Universal Basic Income

Contents

• 1. What Is UBI?
• 2. A Brief History
• 3. The Case For
• 4. The Case Against
• 5. Show Me the Math
• 6. The Libertarian Perspective
• 7. Current Research & Pilot Programs
• 8. Variants & Alternatives
• 9. UBI and the AI Transition
• 10. Further Reading
• Part II: Proof of Benefit — An Alternative Framework

1. What Is UBI?

Universal Basic Income is a government program in which every adult citizen receives a regular, unconditional cash payment—regardless of employment status, wealth, or any other qualifying criteria. The three defining features:

• Universal: Everyone gets it. No means testing, no income thresholds, no application process.
• Basic: Enough to meet minimal living needs—not enough to live lavishly.
• Income: Cash, not vouchers or in-kind benefits. Recipients decide how to spend it.

Most proposals envision something in the range of $500–$1,500 per month per adult. That's enough to prevent absolute destitution but not enough to replace a median salary.

2. A Brief History

The idea is far older than most people assume:

• 1797: Thomas Paine's Agrarian Justice proposed a one-time payment to every citizen at age 21, funded by a ground-rent tax—perhaps the earliest UBI-adjacent proposal.
• 1962: Milton Friedman proposed the Negative Income Tax (NIT) in Capitalism and Freedom, a mechanism that functions identically to UBI at the mathematical level.
• 1967: Martin Luther King Jr. advocated a guaranteed income in Where Do We Go from Here: Chaos or Community?
• 1969–1980: The U.S. ran four large-scale NIT experiments (New Jersey, Iowa/North Carolina, Gary, Seattle/Denver). Results were mixed; modest labor reductions but significant welfare gains.
• 1982: Alaska's Permanent Fund Dividend began paying every resident an annual share of oil revenues—averaging ~$1,600/year. It remains the longest-running quasi-UBI in the developed world.
• 2017: Finland launched its UBI experiment. Andrew Yang's 2020 presidential campaign built around "The Freedom Dividend" brought UBI into mainstream American politics.
• 2020–present: A wave of guaranteed income pilots across the U.S. and developing world, accelerated by COVID-era stimulus checks and growing AI anxiety.

3. The Case For

Economic Security

• Eliminates extreme poverty with a single, simple mechanism. No one falls through the cracks because they didn't fill out the right form.
• Smooths economic shocks. Job loss, medical emergencies, and recessions become survivable without bureaucratic delay.
• Reduces the "welfare cliff"—the perverse incentive where earning $1 more can cost you $2 in lost benefits. UBI has no phase-out trap.

Freedom & Dignity

• Respects individual autonomy. Cash lets people make their own trade-offs instead of being told what to eat, where to live, or how to spend.
• Reduces the power imbalance between employers and employees. Workers can say "no" to exploitative conditions.
• Supports unpaid work: caregiving, volunteering, art, open-source development, and community building—all currently invisible to GDP.

Administrative Simplicity

• Replaces dozens of overlapping programs with a single transfer. No caseworkers, no eligibility reviews, no fraud investigations.
• Dramatically cheaper to administer than the current welfare bureaucracy, which spends significant resources verifying who deserves help.

AI Transition Insurance

• Provides a buffer during the period when AI displaces jobs faster than new ones emerge.
• Enables retraining. People can afford to learn new skills without simultaneously scrambling for rent money.

4. The Case Against

Cost

• Staggeringly expensive at scale. $1,000/month to every U.S. adult (~260 million people) = ~$3.1 trillion per year. That's roughly the entire federal tax revenue.
• Funding remains the hardest question. Most proposals rely on some combination of new taxes (VAT, wealth tax, carbon tax), eliminating existing programs, and deficit spending. None are painless.

Work Incentives

• Modest but real labor supply reductions have appeared in nearly every major study. The AEI's 2026 meta-analysis of 122 programs found a net employment effect of –3.2% in larger pilots (500+ participants).
• The question is magnitude, not direction. Most economists agree some people will work less. The debate is whether that's a crisis or a rounding error.

Inflation

• Injecting new purchasing power without increasing production can drive up prices, especially for inelastic goods like housing and healthcare.
• Landlords could capture gains. In tight housing markets, rents may rise to absorb the UBI payment, effectively transferring the subsidy to property owners.

Political Durability

• Once enacted, impossible to repeal—creating a permanent fiscal commitment regardless of economic conditions.
• Risk of "UBI + everything else." The libertarian selling point is that UBI replaces the welfare state. The political reality is that existing programs have constituencies that will fight to keep them. Result: UBI layered on top, not instead of.

Targeted Needs

• A flat payment ignores unequal needs. A healthy 25-year-old and a disabled single parent with medical bills receive the same amount. If UBI replaces targeted programs, the most vulnerable may end up worse off.

5. Show Me the Math

The most common question about UBI: can we actually pay for it? Here are two scenarios for the United States, modeled in today's dollars, benchmarked against total corporate profits.

The Baseline Numbers

Scenario A: "Welfare Floor" — Poverty-Level UBI

Every adult receives enough to reach the federal poverty line: $15,650/year ($1,304/month).

Scenario B: "Salary Replacement" — Median Income UBI

Every adult receives the median individual income: $53,000/year ($4,417/month). This isn't what most UBI advocates propose, but it models the "what if AI replaces most jobs" endgame.

Scenario C: Negative Income Tax (Same Poverty Floor, Lower Gross Cost)

A Negative Income Tax delivers the same poverty-floor guarantee but only sends money to those who need it. Payments phase out as income rises. Here's the same $15,650/year floor with a 50% phase-out rate (Friedman's preferred structure):

Examples: Someone earning $0 gets the full $15,650. Someone earning $20,000 gets $5,650 ($15,650 − 50% × $20K). Someone earning $31,300+ gets nothing.

NIT vs. UBI: Side by Side

What This Tells Us

• The poverty-level UBI is expensive but not impossible. Net new cost of ~$2.6T is roughly 66% of pre-tax corporate profits. For context, current U.S. corporate tax revenue is about $420B (an effective rate around 11%). You'd need to raise the effective corporate tax rate to roughly 77% to fund it from corporate profits alone—obviously unrealistic as a single funding source.
• Salary-replacement UBI from corporate profits alone is arithmetically impossible. It would require more than 3x total corporate profits. The entire U.S. wage bill is $12T, and you'd be trying to replicate that from a $3.9T profit pool.
• This is why most serious proposals blend funding sources: a VAT (like Andrew Yang's 10% proposal, estimated at ~$800B), carbon taxes, financial transaction taxes, partial welfare offsets, plus some corporate tax increase—not any single mechanism.
• The AI optimist's counter-argument: If AI dramatically increases corporate output and profits while reducing the need for human labor, the denominator (corporate profits) could grow substantially. A world where AI triples corporate profits to $12T makes the poverty-level UBI much more feasible at a ~21% effective tax rate.

6. The Libertarian Perspective

The libertarian relationship with UBI is more nuanced than outsiders might expect. It's neither blanket endorsement nor reflexive opposition—it's a conditional, pragmatic argument about second-best options.

The "If We Must" Argument

The strongest libertarian case for UBI, articulated by Milton Friedman, Friedrich Hayek, and Charles Murray, runs roughly:

• The welfare state exists and isn't going away.
• The current system is paternalistic, bureaucratic, and wasteful.
• If government is going to redistribute income anyway, a single unconditional cash transfer is the least bad way to do it.
• Cash respects individual choice. Food stamps tell you what to buy; cash lets you decide.

Murray's 2006 book In Our Hands proposed $10,000/year to every American adult, coupled with the complete elimination of all other welfare, including Social Security, Medicare, Medicaid, and housing subsidies.

The Deal-Breaker

This is the crux of libertarian skepticism, as articulated repeatedly in Reason:

• UBI only works as a replacement, never an addition. The moment it's layered on top of existing programs, you get the worst of both worlds: massive new spending plus the old bureaucracy.
• There is no credible political mechanism to ensure existing programs are permanently eliminated. Every welfare program has a constituency. Congress can always re-create them.
• This isn't hypothetical. Spain's 2020 "UBI" program was means-tested and added on top of existing welfare—violating every principle that made UBI appealing to libertarians in the first place.

As Reason's Veronique de Rugy wrote: "Without a strong guarantee that all anti-poverty measures would be terminated—and that they will not be brought back to life later—UBI is a terrible idea."

The Deeper Objection

Some libertarians go further, arguing that UBI is objectionable on principle regardless of implementation:

• Taxation for redistribution is coercive, full stop. Efficiency gains don't justify the underlying compulsion.
• UBI normalizes the idea that the state is the default provider. Even a "simple" cash transfer embeds the assumption that government should guarantee outcomes.
• The best anti-poverty program is economic growth, deregulation, and free markets—not redistribution in any form.

Where Libertarians Converge

Both pro- and anti-UBI libertarians generally agree on two things:

1. The current welfare system is broken, paternalistic, and creates dependency.
2. If cash transfers happen, they should be unconditional—no means testing, no behavioral requirements, no bureaucratic gatekeeping.

7. Current Research & Pilot Programs

As of 2026, no country has implemented a full nationwide UBI. But dozens of pilots have generated real data:

The Meta-Picture

A 2026 meta-analysis by the American Enterprise Institute reviewed 122 UBI-related programs and found:

• Among 30 randomized pilots with published results, the net employment effect was +0.8%.
• In the four largest programs (500+ participants), the net employment effect was –3.2%.
• Well-being, health, and financial stability improved consistently across nearly all studies.
• No study found significant increases in substance abuse or reckless spending.

The honest summary: UBI makes people's lives measurably better. It may modestly reduce work. Whether that trade-off is acceptable depends on your values and your theory of what work is for.

8. Variants & Alternatives

UBI isn't the only idea on the table. Several alternatives share its goals while addressing specific criticisms:

9. UBI and the AI Transition

The resurgence of UBI interest is inseparable from AI. Here's why:

The Speed Problem

Previous waves of automation (agriculture, manufacturing) displaced workers over decades, giving labor markets time to adjust. AI may compress that timeline dramatically. When an AI can do legal research, medical diagnosis, software engineering, and customer service simultaneously, the "just retrain" argument assumes retraining leads somewhere.

The Breadth Problem

Past automation targeted blue-collar, routine tasks. AI threatens white-collar, cognitive work—the very jobs displaced factory workers were told to retrain for. The usual escape hatch (move up the skill ladder) may not exist if AI occupies the upper rungs too.

The Optimist's Case

If AI drives the cost of goods and services toward zero—as the most optimistic projections suggest—then even a modest UBI could provide a high standard of living. In a world of radical abundance, the question shifts from "can we afford UBI?" to "why would we need anything else?"

The Realist's Concern

The dangerous window is the transition period: AI is productive enough to displace wages but prices haven't fallen yet. During this gap, a lot of people could be in serious trouble. UBI (or something like it) may be the simplest bridge across that gap, regardless of whether you think the destination is utopia or just a different economy.

10. Further Reading

• Reason: Basic Income Archives — The libertarian perspective, extensively documented
• OpenResearch Findings — Full results from the Altman-funded UBI study
• GiveDirectly: Kenya UBI Results — Early findings from the world's largest UBI experiment
• Britannica ProCon: UBI — Balanced pro/con overview with sourced arguments
• Charles Murray, In Our Hands: A Plan to Replace the Welfare State (2006)
• Sam Altman, "Moore's Law for Everything" (2021)
• Philippe Van Parijs & Yannick Vanderborght, Basic Income: A Radical Proposal for a Free Society and a Sane Economy (2017)

The Core Idea

Universal Basic Income answers the right question—how do people survive when AI displaces work?—but it may answer it too bluntly. UBI says: everyone gets the same amount, no questions asked.

Proof of Benefit asks a different question: what if we could measure and reward the value people create, even when that value doesn't fit neatly into a traditional job?

The insight is that "work" has always been broader than "employment." Parenting, mentoring, caregiving, community organizing, open-source software, local journalism, civic participation—all of these generate enormous value. The economy just has no mechanism to compensate them. In a world where AI handles more of the tasks employers currently pay for, these human contributions become more important, not less.

The Blockchain Analogy

The name isn't accidental. In cryptocurrency:

• Proof of Work rewards computational effort—whoever burns the most electricity wins.
• Proof of Stake rewards commitment—whoever has the most invested wins.

Proof of Benefit would reward demonstrable positive impact—whoever creates the most value for others receives the most support.

This doesn't require blockchain technology (though it could use it). The analogy is about the mechanism design: creating a system where the incentives align contribution with compensation, even when there's no employer writing a check.

How It Differs from UBI

Precedents & Precursors

This isn't an idea from nowhere. Several existing models explore the same territory:

The Hard Part: Measurement

The obvious objection: who decides what counts as "benefit"? This is the central challenge, and it's worth being honest about how difficult it is.

What's Easy to Measure

• Education: Tutoring hours, students mentored, courses completed, certifications earned
• Caregiving: Hours of eldercare, childcare, disability support provided
• Civic participation: Volunteer hours, community board service, election work
• Creative output: Works published, performances given, open-source contributions
• Environmental impact: Carbon offset, habitat restoration, cleanup events

What's Hard to Measure

• Quality vs. quantity: 100 hours of bad tutoring isn't worth 10 hours of great mentorship
• Intangible contributions: Being a good neighbor, maintaining social cohesion, emotional labor
• Gaming: Any measurement system creates incentives to optimize for the metric rather than the outcome (Goodhart's Law)
• Power dynamics: Who sets the criteria? A government agency? An algorithm? Community vote?

Where AI Might Help

Ironically, AI—the same technology creating the problem—might help solve the measurement challenge:

• Automated impact assessment: AI could evaluate the downstream effects of contributions (did that tutoring actually improve test scores?) rather than just counting hours.
• Fraud detection: Pattern recognition to identify gaming without requiring a human bureaucracy.
• Peer validation: AI-mediated reputation systems where community members confirm contributions, similar to peer review in science.
• Dynamic pricing of social goods: An AI system could estimate the market-equivalent value of caregiving, mentorship, and community work in real time.

The Hybrid: UBI Floor + Proof of Benefit Bonus

The most pragmatic version may not be either/or. Consider a layered system:

This hybrid gets you:

• The security of UBI (no one falls through the floor)
• The incentive alignment of proof-of-benefit (contribution is recognized and rewarded)
• The political viability of maintaining a connection between effort and reward (the concern that kills pure UBI for most conservatives and many moderates)
• An expanded definition of work that includes the things humans will still do better than AI: care, create, connect, teach

The Artisanal Economy

One compelling vision of how proof-of-benefit could work in practice is the artisanal model: instead of everyone working for large employers, people create tools, content, and services for their immediate communities.

What This Looks Like

• A retired teacher builds an AI-powered tutoring system for her neighborhood and is recognized (and compensated) for student outcomes.
• A developer creates a scheduling app for the local Little League and is rewarded through the community's proof-of-benefit pool.
• A caregiver's daily work supporting an aging parent is formally valued at market rates, funded through the system rather than absorbed as unpaid labor.
• A citizen journalist covering city council meetings receives compensation proportional to readership and civic engagement impact.

Why AI Makes This Possible

Before AI, building a tutoring platform or a scheduling app required specialized skills and significant time investment. AI collapses the skill barrier: anyone with a clear idea of what their community needs can build it. The bottleneck shifts from can you code? to do you understand the problem?—and local knowledge becomes the scarce, valuable resource.

Open Questions

This framework is a starting point, not a finished policy. The hard questions remain:

1. Who administers the system? Government? A decentralized protocol? Community-run DAOs? Each option has different trust and corruption profiles.
2. How do you prevent a social credit system? There's a fine line between "rewarding contribution" and "scoring citizens." The system must reward positive action without punishing inaction.
3. What about people who genuinely can't contribute? Disability, severe illness, age extremes. The UBI floor handles this, but the framing matters: these people aren't "failing" the system.
4. Can measurement scale? Proof-of-benefit might work beautifully in a community of 500 where everyone knows each other. Can it work for 330 million people?
5. Does variable reward just recreate inequality? If some contributions are valued more than others, you're back to income stratification—just with different winners and losers.

Why This Matters for the AI Transition

The deepest version of this argument: if AI truly delivers abundance—cheap goods, cheap services, cheap energy—then the scarcity that remains is meaning. People need to feel useful. They need to know their contribution matters.

UBI solves the material problem. Proof of Benefit attempts to solve the existential one: in a world where machines can do most things better, faster, and cheaper, what makes a human contribution worth recognizing?

The answer might be: the things only humans can do. Care. Judgment. Creativity. Local knowledge. Moral reasoning. Community presence. These are precisely the contributions that traditional employment has always undervalued—and that a proof-of-benefit system could finally make visible.

← The Transition  ·  Reading List →