I’ve been working for a while on the idea of Universal Basic Income (UBI), and the closely related alternative of a Guaranteed Basic Income (GBI), in which the payment is phased out as income increases. I’ve now developed a very simple model to illustrate some of the crucial points. Here are three observations. Only Observation 2 requires the model, and the assumption that the distribution of income is broadly similar to that prevailing in Australia today.
Observation 1: Any UBI scheme can be replicated by a GBI with the same effective marginal tax rates, and vice versa
Observation 2: A GBI equal to 40 per cent of average income, with a phaseout rate of 40 per cent, would require additional transfer payments equal to between 8 and 10 per cent of national income.
Observation 3: A UBI equal to 40 per cent of average income, with no phaseout, would require additional transfer payments equal around 30 per cent of national income, but would have the same effective marginal tax rates as a GBI.
27 thoughts on “Three observations on guaranteed and universal basic income”
Interesting. A bit more complicated than a single-EMTR negative income tax approach, but also a bit more flexible.
(I’m a big fan of the NIT approach just to bulldoze the various high-EMTR poverty-trap points. As an aside to an aside, I was astounded to learn that the EMTR for Newstart payments can be well over 50%.)
Despite the obvious churning, a UBI would surely require far less administration.
What would be the advantage of a GBI?
I certainly hope that a UBI/GBI can be achieved – doing so will cut the nexus between “jobs” and the most egregious forms “development”.
Barnaby Joyce wouldn’t be able to say logging jobs are more important than the extinction of some furry critters. Our premier and her federal colleagues would need more than to spruik “10,000 jobs !” in the promotion of Adani’s interests.
Currently, any mention of jobs beats anything else, even if the jobs are ultimately a mirage.
A couple of points.
(1) Why not go further and implement a Job Guarantee beside the UBI?
(2) A UBI would require less administration. The churning is just churning of numbers (data bits) in computers these days. The cost would be negligible.
If you want a job and are capable of doing a job you could get one from the Government Job Guarantee at a minimum wage if you couldn’t find a private enterprise job. If you didn’t want a job you could take the (lower) UBI.
“additional transfer payments equal around 30 per cent of national income”
That is Denmark territory. They finance their public sector and welfare payments with a GST (equivalent) of 25%.
“Why not go further and implement a Job Guarantee beside the UBI?”
Watch this space
10% v 30% of national income seems (political) game over to me. The extra administration cost would have to be horrendous and, even if they were, it would likely be smarter to go GBI first.
I thought the goal of EMTR games at the bottom end was to get as many benefit recipients as possible as close to 100% EMTR as they can? 50% sounds very low for “the highest EMTR” to me. The shock horror headlines of the past were when it was so far over 100% that even the politicians couldn’t pretend things were fair. The other one is thresholds – earn another dollar and lose $100… you don’t have to be Kahneman to realise that’s a bad idea (look at the way Australian taxable incomes cluster just under the various levy cut-ins, for example. I have been burnt once by a ~$500 bonus that meant I paid a $1500 Medicare levy… that didn’t make anyone happy).
I did a quick search for “australia marginal tax benefit over 100%” and came up with a few policy papers and worked examples. Australia is a good test case here because we have so many means tested benefits that it’s easy to end up getting four or five different things that all reduce with income and the percentages add up to a very nasty number.
The cliche example is mothers who work part time, where the “hidden” cost of childcare often means they don’t benefit financially in the short term even if the EMTR is technically below 100% (maintaining some semblance of a career is a long-term benefit that might offset it, and for many women, staying sane is a short-term benefit without price).
What you call a GBI is in fact very close to Milton Friedman’s NIT. Julia and Ikonoclast have it exactly right – the churning with a UBI is purely an accounting artefact, not an actual economic difference from the GBI. The NET transfer is exactly the same percentage of GDP, as of course it must be if the maximum payment and the EMTRs are identical. And a UBI is far simpler to administer – you don’t need fortnightly reporting of income or an annual reconciliation with tax (with the prospect of considerable tax debts owed by lower-middle income people).
Most non-behavioural microsimulations come out with the sort of figures you have. But proponents tend to argue that the removal of high EMTRs on the population segments with the highest labour supply elasticities (part timers/casuals and mothers) and the highest tax avoidance elasticities (the self employed) mean that the real life figures are a bit better than this.
On Moz’ point, most empiric expenditure research finds the costs of paid work are higher than you’d expect – and not only for those needing paid childcare. In fact the FIXED cost of work (in both money and time) is IMO the reason for the well known finding that labour supply elasticities are much higher at the extensive margin (ie taking the first hour of work) than at the intensive margin (ie getting marginal increases in hours).
GBI, NIT and UBI are all essentially identical in terms of net transfer and EMTRs (NIT is usually described without a threshold, but it’s easy to put one in, or to eliminate it from the GBI/UBI proposals if you want).
On the churn point, you’re contradicting yourself. If you administer the UBI payment and the tax system separately then there is, indeed, churn, since high income earners receive the payment and pay high tax rates. If they can avoid taxes you have big problems.
To get to the point where you have only an accounting artefact, you need a tax reconciliation, so that high income earners never actually see the money, just two offsetting lines in their tax assessment. But that gives you exactly the same admin requirements as a GBI.
I’ll be talking about EMTR’s in my next post.
I prefer to think about these issues from the opposite direction. Instead of starting from basic principles, consider how UBI/GBI differs from our current system of income support. There are three key issues:
1) Means testing. This is a minor difference as both UBI or GBI will be effectively means tested via either a withdrawal rate or via the taxes used to finance it. Within this broad framework there are many options for concentrating the effective tax rates at different parts of the income distribution.
2) Family means testing. The current system is means tested based on the income and assets of a person plus their spouse (or plus parents for youth). This is a major reason for the greater cost of a UBI/GBI system.
3) Activity testing. Current payments require people to fall into particular categories for eligibility. Eg over pension age, or looking for work. For working-age people, these requirements are quite stringent. The policy direction over the last 30 years has been to steadily make these more restrictive. UBI/GBI would abolish these – though presumably we could have both a UBI and a higher activity-tested payment operating at the same time.
Most discussion of UBI/GBI focusses on issue 1). However, 2) and particularly 3) are much more important issues. If we are to think about gradual movements in the UBI direction, relaxing activity testing would be an obvious place to start.
With UBI, how do you get around the political problem of “you’re taxing low income people to give welfare to James Packer!”?
I support what Derrida derider posted. My work on Friedman’s NIT suggested that it would be of major benefit to the lowest decile. You don’t need to worry about the top two income deciles, they are very good at looking after themselves. My only added point is on letting Centerlink administer such a scheme. As we have seen with the LNP welfare grab in recent times, conservative government’s can use procedural rules to exclude needy recipients. One advantage of the NIT was the transfer to the taxation office. Canberra beaucratss live in their mind bubble of rules and regulations. They often care more for their paperwork than they do for needy applicants. Herein lies the largest implementation problem.
“If we are to think about gradual movements in the UBI direction, relaxing activity testing would be an obvious place to start.”
This is exactly my view.
Taxmen – at least at the senior operational level – have a different incentive structure than standard bureaucrats. At least in the UK, tax inspectors get a small percentage of tax recoveries. So their decisions in whom.to audit are those of rational economic actors, and they don’t waste time yin teachers and pensioners. Jobbing builders and dentists are better game.
Na, the “churning” is not a problem because all PAYG taxpayers don’t see the money either – the churning is reducing their pay packet and paying it back to them by the UBI. And by assumption there is no tax free threshold and a constant rate of tax, which makes the tax far harder to avoid (it doesn’t matter who you assign the income to, including some fictional corporate entity or a non-employed family member, it gets taxed at exactly the same rate – so the motivation for such assignment disappears).
The point is that the “cash in hand” position of everyone under a UBI and a GBI/NIT is exactly the same (as it must be with the same maximum payment and the same EMTRs) so the net amount as % of GDP must be the same. But because the tax and transfer sides are both radically simpler under a UBI, the UBI system is far simpler both to administer and for people to comply with.
That’s what I mean by those huge gross numbers apparently flowing through government being an accounting reality but an economic fiction – a fiction because the effects on labour supply and saving must be identical. Getting rid of the ‘churning’ is far less efficient than tolerating it.
I’m confused by this, I must admit. I agree that the payments and the EMTRs must be the same, but I don’t know what you have in mind when you talk about a UBI.
So, to clarify, are we talking about a scheme where
1. Everyone receives a UBI as a payment from government
2. Separately, everyone pays tax set at rates sufficient to finance the UBI
In particular, PAYG taxpayers would receive a pay packet with tax deducted at the EMTR for all income above the threshold and a separate bank deposit for the UBI.
Is that correct?
I am looking forward to seeing John Quiggin join forces with Bill Mitchell to promote the economic case for a Job Guarantee.
I rather like the simplicity of a UBI provision but politically, if it were unsaleable without some way of preventing the wealthy from getting it, I think there are reasonable woork arounds.
Currently, average full time ordinary earnings are about $1543 per week. Assuming a UBI of 40% we get a basic income of around $609 per week or about $31,650 p.a. That would become your threshold.
Currently, people on that income pay 19c at the margin until 37k, where you start paying 32.5%, which margin takes you up to 80k. It seems to me though that around $60k (ex-UBI) it’s not unreasonable to start (gently) clawing back income for those now getting that $609 extra each week. So i’d start a stepwise dollar by dollar increment on marginal tax rates to the 37.5% that those of us above $80k start paying. We in turn could start a stepwise move towards the bracket the above 180k folk get (45%) but get there at $140k. Above $140k you’d pay a margin of 75% increasing stepwise to 90% by the time you got to 180k. Everything else above would be 90%.
Naturally. The thresholds would be adjusted to take account of shifts in UBI.
It would be open to any taxpayer to waive getting UBI and take a non-UBI tax rate regime. This would only make sense for those in the top incone band, and for those so electing, I’d abolish the threshold and charge them 30% on the first 50k and 50% on everything else.
Corporate taxation would treat income packages in excess of 3 times average full time ordinary weekly earnings as less only 50% deductible, and those in excess of four times the average as not deductible at all. Corporate tax would be set at 30%. Companies not paying appropriate tax in Australia would be excluded from public tenders, and would not be given the benefit of industry assistance.
With the coming of UBI, the means-tested pension system could be abolished, though those with special needs (e.g. those needing soecific disability support) would get separate assustance based on need.
A jobs guarantee has been raised above and tbis also seems just. The state could ser aside money for targeted public works or community service and soecific prposals could be examined and evaluated for suitability on environmental, social inclusion, training and development and other grounds. Proponents would need to meet relevant WHS and RTO standards but would receive funds for wage costs at the relevant benchmarked industry average.
OK, lets work backwards from the constant EMTR (say, 50 cents in the dollar) and some form of a minimum income guarantee (ie no-one’s net income falls below, say, $20k a year). There are at least two ways you can deliver this:
– the classic non-taxable demogrant (UBI) of $20k to everyone with a flat (ie constant rate with no tax-free threshold) income tax rate of 50 cents. The rich get their $20k grant from the government like everyone else, but remember they are handing over half of ALL their (non-UBI) income to the government – not just the margin above a top tax bracket threshold; that’s how the UBI is funded. The poor get their $20k – a much bigger proportion of their income so it is progressive – but also have to hand over half of any wage.
– We have an NIT. If you have no taxable income you get an annual tax refund of $20k (if cashflow is a problem the system ensures you can effectively borrow against your refund, either privately or through a government fund). If you have $20k in earnings, you get a refund of (20k-(20k*0.5)) = $10k. If you have $40k in earnings, you owe no tax but you get no refund. If you have $80k in income you pay (80k*0.5-20k)=$20k in tax (taken out through the normal PAYG arrangements, though these become messy if your earnings through the year are variable – you could easily end up owing the taxman at the end of the year).
The point is that because total after-tax and after-benefit incomes of everyone are identical between these two schemes, the distributional and incentive effects are identical. But the first is far easier to administer and has far lower compliance costs, despite the fact that, unlike the NIT, you are ‘churning’ a great whack (on your figures well over 20%) of GDP through the scheme.
So if the distributional and incentive (ie economic) effects are identical, in what sense other than an accounting one is the bloated Budget of the first scheme a problem?
I’m not disputing the point about being identical in distributional and incentive effects, just the claim about administration The absence of any end-of-year discrepancy only works for people who have only PAYG income and no deductions.
More implortantly, the UBI seems to involve two systems (grants and taxes) rather than one. If you can hide from the tax system by not submitting a return, while collecting the UBI, you gain more than if you hide from an NIT system. So, it seems to me, you need more administrative/compliance effort to prevent this.
A GBI is different again, but I’ll think more about that.
“If you can hide from the tax system by not submitting a return,while collecting the UBI”
How is that possible? If you get the UBI, you are in the system. If you get any private income at all, you are in the system. If there’s just one tax rate, there’s no incentive to shift income to family members.
Tax is negative welfare and welfare is negative tax. I’m not the first to make this observation. The answer would be to have one system which integrates tax and welfare. When I say “system” here, I mean one computer/accounting/copliance system – an integrated database / mainframe system with all the usual back-ups and site redundancy.
When it comes to administration you keep two administrations, tax and welfare, which are fully walled from each other in administrative terms but each of which feeds their data into the main computer system. Why do this? Well frankly, the “giving” and the “taking” administrations involve different mindsets and require (oftentimes) different kinds of professionals and different kinds of people. They have very different briefs and the mindset by which each administration is run must be different. For example, tax will have a larger compliance section (and suitably tough-minded staff) and welfare will have a larger social work section (and suitably sympathetic staff).
How the one system (computer and financial/accounting) would work is a post all on its own. But to register for UBI and/or the Job Guarantee you are required to register in the tax system too. It’s all the one registration of course. So you do not have a welfare number and a TFN but simply a CRN (citizen registration number). In fact, turn 18 and you must be registered with a CRN. As an aside, we would need a Bill of Citizen Rights to put in safeguards for this system and for other reasons.
Beyond the above, spelling out reporting and compliance does get complex but it is do-able, less complex and more integrated than our current systems, administrative and computer. But as I said that’s a whole extra post to flesh it out.
I am a bit worried that we may be endorsing the introduction of a stealth Australia Card. Reading The Saturday Paper today about the leaders of this country jackbooting civil rights by this rediculous idea of a photo indentity database, gives me even greater concerns. It is one thing to help those in low deciles break out of the piverty cycle, that is always laudable. But it is something else when this is used to deny civil rights. Economics does not amtter if you are living under the sort of state control George Orwell described and warned about thirty years ago. We risk doing what our Prime Minister, aided and abbetted by all the Premiers, is now attempting to sneek past a sleeping public. The comments of some of the State Premiers was simply stupid. The WA Premier takes the price for bonehead comment of the week, but he is not alone among our braindead leadership group. Civil liberties are the social benefits of democracy. To throw them away is sheer madness.
I saw an, um, interesting talk by former Supreme Court judge Howard Nathan, drawing parallels between unemployment levels now and before WW1.
A Universal or guaranteed basic income is probably good policy given high youth unemployment in multiple countries around the world at present.
I wonder if it could be combined with action on climate change and sustainability. I know that sort of reeks of the Green Army stuff, but it would provide public benefit as well as facilitating skills development of the unemployed or under-employed.
The Conversation today has an article about the proposed/new HELP system, pointing out that for single-income families with children the new scam would have EMTR’s around 80% for incomes between $50k and $120k. There are also some ugly drops when benefits are abruptly withdrawn.
Wonderful, what a website it is!This web site presents valuable information to us, keep it up. http://astroaula.net/blog/2015/09/24/what-is-the-most-curious-obstacle_that-prevents-people-from-being-able-to-live-on-the-moon/