Haute Lumière

Commerce · V.07 · MMXXVI · daylight

La Bourse  /  Volume V  /  Nº V.07

Income, Floor, and Dignity

Volume V — Labour, Value, Flourishing

Nine movements, one floor.

---

THE PLATE

---

THE LETTER

There is a version of this argument you have already heard, probably more than once, and it is conducted entirely in adjectives. One side says a guaranteed income is humane and liberating. The other says it is unaffordable and corrosive. Both sides are certain, neither side has done the arithmetic in public, and the conversation has been stuck in that shape for about sixty years.

This chapter does the arithmetic in public.

We will take every serious income-floor trial there has been — Finland, Stockton, GiveDirectly's Kenyan study, the four American negative income tax experiments of the 1970s, Manitoba's Mincome, Alaska's permanent dividend — and say exactly what each of them was designed to measure, what it found, and what it cannot possibly tell you. Then we will cost three real instruments for a real country, with the tax schedule written out, the claw-back written out, and the effective marginal rate in the withdrawal band computed rather than skipped.

That last number is the point of the whole chapter. Every floor is won or lost in the band where it is being taken away. A guarantee is easy; a guarantee has never been the hard part. The hard part is what happens to the next pound a person earns while the guarantee is still tapering, and if you cannot state that rate to one decimal place, you are not yet in the conversation — you are in the adjectives.

You will find, when we get there, that the number is uncomfortable, that it is uncomfortable in every design including the ones this house prefers, and that it is already uncomfortable in the system running today. That last discovery is the one worth carrying out of this chapter.

A word about what this is for. Volume V asks what a person is worth and to whom. A floor is the most direct answer any economy ever gives to that question, because it is the amount below which the society has decided a person will not be allowed to fall regardless of what they produce. Whatever number that is — including zero — it is a statement about worth, and it is already being made.

— The Editors

---

DISCOVERY

What is already working

Start with the thing that is easiest to forget: income floors are not hypothetical. Several have been running for decades, and one of them has been paying every resident of an American state, unconditionally, every year since 1982.

Alaska's Permanent Fund Dividend. Every resident, no means test, no work test, no application beyond residency, funded from the returns on a sovereign fund built out of oil royalties. It is permanent, it is universal within the state, and it is general equilibrium — the whole economy has adjusted around it for four decades. Damon Jones and Ioana Marinescu examined its labour-market effect and found aggregate employment not distinguishable from zero, alongside a 1.8 percentage-point rise in part-time work. That is the single most valuable result in this literature, because it is the only one with all three properties every trial below is missing: permanent, universal, and economy-wide.

Manitoba's Mincome, 1974 to 1979. The town of Dauphin received a saturation guarantee — the whole town eligible, not a sample. Evelyn Forget recovered the records and compared Dauphin against matched communities. Hospitalisations fell 8.5 percent relative to the comparison group, with accidents, injuries and mental-health admissions carrying most of the fall, and high-school completion rose. The saturation design matters: it is the only one in which a guarantee could change what a community did, rather than what individuals in it did.

GiveDirectly's Kenyan study is the best-designed income experiment ever run, and the reason is that it randomised the thing everybody argues about. In 195 villages across Siaya and Bomet counties, adults receive roughly $0.75 PPP per day — about $22.81 a month. One arm receives it for two years. One arm receives it for twelve. One arm receives the entire two-year total, $548 PPP, in a single payment. One arm receives nothing. The same monthly amount, promised for two years and for twelve, against the same money handed over at once: three different instruments, in one study, with a control.

The short-run results, reported by Abhijit Banerjee, Michael Faye, Alan Krueger, Paul Niehaus and Tavneet Suri, are that labour supply did not fall, and that the lump-sum and long-horizon arms produced the most enterprise formation. The money did not stop people working. What it changed was what kind of work they could risk.

The general-equilibrium evidence, which exists exactly once. Dennis Egger, Johannes Haushofer, Edward Miguel, Paul Niehaus and Michael Walker studied the effect of GiveDirectly transfers on entire local economies — roughly $1,000 per recipient household, about $10,500 per treated village — and measured what happened to households who received nothing. They found a local fiscal multiplier of about 2.4, implying some $25,200 of local income created per village, with a price response of roughly 0.1 percent. The money circulated. It did not simply inflate.

Finland, 2017 to 2018. Two thousand people, drawn at random from those already receiving the basic unemployment allowance, were paid €560 a month, tax-free and unconditional, for 24 months — €13,440 each, €26,880,000 in total, against a control register of 173,000. In the second year the treated group worked 78 days on average against 72 for the control: six days more, 8.3 percent. Life satisfaction was 7.32 against 6.76 on a ten-point scale.

Stockton, California, 2019 to 2021. One hundred and twenty-five residents of neighbourhoods at or below the city's median household income of $46,033 received $500 a month for 24 months — $12,000 each — against a control of 200. Full-time employment among recipients moved from 28 to 40 percent while the control moved from 32 to 37: a difference in differences of seven percentage points. Recipients spent the money on food, utilities and goods; the share going to alcohol and tobacco was under one percent, which is the finding that has now been replicated so many times in so many countries that continuing to ask the question is itself the interesting behaviour.

Six studies, four continents, sixty years, and a pattern that is much narrower than either side of the argument usually claims: the money is not wasted, it is not spent on vice, it does not produce a withdrawal from work, and the largest measured effects are not economic at all — they are in health, in schooling, and in what people are able to attempt.

That is the positive core. Now we cost it, and we find out what it cannot do.

---

THE ARITHMETIC

What works, what does not, and where the band sits

First, what the trials can and cannot support. This is not a caveat paragraph. It is the load-bearing part, because a policy costed off a misread trial is a policy that fails in production.

They are small. Stockton treated 125 people. From the published proportions, the standard error on the treated arm's endpoint is 4.38 percentage points, on the control arm 3.41, and on the difference between them 5.55. The gap between the two endpoint levels is about 0.54 standard errors. The result is real, it points the right way, and it is not precise. Both sentences are true and they belong in the same paragraph.

They are time-limited, and a time-limited payment is a different instrument from a permanent one. This is not a theoretical worry; the experiments themselves measured it. Seattle–Denver — the largest of the four American negative income tax experiments, some 4,800 families — randomised the length of the guarantee as well as its level, running three-year and five-year arms. The five-year arm reduced hours by more than the three-year arm. People respond to the horizon they are offered, and a two-year trial therefore understates the response to a permanent programme. GiveDirectly's twelve-year arm exists precisely because of this finding.

They are not general equilibrium. Pay 125 people in a city of 300,000 and nothing moves in the labour market, the rent, or the price of anything. Pay everyone, and wages, rents and prices are all endogenous. Only Egger and colleagues have measured a whole local economy, and they measured a village economy with idle capacity and elastic local supply. That 2.4 multiplier does not transfer to a national programme facing a labour market that can bind, and anyone quoting it as though it does is quoting it wrongly.

They are often not testing what their headline says. Finland drew its sample exclusively from people already unemployed and already on benefit. It is an excellent test of removing conditions from an existing payment. It is not a test of a universal basic income, and its own evaluators say so.

Second, the labour-supply question, answered with the only large evidence there is. Four American negative income tax experiments ran between 1968 and 1980 — New Jersey and Pennsylvania with 1,357 families, rural Iowa and North Carolina, Gary, and Seattle–Denver. Philip Robins's comparison of all four found reductions in hours of roughly 7 percent for husbands, 21 percent for wives, and 13 percent for single mothers. At a forty-hour week over 48 weeks that is 2.80 hours a week and 134.4 hours a year for husbands, 8.40 and 403.2 for wives, 5.20 and 249.6 for single mothers.

Those are real reductions and this chapter will not talk them away. Two things about them are worth saying plainly. The first is that the largest responses are concentrated exactly where the competing use of the hour is an infant or a sick parent — which is a statement about what the labour market was pricing at zero, not a statement about idleness. The second is arithmetic. Apply the largest male response, 7 percent, to every one of the 33,000,000 people in UK employment working an average 32.0-hour week — 1,056,000,000 hours a week — and you lose 73,920,000 hours a week. Priced at £20.00 of gross value added an hour over 48 weeks, that is £71.0 billion a year, 2.8 percent of GDP.

That is the upper bound and it is deliberately unkind to the argument. It applies a 1970s time-limited experimental response to every worker in a country, and prices every lost hour at the average product rather than the marginal product of the people who would actually reduce. The true figure is smaller. It is not zero, and a chapter that pretended otherwise would deserve the scepticism it got.

Third, the cost of a real floor for a real country. Take the United Kingdom: 68,300,000 people, of whom 54,200,000 are adults and 14,100,000 children; nominal GDP £2,560,000,000,000; total managed expenditure £1,230,000,000,000; personal allowance £12,570; income tax at 20, 40 and 45 percent; employee National Insurance at 8 percent.

Set the floor at £100 a week per adult and £50 a week per child — £5,200 and £2,600 a year.

  adults     54,200,000  x  £5,200   =   £281.8 bn
  children   14,100,000  x  £2,600   =    £36.7 bn
  -------------------------------------------------
  GROSS                                  £318.5 bn      12.4% of GDP
                                                        25.9% of public spending

Now the offsets, each computed rather than waved at. Abolish the personal allowance: 31,000,000 basic-rate payers × £12,570 × 20 percent is £77.9 billion, and 6,000,000 higher-rate payers × £12,570 × 40 percent is £30.2 billion — £108.1 billion in total, which is deliberately the conservative reading. Abolish the National Insurance threshold: 28,000,000 employees × £12,570 × 8 percent is £28.2 billion. Replace Child Benefit at £12 billion and the standard allowance element of Universal Credit at £28 billion: £40 billion.

  gross                                  £318.5 bn
  less personal allowance                £108.1 bn
  less National Insurance threshold       £28.2 bn
  less benefits replaced                  £40.0 bn
  -------------------------------------------------
  NET FISCAL COST                        £142.2 bn       5.6% of GDP
                                                        11.6% of public spending

Housing support, disability support and carer's support are not folded in and not offset. A floor that swallows the additional-needs system is a cut wearing a new name, and this house will not cost one.

Note the ratio: gross to net is 2.24 to one. The gross number is the one quoted in public and it is the least informative number in this chapter, because most of it is money paid to people who fund it themselves in the same tax year.

Funding it: a penny on the basic rate raises about £6.9 billion today, and abolishing the allowance brings £465.1 billion of income newly into charge, so a penny on the widened base raises £11.6 billion. £142.2 billion needs 12.3 points — an entry rate of income tax of 32.3 percent, or 40.3 percent with National Insurance. Break-even, where what you receive equals what you pay, is £5,200 divided by 0.123: £42,227, which is 1.51 times the median individual income of £28,000. Most adults are net receivers. That is the political arithmetic, and it is computable.

Fourth — and this is the movement that does the real work — the withdrawal band.

Here is the honest negative, and it applies to every design including the ones this chapter recommends. Claw a £5,200 guarantee back at 50 pence in the pound and break-even arrives at £10,400. Inside that band, with National Insurance, the effective marginal rate is 58.0 percent — 13.0 points above the 45 percent top rate of income tax. Buy it down to the top rate and you must claw back at 37.0 percent, which pushes break-even out to £14,054 and raises the cost from £62.8 billion to £69.6 billion. The price of that comfort is £6.8 billion.

The general rule, and it does not negotiate:

  claw-back rate  x  width of the band  =  the guarantee

You cannot lower the rate without widening the band, and you cannot widen the band without paying for it. Here is the frontier for a £5,200 guarantee:

The payment leg is what it costs to pay everyone below break-even with the ordinary tax schedule left standing above it. It is comparable to the universal scheme's gross adult cost and to nothing else. Universal costs £281.8 billion gross for adults; clawed back at 50 percent it costs £62.8 billion. The claw-back saves £219.0 billion, or 77.7 percent — and every pound of that saving is bought with the withdrawal band. That is the trade, in one line.

Now the cut. Every objection above is routinely made against basic income proposals as though the alternative were a system without a withdrawal band. Compute the rate a British worker on Universal Credit faces today. Income tax and National Insurance come off first; Universal Credit then tapers 55 percent of what is left:

  1 − (1 − 0.20 − 0.08) × (1 − 0.55)  =  67.6%

With a student loan repaying at 9 percent it is 71.6 percent. And at the other end, between £100,000 and £125,140 the personal allowance withdraws at £1 for every £2 earned, making the marginal rate 1.5 × 40 = 60.0 percent, or 62.0 percent with National Insurance.

The highest marginal tax rate in Britain is 67.6 percent and it is not paid by the richest person in the country. It is paid by someone earning about fifteen thousand pounds. The top rate of income tax is 45 percent — 22.6 points lower.

So the withdrawal band is not an objection to an income floor. It is a description of the system already running, and the only question any of these proposals actually asks is where you put the band and who has to stand in it. A universal payment does not abolish the high marginal rate. It moves it off the poorest and spreads it across everybody, at 40.3 percent, in a form people can see. That is a real choice with a real price, and it is the first time in this argument that anyone has been offered one.

---

DREAM

What becomes ordinary

In the economy that has done this, the floor is boring. That is the whole achievement.

It arrives on the same date. It does not ask what you did last month. It does not stop because you took three shifts, and it does not need to be reclaimed because you took four. Nobody in the country has a sentence beginning if I earn another eighty pounds I lose—, because the arithmetic that produced that sentence has been retired. The rate on the next pound earned is the same rate at the bottom of the distribution as in the middle of it, and everybody knows what it is, because it is one number and it is printed on the payslip.

Hiring looks different, and it looks different in a way employers like. The conversation about wages stops carrying the whole weight of survival, so it can be about the work. A low-paid job that is genuinely worth doing gets done because the pay is fair for the hours, not because refusing it means destitution. Jobs that only existed because people had no ability to decline them are re-priced, and some of them disappear, and the things they were doing turn out to be either worth automating or worth paying properly for. Both of those are good outcomes and the second is more common than the first.

Care becomes visible in the accounts. The person who leaves the labour market for three years to raise a child or nurse a parent is not, during those years, producing nothing — they are producing the thing the labour market never learned to price. Under a floor they are producing it with an income. The measured hours of paid work fall, and the measured output of the society does not, and a generation of economists have to explain the gap, which is an improvement on a generation of carers absorbing it.

Enterprise formation rises, quietly and at the small end. The person who has wanted to try something for six years tries it, because the downside has a bottom now. Most of these attempts fail, as most always have, and the failures are survivable, which is the change. Risk-taking is not a personality trait. It is a function of what happens to you if you are wrong, and an economy that wants more of it has exactly one lever.

And the administrative machinery shrinks. The assessment, the conditionality, the sanction, the appeal, the tribunal, the reassessment — an apparatus whose purpose was to determine who deserved what — has much less to do, because the question it existed to answer is no longer being asked.

---

DESIGN

The three instruments, and how to choose between them

There are exactly three mechanisms. Everything else is a variant of one of them.

One — the universal payment. Everyone receives it; nobody is assessed; nothing is withdrawn. It is funded through the ordinary tax schedule, which means the rate on the first pound of income is high and flat. For our worked case: £142.2 billion net, an entry rate of 32.3 percent, 40.3 percent with National Insurance, and break-even at £42,227.

Its virtue is that it has no band at all — the marginal rate is the same everywhere, so nobody is ever punished for the next hour. Its cost is visibility: the tax rise is large, legible and lands in one line on everyone's payslip, which is a political property rather than an economic one, and it has sunk more proposals than any arithmetic ever has.

Two — the negative income tax. A guarantee withdrawn as income rises, so only those below break-even receive anything. For our case at 50 percent: £62.8 billion for the payment leg, break-even at £10,400, and a band rate of 58.0 percent.

Its virtue is that it costs less than a third as much in gross outlay. Its cost is the band, which is above the top rate of tax, lands on the lowest earners, and is the reason this design has failed politically every time it has been proposed by someone who did not disclose the rate.

Three — the job guarantee. A standing public offer of employment at a fixed wage to anyone who wants it. Cost it the same way: at the National Living Wage of £11.44 an hour over a 35-hour week, the annual wage is £20,821; employer National Insurance at 13.8 percent above £9,100 is £1,617; materials and supervision at 25 percent of the wage is £5,205; gross cost per post is £27,643. Against that, £8,251 of taxable pay returns £1,650 of income tax and £660 of employee National Insurance, the employer's £1,617 returns to the Treasury, and about £5,000 of benefit is no longer paid. Net cost per post: £18,716. One million posts — 69.4 percent of the UK's 1,440,000 ILO unemployed — costs £18.7 billion.

Its virtue is that it is cheap, it sets a wage floor that private employers must match, and it produces public output. Its cost is that it does nothing whatsoever for the carer, the student, the disabled person, or anyone else for whom the binding constraint is not the absence of a job offer.

Notice the per-person figures and then refuse to use them as a verdict. The floor costs £2,083 per person reached; the job guarantee costs £18,716 per person reached, about nine times more. That ratio is not an argument, because the two instruments are not buying the same thing. The job guarantee buys a wage floor, a price anchor and a public output. The floor buys an exit option, and the exit option is what the carer, the student and the person who cannot work at any wage are actually asking for.

The sequence this chapter recommends, and it is a sequence rather than a choice, because all three are affordable and none is sufficient:

1. Start with the band you already have. Before legislating anything new, lower the 67.6 percent. Every point off the Universal Credit taper is a point off the highest marginal rate in the country and costs a fraction of any new programme.
2. Add a job guarantee at £18.7 billion. It is the cheapest of the three, it is the easiest to administer, and it sets the wage floor that makes everything after it cheaper.
3. Then the floor, partial and universal, and fund it visibly. £142.2 billion net, 12.3 points on a widened base, break-even above the median. Take the argument about the tax rise rather than hiding it, because hiding it is what has lost every previous attempt.

---

DESTINY

How it holds when nobody is pushing

A floor sustains itself or it erodes, and which one happens is decided by three design choices made at the beginning.

Indexation, or it dies of arithmetic. A floor fixed in nominal terms falls by whatever inflation is, every year, without anyone deciding anything. That is how most of them have ended. Index it to median earnings rather than to prices: prices keep it constant against the past, earnings keep it constant against the society, and the second is what a floor is for.

Universality, or it dies of politics. A payment that goes to everyone has everyone as its constituency. A payment that goes only to the poor is defended only by the poor, which is the weakest coalition available. Alaska's dividend has survived four decades of fiscal pressure and multiple attempts to cut it because every voter in the state receives it. That is not sentiment; it is structure.

A ring-fenced source, or it dies of competition. A floor funded from general revenue competes every year with health and schools, and it loses, because its beneficiaries are diffuse and its rivals have hospitals. A floor funded from a named source — a sovereign fund, a land value charge, a carbon dividend — is defended by the source.

Now the honest part. Here is where this fails.

It fails when the floor is set too low to change any decision, which is the most common outcome: a payment large enough to be expensive and small enough to be irrelevant, purchased at full political cost for no behavioural return.

It fails when the additional-needs system is folded in to pay for it. A disabled person with £5,200 and no disability support is poorer than before, and the programme has bought its own fiercest opposition with its own money.

It fails when the withdrawal band is not disclosed. Every negative income tax proposal that has been defeated was defeated at the moment somebody else computed the band rate and announced it. Compute it yourself, publish it in the first paragraph, and you have removed the only weapon that has ever worked.

And it fails when supply is inelastic where the money lands. A floor in a housing market with fixed supply is a transfer to landlords with extra steps. Egger and colleagues found a price response of only 0.1 percent, but that was a village economy with slack, and rent in a supply-constrained city is the exact opposite case. A floor without a supply policy on housing is a subsidy to whoever owns the scarce thing — and that is the single most important sentence in this movement.

---

DELIGHT

What it feels like

The thing people report is not what the arguments predict. Nobody in these studies describes feeling rich. What they describe, over and over and in several languages, is a kind of quiet returning to the room.

The Finnish participants' life satisfaction moved from 6.76 to 7.32 — half a point on a ten-point scale, which sounds like almost nothing until you ask what it took to move it. It was not the €560. It was the end of proving. No form, no appointment, no explaining yourself to someone with the power to stop the money. The administrative burden of being poor is a real cost paid in real hours and real attention, and lifting it is felt immediately, before any money is spent.

There is a specific pleasure in a standing order that does not ask questions. It turns the first week of the month from an arithmetic problem into an ordinary week. People describe sleeping differently. They describe being better company. In Dauphin the measurable version of this was an 8.5 percent fall in hospitalisations, and behind that number is something simpler: a person who is not frightened makes fewer mistakes with machinery, and drinks less, and goes to the doctor earlier.

And there is the pleasure of being able to say no. Not dramatically — once, quietly, to one bad offer, without rehearsing it for a week first. An economy in which people can decline things is a more courteous economy, and courtesy between strangers turns out to be one of the things a floor buys.

---

OPERATIONALIZE THIS

At the level of finance — the Floor Facility

A national floor requires a legislature. An employer can build one this quarter, and the structure is a standard insurance liability that a treasurer will recognise on sight.

The instrument: a pay-floor guarantee, reserve-funded, triggered on hours volatility.

The problem it solves is not low pay. It is variance in pay. An hourly workforce on a fluctuating rota faces monthly income that moves by ten to twenty percent with no warning, and that variance — not the level — is what drives the payday loan, the missed rent and the resignation. Guarantee the floor and you have bought the most valuable thing the employee does not currently have, usually for less than the turnover it prevents.

The structure. The employer guarantees each hourly employee a minimum net monthly pay equal to a fixed share of their own trailing three-month median. If scheduled hours fall short, the difference is topped up. It is not a loan, it is not advanced against future hours, and it is never recovered.

The pricing, worked. A workforce of 2,000 with median net monthly pay of £1,850 and a monthly pay standard deviation of 12.0 percent of that median — £222.00. Set the floor at 90.0 percent of trailing median: £1,665.00 a month. The expected top-up is the standard shortfall integral,

  E[(a − X)⁺]  =  σ · ( z·Φ(z) + φ(z) ),   z = (a − μ) / σ

with z = −0.8333, Φ(z) = 0.2023 and φ(z) = 0.2819, giving £25.15 per person per month — £301.84 a year, and £603,688 a year across 2,000 people.

The balance-sheet treatment. It is a constructive obligation with a reliably estimable expected outflow, so it is provisioned, not disclosed as a contingency. Carry a reserve at twelve months of expected top-up plus a buffer sized to the historic worst month, and recognise the expense as it accrues against payroll rather than as an exceptional item. Your auditors will ask whether it is a defined benefit; it is not, because the obligation is capped at the shortfall and terminates with the employment.

The counterparty. Yourself, for the first year — this is a self-insured retention, not a product, and it needs no broker. Once you have twelve months of loss data, a stop-loss wrapper above the worst month becomes cheap, and at that point it can be syndicated across a sector.

The number that decides it. Annual turnover is 60.0 percent — 1,200 leavers — at a loaded replacement cost of £3,000 each: £3,600,000 a year. So:

        annual cost of the guarantee            £603,688
   ------------------------------------  =  --------------  =  10.06 points
      headcount × replacement cost          2,000 × £3,000

If the guarantee reduces annual turnover by more than 10.06 percentage points, it is free. At a 5-point fall it saves £300,000 and costs £303,688 net. At 10 points it saves £600,000 and is within £3,688 of break-even. At 15 points it saves £900,000 and returns £296,312.

Put that single inequality on the front page of the paper and the conversation changes register entirely: you are no longer proposing a benefit, you are proposing a retention instrument with a stated break-even.

The first ninety days.

Run it on one site against a matched site. You will have a number in two quarters, and it will be yours rather than borrowed from a trial in California with 125 people in it.

---

APPRECIATIVE QUESTIONS

Twelve, for a room

Discovery — what is already working

1. Where in this organisation does someone already have a floor — a retainer, a guaranteed minimum, a salaried role in a mostly hourly team — and what do we observe about how they work that we would want more of?
2. Think of a time you or a colleague took a real risk at work that paid off. What was underneath you that made it possible to take?
3. Which of our people have the steadiest income, and what does that steadiness let them do for us that we have never counted?

Dream — what becomes possible

1. If nobody here had to take a shift out of fear, what would we have to change about how we ask for one — and what would improve?
2. Imagine our pay in three years contains one guarantee we are proud of. What is it, and what does it say about us to someone deciding whether to join?
3. If every person here could decline one thing a month without consequence, what would we stop doing that nobody actually wanted done?

Design — what we build

1. What is the smallest guarantee we could offer that would change a real decision for a real person here — and what would it cost us to find out?
2. Where in our own pay and benefit rules is the rate on someone's next hour higher than we intend? Who would know?
3. What would we need to measure for two quarters, starting this month, to be able to price a floor properly rather than argue about one?

Destiny — how it holds

1. What would keep this guarantee in place through a bad year, and who would defend it if the person who built it had left?
2. What is the first sign we would see that the floor was quietly being eroded by inflation or by drift, and who would notice it first?
3. If this worked here, who else in our sector would want it — and what would we be willing to share to make that easy?

---

WORKS CITED

Jones, D. and Marinescu, I. (2022). "The Labor Market Impacts of Universal and Permanent Cash Transfers: Evidence from the Alaska Permanent Fund." American Economic Journal: Economic Policy, 14(2), 315–340.

Forget, E. L. (2011). "The Town with No Poverty: The Health Effects of a Canadian Guaranteed Annual Income Field Experiment." Canadian Public Policy, 37(3), 283–305.

Banerjee, A., Faye, M., Krueger, A., Niehaus, P. and Suri, T. (2023). "Universal Basic Income: Short-Term Results from a Long-Term Experiment in Kenya." NBER Working Paper.

Egger, D., Haushofer, J., Miguel, E., Niehaus, P. and Walker, M. (2022). "General Equilibrium Effects of Cash Transfers: Experimental Evidence from Kenya." Econometrica, 90(6), 2603–2643.

Haushofer, J. and Shapiro, J. (2016). "The Short-Term Impact of Unconditional Cash Transfers to the Poor: Experimental Evidence from Kenya." Quarterly Journal of Economics, 131(4), 1973–2042.

Kangas, O., Jauhiainen, S., Simanainen, M. and Ylikännö, M. (eds) (2020). Experimenting with Unconditional Basic Income: Lessons from the Finnish BI Experiment 2017–2018. Ministry of Social Affairs and Health, Helsinki.

West, S., Castro Baker, A., Samra, S. and Coltrera, E. (2021). Preliminary Analysis: SEED's First Year. Stockton Economic Empowerment Demonstration.

Robins, P. K. (1985). "A Comparison of the Labor Supply Findings from the Four Negative Income Tax Experiments." Journal of Human Resources, 20(4), 567–582.

Robins, P. K. and West, R. W. (1980). "Program Participation and Labor-Supply Response." Journal of Human Resources, 15(4), 499–523.

Cain, G. G. and Wissoker, D. A. (1990). "A Reanalysis of Marital Stability in the Seattle–Denver Income-Maintenance Experiment." American Journal of Sociology, 95(5), 1235–1269.

Friedman, M. (1962). Capitalism and Freedom. University of Chicago Press.

Tobin, J., Pechman, J. A. and Mieszkowski, P. M. (1967). "Is a Negative Income Tax Practical?" Yale Law Journal, 77(1), 1–27.

Van Parijs, P. and Vanderborght, Y. (2017). Basic Income: A Radical Proposal for a Free Society and a Sane Economy. Harvard University Press.

Standing, G. (2017). Basic Income: And How We Can Make It Happen. Pelican.

Torry, M. (2018). Why We Need a Citizen's Basic Income. Policy Press.

Tcherneva, P. R. (2020). The Case for a Job Guarantee. Polity Press.

Mitchell, W. and Muysken, J. (2008). Full Employment Abandoned: Shifting Sands and Policy Failures. Edward Elgar.

Costello, E. J., Compton, S. N., Keeler, G. and Angold, A. (2003). "Relationships Between Poverty and Psychopathology: A Natural Experiment." JAMA, 290(15), 2023–2029.

Ostrom, E. (1990). Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge University Press.

Cooperrider, D. L. and Whitney, D. (2005). Appreciative Inquiry: A Positive Revolution in Change. Berrett-Koehler.

Note on figures. UK population, GDP, public expenditure and tax parameters are 2024–25 published values; the floor, the offsets, the funding rate, the break-even income, every effective marginal rate, the negative income tax frontier, the job guarantee cost and the Floor Facility pricing are computed in lib/verify/V_07.py and reproducible there. The adult income distribution used for the negative income tax costing is a two-parameter model calibrated to published median and mean individual income, and is marked as such in the module; a policy costing requires a microsimulation and this is not one.