Universal Credit is a bad idea badly implemented

Recently I complained about how government incompetence was now routine. One example of this is the government’s reform of benefits, known as Universal Credit. It is a particularly good example of what I was talking about. It is a misconceived idea, poorly executed.

Interestingly, the political conventional wisdom is that Universal Credit is a sound idea, but lamentably executed. And yet problems with execution usually reflect problems of design. Universal Credit was originally conceived by the Coalition government’s Work and Pensions Secretary, Iain Duncan Smith (known as IDS). It is designed to replace a series of benefits, including unemployment benefit, tax credits for those on low incomes, and housing benefit to help those with low incomes pay their rent. The idea is that the level of benefit is scaled back as income rises, so that it is always worth a claimant’s while to take a job and increase their hours or rate of pay. Previously many of these benefits (though not tax credits) had a cliff-edge effect: if your income exceeded a threshold then you lost all of your benefit. It was sold as a simplification that improved incentives for people to be self-sufficient, and welcomed across the political spectrum. But it is only simple in the sense that it is relatively easy to understand what your entitlement is if you know your income. There is some law of government which runs that there is nothing more complicated than a simplification.

The problem is knowing what a claimant’s income actually is at any point in time and therefore how much they should get. This faces two problems. First is that claimants are by definition poor and struggle to manage cash flow. Keeping them waiting for their money while you work things out is not a good idea; paying them an advance runs into difficulty if the advance is too high and so the clawback becomes onerous – an issue that led to notorious problems for Labour’s tax credit system. The second problem is a complete lack of trust: the idea that claimants should report their own income would not be taken seriously by any politician – probably with some reason, though the contrast with the way better off people are treated is striking; the benefit had to be linked to hard data on income. IDS’s answer to this was to link the information gathering to systems for Pay As You Earn (PAYE) for administering income tax and national insurance deducted at source. But this was a highly complex undertaking, and, since mostly PAYE works on a monthly cycle, still leaves a six-week wait for new claimants, and a monthly payment cycle, for people used to budgeting by the week.

The complexities associated with this should have been obvious from the outset, but IDS is a particularly ideological and obstinate character, and refused to accept the scale of the problems. Implementation deadlines kept being put back.  When he resigned last year, I predicted that Universal Credit would have to be killed off, along with the fact that Theresa May would be the next Prime Minister. But although Mrs May is now PM, Universal Credit lives on. But the six week wait and monthly cycle are now becoming a hot political topic. A weak government will have to give ground, and doubtless a fudge is in the works. But that fudge will lead to further difficulties.

That’s not the only problem with Universal Credit. I haven’t mentioned issues that arise with self-employed people, of which there are increasing numbers amongst low income earners (known as the gig economy). Another issue is not with the fundamental design, but the way in which the government has turned the relatively gentle withdrawal of benefit into something much sharper, undermining the incentives for people to increase income. This was part of the government’s promise to reduce the costs of welfare. That is another aspect of the self-defeating nature of austerity – savings that build up future costs.

But it is the administrative architecture that is the real problem with the benefit, and why it will never work satisfactorily. The DWP wants a giant database for everybody with all the real time information on it that you might need to administer tax, benefits and any other entitlements, which can then be automated. This is a wildly impractical idea, but promoted by IT consultants on the basis that they will be able to get out with their fees long before the inevitable collapse. To be fair, I suspect that many IT professionals are just as deluded about the capabilities of IT as the shallow think-tankers who promote this sort of thinking in government. Even in a world of massively more capable IT, trying to get accurate, real-time data on incomes in a modern, free and diverse economy is an impossibility.

Are there other approaches? The US operates a rather similar idea in its tax credit system. This is based on the annual filing of tax returns by all citizens. This means that any amount owing comes back as an annual payment – something that makes the current discomfort with six-week waits and monthly payment cycles look laughable. Poorer citizens have to borrow against the expected credit, inviting high financing costs. But whatever the virtues of the American approach, it is incompatible with the way the current British tax system works. We deduct tax as much as possible at source, and tax free allowances scoop up a host of small payments that would otherwise be within scope. Only a minority of citizens have to file tax returns, with all the hassle and expense this entails. Extending Self Assessment (as it is known) to the whole population is not an inviting prospect – though the system is much simpler than its American equivalent.

So what to do? A popular idea on the left is to replace universal credit with a universal basic income to which everybody is entitled – so eliminating the need to gather data on incomes. The problem here is that it would require a big expansion of tax if this income is going to be at the level of maximum universal credit (up to £3,800 per year). My back of the envelope calculation is that the cost would be about £200bn per annum (excluding pensioners) – about the same as the entire income tax take. This is not impossible: overall tax receipts are currently £700bn, and a good proportion is being paid out already (especially if you include the tax free allowance). But it brings with it a host of issues around consent and enforcement. And would that level of income really be enough?

The alternative is to move away from these centralised, universal systems to ones that are more piecemeal and community-oriented. That would mean that each local area would need to work out its own system for alleviating poverty, including social housing, and job guarantees as well as cash benefits. All this to be administered by one-stop-shops that are able to deal with each person’s needs in the round.

But that would require such a radical change of thinking that I cannot see any sign of it happening. Our politicians, and political activists generally, are addicted to creating giant, universal systems administered through online portals, that have failure designed into them.