Interest in a universal basic income is growing at both ends of the political spectrum, albeit for very different reasons. It is also attracting support, and some financial backing, from tech entrepreneurs. The flexibility of the concept explains why it has created such broad interest. It can be regressive or progressive, depending on the level, who pays and what it means for other welfare benefits. But that may also make it politically impossible to implement for now. A radical version – set either at a high level or which replaces a large part of a welfare system – could appeal to left or right, but not both. As a policy, it is also risky, with clear and tangible costs, but mostly theoretical benefits, explaining why mainstream political parties are reluctant to take the idea seriously. Experiments with a basic income are important, both to inform design choices and to provide hard evidence of the benefits. The political saleability of a basic income may turn on this. It may also depend on whether the future of work is as insecure and socially disruptive as many people now fear it will be.
Interest in a universal basic income is spreading across the political spectrum. It comes in many forms and goes by several names (such as a basic income guarantee or BIG), but the concept is essentially a simple one: a universal commitment by the state to pay a basic income to all citizens regardless of their circumstances. Despite drawing support from the left and the right, both in Europe and the US, the idea has not quite made it into the political mainstream, no doubt because of its radicalism and its cost. But at a time when mainstream political parties are threatened by new, populist groups and when policy orthodoxy is increasingly challenged, could this be the moment when policymakers start to take the BIG idea seriously?
An old idea back in fashion
Basic income as a concept is not new. As has been documented elsewhere, it has over time been championed by a wide group of intellectuals, including Bertrand Russell, Friedrich von Hayek, Paul Samuelson, JK Galbraith and Milton Friedman. It was taken seriously by the Nixon administration in the US and was the subject of policy experiments in Manitoba in Canada in the 1970s. More recently, versions have been tried in India and Brazil, while public authorities in several European countries are toying with the concept, including in the Netherlands and Italy. At the start of 2017, Finland will begin a new experiment with a basic income. In the US, its merits are being discussed by conservatives and progressives alike, although not necessarily in the same room. Perhaps most interestingly, it has also become fashionable among Silicon Valley tech entrepreneurs, with Y Combinator even financing a small-scale basic income experiment in Oakland, California.
For progressives on the left it is a means to reduce poverty, tackle inequality in opportunities and outcomes, and create economic security at a time when more low and middle income households feel their financial situations are increasingly precarious. By not penalising those that choose not to work, it promotes better, more balanced lifestyles, supporting people moving in and out of employment as their circumstances change, allowing them to spend more time in education or caring for dependents where necessary. By ensuring men and women are paid on equal terms, it increases the financial independence of women and therefore improves gender equality. More fundamentally, it is a way to spread economic power that has become increasingly concentrated in a small wealthy elite.
For supporters on the right, the motives are different. Its simplicity is a way to reduce government bureaucracy. The absence of means-testing alleviates the poverty trap and many of the disincentives to work hard and earn more, or other distortions to choices that are a feature of complicated, modern-day welfare systems. For some it is also simply a means to reduce the size of the state.
For tech entrepreneurs, the motives are less ideological. It is a means to address the social consequences of increasing automation, which is making it harder for many people to find and hold onto good jobs. In a world where the future of work may be about ‘gigging’, or taking multiple, insecure, low-paid jobs, it is a means to provide social protection. This is not necessarily altruistic. Backing a basic income could be the price to pay for an economy where innovation really is disruptive, not only to sleepy incumbent businesses, but for people at work.
With such starkly differing objectives, it is odd that an apparently simple policy idea can attract such broad interest. This raises the question of whether people are talking about the same thing or at cross-purposes? And that depends on how three questions are answered about the way a basic income guarantee works.
Three BIG design questions
What form? In its purest form, a basic income guarantee must be truly universal and paid to all citizens regardless of their income level or personal circumstances. The problem is cost, leading some to propose tapers or cut-off points for higher earners, which in turn has implications for incentives where these begins to bite. Linked to this is the broader question of whether payments should be contingent, perhaps linked to age or personal circumstances that may indicate a greater need for state support.
This is related to another issue – and an apparent dividing line between left and right – of whether a basic income guarantee is a substitute or complement to the existing welfare state. Here it is helpful to separate questions that are really matters of political preference from those about the effective functioning of a basic income system. In its purest form, a basic income should replace all means-tested income support, but not public services made available based on a need other than for income. The latter would, for example, include different forms of disability support, even where this is in the form of a monetary payment.
Who pays? The high potential cost of a universal basic income leads to the obvious question of who will finance it? There is no ‘correct’ answer to this question, only preferences. But it is critical to determining the political feasibility of a basic income, whether it is ultimately regressive or progressive in its impact, its wider consequences on incentives and its social impact. Those on the left and the right will have different preferences for who pays. For tech entrepreneurs, if a basic income guarantee really is a licence for disruptive innovation, then the obvious answer is that those who benefit from the license make a commensurate contribution towards it.
How high? This question cannot be separated from the others. The higher the basic income, the more acceptable it becomes to replace all existing forms of means-tested income support. But it clearly also makes the question of who pays more sensitive. In short, the higher the basic income, the more radical the political proposition, which is why most fully-costed proposals for a basic income are quite modest regarding this central design feature. The more politically savvy proponents, however, recognise that a basic income can start small and grow, once it has become established.
Then there are the unknowns. The most important is the impact on incentives, as this affects the affordability of any scheme and the efficacy of its design features. For some, an attraction of a basic income is that it stops people being forced into paid employment, allowing them to make more careful judgements about how to balance their time between home and work. For others, though, it is about sharpening incentives to work, as scrapping the means-testing of benefits allows those on low incomes to keep a higher proportion of their extra earnings. For economists, these are income and substitution effects that work in opposite directions, and which dominates is an empirical question.
The wider social consequences of a basic income are also hard to predict. Much has been made about its impact on the ability of people to move in and out of education during their lifetime. This raises quantitative questions about the strength of this effect and qualitative questions about the type of education people might seek. For example, might it be less vocational and more enriching? It has also been suggested that a basic income is good for health, as it removes the need for people to work when they are stressed or suffering other forms of work-related illness. The significance of this effect is almost unknowable in advance as it likely depends on the detail of any scheme and how it interacts with existing systems of welfare and healthcare, which vary considerably from one country to another.
There are also questions about the impact of a basic income in aging societies, which have not yet been given much attention in the public debate. The most important is the effect on the incentive to save for retirement. A basic income guarantee effectively creates a contingent liability for the state, which must either be funded, or if unfunded, becomes vulnerable to a shrinking tax base in an aging society. If a basic income guarantee – and the promise of a basic income in later years – reduces the incentive of individuals to save for retirement, this exacerbates the problem, as the lower net worth of retirees narrows the funding base for the basic income in later years.
These are issues that can potentially be explored through policy experiments, of the sort being conducted by the government in Finland and by Y Combinator in California. But there are limitations to these exercises. Well-designed experiments may shed some light on well-defined issues, such as how scheme design choices affect the incentive to work. But they are unlikely to provide universally applicable insights into the wider social consequences of a basic income, as these will be highly context-dependent. And they would need to be run for several years with credible promises to continue paying a basic income far into the future if they are to provide robust insights into the impact on behaviour over the lifecycle in aging societies.
Some basic maths
In the UK, the annual income of someone working 35 hours a week while earning the £7.20 minimum wage is around £13,000. There are 52 million over-18s in the UK, implying an annual cost of around £680 billion, which is approximately one third of UK GDP. And that is before any payments for children are factored in, which most proponents of a basic income would support.
Alternatively, the estimated £217 billion currently spent on benefits, state pensions and tax credits could finance a basic annual income for over 18s of around £4,150 per person or a monthly income of just £345. This is significantly below the €560 per month that will be paid to 2,000 unemployed Finns as that country experiments with a basic income in 2017.
The first scheme above would be unaffordable, even though the income is below the ‘living wage’ as assessed by the Resolution Foundation of £9.75 in London and £8.45 elsewhere in the UK. The second would be deeply regressive, as it implies a substantial transfer of income from those that currently receive benefits to those that do not.
Similar calculations can be done for other advanced economies. They all lead to the same conclusions – a basic income guarantee is hugely expensive, unless it is also deeply regressive.
For this reason, many fully-costed basic income proposals are incremental in their design – hybrids that borrow from existing welfare systems, while typically financing modest and not-quite-universal basic incomes.
A universal basic income can mean many things. It can be regressive or progressive, depending on the level, who pays and what it means for other benefits. It can shrink the role of the state or expand it, although it would most likely simplify it by reducing means-testing and combining different forms of welfare into a single payment.
The flexibility of the concept is why it has created interest from across the political spectrum. But that may also mean it is politically impossible to implement, at least not yet. A radical version – set either at a high level or which replaces a very large part of an existing welfare system – could appeal to the left or right, but is unlikely to appeal to both. The objectives set for a basic income are just too far apart for this to be possible. A more modest, hybrid system – set at a low level, but perhaps not quite universal and replacing only some welfare payments – might be more politically acceptable, but would have a much blunter impact and lose many of the perceived benefits.
Experiments of the sort being run in Finland and California may tell us much about the impact on incentives and therefore about policy design. The evidence could prove essential for creating the political case for a basic income. For many, the argument for universality is not yet convincing enough to tear up welfare systems that have existed for generations, evolving only gradually, and which deliberately target support to where it is most needed. Basic income schemes involve very clear, tangible costs, but mostly theoretical benefits, which mean they are too risky to be taken seriously by mainstream political parties. This is one reason why an attempt in Switzerland in June 2016 to win approval for a basic income in a referendum ultimately failed, decisively, with a three-to-one vote against. Voters were sceptical about the case for universality and the lack of clarity about who would pay, while none of the parties in the Swiss parliament backed it.
The most innovative and least ideological arguments for a basic income are being made by tech entrepreneurs. The political saleability of a basic income may eventually turn on what the future of work is like. For now, no one really knows whether automation will displace jobs, or produce a more gradual evolution in the types of work that people do, of the sort that has been seen in the past when there have been disruptive changes to technology. The ‘future’ of work may need to become a reality before a convincing political case – one which appeals across the political mainstream – is made.
The views expressed in this note can be attributed to the named author(s) only.