The OBR and Welfare and Housing Benefit Spending Trends
by Ken Gibb
The intrinsic difficulties facing agencies tasked to produce reliable forecasts for future years spend and caseload across different policies, are well known and understood. In the present debate over how future cuts in welfare spending will be achieved and the all too frequent implementation gap between planned savings and actual outturn spend, a degree of skepticism about deliverability is appropriate (before we get to questions of desirability or otherwise).
While scanning through the Budget’s supporting documents, I happened upon the OBR’s Welfare Trends Report, published in the Autumn of 2014. Of course, this turns out to be part of the immediate political narrative about welfare spending cuts following from the Chancellor’s statement, but I am actually interested here more in what it says about underlying factors and expected trends in welfare spending and their ‘forecast-ability’. In particular I want to look at the figures for Housing Benefit. I may be rather late to reading this report but I thought it was still worth exploring a little.
OBR argues that past data on a small number of drivers explain long-term trends in welfare spending (and implicitly will continue to do so in the future). The main candidates are: demographic trends, labour market trends, inflation & earnings growth, and, housing market trends. We can also point to successive incremental and then non-marginal policy changes especially after 2010. How policy change impacts on spend and caseload is of course far from straightforward, particularly in terms of assumptions made about second round behavioural changes conditional on benefit policy changes e.g. on labour supply or landlord behaviour. We return to this below.
Future welfare spending trends are forecast of the period 2013-14 to 2018-19. OBR expects welfare spending to rise in cash terms by 12.5%. However, this is less than the forecast growth in the economy. Thus, as a percentage of GDP, welfare spend will fall from 12.8% to 11.6%.
The forecasts are of course subject to risk. The chief risks are seen to be, first, the uncertainty associated with the future of Housing Benefit, which may vary as a result of how the economy performs (jobs, wages and rents are the key variables). Second, policy change associated with disability and incapacity benefit is also viewed as an important source of risk to the forecast. Third, and believed to be less significant than the first two, the introduction of universal credit is also identified as a risk to the accuracy of future welfare spending forecasts.
Chapter 9 is the Housing Benefit (HB) chapter. Spending on HB in 2013-14 was £23.9 billion (more than £20 billion of which counted within the benefit cap). The OBR forecast that HB will rise in cash terms to £27.4 billion by 2018-19 but that because of forecast larger economic growth this will constitute a falling share of GDP – from 1.5% to 1.3% between 2013-14 and 2018-19.
There are some striking facts in this chapter. The remarkable rise, for instance, in those in employment on HB. This was 0.4 million in 2008 but had risen to 1.1 million in the middle of 2014 (i.e. 3.5% of all employed adults). Second, the important drivers of trend growth in HB has been real rent increases over time, the shifting caseload by tenure into the private rented sector (with its average higher rents) and also the impact of being the recipient of other benefits associated with HB. In this case the forecasts also assume an impact from the cumulative effects of welfare reform and that includes the retention of the bedroom tax.
OBR disaggregates the source of the increase in HB spend up to 2018-19 and much of it is the result of what they call HB only cases – almost entirely, renters in employment. They also point to the importance of the expected growth in Incapacity Benefit caseload linked to higher HB average awards. The growth is offset by expected falls in JSA through anticipated falls in unemployment, and also as a result of a decline to come in the pensioner HB caseload as a result of shifts in their treatment via benefits like pensioner credit.
However, OBR recognises that they have systematically underestimated the growth in HB spending in the (recent) past, and, while they think they have corrected for this, we should be cautious. It turns out that much of the risk or uncertainty in their forecasts will depend on the future pattern of tenure change and in particular how much home ownership grows relative to renting.
The home ownership tenure change outcome is construed as the net effect of different policy impacts on home ownership rates: between help to buy schemes like the new ISA package versus the restrictive impact of the Mortgage Market Review condition. I think there are several other tenure choice variables and market processes, not least likely rising future mortgage rates, that will shape this home ownership figure and perhaps, in the end, housing policy is actually less important to that change than the commentariat believe but that is another story. OBR also recognise that the future path of rents and the way the labour market evolves (in terms of wages and employment) will also matter. It would be helpful and very interesting to see how the model builders at OBR conceive of and operationalise the underlying housing and mortgage market within the wider models they use to construct these key parameters for welfare spend forecasts.