Three questions behind the isms
The system a person lives under tends to feel like weather: pervasive, unchosen, and not obviously one option among several. The words available for talking about alternatives do not help. Capitalism, socialism, and communism arrive pre-loaded, each carrying a century’s worth of flags and casualty figures, and an argument that starts with the labels usually ends there too.
The labels become usable once they are unbundled. Underneath the isms sit three questions that any economy, from a village to a continent, has to answer somehow. Who owns the productive assets: the farms, factories, buildings, and machines that make everything else? Who decides what gets produced and at what price: buyers and sellers discovering prices in markets, or an authority setting quantities by plan? And who has first claim on the surplus, the value left over after costs are paid: private owners, the state, or the workers of the enterprise itself? The familiar isms are bundles of answers. Capitalism, in its textbook form, answers private, market, owners. Communism, in its textbook form, answers state, plan, state. Between and around those bundles lies nearly everything that has actually existed.
Who owns, who prices, who claims
The first axis, ownership, is the one the isms were named for, and it is less binary than the naming suggests. Between a family firm and a state factory lie listed corporations owned by pension funds, state-owned enterprises with private minority shareholders, cooperatives owned by their workers, and utilities owned by municipalities. Ownership also decomposes: the right to use an asset, the right to its income, and the right to sell it can be held by different parties, which is why a heavily regulated private firm sits closer to the middle of the axis than its deed implies. Regulation is a partial public claim on rights the title still calls private.
The second axis, coordination, asks how production decisions find their quantities. Markets do it through prices: shortages raise them, gluts lower them, and producers respond without anyone instructing them. Plans do it through targets: an authority decides that this year requires so many tons of steel and so many pairs of shoes, and allocates inputs accordingly. The deep difference is informational. Prices compress dispersed knowledge about wants and costs into a single moving number; plans must gather that knowledge into an office. What happens to the gathered knowledge as an economy grows more complicated is the subject of Chapter 4.
The third axis, the claim on surplus, is the quiet one, and for the reader’s daily life often the most consequential. Two economies can both be private and market-coordinated while routing their surplus entirely differently: one leaves it with shareholders; the other taxes a large share into pensions, health systems, and transfers. The paycheck question of who keeps how much of the value a workplace creates, introduced in Chapter 1 as the bargaining margin, is this axis experienced from below.
What socialism names
Two of the title’s three isms now have bundles: textbook capitalism answers private, market, owners; textbook communism answers state, plan, state. The word between them has been left undefined so far, and it is the one used most loosely in argument. In the philosophical literature, socialism names an answer to the first axis: social ownership, or democratic control, of the means of production, with productive assets answerable to the people who work them or to the community affected by them (Gilabert & O’Neill, “Socialism,” Stanford Encyclopedia of Philosophy, 2024, §§1, 4). The definition cuts in two directions at once. Against capitalism, it rejects concentrated private ownership. Against statism, it rejects swapping a ministry for a shareholder: an economy whose factories answer to a party apparatus is certainly centralized, but nothing in that arrangement makes the control social. Whether state ownership can count as socialism at all, and under what democratic conditions, is a live argument inside the tradition rather than a settled equation, which is precisely why the axis matters more than the label.
Communism, in turn, is two objects wearing one name, and the bundle given above describes only the second. In Marx’s usage the word named a theory. The Critique of the Gotha Programme (1875) sketches a cooperative society based on common ownership of the means of production, passing through two phases: a lower phase still stamped, in Marx’s words, with the birthmarks of the old society, in which people are paid according to their contribution, and a higher phase in which distribution follows need. On the state the Critique claims less than the slogan later attached to it: freedom, Marx writes there, consists in converting the state from an organ superimposed upon society into one completely subordinate to it. The state that “withers away” is Engels’s formulation, from Anti-Dühring (1877–78). What the twentieth century built and called communism was a different object: administrative planning run by a permanent party-state that owned the assets, set the prices, and claimed the surplus. The theory and the practice share a vocabulary and little else that can be measured. The higher phase was never run anywhere at national scale, so it has no ledger; Soviet-type planning ran for six decades across a dozen countries, so it has an extensive one.
The distinction sets the reading rules for everything that follows. When the Soviet Union of 1980 sits at the planned pole of Figure 2.1, the marker records Soviet-type planning as it actually operated, a practice, not socialism’s definition or communism’s theory. Yugoslavia’s worker-managed firms took the social-ownership answer without the plan; the postwar welfare settlements took a collective claim on the surplus without touching ownership; the Soviet system took state ownership and the plan while the control stayed anything but social. Each wore one of the three labels in argument, and none is described by it. The performance evidence of the next two chapters attaches to the practices, placed on the axes, because practices are what history ran.
Everyone is a hybrid
Set actual economies on those axes and the poles empty out. The United States, the reference case for capitalism, routes a substantial share of national income through government at all levels and owns outright such capital as the interstate highways, the air traffic system, and the largest single buyer of medical services. Across the rich democracies, government spending ranges from roughly a quarter of GDP to more than half, with most members near two-fifths (OECD, Government at a Glance 2023). Spending measures the state’s fiscal weight rather than its claim on any firm’s residual, and much of it is insurance and transfers; but an economy routing two-fifths of its income through the state has plainly left the textbook pole, whichever axis it is read on. Sweden pairs resolutely private, market-coordinated production with one of the heaviest redistributive claims on the surplus anywhere; its capitalism and its welfare state are the same system read on different axes.
The composites run the other direction too. China since 1978 has moved far along the coordination axis toward market prices while the state retains ownership of the commanding heights: energy, banking, rail, and urban land, with rural and suburban land owned by village collectives; enterprises and households hold transferable use rights over both rather than title in fee (PRC Constitution, art. 10; Naughton, The Chinese Economy, 2nd ed., 2018). Yugoslavia spent four decades running worker-managed firms inside a market, answering the ownership question with neither private nor state. Norway leaves production private and market-priced while its state holds majority stakes in the country’s largest firms and banks the hydrocarbon surplus in a fund holding about 1.5% of the shares of the world’s listed companies, spread across more than seven thousand firms (Norges Bank Investment Management). The Gulf monarchies combine market coordination with hydrocarbon rents that make the state the residual claimant of nearly everything. Each of these is routinely filed under one ism or another; none is described by its file label.
Hybridity is the norm for a practical reason: the axes answer different problems. Ownership assigns responsibility for assets, coordination processes information, and the surplus claim settles distribution. A society can want market information-processing, broad distribution, and mixed ownership at the same time, and most electorates, given the chance, have voted for exactly that combination in some proportion.
Why the poles stay empty
The corners of the map are not merely unoccupied; they appear to be uninhabitable, and the failures are instructive in both directions. A pure market order, with the state confined to watching, has never been observed at national scale, because markets themselves run on machinery only a state has proven able to supply at that scale: enforceable contracts, recorded titles, adjudicated disputes, a currency worth quoting prices in. Remove the machinery and exchange retreats to whoever can enforce deals privately, which is a description of a protection racket rather than a market. What economies look like where that machinery has collapsed is taken up with state capacity in Volume III: Countries.
The planned pole proved equally unreachable from the other side. The Soviet economy never actually eliminated market exchange; it relegated it. Collective-farm markets sold food at uncontrolled prices throughout the system’s life, and beneath the plan ran what Gregory Grossman named the second economy, the sprawling unofficial layer of private production, barter among enterprises, and paid favors through which the official economy obtained what the plan had failed to deliver (Grossman, “The ‘Second Economy’ of the USSR,” Problems of Communism, 1977). Planners tolerated it because the plan could not run without it. Each pole, pressed to purity, quietly re-imported the other; the hybrids of the previous section are not compromises of principle so much as the only stable configurations anyone has found.
Where the systems came from
The isms also read differently once dated. They are not a menu that societies have always chosen from; they are layers, each assembled in response to the failures of the one before.
Mercantilism, the operating system of the seventeenth and eighteenth centuries, treated trade as a contest for a fixed stock of treasure and organized it through chartered monopolies. Adam Smith’s Wealth of Nations (1776) was an attack on that arrangement, and the market order it argued for became the nineteenth century’s default. Comprehensive planning arrived as a critique of that default’s crises and inequities, made operational in the Soviet Union’s first Five-Year Plan (1928). The mixed economy, built across the postwar West, was an explicit settlement between the two: market coordination, private ownership, collective claim on a large share of the surplus. Britain’s Beveridge Report (1942) supplied its social-insurance blueprint; the nationalizations and full-employment commitments that completed the settlement were assembled separately by the governments that followed. The last two arrivals moved the other way: China’s reform opening (1978) and Vietnam’s Đội Mới (1986) reintroduced market coordination inside state-owned shells, producing the hybrids that complicate every simple story about the Cold War’s verdict.
What the labels miss
Reading systems through the three axes clears away the two most common confusions in everyday argument. The first confusion treats any expansion of the collective surplus claim as a step toward communism, as if the axes were one lever. They move independently: Sweden’s tax share and the Soviet plan were different axes entirely, and an economy can slide along one for decades without stirring on the others. The second confusion runs the opposite direction, treating market coordination as if it entailed private ownership and thin redistribution. China’s four decades of market prices under state ownership are the standing counterexample.
What the unbundling cannot do is rank the bundles. Knowing that answers can be mixed says nothing about which mixtures deliver growth, health, or freedom, and that question has evidence rather than definitions: the split countries that ran different bundles on the same people. That evidence is the next chapter. Two bundles get closer examination later in the volume: the planned pole and its information problem in Chapter 4, and the worker-owned answer to the ownership question, rarer than its survival record suggests it should be, in Chapter 9. Whether any genuinely new bundle is still possible is the volume’s closing question in Chapter 10.