The 1894 Michigan Veteran Census: What It Is and How to Use It

What this census is, and what it isn’t

It’s a name-by-name state census of every living Civil War soldier residing in Michigan on June 1, 1894, published two years later as Volume III of the Census of the State of Michigan, 1894, under the title “Soldiers, Sailors and Marines.” Washington Gardner, then Michigan’s Secretary of State, compiled it under authority of an act of the legislature approved May 31, 1893, and transmitted the finished volume to Governor John T. Rich in June 1896.

I’ve pulled this volume more than once while chasing a Michigan ancestor who wasn’t showing up anywhere else for the early 1890s, and it’s worth understanding why it exists at all. Michigan periodically ran its own state censuses between the federal ones (1854, 1864, 1874, 1884, 1894, and 1904 were all state census years), and 1894 happened to land at a moment when Union veterans were aging and dying at increasing rates. That made their numbers a matter of real interest to state pension boards and Grand Army of the Republic posts alike. Counting them by name wasn’t an afterthought; it was the entire purpose of the volume.

The numbers give a sense of scale. The published aggregate lists 42,544 United States soldiers of the Civil War living in Michigan that June — 34,946 native-born, 7,598 foreign-born. A separate table counts 359 U.S. sailors and marines, and, almost as a footnote, 148 Confederate soldiers then residing in the state (102 native-born, 46 foreign-born) and 41 self-reported Mexican War veterans. The average age of the Civil War soldiers counted was 55.4 years.

Why this record exists separately from the federal 1890 count

The short version: the federal government tried something similar in 1890, and most of what it collected didn’t survive.

The 1890 federal census included a special schedule, officially titled “Surviving Soldiers, Sailors and Marines and Widows, etc.,” covering nearly 75,000 households of Union veterans and widows nationwide. The Pension Office had requested it specifically to help veterans locate comrades for pension testimony and to gauge how many survivors and widows remained. That schedule mostly survives; it’s the general population schedules of the 1890 census that suffered the catastrophic loss. A fire broke out in the basement of the Commerce Department building in Washington on January 10, 1921, and destroyed roughly a quarter of the 1890 census volumes outright, with most of the rest damaged beyond practical use by water and smoke. Congress eventually authorized destruction of what remained in 1933.

Michigan’s own veterans schedules from that federal 1890 count do survive and are reproduced on National Archives microfilm publication M123. But they’re a single snapshot, taken by federal enumerators working from a national instruction sheet, and they cover the population only as of 1890. Michigan’s 1894 state census, taken independently four years later under a state law, is a second data point — one gathered by different people, at a different moment, using a Michigan-specific instrument built for this exact purpose. If your ancestor’s name is misspelled, missing, or contradicted in one record, the other is often where you catch the error.

What a record in the 1894 census contains

The published volume is organized as statistical tables by county, broken out by nativity (native vs. foreign-born), age range, and marital status, with the underlying enumeration sheets naming each soldier. A typical entry, drawn from the original schedules, gives:

• Soldier’s name, as reported to the enumerator
• Township or ward and county of residence, June 1, 1894
• Age
• Whether native-born or foreign-born
• Marital status (single, married, widowed, or divorced)
• Any disease or disability reported as temporarily or permanently disabling him from ordinary business on the day of the enumerator’s visit

That last field is a detail people miss. Table 8 of the published volume specifically tabulates soldiers who were “sick or disabled” at the moment the census taker showed up — 6,105 men statewide, just over 14 percent of those counted. If you’re trying to establish a health condition around a pension claim’s timeline, that field is worth cross-checking against the claim file itself.

How to read and verify an entry

1. Confirm the county and township, not just the name. Michigan had dozens of townships that shared or nearly shared names across counties in the 1890s; a “John Miller” in the wrong county isn’t your John Miller.
2. Check for bracketed names. Gardner’s office sent the completed list back to enumerators and to Grand Army of the Republic posts in each county in March 1896 for correction before final printing. Names added at that stage, rather than caught in the original 1894 enumeration, appear in brackets in the original volume — a signal the record went through a second round of verification.
3. Cross-reference against the county’s GAR post roster if one survives. Local posts were the ones asked to vouch for or add names, so a soldier missing from the state census but present in a post’s own membership book may still show up once you check that source too.
4. Don’t stop at one census year. If your ancestor doesn’t appear in 1894, check whether he’d already died, moved out of state, or was simply missed — then try the federal 1890 veterans schedule or a pension file to bracket the years around it.
5. Treat the age and marital status fields as approximate. Self-reported ages in 19th-century censuses drift by a year or two routinely; use them to narrow a birth year range, not to fix one.

Where to find the 1894 census now

The full published volume, including the county-by-county statistical tables and the underlying name lists, is digitized and freely browsable through the Internet Archive and Wikimedia Commons — page images scanned from the original held by Cornell University Library. Ancestry hosts a separately indexed version of the broader 1894 Michigan state census (not limited to veterans) as part of its state census collections, and the Library of Michigan’s census records research guide is a good next stop for the other Michigan state census years if 1894 alone doesn’t resolve your search.

Worth knowing: the veterans’ volume is only one piece of the larger 1894 state census. That same enumeration also recorded Michigan’s general population (households, ages, occupations, birthplaces) across every county, separate from the federal count. The originals from that count, and from Michigan’s other state census years, are held by the State Archives of Michigan in Lansing. The soldiers-sailors-marines volume was compiled as a special supplement to that population count, using the same statewide enumeration infrastructure, which is part of why it’s as complete as it is.

Frequently asked questions

How many soldiers were counted in the 1894 Michigan veteran census? The published volume gives an aggregate of 42,544 United States soldiers of the Civil War residing in Michigan on June 1, 1894 — 34,946 native-born and 7,598 foreign-born. A separate count of 359 U.S. sailors and marines and 41 Mexican War soldiers appears in the same volume, along with 148 Confederate soldiers then living in the state.

Why does this Michigan census matter more than the federal 1890 veterans schedule? Because the federal schedule barely survived. The 1890 federal census included a special enumeration of Union veterans and widows, but roughly 25 percent of the schedules burned outright in a January 1921 fire at the Commerce Department, and most of the remainder was damaged beyond use. Michigan’s schedules for that federal count do survive, but the state’s own 1894 census gives a second, independently gathered snapshot of exactly the same population four years later, filling gaps the federal record can’t.

Does the 1894 census list Confederate soldiers or widows? It lists 148 Confederate soldiers living in Michigan at the time, broken out separately from the main Union soldier count, but it was not designed as a widows’ schedule. If you’re chasing a widow’s record instead of a living veteran, a federal pension file is the better source.

How accurate are the names in the 1894 census? Imperfect, by the compilers’ own admission. Washington Gardner, the Secretary of State who oversaw publication, sent the completed list back to enumerators and to local Grand Army of the Republic posts in 1896 for correction before printing, specifically because names had been misspelled or missed. Corrected or added names appear bracketed in the original volume. Treat a name that doesn’t match your ancestor’s usual spelling as a lead worth checking, not a dead end.