Welfare and work: continuity and change.

Portes, Jonathan

Introduction

Shortly after taking office, the Minister for Employment, Chris Grayling, argued:

"Under Labour employment in the United Kingdom rose by nearly four million. And yet the vast majority of the new jobs created were taken up by people moving into this country from overseas. Yet through all of those years we had around five million people on out-of-work benefits in this country--and many of them could and should have been working. It was a huge waste of human talent, and a huge failing for Britain--and that failing came at a vast cost to all of us. We can't afford to make the same mistake again." (1)

What explains the persistently high numbers on out-of-work benefits? And will current policies ensure, as the Minister says, that we don't "make the same mistake again"? At the beginning of the current economic downturn, the overall UK employment rate had been above 72 per cent for a decade. Although data before the 1970s do not appear to exist on a comparable basis, this appears to be unprecedented in the history of the modern UK labour market.

At the same time, the number of people on out-of-work benefits remained at historically unprecedented levels. In 1979 there were about 750,000 people on Invalidity Benefit (IVB), about 300,000 lone parents on Income Support (IS), and 1.1 million on unemployment benefit, for a total of somewhat over 2 million--and this was of course far higher than the levels prevailing in the 1950s and 1960s. As shown in figure 2, in 1994 the total number on these benefits peaked at about 6 million, and in 2008 it was still well above 4 million. Even given the growth in the labour force over this period, this is a remarkable rise, particularly in the 'inactive' benefits which are much less closely related to the economic cycle.

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

Another way of looking at it is to observe that while overall 'inactivity' rates--people who don't have a job and aren't looking for one- have not changed that much, far fewer of those who are inactive are married women relying on their husbands' earnings, and far more are on out-of-work benefits, especially incapacity benefits.

The origins of the problem: a passive benefit system

So what happened? After three decades during which the economy, and especially the labour market, performed generally very well (Corry, Valero and Van Reenen, 2012), why were so many people still claiming benefits designed for people who were either expected to be out of the labour market for a relatively short time (the claimant count), or for a small group who were incapable of work for an extended period (the 'inactive' benefits)? And to what extent will current 'welfare reform' policies reduce these numbers?

As figure 2 shows, the expansion in the numbers on out-of-work benefits took place in the period up to about 1995, and was driven by three main factors:

* First, and most importantly, the explosion in the number of people on what is now Employment and Support Allowance (ESA- previously IVB, previously Incapacity Benefit). The initial story of this expansion is reasonably well known; the rise in these benefits began in the late-1970s and early-1980s, as deindustrialisation led to large numbers of middle-aged and older male manual workers losing their jobs. A combination of factors Jobcentres overwhelmed by the sheer volume of the unemployed, the relative and increasing generosity of Invalidity Benefit compared to unemployment benefit, political pressure to reduce the claimant count, and a prevailing view that many of these men would never get another job--meant that pushing unemployed people onto Invalidity Benefit, and, even worse and perhaps more important, leaving them there with virtually no attempt to return them to the labour market, seemed like a win for everybody. Arithmetically, the latter effect--increased durations on benefit--appears to have been even more important than increases in benefit inflow rates. (2)

* What is less well known, but is again clear from the figure, is that the real explosion in IVB took place in the late-1980s and early-1990s, much of the rise taking place well after the end of that recession. Between 1990 and 1995 claimant numbers rose by nearly a million. Moreover, in contrast to the earlier rise, the increase was less concentrated on older men in particular areas. The explanations for this increase are far less obvious--but it is clear in retrospect that allowing this to happen was one of the major social policy mistakes of the postwar era.

* The second driver was the increase in the number of lone parents on IS; this number more than tripled between 1980 and 1995. This was the result of the combination of demographic trends towards much higher levels of lone parenthood (common in many northern European countries) with a benefit system that, unlike most other such countries, essentially allowed lone parents to remain on benefit until their youngest child reached the age of sixteen. Again there was virtually no effort to support lone parents back into the labour market.

* The third driver was the operation of the basic unemployment benefit system. As previously noted, the system essentially buckled under the weight of unprecedented inflows in the late-1970s and early-1980s. The emphasis on jobsearch was significantly reduced; long-term unemployment, in particular, grew sharply. While moves in the other direction began in the late 1980s, in particular with the introduction of Restart in 1986, the impact does not appear to have been significant overall until some years later.

A further contributory factor was the increasingly dysfunctional nature of the Housing Benefit (HB) system. The combination in the 1980s and 1990s of a reduction in the social housing stock as the result of Right-to-buy, much lower levels of housebuilding in the social sector, and higher rents in the social sector, led to higher rents both in the private and social sectors. This both pushed up HB bills-and policy was explicitly designed to do this, by transferring money from housing and social rented sector subsidies to HB--and substantially worsened work incentives for those on out-of-work benefits who were also claiming HB.

Meanwhile, as noted above, overall labour market performance was (outside of recessions) excellent, demographic and labour market trends were also important. Female labour force participation, as illustrated in figure 1, rose sharply, and there were far more single-person households. The number of jobs for low-skilled men fell, as did their relative wages.

The shift to active labour market policy

So by the mid-to-late 1990s, the picture was one of a buoyant economy and labour market--but a largely passive benefit system, with an increasing proportion of people who had been on benefits for a substantial period of time. But then--influenced strongly by the work of Layard, Jackman and Nickell (1991)--policy began to move back in the other direction. This had begun with the introduction of Restart in 1986, but picked up pace in the mid-1990s. IVB was reformed and replaced by a more restrictive IB in 1995, but far more important was the introduction of Jobseekers' Allowance (JSA) to replace Unemployment Benefit in 1996. With a new government (and a Chancellor, Gordon Brown, much influenced by the LSE economists mentioned above) this approach took center stage, backed with both resource and political attention.

The economic and social philosophy underlying the new approach, generally described as "active labour market policy", was threefold:

* Unemployment was fundamentally a supply-side phenomenon; if individuals looked for work, they were likely to find it. The demand side of the labour market could safely be left to the Treasury and those running the macroeconomy;

* But this did not mean simply telling people to look for jobs and cutting off their benefits if they didn't get one, although sometimes this might be necessary. Active help and support, through assisted jobsearch, personal advisor support, and for those farther from the labour market wage subsidies or training, was essential;

* And (a somewhat later development) all this applied not just to the unemployed, but (at least to some extent) to those on inactive benefits, in particular lone parents and those on incapacity benefits.

All this was underpinned by the introduction of tax credits for low income people in work, to help "make work pay". So "active labour market policy" was from the beginning a combination of carrots and sticks. Incentives, help and support; but also activation, conditionality, backed up if necessary through the threat of benefit withdrawal. It consisted of a set of programmes that were eventually extended to all of those on out-of-work benefits. For the unemployed, there was the basic JSA regime, supplemented for young unemployed people with the New Deal for Young People and for other long-term unemployed the New Deal 25+; for lone parents, the New Deal for Lone Parents (NDLP); and for those on IB/ESA, first the New Deal for Disabled People and then Pathways to Work.

Substantial resources and a tremendous amount of policy development were invested in trying to reverse the buildup in the numbers on benefits. Did it work? As the figure shows, up to a point. There was progress, especially with lone parents, resulting from a combination of a successful active labour market programme based primarily on 'carrots' (NDLP), improved work incentives as a result of tax credits, and a labour market that was increasingly friendly to part-time work. However, it was very slow for the bulk of those on IB/ESA where policy was slower and less successful and financial incentives considerably weaker. There was also one extremely important 'negative' success: in sharp contrast to the previous two recessions, the current period of very slow growth has not led to any increase in those on inactive benefits. This reflects a clear determination by policymakers not to repeat the mistakes of the past, and should mean that the lasting social and economic damage done is not nearly as great.

This returns us to Chris Grayling's quote at the beginning of this article; how does the new government's approach differ, and will it be notably more successful?

Welfare reform under the coalition

There are three main planks to the new government's approach:

* Universal Credit (UC) aims to improve work incentives and to make the transition to work easier, simpler and more attractive financially; The Work Programme (WP) uses private and voluntary sector (PVS)providers, on a payments by results basis, to move long-term unemployed people into work;

* And changes to both ESA (a much tougher Work Capability Assessment-WCA) and IS for lone parents (now only available to those with children of five and under) will aim to move many of those currently on inactive benefits onto JSA and ultimately into work.

The first important point is that there is a substantial degree of policy continuity underlying this approach:

* UC represents an integration of the benefit system with the tax credit system. In this sense it merely represents a further evolution of an approach which began with the introduction of Family Income Supplement, later Family Credit, as long ago as 1971;

* The use of the PVS to deliver employment support, with payment on the basis of outcomes, began with Employment Zones and the New Deal for Disabled People in the early 2000s. The WP ostensibly "replaces" the Flexible New Deal, but is delivered by many of the same providers under a contractual and procurement structure administered in a very similar fashion by DWP;

* The moves to reduce the age of eligibility for IS for lone parents began in 2008; similarly the WCA was already planned by the previous government. In both cases the speed and extent of change have accelerated, but the direction is almost identical.

There is more emphasis, both actual and rhetorical, on sticks rather than carrots, and on the use of private sector providers. This in part reflects financial realities carrots cost money- rather than a substantive shift. Some policies that have got considerable media and political attention (for example, Mandatory Work Activity, or the restrictions on overall benefit payments to very large families) are largely peripheral, and do not affect a large number of claimants.

What are the likely impacts?

So will this acceleration, rather than directional change, achieve the desired result? Much will depend on the wider macroeconomy. But I am sceptical that the reforms which have got the most attention--universal credit and the work programme--will have a huge impact.

The government has argued (DWP, 2010) that, "Based on available evidence, a plausible estimate is that the Universal Credit will reduce the number of workless households by up to 300,000". The evidence base for this assertion (DWP "internal analysis", based on "plausible assumptions") seems exceptionally thin, and certainly the DWP has not been willing either to publish it or to include the financial implications of this estimate in their published impact assessment. In fact, the direct labour market impact of UC--which in any case will only be introduced gradually, and from 2014--is likely to be relatively small. IFS research (Browne, 2011) shows that for the vast majority of households, the direct impact on work incentives is marginal, with the negatives and positives roughly balancing. Given this, it seems highly unlikely to have a substantial impact on the numbers on benefit.

As for the WP, it is as yet unclear whether it will be substantially more effective than the programmes it replaces, such as the Flexible New Deal; given the similarities, most analysts are not expecting a major change. But even if it is more successful, it simply isn't likely to get enough extra people into work to have a large impact, as can be seen from simple back-of-the-envelope calculations. At the moment, the DWP forecasts that about 700,000 people a year will join the WP. But of course total volumes flowing through are very different from additional employment; many people flowing through the programme would have got jobs anyway (deadweight); others, even if the programme is successful, won't get jobs; and others will get jobs but return to benefit at some point. In order to calculate the impact on employment, we need to know what the additionality is (that is, how many more people the programme gets into jobs, relative to the expected deadweight); how long they stay in those jobs; and what, if any, displacement effects occur (that is, do they take jobs that otherwise might go to other unemployed or inactive people).

We have considerable experience from existing labour market programmes, delivered both by Jobcentre Plus and the private and voluntary sector, that suggest that--for a successful programme--we might hope to see additionality of 5-10 percentage points (see Portes, 2011, for more detail).

So let's make the following--on balance, quite optimistic --assumptions:

* additionality of 10 percentage points for JSA customers and 5 per cent for other customers (this reflects the fact that JSA customers are likely to be somewhat less disadvantaged in labour market terms);

* for these 'extra' people into work, a duration off benefit of one year for JSA customers and two years for other customers (compared to what would otherwise occur). This might be somewhat pessimistic, since some DWP work on NDYP and NDDP has shown longer lasting impacts, but these were relative to a counterfactual of no intervention at all (or just the basic JSA regime);

* no displacement impacts at all; that is, that the WP has no negative impacts on the employment probabilities of non-participants. This is obviously an optimistic assumption, and might seem counterintuitive--surely some participants will take jobs that otherwise would have gone to other people--but analysis in the UK, albeit not conclusive, has consistently failed to show evidence of displacement.

On these assumptions, the net impact of the WP on the equilibrium level of employment and the number of benefit claimants will be about 70,000. It is emphasised that this is an illustrative calculation--it is emphatically not a prediction; but it is likely to be the right order of magnitude. This successful outcome would represent something over 1 per cent of the number of people currently on out-of-work benefits, and about 0.2 per cent of the number of people currently in employment. So we should not expect the Work Programme to transform the operation of the labour market.

[FIGURE 3 OMITTED]

What about the changes to eligibility for IS and ESA? This is the biggest unknown. These changes will affect a large number of claimants. There are now over 100,000 lone parents claiming JSA rather than income support:

The sharp rises in the number reflects policy changes, as lone parents with successively younger children were moved from IS to JSA.

Even more people will be affected by retesting and tighter eligibility for ESA; so far, this has led to only a modest reduction in the numbers on ESA, which have fallen by 150,000--about 6 per cent--over the past five years. This relatively modest progress to date partly reflects the speed of retesting, and partly that many of those whose ESA claims were rejected are stuck in a lengthy appeals process. Over time, the impact is likely to be much larger.

The big question is whether making lone parents and those who would previously have claimed ESA claim JSA will in fact lead to them moving off benefits into work. The result of expenditure reductions--as well as far more claimants in this position--means that such people are likely to get less, not more, support than they would have done under previous programmes like NDLP and Pathways to Work. Moreover, an increasing number of those moving to JSA will have been on benefit for a very long time, and may possess relatively little in the way of skills and qualifications that will make them attractive to employers. However, the benefit regime is tougher, and getting more so, and the financial incentives will be progressively strengthened by the eroding value of JSA. This will be a real test of whether the use of sticks, without much in the way of carrots, is an effective approach to active labour market policy--particularly in a period of weak economic growth, when competition for jobs is likely to be relatively intense.

To conclude, the current very high level of claimants on out-of-work benefits is the result of a combination of policy mistakes and demographic and social change, mostly in the 1980s and 1990s. Active labour market policies, generously funded and consisting of a mixture of carrots and sticks, reduced these numbers somewhat in the 2000s, but progress was modest compared to the size of the problem. In a much more constrained environment, the government is undertaking two major reform programmes--Universal Credit, and the Work Programme--but while each has commendable features, in themselves they are unlikely to have much impact on the numbers on benefit. The bigger impact will be that of simply restricting access to the benefit system; but whether that leads to substantial reductions in worklessness, or simply to reduced expenditures and more people on JSA, or worse still out of work and outside the benefit system with little or no income, remains to be seen.

REFERENCES

Browne, J. (2011), 'Universal credit: impact on work incentives', Institute for Fiscal Studies, http://www.ifs.org.uk/conferences/uc2011_browne.pdf.

Corry, D., Valero, A. and Van Reenen, J. (2012), 'Economic performance since 1997: growth, productivity and jobs', LSE, Centre Piece, 17(1), Spring.

Department for Work and Pensions (2010), Universal Credit: Welfare that Works, November.

Layard, R., Nickell, S. and Jackman, R. (1991), Unemployment: Macroeconomic Performance and the Labour Market, Oxford, Oxford University Press.

Portes, J. (201 I), 'Will the work programme succeed', Working Brief, Summer.

NOTES

(1) http://conservativehome.blogs.com/platform/2010/12/chrisgrayling-mp- how-the-governments-work-programme-will-tackle-long-term-worklessness- .html.

(2) Analysis by my former DWP colleague, Rebecca Endean.

Jonathan Portes, I am deeply indebted to Bill Wells of the Department of Business for the data underlying figure 2 and for very helpful discussions about these issues, both in the context of this article and over a number of years; and to Rebecca Endean of the Ministry of Justice. Neither of course has any responsibility for the analysis or conclusions.