A ballooning welfare bill? It’s not that simple

By Declan Gaffney

The chancellor and the chief secretary to the treasury have both adopted a simplistic narrative of out of control welfare spending in framing the June budget decisions and the coming spending review. George Osborne said in his budget speech:

“Total welfare spending has increased from £132 billion ten years ago to £192 billion today. That represents a real terms increase of a staggering 45 per cent. It’s one reason why there is no money left. It has also left an increasing number of our fellow citizens trapped on out-of-work benefits for the whole of their lives.”

Danny Alexander took up the message in an Observer article:

“The welfare bill has ballooned by £60bn in 10 years. Failing to tackle this would make it impossible to deal with our debts without cutting essential services.”

That £60 bn figure has received surprisingly little attention from commentators. What was this money used for? The chancellor’s statement referred only to long term unemployment and economic inactivity. But £60bn is a very big figure in terms of UK government expenditure- in fact it’s bigger than the total DWP spend on the working age client group, so the growth can’t be solely due to this factor.

One can see why some may want to believe that growth in welfare expenditure is accounted for by people being ‘trapped on out-of-work benefits’ under the previous government, as that would help in presenting cuts to the welfare bill in terms of what sound like socially desirable objectives rather than as pure pain. But the main programmes contributing to growth in welfare spending over the last ten years were payments to the retired (pensions, pension credit and disability benefits), to families with children (child tax credit) and to low-income workers (working tax credit). Claims that expenditure was ‘out of control’ before the recession are exaggerated at best, as the main drivers of the increase over the longer term have been demographic change (an ageing population) and conscious policy choices (reducing child and pensioner poverty).

The contribution of working age social security benefits to overall expenditure growth is largely concentrated in the last two years- reflecting the impact of recession rather than long term structural factors. There was growth in some out-of-work benefits expenditure before the recession, but there were also falls in others, notably in incapacity benefit, so net growth from 2000/01 was less than £3bn up to the crisis. The chart below gives a broad picture of how welfare expenditure has developed since 2000/01. The figures won’t match up exactly to those cited by the government, as there is no published time series which gives detailed and consistent figures for both types of expenditure back to 2000. Details on the sources are at the foot of this article. The green column shows change up to 2008/09, the red from 2008/9 to the present:

Between 2000/1 and 2008/9, social security expenditure on those over working age increased by £20bn, while expenditure on those of working age increased by just under £3bn. The great bulk of the increase in social security over this period was thus accounted for by pensions and benefits paid to those above working age. The working age total reflects increases in some programmes (housing benefit and DLA) and falls in others, including incapacity benefit, which began to fall long before the introduction of the employment and support allowance in late 2008. Social security expenditure on children fell but this needs to be seen in the light of the £14bn increase in tax credit expenditure over the same period, which saw the introduction of child tax credit.

DWP’s forecast for 2010/11 gives an indication of the impact of the recession on expenditure: growth in working age benefits since 2008/9 is more than twice the growth in the previous eight years: £6bn compared to £3bn. The main drivers are rises in housing benefit and JSA payments and, ironically, the new employment and support allowance (partly offset by continuing falls in incapacity benefit). Spending on the retired has also grown significantly- indeed, on these figures it increased more in the last two years than working age benefits increased over the previous eight years. And tax credit expenditure has increased by almost as much, reflecting both policy decisions and the effects of recession on employment and hours worked.

So where did the £60bn referred to by the Chancellor and Chief Secretary go? It isn’t possible to give a definitive answer to that question from these figures, but we can put some orders of magnitude on the shares going to different groups based on the best information available at the moment. Social security benefits to those of working age before 2008/9 contributed only 6% of the growth on these figures, which should dispose of any claims that structural problems in the social security system or the labour market are the major driver of the growth in welfare expenditure over the last ten years (needless to say, this does not mean that there are no such structural problems, just that they don’t explain why expenditure has grown). Growth in pensions prior to 2008/9 accounts for 40%, and tax credits for 27% (offset by reductions in DWP childrens’ benefits of -3%).

On a more tentative basis, it looks as if increases in working age benefits over the entire period since 2000/01 contributed about 18% of the expenditure growth that we can account for, but as we have seen this is heavily concentrated in the last two years. Over the entire period 2000-2011, pensions and benefits to the retired account for about half of the total increase, and tax credits about a third taking account of falls in DWP children’s benefits. There will be other ways of doing the calculation and different data sources, but I doubt that other approaches would yield substantially different broad results.

Of course expenditure would have risen less if working age social security had fallen prior to 2008/9, and tax credit expenditure would also be lower if there were fewer out-of-work recipient families. But those who want to argue this way need to specify their counterfactual: exactly where should expenditure have fallen, and what policies would have brought this about? Unfortunately, it looks more likely that distorted accounts of trends in benefit expenditure will be used to make the case for cuts in the absence of coherent policy rationales, as with the budget announcements of arbitrary reductions to disability benefits and the savage proposal to cut housing benefit entitlement for those who have been out of work for more than a year.

Data sources

Social security spending is from DWP’s benefit expenditure tables, last updated in March, and in particular their client group analysis, which divides expenditure between children, working age adults and adults over working age (Table 5). Tax credit figures back to 2004/5 come from July’s Public Expenditure Statistical Analyses: figures for earlier years are from earlier editions of PESA, in flagrant disregard of PESA’s strictures on combining data from different publications. Child benefit, taken from HMRC’s time series from 2003 to 2008/9, is allocated to the DWP ‘children’ category for simplicity throughout, even though DWP no longer administers this benefit. For years after 2008/9 that year’s real terms expenditure on child benefit is rolled forward as HMRC don’t seem to have published estimates for those years. All figures are in 2008/9 prices, calculated using the current GDP deflator.

Combining data from different sources into a single times series is hazardous because of differences in reporting conventions, and double counting of some minor items in DWP and PESA sources can’t completely be ruled out. Total spend is calculated from the individual series, and won’t correspond to PESA figures. For the period up to 2008/9 the individual data series are based on outturn expenditure reported by government. For the period 2008/9 to 2010/11 figures are provisional: the only detailed data on social security expenditure is from a forecast which precedes the budget. So the figures for change up to 2010/11 need to be heavily caveated: they are shown only to illustrate our best understanding of the broad trend: in this chart, red is for danger. At the same time the risks shouldn’t be exaggerated- these figures weren’t plucked out of the air.