Participation measured against the same-year citizen voting-age population. Audited cycles count total general-election participants — ballots or participating voters, not presidential candidate votes. Cycles before 2016 predate that series and are derived from the certified presidential returns instead; they are marked everywhere they appear and are excluded from the figures below.
2004–20248 cycles · 5 audited
2024 turnout
86.2%3,894 participants ÷ 4,516 CVAP
2020→2022 drop-off
+17.2 ptsVolatile
Drop-off vs state
+2.1 ptsPositive means a larger decline than the statewide pattern
Mechanical 2026
69.0%–72.4%Recent-cycle and two-cycle persistence scenarios; not a forecast
Participation by cycle
Presidential Midterm Derived, pre-2016
0%20%40%60%80%100%
73.6%
70.4%
20042008201220162018202020222024
Select a presidential cycle to show it on the map and results ledger above.
Cycle
Participants
CVAP
Turnout
Series
Quality
3,894
4,516 ± 307
86.2%
Audited
E/A
2022Mid
3,155
4,481 ± 312
70.4%
Audited
E/A
3,776
4,308 ± 319
87.7%
Audited
E/A
2018Mid
2,959
4,019 ± 214
73.6%
Audited
E/A
3,236
3,854 ± 229
84.0%
Audited
E/A
2,568
3,900
65.8%
Derived
—
3,032
3,650
83.1%
Derived
—
2,889
3,400 extrapolated
85.0%
Derived
—
Sources and methods
Audited cycles (2016, 2018, 2020, 2022, 2024) count total general-election participants against the same-year CVAP special tabulation, each cell carrying its own numerator and denominator quality grade. Derived cycles (2004, 2008, 2012) come from the certified presidential returns on this page over the ACS 5-year CVAP series, so they measure presidential votes cast rather than ballots — a floor on participation. Where a derived cycle predates the ACS its denominator is extrapolated rather than read, and says so on its own line below. They are shown for the longer arc and are excluded from the drop-off, excess-drop-off, and 2026 scenario figures above.
2024: numerator source 160, denominator source 157; estimated total ballots; preferred-county:block-reference-presidential-or-cvap-weight:v2; allocation share 100.0%; series 323/229.
2022: numerator source 160, denominator source 157; estimated total ballots; preferred-county:block-reference-presidential-or-cvap-weight:v2; allocation share 100.0%; series 322/184.
2020: numerator source 160, denominator source 157; estimated total ballots; preferred-county:block-reference-presidential-or-cvap-weight:v2; allocation share 100.0%; series 321/4.
2018: numerator source 160, denominator source 157; estimated total ballots; preferred-county:block-reference-presidential-or-cvap-weight:v2; allocation share 100.0%; series 320/139.
2016: numerator source 160, denominator source 157; estimated total ballots; preferred-county:block-reference-presidential-or-cvap-weight:v2; allocation share 100.0%; series 319/133.
2012 (derived): Certified presidential returns · ACS 5-year CVAP (B05003); presidential votes cast — a floor on participation, not a ballot count.
2008 (derived): Certified presidential returns · ACS 5-year CVAP (B05003); presidential votes cast — a floor on participation, not a ballot count.
2004 (derived): Certified presidential returns · extrapolated CVAP denominator; presidential votes cast — a floor on participation, not a ballot count; 2004 predates the ACS, so the denominator is a linear back-extrapolation from the 2008 and 2012 CVAP values, not a read figure.
presidential history
Party record book
Chelsea, MI · the strongest Democratic and Republican presidential performances recorded here, measured three ways.
2004–2024 · 6 presidential cycles
Highest vote share
Largest share of the presidential candidate vote cast here.
Democratic
Kamala Harris2024
66.2%
2,578 votes · D+34.6
Runner-up · Joe Biden, 2020 · 66.0%
Republican
George W. Bush2004
48.6%
1,403 votes · D+2.6
Runner-up · Mitt Romney, 2012 · 42.4%
Most votes
Largest number of votes cast for the party’s presidential nominee here.
Democratic
Kamala Harris2024
2,578votes
66.2% of the vote · D+34.6
Runner-up · Joe Biden, 2020 · 2,475 votes
Republican
George W. Bush2004
1,403votes
48.6% of the vote · D+2.6
Runner-up · Donald Trump, 2020 · 1,233 votes
Best relative to the U.S.
This Place’s margin against the national margin in the same election — positive when it leaned toward the party more than the country did.
Democratic
Kamala Harris2024
+36.0pts D vs U.S.
Place D+34.6 · U.S. R+1.5
Runner-up · Joe Biden, 2020 · +28.7 pts
Republican
George W. Bush2004
−5.1pts R vs U.S.
Place D+2.6 · U.S. R+2.5
Runner-up · Mitt Romney, 2012 · −9.7 pts
U.S. House elections · 2 House
Senate and governor results are in the Office selector at the top of this page.
U.S. House
Each row shows U.S. House votes counted inside this Place. “Vote lead” is the local lead here, not necessarily the winner of the full contest.
Chelsea · which candidate this place backed in each party’s presidential nominating contest, cycle by cycle. A primary is a contest inside one party, so it has no Democratic–Republican margin and is not part of the general-election record above.
Akashic reports votes cast, not delegates awarded. The two can disagree: Clinton led the 2008 national primary popular vote by less than half a point and did not win the nomination.
Primary returns are published at county grain. The figures below are Washtenaw County, Michigan, the county containing Chelsea — Chelsea’s own primary vote sits inside those totals and is not separately reported.
Presidential nominating contests in Chelsea, by party and cycle.
○no contest record — Akashic holds no result for that party and cycle here. It is not a claim that no contest was held. Telling apart a contest that was cancelled, one that went uncontested, and one simply not yet ingested needs a nominating-contest calendar, which is a separate ingest from results and does not exist yet.
Republican presidential primaries are not yet ingested — the Republican series runs from 1856 and is in scope. The row is kept so its absence is visible: a Democratic-only panel would otherwise read as though no one else held a contest.
Where a cell reads All others, no named candidate led — the source’s residual bucket did. That bucket holds every name past the ones the contest itemises, including candidates who had withdrawn but still drew votes. It is a total, not a person, so it carries no candidate colour.
Akashic carries 10 states for 2012 and 3 states for 2024. Cycles in which an incumbent president seeks renomination are thin by nature — few states run a contested primary — so a sparse column is politics, not a hole in the record.
County returns carry a leader and a lead over the runner-up, and nothing more. The full candidate field with vote shares is published at state grain — follow a cycle above to the state contest for it.
Chelsea sits in the western edge of Washtenaw County, buffered from Ann Arbor's university influence, giving it a more competitive partisan profile than its neighbors despite a similarly college-educated small-town demographic.
Across the recorded series it reached a Democratic high of 34.6 points in 2024. Between 2020 and 2024 the city moved 1.5 points toward the Democratic candidate; the 2024 margin was 34.6 points.
A population of 5,426, a 96% non-Hispanic-white share, and a median household income of $72,098 describe the city. The city's voting pattern over the last decade is most similar to that of Saline and Estes Park.
Akashic
Political twins — cities
The cities whose fingerprint — 2000–2024 trajectory, demographics, ancestry, religion — sits closest to this one, and why.
Twin score (0–100) = weighted cosine between tier-standardized fingerprints: presidential margins 2000–2024 with 12-yr trend and elasticity, ACS income, age, education, population and race, top reported ancestries, and religious adherence — chips name the closest-shared dimensions and the widest gap. Re-weight the groups in the twins explorer.
In 2024, Chelsea, Michigan voted Democratic by 34.6 points (D+34.6), carried by the Democratic candidate. Out of 3,894 votes cast, 2,578 went Democratic and 1,232 went Republican.
How many people live in Chelsea, Michigan?
Chelsea, Michigan has a population of 5,426 according to the 2024 American Community Survey 5-year estimates from the US Census Bureau.
What is the median household income in Chelsea, Michigan?
Median household income in Chelsea, Michigan is $72,098 — below the national median of $80,734. The Michigan state median is $72,875.
What is the political history of Chelsea, Michigan?
Akashic tracks 6 presidential elections in Chelsea, Michigan from 2004 to 2024. Of those, 6 went Democratic and 0 went Republican.
