Flying a kite: a convertible personal tax allowance

This is a partly-baked idea, based on reform of the Personal Tax Allowance.  It would allow people to take the allowance in cash rather than by setting it against their liability for tax.  The current allowance of £11500 is worth £2300 (that is, 20% of the allowance) to everyone who pays income tax.  If it was possible to convert it to a cash payment, it could benefit many more people.

In round numbers, there are about 65 million people in the United Kingdom.  30 million (24 million people of working age, 6 million pensioners) of them get the personal tax allowance.  For most of these taxpayers, tax and allowances are calculated at source, so the benefit is invisible.  The cost to the Treasury is something in the region of £70 billion.

If 30 million people receive the allowance, 35 million don’t. 15 million of them are 19 or under, so there are roughly 20 million adults (both pensioners and people of working age) who don’t currently get the benefit of the Personal Tax Allowance.  The proposal is simple:  to make it possible for those people to elect to receive the equivalent cash value of the allowance (that is, £2300 pa, just under £200 per month).

The idea may be off the wall, but it’s not outlandish: it asks for nothing better than what is already received by the other 60% of the adult population.  The only net beneficiaries would be people with incomes lower than the tax threshold.  About 6 million of them are people over 65 on lower pensions, 7 million are adults of working age receiving benefits, and the remaining 7 million are either people on low pay or non-workers.  There is no obstacle to people in work receiving the same option – there would be no net cost, apart from the mechanical administration of the payment – but there is no financial advantage for people who are already in receipt of secure incomes to make the switch.

The basic model should be easy to operate.  A tax year runs roughly from April to March.  Records of tax allowances are already kept for everyone with a National Insurance number.  People who wanted to convert their tax allowance into a cash payment would need to make their election in good time before the tax year – the obvious deadline is the time of the submission of tax returns by the end of January.  A person who made that election would then have a tax allowance for the forthcoming year of zero. No further administrative test is necessary.

Many Basic Income schemes have suggested that the Personal Tax Allowance should simply be abolished.  I think it’s important that there should be a choice, for three reasons.  First, PAYE works, and it works for millions of people.  Getting rid of it would complicate people’s lives and breed resentment; it would also more than double the administrative burden.  Second, there are some people, especially self-employed workers, for whom a zero tax allowance would be difficult to manage or live with. Third, choice matters in its own right. If the choice proved popular among workers, it could pave the way for other, more ambitious schemes; we can cross that bridge if we come to it.

This scheme could all get more complicated, but it’s important to avoid that.  Changing back and forth between PAYE and cash payment, week by week or month by month,  would be complicated and could be error-prone.  Disregarding the payment in means-tested benefits would be inequitable, because if the benefit is allocated continuously for at least a tax year, some people would have the payment and others wouldn’t; and if it’s not allocated continuously, there would have to be a complex set of mechanisms to cope. People who owe back tax typically get an adjustment in their tax allowance; if people do not have a full personal tax allowance they should not be eligible.

In principle this scheme could cost up to £46 billion (20 million times £2300), but there are reasons why it wouldn’t.   First, anyone who is currently getting more than the tax threshold in the course of a year already receives the equivalent benefit.  Most people would remain on PAYE – the main reason for electing for a change would be to secure a more stable income.  Most people who become unemployed pass through unemployment: about half the people who drop out of the PAYE system return within six months.  They will typically keep their tax allowance in PAYE.  So not everyone who receives benefits will be getting a converted allowance.  Second, those who earn but are below the tax threshold already receive part of the equivalent benefit.  Third, most important, the cash payment will have to be offset against benefits for about half the relevant population – that is, pensioners and adults of working age on means-tested benefits.  I’m not going to pretend that the interactions with benefits are straightforward – far too many pensioners who are entitled to Pension Credit don’t actually get it – but it should all mean that the net cost is likely to be in the region of half the maximum, something over £20 billion.  That’s still not cheap, but it’s both practical and much less expensive than many alternative proposals that have been canvassed.

This is a proposal, then, for a relatively small, limited benefit that would be unconditional and easy to administer.  It would mainly work in favour of people who are not well protected in the current system: people on unstable incomes, students, people who are not receiving benefits, people with home responsibilities. The argument for a scheme like this is that it would make a difference, not that it would answer every problem.  It would not settle the problems of the benefits system, and it is not big enough to make any very large difference to poverty.  However, money from one source can be mixed together with money from other sources, and a small benefit like this would make sense alongside other benefits.  It would offer a limited but stable source of supplementary income to many people on insecure incomes, and it would offer some support to people who currently have none.  What set me off on this week was a challenge made by Kathy Mohan to Theresa May in the street: “I can’t live on £100 a month”.  She’s right; she can’t.  We need to make sure that part of the benefits system, at least, is stable, secure, guaranteed and permanent.

The cost is the main reservation.  It’s much less than the cost of many Basic Income schemes, and it’s better targeted – getting money to just the people that Basic Incomes are supposed to reach, and doing it without making any of them worse off.  But there is always an opportunity cost.  I’ve not run the figures, but I suspect putting the same money into Child Benefit would have a much greater impact on poverty than this proposal, which offers more to pensioners with moderate incomes, students and women with home responsibility. On the other hand, it probably compares favorably in distributive terms to current proposals for supporting university students and FE, which will cost over £14bn, but have other arguments running for them.  It’s all a question of priorities.  Nor is this the only half-baked scheme I’ve canvassed on this blog.  Think of this as the glimmer of an idea, rather than a personal recommendation.

I also have to admit to a certain apprehension about putting potentially dangerous ideas into the public domain.  One of my first supervisors, Della Nevitt, once scribbled a back of an envelope calculation in the Ministry for Housing and Local Government and it mushroomed into the Rate Rebate,  Rent Rebate and Housing Benefit schemes, the last one in its day described as the greatest administrative fiasco in the history of the welfare state.  I need to think more about this, and it’s possible that other people can see a lurking catastrophe in the idea I’ve not thought about yet.  I’d be grateful for comments.


    • Paul Spicker

      There are some economists who’d argue that all taxation has a disincentive effect, and so that the slightly lower rates experienced on low pay ought to have slightly less. The empirical evidence doesn’t bear out either the general proposition, or the subordinate one about low pay. In relation to the general argument, studies of incentive effects have been inconclusive: for everyone who’s put off working by higher marginal tax rates, there’s someone else who works more to reach a target income. In relation to the specific case of low paid employees, I’m not aware of any study that offers evidence that varying the personal tax allowance makes any difference. For there to be such an an effect, people would need to know and recognise the relative importance of the tax code compared to other factor; but any effect it had would be confounded by the withdrawal of benefits, by the impact of national insurance contributions, and by the tax regime – people who move into employment will usually have a period with an ’emergency’ tax code, meaning they are not ‘taken out of tax’ – rather, they pay tax for a couple of months even if their previous income doesn’t merit it. There’s really no reason to suppose that changing the level of the tax allowance would have any direct effect on behaviour – but this proposal isn’t doing that anyway. It’s changing how it gets paid, which is something quite different.

      All of that comes from taking your question literally. However, when most people talk about ‘incentives’ they usually mean something else. Some people are concerned with giving rewards for correct behaviour, and there have been tax allowances that ‘send a message’ – the married person’s allowance was supposed to do that – but the personal tax allowance doesn’t. It has much more to do with fairness. It’s a method of calculation – a way of gearing the amount people pay in tax so that people on lower incomes pay less and people on higher incomes pay more. (It doesn’t do this perfectly, but that’s not the main point here.) I think in this case that an argument for fairness becomes an argument for extending the advantage to people on lower incomes more widely.

      This argument only relates to work, not saving. I haven’t made any proposal about the personal savings allowance, or any other tax allowance.

  1. Gareth Morgan

    Personal allowances apply against taxation of pensions and other savings as well as earnings. As most MTBs are based on net, post tax, earnings, the allowances do have a direct effect on the return from earnings. I’ve long pointed out, with many others, taht increasing personal allowances doesn’t help the lower paid in the way that politicians trumpet. An increase in allowances increases net earnings which reduces benefits. The MDR is therefore much higher for low paid than higher paid individuals. from that perspective the allowances are a part of the support system and a part of any incentive / disincentive as well. Distributing them as, effectively, a form of BI must open up that argument.

  2. Paul Spicker

    I take the point about savings, but once again I have to say that I’m not aware of any work that establishes that the allowances have any incentive effect. You suggest that “the allowances are a part of the support system and part of any incentive/disincentive as well”. You can’t assume an incentive or disincentive effect; it has to be demonstrated. As I said out the outset, many economists assume that financial or distributive outcomes must change behaviour; that’s a common mistake, but it is a mistake. In the first place, incentives have to be understood at least in terms of costs and benefits, and any statement that something is an incentive ‘other things being equal’ leads to absurd conclusions: a funeral grant is not an incentive, other things being equal, to become dead. You can’t tell what the incentive to work or save is without knowing what the alternatives are. Second, responsiveness to economic stimuli – usually expressed in terms of elasticity – varies. You can’t tell how much effect an incentive might have without that information. You’ll find a detailed set of arguments about the nature and operation of incentives on my open access page.

    It follows from that that I also think most of the incentive arguments raised about BI are bunkum. The incentive argument that’s usually applied to BI is that given the choice between work and leisure, people will not choose to work. There is some contrary evidence – the reports from India suggest that people might be more likely to engage in economic activity, not less – but there is also evidence to back it up: over the course of a century, we have moved from more than 80% of old people working to about 90% not working, and it’s the existence of the pension that has made that possible. From the point of view of economic theory, the problem with that second observation is that it’s taken more than seventy years for it to happen. Incentives are usually conceptualised as something timeless – people maximise their utility instantaneously. The fact that withdrawal from the labour market has taken more so long to materialise suggests that something else is going on – that the existence of pensions has changed the culture and (from the point of view of economic theory) that it’s ‘shifted the curves’. If that’s right, current experiments designed to assess work incentives of BI are probably not going to be able to assess the impact – they won’t be able to predict outcomes unless they run for a few decades.

Leave a Reply