House prices and how to measure them

Every time house prices hit the news, someone at the weekend BBQ asks the same question: why are all the numbers different? One headline says the median is up. Another says values are flat. A third says prices are still well below peak. So, who’s right?

The short answer is that there is no single stock answer for house prices because they are measuring different things. Some look only at the homes that sold. Others try to estimate what the wider housing stock is worth. Some are snapshots, while others are built to track change over time. If you do not keep that distinction in mind, it is easy to compare apples, oranges, and the occasional kiwifruit.

A simple way to think about this is to change the unit of measurement altogether. In 1966, it would have taken about 3,100 sheep to buy a typical house. Today, it would be about 2,900. On that measure, houses have become slightly cheaper. So, should we all start tracking house prices in sheep because it sounds like house prices are falling? Clearly not. But it is a useful reminder that the way we measure something shapes the story ewe tell about it.

That’s the heart of the issue. Before we argue about whether house prices are up or down, we need to ask what exactly we are measuring, and why. The right measure depends on the question. For example, if you want a quick snapshot of what homes sold for, medians and averages can be useful. But there is a catch.

Average and median sale prices are sensitive to the mix of homes that happened to sell in that month or quarter. If more expensive homes sell than usual, the average can jump even if the underlying market has barely moved. If cheaper homes dominate sales, the average can fall even if like for like values are broadly steady. Medians are less affected, but they aren’t immune. That is why two months with very similar market conditions can produce quite different results. It does not mean one number is wrong. It means each measure is answering a different question.

It gets even trickier when media releases talk about changes in “average asking prices”. Those aren’t realised sale prices. They reflect what vendors hope to get, and they are also shaped by the mix of properties being advertised. With a high number of listings on the market, you could cheekily call them “unsold prices”.

But if your question is how much house prices have risen or fallen over time, an index is usually a better tool. A good house price index aims to track what happened to prices without being affected by changes in quality and types of houses selling.

That does not mean indexes are simple. They use additional property information to control for the changing mix of homes being sold, which is why they are often better for tracking underlying movement. But I’ll leave the detailed methodology for another time.

Even then, indexes come with complications. Different organisations use different data sources and methods. Some use settled sales from councils, while others use data from real estate agents when contracts go unconditional. They can also report changes over different timeframes, such as monthly, quarterly, or annually.

There is another complexity too. Some indexes measure changes in sale prices, while others measure changes in the value of the whole housing stock — not just what sold. That may sound like a technical distinction, but again it does matter. Whilst an index is not perfect, it is better suited to tracking change over time in a messy real-world market.

That is not to say medians and averages are useless. Far from it. If you want a simple sense of the dollar amounts homes are selling for right now, they are still very useful. Journalists like them. Buyers and sellers relate to them. They are concrete in a way index numbers often are not. That’s why HUD has just released its new Property and Sales Statistics dashboard with comprehensive property sales data.

But if the goal is analysis rather than headlines, it’s better to use an index. When people are making big calls about momentum, inflection points, affordability pressures, or policy impacts, a median can be noisy and an average noisier still. A well-built index gives a cleaner read on the underlying trend.

So next time you see someone confidently declaring that house prices are up, down, flat, booming, collapsing, or somehow all four at once, the first question to ask is not “who is right?” It is “what exactly are they measuring?”

Because if you choose the wrong measure, you might be left looking sheepish.