Guides
Part of People by US city: records, categories and updates for 2027
21 people by US city facts with sources and useful context
People by US city facts: a dozen custodians hold the municipal record, each with its own mandate, retention rule and access rule, and none holds all of it.
Cities keep more records about people than any other level of American government, and they keep almost none of them in one place. The single most useful thing to know about city research is which office holds what, because a document you cannot find is usually a document you asked the wrong custodian for.
Below are twenty one points that decide whether a city attribution survives contact with the records.
What to take away
- There is no municipal archive in the general sense. There are a dozen custodians with separate mandates, separate retention rules and separate access rules.
- A city is a legal entity with a start date and a changing footprint. Records before incorporation and outside the line are somewhere else.
- Most "famous residents" claims made by cities are promotional, and the standard behind them is usually not published at all.
Who holds what
The city clerk holds the city's own acts: council minutes, ordinances, licenses, and often the register of municipal officers. This is a record of the government, and it touches individuals only where they dealt with it.
Deeds and mortgages are usually a county function, held by a recorder or register of deeds, not by the city. A property chain crosses the city line without changing custodian.
Vital records are state and local, and which local office holds them varies by state. City health departments in some places, county offices in others, and the split has changed over time within a single state.
Death investigation is its own office, a coroner or medical examiner, with its own files that are separate from the death certificate and often far more detailed.
Courts index by case, not by person. A person appears inside a case file. Finding them means finding the case first, through a docket, a judgment index or a plaintiff and defendant index, and each of those covers a different span.
Tax assessment rolls and building permits are the most underused city records there are. They place a person at a property year by year, and permits name the owner, the builder and sometimes the occupant.
Voter registration rolls are public in many jurisdictions and restricted in others, with different rules for current and historical registers. Treat them as a source with an access question attached rather than an open file.
Schools keep records under district retention schedules, and districts consolidate and dissolve. A closed school's files may sit with a successor district, a county office, or a historical society, and nobody will tell you which unless you ask all three.
Congregational records belong to the congregation, which may have merged, moved or closed. Denominational archives collect some of these and not others, and access is at the archive's discretion.
Cemetery records are held by the cemetery, which is frequently a private company or an association. The burial register, the plot map and the monument can disagree, and the monument is the least reliable of the three.
Hospital records are closed. A hospital appears in this research as a place of an event on somebody else's document, never as a source you can consult.
Local historical societies keep vertical files: clippings, photographs, and correspondence filed by subject rather than cataloged item by item. They are invisible to catalog search and often hold the only surviving copy of something.
What the city itself is
- Incorporation has a date. Before it, the place was a township, a village, an unincorporated settlement or nothing at all, and the records are with whatever body governed it then. The city's own files start when the city does, and the entry in the Geographic Names Information System records the place rather than the government.
A city plan from one year shows the boundary of that year, which is the boundary the records of that year were filed under.
City and county can be the same government. Consolidated city-county jurisdictions collapse two custodians into one, which changes where you look and what survives.
Wards and precincts were record geographies, and they moved. An enumeration or a poll book organized by ward requires the ward map for that year, not the current one.
Streets are renamed and renumbered, sometimes across a whole district at once. An address in an old source may not exist now, and the modern address at that number may be a different building. Historic fire insurance maps and old plat books are the usual way back.
Neighborhood names have no legal boundary and no custodian. They are real to residents and unusable as a data field, because no two sources draw them the same way.
City hall losses are documented. Fires, floods and simple disposal destroyed identifiable series in identifiable places. Ask the custodian what survives before drawing anything from a gap. The same reasoning about absence appears on the state facts page.
What is not evidence
A city's claim on a person is marketing. Welcome signs, hall of fame inductions, honorary titles and "notable residents" pages apply whatever standard suits the occasion, and the standard is rarely written down. Cite the claim to the body that made it.
A local newspaper mention establishes that a paper published something, which is a fact about the paper, and library guidance on reading a digitized newspaper run is explicit about how little more it establishes. Small papers ran items supplied by families and by publicists without checking them, and the same item was syndicated onward. This is how one unverified line becomes a dozen apparent sources.
Do not publish a living person's neighborhood, street or current city. Where someone has publicly described where they live, attribute the statement and date it, and stop there. Precise location information about a living person is a safety matter before it is an accuracy matter, which is why the city pillar treats it as a hard rule rather than a preference.
Assembling a city attribution
Work outward from the event rather than inward from the place:
| Step | Question | Where it lands |
|---|---|---|
| 1 | What event am I placing? | Birth, schooling, residence, work, burial |
| 2 | What document records that event? | The custodian follows from the document |
| 3 | What did the place look like on that date? | Boundaries, ward, street name, incorporation status |
| 4 | Which office holds it now? | Successor bodies, archives, private holders |
| 5 | What survives? | Ask, and record the answer including the gaps |
Done in that order, the place falls out of the document. Done in reverse, you pick a city first and then hunt for support, which is how weak sources get promoted.
Bottom line
City research is custodian research. Name the event, name the document, name the office, and ask what survives. Everything else, including every list of famous residents you will find published by a city, is downstream of somebody having done that or having skipped it.
Common questions
A directory and a tax roll disagree about an address. Which wins?
Neither. Record both with their dates. The directory is a canvass reported by whoever answered the door, described on the directory page; the tax roll is an assessment against an owner, who may not be the occupant. They can both be accurate about different things.
Can a birth be attributed to a city from a hospital name alone?
It attributes the birth to the hospital's location, which is often not where the family lived and sometimes not even in the same county. Treat the hospital as a fact about the event, not about the person's origins, in the same way that maps built from these fields show records rather than lives.