The reading
The entry wave is real and it is locked in. 4.79 million Americans turn 21 in 2028, more than in any year on record. Everyone in that number has already been born.
The wave is smaller than it sounds. 45.2 million people turn 21 between 2027 and 2036. The previous decade delivered 44.3 million. That's 2.2% more bodies over ten years.
A new legal drinker is worth about 46 cents on the peak-age dollar. Two datasets built from different inputs land in the same place.
The good demographic news belongs to Millennials. The 40 to 59 band, where household alcohol spending peaks, grows 18.8% between 2026 and 2046.
The gate opens widest in 2028
Everyone turning 21 in a given year was born 21 years earlier, which makes this curve unusually knowable. No survey, no model of intent. Just arithmetic on people who already exist.
The peak is 2028, at 4.79 million. It never gets back there.
The headline hides the size of it. 45.2 million Americans turn 21 between 2027 and 2036. The ten years before that delivered 44.3 million. Call it 0.2% a year of extra bodies. Real, and small.
The good news is 44 years old
The 40 to 59 band grows from 70.7 million in 2026 to 84.0 million in 2046, with no peak inside the window. Household spending on alcohol peaks in that band. The tailwind in this category is already in its forties.
The catch
Gallup puts US adult drinking at 54% in July 2026, the lowest reading in about 90 years of asking. It was 62% in 2023.
The cohort walking through the gate is the least likely of any to drink. Demography hands the category 0.7% a year, and a two-point move in participation spends all of it.
Build the entry ladder on purpose, and stop discounting the flagship
The entering cohort buys small. Formats under 375ml grew 26% in two years while the 750ml fell 7.2%. A 200ml at full price per ounce recruits a drinker. A discounted 750ml teaches a price you'll spend five years trying to unteach.
Plan revenue against 45 and recruitment against 25
The 40 to 59 band spends 1.3 times the household average. Gen Z spends 0.70 times. Budget them separately, so spend against a cohort that converges in the 2040s isn't judged on this year's revenue.