Government’s flagship Welfare-to-Work scheme – failing the ‘harder to help’ by design?

Eleanor Carter (PhD researcher) and Dr Adam Whitworth, Dept of Geography

The Work Programme is Government’s flagship welfare-to-work initiative and uses non-state providers to support long-term unemployed people into paid employment under a payment by results funding regime. Within such quasi-marketised systems the problem of inequitable and/or inadequate treatment of claimants with multiple or complex needs (typically referred to as ‘parking’) is notoriously pervasive and difficult to mitigate. Such claimants are especially vulnerable to being parked when schemes also operate a light touch minimum standards and monitoring framework, face tight performance targets and costs pressures, and/or are dominated by for-profit providers – all of which apply in the case of the UK Work Programme.

Policy makers, however, have suggested that the Work Programme is explicitly intended to close the ‘performance gap’ between disadvantaged and easier-to-help claimants. To counter tendencies for providers to neglect those participants whose barriers to work are greater the Department for Work and Pensions relies on a differential payment structure with nine payment groups each with different outcome payment levels for providers. These payment groups are based relatively simplistically on claimants’ prior benefit type alone and from the outset there has been criticism that these groups are overly crude and a poor proxy for distance to the labour market, hiding more than they reveal. With income for providers following payment groups the concern is that perverse incentives are built into Work Programme, leaving ‘harder to help’ claimants inside each payment group vulnerable to being under-supported relative to claimants with simpler employment support needs inside that same payment group. Our research set out to explore claimant variation within payment groups and hence whether they are likely to mitigate provider incentives to ‘cream’ and ‘park’.

Our analyses used detailed survey data (Understanding Society) to build a statistical model predicting which participants enter the sustained employment needed to trigger a Work Programme job outcome payment. Our analysis leads to two important findings for policy makers to consider as they think through the next phase of Work Programme commissioning. Firstly, under the current payment group model the probabilities of participants entering paid work vary more within payment groups than between them – contrary to the government’s intention, the payment groups seem more likely to design in rather than design out ‘creaming’ and ‘parking’. The chart below shows this. We model here the probability that our sample of matched Work Programme participants move into the sustained employment that Department for Work and Pensions (DWP) consider a ‘successful job outcome’ and that triggers an outcome payment for providers. The chart on the left shows these probabilities separately for four of the main payment groups within the programme, with the predicted probability of moving into sustained work increasing as one moves vertically up the y-axis. The triangles show the DWP’s view of the average likelihood of each payment group moving into sustained work and it is around this that the differential payment levels have been set. In contrast, what our analyses highlight is the large degree of vertical spread in the modelled likelihood of moving into sustained work within each payment group. Indeed, virtually all of the overall variation in these predicted probabilities is seen within the payment groups rathat than between the payment groups as designed by DWP and as implemented through the current differential payment model. The result is the designing in rather than out of perverse incentives for providers to ‘park’ (i.e. offer little real support to) claimants with more complex employment support needs (those with the lowest predicted probabilities) – precisely the opposite to what DWP intended to achieve with these various payment groups and differential pricing model.


The second key implication of our work is that whilst statistical profiling of claimants is by no means perfect, and will never be a single cure for risks of creaming and parking, they can offer part of a sensible, joined-up response: it is possible to use the information that we have – or could easily collect – about claimants to better predict which participants are likely to enter sustained work. Thinking about payment groups according to claimant need (cutting horizontally across benefit types) offers claimants a better chance of receiving the level of support that they need, policy makers a better chance of realising performance targets, and taxpayers a better chance of seeing value-for-money. We outline in the paper a possible way to do this.

Irrespective of the outcome of the 2015 election DWP Ministers will be considering the future of the Work Programme. Despite the difficulties that Work Programme has faced there is broad cross-party agreement on the scheme’s overarching design principles and, indeed, some good reasons for such a design approach. But there are also various real concerns with Work Programme performance and claimant experiences so far and several ways in which policy makers will be looking to make changes and improvements. The disappointing performance of ESA claimants is perhaps the area attracting the most noise, and rightly so. Work Programme Leavers schemes separate ESA claimants back out into their own separate scheme and such an approach has obvious potential for Work Programme itself. Our work however suggests that whilst the focus is right (current Work Programme design is not delivering as we would like for ESA claimants) the solution is not and that we need to move beyond a fixation with benefit types and to instead think about employment support designed in a personalised way around the individual needs of each claimant, whatever their benefit group.

The full article is published in the Journal of Social Policy and is available freely to all via open access.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s