Concealed households — what’s the real story?
Liam Halligan is always a man you’d want to have on your side in a debate. Unfortunately on housing he disagrees with me. But his rebuttal sidesteps the central case I’ve been making: that benign day-to-day housing costs and 25 years’ growth in the surplus stock of housing tell us that a lack of houses is not the reason why house prices have become unaffordable. He focuses instead on what increasingly seems to be the only piece of circumstantial evidence that housing shortage proponents have left to point to: the evidence of increasing numbers of ‘hidden’ households.
Also referred to as ‘concealed’ or ‘multi-family’ households, these are households containing more than one family unit. While there are many possible reasons why two family units might choose to share a house, one of those is that housing costs could be forcing them to share when they would otherwise live separately.
In practice it’s hard to see how housing supply can be the explanation when average housing costs are falling relative to average earnings almost everywhere, and stable in London.* What’s more, distributional changes like the erosion of social housing, weak wage growth for younger people, and more recently the benefits squeeze, seem like more probable causes of such a trend, and those things have a far more potent effect on affordability than changes in supply ever could. But nevertheless, it’s evidence that is at least consistent with a shortage hypothesis and is worth exploring.
Why is the number of concealed households rising?
First the numbers. Liam claims that concealed households have risen by 50% in the decade to 2016, reaching 2.5 million. James Gleeson offers another set of LFS-based figures here, showing the number rose by just under 50%, from 1.65 million in 1996 to 2.45 million in 2016 — similar growth and level to the numbers Liam cites, but over 20 years rather than 10. In neither case is it entirely clear how the figures have been derived — Liam cites an unpublished MHCLG memo — but what is clear is that the number of concealed households has indeed been on the rise.
What’s driving this? If housing costs are a problem we’d expect to see a general increase in the rate of concealed households across all families. So we need to get underneath the headline figures to understand who makes up this population of hidden households.
I’ve used the Labour Force Survey to recreate a measure of concealed households (see below for the variables used).** It shows that the proportion of family units that are ‘concealed’ increased from 7.2% of all families in 1997 to 8.6% by 2017. For comparison with the above figures the absolute numbers work out as an increase from 1.86 million to 2.57 million by 2017, so not far off James Gleeson’s estimates. It’s worth noting that, despite my referring to them as ‘families’, it appears that the majority of these are unrelated adults sharing a house.
The UK population has undergone changed in the past 20 years owing to higher levels of immigration, particularly in London. And from the LFS it’s apparent that migrant families are about twice as likely to share a household than UK-born ones — perhaps because many are single adults who come to work and save.
So if we split concealed households into those headed by someone born overseas and those with a UK-born head, it reveals that there has actually been no change in the UK-born proportion of families that are ‘concealed’. That number starts at 6.0% in 1997 and ends at 5.8% in 2017 — a higher absolute number due only to the growth in population. In other words a family headed by a UK-born person seems no more likely to be sharing a house with another family than it was 20 years ago.
Meanwhile all the growth has been driven by concealed migrant households — here again the majority appear to be unrelated adults sharing a house — who used to account for 1.2% of all families in 1997 but by last year accounted for 2.9%. That’s an increase from 310,000 families in 1997 to about 850,000 by 2017.
So it seems that the concealed households argument for a shortage of houses — always somewhat tenuous because of its many possible causes — is actually an artefact of recent immigration. There is no evidence, on this simple analysis at least, to suggest that anything has changed for the UK-born population. We don’t know why these families share households and there are no doubt many different reasons. Nor do higher rates of sharing among migrants mean that everything is fine. But what this compositional explanation does imply is that housing supply isn’t the cause of the growth in concealed households.
— — — — — -
*I use a simple definition based on the LFS variables for the number of family units within a household (variable FAMUNIT), identifying concealed households among those households containing more than one family unit. I then count the number of ‘heads of family unit’ (variable RELHFU) who live in the second or subsequent family unit within the house.
**Incidentally, Liam also argues that the housing shortage is based entirely on national numbers. Here is my analysis of the regional numbers on rent and stock, showing no evidence of a deterioration in the availability of places to live over the past 20 years, even in London.