Failing to measure success in the Work Programme

26 September 2014

The Work Programme aims to increase the number of unemployed individuals finding long term work through the introduction of payment by results. This means that finding a way to evaluate performance is vital. Currently, the success of the individual service providers is based upon the number of participants who receive a job outcome per year as a proportion of the number of unemployed individuals referred to them. The provider is then compared to set minimum performance levels (MPLs). This method of assessing success is flawed for a number of reasons.

Firstly, the method used to evaluate the success of the providers is problematic, as outlined by Paul Lester CBE in his government commissioned review on MPLs. Providers’ performance is calculated as the number of job outcomes per year as a proportion of new referrals that year, despite there being no obvious relationship between the two. For example, Candidate A receiving a job outcome at the start of the year will be included, despite perhaps being referred two years previously. Candidate B who has been referred at the end of the year will also be included, despite there being very little chance of him yet receiving a job outcome. The use of one year as a time frame to compare the two is arbitrary, and influxes of referrals and fluctuations in outcomes have the potential to seriously distort the figures. Assessing providers on a cohort basis would remove this fluctuation, and create a more intuitive and representative assessment of each provider.

Further, pre-determined MPLs do not respond to changes in the job market. For example, a boom in the market would lead to a higher success rate for the provider, but no increase in the static MPL. Incentive payments are given to the contractors for performances higher than the MPL and there have been concerns companies could then make potentially unreasonable profits. What is more likely is a lack of profit for the contractors – Deloitte’s selling off of its contract last year was widely reported as a strategic exit for this reason. Payment by results was introduced by the government to ensure that contractors were providing a high quality service. However, when these financial incentives are so greatly reduced by hard to reach targets, the motivation to invest the best time and effort is removed.

This is exacerbated by the fact that MPLs are standardised across the country. Latest ONS statistics show unemployment in the North East as 9.4 percent, but at only 4.4 percent in the South East. When contractors are so clearly facing a harder task in different parts of the country, to assess and reward them based on the same target is unreasonable. Cumulative figures over the whole Work Programme show that in the North East, job outcomes have been found for 20.5 percent of referrals, in comparison to 22.7 percent in the South East. Regional differences are then reinforced, as the government pays the most money to the companies doing well, which are often in the easiest areas to succeed in, whilst the hardest parts of the country suffer from lack of funds and lack of incentive, leading to worsening outcomes.

It is hard to truly assess the success of the Work Programme when the methods used to do so are so flawed. If the government is committed to using payment by results in welfare as a way to achieve more for less, then the models used to assess the performance of their contracts must be reformed. The current model is not representative or effective, and the incentive structures lost by these errors have far-reaching consequences for those that need help the most.

Catherine Wyatt, intern at Reform



Failing to Measure Success in the Work Programme | Catherine Wyatt

29 July, 2015

[…] This was published on The Reformer Blog, whilst I worked at Reform, an independent, non-party think … […]