and cannot possibly be financially sustainable.

The areas I will discuss are:

1.      Costs of leadership, planning, and pre-studies

2.      The opportunity cost of funds invested over time

3.      The amount of capital required for initial equipment and basic software, i.e., initial equipment costs. 

4.      The costs of the specialized, localized programming,--i.e., the “applications” required to begin the project and make it relevant and useful to its potential users

5.      Personnel training and staffing costs

6.      Initial and ongoing maintenance costs

7.      Operating costs per unit time

8.      Depreciation on initial capital investment and subsequent expenses

9.      Shut down costs and capital recovery in case of project non-sustainability


Let me now discuss each of these considerations at somewhat greater length.

1.  Leadership, Planning, and Pre-Studies

The most publicized and successful projects in India are characterized by outstanding leadership, usually on the part of unusually dedicated individuals with a thorough knowledge of the region in which the project is installed.  “I began by spending a lot of time in the villages,” is heard from IAS officers, project directors from NGOs, and officers of private companies.  Such men (unfortunately few women are as yet involved in the planning of major projects), while they rarely receive international-level salaries, are usually well regarded by their organizations and paid well by Indian standards.  The “best” projects involve protracted periods of planning, working with local communities, pre-studies, site visits, etc.  In short, much valuable time on the part of the leaders--and often their assistants--is involved. 

Among the initial costs of any project, then, are the salaries of those involved in the planning, the percentage of their time involved, and the total amount of time involved, taken as a percentage of their salary.  In many cases, this would be a considerable sum.  In no case that I know of is this sum mentioned among the costs of the project. 

            It might be thought that once a project is begun, it could be repeated indefinitely, such that the leadership and planning cost would be incurred once and only once for the first project in a set of infokiosks.  In fact, however, the advice that emerges from existing projects indicates that in each new situation, extensive new local consultation, involvement, and discussion is involved.  Often the success of the project depends, for example, on the ability to mobilize the community in new ways before even beginning to install an infokiosk.  To be sure, projects may not have to be planned* completely* from the start.  But in every case, there must be a period of planning, local consultation, education or mobilization of local users, advertising, pre-study, and so on.  Thus, initial leadership and planning costs will exist in all projects that are likely to succeed.

2.  Opportunity Costs of the Funds Invested

If I am a wiley investor, I will realize that, for example, one crore invested in one ICT4D project could also be invested in the bond market, and that it would yield a defined rate of interest.  By investing in the ICT4D I forego this interest.  The amount foregone is the opportunity cost of my investment.  It needs to be calculated not only for the first year, but for the entire period for which my funds are tied up in the ICT4D project.  Such calculations are likely to be difficult and involve much guess work.  Will the project last five years?  ten years?  Forever?  Let us assume five years, and let us assume that the interest rate on a five year bond in India is 8%.  By investing rupees ten million in the ICT4D project, I forego compounded 8% a year for five years, in short, more than 40% of my initial investment, or about 4,000,000 rupees.  I also need to estimate the likely rate of inflation.  If I assume 4% a year, I will have an opportunity cost of 5%-4% or 7,000,000 rupees over five years.

These are important costs, but the opportunity costs of funds invested are to my knowledge never included in calculations of sustainability.

3.  Initial Capital Investment in Equipment and Other Tangible Materials

Virtually every ICT4D project in India and elsewhere mentions the start up cost of equipment, and many have made strenuous efforts to reduce these costs.  These costs vary from place to place, and from situation to situation, depending in part on the nature of the installation, and in part on the quality of the existing infrastructure.  It is essential to specify the requirements of the village infokiosk:  e.g., a computer, a printer, battery back up, usually a solar powered supply because of erratic (or non-existent) electric supplies, grounding, switching, and so on.  If a VSAT is required for reliability or because of the absence of trustworthy telephone lines, the costs of the VSAT, including installation, further solar power, grounding and wiring must be added.  In some parts of India, the modification of existing telephone switches is required so that they will accept data in addition to voice.  If an intranet is envisioned, the costs of cabling and wiring should be included.  The point here is that the “computer set up” (computer, printer, battery backup), may only be one part of the total equipment requirements. 

Furthermore, initial set up investments require core software, i.e., an operating system, which will usually be either an MS product or a Linux system, plus installation and testing of this basic operating system.  These, too, cost money.  (more)

4.  Costs of Local Software Development

  More important and usually more costly is the development of specific applications that respond to the needs of the local community and are available in the local language. Local language software in India, as in many other developing countries, is usually either missing or inadequate.  For example, one friend in a northern Indian state reports that he, an experienced software engineer, along with an equally experienced colleague, spent the first three months of a local infokiosk project in refining the coding for the local language--since the coding system provided by the Government of India agency involved proved to be “buggy,” expensive, and inadequate. 

This basic but necessary work developing a usuable local language coding scheme must often precede the even more important work of developing specific applications that enable the infokiosk to respond (in the local language) to queries by local citizens on such matters as health, government certificates, agricultural inputs, mandi prices, matrimonial advertisements, horoscopes, and the like. 

The costs of software development recur to some degree with each new installation.  Even if standardized local language coding were universally available in India (and it is not), each community has its own local needs, customs, and requirement based on its topography, its traditions, its agriculture, and its communal and caste composition.  All of these factors need to be built into the specifically localized applications written for that community and for others nearby it.  Experienced and devoted programmers are required for this task: they are usually in short supply.

Initial investment also includes housing for the technical equipment that will be used.  In some cases, this may simply be the home of a prosperous local farmer, and the only added cost is posting a sign or painting a logo on the wall of a house.  But in other cases, like the Sindhudurg kiosks designed by NIIT for children in an area of heavy monsoons, special buildings that enable computers to resist the high humidity of the monsoon season were necessary, as were substitutes for the customary mouse, which, experience showed, failed rapidly under the intense use of children.  Once again, the costs of providing a secure and suitable physical place for the infokiosk vary greatly from place to place; but in all cases, they must be included in the estimation of sustainability. 

5.  Training Personnel and Paying Operators

Every ICT installation in India requires the presence of an operator, usually full time, when the infokiosk is working.  The expense involved can be considerable, and is indeed usually mentioned prominently in discussions of sustainability.  To begin with, operators must be literate, must in general have at least a tenth form education, and must be broadly familiar both with the computer technology and with the applications available in the infokiosk.  Furthermore, they must be familiar with the local community, speak the local language, and have good rapport with local people.  In much of rural India, this creates requirements concerning not only education, but caste, gender, and religious community.  Given individuals who fulfill these requirements, further screening for motivation and reliability is necessary, followed by a period of training, which is never less than two weeks and is often longer, and which requires frequent refresher courses. 

Most important, almost any individual who commits himself or herself to be the operator of an infokiosk expects to earn a living—or at least an income--from that commitment.  Large salaries are rarely promised.  But individuals with a tenth form education and further training in the use of computers reasonably expect middle level incomes once the infokiosk is in operation.  In practice, many installations also expect operators to borrow funds to purchase the basic equipment, to repay these loans, to pay operating costs, and to live on the residual income derived from fees for services rendered at the infokiosk.  Typically, this requires an after-cost income of 2,000-3,000 rupees per month (or more) once the installation is fully operative.  Less than that means rapid attrition on the part of operators. 

6.  Maintenance

In addition, equipment needs to be maintained, and maintenance may be a particular challenge in regions of India where climatic conditions are harsh.  In monsoon-driven areas of Kerala or desert areas of Rajasthan, imported computers (and India assembles, but does not manufacture many of the key components of computers like the CPUs) are operating under conditions for which they were not designed.  Furthermore, rodents, insects and animals will do unpredicted damage, high winds will affect intranet cables, and the only sure prediction is that accidents will happen.  Software inevitably crashes; worms and viruses disable whole systems.  Thus, every project must have available skilled maintenance personnel who can travel to the local infokiosk on short notice to make repairs of either software or hardware or both.  Their wages and transportation costs must be covered.  The cost of maintenance must either be included in the purchase price of the equipment, or added as a supplement.

Another personnel cost is the retraining, further training, and replacement of operators.  Not everyone who commits himself or herself stays the course.  New operators must be located, recruited, motivated and trained.  If software or hardware are upgraded, retraining is again needed.  Finally, project morale is generally maintained or boosted by meetings at some central point: transportation and housing must be provided.  These factors need to be costed to assess sustainability.

Maintenance costs, like installation costs, may escalate if complex technologies are used.  For example, the official description of the Tarahaat project has told of the many unexpected steps required before the complex VSAT installation could be properly installed in the unforgiving (Bundelkund) terrain. In addition, maintenance costs for VSATs generally run high and require specialized maintenance personnel.  As a rule, the more complex and sophisticated the installation, the greater the likelihood of problems, and the higher the cost of maintenance.

7.  Operating Costs

Maintaining the back end.  Infokiosks invariably attempt to provide information about something that will prove locally useful: e.g., government records, health problems, job opportunities, land purchase and sales, records, agricultural conditions, weather forecasts, market prices for local produce.  All of these data must be maintained in digital form in some central server or repository, and must be accessible instantaneously (or nearly so) by the local infokiosk operators. 

Local weather forecasts, market prices, or even updated government records do not appear magically in digital form at a central server.  They constitute the “back end” of any ICT4D installation, and they must be constantly maintained and updated.  Doing so requires skilled personnel who are capable of inputting data that arrives in many different forms into a common digital format, such that it is accessible to the infokiosks in the project.

Failure to maintain the back end has spelled doom for some projects.  For example, according to several visitors, one well-known project became a laughing stock to the local people because mandi (market) prices for local agricultural products posted at the local infokiosk were two to three weeks old and hence completely useless.  Indeed, only a few days of failure to maintain the back end is sufficient to discourage and turn away even the most enthusiastic user.  This means that the operators entrusted with maintaining the back end, keeping information up to date, etc. must be well supervised and highly trustworthy.  In a similar way, if a District Collector or other magistrate promises a two week response to citizens’ complaints, but if these complaints are allowed to pile up without reply for months, citizen disillusionment with government--at the ready in every country--is quickly activated, and citizens cease to use the infokiosk as a means of registering their requests and dissatisfactions.

Often neglected, and rarely included among costs, “maintaining the back end” is among the most essential tasks of any ICT4D system, one that requires constant oversight and supervision, and one which, if it fails, condemns the entire project to ridicule and failure.  The costs of maintaining the back end may well be considerable, but they are vital. 

Other Operating Costs

Almost every project in India includes among its stated costs those of electricity, line leasing where telephones exist, VSAT charges, etc.  Sometimes these costs are carried by the operators of the infokiosks themselves out of their infokiosk income.  At other times, they are born by the organization of which the infokiosk forms a part.  In either case, they are constant, predictable, and, in some cases, fixed. 

            Some variation occurs, of course, depending on usage: where land line telephone use is metered, variability in daily or weekly cost is inevitable depending on usage.  Similarly, VSAT cost usually depends on usage.  Finally, electricity cost – when electricity is available—will also be metered according to usage.

            One variable rarely included, but important, involves technologies of connection like wireless local loop.  Some wireless local loop, 802.11, and other wireless technologies permit, in theory, the connection at low cost of very large numbers of satellite receivers.  If, for example, a given sending tower for wireless local loop is connected to 100 receiving stations, the total cost of building the tower and the equipment associated with it is amortized over 100 units, and thus reduced to 1% of the total capital cost of the base station—usually a small figure.  If that figure is used to calculate for the cost of connectivity, the total operating cost of the infokiosk is much reduced.

            But in fact, many of these technologies are new, and exceedingly few are utilized to the maximum.  To take an extreme opposite example, if a wireless local loop tower is constructed at considerable cost, but during the early stages of the project serves only five local stations, each station is in some sense financially “responsible” for 20% of the total cost of the tower.  This may turn out to be a very high figure, which importantly affects the profitability or non-profitability of the infokiosk.  Thus it is important to know precisely how, in the case of promising new technologies for connectivity, these costs have been allocated and calculated for each infokiosk.

8.  Depreciation on Capital Costs

Under ordinary conditions, computers, printers, and switches are relatively long-lived technologies, whose life span – given adequate maintenance--can be calculated over a period of years.  But the shrewd investor will realize that no technology lasts forever, that computers, printers, solar panels, and all other technologies eventually collapse and must be replaced or rebuilt, that technologies change, that the demands of software tend to increase to the point of rendering older computers obsolete, and so on. 

As a result, some calculation for the rate of depreciation of the initial capital investment must be factored into the calculus of sustainability.  The tax code of the Government in India allows 66% depreciation on computers the first year, but this high figure is surely determined in order to encourage the information technology industry.  The tax code of the United States allows depreciation of computers over three years.  At the university where I teach, computers are generally considered obsolete and replacement warranted at the end of three or four years.  In fact, of course, it may be that computers in the rural world, used largely for email and basic internet functions, do not require frequent upgrading, that the use of free software (e.g., Linux) will reduce the hardware obsolescence built into ever more complex and demanding MicroSoft programs, so that a life span of five to ten years is more reasonable. 

But whatever the figure chosen, depreciation on initial capital cost must be included as a factor in calculating sustainability.  It almost always goes unmentioned.

9.  Shutdown and Cost of Recovery

In addition to calculating the costs of the eight factors mentioned above, the prudent investor will consider the possibility that his investment will be for naught, and that the project will fail.  In that case, he will seek a clear exit strategy and a plan for maximum cost recovery.

By exit strategy I mean a set of criteria by which the projects organizers judge that the project has not succeeded, and a process for closing it down.  The factors involved in the judgment may include issues of financial sustainability and cost recovery mentioned above, but they may also include civil unrest, warfare, migration, disastrous climatic events, unexpected local hostilities, etc.  Such eventualities should be spelled out in the initial plan, and a more or less automatic decision point and exit procedure should define the circumstances under which the project is declared closed. 

If that happens, the prudent investor will seek to recover as much as possible of his or her initial investment.  This means pre-arranged plans for the recovery of invested assets, the sale of still-usable equipment, the reutilization of buildings, constructions, wiring, etc. developed for the project, and so on.  In most cases, if a project fails after several years, only a small proportion of the initial investment can be recovered: but given a large investment, even a 10%-20% recovery rate is worth a plan.

II.  Sources of “ICT4D Revenue

            In my comments so far, I have only noted the financial outflows of what I am calling infokiosks.  The other side of the ledger is, of course, equally important.  It is the actual and/or potential sources of income, the revenue shown for each infokiosk and/or for the project as a whole. 

            It is possible to be somewhat categorical about the costs of ICT4D.  But it is more difficult to be inclusive about the actual and potential sources of income.  For one, there may be sources of income and revenue that so far have not been tapped, or at least not successfully so.  For example, some observers hope that local infokiosks in India (as is claimed in some other countries) could be the selling point for local crafts, which would be sold either to the great Indian market or abroad.  Although these hopes have not to my knowledge so far eventuated in actually successful revenue models, they could conceivably do so at some future date, despite the problems involved.  Since so many existing ICT4D projects in India depend heavily on support from government, NGOs, or private individuals in general, exploration of revenue sources for infokiosks is so far limited, and, unfortunately, has produced mixed and confusing results.

            With these qualifications, however, we can note a number of categories or initiatives which either have produced revenue streams, or else might do so in the future.  Finally, we will touch upon one commercial venture which, of those so far studied, seems closest to achieving financial sustainability. 

            I will briefly outline the following potential sources of revenue for ITC4D:

1.  Government Records, Entitlements, and Responses

2.      E-Mail

3.  Transparency and the Elimination of Corruption (e-Governance)

4.  Agricultural information. 

5.  Forecasting Weather

6.  Infokiosks as Local Market Places

7.      E-Commerce

8.      Health and Medical Attention

9.  Teaching Computing

10.  The Role of Entrepreneurship

11.  The Sum of Incomes

1.  Government Records, Entitlements, and Responses

            Like all modern societies, and perhaps more than some, India has a highly developed bureaucracy, many official records, and many entitlements, especially for the weaker sections of society.  Land records, birth and death certificates, licenses of all kinds, permits, tenders, etc. are available and often necessary for citizens to conduct commonplace transactions.  Given this fact, one obvious use to which infokiosks have been put is making available to ordinary citizens, in real time, documents, certificates, and entitlements which previously could be obtained only by going to distant record offices, waiting in lengthy queues, often paying a commission or bribe in addition to a stamp tax or official fee, returning some time later, and perhaps finally receiving a stamped paper certificate with the information or documentation required.  The costs of such transactions are considerable: transportation to the Records Office, the value of time spent traveling and waiting, fees, commissions, and bribes, often a stay overnight or a return visit to await the delivery of records, the cost of transportation back to one’s own village, and so on.

            One of the first, and initially, most successful Indian projects, the Gyandoot Project, initiated in the district of Dhar (Madhya Pradesh) by Amit Agrawal, IAS, and Rajesh Rajora, IAS, built upon the need of the very poor, disproportionately tribal, and disproportionately illiterate population of that district for documents.  For relatively low fees—five to twenty rupees—individual farmers or members of their families could go to the local infokiosk operators (soochalaks), request the information, have the operator fill out the necessary forms on the computer, send them to the central data bank on which records had been digitized, and receive in short order an officially valid document containing the land record, birth certificate, caste certificate, or other document required. 

            For this system to work well obviously requires the digitization of records and a central data bank, access to that data bank, and a legal structure that eliminates fraud and protects privacy, yet permits the production and printing of legally valid documents in local infokiosks.  Thus, the enthusiastic collaboration of the District Collector and his deputy was necessary, as was the support of the Chief Minister of Madhya Pradesh and the ruling Congress Party in the State Legislative Assembly.  Finally, this scheme, which was honored with the Stockholm Prize for the outstanding IT project in the world, required constant maintenance of the back end to provide updated records, mandi (market) prices, meteorological information, and a variety of other kinds of information.3 

The heart of this government-sponsored project, like some of its successor projects like Drishtee, was the provision of government records and services.  Among other things, complaints by local citizens could be lodged with the District Collector, who promised to respond within two weeks.  In case of extreme urgency (such as the running dry of a hand dug well), faster response was possible. The excellent software that made these data and responses available was written in Bhopal, the state capital of Madhya Pradesh, by Satyan Mishtra and his outstanding Drishtee group. 

            For a time, and as long as its two IAS initiators remained in the Dhar district, the Gyandoot project appeared to be largely successful from a financial point of view.  To be sure, many of the initial capital costs, depreciation, etc., were not included in estimates of sustainability.  The State of Madya Pradesh assumed the cost of establishing the network that linked village kiosks to a central hub at the District headquarters.  It is not clear how the programmers at Drishtee were compensated, and by whom.  But local operators in general claimed that, if they were entrepreneurial and collected the appropriate fees for their services, they were able to repay their loans to the banks, grossing two thousand to three thousand rupees a month after expenses, a relatively high income for their region. 

While the provision to ordinary citizens of government documents is undoubtedly one of the most useful services that infokiosks can perform, relying on government documents as a steady source of substantial revenue appears to be somewhat limited.  As some observers have noted, the average citizen only rarely needs valid copies of land records or caste certificates--once a year would be frequent.  Thus, most Indians, most of the time, have little use for this core function of obtaining government records and licenses, and hence, will provide only a small income to each infokiosk. 

            Provision of government services is also the core of the remarkable project Bhoomi in Karnataka, which recently succeeded in computerizing the land records of the entire state of more than 40 million people and closing all of the many Land Record Offices that depended on written paper records.  Bhoomi is now planning on integrating into one central file the land records from each of the mandals (counties) in the State of Karnataka, which Rajir Chawla, IAS, computerized into one central system.  Landowners seeking records go to Bhoomi computer stations and, for a fee, obtain them almost instantaneously. 

Although exact figures are as usual not available, the project is said to be financially viable—although the precise meaning of financial viability, like the meaning of sustainability, has not yet been fully spelled out.  Be that as it may, the accomplishment of computerizing land records for an entire state, by digitizing what are surely the most vital documents in an agricultural nation, is an heroic achievement where much is owed to its prime mover.  Furthermore, providing accurate land records is a legitimate and indeed essential government function, so it is reasonable to expect a steady revenue from the public purse as well as from landowners’ fees.

A second and highly publicized provision of government services occurs in the state of Andhra Pradesh, where projects like CARDS (computer-assisted registration of deeds), and E-Seva--a complex system that in theory integrates 30-40 government services, banking functions, bill paying, licensing, etc. into the single database accessible from a single kiosk.  The Municipal Government of Visag, also in Andhra Pradesh, has developed a system whereby citizens can access records, register complaints, etc.  Precise information on the costs and sustainability on these projects is not readily available, although the Visag project, from the start aware of the unavailability of municipal funds, entered into a series of private partnerships (e.g., with local banks), to pay for the basic infrastructure for their system. 

As noted, the difficulty with relying heavily for infokiosk revenue on the provision of government records and services is that the average citizen so rarely needs government records, licenses, certificates, etc., and so rarely registers complaints.  None of this is specific to India: one would find the same pattern in other developing nations and in the highly computerized nations of the North.  Indeed, some Indian projects like E-Seva involve actual or projected integrations of services which are not achieved in any highly industrialized country with which I am familiar.

2.  E-Mail

            A common use of infokiosks, and a frequent source of revenue as well, is sending electronic mail.  In the most frequently—mention case, rural farmers have relatives who work in other parts of India or abroad, typically in the Gulf States.  Communications are by telephone are erratic or impossible.  E-mail, by contrast, promises lower cost and availability to the recipient at whatever time he or she chooses to consult an e-mail sourcer: e.g., an operator in Kuwait, an infokiosk in Mumbai, a family member in Dhaka.

            The possibility of sending electronic messages of course depends on there being not only an infokiosk from which the messages are sent, but a recipient with access to e-mail on the other end.  This second condition is often not met.  The migrant agricultural worker moving from Bihar to Tamil Nadu for a season may have little or no access to e-mail, and is in all likelihood illiterate in any case.  Thus, complex prior arrangements are necessary, and many or most of those family members or friends who are the likely recipients of infokiosks-sent e-mail are unlikely to be able to receive it.  Like the Grameen Bank cell-phone work which has often been studied and described, e-mail requires both parties with access to electronic communications.  In India, where computers are scarce and about half the population is illiterate and where, furthermore, local language software is frequently unavailable, sending e-mail communications to relatives working in Abu Dhabi or Kolkata often involve infrastructures which does not exist. 

            Another hypothesized use of e-mail and source of infokiosk revenue, is local communications, for example, within in the area served by a network of infokiosks.  Here, the vision is of villagers communicating with each other and between villages using e-mail as their medium.  And indeed, there are examples of one infokiosk sending messages to another, and even of a few, usually highly affluent users, sending e-mail messages to family and friends in other Indian cities. 

            But here, too, infra-structural constraints currently limit the possibilities of e-mail becoming a major source of revenue.  In one study of the Grameen Bank cell-phone experiment, the most common reason given by local Bangladeshis for not using the cell-phone is that they had no one to call.  Simply put, unless the intended recipient of the telephone call had a cell-phone, very complex arrangements had to be made to locate them in order to communicate with them.  Such arrangements are indeed possible, but they are not equivalent to the standard in the industrialized nations where “everyone” has a telephone and can be reached telephonically.  Typically in the Grameen cell-phone work, a telephone call would be made at a precisely pre-arranged time from one cell-phone operator to another: at just that time, the intended recipient of the call would be present at the home of the cell-phone operator.  Alternatively, and somewhat more wastefully, a telephone call to a village telephone operator could result in sending a boy running rapidly to the home of the recipient of the call to fetch them to the cell-phone so that they could speak with the person on the other end.

            In short, e-mail is an actual and potential source of revenue, but its use is constrained by the very low availability of access to Internet communications both within India and by Indians working in other nations.  It is clear that people are willing to pay reasonable sums in order to send e-mail messages, but only if they are certain that these messages will in fact be received.  A guaranteeing receipt may involve creating whole new systems on the receiving end at great expense.  For the moment, at least, e-mail does not appear to be a likely source of major revenues for most Indian infokiosks. 

3.  Transparency and the Elimination of Corruption (e-Governance)

            It is often argued that the computerization and availability at low cost of government documents is an antidote to the common practice of demanding “commissions” or “fees” for the obtaining of these documents.  Such claims, I believe, must be taken with skepticism.  One unpublished study of the CARDS project in Andhra Pradesh suggests that the payment of bribes is still expected at two levels, even though the system is computerized.  (Informal conversation with several citizens of Hyderabad confirmed this finding.)  The benefit, however, it was noted, is that transactions that used to take weeks or even months (with bribes) are now accomplished in a few hours (with bribes).  Thus there is indeed a saving of time and whatever money the time is worth.

In general, however, there is no reason whatsoever why a computerized system should be inherently free from corruption: it is simple enough for a computer operator to demand a bribe in return for entering data onto a database or printing out a computerized record, or to add a (hidden) 20% commission for himself for every transaction he conducts.  Corruption is a political and social phenomenon, not a technological problem to be solved by better technologies.  Similarly, no system that offers citizens an opportunity to address complaints via computers to local magistrates is any better than the local magistrate’s willingness to respond in a timely fashion.  In short, computerization of government services has many virtues, chief among them the increase in speed and ease of access, but the often-made claim that “transparency eliminates corruption” is not necessarily one of them. 

4.  Agricultural information. 

Six hundred to seven hundred million Indians live on the land, deriving their livelihood from agriculture.  Thus, any system of infokiosks or other uses of ICT that improves agricultural productivity, that increases profitability, that reduces crop diseases, that eliminates spoilage, that increases efficiency, that enables farmers to obtain a better price, is likely, other things equal, to alleviate poverty and to improve living conditions for the ordinary Indian.  As a result, almost all ICT4D projects in India--as in other developing countries—have an important component of processing and making available agricultural information. 

Most of the factors that determine the agricultural income of the typical Indian farmer are, of course, beyond the control of any system of information and communication.  In India, the date and nature of the monsoon is vital to farmers in most of that nation; if the monsoon fails, the crop fails, and farmers grow more indebted, hungrier, and more desperate—to the point of all-too-common suicides.  As India joins the World Trade Organization and the international commodity market, world prices of agricultural products like soya, cotton, and wheat, along with subsidies within India and outside of India for agricultural producers, variable costs of shipping and oversea demand—all these factors are among the principal factors that determine Indian agricultural incomes.

But above and beyond these factors, uncontrollable by the average farmer or by the ICT4D project, lie others where information can sometimes make a difference.  Probably the most common use of ICTs is to provide local farmers, through infokiosks, with mandi (market) prices for their primary agricultural products.  Typically, the previous day’s prices in nearby cities (and some not so nearby) are posted at the infokiosk each morning.  With perishable crops like tomatoes, this may allow the farmer to move his produce to the town or city where the price of tomatoes is highest; for storable products like wheat, coffee, or soya, it may allow farmers to hold their produce back from the market in hopes of higher prices later, or to truck their produce to a distant market where prices are higher.  The availability of mandi prices in village infokiosks also in effect bypasses middlemen who have traditionally had a monopoly on knowledge of produce prices in distant markets, and thus have been able to at times to extract exorbitant profits by buying low and selling high. 

In a creative variant of this practice, India’s largest agricultural exporter, the International Tobacco Company-International Business Division, is instituting in central India a system of direct purchases from villages of soybeans, whereby the farmer takes the product in trolleys (small trucks) directly to the processing plant, bypassing the intermediaries who manage local markets, repeatedly bagging, mixing of good and bad soya, and engaging in a variety of other cost-increasing practices.  A preliminary study (R. Kumar) suggests that the combination of the improved quality of soya thus delivered, together with the elimination of transaction costs incurred at the local market, creates a win-win situation whereby farmers receive, in general, higher prices and ITC receives a higher grade of soya at a lower overall cost.  Although this system is nominally based on India’s largest network of village infokiosks, in fact, transactions are largely conducted by telephone, and the savings so far achieved result not from the use of sophisticated information and communication technology, but largely from disintermediation, a new information chain, improved quality of delivered soya, and a more rational system of production, sale and purchasing. 

Another common promise—and actual use—of agricultural information is in providing local farmers with information about best agricultural practices through infokiosks.  The availability of hybrid seeds, new pesticides, of antidotes for crop diseases, of the best strains to plant in soils of differing quality and in regions of differing rainfall—all these kinds of information could, in theory, be available in one of India’s several websites that provide “best practice” information. 

Relatedly, some infokiosks offer special consultations with local agricultural colleges in case of unusual diseases or unsolvable problems.  Here we must rely on the “stories” of crops saved by the rapid intervention of agricultural experts consulted by email at distant agricultural colleges.  Skeptics with experience in the field, note that the effectiveness of this system depends on there being, at the agricultural college, one or more people knowledgeable, available, and quick to respond to requests from local farmers.  Needless to say, such people cannot be guaranteed.  Nor, in most cases, are there funds to provide them consultants’ fees for their advice.

Yet another use of ICTs in agriculture has to do with the sale of agricultural inputs.  The ITC-IBD e-choupals mentioned above sell soy oil at low cost to the villages involved in their work.  More important, they encourage the bulk buying at wholesale prices of agricultural inputs like fertilizers and pesticides, taking a commission for ITC-IBD, and providing the agricultural inputs to farmers at costs lower than those charged if they bought the same inputs on the retail market.  Other schemes contemplate the sale of everything from motorcycles to tractors to trolleys to insurance via local infokiosks.  It is too soon to evaluate the success and profitability of these plans. 

The Special Case of Sugar Cane.  It is no accident that two of the most publicized and possibly most successful Indian ICT4D projects involve sugar refineries.  Sugar is an almost unique crop in that once the cane is cut, the sugar in the cane begins to convert itself to starch within a matter of minutes, and, within a few days, the sugar in the cane is almost entirely gone and the cane is useless.  (Partly because of inefficiencies in delivery and refining, India has the world’s largest plantations of sugar cane, but Brazil, with smaller plantations, produces more raw sugar.) Furthermore, sugar refineries are large and capital intensive installations, often coupled with a secondary refinery that produces alcohol.  For maximum efficiency, this refinery should operate twenty-four hours a day for as many months a year as possible.  The nightmare of every sugar refiner is the arrival on a single day of the entire sugar crop.  The optimal goal is the steady arrival, hour by hour, day and night of a constant stream of just-cut cane, a stream that lasts twelve months a year without interruption—and, of course, a refinery that never breaks down or needs maintenance.

To approximate this latter goal requires, for the benefit of both the farmer and the refiner, a system of iron discipline over the planting, cutting, delivery, and refining of sugar.  Cane that sits outside the refinery waiting for refining for twenty-four hours has its sugar content fall from 15-10%, an almost unacceptable loss.  To avoid such eventualities, major sugar refiners seek to control virtually the entire surrounding region of sugar cane producers, and to dictate or determine dates of planting, days of harvesting, hours of pick up and delivery.  To plan and accomplish the steady, even delivery of cane to the refinery over a period of, say, ten months is a task of great complexity; to convey the requisite orders to local farmers entails the use of hundreds of agents who work closely with cooperating farmers (and all farmers must be cooperative if they are to sell to the refinery). 

Given the special characteristics of sugar planting, growing, harvesting, and refining, the use of computers in planning the process, and the use of infokiosks in disseminating instruction to farmers about harvesting and delivery, are not only tempting but potentially highly rational.  From the companies’ point of view, well designed computerized planning could avoid mistakes and could permit rapid readjustments in the case of unusual climatic conditions.  Moreover, the army of agents ordinarily employed by a large sugar refinery to communicate with local farmers can be significantly reduced by the creation of local infokiosks to which local farmers go to receive instructions, either verbally or in writing, about planting, harvesting, etc.  Finally, the same infokiosks can, in theory, be used for disseminating information about best agricultural practices, productive strains, pest control, and other agricultural inputs.

Probably as a result of these two factors, two of the best known Indian ICT4D projects involve sugar refineries: one, subsidized by the Government of India and the state of Maharashtra at Warananagar, the second, the creation of E.I.D.Parry, a major Indian agribusiness at Nellikuppam in the Cuddalore area of Tamil Nadu.  In neither case do we yet have precise figures about initial investment, return, etc.  In both cases, the initial hopes that the infokiosks, once established, could be used for purposes other than sugar cane payments, deliveries, etc.--have not yet been fulfilled for lack of the necessary resources to write and install the software.  In the Nellikuppam experiment, one student studying the project was told informally by company executives that the company currently expected to recover approximately 80% of its initial investment in ICT.  New technologies of communication involving wireless local loop might increase that percentage to 100%. 

The sugar cane example underlines a critical point with regard to the agricultural uses of ICT.  Given the enormous size, meteorological variety, agricultural complexity, linguistic diversity, and variety of peoples and crops in India, it is impossible to imagine a “uniform” agricultural website or agricultural system.  There might, perhaps, be a central database that covered a great variety of crops, conditions, and regions.  But every region and perhaps every village would require a kind of “customization”—for the local language and for local crops and for local social organization and for local soils and climates.  Achieving such customization--even in the presence of a single agricultural database for all of India--would be a task for thousands and thousands of expert agricultural scientists and programmers. 

5.  Forecasting Weather

The weather—although it cannot be controlled—could at least be predicted with useful results to farmers as far as planting and harvesting go.  The Government of India Meteorological Service has available data, which, in theory, could be analyzed to provide forecasts for each of India’s more than 500 districts.  In fact, however, limited resources mean that weather forecasts are generally available only for entire States.  Since some Indian States have populations of up to one hundred million people, and run distances of a thousand kilometers from end to end, weather forecasts for the entire State may be virtually useless.  Some ICT4D projects have begun to develop the software and technology to derive district-specific weather forecasts that have relevance to local farmers.  Thus, a local forecast of heavy rain within two days may permit farmers to plant seeds; a local forecast of heavy rain, hail, and winds may encourage the harvesting of wheat earlier than otherwise.  These forecasting systems are far from universal in India, but they hold promise for local agriculture. 

6.  Infokiosks as Local Market Places

Several of the ICT4D projects in India offer the possibility of local purchases and sales using the infokiosks.  In a story often repeated by journalists with regard to several different projects, the local farmer’s wife wishes to sell a newborn calf.  She posts the wished-for sale on the local intranet or information system, describing the calf’s characteristics and the desired price.  Nearby operators at other networked infokiosks inform their clients of the availability of the calf.  Eventually, a customer comes on foot or on bicycle to see the calf, decides to purchase it, and all parties leave happy.  The operator of the infokiosk in return receives a percentage—usually 10%--of the cost of the purchase.

In another variant of the story, infokiosk operators post on a chalkboard items available for sale in their locations, and those rural Indians who are literate read (and translate for those who are not) notices about available bicycles, cows, chairs, or whatever may be for sale.  The network of computers acts as a virtual market place, a kind of local Indian e-Bay, facilitating transactions between nearby villages and, at the same time, giving a modest profit to the operator.  No one knows how widespread or successful such intra-kiosk sales are; no data are available on the amounts transacted or the profits for the operators. 

7.  E-Commerce

            The fashionable notion of “e-commerce” covers a multitude of distinct activities.  At one extreme it includes the practice of multinational corporations of placing on the internet detailed specifications and bids for component parts, along with quality specifications, and guarantees of transportation costs.  E-commerce at this level involves billions of dollars of transactions, and presupposes large flows of goods and capital from nation to nation.  At the opposite end, e-commerce involves the local virtual markets I have just described, in which a bicycle, a cow, or a chair is sold from one village to the next using a local intranet.

            Between these two extremes lie a number of other forms of e-commerce about which optimism has often been expressed.  Two in particular are relevant here.

The Sale of Local Handicrafts.  Within the last several years, the government of one Indian State announced that the exquisite handicrafts made by women in that State would soon be available on the web to an international market, thus, increasing sales throughout the world and providing a much-needed source of income to rural women. 

            So far, however, this promise has not yet been fulfilled, and for reasons that are not hard to understand.  For one, the infrastructure necessary to manufacture, gather, transport, label, package, bill, advertise, and sell local handicrafts is extremely complex and not easy to set up.  At present, most local handicrafts in India are sold at the roadside or through a complex system of brokers to shops that cater to tourists and newlyweds in the fashionable sections of the larger Indian cities.  Even in domestic Indian cases, networks of brokers, purchasers, and salesmen are involved, along with mechanisms of quality control, returns, etc.  When creating these networks for sales throughout India via the internet and web, or—even more difficult—for sales to other countries where complex currency regulations and credit card transactions, shipping and returns are entailed--all of this requires an infrastructure that does not in most cases as yet exist in India or in any developing country.  Indeed, in the United States, and in the North European countries which are highly computerized, it still remains true that relatively few customers make significant purchases on the web, apart from a few standard objects like books, records, household electronic equipment, and a few standard garments.  Most people, most of the time, prefer to deal with a local merchant, to visit a local shop, to touch and feel the object they are purchasing, and to deal directly with a local entrepreneur in case of repairs or returns. 

            Equally important, the market for even the most exquisite artisanal objects produced in India or other developing countries is, in the rich nations which are the targets of such e-commerce, a highly limited niche market.  Moreover, that market tends to be close to saturation even today.  A cursory examination of the e-commerce sites in India that sell overseas shows a proliferation of statues of Ganesh, candle holders, and a few kinds of sandalwood baskets and articles, largely aimed at NRIs and a handful of ?Indiophiles.  Even more discouraging, in the city of Boston where I live, there exist already at least a half dozen shops owned by Indians that sell largely Indian goods.  Their owners and their agents comb India for objects for the usefulness and beauty--everything from jewelry to pots to baskets to spices is available.  But, alas, these shops are not crammed with purchasers: non-resident Indians constitute a disproportionate number of those who buy spices; Americans who wish to purchase handicrafts from developing nations have a very large choice of nations to choose from; and what is true of Boston, is, I fear, true of Osaka, Singapore, Oslo, Paris and Madrid.  In short, the dream of selling the exquisite handicrafts of India abroad is worth pursuing, but exaggerated hopes on the ease with which this can be done, or on its profitability should be discouraged.  The niche market which exists is, alas, largely saturated. 

Electronic Micro-credit.  Inspired in part by the Grameen Bank micro-credit work in Bangladesh, the idea of organizing what are essentially credit unions or mini-banks led by local people, usually women, in the weaker sections of society has spread like wild fire across the world.  This is not the place to assess the feasibility or profitability of these micro-credit schemes.  It is enough here to note that in the world of development organizations, setting up micro-credit schemes is one of the most favored activities.

            Since the circle of people, usually women, who pool their finances in order to share financial resources, guarantee loans and initiate new enterprises, involves relatively complicated record keeping concerning amounts deposited, loaned, borrowed, and owed.  It is only natural that inventive development experts and enthusiasts for ICT would view micro-credit for potential use of newer information technologies, as well as a potential source of revenues for infokiosks and the like.  In fact, several experiments are currently under way to develop usable software for the often-illiterate micro-credit group members (for example, iconic screens), and to permit the electronic recording of deposits, loans, debts and payments, often using simple devices like PDAs.

            These experiments are indeed promising, and deserve the attention they receive.

            As a major source of revenue, however, micro-credit circles do not promise great returns.  Often aimed at the “poorest of the poor,” these micro-credit groups frequently involve very small sums of money.  Moreover, they may depend on highly subsidized government or bank loans, which may or may not be repaid.  In a word, the whole issue of the financial viability of micro-credit schemes is itself controversial.  It is clear that many micro-credit groups are unable to repay their loans, although others are successful.  It also seems clear that, under ordinary circumstances, it is sufficient for each micro-credit group to have one or more members who are numerate (though not necessarily literate), in order to keep records.  Whether PDAs—themselves an expensive technology—or computers—an even more expensive technology--which so simplify and speed the process as to make the investment worthwhile remains an open question.  As a result, counting on micro-credit circles as an important source of revenue for infokiosks and other uses of IT for ordinary people seems premature, to say the least.

E-Commerce From Indian Companies Abroad.  Another hope is that the “information age” would permit the development of a flourishing new trade network involving Indian companies with companies abroad.  Here, the dream is that major Indian producers would find it easier to outsource their inputs to nations with lower raw material costs, and to export their products to countries with higher levels of demand.  The result, then, for all, would be flourishing international e-commerce and greater prosperity for India.

            A recent study of the impact of e-commerce, however, indicates that these hopes have not so far been fulfilled.  Where Indian firms have used email and the web actively, it has not been to generate new business, but in the context in continuing in electronic form pre-existing and established relationships with suppliers or purchasers abroad.  Little if any new business is generated: the change is largely a change from telephone and telex to email and web transactions.  The net aggregation of wealth to India—or the net increase in commerce between India and other countries--appears unaffected by the advent by the newer communication technologies.

            All of this could, of course, change in the future.  But the fact that it has not changed so far means that the burden of proof is upon those who argue that e-commerce will provide India businesses with large new opportunities for flourishing businesses with firms in other nations.  If such opportunities exist, they are likely to be facilitated by the more traditional means of person-to-person contacts, of advertising, and the slow, traditional development of relationships of trust between individuals and companies.

8. Health and Medical Attention

            Indian authorities and citizens have long deplored the limitations of health care in the Indian countryside.  Despite fifty years of efforts, many Indians lack basic health care, child immunization is not yet universal, and some physicians in the public employ allegedly devote most of their time to private practice.  The result is a rural population significant parts of which are medically underserved.             

            Information and communication technology might seem to be a good way of disseminating health information to rural populations that otherwise would remain ignorant of both diagnoses and cures of common diseases.  Thus, a number of projects have attempted to make available on-line information describing the symptoms of common diseases like malaria, tuberculosis, severe dysentery, or HIV-- at the same time emphasizing the importance of, and availability of, immunizations against childhood diseases for the youthful population of India. 

Finally, in regions where auxiliary nurse midwives conduct canvases of the health conditions and the health status of rural populations, a few projects have substituted PDAs (personal digital assistants) for the traditional sheets of paper on which auxiliary nurse midwives have recorded their data.  They claim economies of time (if not of funding) and greater accuracy in the gathering of data about the health status of rural populations, and especially of newborn and young rural children.  (These claims, however plausible, remain to be backed up with data.)

            The most publicized uses of IT for health in India, however, have been relatively high-tech uses.  For example, where adequate connectivity exists, it is possible in India (as elsewhere) to send digitized X-rays from local IT-enabled sites to specialty hospitals in the so-called Class-I cities like Delhi, Mumbai, Chennai, and Kolkata.  There, highly trained cardiac specialists are able to diagnose otherwise obscure heart conditions and make appropriate recommendations for surgery or medication.  In another publicized use of telemedicine, the SARI project near Madurai uses infokiosks with a small digital camera on top of each monitor.  Local citizens with eye problems have a low-resolution photograph of their eye sent to the excellent Arvind Eye Hospital many kilometers away.  At the hospital, expert ophthalmologists are able, in some cases, to make an immediate diagnosis on the basis of the photograph and to recommend, for example, cataract surgery at the hospital at a date which can be prearranged. 

            Sophisticated medical uses of IT have much promise in the very long term.  Their feasibility has been demonstrated in a number of Indian projects.  It is questionable, however, whether they contribute to the sustainability of these projects, just as it is uncertain whether funds available for health care in India are better spent sending computerized cardiac X-rays to experts in Class-I cities than in providing basic immunizations to Indian children against childhood diseases.  It may be, for example, that the most effective use of IT to improve the health of the average Indian would be via the intermediary of such auxiliary health workers such as the auxiliary nurse midwives, who could be kept up to date on the latest forms of treatment of common diseases and on new epidemics or ailments about which they should be informed.  In any case, however, no study as yet shows that the provision of health information, however useful and necessary this may be, is a good source of revenue or contributes significantly to the income of Indian ICT4D projects.

9.  Teaching Computing

            The activity most uniformly and unequivocally profitable for local infokiosks is teaching computing.  Every visitor to India today must be impressed with the enormous cultural and economic significance attributed to modern information technology, ranging from computers to websites to cell-phones.  Even though only a small percentage of the Indian population may be “personally” involved-- in the sense of owning a computer (about six million computers, including business computers, in a nation of more than one billion) or having an internet connection (perhaps two million internet connections, many of them in businesses)-- the idea of being computer-savvy, the desire to develop a local computer industry, and, for the individual, the desire to learn about computers is very widespread in India.

            This is not to say that every Indian villager has ever seen a computer or even knows the word “computer,” whether in Hindi, Tamil, or Marathi.  But it is to claim that computers have a cachet and attractiveness for families and children unlike that of any other technology in contemporary India.

            One result, then, is that ICT4D operators can offer basic computer instruction to children of middle and upper school age in their villages for a modest fee that adds to their income and their ability to sustain a reasonable standard of life.  The level of instruction provided may not be the highest imaginable, since some operators themselves have only a few weeks’ training.  And with only one computer to practice on, a group of five to ten students must inevitably take turns.  Nevertheless, for the operator as well as for some of the children involved, new doors are opened, new possibilities emerge, and some of the trained manpower and womanpower that India needs if it is to continue its phenomenal growth in information technology is thus recruited.  Beyond village infokiosks and in larger urban centers lie more advanced commercial training establishments like NIIT and Aptech which provide, for a price, one, two, or three years further instruction in computer use. 

At the end of this instruction, students are by no means system programmers, but they are ready to move into basic jobs in the computer industry, be it as operators in call services, or as writers of code, or, in the case of creative individuals who work on special projects, as co-developers of the hundreds of thousands of local applications which India will need if the ICT4D enterprise is to flourish.  And whatever the individual outcomes for the students who enroll in computer classes at local infokiosks, the modest sums paid by their parents for five, ten, or twenty hours of instruction add significantly to the incomes of the operators.  Other sources of income may be unreliable or depend entirely on the entrepreneurial skills and persuasiveness of the operators.  Still others may depend on external authorities who do or do not maintain the back end of the IT project.  But “teaching computers” is something which any reasonably qualified operator can provide for young children, and it constitutes an important and relatively reliable source of income for local infokiosks. 

10.  The Role of Entrepreneurship

            Several projects in India are attempting to develop fairly large networks of infokiosks concentrated in a defined geographic region: the Madurai-Melur region in the case of SARI, the Sirsa region (Haryana) and parts of Rajasthan in the case of Drishtree, for example.  Both projects agree on a point that is often neglected.  The economic and societal revenue stream of any infokiosk depends not only on the factors mentioned above, but on the entrepreneurial ability, energy, and persuasiveness of its operators. 

            At the extreme, some projects find that many of their operators, having borrowed heavily to invest in the equipment necessary, after a few months abandon the project and default their loans because of the initial unprofitability of the enterprise.  But others in apparently comparable villages appear to succeed and prosper.  No study has carefully examined the differences, but anecdotal reports of the organizers of the project are very consistent.

            The ideal operator is well respected, youthful, energetic, outgoing, initiating and firmly grounded in the community.  He/she does not sit waiting for clients to come to use a service whose advantages they know nothing of.  On the contrary, the operator moves throughout the village, explaining to one and all the wonders of computers, email and intranet, the information available, the ways in which they could profit from it, and the advantages to their children of “learning computers”.  Such operators typically have low operating expenses: they operate from the back room of their home or from the second floor of their house, if it has two floors.  They are agreeable, well liked, and their social and caste status makes them acceptable to one and all.  They feel themselves to be part both of their local community and also of a larger enterprise, and they contribute to both with suggestions for improvement.  If they do not turn a profit in the first month, they redouble their efforts, imagine new uses for the computer, recruit new students, and demonstrate to their friends, neighbors, and fellow villagers that they can use the computers to communicate by email with their relatives who are working in Kuwait or Dubai.

            In short, the entrepreneurship of the operator is a crucial, and perhaps the most crucial, factor in determining the profitability of the infokiosk.  An inert, discouraged, or lazy operator can condemn a potentially profitable infokiosk into a miserable failure, the butt of every village joker.  One of the challenges of those establishing infokiosks in new rural settings is identifying young, vigorous, committed, energetic and entrepreneurial operators.

11.  The Sum of Incomes

            If we aggregate all of the revenue factors listed above—and doubtless others that have not been mentioned—we do not arrive at large sums for most infokiosks, especially if we balance them against the costs discussed earlier.  People will pay for government services, but not very much and not very often.  Agricultural information is useful, and in the case of certain specialized products like sugar, ICT can be used effectively, indeed, essentially.  But in most cases, the need for highly localized information, and the cost of providing such information, together with the apparent reluctance of most users to pay for it means that agricultural information—although absolutely vital to the Indian farmer—is rarely a profit center for the IT operator.  Wholesale purchases of agricultural inputs may, in time, yield a profit for the operator, but it is too early to know, and it is easy to imagine that agricultural supply companies will cut prices rather than see their existing networks of distribution bypassed by “mere” ICT operators.  ICTs can indeed, in certain times and places, serve the function of virtual local markets, but the products bought and sold are usually of small value and limited in number, so the profits to the operator are small.  Health information could potentially be distributed via computers, and highly sophisticated telemedicine has shown itself possible.  But since health is in India, as in most countries, a public responsibility, it is questionable whether the average Indian is willing to pay for basic health information from a computer screen or operator.  Teaching computing does, indeed, seem to be a way of making money, and holds promise for the future.  Most important of all in determining the profitability of infokiosks is the entrepreneurial energy of the operator, which can often only be assessed by trial and error, and where there are many failures as well as successes. 

            Of all the experiments with which I am familiar, the work of ITC-IBD comes closest to, and may well exceeds, the balance of profitability over cost.  This firm is not an NGO and not a government: it is in a large agri-business attempting to compete, largely on the international market, with the huge subsidized agricultural conglomerates of nations like Canada, Argentina, the United States, and north Europe.  Yet it is planning to expand its current number of e-chaupals (village infokiosks) from 800 to more than 2,000 because its data suggests a high rate of return and of the recovery of its initial investments.  Although many questions remain to be answered, those questions are being asked by the thoughtful managers of ITC-IBD themselves.  (In another paper these issues are dealt with at greater length and in greater detail.)

            But with this possible exception apart, the most accurate answer on sustainability, despite much discussion and many claims, is that we simply do not know, that we do not have any figures, that we do not have the basic data, that even the most enthusiastic initiators of projects themselves do not know whether ICT4D projects in India are or are not sustainable.  My own guess is that if we factor in all of the costs mentioned in section I of this paper—including recovery of initial capital, depreciation, etc.--almost none of the projects currently under way will turn out to be sustainable, and, as a result, almost none are independently scalable on a large financial canvas. 

III.  Sustainability and Scalability [This section remains to be written]

            Scalability and sustainability are often confounded.  On the one hand, it is assumed that once we have established sustainability, scalability follows automatically.  On the other hand, it is sometimes assumed that all the fundamental costs, apart from equipment costs, are those of establishing the first in a series of related projects, so that the non-sustainability of the initial project may be counterbalanced by the low start up costs of the rest.  Finally, various theories about viral networks, or the value added by developing intranets, complicate the relationship between the sustainability of one project and the scalability of that project to a larger scale.

            Although the issues are complicated, this section needs to try to pull them apart.  There are indeed some non-recurring costs for an initial project, if that project is one of several or a network of projects, that do approximately the same thing in approximately the same area.  To take an obvious example, developing the basic coding for the local language need only be done once; subsequent programmers can use that coding without major modifications.  Similarly, the costs of training twenty operators are not identical to twenty times the cost of training one operator: there are economies of scale in training.  [It will be necessary to go through the various cost and revenue factors one by one to look at whether or not they are recurring or non-recurring, and whether there are economies of scale imaginable.]

            The overall conclusion should be cautionary.  Successful ICT projects almost invariably involve intensive work with the local population, adaptation of the software, however “generic,” to the needs of the local population, work recruiting and supporting operators one by one, and a variety of other recurring expenses.  In addition, of course, there is a recurring expense of installation of equipment.  Overall, it is a mistake to think that having established one well-functioning infokiosk it is possible to open up a large series with only the added cost of the equipment included among costs.  One of the reasons for the success of the Pondicherry work done by V. Balaji at the Swaminathan Foundation was the great flexibility of the organizers in adapting infokiosks to local needs: e.g., broadcast news about wave heights for fishing villages; information about rice diseases for rice growers; special infokiosks for dalit villagers unwelcome in upper caste villages; and so on.  In a similar way, the cost of setting up e-governance sites in a major city with good telephone connectivity and reliable electricity may be much lower than the costs of setting up similar sites in rural areas where neither continuous power supply nor connectivity can be guaranteed. 

            Finally, the naïve interpretation of “viral networks,” which suggests that the value of a network is a function of the square of the number of interconnected units in it, is an oversimplification of the on-the-ground realities.  It presupposes, for example, that in a network of one hundred infokiosks, each infokiosk will experience one hundred times more “business” than it would, had there been only one infokiosk.  This assumption is clearly incorrect.  Intra-project traffic is often low or nonexistent.  For example, in the Grameen Bank cell-phone experiment, very few calls were “local,” the stated reason being that “I do not know anyone who has a telephone.”  Similarly, in the NepalNet experiment connecting NGOs via email and the web, the project was eventually abandoned because it turned out to be that NGOs had little to say to each other, and rarely used the NepalNet, even though it provided excellent connectivity for NGOs in or near the major cities of Pokhara and Kathmandu. 

            In sum, then, there are indeed economies of scale and non-recurring costs.  Core software can be used and reused, as, for example, in the reuse of the core Dhar-Gyandoot software by the Drishtee Project.  Lessons learned in one place can be applied in another.  But at the same time, every location is unique, every community has its own problems and its own need for organization, and the mechanical multiplication of infokiosk sites without modification and without serious involvement of the local community is an almost certain prescription for failure. 

            All of this will need to be developed at greater length, since a growing number of projects (e.g., SARI, Drishtee, etc.) appear to be relying in part on the savings imagined from scaling up from a single successful model.  [Earlier, the complicated issue of maintaining the back-end for a large project needs to be discussed.]

IV.  Sustainability:  What Should We Conclude? 

            If I am correct in arguing that very few, of even the most imaginative Indian projects are sustainable or scalable as defined in this paper, what should be our response?

            Several responses are possible.

            To return to the imagined position of the profit-seeking individual investor, the most likely response is simply not to invest at all until or unless detailed figures are available for specific projects showing sustainability and involving careful plans for scalability, including a clear exit strategy.  ICT4D is, in short, not a likely candidate for venture capital investors looking for a quick rupee--or even a long-term return. 

            But, in fact, there are to my knowledge no ICT4D projects in India—or in any other country for that matter—which were set up purely as profit-making enterprises by venture capitalists or their equivalents.  Despite the recent argument that enlightened capitalists can make profits by serving the poorest of the poor, the cases cited to support this argument in India, at least, are cases of non-sustainable and/or subsidized projects.  So, in my view, we can put to one side the argument that a good way to make money is to set up infokiosks for the poorest of the poor.  It may, of course, be a good way to serve their needs—although it remains to be seen if Bill Gates is right in believing that the poor do not need computers but healthcare, food, and a variety of other necessities.  But until we have clearer evidence of the benefits to local people of ICTs, Gates’ comments must remain a speculation and not a statement of fact. 

            As repeatedly noted, ICT4D projects in India invariably have had at least the partial support of some external, not-for-profit benefactor, an NGO, a government, an individual, or a profit-making corporation with a philanthropic arm. (See TCS and its Adult Literacy program; ICICI and its many philanthropic activities).  Even the prize money won by several imaginative and well-run Indian projects from international groups can, if translated from dollars into rupees, make a significant contribution to the success and expansion of Indian ICT4D projects.

            But if we try to imagine an India that is “wired”—not only in the sense of having telephones in every village (a goal still far from attainment), but in the sense of having an infokiosk in every village--then a few generous men, prize money, philanthropy, foundations, international agencies and local charities will not be enough.  We need 600,000 infokiosks if every village is to be served; 100,000 if every sixth village is to be served.  “Scaling up” to this level, given present costs and even assuming steadily falling costs of equipment, is almost unimaginable if infokiosks profits alone are to be the main sustenance.  As noted, the costs of equipment constitute but a small percentage of the total cost of the project, most of which are the costs of the people involved, their enterprise, their imagination, their dedication, their entrepreneurship, and their consistency in maintaining and sustaining the projects.  We need, then, to imagine solutions other than the freestanding, self-supporting profit-making village infokiosks, supported solely by revenues gathered from the citizens of the local community.  What are some of the alternatives?

            One has been repeatedly mentioned: it is the model of ITC-IBD, a commercial model that seeks greater profitability for the company, greater competitiveness in the international markets, and a higher price for the local farmer.  The model can surely be criticized, and has been:  it provides a public utility for private profit; it has promised but has not yet provided other utilities apart from those involved in the purchase and sale of agricultural products; although it has been extraordinarily open in calculating and providing its costs, the company itself especially states that it does not know how its competition will react, how its expansion will progress, or what profits it may yet gain from the sale of agricultural inputs through its e-chaupals.  Yet ITC-IBD, together with the near-profitability of Warananagar and Nellikuppam, suggests that the commercial agricultural model is worth exploring, and more than exploring, pursuing vigorously.

            A second model is that of the municipality of Vishnakapatham (or Visag), which early confronted its lack of capital resources and joined forces with private banks in order to provide a gamut of municipal services to citizens, along with advertising and other facilities for private enterprises.  The upfront capital was provided by private sources; the city provided the back end for municipal information and services; and the system, though far from complete, is said to be currently solvent.  Those who have a principled or ideological opposition to collaboration of the public sector with the private sector will, of course, object.  But the alternative, in this case, would have been no municipal project at all.  Similarly, the philanthropic generosity of companies such like ICICI in rural IT projects has complicated motives: on the one hand, it is part of ICICI’s many disinterested public service activities; but on the other hand, it is congruent with ICICI’s long term interest in gaining greater access to the expanding rural market in banking and financial services. 

Thirdly, and perhaps most promising, there is the public plus private model.  Project Bhoomi, the computerization of land records in Karnataka has not to my knowledge released a full accounting of its expenses and revenues.  It claims that it operates at a profit because of small fees charged to individual citizens who wish to access and have copies of their land records.  Several considerations would be involved in calculating the real cost of project Bhoomi.  First, before Project Bhoomi, the State already maintained and paid for a system of depositories for land records in each of the XX mandals (counties) in the state: they cost the State money, not least of all the salaries of those employed in them.  Moreover, in so far as public officials were corrupt, individual citizens incurred not only official charges for documents, but additional charges for bribes to obtain their documents.  Under Project Bhoomi, it is claimed, all these costs are eliminated or reduced.  Second, the extraordinary achievement of digitizing the land records of an entire State—land records that often go back more than 100 years, that are worm-eaten, dog-eared, written in archaic Kannada, and that frequently involve contested records or records in the names of persons who have died years ago—such computerization was a complex, demanding and challenging process which itself costs money.  It is not clear how these expenses were born, nor how long it will take to recover the costs, if indeed they are recoverable.  Yet the convenience to the public—and to the State—of having digitized land records is a public value for which the State might well be expected to pay. 

Or, to take another example, the “Head Start” program in Madhya Pradesh uses relatively old computers in rural schools as a teaching aide, used by regular teachers, during those hours when electricity is available to the typical school.  A dedicated team of young, talented programmers in Bhopal has produced a series of simple but imaginative CDs used in the freestanding computers in the schools.  Several thousand schools are involved.  For this and other reasons, the State of Madhya Pradesh, formerly far behind the Indian average in literacy, has now reached and in some areas surpassed average Indian literacy levels.  In my own brief observations of several isolated, impoverished rural schools, both teachers and students were delighted when the electricity suddenly went on and they were able to use their Head Start CDs as an instructional aid.  Here again, public resources are used in several ways--to assist with the basic programming of the instructional CDs, to provide actual computers used, to accelerate a project of public schooling and universal literacy which is defined as a right of all Indians in the Constitution of the Republic of India.

A third example, in fact a failure, illustrates yet another possibility.  Several years ago, the government of the State of Tamil Nadu entered into an agreement with Aptech? (“NIIT”) to provide at least one networked computer to all the secondary schools in that large and extended state.  The agreement called for Aptech to provide the basic wiring and equipment, along with instruction to the secondary school teachers who would use the computers during the school day to impart “computer literacy” to future secondary school graduates.  In return, Aptech would gain access to the schools and the computers in after-school hours (late afternoons and evenings) for instructional purposes, for which they would charge their customary fees.  As originally defined, the project was to benefit the children, schools and teachers during the daytime, and Aptech’s profit-making computer instruction in the evenings.

The project eventually failed, allegedly because of the lack of demand for Aptech services in the evenings and because of the poor quality and the often remote location of the schools.  (Aptech characteristically builds bright, modern, new, air-conditioned facilities for its instructional purposes; Tamil Nadu secondary schools do not always have these characteristics).  But although the project was abandoned, the concept of an alliance of mutual benefit between public authorities and private enterprises is promising.  NIIT, the larger competitor of Aptech, known for its “computer in the wall” experiment in Delhi, also has major projects in Sindhudurg, a region in the southernmost portion of Maharashtra, on the Indian Ocean.  In Sindhudurg, NIIT researchers are collaborating with school teachers in an effort to ascertain how much students can learn about the operation of computers without any external aid, and at what point instruction from school teachers is needed to assist them to move forward.  Here, too, a public-private alliance is advancing knowledge.

Thus, the answer to the question of sustainability of ICT4D is, as one might expect, complicated.  Like many such questions, it depends on how we define sustainability, how we define cost, how we define profits, and on the data. 

But, sadly, one thing is clear: the data are almost completely missing.  We know anecdotally that project X, once externally funded, has had its funding reduced or withdrawn and is desperately seeking new sources.  We know anecdotally that projects that were once successful have ceased to be so after the departure of their creative, enterprising initiators.  We know anecdotally that several prize-winning projects are in serious financial difficulties, and are trying to develop new revenue models, which might move them toward greater financial sustainability.  But, in any case, we know all too little systematically of the actual benefits to ordinary people of ICT4D projects.

It may also be entirely unrealistic to define sustainability through they eyes of an imaginary venture capitalist whose goal is short term and long-term profit.  I know of no project sponsored by an NGO or by a government that ever sought to recover its initial capital costs: only ITC-IBD has included cost recovery among its goals.  I know of no project that has ever explicitly mentioned annual depreciation of equipment as a cost of the project.  Few include the costs of writing and maintaining locally appropriate software among their start-up costs; most simply mention the expense of the technological equipment as a start-up cost.  However “rational” many considerations may be to the venture capitalist, they simply do not enter into the calculations of most ICT4D projects, and perhaps it is wrong for us to include them.  But enthusiasts of ICT4D who envision an infokiosk in every Indian village need to consider such crass factors as cost recovery, maintaining the back end, the opportunity cost of funds invested, writing localized software, depreciation, and exit strategies as they project grand plans for the wiring of India, or, for that matter, of any other country. 

To conclude:  perhaps the most striking feature of the ICT4D boom in the world today is that so much enthusiasm and money is being committed on the basis of so little knowledge. The same story about the lady selling her newborn calf appears in many nations; enthusiastic reports of wonderful benefits to individuals are written on the basis of a visit of a few hours or, perhaps more commonly, on the basis of the public relations releases of the project leaders.  Instead of evaluations, we have “stories”—doubtless true stories of wonders and marvels accomplished through the use of ICTs.  But in addition to “stories” we need something closer to cost benefit analyses, as crass and abusive as these can be. 

At one meeting I once attended, for example, an enthusiast for ICTs talked about the marvelous fact that villagers who lived ten to fifteen kilometers away from a port were able, through a new infokiosk system, to obtain market prices at the port with the click of a mouse.  No one gave precise figures for the cost of the ITC installation in the village, but at the time it would have been, in dollar terms, at least two thousand dollars, not including the payment to the operator.  One cynical participant in the conference asked, “What was the matter with a bicycle?”  The import of his question was obvious: he meant that for a few rupees one of the many unemployed youths of the village could have bicycled to the port, obtained the mandi prices, and bicycled back in three or four hours.  Not instantaneous, not computerized, not high tech, but quite adequate and at an infinitesimal fraction of the cost.  Such skepticism is needed about many of the alleged benefits of ICT.  Skepticism should not deny benefits, which in many cases are real and cannot be obtained in any other way.  But we need more than “stories”: we need studies, research, analyses and comparisons.

In a word, then, if sustainability is described in crassly monetary terms, by a venture capitalist seeking maximum profits, it is very unlikely that most ICT4D profits are sustainable, and by that token, scalable.  The real total costs of installing an infokiosk or a network of infokiosks are very likely to exceed the revenues, even over time, which might enable us to cover and amortize both the initial and running cost of the installation. 

But if the benefits of ICTs are really great, as at times they clearly are, then we may wish to consider local infokiosks or related installations as public goods, to be supported by public taxation, or by a mix of public and private enterprises.  Nor is there any reason why private philanthropy, whether from within India, NGOs outside India, or from the great international organizations, should diminish or cease.  Finally, it is not legitimate to expect of a pilot and demonstration project that it should turn a profit.  These are projects from which others can learn, projects from which lessons can be drawn, projects, which can provide models, both positive and negative, for the next generation.

Demonstration projects exist to provide lessons for those who follow, not to make money, much less to pay back their donors.  The problem in every country in the world is that these “lessons” are not being learned, and that the questions which we need to answer to establish the benefits or lack of benefits of ICT4Ds are not being asked, nor are facts and figures available which we need to assess the sustainability of ICT4D projects.  As ICT4D projects become more extensive, as more and more rupees and dollars are committed to them, we must work for more extensive scrutiny of their benefits, their costs, and their revenues, and thus hope for sounder judgments of their sustainability. 

Sites Visited


Sites Visited as of February 2003:


SARI (Melur-Madurai), Kuppam (Tenet, now HP, Andhra Pradesh), Nalgonda (A.P., CMC), Nellikuppam (T.N., EID Parry), Pondicherry (Swaminathan Foundation), Bhoomi (Karnataka); CARDS and eSeva (Hyderabad), Sindhudurg (Maharastra, NIIT), Head Start (M.P.), ITC-IBD e-chaupals (near Bhopal M.P), CHIPS and other sites in Chattisgarh, Bathinda (Development Alternatives, Punjab), Gyandoot (Dhar, M.P.), Sirsa (Drishtee, Haryana), Panchkula (outside of Chandigarh), Honeybee-Sristi (Ahmedabad, Gujarat), Warananagar (Maharashtra, Warana Sugar Cooperative, NIC), Nilgiri Networks, (Ooty, T.N.).


Possible Summer 2003:


Dak-Net, SARI (repeat), Visag Municipality, GramChitra, CA:SH, Panchkula (repeat), Kuppam (repeat), Akshaya (Kerala).



These articles above are based on visits to the following sites, among others:


SARI (Melur-Madurai, nCore), Kuppam (Tenet, now HP, Andhra Pradesh), Malgonda (A.P., CMC), Nellikuppam (T.N., EID Parry), Pondicherry (Swaminathan Foundation), Bhoomi (Karnataka); CARDS and eSeva (Hyderabad), Sindhudurg (Maharashtra, NIIT), Head Start (M.P.), ITC-IBD e-chaupals (near Bhopal M.P), CHIPS and other sites in Chhattisgarh, Bathinda (Development Alternatives, Punjab), Gyandoot (Dhar, M.P.), Sirsa (Drishtee, Punjab), Panchkula (outside of Chandigarh), Honeybee-Sristi (Ahmedabad, Gujarat), Warananagar (Maharashtra, Warana Sugar Cooperative, NIC), Nilgiri Networks, (Ooty, T.N.).


I am indebted to many colleagues for their advice, help, and criticisms.  Among them are Ashok Jhunjhunwala, Joe Thomas, Vijay Chandru, Rajiv Sanyal, V. Balaji, S.S. Sadagopan, Roddam Narasimha, E.A.S. Sarma, Rajesh Kaley, Timothy Gonsalves, Venkatesh Hariharan, Swami Manohar, Tapan Parikh, Bikram Duggal, Roy Sastry, Maj. Gen. (Retd) Dr. R. K. Bagga, G.N. Murthy, the officers of NIC Pune, K.V. Nori, Sugata Mitra, Amit Agrawal, Rajesh Rajora, M.D., Rajesh Khanna, Anil Gupta, Satyam Mishra, and the IT Secretaries of the States of Tamil Nadu, Andhra Pradesh, Karnataka, Maharashtra, and Haryana.


Many friends have helped me throughout this work:  one is Frederic Noronha, whose knowledge of South Asian “bytesforall” is unsurpassed; others are Kavas and Paula Petigara, who have been my guides to what is best in India.


Funding for this research has come from the following sources:  The Nippon Electric Company grant to MIT, administered by the Provost’s Fund, a grant from the Ford Foundation, Delhi, for a conference and book on “Equity, Diversity and Information Technology”; support for a research leave and research expenses from Media Lab Asia during the fall semester, 2001.  My thanks to the generosity of all of these agencies: without it, travel to and within India would have been completely impossible.





[1] All of these topics deserve a far more extensive analysis than I can provide.  We need to understand how the notion of “progress” in the West was transformed from the notion of moral and social improvement to a notion of technological and scientific advance.  We need to fathom how the “white man’s burden” in the colonialist discourse was transformed, as in India, to “civilizing the natives,” then to creating “a race of brown-skinned Englishmen,” subsequently to “modernization,” and to finally “development.”  We need to understand how the concept of “development” obscures the enormous differences in the “development” of different societies, implicitly assuming there is one path—usually through science and technology—to advanced status.  We need to understand how, especially after the dissolution of the Soviet Union, a new triumphant capitalism has emerged, with the claim today that profit-seeking, intelligently applied, is adequate to meet the needs of the 2 billion people in the world who live on less than two dollars a day. 


We should also analyze the concepts of the “information age” and the “knowledge society,” largely seen as prerogatives of the elites of the world, and especially of the Northern nations.  Here, it may be enough to note the observation of a young Indian-American writer, Paven Malhotra, who commented on the injunction of the Chief Minister of one of the more aggressive Indian states, to the effect that his state must become a “knowledge society.”  What does this mean, Malhotra asked, in an India where, for more than two thousand years, the highest prestige has been held by a Brahman caste which possessed little mercantile skill and no military prowess, but had a monopoly on knowledge of sacred texts? 

     Most germane to this topic, and absent from this paper, we need a careful analysis of the actual benefits of information and communication technology for development.  I have written elsewhere (Srinivas Lecture) that few studies in any country have systematically shown the benefits of ICTs, much less compared these benefits with those that might have been obtained by spending the same monies and energies for other socially useful purposes.  Lacking an assessment of human, social, political, and other benefits of ICTs, it is impossible to judge whether or not funds spent on non-sustainable projects are spent wisely or foolishly.  This, too, is a subject for another paper.  Here, I simply assume that ICT4D has benefits, and do not attempt to assign a particular monetary, social, or political value to them.  To my knowledge, few, if any, studies, have attempted this difficult task. 

   Finally, with regard to the notion that some communities and some people are “information rich” and others are “information poor,” some are “knowledge rich” while others are “knowledge poor,” critical analysis is needed.  In my experience, information and knowledge are poorly correlated, if at all, with such factors as literacy and education.  Illiterate farmers in India and elsewhere possess vast stores of local knowledge and information, information that tends to be exact, and thoroughly tested.  As compared to the urban, literate, educated, cosmopolitan workers who come to “help” them, they are far better informed about local conditions than their helpers.  Thus, the wisdom of the social organizer, Saul Alinsky, applies to almost all ICT4D projects that are initiated from outside: the goal of the project organizers must be to empower local individuals to the point where they tell the organizer that he is no longer wanted in a community that can now take care of its own needs.  A thorough examination of the concept of “knowledge poor” and “information poor” communities and individuals would do much to “localize” ICT4D projects, many of which assume that outside organizers have more and better information and knowledge than the local people who are being organized. 

[2]   For purposes of simplicity, I will use two words consistently to refer to a large variety of distinct  phenomena.  I will use the word “infokiosk” to refer to all the many community knowledge centers, web stations, information centers, that have in common the use of computers and/or connectivity to email and to the worldwide web, whether through fixed lines, through wireless local loop, or through satellite.  Similarly, I will use the word “operator” for the individual who runs the infokiosk, often the same one who makes the initial investment, sets up the equipment, recruits customers, answers questions, etc.  Again, this term covers a great variety of distinct kinds of individuals,  ranging from dalit women in the Pondicherry-Swaminathan Project to lively entrepreneurs in the Sirsa-Drishtee Project to respected farmers in the ITC-IBD e-chaupals in Madhya Pradesh.  Their financial and other responsibilities vary greatly: in the case of ITC-IBD, it is the agricultural company itself which makes the entire investment and provides free training to the operators; in the case of Drishtee, the operators are expected to negotiate loans, to establish infokiosks with help from central Drishtee organization, and eventually to attain financial sustainability and profitability.  Nevertheless, the term “operator” will have to serve for all these diverse individuals. 

3   The local operators (soochaks) were also permitted to use the equipment (which they had often purchased with government-subsidized loans), for “private” purposes such as sending email, local e-commerce (e.g., selling a cow or a bicycle within the range of the intranet), astrological predictions and forecasts, typing of letters, and so on.  Finally, local operators were allowed to give private computer lessons for a fee to local children or adults as a means of supplementing their incomes.  (See below.)