Blog

Three countries, three public registers, three different things recorded when a home changes hands. What each one carries, what it leaves out, and what that means for the figures.

Price per square foot, HDB versus private, remaining lease and the mix adjustment: the four things that make Singapore's figures legible.

A register that publishes the price of every sale and the size of none of them. What that separation costs, why it forces a second dataset into the calculation, and how it compares with the two registers that do not need one.

The most complete open property register in Europe, published by the tax administration, with a legal condition attached that shapes what anyone may build on it. What it carries, what it leaves out, and the three départements it cannot cover.

Five specific obstacles between three honest national datasets and one honest cross-country number, and the account of which of them a third market removed.

Every price per square metre divides by a floor area, and no two countries measure one the same way. What is counted, what is excluded, and why a rate cannot cross a border unchanged.

Three kinds of number get called “house prices” and they answer different questions. Which one moves when a mansion sells, which one survives a change in what is selling, and which one cannot tell you what anything costs.

The most common way a property headline misleads, and it requires nobody to lie. When what is selling changes, an average moves without any home changing price. What mix adjustment does about it, and why one market here refuses to attempt it.

Measure price change using only homes that sold twice, and the property itself becomes the control. What that fixes, what it costs in coverage, and the one bias every repeat-sales index in the world carries.

Every published local price rests on a sample, and below some size it stops describing a market and starts describing a handful of houses. Where three national registers put that line, and why the answer is partly a legal question in one of them.

Property statistics are published in the money of the day, so a decade of flat prices reads as flat when it was a substantial fall. What the correction is, why publishers skip it, and where it changes the answer most.

Every property statistic is one of three things, and most of them do not say which. What separates the number a seller hoped for, the number a model produced, and the number that was actually paid.

Every register of achieved prices is behind, and the amount it is behind by is a property of the country’s conveyancing system rather than of the publisher. What the lag is made of, and how to read a figure that has not finished arriving.

A directory of national registers that publish what homes actually sold for, with what each one carries, what it costs, and the condition attached to reusing it. Short, and only registers we have read.

Three open licences, three different answers about what a reuser may do. The permissive base terms, the sector conditions layered on top, and the one condition that constrains search engines rather than people.