When books leave
Of 352,551 books that passed through the stream between 2021-04 and 2026-07, 333,355 (94.6%) carry a complete record: the year they were printed and the month they left a household for good. A quarter were gone by age 12. Half by 20. A quarter outlasted 29 years on somebody’s shelf.
The fate fork
Demand favors the young. Books rescued into new hands through resale have a median age of 17 at exit; books released into the reuse & recycling track run 22. Overall, 32.1% of the whole stream (143,631 items) found a next reader.
Age at exit by decade of printing
| Printed in | books observed | median age at exit |
|---|---|---|
| 1920s | 116 | 102.0 |
| 1930s | 163 | 87 |
| 1940s | 261 | 77 |
| 1950s | 1,007 | 68 |
| 1960s | 2,635 | 58 |
| 1970s | 13,912 | 48.0 |
| 1980s | 35,318 | 38.0 |
| 1990s | 80,216 | 28.0 |
| 2000s | 102,678 | 19.0 |
| 2010s | 81,413 | 9 |
| 2020s | 15,519 | 2 |
Read this column honestly: the observation window is fixed (2021–2026), so a 1940s book that exits must be old — this table describes the ages of the leaving, not the survival odds of the printed.
Methods, plainly
- What this is: a death register. Every record is a real physical item scanned at intake by one Albuquerque donation operation, 2021–2026 — 447,734 items, of which 352,551 are books.
- What this is not: survival curves. There is no census of living books at risk, so these are age-at-exit distributions of the observed dead — the tradition of John Graunt’s 1662 Bills of Mortality, three centuries before Kaplan–Meier.
- Birth years join from Library of Congress and Open Library open records by the book’s own identifier; 94.6% of books joined. Earliest recorded print year wins, so ages skew, if anything, older.
- “Released” means an item left the resale stream into the reuse/recycling track. When a title is over-printed, supply exceeds demand and recycling is the honest last stop; nothing here is a per-item destruction claim.
- Selection bias, stated: this is a household donation stream seen through a resale operation’s intake — not library circulation, not bookstore sales, not a probability sample of shelves.
- Privacy: the register contains no donor names, addresses, or dates finer than the month. It records objects, never people.
- Data: aggregates and the monthly register are CC BY 4.0 — summary.json. Cite: New Mexico Literacy Project, The Life Tables of Print (2026).
Walk the names themselves on the Book Necrology →
Questions people ask
How long does a book live in a home?
Measured across 333,355 books with complete records from a real donation stream (2021–2026): the median book leaving an Albuquerque home is 20 years old. A quarter leave by age 12; a quarter last past 29.
Are these survival curves?
No — and the methods say so plainly. There is no census of living books at risk, so these are age-at-exit distributions of the observed dead, in the tradition of John Graunt's 1662 Bills of Mortality, not Kaplan–Meier survival estimates.
Where does the data come from?
Every item is a physical book scanned by one Albuquerque donation operation at intake, 2021–2026. Print years join from Library of Congress and Open Library open records; 94.6% of books joined. The aggregate data is CC BY 4.0.