The missing number 13: the house number new streets in England and Wales leave out

How often do new streets skip number 13?

When a street is built, every new home is sold at least once, and HM Land Registry records the sale with the address. So on a new street the sales record is close to a full list of the door numbers.

Take the 45,244 streets in England and Wales where numbers 11 and 15 were both first sold as new homes since 1995. On 67% of them, no house numbered 13 has been sold at all, new or later. That is the number this page means by a street with no 13; streets too short to reach 15 are not counted.

Any number can be absent for ordinary reasons: a home kept by a housing association, a sale not recorded, a plot built later. The same test for numbers 9, 11, 15, 17, 19 and 21 measures that background: they are missing on 2.4% of streets. Thirteen is missing 28 times as often. Allowing for the background, roughly two in three new streets leave it out on purpose.

A second register agrees. Every new home since 2008 needs an energy certificate, sold or not, and the government's certificate search lists them by postcode. On 25 of these streets built since 2010, picked at random, it lists no number 13 at all; on six streets where a 13 was sold, the same search found every one.

What takes its place? Almost always nothing: the numbers simply jump. Of the 30,162 new streets with no 13, 204 have an 11A and 184 a 12A instead.

Has it always been like this?

It was already the norm when the record begins: 73% of the streets first sold in 1995–1999 have no 13. It has been falling since the mid-2000s, to 56% for streets first sold in 2019–2023. Still more than half.

The last two years are shaded: their streets are still being sold, so some numbers, 13 included, may yet appear.

Which councils let builders leave out 13?

The decision belongs to the council: outside London under the Towns Improvement Clauses Act 1847 and the Public Health Act 1925, in London under the London Building Acts. Some councils rule out skipping 13 in writing. Hastings: “All numbers, including the number 13, must be used in the proper sequence.” Others leave it out by default; for new homes in an existing street, Adur and Worthing say: “All numbers (except the number 13, as this is excluded) must be used and allocated accordingly in sequential order.”

Before looking at any council, the expected result was written down: councils that forbid skipping should look like the background rate (under 25%), councils that exclude 13 should be high (over 70%).

no 13has a 13

Each dot is a new-build street where 11 and 15 were sold new, placed at its commonest postcode: no 13 sold (bright) or a 13 sold (grey).

The prediction mostly failed. South Derbyshire (99%), Welwyn Hatfield (10%) and Neath Port Talbot (6%) behaved as written down; Worthing (62%) fell short of the 70% expected, and Adur had too few streets to count. Four councils whose policies forbid gaps skip 13 most of the time anyway: Warwick 96%, Hastings 83%, Leeds 78% and Oxford 68%. Leeds's policy bans gaps without naming 13. Timing explains some of it. Hastings adopted its ban in July 2013: 29 of its 30 new streets first sold before 2014 lack a 13; all five since have one. Leeds fell from 98% (streets first sold up to 2008) to 31% (since 2016). Warwick barely moved (100% to 89%).

Councils differ enormously. Of the 284 councils with at least 30 new streets in the record, 89 left out 13 on nine in ten or more of them, 39 on one in ten or fewer, and the other 156 somewhere between.

Councils that changed their minds

Comparing streets first sold up to 2008 with streets first sold since 2016, the share of new streets without a 13 changed by more than 50 percentage points in 48 of the 204 councils with at least 20 new streets in each period. 45 now leave out 13 far less often; three began leaving it out. That is much of why the national share has been falling.

Does a number 13 sell for less?

Every Halloween and every Friday 13th, the same story goes round: number 13s sell for thousands less. In October 2024 an estate agent put the gap at £36,000, comparing the average number-13 sale in January–August 2024 with the average of all sales.

Run that method on the same months of the same Land Registry file and number-13 sales averaged £333,297, against £353,768 for everything: 5.8% less. By medians, which the release's table names, the gap is 1.8%. Neither reproduces the 12.2% published, whose exact selection was not given. The bigger problem is what 'everything' contains. 14% of sales have no plain door number: mostly flats in named blocks, but also houses with a name instead of a number, cottages and farmhouses among them, which are 3% of all sales and average £656,000. Against homes with a plain door number, number 13 is −2.3%. And by the same arithmetic numbers 11, 12 and 15 'sell for less' too (11 −4.5%, 12 −6.8%, 15 −6.5%).

From the whole country to the house next door

The fair comparison is with the houses either side, on the same street, of the same type, sold in the same year. Over 24,595 streets, number 13 sold for +0.1% compared with numbers 11 and 15. A discount larger than 0.16% (about £450 on a typical number-13 house) lies outside the 95% interval. Every other odd number tested the same way, from 5 to 41, is also within a fraction of a percent of its neighbours.

Nor do owners leave number 13 sooner. On streets where 11, 13 and 15 have all been sold, each number 13 changed hands 2.04 times on average since 1995; its neighbours, 2.05.

Do people avoid moving on Friday the 13th?

Yes, and here the superstition shows up in the record. Of 54 Fridays the 13th since 1995 (Good Friday 2001 left out, as are all bank holidays), 53 had fewer completed sales than the typical Friday of the same month.

Sales pile up towards the end of each month, so the fair test is the Fridays either side in the calendar. A Friday on the 12th runs at 95% of the month's typical Friday, a Friday on the 14th at 102%. A Friday the 13th runs at 66%: about a third fewer. The next quietest date is the 2nd, at 82%, just after New Year and month-end.

So the superstition moves the moving date. It does not move the price.