How Old Are the Books America’s Schools Assign?

The assigned canon is younger than you think

Ask most people what schools assign and they picture a shelf of classics — To Kill a Mockingbird, Shakespeare, The Great Gatsby. Those books are on the lists. But they are not what a typical American student is handed. The median book across 13,250 source-cited assignments was first published in 2012. Half of everything states assign is newer than that.

More than a third of assignments — 36.4% — are books published in the last ten years (2016 or later): 4,823 assignment rows spread across 671 distinct recent titles. The mean publication year is a little older, 2006, because a thin tail of genuine classics (the oldest assigned title is Romeo and Juliet, 1597) pulls the average back. The median ignores that tail, which is exactly why it’s the honest headline: it describes the middle of the list, not its edges.

What states assign, by decade of publication

The mass sits recent. The 2010s alone account for 38.75% of all assignments; the 2000s and 2020s add most of the rest. Everything before 1920 — every Shakespeare play, every 19th-century novel combined — is 0.6% of the list.

“Distinct titles” counts each book once; “assignments” counts each source-cited list appearance, so a book assigned in many states counts many times. The full per-decade file (back to the 1590s) is in the download below.

The classics that survive

A pre-1980 canon still exists — it’s just small and specific. Books published before 1950 are 4.9% of the catalog (82 distinct titles) but only 2.5% of assignments. And the pre-1980 books that doget assigned most are not the high-school syllabus classics — they’re mid-century children’s and middle-grade staples:

The books replacing them

The most-assigned books of the last decade look nothing like the classics — they’re contemporary, diverse, and graphic-novel-heavy middle-grade titles. This is the canon actually being built, in real time, on the lists:

Every grade gets a modern book — even high school

The intuition is that little kids read new picture books and high schoolers get the old classics. The data says the opposite: high schoolers are assigned the newest books of any band — a median publication year of 2015, newer than the elementary and middle-grade medians. Every band’s median lands in the 2010s.

What this does and doesn’t say

• “Publication year” is the work’s first publication, not the assigned edition. Romeo and Juliet is dated 1597, not the year of the paperback in a backpack. That is the right frame for “how old is this book” — but it means a modern translation or reissue of an old work still counts as old.
• This is assignment-weighted. A book assigned in 40 states counts 40 times, because the question is what a student is typically handed, not how many distinct titles exist. The distinct-title view (each book once) has a longer classic tail, and the decade table shows both columns so you can see the difference.
• Newer isn’t a verdict. A contemporary median says nothing about whether the books are good, rigorous or durable — only about when they were written. Many will not still be assigned in twenty years; some of the 1970s titles here have already lasted fifty.
• Coverage is 99.9%. 8 of 13,258 assignment rows have no parseable publication year and are excluded — never guessed. At 99.9% the excluded slice cannot move the medians.
• These are lists, not classrooms. The corpus is what states and districts publish as recommended or required reading. What any individual teacher assigns may differ.

Methodology

Every figure on this page is computed from ReadingList.school’s assignment corpus: 13,258 rows, each carrying a source_url and source_document pointing at the state or district reading list it came from. Nothing here is estimated, modelled or inferred.

1. Join each assignment row to its book’s first_published year and (for the grade-band table) to its grade band (100% grade coverage).
2. Count a publication year only when it parses to a plausible four-digit year (1400 < year ≤ 2026). Nulls and garbage are excluded, not guessed. This leaves 13,250 rows (99.9%) across 1,687 distinct books.
3. Take the median publication year, assignment-weighted (a book assigned in twelve states counts twelve times). The median leads because a long classical tail pulls the mean back ~6 years; both are published so the skew is visible.
4. Bucket by decade of first publication for the distribution, and by grade band for the age-by-grade table.

The generator is committed and runnable: scripts/compute-how-old-are-assigned-books-2026.py. It emits this page’s data file and the CSV in one pass, so the two can never disagree. Figures are point-in-time as of 2026-07-19 and are committed rather than recomputed per build, so a number quoted from this page stays quotable.

Download the data

Free to use, republish and chart with attribution (CC BY 4.0).

• how-old-are-assigned-books-2026.csv — every decade of first publication, with distinct-title counts, assignment counts and each decade’s share of the list.

Cite this study

Explore the lists behind the numbers

Every figure above comes from lists you can read:

• The most-assigned books in America’s state reading lists — the companion study: which books, not how old.
• How long are the books American schools assign? — the sibling study: page length by grade, and the 3rd-grade cliff.
• Browse every state and grade — the full corpus this study is computed from.

A Figures Bureau study. Data CC BY 4.0 — republish it, chart it, quote it; a link back to this page is all we ask. Last verified 2026-07-19.