WeCampaign
Hawaiʻi's Ballot Returns Have Not Peaked: What the D-13 Numbers Say About the Final Two Weeks
The general-cycle return curve went vertical this week and is still climbing. Here is what the official Office of Elections data shows at D-13, how it compares to the primary at the same days-out, and what it means for the last two weeks.
WeCampaign Team
The 2026 general election is on November 3. That is 13 days from the most recent official report.
The single most useful thing in this post: the ballot return curve has not peaked. It is still accelerating, and anyone who plans against a “closing window” is planning against a shape the data no longer supports.
Live data. This post quotes the October 21, 2026 (D-13) reconciliation report. For current official numbers, see Hawaiʻi 2026 Ballot Returns by Precinct. This cycle is moving fast enough that these figures will be well out of date within days — treat the shape as the durable part and the numbers as a snapshot.
What changed this week
Four days ago, on the October 16 report, 829 ballots had come back statewide. Yesterday it was 58,260.
| Report date | Days out | Ballots returned | New since prior report | Cumulative rate |
|---|---|---|---|---|
| Oct 16 | D-18 | 829 | — | 0.11% |
| Oct 19 | D-15 | 14,801 | +13,972 (3-day gap) | 2.05% |
| Oct 20 | D-14 | 34,321 | +19,520 | 4.74% |
| Oct 21 | D-13 | 58,260 | +23,939 | 8.03% |
No report fell on October 17 or 18, so the +13,972 on the October 19 row covers three days. The two later rows are single days. Every rate below is labelled with which it is.
Three things are true at once here, and the third is the one people get wrong.
The pace is still rising. Statewide returns averaged 4,657/day over the three days from the October 16 report to the October 19 report, then came in at 19,520 on October 20 and 23,939 on October 21. The first figure is a three-day average, not a single day; the second and third are single days. An earlier draft of this post treated that climb as a closing window — the reasoning being that ballots either come back in the final fortnight or they do not come back at all. That framing is now wrong, and it was wrong in the direction that would have pulled campaign schedules forward too early. A curve that is still steepening is not a window that is closing.
The cumulative rate has passed 8 percent with two weeks still on the clock. For scale, the 2026 primary finished at 31.73%.
The curve has caught up to the primary. At equal days-out the general cycle ran behind — at D-15 the primary was at 7.52% and the general at 2.05%, a 5.5-point gap. But the general reached 8.03% on D-13, which is above the level the primary was at on D-15. Two cycles that looked very different three days ago now track each other closely.
If your ballot-chase plan assumed a slow, steady trickle through mid-October, that assumption is wrong, and it is wrong in the direction that costs you time.
The D-13 snapshot
From the Hawaiʻi Office of Elections absentee-reconciliation report, report date October 21, 2026:
| County | Ballots sent | Returned | Return rate | Invalid | Rejection rate |
|---|---|---|---|---|---|
| Hawaiʻi | 111,852 | 11,032 | 9.86% | 73 | 0.66% |
| Honolulu | 475,220 | 36,679 | 7.72% | 287 | 0.78% |
| Kauaʻi | 41,265 | 3,962 | 9.60% | 64 | 1.62% |
| Maui | 97,107 | 6,587 | 6.78% | 100 | 1.52% |
| Statewide | 725,444 | 58,260 | 8.03% | 524 | 0.90% |
All 247 precincts are reporting. Sixteen are still at zero returns, down from 85 three days ago.
(One ballot: the statewide report and the Hawaiʻi county report disagree by a single ballot on sent for that county — 111,853 versus 111,852. Return and invalid counts agree exactly. This is a real, known disagreement between two official documents, it is well inside our reconciliation tolerance, and we disclose it rather than silently picking a number.)
What this means for the last two weeks
1. Do not plan for a cutoff. Plan for a slope. The daily pace has risen for three consecutive days. Nothing in the official data yet indicates the crest has passed. Operations that assume returns flatten out now — signature-verification staffing, ballot-processing capacity, chase-list pacing — are sized for the wrong week.
2. Volume is where the operational risk lives, and it is still climbing. Invalid ballots went from 1 on October 16 to 524 yesterday. The rejection rate is 0.90% — inside the range national vote-by-mail research treats as normal, and falling slightly as volume grows, which is what a well-run program looks like. In absolute terms that is 524 real ballots needing signature work, and the count grows with the denominator.
3. Rejection rates still split sharply by county, and the small counties are worse. Kauaʻi is at 1.62% and Maui 1.52%, against 0.66% on Hawaiʻi and 0.78% on Honolulu. Kauaʻi has 16 precincts and Maui 35, so a handful of ballots moves their percentages hard — but the gap has persisted across four consecutive reports, which makes it a property of how mail moves on those islands rather than noise. Small sample, long distances, late transit. Staff signature follow-up accordingly.
4. Honolulu is still the volume problem, but the gap is closing. Honolulu accounts for 65.5% of mail ballots sent statewide, but only 63.0% of the returns. Largest chase list, deepest bench, most slack — and the neighbor islands are now returning proportionally faster than they were.
5. Neighbor-island timing matters more than neighbor-island volume. Maui and Kauaʻi together account for 138,372 sent ballots against Honolulu’s 475,220. Their combined return rate is 7.62%, close to the statewide 8.03%, and their combined rejection rate is 1.56% against 0.90% statewide. The constraint out there is signature handling and mail transit time, not the size of the list.
Where to check the numbers yourself
- Ballot returns by precinct — county and precinct table
- Precinct map — where returns are landing
- Office of Elections absentee reporting — the primary source
No aggregator, no third-party dashboard. Just the state’s own reports, parsed and validated.
A note on what this data is for
Ballot-return data is a turnout-planning tool, not a campaign weapon. It tells you where voters are not yet participating so you can help them participate. Using precinct-level return data to target persuasion against a specific demographic or neighborhood crosses a line that this site will not cross, and we would rather lose a metric than cross it.
The headline numbers here were correct on October 21, 2026. The reasoning above them holds for the rest of the cycle.
Snapshot from the October 21, 2026 Office of Elections reconciliation report. For current numbers, see the live ballot-returns page linked above.