The Guardian has a slightly strange write-up of a piece of research around the government's flagship Work Programme, which is aimed at getting long-term unemployed people back into work (although the initial statistics imply it is less than efficient at doing so [2]). The programme is structured in a "black box" manner; providers are entitled to, within reason, offer whichever schemes they think will work best to participants, and are paid by result.

But, Patrick Butler writes [3]:

Welfare firms are involved in widespread "gaming" of the Work Programme, with the most vulnerable jobseekers often ignored because they are too costly to help, according to new research into how the government's flagship employment initiative is working in practice.
Providers privately admit they are focusing resources on the "easy customers" who are more likely to generate a fee, and sidelining jobless clients who require more time and investment to become ready for work, a process known as "creaming and parking," the study says.
It concludes that the quality of services offered to jobseekers is being undermined because the design of the Work Programme, in which companies are not paid until customers have been in work for two years, creates such huge financial stresses that many providers have little option but to cut corners.

The last paragraph is not strictly true. Providers are paid periodically throughout the two years that participants are said to be active. They receive a referral fee when an unemployed person arrives on their books, and then further payments when they find that person work, and periodically while that person is in work up to the two year deadline, when they are deemed to be back in stable employment.

The most obvious way of gaming that system is indeed relatively frequent: taking the referral fee for a new "customer", and then proceeding to ignore them entirely. Since there is not a huge amount of variation in the fees depending on how difficult it might be to find work, that usually results in people who are scarred from the effects of long-term unemployment being taken into the programme and left languishing while their referral fees are used to subsidise training for more easy-to-help participants.

That wheeze is likely to be short lived, for a couple of reasons. The first is that it won't result in many people actually getting work, and so the workfare provider's overall statistics will look terrible. If the government has a modicum of competency, that will be taken into account when the next contracts go out.

Competency, of course, is not guaranteed, but luckily the referral fees were only ever intended to be short term. They are important to getting the scheme going, but the intention is that the training for one cohort of jobseekers should be paid for, not with the referral fees, but with the profit from the previous cohort. That way, the system is true payment by results: if you don't find someone a job, you don't get anything.

The gaming Butler describes is a different sort. The black-box model the Programme runs on allows providers to subcontract work; and that seems to be where the trouble is starting:

The study cites a small private-sector provider which complained that big corporate providers, known as "primes", would keep "job-ready" customers for themselves while passing on more difficult cases to subcontractors. "It's not being PC but I'll just say it as it is … you tend to get left with the rubbish; people who aren't going to get a job … If the [prime] thought they could get them a job, they wouldn't [refer them to] someone else to get a job."

That doesn't seem to be symptomatic of anything other than bad business on the part of the small private-sector provider. There's nothing making them subcontract with the big corporate providers. Presumably they thought they could make a profit. The fact that they can't on the terms they'd agreed just means they should draw up a better contract.

Just as we can hope the government will take performance into account when offering the next set of Work Programme contracts, the best situation for the subcontractors is to stop taking work from corporate providers who offer them bad terms. That is, after all, how capitalism works.

Not that it has to be that way. Not all of the workfare providers view their job as purely extracting profit from a badly designed system, and at least one major one largely foreswears the possibility of boosting income by gaming it. Unsurprisingly, it is not one of the providers backed by private equity.

The Work Programme is frequently poorly designed, and many — but not every — provider is out to milk it for all it's worth. But the problem with it isn't that there is isn't enough money floating in the system.