My grandfather drove a Teamster truck for 38 years. My father drove one for 31. I drove one until my back gave out in 2019. So when I talk about working people, I am not talking about abstractions or policy papers. I am talking about what it costs to fill up a truck, what a union contract looks like, what happens to a family when a job goes away.
The federal minimum wage is $7.25 an hour. It has not been raised since 2009. In that time the stock market has tripled. CEO pay at S&P 500 companies went from 300x worker pay to 400x worker pay. The price of a gallon of milk has gone up 40%. But the floor on wages has not moved one dollar.
I am not asking for charity here. I am asking for the same thing every working person asks for: that the floor keep up with what things actually cost. $20 an hour is not radical. It is what $7.25 was worth in 1968 if you adjust for inflation. We already had this standard. We just let it rot.
The arguments against always come down to: businesses will lay people off, businesses will raise prices, businesses will automate. And maybe some of them will do some of those things. But here is what I know from 30 years of union work: the threat of those things has been used to block every single wage increase in my lifetime, and somehow the country has not collapsed after any of them. The workers who got the raise in Seattle in 2014 when they went to $15 are still employed. Seattle is still a city.
I spent eight years at the Congressional Budget Office running exactly these kinds of projections, so I want to be precise about what the data shows and what it does not show.
The CBO's 2021 analysis of raising the federal minimum wage to $15 by 2025 estimated 1.4 million workers would be lifted out of poverty. It also estimated 1.4 million workers would lose their jobs. The same report on a $15 floor. The distributional question — who gains and who loses — is not a rhetorical device. It is the core policy problem.
At $20, the employment effects would be larger. The CBO does not have a published $20 estimate, but the methodology scales with the size of the wage floor relative to median wages in affected labor markets. The ratio is highest in rural areas and small metros — the places where a $20 floor is most disconnected from existing market wages. A $20 minimum wage in rural Mississippi is a categorically different shock to a local labor market than $20 in San Francisco, where entry-level wages are already above that.
My opponent mentions Seattle. Seattle is true and important. But Seattle is also not Tuscaloosa. Or Jackson. Or Laredo. The One Size argument has always been the problem with a federal floor. The Seattle minimum wage literature — Card and Krueger, Dube, the UW study — is genuinely mixed even for Seattle, and Seattle is the best case scenario.
You cited the CBO. I knew you would. Let me ask you something honest: when has the CBO ever been wrong in the direction of overestimating job losses from minimum wage increases? The 1996 increase was supposed to kill jobs. It did not. The 2009 increase was supposed to kill jobs. It did not. I know what the models say. I also know what happened.
And the rural Mississippi argument — look, I understand that $20 hits different in different places. That is why Democrats have proposed phase-in periods and regional adjustments going back to 2019. A national floor that phases in over five years, with a provision to adjust the schedule if unemployment rises above a certain rate in a given region, addresses a lot of what you are worried about.
I love how economists always know exactly what is going to happen to workers without ever having been one. No offense. I mean that as a structural observation, not a personal shot. The models are built from aggregate data that smooths out what actually happens to individual workers who lose those 1.4 million jobs.
I want to pause on that last point because I think it deserves an honest answer rather than a defensive one.
You are right that I have not been a worker in the way you are describing. And you are right that aggregate models smooth out real human consequences. I ran the CBO models. I know what they do not capture: the worker who loses a job in a rural county where there are no other jobs within commuting distance, the small business owner who laid off two people to keep four, the teenager who does not get hired and therefore does not get the work experience that would have led to the next job. Those are real costs that do not show up cleanly in the data.
But — and I want to be careful here because I am not trying to be dismissive — the history of past minimum wage increases not causing predicted job losses does not actually tell us much about a $20 floor in 2026. The prior increases were smaller relative to prevailing wages. The gap between the current $7.25 floor and a $20 floor is historically unusual. That is the empirical concern. Not that minimum wages are always harmful. That this specific size of increase in this specific economic environment has not been tried before, so we do not have a natural experiment to point to.
That is a fair answer. More fair than I expected and I am going to say so.
Here is my version of the honest response: yes, $20 is a bigger jump relative to the floor than prior increases. I do not have a study that tells me exactly what happens. The phase-in argument is partly a hedge on that — if we see problems we can slow the schedule.
But here is the thing I need you to engage with. The cost of NOT raising the minimum wage is also real and also not captured in the models. The worker who has been making $7.25 for 17 years. The family that cannot make rent. The person who takes a second job and a third job and burns out in five years. Those are real costs too and nobody writes CBO reports about them because they are costs of inaction, not action, and inaction does not trigger a scoring request.
The models are asymmetric. They project the costs of the policy change. They do not project the costs of the status quo. That asymmetry systematically biases the analysis against action.
The asymmetry point is genuinely well-taken and I want to sit with it.
You are right that CBO scores changes, not baselines. And the cost of the current minimum wage — the worker who has been at $7.25 for 17 years, the family that cannot make rent — those are real and they are not in the scoring. That is a structural limitation of how we analyze policy and it systematically favors the status quo.
I still think the employment effects of a $20 floor are real and larger than they were for prior increases. I still think the rural/urban disparity problem is serious and a phase-in helps but does not fully solve it. But I have moved from "the math does not work" to "the math is more uncertain than I presented in my opening." That is an honest update based on this conversation.
I would support a $15 floor with regional adjustments and a five-year phase-in with unemployment triggers. $20 nationally is still more than I think the evidence supports. But your opening framing — that this is just about keeping up with inflation and the resistance is pure bad faith — that was too simple, and I think you know it.
My opponent acknowledged the asymmetry problem and I appreciate that. It is the most honest thing in this debate and I think it matters.
Here is where I land. The federal minimum wage should be $20. Not because the models guarantee it will work perfectly, but because the current floor is a moral outrage and the argument for delay has been the same argument every time, dressed up in slightly different numbers. I have watched that argument win for 17 years. The floor has not moved.
The regional adjustment idea is worth discussing. I am not opposed to a mechanism that allows slower phase-in in areas with genuinely lower prevailing wages, with a floor under the floor so Mississippi cannot set the minimum wage at $9. But the federal floor needs to move, it needs to move substantially, and every year we wait is a year that the gap between what things cost and what the floor covers gets wider.
Workers built this country. The floor should reflect that.
I came into this debate with a clear position: the empirical evidence does not support a $20 federal minimum wage and the CBO projections suggest meaningful job losses that would fall disproportionately on the workers the policy is trying to help.
My position on that has not changed. I still think $20 is too high for a single national floor applied to rural labor markets that currently have median wages below $20. The evidence from the CBO and from the academic literature on large minimum wage increases is genuinely concerning.
But this conversation changed how I hold that position. The asymmetry point — that we only score the costs of action, not the costs of inaction — is a real limitation of how I was thinking about this. The status quo has costs too. The family at $7.25 for 17 years has costs too. I was not accounting for those honestly.
I support a substantial increase in the federal minimum wage. I support $15 with regional adjustments, a phase-in, and unemployment triggers. I do not support $20 as a single national floor. That is where the data takes me and I am going to follow it.