Summary

  • Nancy Hafkin’s work at UNECA joined development databases, common standards, local centres and trained operators to low-cost store-and-forward communications; the network’s institutional parts existed before reliable full Internet access.
  • CABECA’s technical gains and political vulnerabilities helped turn a pilot into AISI, a ministerially adopted framework that put national policy, skills, information resources and inclusion alongside connectivity.

The most revealing packet in the early history of African networking may be one that took nine years to arrive. UNECA later recorded that a PADIS newsletter sent to Niamey in 1991 came back to Addis Ababa in 2000. Nancy Hafkin recalled the episode as proof of an information system defeated by its means of delivery. Countries supplied records to shared databases, yet the people meant to use them could wait so long that the answer lost its value.

It is tempting to begin the repair with a modem. Hafkin began elsewhere: with the system that would give a modem something useful to connect. The Pan African Development Information System, established by ECA in 1980, was designed as a cooperative structure. A coordinating node in Addis Ababa worked with national and institutional centres. Bibliographic, statistical and referral databases held African development information. Common processing standards were meant to let records travel between centres without becoming unintelligible.

By 1994, ECA counted 40 national participating centres and 15 regional or subregional institutional centres. PADIS had trained more than 100 information specialists. These numbers describe an information network, not yet a dependable telecommunications network. Its original vision assumed satellite links that the continent’s practical infrastructure could not supply. Requests therefore fell back to post, fax and costly calls. The architecture was ahead of the transport.

Making delay part of the design

CABECA—Capacity Building for Electronic Communication in Africa—was the bridge. Begun with IDRC support, it sought low-cost Fido-based nodes in roughly 24 countries. A store-and-forward system did not pretend that a fragile telephone line was permanent. A local computer collected messages, made a dial-up call when it could, exchanged a batch and passed the mail onward. Interruption became a condition to manage rather than a reason not to communicate.

APC’s retrospective describes 1,200-bps modems, FidoNet, UUCP-style links and a principal route through GreenNet in London. Email, conference mail, file transfer and database queries could move without a full-time TCP/IP connection. Internet Society records a 1990–1993 pilot linking about 18 African institutions and operating a bulletin board and an electronic conference. The figures refer to different programme stages, but the mechanism is consistent: scarcity was answered with asynchronous movement, local operation and shared services.

Hardware was only one input. ECA says CABECA trained more than 100 system operators who became trainers and troubleshooters. It supplied nodes run by local operators, promoted discussion lists and online databases, and cultivated policymakers as well as users. APC says the wider effort helped establish early email networks in 24 countries and that PADIS reached about 1,500 subscribers before the Ethiopian telecommunications authority closed it in 1996. Those retrospective counts deserve attribution, but they illuminate what “connection” meant: a community capable of keeping a modest node useful.

When a technical pilot meets a monopoly

CABECA also exposed what clever engineering could not route around. ECA’s own review is unusually direct: telecommunications operators, ministries and regulators were not participants at the outset. State-owned postal and telecommunications operators could jeopardise the project. Import rules, tariffs, scarce lines, expensive international calls, power failures and server breakdowns all sat outside the protocol.

This was not simply friction around an otherwise technical programme. It was the programme’s control surface. A node could demonstrate demand, but an incumbent could deny permission, price the call or close the service. A trained operator could repair a server, but not repeal a modem ban. Sustainability required the institutions that controlled networks, customs and national planning to treat communications as development infrastructure.

The pivot from CABECA to the African Information Society Initiative therefore matters more than a story of technical succession. In 1995, African ministers called for an action plan. On 8 May 1996, ECA’s Conference of Ministers adopted AISI through Resolution 812 (XXXI). The framework joined decision-support systems, African information resources, human capability and regional and global links. It asked member states to place the initiative inside national development plans and distributed work among governments, ECA, international agencies, regional bodies, civil society and the private sector.

Hafkin’s earlier work on women and development also remained visible. The resolution required a gender-balanced African technical committee and integration of gender into the policy. A clause did not dissolve unequal access. It did, however, make inclusion part of the institution’s stated design rather than a benefit to be considered after cables arrived.

ECA later reported that roughly three quarters of its 53 member states had national e-strategies or NICI plans during AISI’s first decade. That is not a clean causal measure of one programme. It is evidence of a change in scale: networking had moved from proving that messages could cross weak lines to influencing how states planned information infrastructure.

Hafkin’s distinctive contribution was to hold the layers together. Information had to be collected and compatible. Operators had to be local and able to teach. Transport had to fit scarcity. Rules had to permit the equipment and the service. Governments had to understand why the network mattered. The connection was never the whole network; it was the interval in which a prepared network could finally move.

Sources