In a little over a decade after the 1917 Revolution, the Soviet Union vaulted from a mostly agrarian economy to an industrial power. By 1940, overall industrial output was several times its late-1920s level; steel rose from about 4 million tons to roughly 18 million; coal from the tens of millions to well over 150 million tons; electricity generation multiplied manyfold; and the urban share of the population jumped from about 18 percent (1926) to a third (1939). The exact growth rates are still argued over because the historical series are messy, but the broad picture isn’t: the 1930s brought rapid, transformational industrial growth.
That sets up the real question: could the same goals have been reached with less coercion, fewer broken lives, and a more balanced economy? Or, as Alec Nove famously asked, was Stalin really necessary?
In this post I sketch three ways people answer that question and what each implies about speed, risk, and long-term damage.
No Revolution Required?
This view says the Bolshevik rupture wasn’t economically required. By the 1890s–1914, Russia already had a (messy) capitalist modernization underway: output was rising, railways knit the empire together, heavy industry grew around new hubs, literacy climbed, and reformers like Witte and Stolypin pushed finance, infrastructure, and farm restructuring. On this reading, war and revolution shattered a path that could have delivered factories and cities without collectivization or a police state.
To give it some empirical weight, economists have tried counterfactuals. One prominent exercise projects pre-1914 trends forward and finds that, by around 1940, a capitalist Russia might have achieved GDP per capita in the same neighborhood as Stalin’s results but with far lower human costs (NBER paper by Cheremukhin, Golosov, Guriev and Tsyvinski). Others are more skeptical. Robert Allen’s simulations summarized in his book Farm to Factory: A Reinterpretation of the Soviet Industrial Revolution suggest a hard-budget capitalist path struggles to mobilize enough savings and investment quickly, implying slower heavy-industry build-up and weaker capacity in the critical 1930s. Same question, different model, different answer.
Output, however, is only half the story. Even if a capitalist path had matched Soviet output by 1940, the distribution of gains would likely have been very different. Under Stalin, money incomes for most people rose only modestly and mostly after the mid-1930s, but the state redirected a huge share of resources into mass schooling, basic health care, urban services, and new avenues of social mobility, especially for workers and women entering skilled jobs. Quality was often poor and shortages constant, but the “social wage” expanded rapidly: literacy surged, infant mortality fell, and whole regions gained electricity and clinics that private capital had little incentive to supply at that speed.
There’s also the structure of growth. Russia’s pre-1914 insertion into the world economy was peripheral: primary exports out, capital goods and technology in. Following Samir Amin, sustained catch-up typically requires some degree of “delinking” and a self-centered development strategy that biases resources toward domestic capital-goods industries rather than export enclaves. In a no-revolution counterfactual, the question isn’t only “how much output?” but “which output?” Would a capitalist Russia have built the same heavy-industry backbone, or tilted toward extractives, processing, light industry, and assembly of imported parts to serve core markets? Without strong protection, state banking, and procurement geared to producer goods, moving up the industrial ladder fast is hard.
A glance at the 1990s is a cautionary mirror, not proof: when the post-Soviet economy reintegrated on liberal terms, specialization slid toward hydrocarbons, metals, and assembly while many complex capital-goods chains withered. That experience doesn’t settle the 1930s counterfactual, but it shows how global hierarchies tug structure as much as they set prices.
So the continuity thesis poses a trade-off in plain terms. Maybe you could get comparable factories and rails without the terror. But you would also get a different social map: likely higher inequality, slower rollout of universal services, and a narrower ladder of mobility. The claim, at its strongest, is not just that a revolution wasn’t necessary for industrialization. It is that the mix of growth and social outcomes would have been fundamentally different.
A more general question, however, is whether continuity was politically possible under those conditions. As Kagarlitsky argues in Empire of the Periphery, had the prewar modernization been sufficient, there would have been no 1917; the revolution is better read as evidence of that route’s failure than its promise.
A Gentler Socialist Road
This second view takes the revolution as a given but rejects the claim that only a crash program could deliver modernization. The core idea: with a mixed economy and a strategic state, the USSR could have industrialized more gradually, with far less coercion and fewer long-term distortions.
NEP is the exhibit A. In the 1920s the state kept control of the “commanding heights” while letting markets work in agriculture, small industry, and trade. Peasants had incentives, shops reopened, light industry revived, and tax revenues stabilized. Markets and planning didn’t cancel each other out; they complemented each other.
Crucially, even the leading advocate of this gradualist line, Nikolai Bukharin, knew what that implied for speed. He warned that sticking to balanced accumulation would move “at a turtle’s pace.” In other words, the bargain was explicit: slower headline growth in exchange for social stability, peasant buy-in, and fewer economic distortions.
There was also a political-ideological brake. In the leadership’s reading of Marxism, small private farming tended to regenerate “capitalist relations.” Lenin had framed that logic early on in The Development of Capitalism in Russia (1899), and by the late 1920s the grain procurement crises and the visibility of NEP “dealers” seemed to confirm it. With a thin urban proletariat and a countryside still dominant, continued NEP looked not just slow but dangerous: a possible road to capitalist restoration. That fear made a gradual, market-friendly socialism hard to defend inside the party.
Did collectivization actually finance the industrial push? Investment certainly surged in the early 1930s, but once you count the flow of industrial goods back to the countryside and the costs of disruption, classic studies show that agriculture was at best a modest net contributor and in some calculations a net drain. Resistance and dislocation forced the state to recycle much of what it “extracted” just to keep production from collapsing. Economically, collectivization didn’t unlock a hidden surplus so much as create losses that had to be patched.
What about the models that ask, “Could NEP have been enough?” Counterfactuals help frame the trade-offs. One prominent exercise projects pre-1914 or NEP-era dynamics forward and finds that Russia could plausibly match Stalin’s broad growth outcomes by 1940 under alternative policies, at far lower human cost. Holding the overall investment rate constant, Allen’s simulations push in a different, more uncomfortable direction: if you hold the overall investment rate constant and compare like with like, a collectivized economy delivers slightly higher non-agricultural value added and an 8 percent larger non-agricultural capital stock by 1939 than a NEP regime with soft budgets. By contrast, when comparing a high-investment NEP counterfactual to the actual crash-industrialization path (with even higher effective investment plus soft budgets and forced reallocation), NEP still ends notably lower in non-agricultural output (22 percent) and somewhat lower in capital stock.
So the balanced-socialist case isn’t nostalgic. It concedes slower speed up front, even “turtle-like” at times, but claims better distributional and institutional outcomes: peasant incentives preserved, famine avoided, a steadier expansion of the social wage (schools, clinics, housing), and a planning system less deformed by emergency controls. On this view, scrapping NEP rather than reforming it was the real fork in the road. A regulated market socialism was possible. The leadership chose something else.
The No-Choice Argument
This third view says the choices weren’t really choices. Start with the constraints: a late-industrializing, mostly agrarian economy; thin domestic capital markets; hostile geopolitics; and crisis-era terms of trade that cut the wrong way. As Kagarlitsky notes in Empire of the Periphery, during the late-1920s–early-1930s the volume of world trade fell by roughly two-thirds; imported machinery got cheaper, but Soviet export prices fell faster. Official tallies put the net price shock at roughly a 1.1-billion-ruble loss. The cheaper the grain, the more of it had to be shipped. As Alexander Gerschenkron argued in Economic Backwardness in Historical Perspective (1962), the greater the initial backwardness, the more you rely on ‘substitute’ institutions — powerful state banks, tariff walls, planning ministries, and, when convenient, coercion — to compress time, bias investment toward producer goods, and finance growth through forced saving rather than consumer markets.
Read the 1930s through that lens and the “big push” looks less like madness and more like a brutal kind of rationality. The leadership expected war and prioritized steel, machines, and arms. Investment was driven upward, resources were shoved into heavy industry, and planners accepted a collapse in living standards as the price of speed. Archival histories of the period tie this directly to defense planning and the international environment, not just ideology. As Stalin put it in 1931, “Either we do it, or they will crush us,” a motto for a decade of defense-driven planning shaped by the international environment as much as ideology.
Do the numbers back that up? Some model exercises say yes, with asterisks. Cheremukhin, Golosov, Guriev, and Tsyvinski estimate in their paper that Stalin’s policies inflicted a huge short-run welfare loss during 1928–40 (on the order of a quarter of consumption), yet by reducing frictions in labor and capital flows they yielded modest long-run gains for later cohorts; children born around 1940, for example, look better off in the model’s lifetime calculus. Robert Allen’s simulations pull in a similar direction: if the goal is to maximize industrial capacity within a decade, the collectivized high-investment path beats both a reformed NEP and a capitalist alternative by 1939 on non-agricultural output and even on consumption. In short, if survival and armaments in the 1930s were the overriding objective, the crash path “works” on its own terms.
But the asterisks matter. The same evidence also records what made that speed possible: a catastrophic hit to consumption at the start, the destruction of peasant autonomy, and a command system that hardened into long-term distortions. In this reading, inevitability is not a moral defense; it’s a description of a strategy under binding constraints. The claim is simple and harsh: given where Russia started and what it feared, a slower, gentler route probably wouldn’t have delivered the steel and machines in time. The bill was paid by society first and the economy later.
Conclusion
These three camps don’t really fight over the basic record so much as over feasibility and choice. You can grant all three premises at once: late-imperial Russia had real momentum; a mixed socialist model was workable; and backwardness plus looming war narrowed the menu. The argument is about the shadow price of speed: how much coercion, institutional lock-in, and foregone consumption it took to compress decades of industrialization into one.
My own view sits between the second and third positions. Given what we know about the limited net surplus actually squeezed from agriculture and the scale of human loss, a softer, more balanced socialist strategy looks economically plausible and morally preferable. A reformed NEP with hard investment targets could have delivered comparable industrial capacity by the early 1940s, with fewer long-term distortions and a very different distribution of gains.
But plausibility on paper isn’t the same as political feasibility in 1928–32. Information was thin; the leadership’s worldview was siege-minded, which was hardly irrational given the collapse of collective-security efforts after 1933 and, ultimately, the 1941 German invasion; and incentives inside the party-state pushed toward command solutions. From that vantage point, the crash path felt “inevitable,” even if it wasn’t in a strict economic sense. The late-1980s debates in the Soviet Union revived the counterfactuals, including the idea that Lenin might have kept a mixed economy longer. As Alec Nove cautioned in the last edition of An Economic History of the USSR, it’s far from obvious he would have chosen differently had he lived into the 1930s.
So, I land here: a more humane path was possible, but not likely under the beliefs, threats, and institutions of the time. The lesson isn’t that speed always justifies the means. It’s that the means shape the society you end up with, long after the steel is rolled and the machines are built.