Don’t panic!
What the ONS’s new population projections mean for household formation and housing need
Earlier this week the ONS published updated population projections. These led to some hyperventilating about housing need, with associated shocking numbers, and in some quarters tough talk about immigration. So what really are the implications for household formation in England out to 2036 and how worried should we be?
The short version is that the new population outlook implies a significant increase in the pace of household formation in England compared to the ONS’s most recent official household projection. But it’s not something to panic about, not least because the rate of household growth in recent decades has consistently lagged behind housing supply — not been constrained by it.
First the headline population numbers. The increase compared to last year’s ‘2020-based’ population projection is big, with the population of England anticipated to hit 62.5 million by 2036 compared to last year’s expectation of 59.3 million by that year.
Most of that increase is driven by a jump in the rate of net migration, roughly doubling to 4.0 million between 2023 and 2036. Together with an increase of around 1 million in the pre-existing population, driven primarily by an upward revision to the birth rate, the result is an increase of 3.2 million people in 2036.
What does that mean for household formation? The ONS publishes household projections based on a combination of its population forecast and estimates of household size stratified by age. The household count is simply calculated as the population divided by the average household size (for more on past methodological shenanigans on this front see this). *
Applying the new population projection to the ONS’s existing outlook for average household size gives an implied household formation rate for England of 238,000 per year between 2021 and 2036 (the blue line). That’s an increase of 53% on the official projection published in 2020, which anticipated 156,000 new households each year (the green line). In practice, given that migrants tend to live in larger households and the domestic population increase is due to a higher birth rate, a more sophisticated projection — when ONS produces one next year — will probably come in somewhat below 238,000 because the average household size will be higher.
If the new population numbers turn out to be accurate they imply we need a more houses over the next decade than past iterations implied. How concerned should we be? It’s important to put the implied household numbers into historical context. 238,000 households per year (green diamond on the chart) would be a rate of increase far higher than we’ve been used to over the past generation (147,000 per year — the red line). And while net additions to the housing stock have averaged 235,000 per year over the past 5 years, those are high rates of building compared to recent decades (blue line).
On the other hand, there’s definitely no need to panic. Since 1996 we’ve added 4.7 million houses (25.2 million in 2022) and seen 3.8 million more households (23.6 million in 2022) — a steady improvement in housing availability over a generation, if not the revolution some advocate. Assertions that there’s a ‘decades-long backlog’ of building that this ignores are hard to square with the historical data. There is capacity to manage short-term volatility in the housing stock relative to the population without it materially affecting affordability.
The other reason not to panic is that population projections are notoriously volatile. It could easily be the case that they are revised down again before long, especially if policy acts to reduce gross migration inflows. As Yogi Berra would surely have said: it’s difficult to make population predictions, especially about the future.
All in all, sustaining current rates of housebuilding across the economic cycle would be sufficient on these projections to hold affordability steady, all else equal. But what will matter far more for the affordability of housing over the coming decade is what happens to interest rates, the generosity of housing benefit and the supply of social housing. As we’ve seen over two generations, these policies, far more than market housing supply, are the critical determinants of housing affordability.
— — — — — — — — — — — — — — — — — — — —
*There are obviously lots of legitimate debates to be had about what factors — such as housing costs, changing social norms or parents with spare bedrooms and more liberal attitudes — have influenced past changes in household formation. I won’t get into those here as the purpose is just to draw out the implications of the new population forecast for the ONS’s household projection.
