AfricaLink Program Assessment and Recommendations
Executive Summary

Jeffrey A. Cochrane
USAID/AfricaLink Advisor
June 3, 1998
Revised January 8, 1999

The USAID/AfricaLink Advisor undertook a series of three visits to countries with sites that have received support under the AfricaLink programs in East and West Africa. The purpose of the visits was to assess the effectiveness of connectivity interventions and to determine how further AfricaLink assistance might facilitate more effective information management. The following recommendations have been formulated as a result of the visits and in consultation with AfricaLink's implementing partners:

• Greater "networking" is clearly a priority, and this will require an increasing focus not just on access to the Internet but on the quality of that access, particularly access within an organization.⁽¹⁾ Direct access to a keyboard, rather than access through a receptionist, should be a minimal goal.
  
  
• A continued USAID role as catalyst is appropriate, funding primarily initial capital investments rather than maintenance costs. Capital investments should ordinarily at a minimum assure scientists direct access to a keyboard, which will often mean the financing of a dedicated computer for a computer room.
  
  
• Technical support should be focused on national information technology coordinators in the NARS institutions and on regional network leaders. Training programs and technical support organized in the past have not relied directly on these resource people. A series of regional workshops is proposed to assist key resource people in formulating comprehensive information management plans.
  
  
• For funding purposes, networking (as opposed to connectivity) support should be targeted to major international research institutions (ICRAF, WARDA, IITA, ICRISAT, SACCAR, etc.) that demonstrate a capacity and a willingness to foster greater information sharing among their related regional research networks.
  
  
• No specific recommendation is offered about the structuring of tasks presently undertaken by the AfricaLink Advisor in Washington. However, options include the status quo, a replication/integration strategy, and decentralization.

AfricaLink Program Assessment and Recommendations

Jeffrey A. Cochrane
USAID/AfricaLink Advisor
June 3, 1998

The quality of networking depends greatly on the quality of access to the Internet. If greater networking is a priority, then the enhancement of the quality of access to the Internet within an organization, down to the last few meters to the scientist's desktop, must also be a priority.

Having access to the Internet through a receptionist in a distant building is not at all the same as having instantaneous access to the Internet through a desktop computer. Consider for example that while it is technically feasible to participate in an electronic conference (list server) using an information system that passes messages through a receptionist and a printer, the sheer inconvenience of such a system will require enormous desire, dedication, and perseverance on the part of the participating scientist. Few scientists visited by us appear sufficiently willing to make such an effort.

Initial work under AfricaLink has focused on minimal organizational access to the Internet, generally sufficient for email, essentially as a replacement for an organization's fax machine. Under a standard package of assistance, the objective has been the broadest possible access to the Internet for organizations that typically have severe fiscal constraints. The standard package assumes the organization already has a computer and a telephone line, and therefore budgets for the addition of a modem, software installation, Internet service startup, and basic training.

At organizations with very limited finances, staff are typically afforded the ability to send/receive email via a centrally managed terminal, often through a receptionist. Administratively, organizations have treated this new technology, email, more or less as they treated the older technology, fax. Where receptionists formerly managed the fax machine, they now manage the email machine. Received messages are printed and delivered as with faxes. Outgoing messages are transcribed or passed to the receptionist via diskette. Because there is typically no dedicated email computer, the receptionist's terminal is used not only for Internet access but also for routine word processing. Scientists routinely complain that they must compete with their director's word processing demands on the receptionist in order to access their email.

One alternative would have been to provide a somewhat higher level of AfricaLink startup support sufficient to install a dedicated email-access computer in a computer room. Of course the alternative would also have required organizations to allocate additional resources to sustain a higher level or Internet access, including a dedicated place to install the computer, and a telephone line specifically for that computer. Many organizations with which we work under AfricaLink cannot afford the telephone line in particular.

While no computer was included in our initial "standard" package, actual packages of assistance often included a computer. AfricaLink's implementing partners (ICRAF, WARDA, SACCAR) have been regularly encouraged to adjust the standard package to suit local circumstances, and they have indeed regularly made such adjustments. Our implementing partners in the field have seen the need for this higher level of support, and have acted accordingly.

We have seen that email access through a receptionist's computer is the lowest cost means of Internet access at a research station, but we have also seen how this arrangement for email access discourages usage. Email access through a receptionist essentially provides a "fax" level of access to the Internet. With a "fax" level of access, the frequency of email correspondence is not surprisingly found to be comparable to that found with fax machines, though of course at a substantially lower cost. Email enables organizations to do what they were doing before (faxes), but at a lower cost (email). For scientific research, however, we are interested in taking advantage of the Internet to do more than was done before.

While replacing the fax machine with email is an excellent and efficient first step, more can be done, often for only a very modest additional investment. By world standards, using email at a rate comparable to fax machines is not sufficient. For effective collaboration among scientists, we might expect to see at least daily exchanges of email. Such a high frequency of correspondence was not seen at any of the research stations visited, though some scientists did observe that they had corresponded via email at that level when they were studying or working in Europe or America. This suggests that the constraint is not knowledge or interest, but rather access.

Facilitating High Quality Access to the Internet

Access in Africa today is largely a matter of cost and affordability. With unlimited budgets, we would want to place a computer with a dedicated link to the Internet on every scientist's desktop. With limited budgets, creative solutions must be found to facilitate the best possible access. Most of the scientists at stations we visited understood this problem, and many were working toward an efficient solution.

In the past, it has not been AfricaLink policy to intervene directly in an organization's decision about where to place newly acquired Internet access facilities. While we believe that each organization must decide for itself what it can afford and how it shall organize its internal information management, we are now prepared to make a stronger case with organizations about specific configurations. We can make this case based on experience, and we can present evidence of this experience as guidance to organizations as they make their choices.

In the coming weeks, a discussion will be initiated with the AfricaLink virtual team, using AfricaLink-L, about the generation of a small guide to organization managers about issues they should consider in reconfiguring their information management systems to take advantage of Internet access. The idea is to print such a guide and distribute it directly to organization managers, and also to place it on the AfricaLink web site.

The team will consider the proposition that stations with limited resources should generally strive to place an open-access terminal within a computer room, rather than on the desk of a receptionist. The computer should be dedicated for Internet use if possible. Scientists should have the ability to sit at the computer, log into a personal account, read messages addressed privately to them, and compose their own responses directly. This is technically feasible in most countries on the continent.

We can in fact specify a range of hardware deployment options for organizations:

Low Cost

• Single terminal with receptionist
• Multiple use computer (e.g. Internet access and word processing)
• Periodic exchanges of email with ISP via non-dedicated telephone line
• Single organizational email address, shared by all users
• Distribution of messages on paper or diskette
• Limited web browsing, high-priority use only

Minimal Networking

• Single terminal in a computer room or library
• Direct access to the keyboard by users
• Private email accounts for each high-priority user, password protected
• Shared organizational email account for low-priority users
• On-demand dial-up for web browsing, online time individually charged
• Oversight by computer room manager or librarian

Intermediate Networking

• Multiple terminals in a computer room
• Locally networked
• Subdomain enterprise email with private email accounts, password protected
• Dedicated Internet link for unlimited, multi-user simultaneous web browsing
• Oversight by computer room manager or librarian

Enhanced Networking

• Desktop computers in individual offices
• Locally networked
• Subdomain enterprise email with private email accounts, password protected
• Dedicated Internet link for unlimited, simultaneous web browsing, paid centrally
• Oversight by computer room manager or librarian

Low-cost networking has been the standard recommendation in the past, particularly for organizations under severe fiscal constraints. We are now suggesting that organizations deploy resources to accomplish at least minimal networking, and that they restructure their internal spending if necessary to finance it. This will require an increase in the package of startup assistance usually provided under AfricaLink.

Note, however, that it will also require organizations to set aside not only a special place for email access but also a special telephone line. At a minimum, organizations will have to connect an extension, dual-use telephone line to the dedicated email computer. Ideally a dedicated or "fax" line would be allocated for this purpose so as not to compete with voice and other data uses. This will pose a significant administrative challenge for organizations that might already have difficulty paying their phone charges, or that are situated in countries where the installation of a new telephone line is problematic.

This represents a preliminary recommendation for guidance to the managers of organizations assisted under AfricaLink. It will be the subject of discussion among the members of the AfricaLink team in the coming weeks.

Financing Better Networking

The AfricaLink policy with regard to the investment of resources for connectivity has been that USAID should play only a catalytic role, providing connectivity financing through AfricaLink in most cases simply to help secure initial access to the Internet. In so doing, we have sought simply to clarify a need and help organizations see the potential of information technologies, the better then to judge how to incorporate information technologies in their overall administration of information. Organizations that have tasted access to the Internet now typically manifest a growing demand for better access. Seeing the potential for networking, scientists demand the tools to accomplish it, and they are increasingly asking USAID and AfricaLink to pay for it.

The AfricaLink policy has been to provide initial capital investments, but to avoid paying running costs. There has been disagreement among the members of the AfricaLink team about this. The recommendation from Washington has been to fund an initial three months of access, assuring that the monthly expenses are seen by the assisted organization so that they can prepare to assume those monthly charges after the initial subsidized period expires. In East Africa, under the ICRAF/AfricaLink program, a year of access has typically been paid, and payment beyond the initial year is now under consideration. In West and Southern Africa, six months paid access has been more the standard. Our partners in Africa have argued that it takes organizations a longer period to readjust their information expenditures to cover the new costs of Internet access. The alternative view is that reductions in fax expenditures should cover the new Internet costs immediately.

Ultimately, Internet access should be understood as a utility, just like the telephone, electricity, and water. Treating Internet access as a utility like any other, the monthly expense for Internet access should be paid out of a core operational budget. Monthly running costs beyond an introductory period should not be paid from an independent source, e.g. AfricaLink, lest the organization fail to weigh the proper costs and benefits of information along with the other factors affecting the organization's operations. The disagreement among the members of the AfricaLink team has simply been about the appropriate length of the introductory period. However, there is agreement on the general principle.

As organizations structure their information management facilities, there should be only the most prudent of controls placed on individual use. Wherever feasible, time spent online should not be metered. Managers concerned about staff efficiency should if possible judge performance based on results, and allow staff whatever access to the Internet the staff themselves feel appropriate to achieve the results required of them by management.

The implication of unmetered access to the Internet is that organizations, their governments, and their donors should be willing if feasible to consider investing in Internet accounts providing unlimited access, at least for a central library computer, if not for desktop access. AfricaLink has in all regions been willing to pay for unlimited access for an initial period. It remains with organization managers to convince their governments and donors that this is a worthwhile investment for the longer term.

Technical Support for Better Networking

All of these points pertain to strategies for an organizational information management system. Deciding whether computers should be purchased for desktops or for a computer room requires consideration of complex interactions among staff and broader organizational objectives. It would not be prudent to make such a decision from Washington or even from Nairobi or Bouaké. Organizations must make such decisions themselves.

We have considered a variety of mechanisms by which to assist organizations in making appropriate decisions about information management and in effectively utilizing the information resources at their disposal:

• In Southern Africa we contracted with an NGO to organize technical training seminars at meetings of scientists, which were particularly successful where a network's leader took a particular interest. As a result, it was decided to focus greater training and connectivity resources on network leaders.
  
  
• In East Africa we contracted with a regional "help desk" to provide training and monitor connectivity status, though we observe that the help desk is most effective only in the country where it is based (Uganda). We are now considering restructuring the help-desk program, in collaboration with the IDRC, on a national rather than regional basis.
  
  
• We teamed the help desk in Uganda with a reference librarian, and were impressed by the ability of the librarian, Esther Lwanga of NARO/ARIS to monitor and evaluate needs at a national level. Focusing training and other resources on national resource centers seems appropriate.
  
  
• Most recently in Mali, we were impressed with the ability of the IER "Informatics Engineer" to coordinate national implementation of connectivity assistance for research institutes. As with Ms. Lwanga in Uganda, Ouleymatou Traoré Sissoko understands local circumstances and knows the needs of individual scientists.

In consideration of these experiences, and in discussions with Mr. Diallo of WARDA and with Mr. Hailu of ICRAF, it was agreed that we should focus resources where appropriate on key information management individuals, such as Ms. Lwanga in Uganda and Ms. Sissoko in Mali. Mr. Diallo, for example, suggested inviting Ms. Sissoko to Bouaké to spend time at the WARDA documentation center to work on information management strategies. Specific technical training, e.g. for WinNT LAN administration, might also be arranged for her in Abidjan. Similarly with Ms. Lwanga in Uganda, Romaine Williams of USAID/Washington is assisting in the drafting of a national information management plan to specify how any resources concentrated at NARO/ARIS will be applied to benefit scientists nationwide. Similarly, support is now being arranged for Fatou Ndoye of NESDA to be trained in the management of an electronic information resource center. USAID's CDIE has been contacted to provide mentoring and training for Ms. Ndoye.

We should proceed immediately to identify key resource people, such as Ms. Lwanga, Ms. Sissoko, and Ms. Ndoye in each of the AfricaLink countries. It seems prudent and cost effective to schedule a series of regional workshops, perhaps one each at ICRAF, WARDA, and SACCAR, or at other appropriate regional institutions. The purpose of these workshops would be for the resource people to explore strategies for facilitating information management within their national organizations and networks. The principal output of each workshop would be a complete national (or network) information management strategy, including a specification of requirements.

Specific agenda items to be considered by the AfricaLink team are:

• Comparison of national Internet access experiences among participants
• Technical issues, possibly to include training for
• Managing an effective web site
• Managing an organizational email server
• Managing a local area network
• Incorporating Internet access in a national (or network) information management system
• Strategies for facilitating sub-regional networking
• Constructing an information management plan

Experts might be contracted locally for particular training topics, based on needs identified by the AfricaLink team in consultation with participants prior to the workshops. CDIE under the Leland Initiative has staff with expertise in facilitating such workshops (e.g. Zoey Breslar) who might be tapped to assist. It may be appropriate to include at these workshops some of the leadership of key regional networks (e.g. SARRNeT, PEDUNE), as well as staff from the major regional coordinating bodies (e.g. ASARECA, CORAF). If so, then double the estimate for subregional participants.

Funds do remain in the grants to ICRAF and WARDA to cover costs if our implementing partners agree that such workshops would be worthwhile. The RCSA may wish to fund a workshop hosted by SACCAR. AfricaLink/Washington and/or the Leland Initiative can pay for international participants. It remains then to determine how USAID support shall be structured subsequent to these workshops. There are essentially two issues to be considered. One has to do with the AfricaLink roles of the international research institutions in Africa. The other concerns the role played by the AfricaLink Advisor.

The Role of IARCs in Africa

The critical question with respect to the international agricultural research centers (IARCs) in Africa (ICRAF, WARDA, IITA, ICRISAT, SACCAR) is whether they are well placed to play an effective role in implementing the above recommendations. Two of these, ICRAF and WARDA, have been key implementing partners for AfricaLink, serving primarily to:

• Process requests for connectivity assistance from African organizations, in accordance with guidelines suggested by the AfricaLink Advisor in Washington
• Evaluate the validity of connectivity assistance proposals and process payments to vendors.
• Report on the effectiveness of connectivity assistance grants to African organizations.

All the IARCs, not just those presently implementing AfricaLink, play a role in supporting African agricultural research networks and NARS organizations. An important working relationship is already in place. As custodians themselves of substantial quantities of important research information about Africa, they already provide reference and research services to African scientists. Most are also well connected to the Internet, often with high-bandwidth satellite or fiber links. The IARCs are thus well placed to facilitate more effective networking in Africa by working closely with African networks and NARS organizations. It is therefore recommended that AfricaLink support for networking be channeled through these IARCs in support of the specific networks with which they have an existing working relationship.

In the past, AfricaLink funds have been channeled to one IARC in each region in support of connectivity for all agricultural networks. The recommendation is that this funding be diversified among more of the IARCs, and more closely tied to research networking. This should be done, however, only with those IARCs that actively seek to integrate their information resources into African network information systems.

¹ We define "networking" to be the regular sharing of information among a defined group of individuals working toward a common objective.