Basically unaffordable: The Economist gets its sums wrong on Basic Income

An article in The Economist last week condemned Basic Income as being “prohibitively expensive”; but the criticism  makes some very basic mistakes.

In 1970 James Tobin, an economist, produced a simple formula for calculating their cost. Suppose the government needs to levy tax of 25% of national income to fund public services such as education, policing and infrastructure. Paying for a basic income worth 10% of the average income requires average taxes to rise by ten percentage points, to 35%. A basic income worth 20% of the average income requires average taxes to be 20 percentage points higher, at 45%, and so on. Eradicating relative poverty, defined as income beneath 60% of the median, would require tax rates approaching 85%.

Tobin’s calculation is based on national income. The figure of 35% comes from adding 10% in Basic Income to 25% for public services.  The article’s key mistake comes from confusing national income with average income.  In the first place, almost discussions of average income refer to the median, not the mean.  The median average income accounts for  less than the mean average,  because people in the top half of the income distribution get more income share than the bottom half.   In the UK,  the population median income is £404 pw; the population mean income is £504.  60% of the median share distributed to everyone would be 48% of all personal income, not 60%.

This in turn is less than GNI because GNI is not held exclusively by individuals – a good quarter is corporate.  48% of personal income is 36% of GNI.  Add that figure to the basic 25% tax and we get to 51% of national income, not 85%.  The three figures in the Economist article should not be 35%, 45% and 85%, but 31%, 37% and 51%.

These are very broadly rounded figures, but I don’t think there’s much point in trying to refine them more closely at this stage: the actual cost would depend on the design of the benefit. Costs should be offset against social security reductions (which is one of the key reservations to make about Basic Income schemes) and the abolition of tax allowances.

I’m not an advocate of Basic Income; the apparent simplicity of the scheme is liable to be undermined in practice by the terrifying complexity of people’s daily lives. If we’re looking for a Basic Income to be an element in people’s income packages, however, the problem is not the cost.

Further note, 18th May 2017.  I’ve just been reading Basic Income, by Phillipe van Parijs and Yannick Vanderborght, published 2017 by Harvard University Press.  They do suggest, several times, an indicative level of basic income set at 25% of GDP, and that would be indeed be open to the objection that Tobin raises. 

Arguments about Citizens Income

The proponents of a Citizens Income are getting a rough ride at the moment.  Natalie Bennett, the Green Party leader, is supposed to support the idea, but her truly awful interview with Andrew Neil showed up her reluctance to engage with the core arguments.  She just wasn’t ready to say that much of the funding comes from replacing existing benefits, scrapping basic tax allowances and wiping out Tax Credits.  Now the Joseph Rowntree Foundation has published a sceptical report by Donald Hirsch, complaining that the scheme would cost too much, that people wouldn’t approve of it, and that it’s never been done.  Well, it doesn’t have to be paid for entirely in direct taxation – why would you want to do that?  – and as for practical experience, the precedent already exists in Child Benefit.

I’d question many of the arguments made for Citizens Income, and I detailed some of the problems in a previous post.  Any Citizens Income would need to be partial – that’s not a bad thing – and introduced to cover segments of the population.  Pensions can’t be left out; I’m not sure it makes sense to leave out Housing Benefit either.   The main problem with a Citizens Income is common to any scheme that over-simplifies; if you’re replacing complicated benefits with simpler ones, you will have to be ready to cut benefits to some people with complex special needs.   That doesn’t  mean, however,  that the system is either unworkable or unaffordable.  The question is whether we should do anything like it, and if so how far it should go.

A proposal for Basic Income in Scotland

There are those who’d like to see a Citizens Income in Scotland, or in the UK, and there were several submissions on that basis to the Expert Working Group on Welfare, who made a nod in that direction after representations from Ailsa McKay, sadly missed, who had been firmly committed to the principle.   A report this month from the Centre for Welfare Reform develops many of the arguments.    They don’t offer costings, but they do write:

“In the current UK system the least generous rate of Income Support is £2780 per year.  If this was extended as a benefit to all 63 million people in the UK the total cost would be £169 bn.  This is less than the current costs of all benefits and pensions in the UK (£185 bn).”

I ‘ve been doing some back-of-the-envelope calculations for Scotland.  As things stand, benefits in Scotland cost £17.7 billion, and personal tax allowances – an important part of the potential funding pool – cost probably £6.8 billion.  That makes £24.5 billion, but from that we probably have to knock off some benefits that can’t be replaced by a basic income, such as £0.9m disability benefits or £0.3m payments to expatriates.  The population of Scotland has roughly 1.1 million pensioners, 3.3 million people of working age (16+) and 0.9 million children.  I’ve argued in previous writing for a simple equivalence scale:  1 for the first adult in a household, 0.5 for each subsequent adult and children.  There are 2.4m households, so there are 2.4m first adults.  The pot needs then to be divided into 7.4 million shares, and comes to £3135 pa, or £260 pcm.  Let’s call it £250 a month, because this is after all a quick demo.  There’d be £250 for each household, £250 for each adult, and £250 for each child.

It sounds attractive at first, but I’m not at all convinced.   This would be good for lots of people, but bad for others.  As the figure includes a replacement for Housing Benefit, people on high rents wouldn’t get their rent paid.   Getting rid of personal allowances means that all income will be liable to tax.  Getting rid of Tax Credits may appeal but there have to be different arrangements for low paid work (a higher minimum wage)  and for child care.  On £500 pcm for a single pensioner, and £750 for a couple, pensioners would lose out – Scotland’s Future promises a single pension of £160 pw – and try to explain to anyone  why a blind pensioner should get less benefit so that we can have a simpler system.  Benefits are complicated for good reasons, and while there is certainly a case to offer more benefits like Child Benefit on a simple, unconditional basis, this kind of short-cut won’t meet people’s needs.

Arguing for universal benefits

In an article in Scotland on Sunday, Dani Garavelli considers the arguments for free school meals as a universalist measure.   I contributed some points, which the article refers to, but my contribution is immediately followed by comments from  John McTernan, a former political secretary to Tony Blair.  He said this:

“There’s not a single person in the country who believes everyone should get a housing allowance, no-one believes everyone should get tax credits – everyone believes those benefits should be means-tested because that’s the way you focus the most help to those in the most need.  The state pension is universal and that’s correct. But anti-poverty measures have always been targeted at those who need them most.”

Stuff and nonsense.   There are loads of people who do think that all our basic benefits should be universal:  you can find them at the Citizens Income Trust or the Basic Income Earth Network.  The state pension is contributory, not universal, and consequently it fails to provide sufficient support for more than two million pensioners who are entitled to Pension Credit.   Anti-poverty measures  take many forms, some of which are targeted on those on lowest income, and some – like early years intervention – which aren’t.

I don’t agree wholly with the arguments of people who argue for a Citizens Income, for just the same reason that I don’t agree with those who argue that all our benefits should be means-tested:  our benefits systems have to deal with multiple objectives and complex circumstances, and one size can’t be expected to fit every case.  Universal benefits have to be combined with lots of others, including some contributory benefits, some means testing and some discretionary provision,  so as to provide people with a stable income in an unpredictable environment.    But within that framework,  I’d certainly like to see both a universal housing allowance, and a universal benefit rather than  tax credits – so that’s one person, at least, who thinks so.