Digital rights advocacy after an internet shutdown succeeds or fails on evidence collected during the shutdown itself. Once connectivity returns, attention collapses within days, and an organisation that starts gathering documentation at that point has already missed its window.
The uncomfortable implication is that the advocacy work begins while the network is still down, when it is hardest to do.
What counts as evidence in digital rights advocacy
Technical measurement establishes that a disruption occurred. It does not establish what it cost. Both are needed, and organisations consistently over-invest in the first.
Technical record. Independent measurement from NetBlocks or similar, plus your own timestamped notes on what was unreachable and from which networks. Note the granularity — full blackout, mobile data only, specific platforms — because remedies differ.
Legal record. Any directive, licence condition or public statement authorising the restriction, and the statutory basis claimed. Where no instrument exists, that absence is itself the finding.
Human record. This is the part that persuades and the part that is almost always missing. Specific, attributable accounts of what could not be done: the clinic that could not confirm a referral, the trader who lost perishable stock, the student who missed an examination registration, the family who could not receive a transfer.
Collect the human record contemporaneously. Reconstructed six weeks later it is vague, and vagueness is what allows a disruption to be characterised as a minor inconvenience.
Cost it in terms decision-makers already use
Rights arguments alone rarely move a finance ministry. Economic arguments frequently do, and the two are not in tension.
You do not need a sophisticated model. Document what closed, for how long, and what those operations turn over on a normal day. Mobile money agents, transport operators, market traders and small businesses that depend on digital payments are the clearest cases, and their operators can usually tell you precisely. ITU connectivity data and GSMA's mobile economy reporting give you the national context to place those figures in.
An estimate presented with its method and its uncertainty stated is far stronger than a confident number nobody can check.
Escalation routes that exist
Documentation without a destination is an archive, not advocacy. Decide the route before you collect.
- Coalition submission. The #KeepItOn coalition aggregates national evidence into international pressure and has an established process.
- Regional mechanisms. Continental and sub-regional human rights bodies accept communications, and regional courts have issued relevant judgments on network disruptions.
- Domestic litigation. Slow, expensive, and the only route that produces binding precedent. Requires exactly the contemporaneous record described above.
- Operator engagement. Telecommunications companies executing an order are a pressure point, particularly those with international parent companies subject to human rights reporting obligations.
Paradigm Initiative and CIPESA both run established programmes on this across the continent, and coordinating with an organisation that already has standing saves considerable time.
Publish before the attention goes
The first seventy-two hours after restoration are worth more than the following three months. People are online, the experience is recent, and the emotional memory is intact.
Have the summary drafted before the network returns. Publish the human record first and the technical analysis second — the technical piece is for the eventual submission, the human piece is what makes anyone care enough to read it. Newsrooms that kept a verification ledger through the disruption, as described in verifying information during internet shutdowns, already have most of this written.
Connect it to the wider pattern
A single shutdown is an incident. A documented sequence is a pattern, and patterns are what change policy. Keep a running national record with consistent fields — dates, scope, stated justification, legal basis, measured duration, documented costs — so each new event strengthens the case rather than restarting it.
The same discipline applies to platform-level restrictions on speech, which often accompany network disruptions; see the quiet cost of platform moderation in African languages.
Common questions
Is it worth documenting a short shutdown?
Yes. Short disruptions are how longer ones become normalised, and they are the cheapest entries to add to a pattern record.
What if the shutdown was officially denied?
Independent measurement data becomes the centre of the case rather than a supporting exhibit, and the denial itself becomes part of the record.
Who should hold the documentation?
More than one organisation, in more than one jurisdiction. Records held only where the disruption occurred are vulnerable to the same pressures that produced it.