UK average house prices have risen by over 160% in real terms since the middle of 1996. Home ownership remains around its lowest level for a generation. Among political leaders, policymakers and commentators there is a broad consensus that these problems are largely down to one failing: decades of undersupply of housing.
But the housing shortage story is unconvincing. Official data on housing volumes and the day-to-day cost of putting a roof over one’s head suggest that supply growth has been sufficient to restrain upward pressure on house prices. However, much more powerful countervailing forces have driven them to record multiples of income. This implies that the current policy focus on boosting supply does not offer a solution to the housing crisis and a fundamental rethink is badly needed.
Housing supply has outstripped household formation for decades
It is commonly claimed that we have failed to build enough houses to meet the demand for places to live. But official data suggest this is not the case: since the 1996 nadir of house prices, the English housing stock has grown by 168,000 units per year on average, while growth in the number of households has averaged 147,000 per year. As a result, while there were 660,000 more dwellings than households in England in 1996, this surplus has since grown to over 1.1 million by 2018 (see Fig. 1). Similar trends are apparent in Scotland and Wales. This story – based on the official count of dwelling stock and the Labour Force Survey – is also visible in the English Housing Survey and corroborated by alternative measures such as the growth in total residential floor space and bedrooms per head of population.
It’s possible that the growing surplus of houses nationally could obscure tighter supply at regional level. But here too, the available data suggest that even in London and the South East, the number of houses has grown faster than the household count.
These figures are often obscured as commentators tend to compare house building numbers with projections of household formation. Both elements of this comparison have been misleading. First, ‘net additions’ to the housing stock tend to be about 30% higher than the number of new build completions each year, so reliance on the latter understates housing supply. Second, household projections have for many years significantly overestimated the rate of household formation due to questionable assumptions in the official methodology. The Office for National Statistics (ONS) has recently taken action to address these, resulting in household projections that are now much closer to past trends. Had the ONS’s new methodology applied for the past decade we would be used to hearing housing need figures of around 160,000 per year, rather than the 250,000 number that appeared in the government’s 2017 housing white paper.
The affordability of day-to-day housing costs
Despite the benign picture on housing volumes, it’s important to consider the cost of housing since that affects the rate at which people form new households. Here it is essential to distinguish between the cost of housing services and the price of housing assets. The day-to-day cost of housing services is measured by national statistical agencies around the world by the market rent on a rented house or the equivalent ‘imputed rent’ on an owner-occupied house. This cost can move independently of the purchase price of houses.
Any shortage of housing should therefore be reflected in rents and imputed rents rising faster than average household incomes. But ONS and Department for Work and Pensions (DWP) data show that rents have risen slower than median household incomes since 1996 (see Fig. 2). Consequently, affordability constraints on households have also eased, on average. Breaking the data down to the regional level shows a similar picture since 2005 (when a reliable regional rent index became available). In London, affordability has not worsened for the average household since 2005, and in all other regions rents have lagged household income growth. This is the opposite of what we would expect to see if housing supply had been inadequate and chimes with the evidence of a growing surplus housing stock.