What do we know about the demand for private renting?
Regular readers will know that I have previously argued that we know surprisingly little about the UK rental market, despite the magnitude of growth in the sector in recent years. Evidence is hard to come by because of the disparate diffuse nature of the sector, which is really a series of well-defined demand segments. But this does not stop many people writing about the sector as if it was homogenous, simple to understand, monitor, evidence and, consequently, intervene in to moderate excesses.
Unfortunately, this over-generalisation has extended to the former Chancellor of the Exchequer who introduced tax changes that will reduce returns and increase borrowing and transactions costs (but without any proper analysis of the impact of these changes). But the sector is not simply an investor question – it is also about who lives in the sector and their aspirations and experiences; and it is all the more important because at the margin it is the sector where the action is because of credit constraints on entering owner occupation and, for others, because of the insufficient level of affordable housing supply of social housing.
It is against this backdrop that a couple of weeks ago the Bank of England’s blog site Bank Underground issued an interesting paper on the drivers of the demand for private renting by Mariana Gimpeliewicz and Tom Stratton. In what follows I will summarise their paper and its findings and then reflect a little on what they say.
Noting the observed growth in the share of total dwellings privately rented since 2002 and the growth since 2008 of buy to let mortgages (BTL) rising at more than 6% per annum (a subset of all private renting), the authors argued that it made sense to try to understand the growth in PRS demand, what its main components are and then divide the analysis between pre-crisis and post-crisis and also make some projections about future demand levels in the sector. The authors note that BTL loans explained about 70% of PRS expansion before the crisis but that tighter underwriting may explain why BTL only supported about a third of the growth after the crisis (most of which came from ‘cash purchasers’).
The main drivers of demand for the PRS are (1) demographic sub-groups i.e. students, in-migrants and younger people, who all have a higher propensity to rent privately; (2) where the provision of social housing falls, households will be funneled to the rental market; and, (3) potential first time buyers who are credit-constrained or face affordability shortcomings will be obliged to either rent or live with parents. In this study, the lack of affordability data means the focus instead under (3) is on credit conditions.
The authors found that pre-crisis, net inward migration was the largest contributor of PRS demand ie about a quarter of all sector growth (2002-07). Additional students helped explain another 12.5% of growth, with slightly less growth due to the reduction in social housing available. Their analysis explained about half of the growth in the sector n this period. Between 2008-13 (the credit crunch period), migration remained an important factor, with a declining role for social housing and student growth. Much more important was tighter credit conditions for first time buyers ie the withdrawal of high LTV loans – they found that this accounted for two thirds of the growth in the private rented sector in this period. Overall, their drivers account for 82% of the growth in the sector and all of the BTL segment’s growth. The authors recognise that affordability may be driving the unexplained growth – in other words that by driving up house prices new BTL supply may be making owning less affordable and hence more are directed towards the rental market (but this is supposition).
Looking ahead, the authors expect demand to grow to 2019 but much less slowly than in recent years. This amounts to as much as a million additional properties in the PRS in this period. The key to how this turns out will turn on the extent to which frustrated home-owners actually manage to purchase their first home. Buy to let growth is forecast to slow.
This paper is on the one hand an interesting thought experiment that allows us to see how the different segments are growing at varying rates and have done in the recent past. It shows also that buy to let is far from all of the story. One just has to look around our University cities, for instance, to see the growth in bespoke purpose-built student accommodation. The split between pre and post crisis is also intrinsically sensible and a useful exercise.
However, despite the value of the disaggregation of the sector, the use of different data and its creative manipulation, it is worth noting that there is not much underlying economics in the construction of these different numbers. So, relative rents and housing costs and income play no direct role in coming up with these numbers. Instead, a series of generally reasonable assumptions, extrapolations and extraneous propensities are adopted. This is reasonable in the absence of the economic data of the right kind but we should be cautious.
We need more work more on these sub-sectors – what are the economic drivers in different segments; how is individual tenure choice governed by things like relative prices, credit availability, income, down-payment constraints and household variables? These models may well fluctuate over the credit and wider business cycle but do they also vary across space in different regional labour markets? What about the industrial economics or theory of the firm of different kinds of landlords – what are their motivations and how responsive are decision makers to changes in financial and economic variables (this latter question is vital to understanding the impact of policy interventions?