DSA Delusions #2: "Free" Is Not Free!

The DSA's platform doesn't read like a set of policy proposals. It reads like a wish list handed to a genie: paid leave, a shorter week, a living wage, free school from age four through a doctorate, debt wiped clean, health care that costs nothing, and a house guaranteed to everyone with rents capped by law. Nowhere in the document is there a price tag, and that omission is the whole trick. Every one of these promises is achievable in isolation, at some cost, for some people, for some time. None of them is achievable simultaneously, for everyone, forever, at zero cost — which is what "guarantee," "free," and "for all" actually mean once a platform prints them.

Three planks from the DSA's "Workers Deserve More" program are on the table this time: Freedom to Flourish, Healthcare for All, and Housing for All. Each promises something real people want, but without any mechanism to pay for it. And the record on what happens when governments have tried pieces of this before is not encouraging.

Freedom to Flourish

Start with the "living minimum wage." The Congressional Budget Office's own analysis of a $15 federal floor found it would lift roughly 1.3 million people out of poverty while directly raising pay for 17 million workers — real gains, worth stating plainly. It also estimated 1.3 million workers would lose jobs they'd otherwise have held, with a two-thirds chance the true number falls anywhere between zero and 3.7 million. Families below the poverty line would gain about $8 billion in income; families above it would lose about $16 billion, as higher prices and thinner business margins ripple outward. If you’ve read many of my prior articles, you know that a crucial part of sound and fair economics relies on trading value for value. Transferring money from one group to another defies the idea of fair exchange. It's also evidence that even the CBO's own model treats "living wage" as a transfer with losers, not a costless correction — something the platform never admits.

The 32-hour week is the plank with the best real-world evidence behind it, and it deserves an honest look rather than a dismissal. Trials in the UK, Germany, Portugal, and South Africa found that many participating employers held revenue steady on a four-day schedule, with burnout and turnover falling sharply. But every one of those trials was voluntary: companies that suspected a shorter week wouldn't work for them either didn't sign up or dropped out mid-trial, as two large German firms did. A retailer, a hospital, a farm, and a manufacturing line don't have the same slack to compress as the software and professional-services firms that dominate these pilots — for a lot of work, output scales with hours worked, not with how motivated the staff feels. "Full pay and full benefits" mandated by law across every industry is a different claim than "some willing employers found a four-day week worked for them," and the platform quietly substitutes the second for the first. Furthermore, it ignores the fact that we already have labor shortages in several areas, and this will make them worse.

Then there's "free, quality public education from pre-K to college and cancel all student debt." Free means someone else pays the tuition; it doesn't make the professors, buildings, or years of a young person's labor market participation free to produce. Again, they are transferring value without reciprocation. But the debt-cancellation half is the more telling error, because the distributional data runs backward from the platform's own framing. Asking those creating value to give up their created value for others who have already received the value they agreed to pay. Research by Sylvain Catherine and Constantine Yannelis found that blanket cancellation sends nearly half of all relief to the top 30% of earners, with the top income decile collecting more than five times what the bottom decile gets — because lower-income borrowers were, on average, never going to repay the full balance anyway, so the balance on paper overstates what cancellation is actually worth to them. An income-driven alternative flips that ratio, sending the bottom decile four times what the top decile gets. "Cancel all student debt" sounds like a policy for workers. Measured in dollars delivered, it's tilted toward the people who can already afford a closet full of workers-deserve-more T-shirts.

Healthcare for All

"At no cost to individuals" is doing a lot of quiet work in this plank. Care is never actually free — doctors, nurses, hospital beds, drugs, and equipment all consume real resources that have to come from somewhere. "At no cost" just means the bill is paid through taxes instead of a price at the point of service, and the biggest independent estimate of what that bill looks like comes from the Mercatus Center: roughly $32 trillion in new federal spending over ten years. That figure already assumes a roughly 40% cut to what Medicare pays hospitals and doctors — an assumption the study's own author calls unrealistic, meaning actual costs are more likely to run higher, not lower. Even doubling every federal individual and corporate income tax rate would not fully cover it. Whatever else is true about the merits of universal coverage, "at no cost" is not a description of the program; it's a description of who's being asked not to look at the invoice.

Setting the price to zero doesn't make scarcity disappear — it just changes how care gets rationed, from a price at the counter to a wait in line. Canada, which the platform's supporters often point to, illustrates the trade-off: the Fraser Institute's 2025 report puts the median wait from a GP referral to actual treatment by a specialist at 28.6 weeks, the second-longest on record. That's not a knock on the idea of broad coverage; it's a reminder that "complete access" and "no cost" and "no delay" can't all be guaranteed at once when the number of specialists, MRI machines, and operating rooms doesn't expand just because the law says access is guaranteed. Combine this with the 32-hour work week, and you have 80% of current healthcare hours having to treat more patients somehow.

The plan's other promise — making all medical education and training "public and free" — has the same blind spot as free college generally, plus a sharper one: the binding constraint on the U.S. physician supply isn't tuition; it's the fixed number of residency slots hospitals are funded to offer. Waiving tuition doesn't create a single additional resident training position; it just changes who pays for the education of the same number of doctors the system was already going to produce. And remember, cutting physicians to a 32-hour work week means we'll need 25% more immediately, not after 9 years of education. If the goal is genuinely more doctors and shorter waits, the bottleneck is capacity, not price — and a platform that names only the price is treating the symptom it can put a number on instead of the constraint that actually decides how many people get seen.

Housing for All

"Universal rent control" has already been run as a natural experiment, and economists have measured the result. When San Francisco expanded rent control in 1994, Stanford researchers Rebecca Diamond, Timothy McQuade, and Franklin Qian found that covered buildings saw a 15% decline in rental housing supply, as landlords converted units to condos (an 8-percentage-point jump in conversions) or redeveloped properties to exit rent control entirely; renters actually living in rent-controlled units fell 25%. Existing tenants who kept their units benefited, especially minority renters, who were far more likely to stay in the city. But the researchers also found the policy pushed up market rents for everyone else, because the housing supply it protected for incumbents is housing it removed from the market for newcomers. Rent control doesn't lower the cost of housing; it redistributes who bears that cost, protecting whoever already has a lease at the expense of whoever is still looking for one — exactly the trade-off "universal" rent control would lock in nationwide.

Social housing has a real example too, and it's more complicated than the platform's one-line pitch. Vienna is the model most often cited, and a closer look, laid out by the American Enterprise Institute, shows why it's a hard one to copy. It runs on a 1% wage tax charged to every worker in Austria, whether or not they ever live in social housing, and even that wasn't enough: the system ran a deficit equal to 1.6 times its annual rental income in 2016. New tenants pay a 17% premium over long-term residents, and getting into a cooperative unit can require roughly $38,000 upfront — a barrier that has left wealthier applicants disproportionately in the nicest, most subsidized units while lower-income families are effectively priced out of the program built in their name. And Vienna's stock was built under conditions — postwar hyperinflation, political upheaval, the collapse of an empire — that don't exist anywhere the DSA is proposing to replicate it. Cheering the outcome while skipping the financing and the fine print isn't a housing policy; it's a bumper sticker.

The two remaining pieces — strictly regulating investment properties and guaranteeing counsel for every tenant — run into a version of the same problem from the other direction. Capping the return on rental property makes owning rental property less attractive, which tends to shrink the supply of it over time, the same mechanism rent control demonstrates on its own. And a national right to counsel in every eviction case would mean building, essentially from scratch, a public-defender-scale legal system for housing court — a real and expensive undertaking that the platform doesn't cost out any more than it costs out the rest of the program. None of that makes tenant protections a bad goal. It just means "guarantee" is carrying weight that the actual mechanics of housing supply and legal capacity haven't been asked to carry.

The pattern underneath the planks

Three different topics, one recurring error. Set a price at zero, and the cost doesn't vanish — it shows up as a tax bill, a wait list, or a shrinking supply of the very thing being guaranteed. Cap a return and the resource tied to that return — rental units, workable hours, willing lenders — tends to shrink rather than redistribute cleanly. And a transfer dressed up as a universal guarantee, like blanket debt cancellation, often moves money in exactly the opposite direction its name implies. None of this is an argument that family leaves, health coverage, or housing security don't matter, or that markets get every outcome right. Still, the evidence shows that when the government actively guards the market without interfering in it, the result is more benefit for more people. In fact, no socialist system even comes close to the benefits capitalism provides. It's an argument that scarcity and incentives don't take a day off because a platform used the word "guarantee." Every economy ever has run into that arithmetic, which is a large part of why I found it worth writing an entire book about it.