"Half your age plus seven" is a rule of thumb about age gaps: halve the older partner's age, add seven, and the result is said to be the youngest the other partner should be. Read as a floor that works in both directions, it takes in between 86.0% and 98.7% of the 24,688 marriages registered in Singapore in 2025, and about 95.4% if ages are spread evenly within SingStat's five-year groups. The 86.0% assumes nobody in the 60-and-over groups is older than 90. Read the way Frederick Locker printed it in 1879, as the ideal age for a wife, the rule misses badly. His ideal bride for 2025's median groom of 32.0 would be 23.0, and the median bride was 30.4.
Most marriages registered in Singapore in 2025 fall within the band's orbit
Calculators ranking for this rule treat it as a band: half a person's age plus seven below, and the sum run backwards above (subtract seven, then double). We applied that reading to whichever partner is older.
SingStat's table M830164, last updated 10 July 2026, counts marriages by each partner's five-year age group. We checked every whole-year age combination in each cell: all passing is certainly inside, none passing certainly outside, the rest mixed.
Two figures barely move. In 2025 at most 98.7% of marriages sat inside the band and at least 1.3% sat outside, under every bound we tried for the open-ended youngest and oldest groups. The lower bound is softer: 86.0% with the oldest capped at 90, 85.6% at a cap of 113. Of the couples our even-spread estimate puts outside, about 970 have the older groom and fewer than 200 the older bride. The unit is registered marriages, civil and Muslim combined, remarriages included. We did not examine residence.
Clear misses of the band waned through the 2010s
The same method runs back to 1983, when the lower bound was 54.9%. Much of its rise since is resolution rather than behaviour: 8,810 couples sat in mixed cells in 1983 against 3,127 in 2025, so the 1983 figure is only a bound.
Couples certainly outside were 4.5% of marriages in 1983 and 5.3% in 2010, then 1.6% in 2020 and 1.3% in 2025. The most that could be inside went from 95.5% to 98.7%. That is coincidence with a band, and says nothing about how anyone chose a partner.
Sightings of the rule in the books we opened date from 1879
The earliest printed version we found is in Patchwork by Frederick Locker (later Locker-Lampson), London, 1879. On page 88 of the Internet Archive scan he writes: "A wife should be half the age of her husband with seven years added." His worked examples all check, including a lady of 57 who should marry nobody "less than a century". The passage is comic, and Locker gives it as his own opinion.
Max O'Rell's Her Royal Highness Woman (1901) addresses it to men choosing a wife. In Chapter IV of the Project Gutenberg text he says he "heard the other day" the advice and is repeating it. In John Fox Jr.'s novel The Little Shepherd of Kingdom Come (1903), Miss Jennie teases Chad about a match she sees coming: "the French say" a woman should be half a man's age plus seven, which would make the woman in that match "only a few years too young" (page 222). That is no evidence of a French origin, though it shows a woman below the rule already being called too young.
None of this is new. Wikipedia's article on age disparity in sexual relationships already cites Locker's page and notes that early sources framed the rule as an ideal. A newspaper archive we tried was closed to us, so older printings may exist.
The median bride's age eclipses Locker's ideal by about seven years
SingStat's median-age table, M830169, starts in 1961, when the median groom was 27.3 and the median bride 22.8. Locker's ideal for that groom is 20.65 by our arithmetic, already about two years below the real median. In 2025 the median groom was 32.0, giving an ideal of 23.0, and the median bride was 30.4, 7.4 years above it. These are medians of all grooms and all brides, never of the same couples, but because the rule is a straight line the comparison holds.
Couples a constant gap apart come into conjunction with the band at a fixed age
A gap between two birth dates never changes, yet a couple's position against the band does. For an exact gap of g years, they reach the edge when the older partner is 2g + 14 and the younger g + 14: at 30 and 22 for a gap of eight years, at 34 and 24 for ten. The band widens faster than the gap, so once inside, a couple stays.
Month ends shift the phase of a gap counted in months and days
An exact gap in months and days needs a month-end rule. As revised on 8 November 2026, our age difference calculator counts whole months forward from the older date and stops on the last day of a shorter month. So 28, 29, 30 and 31 January to 1 March 2023 all read "1 month, 1 day", over 32, 31, 30 and 29 total days, because each count lands on 28 February. A 29 February birth date reaches its anniversary on 28 February in common years: 29 February 2000 to 28 February 2001 is exactly 1 year, where a 1 March rule gives 11 months and 30 days.
Before that revision the calculator borrowed days from the month before the younger date, so 31 January to 1 March 2023 read "1 month, -2 days". Over every pair with the older date in 1990 and the younger in 1990 to 1999, 192 of 1,266,550 came out negative. It also rounded weeks, so a 4-day gap showed as 1 week; the revision gives whole weeks and leftover days. The dates carry no time of day, so time zones cannot move a result.
Five-year age groups leave a wide penumbra at the band's edges
In 2025, 3,127 couples (12.7%) sat in cells where some age combinations pass and some fail, and the table cannot show which. Ages are whole years at marriage. Since 1995 SingStat has published brides aged 60 and over only against grooms under 40, so we put the rest of that group into one cell of 148 couples; splitting it by the published totals for grooms lifts the lower bound to 86.6%.
Contains information from Marriages By Age Group Of Grooms And Brides And Type Of Marriage, Annual (SingStat table M830164) accessed on 8 November 2026 from the Singapore Department of Statistics which is made available under the terms of the Singapore Open Data Licence version 1.0.
Contains information from Median Age Of Grooms And Brides By Marriage Order And Type Of Marriage, Annual (SingStat table M830169) accessed on 8 November 2026 from the Singapore Department of Statistics which is made available under the terms of the Singapore Open Data Licence version 1.0.
View author profile → · Editorial policy
