Michigan / 2026 general election / Research guide
Michigan’s early vote.
The 2026 research guide.
What the outside analyses conclude. How they get there. And what the ballots still cannot tell us.
Understand the methodsRead the method
before the margin.
Michigan’s early ballots do not come with a party label.
The state records participation. A Republican or Democratic reading requires another layer of analysis: where voters live, a model of their likely affiliation, or assumptions about who will vote next.
This guide brings together 40 outside articles, trackers and social analyses from our research. It distinguishes original analysis from reporting that shares a source. It is a researched collection, not a claim to have found every publication online.
Michigan does not register voters by party. None of the modeled partisan figures in this guide are counted candidate votes. Michigan Department of State.
01 / Three different questions
County patterns, voter models and forecasts answer different questions.
Choose a method to see what it measures—and where the inference begins.
Nine Harris-won counties highlighted
Where are the ballots coming from?
Geography describes places, not individual voters’ choices.
Group the same counties across elections and compare their share of statewide returns. This is the approach in our companion analysis.
- Observed
- Recorded participation within each county.
- Added by the analyst
- A fixed grouping based on a previous election.
- Still unknown
- Which candidates the participating voters chose.
Land area does not represent ballot volume. Both parties’ voters live in each county.
Illustration of a method; dots are not voter data.
What kind of voters have participated?
Voter-file models estimate affiliation from individual records.
Participation histories can identify repeat voters. In Michigan, a party label added by a data provider is an estimate, not a state registration record.
- Observed
- Recorded participation and available voter history.
- Added by the analyst
- A model of likely partisan affiliation.
- Still unknown
- Candidate choice and the composition of the final electorate.
This diagram is explanatory, not a prediction.
What might the final electorate produce?
Forecasts require assumptions beyond ballots already recorded.
A model may allocate ballots by county history or demographic patterns, then estimate the voters still to come. Different assumptions can produce sharply different partisan readings.
- Observed
- Some portion of current participation.
- Added by the analyst
- Party allocation, future turnout and vote-choice assumptions.
- Still unknown
- Whether the assumptions will hold in this election.
02 / Reading the conclusions
Michigan’s partisan readings depend on the method.
These are descriptions of the sources’ arguments. A partisan interpretation is not an endorsement of its forecast.
Descriptive evidence
Participation counts establish volume, not a party’s lead.
Official releases and descriptive trackers establish how much activity has been recorded. A larger total can reflect earlier voting, changes in voting methods or reporting—not necessarily additional voters.
A narrow comparison
ABC’s Michigan comparison points to continuity.
See the dated method and scope below.
An attributed Republican reading
Some Republican arguments depend on county allocation or turnout assumptions.
Election Data Guy’s derivatives and turnout scenarios should be read with their assumptions visible. Multiple articles drawing on one model do not constitute independent confirmation.
An attributed Democratic reading
A Democratic model estimate also needs a data-quality check.
On Point’s partisan display is modeled. Unresolved totals and voting-method coverage prevent us from treating it as a verified measure of Michigan’s electorate.
A date is part of the evidence
A new article can describe an older electorate.
Keep the publication date, data cutoff and review date separate. A later headline does not automatically contain newer participation records.
Example: ABC News’ Michigan comparison. Source and scope ↗
03 / Follow the evidence upstream
Several publications can share the same underlying source.
These examples show dependencies, not an assessment that every tool transforms its data correctly.
Conservative BriefCommon Sense Gay · Oct 15Common Sense Gay · Oct 25
Each library entry identifies its source family where known. Syndication, shared feeds and commentary on an existing model should not be counted as separate studies.
04 / The source library
Compare the conclusions.
Inspect the methods.
40 outside records
Articles · Trackers · Social analysis
Research reviewed October 29, 2026
Open an entry for its method, limitations, data cutoff and access status. Dynamic trackers may change after our review. Where a page could not be re-read, the entry says so.
Showing all 40 sources
Early Vote Tracker and methodology
EDG converts early participation into estimated party allocations and eventual-result projections. A current Michigan output could not be verified in this review.
Method, limitations and source details
- Method
- Where party identification is unavailable, uses county returns and historical voting patterns, then extrapolates. The published method excludes polls.
- Limits
- County estimates are not individual party records. Public documentation omits a full reproducible formula and uncertainty calibration. Dated commentary using EDG should not be mistaken for today’s model output.
- Underlying data date
- Current Michigan output unavailable
- Author / source family
- Harry / Election Data Guy
EDG
Supporting source 1 ↗ Supporting source 2 ↗
The 2026 Early Vote Is Running Ahead of 2022. Who Does That Help?
Michigan absentee returns were modestly ahead of 2022 in its dated comparison. More early ballots can reflect earlier voting, without establishing a partisan advantage.
Method, limitations and source details
- Method
- Michigan state figures, AP and external trackers; separates voting methods and comparison calendars. Broader historical turnout analysis is national.
- Limits
- The Michigan comparison excludes early in-person voting. Its 2022 worksheet is October 31, eight days before Election Day, versus seven days for October 27, 2026—not an exact matched-day comparison. “Mail” includes absentee ballots delivered without mail.
- Underlying data date
- Michigan comparison published 2026-10-27; national sources vary
- Author / source family
- Dave Manuel
MDOS / AP / tracker synthesis
Analyzing Early Voting: North, South, East and West
Interprets older early voters and opening county comparisons as a Republican opportunity.
Method, limitations and source details
- Method
- Age composition, polling and Con ET Takes’ comparisons with 2024.
- Limits
- Early voters are not the final electorate. County winners are not individual parties. Two days of ratios cannot establish a turnout trend.
- Underlying data date
- Opening two in-person days; posts dated 2026-10-25 and 2026-10-26
- Author / source family
- Duane Patterson
HotAir / Con ET Takes / polling
Michigan Absentee Requests Jump 17% as Early Voting Begins — What County Results Reveal
Historical county winners provide context but cannot identify which party benefits from early participation. Expanded voting options complicate comparisons.
Method, limitations and source details
- Method
- October 20 absentee figures and 2022 gubernatorial county winners; explains participation relative to registration.
- Limits
- The article labels October 24, 2022 and October 20, 2026 as both 14 days out. They were 15 and 14 days out, respectively. Its county groups use the governor’s race, not the 2024 presidential contest.
- Underlying data date
- 2026-10-20; historical 2022 county results
- Author / source family
- Mike Brennan
MDOS / MITechNews
The final push: Michigan’s campaign blitz and fears of a ‘November surprise’
The October 26 interview judged it too early to infer a turnout trend from Michigan’s opening in-person days.
Method, limitations and source details
- Method
- Reporter interview about new voting options and conversations with voters changing how they return ballots.
- Limits
- Anecdotes about postal concerns do not measure the scale or party composition of a method shift. No quantified Republican or Democratic advantage.
- Underlying data date
- 2026-10-26 discussion
- Author / source family
- Sophia Saliby and Russ McNamara
MPRN reporting
Where’s all this Democratic enthusiasm? Check the scoreboard
The October 25 commentary challenges Democratic enthusiasm headlines, citing EDG’s narrowing Michigan estimate and anticipated Republican participation later.
Method, limitations and source details
- Method
- EDG state estimates, national registered-party totals and polling about intended voting method.
- Limits
- Michigan remains a county-derived estimate. National totals mix reporting states and voting methods; nationwide voting-intention polls cannot measure Michigan’s remaining electorate. This is the same EDG evidence family, not independent corroboration.
- Underlying data date
- 2026-10-25; trend begins 2026-10-04
- Author / source family
- Common Sense Gay
EDG
The Source: Countdown to the Midterms — Michigan turnout discussion
The broadcast describes Michigan’s early participation as not far from presidential-year levels and discusses turnout incentives for both parties.
Method, limitations and source details
- Method
- Host observation and campaign reporting; no published quantitative comparison in the segment.
- Limits
- The transcript supplies neither counts nor matched dates for the 2024 comparison. Treat it as attributed broadcast commentary, not a validated benchmark or party forecast.
- Underlying data date
- 2026-10-25 broadcast; numerical cutoff unspecified
- Author / source family
- Kaitlan Collins and Arlette Saenz
CNN reporting
Early Voting Expands Saturday in Florida, Michigan, New Jersey, New York
Pairs national early-voting growth with prediction-market expectations and describes Michigan’s Senate contest as competitive.
Method, limitations and source details
- Method
- Synthesis of AP and ABC reporting, national trackers and market prices.
- Limits
- No independent Michigan early-vote model. Market prices are not returns, and national turnout or party patterns do not establish Michigan’s partisan direction.
- Underlying data date
- Sources dated through 2026-10-22; market discussion 2026-10-23
- Author / source family
- Pablo Planovsky
AP / ABC / prediction-market synthesis
Early vote shows Democratic enthusiasm high in some key states, but only one piece of puzzle
Michigan’s modeled party mix resembled 2024.
Method, limitations and source details
- Method
- Matched L2 voter records through October 19.
- Limits
- Party estimates are not votes; subgroup percentages do not describe all early voters or the final electorate.
- Underlying data date
- October 19, 2026
- Author / source family
- Cody Prosperini, Zoe Baker and Oren Oppenheim
L2 / ABC
Michigan Early Voting Tops 861,000, But 17% Jump in Absentee Requests Raises Turnout Questions
Absentee requests grew faster than completed returns. Automatic ballot delivery complicates interpreting requests as enthusiasm or additional turnout.
Method, limitations and source details
- Method
- October 20 state comparison with 2022, followed by October 21 dashboard totals; calculates request growth and return share.
- Limits
- Those are separate snapshots. Historical cutoffs need checking; a press-release date alone does not establish an exact same-stage comparison. No individual party model.
- Underlying data date
- 2026-10-20 comparison; 2026-10-21 snapshot
- Author / source family
- Mike Brennan
MDOS / MITechNews
Many states deciding control of the Senate are seeing a big boost in early voting
AP’s October 21 comparison found Michigan participation high but similar to 2022, unlike several states with large increases.
Method, limitations and source details
- Method
- Early-voting totals through October 20 compared with the corresponding two-week point in 2022.
- Limits
- The national Democratic enthusiasm framing is not a Michigan partisan finding. Volume cannot identify voters’ candidate choices.
- Underlying data date
- October 20, 2026
- Author / source family
- Joey Cappelletti, Mike Catalini and Simran Parwani
AP Election Services
Supporting source 1 ↗
What early voting numbers say about Michigan’s 2026 midterm election
Changing habits complicate turnout predictions.
Method, limitations and source details
- Method
- Expert interview and participation totals.
- Limits
- No reproducible partisan forecast.
- Underlying data date
- October 21, 2026
- Author / source family
- Elaine Rojas-Castillo
CBS interview
Burn the Polls. It’s Turnout Time.
Historical turnout scenarios yield a Republican-leaning Michigan electorate, R+1 to R+3. The article treats current absentee returns as inconclusive.
Method, limitations and source details
- Method
- Applies geographic participation changes from 2016–2018 and 2020–2022 to the 2024 electorate, holding local presidential preferences fixed.
- Limits
- This is a hypothetical turnout scenario, not a measured early-vote margin or the author’s full election forecast. It excludes changes in candidate preference; the article lacks a complete replication dataset.
- Underlying data date
- Historical scenario; current returns discussed through 2026-10-16
- Author / source family
- Patrick Allocco
Zoose historical scenario
Georgia showed up, Michigan is a coin flip, and the gap is closing
The October 15 commentary calls Michigan a coin flip as EDG’s modeled Democratic edge narrows during October.
Method, limitations and source details
- Method
- Compares successive EDG estimated party shares within the 2026 voting period.
- Limits
- Same model family as Conservative Brief. A narrowing early margin may reflect when different voters participate; it is not a matched-cycle test of improvement or a count of candidate votes.
- Underlying data date
- 2026-10-04 through 2026-10-15
- Author / source family
- Common Sense Gay
EDG
What Early Voting Data Is Telling Us About The 2026 Midterms
The October 15 essay reads national data as Democratic enthusiasm but treats Michigan’s Republican-looking mail geography as potentially reflecting a voting-method shift.
Method, limitations and source details
- Method
- Synthesis of trackers, commentary and social posts, including Pete Bray and Nathaniel Rakich.
- Limits
- No independent Michigan model. National conclusions do not establish a Michigan advantage, and the suggested switch from mail to in-person voting remains a hypothesis.
- Underlying data date
- Michigan post dated 2026-10-12; essay 2026-10-15
- Author / source family
- Todd Beeton
Commentary synthesis
Michigan Senate Early Vote Model Shows Positive News For Republicans
The October 13 article interprets EDG’s narrowing modeled Democratic advantage as encouraging for Republicans, while describing a toss-up.
Method, limitations and source details
- Method
- Commentary on EDG’s October 12 party allocations, daily changes and displayed 2024 comparison.
- Limits
- A derivative of EDG, not an independent estimate. Its reported daily Republican improvement and slightly more Democratic mix versus 2024 measure different things. Candidate votes and fully aligned historical benchmarks are not established.
- Underlying data date
- 2026-10-12 EDG output
- Author / source family
- Carmine Sabia
EDG
Supporting source 1 ↗
Michigan early vote, day 2
Again reports higher retention of 2024 volume in Harris-won counties, with a narrower gap.
Method, limitations and source details
- Method
- County-winner comparison, labeled day two.
- Limits
- The embed does not establish daily versus cumulative figures. Precise day-to-day turnout claims remain unsupported.
- Underlying data date
- 2026-10-26 post; day 2 comparison
- Author / source family
- @conettakes
Con ET Takes
Supporting source 1 ↗
Michigan early vote, day 1
Reports Harris-won counties retaining more 2024 opening-day in-person volume than Trump-won counties.
Method, limitations and source details
- Method
- County-winner grouping and opening-day comparison.
- Limits
- Verified through HotAir’s embed; complete inputs were not reproduced. County volume is not a vote margin.
- Underlying data date
- 2026-10-25 post; day 1 comparison
- Author / source family
- @conettakes
Con ET Takes
Supporting source 1 ↗
Trump-county mail share and a voting-method caution
Bray called a larger absentee-return share from Trump counties encouraging for Republicans; Rakich suggested Democrats might be switching to in-person voting.
Method, limitations and source details
- Method
- County-winner geography compared with 2024, followed by qualitative commentary on voting methods.
- Limits
- Verified through The Big Picture’s embedded text; original X post inaccessible. Residence does not identify party, and the suggested method shift is not measured. This is a dated October 12 exchange.
- Underlying data date
- 2026-10-12 post
- Author / source family
- Pete Bray; response by Nathaniel Rakich
Pete Bray county geography
Supporting source 1 ↗
Michigan Early Vote 2026
Compares participation across elections and counties without assigning a Michigan party advantage.
Method, limitations and source details
- Method
- Official daily 2026/2024 counts and four weekly 2022 observations; comparisons align days before the election. County rates adjust for registration, and returns are divided by ballots sent.
- Limits
- Historical interpolation is not an observed daily count. The current chart values did not load in this text retrieval, so no fresh numerical conclusion is assigned.
- Underlying data date
- Current values not exposed in retrieved view
- Author / source family
- ProfessorButts
Michigan Department of State
Tracking the 2026 early vote
No current Michigan partisan reading could be independently verified in this review.
Method, limitations and source details
- Method
- The earlier public-methods review identified TargetSmart affiliation estimates where official party registration is absent. TargetSmart describes matching early-vote activity to its voter file.
- Limits
- The NBC page was unavailable in this refresh; Michigan detail previously reached a subscriber gate. Circulating percentages should not be presented as verified NBC findings.
- Underlying data date
- Michigan cutoff unavailable
- Author / source family
- NBC News / TargetSmart
TargetSmart
Supporting source 1 ↗
2026 Early Vote Tracker
Warns that voting earlier does not necessarily add voters and that mail-heavy counts can look Democratic.
Method, limitations and source details
- Method
- civicAPI counts are compared with final 2022 House turnout; national Cooperative Election Study surveys supply voting-method context. Early counts do not enter its forecast.
- Limits
- Current ballots divided by a final total measure completion, not same-stage growth. National survey patterns do not establish Michigan’s party composition, and the current Michigan row was not available in the retrieved text.
- Underlying data date
- Page updated October 29, 2026; Michigan-specific cutoff unavailable
- Author / source family
- Kurt Wirth
civicAPI + national survey context
Michigan Absentee & Early Voting Tracker
Treats the presidential-year comparison cautiously and highlights automatic absentee delivery; it makes no Michigan party estimate.
Method, limitations and source details
- Method
- Official counts aligned by days before Election Day, using daily 2024/2026 series and three rounded 2022 press-release checkpoints. Return rates divide returns by ballots sent.
- Limits
- Sparse, rounded historical checkpoints limit precision. The tracker’s statement that only those 2022 figures exist does not account for the exact weekly workbooks used in our analysis.
- Underlying data date
- Current values not exposed in retrieved view
- Author / source family
- Seth J. Levy
Michigan Department of State
Supporting source 1 ↗
Michigan Early Vote Tracker
Presents requested and returned ballots as participation, explicitly excluding forecasts and counted votes.
Method, limitations and source details
- Method
- Uses civicAPI data with selectable metrics and county tables.
- Limits
- Its default Michigan view displays requests, which are not completed ballots. Sharing a civicAPI feed with other sites does not create independent corroboration; underlying cutoffs require reconciliation.
- Underlying data date
- Displayed count date October 28, 2026; page update October 28, 8:40 PM ET
- Author / source family
- Election Watch
civicAPI
Who has already voted
Describes early participation as a count and cautions against equating earlier voting with additional turnout.
Method, limitations and source details
- Method
- States that comparisons align days before Election Day, with source and freshness labels and downloadable data. It omits party estimates where registration is absent.
- Limits
- The current Michigan row did not appear in this retrieval. The method description has been reviewed, but the Michigan implementation and values remain unverified.
- Underlying data date
- Michigan-specific cutoff unavailable
- Author / source family
- ElectIndex
State data / national fallbacks
Michigan early voting results 2026
Declines to estimate Michigan party affiliation and excludes early counts from its forecast.
Method, limitations and source details
- Method
- Uses daily civicAPI snapshots and dates the counts to the preceding day, rather than reading Michigan’s source directly.
- Limits
- The retrieved page distinguishes an October 29 check from counts labeled through October 27. A refresh timestamp does not establish a matching official data cutoff or complete coverage of both voting methods.
- Underlying data date
- Through October 27, 2026; checked October 29 at 10:35 am ET
- Author / source family
- Call the Map
civicAPI
Michigan early vote tracker
Its modeled electorate favors Democrats, but its October 29 comparison with its own Senate forecast finds no clear additional signal.
Method, limitations and source details
- Method
- civicAPI counts are combined with 1,152 simulated voter types per county, voting-method estimates and survey-based crossover assumptions.
- Limits
- Party is estimated, not recorded. The retrieved page shows conflicting ballot totals and zero in-person ballots. Historical party comparisons may also be modeled. These unresolved discrepancies prevent treating the output as a verified measure of Michigan’s partisan change.
- Underlying data date
- Page labeled October 29, 2026; underlying cutoffs not reconciled
- Author / source family
- On Point Polling
civicAPI + On Point model
2026 General Election Early Vote — Michigan
A participation reference; no current Michigan partisan finding is established here.
Method, limitations and source details
- Method
- Publishes a state-level early-vote compilation used by other reporting.
- Limits
- The retrieved Michigan page displayed no data. National party-registration totals cannot be assumed to represent Michigan or actual candidate votes.
- Underlying data date
- Not available in retrieved page
- Author / source family
- UF Election Lab
University of Florida compilation
Michigan election guide 2026
Detroit’s opening weekend exceeds 2024.
Method, limitations and source details
- Method
- Local totals.
- Limits
- Not statewide.
- Underlying data date
- October 24–25, 2026
- Author / source family
- Joe Guillen and Annalise Frank
Detroit
Turnout reaches 18.8%
Participation update.
Method, limitations and source details
- Method
- Official counts.
- Limits
- No party estimate.
- Underlying data date
- October 27, 2026
- Author / source family
- Paula Wethington
MDOS
Turnout tops 16%
Participation update.
Method, limitations and source details
- Method
- Official counts.
- Limits
- No party estimate.
- Underlying data date
- October 26, 2026
- Author / source family
- Paula Wethington
MDOS
Detroit Breaks New Opening Day Record For Early In-Person Voting
Reports Detroit’s opening-day early vote at 2,305, above 1,897 in 2024.
Method, limitations and source details
- Method
- City clerk announcement comparing opening days.
- Limits
- A city-level observation cannot establish statewide or final turnout. The clerk’s interpretation of voter motivation is not a representative survey.
- Underlying data date
- October 24, 2026 opening day
- Author / source family
- Allan Lengel
Detroit clerk
Michigan voters cast ballots early as absentee returns climb statewide
Describes Detroit voters choosing early voting for convenience and confidence that their ballot will arrive.
Method, limitations and source details
- Method
- Voter and clerk interviews with state dashboard context. The cited 17% increase concerns requests.
- Limits
- Interviews illustrate individual experiences, not a representative turnout or partisan estimate. The headline varies between the article and publisher archive.
- Underlying data date
- October 24 reporting; exact statewide snapshot time unstated
- Author / source family
- Faraz Javed and Brian Schwartz
WXYZ reporting / Michigan Department of State
Newsletter: nearly 790,000 have voted
Participation nears one-fifth of final 2022 turnout.
Method, limitations and source details
- Method
- UF count comparison.
- Limits
- Completion, not pace.
- Underlying data date
- October 20, 2026
- Author / source family
- David Adkins, Herb Scribner and Joe Guillen
UF
More than 860,000 Michigan absentee ballots returned ahead of Nov. 3 election
Documents rising completed participation and widespread permanent absentee enrollment.
Method, limitations and source details
- Method
- October 21 dashboard totals, an October 13 comparison and county examples.
- Limits
- Reporting changes do not necessarily identify the day ballots were cast. Syndicated versions are one reporting item, and no partisan estimate is offered.
- Underlying data date
- October 21, 2026, 7:05 a.m. dashboard; early votes through October 20
- Author / source family
- Crystal Huggins
Michigan Department of State / Hearst
More than 780,000 Michigan voters have cast their ballot two weeks before Election Day
Reports 789,476 absentee returns and 2,121,458 requests. The widely repeated 17% increase concerns requests.
Method, limitations and source details
- Method
- Official dashboard totals and a linked 2022 comparison.
- Limits
- The cited 2022 worksheet is dated October 24, fifteen days before that election, versus fourteen days for October 20, 2026. These totals cannot identify candidate support.
- Underlying data date
- October 20, 2026, 7 a.m.
- Author / source family
- MDOS
Michigan Department of State
Three weeks before Election Day, 6% of voters have already cast their ballot
Returns were only slightly above the cited 2022 checkpoint while absentee requests had grown much faster.
Method, limitations and source details
- Method
- Official weekly comparison: 440,677 returned ballots versus 432,960 in 2022, alongside requests and permanent-list enrollment.
- Limits
- Automatic delivery changes the request pool. The release reports administrative participation, not voter preferences.
- Underlying data date
- October 13, 2026, 7 a.m.
- Author / source family
- MDOS
Michigan Department of State
Absentee ballot returns in Michigan up by 30% compared to 2022
Reports the state’s four-week rise in absentee returns and notes the new permanent-list context.
Method, limitations and source details
- Method
- MDOS figures plus Mark Grebner’s explanation of automatic delivery in this gubernatorial cycle.
- Limits
- The same underlying official counts also appear in other reports; this is not an independent partisan model.
- Underlying data date
- October 6, 2026 morning figures
- Author / source family
- Michael Arney
Michigan Department of State
Four weeks before Election Day, mail-ballot returns are up 30% over same point in 2022
Reports absentee returns 30% above its four-week 2022 comparison.
Method, limitations and source details
- Method
- Official administrative counts: 201,909 returns on October 6, compared with 155,142 at the cited 2022 checkpoint.
- Limits
- A dated participation comparison, not a partisan estimate or evidence that the initial growth rate will persist.
- Underlying data date
- October 6, 2026, 7 a.m.
- Author / source family
- MDOS
Michigan Department of State
Michigan absentee ballot returns up 30% compared to same time period in 2022
Reports the state’s 30% increase in absentee returns at the four-week checkpoint.
Method, limitations and source details
- Method
- October 6 official figures and state statements.
- Limits
- Shares the underlying state release with WKZO. It does not independently estimate partisan turnout or votes.
- Underlying data date
- October 6, 2026, 7 a.m.
- Author / source family
- Natalia Escalante
Michigan Department of State
No sources match these filters. Try a broader search or clear the filters.
Reporting notes
What this collection covers—and what remains unverified.
We reviewed the prior research catalog and rechecked available original sources on October 29. The collection includes published analysis, public trackers, official updates and relevant social commentary. It excludes our own reporting from the outside-source count.
Titles identify the linked articles, trackers or discussions; some are shortened for clarity. The summaries and method labels are ours. Links may change, require a subscription or display a newer snapshot. Access limitations are shown rather than filled with inferred conclusions.
This guide is not a ranking of forecasters. Source quantity is not a vote for a party, and incompatible estimates should not be averaged.
Download the reviewed source catalog ↘Additional leads and access limitations
New York Times — battleground early-voting tracker
Original page unavailable for review. No Michigan conclusion or method is attributed without access.
Source link ↗New York Times — early-voting data limitations
Located through Quartz. Original article inaccessible; not treated as a fully reviewed item.
Source link ↗Decision Desk HQ — Michigan early voting
Subscriber product. No validated Michigan conclusion available from the public preview. Governor and secretary-of-state wrappers are not independent studies.
Source link ↗WJR — Detroit Voters Head to the Polls in Record Early Numbers
Public page contains a brief participation narrative and embedded podcast. The interview audio and a transcript were not reviewed; do not attribute a partisan model or substantiate its record-turnout wording from the listing.
Source link ↗Pete Bray — later Michigan geography posts
Third-party index surfaced a later comparison with 2022 and 2024. Original permalink/chart not verified, so its numerical conclusions are not included as confirmed findings. October 12 embedded discussion is separately cataloged.
Source link ↗umichvoter — Michigan opening-day observation
Third-party index surfaced an opening-day turnout observation. Original post and complete method were not confirmed. A total alone would be descriptive reporting, not a party forecast.
Source link ↗Reddit — Analysis of first day of in-person early vote in Michigan
User-posted county comparison chart. Chart inputs and timestamp were not fully verified. This is a Reddit user’s work, not reporting published by The Bulwark.
Source link ↗Related national and administrative context
Tom Bonier — Who’s Voting? — October 16. Its party-registration comparison covers 15 states; it does not establish a Michigan partisan trend.
Tom Bonier — Early Notes on Early Votes — October 10. Useful background on early-voter composition; no verified Michigan-specific result from the accessible portion.
Quartz — early-voting data explainer — October 9. Summarizes limitations; cites New York Times reporting. Not an independent Michigan model.
The Guardian — early-voting takeaways — October 23. National context; no verified Michigan party-specific result.
Axios Detroit — readers’ midterm feelings — October 26. Self-selected reader responses about voting and feelings, not a representative survey or returns analysis.
University of Michigan CLOSUP — early-voting strain — Local-official survey about administrative experience and capacity. Background on adoption, not a current party-vote estimate.