Why doesn't the share of income spent on housing rent decline as we all get richer?

A puzzle that sits at the heart of our housing conversation

According to household surveys, an astonishing 20.3% of all household spending in the 1940s was on clothing and footwear.

Today, clothing and footwear spending is just 3.3% of total household spending, a sixth of the share it once was.

But what about housing?

These same spending surveys, which are used to estimate the share of spending on different items to create the consumer price index, show that from 1948 to 2025, the housing rent share of spending has been roughly flat for renter households, being 6.1% in 1948 and 6.6% in 2025.

Why?

Why are people content to spend the same share of their income on housing as they did eight decades ago, but a sixth as much on clothing?

A popular view is that the share of income spent on housing should fall as our collective incomes rise. A flat rent-to-income share is therefore thought to indicate a problem with the housing market.

But is there any economic logic to this claim?

These are the questions I dig into in this deep dive article. I make four key points.

1. A wide range of data shows that the average rent-to-income ratio is flat over long time periods and many locations, but intriguing patterns sit behind these averages.

2. There are economic forces that sustain this flat rent-to-income share, which I call the rental price equilibrium. These forces are composed of what economists call substitution and income effects that result in housing rents tracking the income of renter households (or, in the jargon, people have homothetic preferences for housing). This effect works in both directions, so if we collectively get poorer, we will still spend the same share of our lower incomes on housing rents.

3. A better way to unpack what is happening is to think of housing as a category of spending rather than an item, and within that category, you are buying many housing-related items or attributes—size, quality, bathrooms, bedrooms, pools, etc. There is no economic reason that we should spend less income on the sum of this bundle of attributes as we collectively get richer, as some attributes are luxury items, and in particular, location is a status rank item.

4. The only way to prevent rents reaching the rental equilibrium is to prevent the trades that generate it, which could involve rent stabilisation rules or public housing, benefiting renter households on a home-by-home and dollar-for-dollar basis without affecting the broader rental equilibrium.

---

Please consider a paid subscription to access the full article and many others, including audio versions of my two books, and support Australia’s most independent one-man think-tank.

This is the final deep dive article for the year (as I will be travelling and working abroad for the next month), though a terrific podcast and video will be released in the coming weeks.

Merry Christmas to you all!

By the way, Fresh Economic Thinking is so close to 7,000 free subscribers. If you haven’t subscribed, please do. If you are one of my valued subscribers, please share widely any articles with your networks. Let’s hit the 7,000 subscribers before the year’s end.

---

The data on housing costs as a share of income

The longest comparable surveys of household spending on different items are those used to generate the weighting needed to construct the consumer price index (CPI). The CPI is a weighted average price level for all the items in the basket. To determine how much to weigh, say, clothes, in that average, we need an estimate of the proportion of spending households make on clothes relative to all other items.

The chart below shows the different weights to major categories of spending in Australia’s CPI basket since the inception of the index in 1948.

Notice a few things.

First, the share of spending on clothing and food has fallen as Australian households have become wealthier over the second half of the 20th century.

What else?