Economics

Iain Duncan Smith’s welfare revolution never happened

No radical reform, only retrenchment and continuity with the policies of the last government

April 07, 2015
What kind of record has IDS left for his successor? © Stefan Rousseau/PA Archive/Press Association Images
What kind of record has IDS left for his successor? © Stefan Rousseau/PA Archive/Press Association Images
The piece was written in response to an article by James Bartholomew on Iain Duncan Smith and welfare reform

Perhaps the kindest thing to be said about James Bartholomew’s recent article for Prospect on Iain Duncan Smith is that its partisanship is undisguised. As an account of the coalition’s record on social security, it cannot be taken seriously.

Part of the problem—there’s really no nice way of putting this—is that there are simply too many silly mistakes. We are told that “the rate of unemployment has... fallen by 813,000” since late 2011. That’s a fall in the count, not the rate. We are told that among the reasons for this fall are that more single parents and disabled people are subject to benefit conditionality: but prior to these policy changes, most of those affected didn’t appear in the unemployment figures (because they were economically inactive), so they won’t have contributed to the fall. We are told that prior to the coalition, unemployed claimants  “used to go to a Job Centre, fill in forms and be given their benefits. Now they are required to make a commitment to seek work,” as if conditionality had not been a feature of the UK system for decades (what does he think the introduction of Jobseeker’s Allowance was about?). The aim of the Benefit Cap “has been to make benefits no more than the mean income of a working household.” Wrong: it is to make benefits no more than median earnings, a very different thing.

And so on. This casual attitude to the facts does not inspire confidence, especially when the author is making very big claims indeed. “What Duncan Smith has achieved is the biggest change in welfare to have taken place since the Attlee government.” The basis for this statement, as far as I can see, is the claim that the reduction in unemployment since 2011 has no historical precedent. I’m not sure where to start here, because I’m really not sure what Bartholomew means—or rather, I think I know what he means but I find it hard to believe he can be serious. The argument is that the sort of fall in unemployment we have seen since 2011 usually only occurs during booms. As there hasn’t been a boom since 2010 (I think we can all agree on this) the fall in unemployment must reflect the independent effect of Duncan Smith’s policy changes, which thus represent an unprecedented achievement. Here’s the relevant passage:

“[T]he rate of unemployment has fallen dramatically. Since the latter part of 2011, it has fallen by 813,000. That is an astonishing drop. But most people don’t realise just how astonishing. There are only two other times in the past 60 years when unemployment has fallen at a comparable rate. One was in the late 1980s and the other in the mid-1990s. But on both those two occasions, the British economy was roaring along at well above the trend rate. In the late 1980s, the growth rate was about 5 per cent. In the mid-1990s it was about 4 per cent. But this third, dramatic drop in unemployment has not been helped by any such boom. The growth rate since late 2011 has been below trend, running at 1.6 per cent in 2011, then 0.7, 1.7 and 2.6 per cent. That is an average of only 1.65 per cent. This is unprecedented: a rapid and significant drop in unemployment without a boom.”

Let’s try to make sense of this. Chart 1 shows the International Labour Organisation unemployment rate from the early 1970s to the most recent data for the last quarter of 2014. We look at the rate—the percentage of economically active people out of work—rather than the count because the economically active population has grown since the 1970s, so comparing counts could be misleading. There are indeed three major falls in unemployment. The first marks the end of the early to mid-1980s crisis (both a cyclical and a structural crisis) and the onset of the Lawson boom; the second the recovery from the early Nineties recession and the third the long-delayed recovery of the last two years:




Chart 1



Source: ONS Labour Market Statistics database




If you remember the mid-1990s, you will probably be surprised to hear that this was a boom period, and you’d be right. Why does Bartholomew think it was? He seems to think that a boom occurs whenever the growth rate of GDP is above the long-term average. But it is also possible (and common) for growth rates to exceed the trend when economies are recovering from recessions. The difference is that booms are associated with positive output gaps, where GDP exceeds long-term potential, while during recoveries the output gap is negative—the economy is running at below full capacity, even if the growth rate is relatively high. The latter is what was happening in the mid-90s: there was no boom until the dotcom bubble later in the decade. That periods of falling unemployment should involve a negative (but reducing) output gap is hardly surprising, as both unemployment and the output gap are indicators of under-used capacity. So the idea that big falls in unemployment only occur during booms, the premise of Bartholomew’s argument, is just wrong.

What about the role of policy in reducing unemployment? Bartholomew rightly says that the coalition’s most high-profile policies—Universal Credit, the so-called Bedroom Tax and the benefit cap—will have had little or no impact: only a tiny minority are receiving Universal Credit and the unemployment effects of the other policies are negligible. He is also right to downplay any additional financial incentive effects from benefit cuts, as these are unlikely to have had more than a marginal impact given the already low value of out-of-work benefits. Moreover, as we have already seen, policy (mostly inherited from the previous government) for single parents and disabled people mainly affects people who did not previously appear in the unemployment figures. This leaves what I described in a recent article as the only major new policy fully implemented by the coalition: an unprecedented tightening of benefit conditionality and sanctions. But does this explain the fall in unemployment as Bartholomew (and Duncan Smith, it seems) claim?

The question is easily dealt with. Changes in the unemployment rate typically involve changes in flows into and out of unemployment. As Jonathan Portes has pointed out, the fall in unemployment benefit claims over recent years is overwhelmingly due to fewer people coming on to the claimant count rather than more leaving. If conditionality and sanctions were the main drivers, we would expect the opposite to be the case. The trends in flows and rates for the broader ILO definition of unemployment (the figures Bartholomew cites are based on this definition), are shown in chart 2. If there has been any effect from conditionality policy, it is small beer compared to the reduction in the inflow. (This is before even considering the fact that fewer than half of those who are unemployed on the ILO definition are in receipt of unemployment benefits.) None of this should be seen as surprising: prior to the crash, the UK already had strict conditionality in place, the value of unemployment benefits was the lowest in the OECD and unemployment spells were generally relatively brief.




Chart 2



Source: ONS Labour Market Flows (Experimental Statistics) March 2015




It is worth dwelling on the last point, as the entire thrust of Bartholomew’s argument turns on passing a return to the pre-crash situation off as a revolutionary change. If we look over time at the probability of unemployed people leaving unemployment we can see just how out of touch with reality these claims are. Chart 3 shows the “hazard rates”—gross flows into employment or economic inactivity or remaining unemployed as a percentage of those unemployed in the previous quarter. Prior to the crash, a little more than half of people unemployed in one calendar quarter were still unemployed in the next quarter: with the crash, this rose to over 60 per cent. On the most recent data, the probability is edging back to the pre-crash norm. The same is true for the probability of leaving unemployment for employment and of exiting the labour market altogether. Compared to the pre-crash world, these are not revolutionary changes: they are not even changes.




Chart 3



Source: ONS Labour Market Flows (Experimental Statistics) March 2015




So what is the coalition’s record on social security? The coalition has actually implemented very little in the way of major working age welfare reform. Its flagship policy, Universal Credit, remains becalmed with a question mark over its future whoever forms the next government. The task of reassessing Disability Living Allowance claims for entitlement to the replacement Personal Independence Payment (the only other major reform) has been deferred to the next parliament. The main achieved changes in social security policy since 2010 have involved implementing the last government’s policies on single parents and incapacity benefit, ramping up conditionality and sanctions and straightforward retrenchment measures.

There has been one clear success: implementing Labour’s plans to extend conditionality for single parents to those with younger children has undoubtedly contributed to employment rates for this group continuing their long-term upward trend (but as noted, has made little difference to overall unemployment). And there has been a major disaster: continuing with Labour’s plans to reassess incapacity benefit claimants despite warning signs that the new Work Capability Assessment was not fit for purpose led to the collapse of the system in 2014, a reversal of the previous downward trend in benefit receipt, much higher costs than planned and a loss of confidence in government’s intentions and competence in this area that will take years to repair. (It is extraordinary that an article on Duncan Smith’s record in office should fail to mention this, the biggest welfare policy mistake in decades.)

Retrenchment, I argued in the NIESR Review, was less regressive than it could have been (taking fiscal targets as given) because much of it was delivered through changes to uprating policy which spread the pain relatively widely and thinly. But it was more regressive than it needed to be because it also involved a host of smaller scale measures which delivered little in the way of savings but which had concentrated impacts on vulnerable groups. Despite these extensive cuts, total expenditure (including pensions) at the end of the parliament was scarcely different to what it would have been under Labour’s 2010 budget plans and £12bn higher than forecast in the coalition’s 2010 Autumn Statement (see Table 1, below), partly because stagnating wages increased the cost of in-work benefits, partly because coalition spending forecasts were marred by what the Office For Budget Responsibility referred to in its Economic and Fiscal Outlook of March 2015, as “a history of optimism bias relating to reforms to incapacity benefits, disability benefits and universal credit.”






Note: Council Tax Benefit expenditure was moved from DWP to DCLG from 2013-14, leading to an artefactual reduction of about £4.5bn in DWP spending. The table corrects for this by excluding actual or forecast CTB spending in all years.

Sources: Office for Budget Responsibility Economic and Fiscal Outlook various editions




Retrenchment and continuity with the policies of the previous government, not radical reform, are the major themes in the coalition’s record on welfare. Despite the rhetoric, whoever forms the next government will not inherit any sort of welfare revolution. But they will inherit a set of unresolved problems, not all of the coalition’s creation. What is to be done about Universal Credit? How can a vandalised disability benefit system be repaired when cuts in this area are already built in to the DWP expenditure baseline for the next parliament?  How can the costs of in-work housing benefit and tax credits be contained? All of this before asking where further reductions could come from. Recently leaked proposals for cuts in the next parliament give a sense of how unpalatable the choices to be faced are likely to be. In this context, assertions such as Bartholomew’s that IDS has accomplished “the greatest improvement to have been made by a British government in the last 30 years” are going to sound like a grim joke to his successor.