PENNER STRATEGYRESEARCH

All research articles

Washington's August 2026 primary: late rejection was more common among mailed ballots

Covers: August 2026 primary, final August 25, 2026 snapshot

Penner Strategy · September 12, 2026

Washington's August 25, 2026 ballot-status snapshot for the August 2026 primary shows large gaps between return methods. Ballots returned by mail were far more likely to be marked too late than those left in drop boxes.

A primary is the election held before November's general election. In Washington's August 4, 2026 primary, the state's final ballot-status file, tallied statewide across all 39 counties in an August 25, 2026 snapshot, recorded how many returned ballots arrived too late to count under state law. Among ballots returned by mail, 19,111 of 591,239, about 32 per 1,000, were marked too late. Among ballots left in drop boxes, 1,055 of 1,316,978, less than 1 per 1,000, were marked too late. The file has no postmark field, so it cannot show when or why any ballot was mailed or arrived late.

Washington's August 2026 primary: late rejection was more common among mailed ballots

Mail: 32.3 late ballots per 1,000 returns. Drop Box: 0.8. Rates use 591,239 mail returns and 1,316,978 drop-box returns.
The August 25 snapshot of Washington’s 2026 primary shows 32.3 ballots marked “too late” per 1,000 Mail returns, compared with 0.8 per 1,000 Drop Box returns. These labels do not show when ballots were mailed or why they were late. The full table includes other return categories.Download chart (PNG)
View all return-method data

Scroll horizontally to see all columns.

Recorded returns and late rejections, August 25, 2026 snapshot
MethodReturnsLatePer 1,000
Drop Box1,316,9781,0550.8011
Mail591,23919,11132.3236
Email4,939316.2766
In Person3,67720.5439
Non-Standard Mail1,3373022.4383
Non-Standard Dropbox1,1513026.0643
Fax2100.0000
(blank)136461.5385
Undeliverable100.0000

Download the source data

A ballot marked Too Late was rejected for missing the legal deadline. Washington law requires a ballot to arrive by 8 p.m. on election day or carry a postmark no later than election day. A mailed ballot can arrive after election day and still meet the deadline. If the postmark is missing or unreadable, the date on the voter's signed declaration determines whether it was cast on time.

Mail and Drop Box are the county-recorded categories used in the chart. The file also has separate categories such as Non-Standard Mail, email and in-person returns. Including Non-Standard Mail raises the mail count to 19,141 late rejections out of 592,576 returns. The full table keeps all categories visible.

Across all return methods, the snapshot contains 1,919,356 ballot records, with 1,892,281 accepted and 27,075 rejected. Of those rejected ballots, 20,265 were marked Too Late. A September 2 news account reported slightly different totals. The retained evidence does not explain that difference.

The return labels do not show when voters sent their ballots or how the ballots traveled. The file has no postmark field. It cannot separate late mailing from delivery delays or other causes. The gap is a reason to examine what happened before these ballots were rejected; it is not proof that using the mail caused the rejection.

Print articleWashington late-ballot rejections: 2018-2024 general elections and 2024-2026 primaries

Sources and further reading

Detailed research sections

Full research record

The answer

Material qualifications

Full research record

Full research record

Full research record

Findings

Findings

How often were mail ballots rejected as late in Washington's federal general elections, 2018-2024?

How often were mail ballots rejected as late in Washington's federal general elections, 2018-2024?

Washington's EAVS deadline-rejection rate across four general elections

Washington's EAVS deadline-rejection rate across four general elections

Full research record

How has Washington's EAVS deadline-rejection rate moved across four general elections?

How has Washington's EAVS deadline-rejection rate moved across four general elections?

Full research record

Why does the rate differ depending on which total is used?

Why does the rate differ depending on which total is used?

The EAVS rate uses two different columns of the same federal file

The EAVS rate uses two different columns of the same federal file

Full research record

Does the 2018 EAVS file's own math add up?

Does the 2018 EAVS file's own math add up?

The 2018 EAVS file's own totals do not add up by 15,611

The 2018 EAVS file's own totals do not add up by 15,611

Full research record

What Washington law determines whether a ballot counts as too late?

What Washington law determines whether a ballot counts as too late?

Washington's statutory deadline rule for mail ballots

Washington's statutory deadline rule for mail ballots

Full research record

What changed between the reported late-ballot counts in the August 2024 and August 2026 primaries?

What changed between the reported late-ballot counts in the August 2024 and August 2026 primaries?

Late-ballot counts for two primaries come from a news report, not an official record

Late-ballot counts for two primaries come from a news report, not an official record

Full research record

How do the two primaries' reported late-ballot rates compare?

How do the two primaries' reported late-ballot rates compare?

Full research record

What do the state's own ballot-status records for the August 2026 primary show?

What do the state's own ballot-status records for the August 2026 primary show?

The August 2026 statewide ballot-status snapshot's own totals

The August 2026 statewide ballot-status snapshot's own totals

Full research record

Is distance from a mail facility related to a county's late-ballot rate?

Is distance from a mail facility related to a county's late-ballot rate?

County-level association between distance to a mail facility and late rates

County-level association between distance to a mail facility and late rates

Full research record

Does distance from a mail facility relate to a county's late rate?

Does distance from a mail facility relate to a county's late rate?

Full research record

What do the retained USPS-related sources say about postmark timing?

What do the retained USPS-related sources say about postmark timing?

A postmark-timing statement quoted in a Senate release and a news account, not a directive

A postmark-timing statement quoted in a Senate release and a news account, not a directive

Full research record

What does a county-level model combining distance and a past vote-share covariate show?

What does a county-level model combining distance and a past vote-share covariate show?

A county-level model fitting distance alongside a 2024 vote-share covariate

A county-level model fitting distance alongside a 2024 vote-share covariate

Full research record

How sensitive is the distance association to how the model is built?

How sensitive is the distance association to how the model is built?

The distance association is not stable across every specification tested

The distance association is not stable across every specification tested

Full research record

How do Too Late rates compare across the different ways ballots were returned?

How do Too Late rates compare across the different ways ballots were returned?

Too Late rates differ by the return method a county recorded

Too Late rates differ by the return method a county recorded

Full research record

Study scope

Full research record

What changed

Full research record

Unanswered questions

Follow this question

Full research record

Later changes

Use this research

Full research record

Cite this article

Penner Strategy. Washington's August 2026 primary: late rejection was more common among mailed ballots. September 12, 2026.

Data and sources

Ask a question about this