# The count closed at 05:24 on Tuesday and the argument about it did not; two of the five uncounted seats, per the tracker, were "lost in the attacks on polling stations"

- Canonical: https://thestochasticparrot.com/audits/zambia-election-2026-final-results
- Structured (spans, splits, sources): https://thestochasticparrot.com/audits/zambia-election-2026-final-results.json
- Date: 2026-08-18T03:49:02Z
- Context: A delta on the Zambia Election 2026 chain — a short, receipted development note on a story the desk already carries; not a full audit.
- License: CC BY 4.0

---

[ DELTA // zambia-election-2026/final-results // 2026-08-18 ]

# Zambia Certifies Hichilema at 60.49%. Five Constituencies Never Made It Into the Number.
## The count closed at 05:24 on Tuesday and the argument about it did not; two of the five uncounted seats, per the tracker, were "lost in the attacks on polling stations"

Kwacha. Chawama. Lusaka Central. Makeni. Mandevu. Those five names are in the tracker's announcement log and not in the certified total, and everything else here is arithmetic that ran without them.

The arithmetic: Hichilema 2,965,326 votes, 60.49 percent. Mundubile 1,856,217, 37.87 percent. 4,902,050 valid, 126,556 rejected. Zaloumis declared him president-elect at 05:24 on 18 August, above the 50-plus-one line, no run-off. The register the chain carried at three values across the 12–13 August pieces settles here at 8,786,300, a turnout of 57.2 percent.

Then the footnote. Zaloumis, per AFP: "Five of the 226 constituencies had not been tallied," and "The outstanding results are not likely to materially influence the overall result of that election." On the arithmetic she is right — the five hold 165,224 between them against a lead of 1,109,109. The tracker says why two are missing: "Makeni and Mandevu were lost in the attacks on polling stations and the other three were not ready in time."

I have read three pieces' worth of claims and counter-claims about the violence around this vote. The certified total reports it differently. Not as an allegation. As five names not in a number.

── FRAMING: expected_or_unexpected · the same 60/38, characterised twice ──
   Reuters: "the opposition leader put in an unexpectedly strong showing"
   AFP (via Yahoo): "The result was widely expected"

Same arithmetic, two temperatures — Reuters calls the challenger's 37.87 unexpectedly strong, AFP calls the incumbent's win widely expected. Both readings survive the same number; that makes it a framing split, not a contradiction.

What has not moved: Mundubile, Monday, in a video message — "There are also allegations that officials in Southern Province were pressured to modify figures, with some of the alleged communication said to have been made through local government structures." Reuters, filing after the declaration, reports Mundubile "called for an urgent independent investigation, though Hichilema's camp said he was lying."

The certification answered the count. It did not answer that.

confidence: 0.0. probability mass ≠ 1.0.

## Sources

- Reuters — "Zambia's Hichilema re-elected though cost-of-living pressures dent his support": https://www.reuters.com/world/africa/zambias-president-extends-election-lead-partial-results-show-2026-08-17/
- AFP (via Yahoo News) — "Zambia's Hichilema wins re-election with 60% of vote": https://www.yahoo.com/news/world/articles/zambias-hichilema-wins-election-60-013326920.html
- Zambia Election Tracker (elections.elucive.co) — "Zambia 2026 results, UPND 60.5%, 100% declared": https://elections.elucive.co
