Counting Continued; Publishing Did Not
There is a distinction between 'precincts being counted' and 'republic-wide vote-share aggregate being published.' Counting is the operational activity at each polling station: ballots are tabulated, protocols are signed, precinct totals are transmitted to CEC central. Republic-wide aggregate publication is the consolidated CEC release that combines the precinct totals into national vote-share percentages. The two activities can be unsynchronised.
Between 22:05 UTC and approximately 23:25 UTC, CEC's precinct progression counter moved from 404 to 1,029 -- meaning roughly 625 precincts completed their count and transmitted protocols to CEC central in that 80-minute window. CEC central received the precinct totals but did not consolidate and publish a new republic-wide percentage breakdown. From the public's perspective, the count looks paused. From CEC's perspective, the count is proceeding faster than the publication rhythm.
What the Pause Could Mean and How to Test
Three readings of the publication pause are available. First reading: standard administrative rhythm -- CEC publishes republic-wide aggregates at fixed-interval batches (every 200-300 precincts, every 60-90 minutes), and the 22:05-to-sunrise window simply fell on a natural batch gap. Second reading: protocol verification slowdown -- as more precincts report, the CEC central verification workload grows, and the publication interval lengthens. Third reading: political non-publication -- CEC chose not to publish because new aggregates would have shown the Civil Contract share continuing its downward trajectory through Pashinyan's victory declaration window, undermining the political narrative.
The three readings are not equally probable, but they are also not directly distinguishable from public information alone. The 09:00 Yerevan (~05:00 UTC) Hovakimyan benchmark is the next test: if the morning aggregate covers significantly more precincts than 1,029 and shows CC near 54.8 percent, the standard-administrative reading is supported. If the morning aggregate covers 1,029 to 1,200 precincts and shows CC at significantly different from 54.8 percent, the protocol-slowdown reading gains weight. If the morning aggregate shows CC at substantially higher than the prior trajectory, the political-non-publication reading becomes harder to dismiss. OWL will document the morning aggregate against these tests.
Sources: Azatutyun.am, 8 June 2026 (CEC precinct progression + Karapetyan reaction) · OWL live blog