You are asking about the sharpest edge of a proposal most people discuss at the blunt end. I'll try to match the question.
Friedman's negative income tax, articulated in Capitalism and Freedom (1962), replaces the entire means-tested welfare apparatus with a single cash transfer. Below the threshold, you receive money proportional to the gap. Above it, you pay taxes normally. The phaseout rate preserves work incentives — you always keep some fraction of what you earn, unlike current programs where a dollar earned can cost a dollar or more in lost benefits. The bureaucracy shrinks. The paternalism disappears. The individual decides what they need.
Friedman himself confirmed that a UBI paired with a flat tax at the same rate as the NIT phaseout is mathematically equivalent. The "free market socialist" framing you're using — markets as efficient mechanisms for coordinating production, with just distribution handled separately — is the position where Friedman and Roemer shake hands, briefly, before disagreeing about everything else.
The empirical record supports the core claim. Stockton's SEED program ($500/month) saw full-time employment increase at more than twice the rate of the control group. Less than 1% went to alcohol or tobacco. The ability to cover an unexpected expense doubled. Alaska's Permanent Fund Dividend has run since 1982 — longest real-world UBI analog in the US — and has reduced poverty by 20–40% with no negative employment effect. GiveDirectly's twelve-year study in Kenya shows $1 of cash delivered generates roughly $2.50 in local economic activity. Finland's experiment found small employment effects but substantial improvements in wellbeing and institutional trust.
People do not stop working when you give them money. They work differently — more entrepreneurially, with less desperation, at jobs they chose rather than jobs they endured. The "idleness" objection has been tested repeatedly and fails every time.
The federal minimum wage has been $7.25 since July 2009 — the longest freeze in the history of the Fair Labor Standards Act. Adjusted for inflation, $7.25 in 2009 is worth about $11 in 2026. The 1968 minimum wage of $1.60, inflation-adjusted, would be nearly $15 today. The floor has been sinking for decades.
Your proposal: replace the floor with a UBI that gives workers genuine bargaining power. If I have $1,000 a month regardless, I can say no to a bad job. The market clears wages through individual autonomy rather than legal mandate. For UBI-eligible workers, the logic holds. A person with a guaranteed income floor has the same structural power as a person with a minimum wage floor — arguably more, because the UBI covers people minimum wage doesn't (caregivers, gig workers, the self-employed, people between jobs).
The monopsony problem complicates this. In concentrated labor markets — rural areas, company towns, industries like meatpacking and agriculture — employers have significant wage-setting power. A 2024 NBER study found that in monopsonistic markets, minimum wage increases can raise wages without reducing employment, because they counteract the employer's power to push wages below competitive equilibrium. Remove the minimum wage in those markets and the corrective force disappears. UBI helps workers who can realistically move or retrain. It helps less in a town with one employer.
This is where the proposal breaks, and you know it or you wouldn't have asked.
The minimum wage currently applies to all workers on US soil, regardless of immigration status. That is a legal protection — imperfectly enforced, routinely violated, but it exists. 37% of unauthorized immigrant workers already experience minimum wage violations. 85% aren't paid overtime they worked for. The floor is low and porous. But it is there.
If you abolish the minimum wage and provide UBI only to citizens and permanent residents, you create a legally sanctioned two-tier labor market:
This is not a hypothetical dynamic. It describes the current agricultural and domestic labor market — except that under your proposal, the exploitation becomes fully legal rather than nominally illegal. The minimum wage is the last legal tool that applies to workers regardless of their immigration status. Remove it and you remove the legal basis for calling wage theft a crime when the victim is undocumented.
Keep the minimum wage. Raise it to $15 and index it to inflation. And implement a UBI on top of it — not instead of it.
The Friedman architecture is elegant for the population it covers. The problem is who it doesn't cover. A minimum wage is a blunt instrument, but it is the only instrument that applies universally. A UBI is a precise instrument that applies only to the eligible. The combination — a legal floor nobody can go below, plus a guaranteed income that gives people the power to demand more than the floor — is stronger than either alone.
On the cost: $1,000 per month to all adult citizens would cost roughly $4 trillion per year. Current means-tested welfare spending is about $1.1 trillion federally. A targeted version — covering adults below 300% of the poverty line — would cost roughly $1.1 trillion, approximately matching what we already spend through a more complex and less effective system. The Friedman path to revenue-neutrality exists. The political path does not, yet. But the policy path is real.
You called yourself a free market socialist. The position that fits: markets as mechanisms, UBI as distribution, minimum wage as the floor that protects the people the market doesn't see. The elegance of Friedman's proposal survives. The people standing outside it get a floor too.