Considering and reconsidering…

Some decisions by DWP staff are subject to Mandatory Reconsiderations. This means that adverse decisions like benefit sanctions have to be checked by someone else. Answering an FOI request, the Department for Work and Pensions have said that it is a Key Performance Indicator that 80% of mandatory reconsiderations should uphold the original adverse decision (that is, adverse to the claimant of course!)

This would naturally be cause for real concern if it meant that those staff carrying out the reconsiderations were themselves to be marked on whether they upheld 80% of the decisions sent to them to check. Obviously if that were a Key Performance Indicator in their own individual Performance Agreements then they would be between a rock and a hard place as if in fact they were sent a bundle of dodgy decisions they would have to find some way of upholding them or suffer adverse consequences themselves through a poor performance marking with knock-on effects on their pay, promotability or even continued employment. That’s clearly what some people consider it means

However, on reconsideration I have to say that I think this unlikely, and that DWP has not ‘admitted’ anything like that. My experience is with HMRC rather than DWP but has included responsibility for performance management policy as well as for reviewing policies and procedures for reconsiderations of tax credits decisions and complaints about them. Presuming that DWP is similar to the benefit-administering part of HMRC, the department’s naked self-interest is that initial decisions be made correctly.

The reason is that it is an enormous waste of time and effort if incorrect decisions are made. Remember that in their funding, departments are squeezed by the Treasury on their operational costs not on the amounts of benefits they pay out. The latter are out of their control: for example HMRC and DWP can do nothing to prevent a recession and people becoming unemployed nor can they prevent people having children (or indeed coming to the UK). Only government, through Parliament, can affect the amount spent on benefits by changing the law on eligibility to benefits and entitlement to particular amounts. The broad role of HMRC and DWP is to administer the law efficiently and they are given money for the staff, buildings, IT systems etc necessary to do this. These costs are subject to negotiations with the Treasury, but the departments are not, and cannot be, given performance indicators about the amount of benefits they pay out. In any set of circumstances in the economy and real world there will be an amount that the law provides should be paid, and the departmental performance indicators are couched only in those terms. For example the targets for reducing error and fraud are in terms of a percentage of the amount that the law provides should be paid rather than an actual number of £ and on take-up rates are a percentage of those the law provides are eligible rather an actual number of people. So the departments’ rational self-interest is always in getting as many initial decisions correct as possible so as not to incur the costs of review and re-work of cases.

Now, if one were to ask whether the department’s frontline staff were making a good job, which thus means an efficient job, of their initial decisions about eligibility and entitlement in law to benefits (and related matters like sanctions) then one way of monitoring this will be what checks on those decisions reveal. It seems not at all unreasonable that the department should hope that at least 80% of these checks find that the original decision is correct. One might think that this is fact rather a low bar, and a rather lower bar than in many parts of the private sector. You wouldn’t think it acceptable if 2 out of every 10 loaves of bread you bought (whether from Waitrose or Morrisons) were mouldy or if 2 new cars in 10 were not actually safe to drive, or a bank machine dispensed the wrong amount of cash 20% of the time. On that basis, the higher the percentage the department sets itself for the upholding of adverse decisions the better. Remember too that there is no mandatory reconsideration of decisions that are not adverse to the claimant. So if the DWP Work Coach decides not to sanction you, that decision isn’t reviewed other than as part of the manager’s routine supervision of him or her.

In the circumstances, there’s clearly a risk of misunderstanding what the DWP’s performance indicator might mean. We can’t say on the very limited information available that the DWP is requiring its staff to uphold 80% of adverse decisions, let alone wilder speculation like that 80% of decisions must be adverse. The Social Security Advisory Committee has had a keen interest in this area and consultations have been carried out and reports written. A lengthy piece of work in 2016 included these statistics showing that in the first half of 2014, 30%–40% of adverse JSA decisions were being changed on Mandatory Reconsideration and 63%-75% of adverse ESA decisions. That seems to me to be evidence of a really poor frontline performance with a quite unacceptable number of initial adverse decisions being made incorrectly. It would be quite right for senior management to want to improve this and quite reasonable for them to measure success by a reduction in the number of adverse decisions found to be wrong.

DWP appear to have had some success with this as the number of adverse decisions found to be incorrect on mandatory reconsideration had dropped to only 12.5% by the end of 2015 taking JSA and ESA together i.e. 88% upheld. The FOI information suggests that has been maintained with a rate of 87.5% from April 2016 to March 2017. While the real measure of underlying performance is the overall accuracy of all initial decisions, it is much harder to assess that because it would require checking all of them and not just the adverse ones. This will in fact be done, but by looking only at a sample of all cases. The results are reported in a range of ‘error and fraud’ publications that for DWP include both official error and customer error. As an aside, HMRC report no official error but that is an absurdity for another time.

I’m sure that DWP might well have done lots of things actively to improve. These might have included better training, improved guidance and supervision etc etc. However the actual caseload has diminished significantly, particularly for JSA as the unemployment has come down to record lows, and it might be that the very high rates of error have come down also because staff have been less overwhelmed by high volumes.

In conclusion then, on this evidence, it seems entirely appropriate for DWP to set itself a target for a high proportion of initial adverse decisions being upheld. Personally, I don’t think it goes far enough. An adverse decision is supposed to have a meaningfully adverse impact on the claimant and the press are full of horror stories resulting from such decision ranging from food-bank shame to suicide. I don’t think anyone should find it acceptable that the DWP can get away with as many as 20% of such decisions being incorrect and still meet its performance indicators. Its objective should be that initial decisions are correct, and for initial decisions that are adverse one would surely want an especially high proportion to be correct. To the extent that the ‘upheld rate’ of Mandatory Reconsiderations is an indicator of underlying frontline departmental performance then it should be not 80% but 100%….