Russia / USSR — Econography

Every other economy on this scale argues with its statistics. This one begins by disbelieving them.

The Soviet numbers were an instrument of the plan, not a picture of the country. TsSU, the state statistical committee, reported that the economy grew roughly ninefold from 1928 to 1985. The CIA's painstaking reconstruction said four-to-five-fold. The dissident economist Grigorii Khanin, working from physical output, said perhaps threefold. All three cannot be true, and the honest answer is that the concept itself — value, in an economy without prices that meant anything — was partly undefined. This volume uses Western estimates (CIA and Maddison-tradition) for Soviet-era growth, and treats even those as directional.

Official inflation was near zero, and it was a lie of category rather than arithmetic. Prices were fixed; the adjustment happened in shortage. The queue was the price. A stats line reading "inflation ~0 (repressed)" should be understood as: the money was losing value in hours-of-waiting and in unbuyable goods, not in the index. When the lid came off in 1991–92, forty years of suppressed inflation detonated at once.

The ruble had no price. Multiple exchange rates, certificate stores (beryozka), a black-market rate several times the official one: converting Soviet output to dollars is estimation, not measurement. PPP reconstructions put Soviet living standards far higher relative to America (30–40% in the 1970s) than the market-rate chain used here (~20–25%); the gap between those two numbers — what the system could make versus what its money could buy — is itself one of this volume's findings.

After 1991 the numbers professionalize and the economy hides. Rosstat became competent; meanwhile half of industrial sales went to barter (1998), survival ran through dacha potatoes and shuttle traders, and the measured collapse both overstates (unwanted tanks no longer counted) and understates (uncounted survival work) the lived one. From 2015 Rosstat's totals include Crimea.

Scope: the narrative to 1991 is the Soviet Union's; the stats line prices the Russian republic (RSFSR proxy), for continuity with what follows.

Verdict: trust direction, doubt levels, and read the queues — the CIA and Maddison reconstructions at the direction only, the plan-era stats line as physical output rather than welfare, and Rosstat's post-1991 books as competent bookkeeping of an economy that spent the 1990s largely outside them.

The queue was the inflation. For forty Soviet years the official price index read zero while the true price — hours of waiting, connections cultivated, goods that simply weren't — compounded silently. When prices were finally freed in 1992, four decades of suppressed inflation detonated in one 2,508% year. The lesson generalizes: this volume's Soviet half must be read with shortage as the hidden deflator, and its ratings do so — the 1980s' ten straight 3s are a decade of statistical calm and lived erosion.

Six raids on the household's money — this scale's record outside Argentina. 1947 (10:1 with tiered savings cuts), 1961 (the kopeck shave), 1991 (Pavlov's three-day voiding), 1992 (the hyperinflation wipeout of the sberkassa), 1993 (the vacation-season swap), 1998 (redenomination, then default, eight months apart). Two total savings destructions inside six years (1992, 1998). The behavioral residue is permanent: dollars in drawers, apartments as the only trusted store of value, and the lowest intergenerational financial trust of any economy on this scale — earned six separate times.

Grain is the system's biography. Exporter under the tsars; 1946's famine; 1963's first gold-for-wheat humiliation; 1972's Great Grain Robbery; the 1970s–80s as the world's largest importer, oil paying for Kansas wheat in a permanent carousel; then — the quiet post-Soviet revolution — private farming and the cheap ruble turning Russia back into the world's largest wheat exporter by the 2010s. The one unambiguous structural success of the market era, and almost never the headline.

Oil was the morphine, three times administered. The 1973 windfall arrived exactly in time to cancel the Kosygin reforms' necessity; the 1986 crash arrived exactly in time to bankrupt perestroika; the 2000s superprice arrived exactly in time to fund the restoration and postpone diversification; and 2013 — growth below 2% at $110 oil — was the printed proof that the dose no longer worked. Every stagnation in this volume is downstream of a barrel.

The 1990s were a Great Depression — deeper and longer than the original. Per capita output fell ~44% from 1989 to 1998 (America 1929–33: ~29%, recovered by 1940). Add hyperinflation, two savings wipeouts, male life expectancy collapsing under 58, and a state that stopped paying wages for months at a time. That the society held — through dachas, barter, shuttle trade, and family networks the statistics never saw — is the decade's real economic story, and the origin of everything politics did next.

Jobs held, wages flexed — the Russian adjustment. In every post-Soviet crisis (1998, 2009, 2015, 2020) unemployment barely moved while real wages and hours absorbed the shock — the mirror image of the American model. It bought social peace at the price of zombie firms and frozen productivity: the monotown preserved, the transformation postponed.

Pride and the pantry ran separate accounts. Sputnik and the sausage queue; Gagarin and the communal kitchen; Crimea at 86% approval and seven straight years of falling incomes. No economy on this scale shows a weaker transmission from national triumph to household consumption — or a population more repeatedly willing, for a season, to accept the substitution. The state noticed, and budgeted accordingly.

The fortress was built in plain sight. After 2014: reserves rebuilt past $600bn, debt near zero, a budget balancing at $40 oil, inflation targeting mastered, import substitution in food, SPFS payment plumbing — a decade of visible preparation universally read as mere conservatism. The restoration's final macro position (2021: fat NWF, minimal debt, record harvest, 4.8% unemployment) was the strongest in the volume — assembled, it became clear twelve weeks after this volume closes, for consumption of a different kind.

Competence arrived late and partial. The volume's institutional bright spots cluster oddly: a central bank (Nabiullina's CBR) achieving tsarist-era-first price stability; a finance ministry saving windfalls (Kudrin's funds paying for 2009 and 2020); a tax service digitized to world class — islands of first-rate technocracy inside a state that jailed its richest man by checklist and amended its constitution by plebiscite. The pattern is the volume's answer to why growth stopped: money was finally sound; property and law never were.

The terminus is chosen, and the reader knows why. This volume ends in December 2021 by design — the last year the Russian economy's story can be told as economics. The entries do not know what February 2022 brings; every reader does. The fortress's purpose, the stagnation's tolerance, the pride-over-pantry transmission — the patterns above are the mechanism of what followed, described in the only tense available to the years themselves: the present, before.

The 1946–1961 ratings, challenged and upheld (audit of July 2026). A reader's challenge — that these opening chapters run too low for a rebuilding superpower whose people believed, whose production boomed, and whose rivals privately debated when, not whether, they would be overtaken — was examined against the entries and declined, and the reasoning belongs in the volume. What the challenge asks for, the Khrushchev chapters already grant: 1955–60 stand at 5 across the board and 1961 at 6 — above Spain's same years, level-for-level beside Iran's golden decade — with the housing program called the era's greatest lived gift and the overtake boast recorded as half-believed on both sides; the when-not-whether consensus (CIA growth estimates of 6–7%, the textbook projections that dated the crossover rather than doubting it) is the consensus lens working, and it is why those years rate as high as they do. What the challenge cannot move is the floor. The famine of 1946–47 killed on the order of a million people while grain was still being exported, and these ratings do not grade famine on a national curve: China's 1959–61 and India's 1966 sit at 1 on identical doctrine. And round one held 1948–52 at 3 on the median-is-rural rule: three-fifths of the country lived on kolkhozes where labor-days paid near zero, taxes drove families to fell their own orchards, and grain per head remained below the 1913 level into 1953 — a crisis Khrushchev himself admitted from the plenum rostrum that September. The urban recovery was real and is credited (the April price cuts appear in the entries), but it was skimmed by the compulsory bond levy and housed at a few square meters per head; a 3 — clearly bad — is what the median household lived. On money: rationing ended in December 1947, so the fifties ran on cash priced by the queue, and the volume's own sources rule — the queue was the price — already refuses to read official inflation as a capitalist number; the 1947 confiscation and the 1957 bond freeze stay in the ledger not as monetarist pedantry but because a state that twice took its citizens' paper evidently agreed the paper mattered. Round one closed there: the floor is doctrine, not oversight. Then the weighting question reopened the middle of it — the addendum below.

The weighting addendum (audit round two). The method lists five lenses — the median household's lived year, contemporaneous consensus, world-relative performance, direction, and statistics as cross-check — and round one, like these ratings generally, let the household lead. The challenge's second round asked the sharper question: what happens in the years where the other four lenses diverge furthest from the kitchen? For the Soviet recovery — the maximum-divergence case among all these economies, an economy whose signature was precisely the gap between what the state achieved and what the household received — the honest answer is to score all five and publish the split. The grid: 1948 · household 3 · consensus 4 · world-relative 4 · direction 5 · statistics 5 → blended 4. 1949 · 3 · 4 · 4 · 5 · 5 → 4. 1950 · 3 · 4 · 4 · 5 · 5 → 4. 1951 · 3 · 4 · 4 · 5 · 5 → 4. 1952 · 3 · 3 · 3 · 3 · 4 → 3 — held, by the grid itself: consensus and direction darken with the gathering purge, which is the evidence this exercise is arithmetic and not a ratchet. 1957 · 5 · 7 · 6 · 6 · 6 → 6 — the one year the consensus and world-standing lenses peak together, Sputnik overhead and the housing program below. 1946 and 1947 were re-run and do not move: in a famine every lens is at the floor — the household starved, the contemporaneous verdict (foreign and, privately, domestic) was of a ruined and hungry country, the world-relative position was among the planet's worst, and the within-year direction was down; no weighting scheme rescues a famine, which is as it should be. The 1955–60 fives were re-blended and confirmed. Net effect: 1948–51 and 1957 up one each, the volume's average from 3.57 to 3.63, and a rule the method now states openly: where the five lenses split this sharply, the volume shows the split rather than letting any one lens absorb the verdict.

The lens definitions, sharpened (audit round three). A further challenge asked why direction, consensus, and statistics sit where they do for 1948–51, and it improved the grid twice while moving no rating. Statistics for 1948–49 rise to 5: the hostile recomputations — Bergson's, the CIA's — confirm recovery at world-class pace, and the cross-check lens should say so. Direction for 1950–51 rises to 5: that lens tracks the year's own trajectory, and the economy's climb stayed steep through the Korea rearmament even as the household's turn was postponed — the household has its own column for that. Consensus holds at 4, by definition: the lens records the era's verdict on the economy as lived, foreign and domestic together. Western alarm at Soviet capacity counts toward it without being identical to it — armies fear power; the lens asks about the year — and the domestic half of the contemporaneous record was official triumph over private exhaustion, with the full when-not-whether consensus belonging to the mid-fifties, which is precisely why 1957 stands at 6. Re-blended under every revision — indeed, under the challenger's preferred sub-scores across the board — 1948–51 remain 4s: the anchors are the household (grain per head below 1913) and the world-relative comparator (Britain sat at 3 with rationing and several times the consumption). One proposed revision is declined on the record: the argument that the famine was sound investment — deprivation as the cheaper alternative to spending reserves on grain — fails first on the export ledger, since the Union shipped grain abroad through the famine while holding stocks and declining to declare the emergency, so the marginal bread that would have saved lives was leaving the country, not funding the bomb; and second on accounting, because these ratings credit returns in the years they arrive — 1949 already prices "the mushroom cloud bought with the village's milk," and 1957's 6 is the dividend paid — while the famine year carries the famine. A rule that scored starvation by its future dividends would acquit the Great Leap Forward, and no such rule enters these ratings.

The evidence run (audit round four). The challenge's final request was the right one: stop re-arguing the held facts and fetch more. The run brought back, on the famine: Ellman's archival study (1.0–1.5 million excess deaths, with state grain stocks held throughout), Ó Gráda's newer count (~900,000 — 300,000 in Ukraine, 100,000 in Moldavia), mortality peaking in early summer 1947, the drought the worst since 1891, the September 1946 decree cutting millions from the ration rolls while raising bread prices, exports itemized by destination — 350,000 tons to Romania in 1946, 600,000 to Czechoslovakia and 900,000 to Poland across the famine — and relief declined, in the government's own framing, to avoid the appearance of weakness. On the recovery: Bergson's recomputed GNP growth of roughly 9.3% a year for 1948–50, prewar industrial output regained by 1948–49 — the sourcing beneath the grid's statistics 5s — and, decisive for the year in dispute, the official Soviet index itself recording industrial labour productivity down 24% in 1946: the reconversion year was a measured collapse before it was a famine. The verdict the evidence supports: 1946 stands at 1 on every lens — output falling, the harvest a third below even 1945's, the ration shock in September, the dying begun, the direction pointing down into the winter; the recovering-while-investing economy the challenge describes is real, and it is 1947–50, misdated by one year. And 1947 stands at 2 on this scale's own consistency test: its recovery half is fully credited in the sub-scores — direction and statistics at 5, the ration abolition a year ahead of Britain's first steps — but the calendar year in which Soviet famine mortality peaked cannot outrank India's 1966, a famine-threat year held to a death toll in the low thousands by accepted foreign aid and rated 1, without the scale meaning nothing. The entries now carry the numbers and the sources, and — in the famine's worst month — the F-1 pile going critical in a Moscow courtyard: the future's down payment and the village's bill, recorded side by side, which is all a rating can honestly do with them.

The adjudication (audit round five). The final challenge braided three arguments, and they divide cleanly. The first is conceded, on this scale' own words: the United States' 1946 — the steepest measured contraction in American history — rates a 5 in its volume, whose stats line calls the −11.6% print "almost meaningless," and the identical reconversion discount belongs to the Soviet −15% and the −24% productivity index; the statistics lens for 1946 rises to 3, the industrial artifact forgiven, the drought's one-third harvest collapse remaining real. The second is declined: redefining 1 as systemic collapse alone would un-rate the Great Leap Forward — a famine conducted by a state in full control, holding stocks, exporting grain, its trajectory positive by plan — and the published anchor reads catastrophe with mass suffering, not loss of control. The third is declined on this audit's own sources: "cynicism, not failure" is a false dichotomy, because distribution is economics — Ellman's paper applies Sen's entitlement framework to precisely this famine, and an economy whose median household's entitlements collapse to zero has failed that household completely, whatever the ledgers say; under control describes the state, not the kolkhoz kitchen. What the conceded point earns, this scale's own template then locates: China's record rates the Leap's onset year (1958) a 2 and the unrelieved death-years 1 — applied to a famine that straddled the winter of 1946–47, with mortality peaking in 1947 beside a genuine second-half recovery, the pair yields 2 and 2: floor-band both years, catastrophe-tier neither alone. 1946 moves from 1 to 2 on the republished grid (household 2 · consensus 2 · world-relative 2 · direction 2 · statistics 3 → 2.2); 1947 holds at 2. The famine floor stands restated, not repealed — famine years live at 1–2, and the unrelieved mass-death years keep the 1, China's three untouched. And the volume's remaining 1s are now exactly the set the challenge asked the digit to mean: 1991, 1992, and 1998 — the years the system itself came apart.

A cross-country note (a definition shared with Saudi Arabia's re-scoring). The world-relative lens was subsequently defined for every economy on this scale as gaining-or-losing ground on the world — motion of position, with level confined to the household column. The 1948–51 grids above are re-read with that lens at 4 (recovering the Union's own position at the pace of Marshall-aided Europe, gaining little on the American frontier); every blend and rating is unchanged.

A closing note on method: this volume rates a planned economy on lived experience precisely because its official statistics measured plan-fulfillment, not welfare — Western estimates anchor the Soviet-era chain, and even they are directional. The stats line prices the Russian republic throughout for continuity; the narrative to 1991 is the Union's. Post-1991 figures are World Bank/Rosstat-consistent, chained to the IMF's 2021 anchor (~$2,195bn in 2026 dollars). PPP comparisons would flatter the Soviet decades (30–40% of US living standards against ~20–25% here) — the distance between what the system produced and what its money could buy abroad. Where the index and the queue disagreed, this volume, as always, priced the queue.

— Compiled July 2026. Sources: Rosstat; Bank of Russia; CIA and Maddison-project reconstructions for the Soviet decades; World Bank WDI; IMF WEO (2025 anchor: $1,829bn). All dollar figures in constant 2026 dollars, chained to that anchor; ratings on the same 1–10 scale applied identically across two dozen major world economies. The volume closes at 2021 — see the note above.

Russia's Economy, Year by Year: 1946–2021

Every year, as it was lived

Sources — this volume's central problem

All figures in 2026 dollars

How the ratings work

1946–1952: Reconstruction by Command

1953–1964: Thaw, Sputnik, and the Corn

1965–1979: From Reform to Oil

1980–1991: The Long Stagnation and the Fall

1992–1999: The Great Depression, Russian Edition

2000–2013: The Petro-Restoration

2014–2021: Fortress and Stagnation

Patterns Worth Noticing