How Socialist Were the Nazis Really?
A conversation with Kristian Niemietz
After talking about the socialist credentials of the Nazis with Bob Lawson and Phil Magness, I had this spirited conversation with Kristian Niemietz, author of the astoundingly good Socialism: The Failed Idea That Never Dies. Niemietz pushes back against my thesis more than Lawson or Magness, so if you enjoy a serious argument, don’t miss our exchange.
The basic problem with referencing the ever-shifting definitions of socialism offered by self-proclaimed socialists over the years is that it doesn't provide us with a clear understanding of what socialism is.
Back when the socialist movement got started, the original premise was that private ownership of the means of production was evil because only the owners (the wicked "capitalist" class), not society at large, allegedly benefited from the existence of such private ownership. Earning a positive return on investments was portrayed as being a parasitic extraction of value from production, not as adding any value to production.
"Socialism" in this early context could broadly refer to any system for organizing production that prevents such private earnings on investments, instead reallocating them to the benefit of "society" according to whatever distributive outcome was considered just. As long as most of the public remains blissfully ignorant about how both private thrift and profit-seeking/loss-avoiding private investment of the resulting savings actually increases the quantity of outputs produced, it is possible to popularize all sorts of schemes to replace "capitalism" with a "socialist" alternative.
Socialism, in other words, was defined in a negative way as the antithesis of private ownership of production, which renders it hopelessly vague. Likewise, narrowly focusing on private ownership of the means of production without regards to other important requirements for a productive profit-and-loss driven economic system (like ownership being based on peaceful acquisition, ownership rights being defined in terms of the owner's exclusive control over the use and disposition of the owned thing, and everybody being at liberty to do anything that doesn't violate other people's ownership rights) leads to an overly-broad definition of "capitalism" as well. It conflates the genuinely productive laissez-faire variant of private ownership with the genuinely exploitative interventionist variants where the state rewards specially-favored private interests with subsidies, with privileges that enhance earnings by restricting the liberty of others, and with immunities that enhance earnings with exemptions from liabilities for violations of ownership rights.
Such fuzzy ambiguities give socialists an escape clause when confronted with evidence of how badly real-world socialist experiments work ("but that isn't an example of real socialism!"), but also permits a conflation of interventionism with laissez-faire that incoherently indicts private ownership with evidence of exploitative outcomes arising from what are actually deviations from libertarian/classical liberal principles. With such a foggy conceptualization of the nature of capitalism and socialism, the underlying issue of the economic necessity of private thrift and entrepreneurial investing for successful production is never confronted.
Bryan, I think you've won more of this than your critics will grant, and less than you're claiming.
What you've won: the "posers" story is dead. Anyone who has read Tooze or Barkai knows the Nazi state directed economic life to a degree that makes "it was just marketing to trick the workers" untenable. And the mutual-excommunication point is genuinely devastating. When every socialist faction has denied the credentials of every other one, the people running that machine have forfeited any standing to police the boundary. You're right about that, and it's the strongest thing in all three conversations.
But there's a case neither Lawson nor Niemietz let pass, and in my view it kills the strong version of your thesis: Franco.
Niemietz put it to you directly: run your exercise on Spain in the 1950s and you'll get a very low number too, but nobody calls Franco a socialist. That's not a cheap shot, because Franco's Spain really was extraordinarily state-directed: autarky, the INI running whole swathes of industry as a state holding company, multiple overvalued exchange rates, quantitative import restrictions, pervasive price and quota controls, and the vertical syndicates that replaced independent unions with corporatist bodies. Same decade, same continent, same anti-Bolshevik self-presentation, comparable degree of economic unfreedom. If the index can't tell Hitler from Franco, it isn't measuring what you're asking it to measure.
And Lawson got to the same place independently, from a completely different direction. He noticed that some countries at the bottom of his own index aren't socialist at all: Sudan, Yemen, just gangs fighting over turf with no plan and no ideology. So he proposed a second condition: low economic freedom *plus* an organizing collectivist ideology directing all that control toward some collective end. That's the sharpest single move anyone made across the three conversations, and it's a real improvement.
It also doesn't save you, because Franco passes it. National Catholicism is an organizing ideology. The collective end is stated plainly in the doctrine. INI and the vertical syndicates are the collective control. He clears both bars comfortably.
Which generalizes. Salazar clears them. Perón clears them. Vargas clears them. Any criterion broad enough to catch Hitler catches all of them, and any criterion sharp enough to exclude them excludes Hitler.
But I don't think this should bother you much.
You already said the right thing. It's at the end of the Magness conversation: the real spectrum runs totalitarian at one end to libertarian at the other, and the extremes were never separated in the first place except by cosmetic differences. I think THAT is the correct claim . And it doesn't need the word "socialism" anywhere in it.
What you're actually pointing at is a genus; let's ccall it the "anti-liberal collectivist economic order", systems that subordinate private economic decision to a collective purpose set by the state. Marxist-Leninist socialism is one species: it takes the title, plans the allocation, and justifies itself by the emancipation of the producing class. Fascist corporatism is another: it leaves title in private hands and exercises command instead, and justifies itself by the greatness or survival of a national or racial collective. Different property form, different telos, same genus, converging effects, which is precisely what Orwell noticed and what Niemietz conceded to you without much of a fight.
Drop Franco into that and nothing breaks. He's just a third member of the family, and a fairly ordinary one.
Look at what the reframe buys. The Eastern Front becomes a war inside the collectivist family, which is your real point and the one that matters. The Benatarian ledger stands untouched, neither species earns a halo for defeating the other. Left-versus-right as a back-projection of modern categories onto people who never thought in those terms, which is where you and Magness ended up, survives completely. And the whole thing is now immune to the Franco problem instead of embarrassed by it.
What it costs is the pleasure of watching socialists squirm being compared to Nazis. That's the entire price.
Which brings me back to the first thing Niemietz said to you, that he's uncomfortable with an argument that's emotionally too satisfying to make. It's why he conceded nearly everything about the Nazi economy while still refusing your conclusion. That is a man applying his own standard consistently and swallowing a result he'd rather not.
One practical suggestion, since arguing about this is less fun than settling it. Prados de la Escosura built a historical economic freedom index in the *Economic History Review* that covers Spain well before Fraser's series begins, and the 1940s–50s autarky years show up as a distinct trough. Run Spain 1950–58 against your German numbers. If Franco lands within a point of Nazi peacetime Germany, the instrument has told you it can't do the job. If he lands well above, I'm wrong and your case is stronger than I'm giving you credit for.
I'd also gently note that "Nazi peacetime below Imperial German wartime" is a fine-grained comparison between two extreme states, and I doubt ChatGPT estimates can bear that much weight even if the direction is right. But the Spain test is cheap, and it's the one that would actually move my argument.