Iconic Figures of the Soviet Soft Budget Constraint

The "soft budget constraint" (a concept by Kornai) in the Soviet system was essentially a systemic phenomenon—enterprise managers, industrial ministries, and local party committees all competed for resources through bargaining (торг)—but several individuals stand out as typical examples:

Trofim Lysenko (Трофим Лысенко) is the classic case. With political backing from Stalin and Khrushchev, he dominated Soviet biology and agriculture for three decades from the 1930s to the 1960s, devouring astronomical amounts of research funding, experimental fields, and institutional positions, but his "Michurinism" was essentially pseudoscience with no substantial results. This is a textbook example of the chain: political backing → unlimited resources → no accountability for output.

Efim Slavsky (Ефим Славский) served for a long time as Minister of Medium Machine Building (actually the nuclear industry ministry), holding the position for nearly thirty years from 1957 to 1986. Minsredmash was one of the ministries with the "softest" budget constraints: secret budgets, lack of oversight, and almost everything they asked for was granted. He built a vast nuclear-industrial city complex (Chelyabinsk-65, Tomsk-7, etc., "closed cities"), but also left behind the Kyshtym disaster and widespread nuclear contamination.

Dmitry Ustinov (Дмитрий Устинов) was in charge of the military-industrial complex (VPK) for a long time, becoming Minister of Defense in 1976. Under his leadership, the military sector was almost free from budget constraints, able to mobilize unlimited resources in steel, electronics, machinery, etc. Many Soviet economic historians consider him a key driver of the structural imbalances of the 1980s.

The BAM (Baikal-Amur Mainline) project under Brezhnev is not an individual, but it fits well as an "iconic" example—it absorbed massive investment and labor, but its economic benefits were far below expectations, a typical case of a "long-term unfinished project" (долгострой) under soft budget constraints.

The "cotton lobby" in Central Asia, especially Uzbekistan's first secretary Sharaf Rashidov (Шараф Рашидов), who obtained large amounts of funds and subsidies from the center by falsifying cotton output, while promoting large-scale irrigation from the Amu Darya and Syr Darya (directly leading to the drying up of the Aral Sea). The "Cotton Case" (Хлопковое дело) purged under Andropov and Gorbachev targeted this mechanism.

A somewhat controversial figure is Sergei Korolev (Сергей Королёв) and his space-missile design bureau. Although he achieved real results, his method of obtaining resources (competing with Chelomei and Yangel design bureaus for projects) is a typical manifestation of what Kornai called "investment hunger" (инвестиционный голод)—each design bureau wanted to launch projects, and none wanted to be cut.

Kornai himself repeatedly emphasized that attributing this to personal morality is wrong—as long as the "paternalistic" state bears the ultimate losses, all rational managers will compete for resources. The above individuals are "iconic" because their scale and duration magnified this systemic logic to the extreme.

Slavsky's life was almost the life of the Soviet nuclear industry itself—from Budyonny's cavalry to the atomic bomb, from Chernobyl to his fall, he lived to 93.

A metallurgical engineer from the cavalry

Efim Pavlovich Slavsky was born in 1898 in what is now Makiivka, Ukraine, into a miner's family. As a teenager he worked as a miner. In 1918, at age 20, he joined the Red Army and was assigned to Budyonny's First Cavalry Army (Первая конная армия). This unit was the cradle of Stalin, Voroshilov, and Budyonny's "old guard"—Slavsky's ability to repeatedly weather political storms later in life was inseparable from this cavalry background.

After the Civil War, he entered the Moscow Institute of Non-Ferrous Metals to study metallurgy, then worked as a plant director and chief engineer at aluminum plants in the Urals and along the Dnieper. During the Great Patriotic War, he served as Deputy People's Commissar of the Non-Ferrous Metals Industry, in charge of aluminum—a key material for aircraft. His hard skills in industrial management were honed over these two decades.

Pulled into the atomic project by Beria

In 1946, Beria and Vannikov were setting up the "First Main Directorate" (Первое Главное Управление, the code organization for the atomic project), needing a tough guy who could manage steel and cement and mobilize resources nationwide. Slavsky was appointed as Vannikov's deputy, in charge of isotope production.

His first task was to go to the "Base 40" in Chelyabinsk Oblast (later Ozyorsk/Chelyabinsk-65, the Mayak complex) to oversee the construction and startup of the Soviet Union's first plutonium production reactor, "Reactor A" ("Annushka"). In June 1948, the reactor reached criticality and produced the first batch of weapons-grade plutonium—the core material for the Soviet Union's first atomic bomb in 1949. Kurchatov was the scientific father; Slavsky was the industrial father.

That generation of nuclear workers paid a terrible radiation price. In the early days, Mayak discharged high-level radioactive waste directly into the Techa River, contaminating dozens of downstream villages. In September 1957, the famous Kyshtym disaster occurred (a waste tank explosion, contaminating an area comparable to early Chernobyl). Slavsky was on site throughout. His attitude toward radiation doses was extremely dismissive; there were private rumors that he personally walked near the accident reactor to "take a look," and that he went to the epicenter of the first atomic bomb test shortly after the explosion—these stories are hard to verify, but they fit the tough-guy persona he cultivated.

Minsredmash: A state within a state

In 1957, Khrushchev appointed Slavsky as Minister of Medium Machine Building (Министерство среднего машиностроения, abbreviated Минсредмаш). The name "Medium Machine Building" was deliberately bland as a cover—it was actually the Soviet Union's entire nuclear industry, from uranium mines to nuclear warheads, from nuclear-powered ships to nuclear power plants, all under one roof.

He held this position for 29 years—through four top leaders: Khrushchev, Brezhnev, Andropov, and Chernenko—making him one of the longest-serving ministers in Soviet history.

Under his rule, Minsredmash grew into a "state within a state":

• At least ten closed cities (ЗАТО), not on ordinary maps, requiring special passes to enter: Arzamas-16 (Sarov, nuclear weapons design bureau), Chelyabinsk-70 (Snezhinsk, another weapons design bureau), Mayak, Tomsk-7 (Seversk), Krasnoyarsk-26 (Zheleznogorsk), Sverdlovsk-44 (Novouralsk, uranium enrichment), etc.;
• Its own hospitals, schools, housing, and food supply systems, with wages and goods far above the Soviet average—sausages, cheese, and imported appliances unavailable in Moscow were available in the "Minsredmash stores" of closed cities;
• Its own construction teams, its own academy (Kurchatov Institute, Dubna Joint Institute for Nuclear Research, etc., were associated), its own environmental monitoring and "independent" accident investigations;
• Budgets were highly secret, and even the State Planning Committee (Госплан) could not see all the details.

This was the extreme form of the Soviet soft budget constraint: because the department bore the "highest national security mission," any resource request was hard to refuse; because of secrecy, any waste, accident, or overspending was hard to audit externally.

"Peaceful Nuclear Explosions"

Under his leadership, from 1965 to 1988, the Soviet Union conducted 124 nuclear explosions "for the national economy"—digging canals, opening oil and gas fields, extinguishing oil well fires, creating underground gas storage, and prospecting. This program (codenamed "Plan 7") was second only to the U.S. Plowshare program (35 tests) in the world. Most of them were not economically viable; they were essentially finding uses for nuclear weapons. This was the strangest byproduct of the soft budget constraint—when a department has almost unlimited explosives, it invents demands to consume them.

The man—this is the best part of the story

Nearly two meters tall, barrel-chested, with a cavalryman's grip that reportedly made others wince. In his eighties, he could still climb the maintenance ladders of reactor buildings. He addressed everyone—including Politburo members—with "ты" (informal "you") instead of "вы" (formal). He spoke bluntly and didn't shy away from cursing.

After Kurchatov's death in 1960, Slavsky became the spiritual leader of the nuclear industry's old guard. When Sakharov publicly opposed nuclear testing in 1968 and was expelled from weapons projects, it was technically Slavsky who executed the decision; but according to Sakharov's memoirs, Slavsky always respected Sakharov's scientific contributions and handled the matter relatively decently—he didn't send him to prison, just made him "leave."

He was awarded the title "Hero of Socialist Labor" three times (the highest civilian honor in the USSR, and three times was almost the maximum), and received ten Orders of Lenin.

After Chernobyl

On April 26, 1986, Reactor 4 at Chernobyl exploded. Strictly speaking, the Chernobyl nuclear power plant was operated by the Ministry of Power and Electrification (Минэнерго), not Minsredmash—but the RBMK-1000 reactor was designed by Minsredmash's Kurchatov Institute.

The cleanup ("ликвидация") after the accident was mainly carried out by Slavsky's Minsredmash—workers from closed cities, Mayak's old-timers experienced in handling radioactive accidents, and the entire nuclear industry's resources were mobilized. However, the political reckoning was inevitable. The design flaws of the RBMK (positive void coefficient, graphite displacer problem) had been known internally at Minsredmash for years but not made public, and Gorbachev's faction seized on this.

In November 1986, at age 88, Slavsky was "asked to retire." On the day he left his ministerial office, he reportedly said only one sentence: "Меня сняли"—"They removed me."

He lived in retirement for nearly five more years, dying of illness in November 1991, less than a month before the dissolution of the Soviet Union. The state within a state he had built outlived him by a few weeks.

How to view this man

If you look at only one side: he was one of the key executors who turned an agricultural country into a nuclear superpower, a spiritual totem for generations of Soviet nuclear workers. To this day, the official biography of Rosatom portrays him almost entirely positively.

If you look at the other side: the Techa River basin remains one of the most radiation-contaminated areas on Earth, and cancer statistics for several Ural oblasts will never be fully published; the system he led exposed hundreds of thousands of people to doses far above safety limits without their knowledge; the budget black box he built under the shield of "national security" is a hallmark of the Soviet economy's distortion.

Kornai said soft budget constraints turn managers into "endlessly demanding children"—Slavsky was not a child; he was the most capable, most loyal, and most dangerous kind of adult that system could produce.

Many people who read Soviet history feel this way, but it's worth thinking about where the charisma comes from, and then the cost.

Why these figures are "dramatic"

This generation—Slavsky, Ustinov, Kurchatov, Korolev, even Lysenko—share several sources of drama:

First, the scale of the era. They experienced the Civil War, the Great Purge, the Great Patriotic War, the atomic bomb, the space race—each person's life story covers almost all the major events of the 20th century. Someone born in 1898 saw the tsar as a child, built the atomic bomb in middle age, and watched the Soviet Union collapse in old age. Such a life has an epic quality.

Second, the unity of power and execution. Slavsky was both a minister and an engineer; he approved budgets and understood reactor physics. He could bang his fist on the Politburo table and then climb onto a reactor roof to inspect it. In today's large organizations, decision-makers and technical experts are almost always separate—this separation eliminates the heroic narrative of "one person bearing a great task alone."

Third, the stakes were real. The cost of their mistakes was not a stock price drop, but tens of thousands of people exposed to radiation, the outcome of wars, the survival of the state. Under that pressure, personalities were compressed into a few extreme traits—this is what makes a literary "character."

Fourth, survivorship bias and narrative curation. We remember those who survived and left a name. The engineers and managers who died in the Great Purge, in the Gulag, or from radiation in the same era—most didn't even leave a name. Tukhachevsky, Pyatakov, Ordzhonikidze... these names are "charismatic" partly because they died early and were frozen in time.

Why they "don't exist" now

But actually, they do exist today, just in different forms, and we naturally lack the aesthetic distance for contemporary figures.

Major projects in developed countries today—like ITER, the LHC, or NASA after the Apollo era—are advanced by committees. Decisions are sliced into hundreds of small decisions, distributed across dozens of institutions, each subject to budget audits, parliamentary oversight, and media scrutiny. This system is more stable and has fewer disasters, but it also means no single individual can accumulate the dramatic weight of "one person equals an entire industrial sector." This is a sign of civilization's progress, but it does thin the narrative.

Contemporary Russia has tried to create such figures—for example, Rogozin in space, or Kiriyenko at Rosatom—but they lack the essential elements: no longer a real epochal scale, executive authority diluted by bureaucratization, and downgraded stakes.

China may be one of the few places with similar conditions. Some industrial founders from the first three decades of reform and opening up—Ren Zhengfei, Wang Xuan, Yu Min, Huang Kun—still show some of these traits, but you'll notice most are already historical figures, and those still alive are in their later years.

Another type of substitute is tech company founders—Musk comes closest, because he spans engineering, decision-making, politics, and risk-taking, and is a single personality. But the fundamental difference from Slavsky is: he risks shareholders' money and his own reputation, not the fate of a nation; the cost of his failure is bankruptcy, not radiation patients.

One more layer worth considering

The "charisma" of that Soviet generation is partly a nostalgia projected from dissatisfaction with our own era—we feel the present world is too mediocre, too procedural, and individuals are insignificant, so we look back and admire an era when individuals could stake everything.

But to be honest: neither you nor I would actually want to live in a closed city ruled by Slavsky, or be assigned to Mayak's waste processing shop, or have our children grow up by the Techa River. What we like is as readers standing on the safe shore of history, watching that kind of personality burn intensely—and the fuel for that burning, most of the time, was not themselves, but others.

The charisma is real, and the cost is real. Both must be remembered.

Your intuition is right—these figures indeed "feel a bit lacking," and the specific differences are worth unpacking.

Yu Min comes closest to "that type" in terms of scientific contribution—the physical breakthrough of the "Yu Min configuration" for the hydrogen bomb is comparable to Teller-Ulam and Sakharov. But notice: when we talk about him, we can only talk about his physics work; we can't talk about his "empire," his personality, or how he banged the table with leaders. He spent his life refusing media attention and biographies; his family emphasized "low profile" even at his funeral. This temperament—clean, restrained, minimizing oneself—is the most admired posture among Chinese intellectuals, but it is aesthetically the opposite direction from Slavsky's "two-meter-tall, cavalry-accented, addressing Politburo members with ты" persona.

Wang Xuan and Huang Kun are even more obvious: they were pure scholars. Wang Xuan worked on laser phototypesetting, Huang Kun on solid-state physics theory (Born-Huang equations), each focusing on a single technology or theory for a lifetime. Their vertical depth can compare with Soviet scientists, but their horizontal span is incomparable—they neither had the opportunity nor the temperament to become ministers.

Ren Zhengfei has horizontal span—Huawei is in some sense a "state within a state," with its own research institutes, supply chain, and overseas network—and he himself has real personality strength. Among all contemporary Chinese figures, he is the closest to Slavsky. But he is a private entrepreneur, which differs in three fundamental ways: he uses shareholders' and employees' capital, not the state's coercive power; the cost of his failure is bankruptcy, not radiation patients; his opponents are markets, not geopolitics. Without this life-and-death stake, the narrative density decreases.

---

The Chinese figures that structurally correspond to Slavsky are actually a different group.

The closest is Qian Xuesen: Caltech-trained, under house arrest in the US for five years before being repatriated, in charge of missiles and space for nearly thirty years, with real political weight (directly writing to Mao and Zhou to participate in decision-making), and later a controversial period (defending "10,000 jin per mu" with physics). But he was ultimately a scientist, not a minister—he commanded a research institute system, not a full industrial ministry.

Structurally more corresponding is Nie Rongzhen: a founding marshal, Vice Premier in charge of the "Two Bombs, One Satellite" program. A military man turned industrial manager—this career line almost exactly matches Slavsky (cavalry → metallurgy → nuclear industry). But in Chinese narratives, he is primarily remembered as a marshal, not an industrial magnate.

Most similar and most underrated might be Yu Qiuli—the "one-armed general" who lost an arm during the Long March, became Minister of Petroleum in 1958 to develop Daqing, and later headed the State Planning Commission. His personality strength (iron fist, hot temper, on-site at oil fields), way of mobilizing resources (Daqing campaign mobilized the whole country), and length of tenure—very Slavsky. But he is almost blank in public memory, because the oil narrative was absorbed by the collectivist symbol (Wang Jinxi).

---

So what is lacking?

I think it's three layers.

First, a rupture. China was producing such figures in the 1950s—on the lineage of Gao Gang, Peng Dehuai, Tao Zhu, He Long, Liu Shaoqi, there were quite a few "red baron" type cadres with both military merit and industrial management experience. But this lineage was systematically purged by the Anti-Rightist Campaign, Lushan Conference, and Cultural Revolution. When the 1980s reorganized, Deng Xiaoping's clear intention was not to have that kind of person again: abolish life tenure, set retirement ages, emphasize collective leadership, and technocratize. The result was a significant increase in governance stability, but the "barons" disappeared. This was a deliberate institutional choice.

Second, aesthetics. Russian narratives favor rough, hard-drinking, violent, flesh-and-blood heroes (Chapaev, Zhukov, Gagarin, Vysotsky all fit this tone). Chinese mainstream narratives—whether Confucian or Party propaganda—favor restrained, dedicated, self-effacing servants (Zhuge Liang, Jiao Yulu, the Yu Min model). The same person would be told as a completely different personality in the two narrative systems. If Slavsky were Chinese, we would remember the details of him organizing emergency response at Mayak, not the details of him banging the table with Brezhnev—the latter would be edited out.

Third, a real gap that must be honestly acknowledged. That Soviet generation turned an agricultural country into a nuclear and space superpower, with stakes at the level of national survival. China's "Two Bombs, One Satellite" is real, but the story after the 1980s is peaceful, commercial, and incremental—it inherently lacks the density of stakes at Slavsky's level. Ren Zhengfei is already the most dramatic figure possible under contemporary conditions, but contemporary conditions themselves have been tamed.

---

This is not necessarily a bad thing. The system that produced Slavsky also produced Mayak, Kyshtym, and Chernobyl. Our nostalgia for that kind of personality is probably nostalgia for the stakes themselves—for an era when one person's decision could change the shape of the world—rather than nostalgia for that kind of person.

And why that era disappeared, you've already indirectly answered: it disappeared partly because humanity survived and learned not to play such big stakes.

Two questions to answer separately, as they are in completely different directions.

---

On Huawei and "government relations company"

We need to distinguish several things.

Did Huawei grow up on government contracts?—Basically no, at least not as a core driver. Huawei started in 1987 as a distributor of switches, and in the 1990s opened up the market with its C&C08 digital switch. Its real foothold was encircling the cities from the countryside—selling to prefectural and municipal telecom bureaus, because it couldn't break into the territories of Ericsson and Siemens in Beijing, Shanghai, and Guangzhou. Its early customers were telecom operators (China Telecom, China Mobile), not government finance. Operators were state-owned, but their procurement was commercial, based on technical specs and prices. Huawei succeeded because it truly achieved competitive performance and price in switches, base stations, optical transmission, etc., compared to Western rivals. This thread is relatively clear from Ren Zhengfei's early speeches and internal Huawei materials.

Does Huawei have government relations? Of course. Any company that grows big in China must have them; it's basic ecology. But "having government relations" and "living off government relations" are two different things. The criterion is: if all government-background advantages disappeared overnight, could the company still survive? The revenue composition of Huawei's carrier business, overseas business, and handset business is mostly the result of market competition—its market share in Europe, the Middle East, Africa, and Southeast Asia cannot rely on Chinese government relations. After the 2019 US sanctions, what hurt Huawei most was not losing government orders, but losing ARM architecture licenses, Google GMS, and TSMC foundry services—these are market contracts, not something the government could compensate for.

After the US sanctions, the situation changed. This is a fair statement. Projects like HarmonyOS, EulerOS, HiSilicon, and the Ascend AI chips—these "backup plans turned mainstream"—partly rely on state investment and state markets: government-enterprise customers, operators, and state-owned enterprises were semi-overfly "suggested" to buy domestic solutions. So after 2020, Huawei's government-enterprise business and domestic substitution component became heavier than before 2019. This was not Huawei's active choice; it was a survival path forced by US sanctions, but the objective result is that Huawei's dependence on national policy environment has indeed increased.

---

On Huawei's car business—your intuition has merit, but needs refinement

Huawei's automotive business has three models, with completely different economic natures:

First: Component supplier (Bosch model)—selling electric drives, LiDAR, and in-vehicle chips to all carmakers. This is a legitimate market business, selling to BYD, Changan, Chery, whoever wants to buy. The logic here is fine.

Second: HI model (Huawei Inside)—Huawei provides the full suite of intelligent driving and intelligent cockpit solutions, while the carmaker builds the car and puts its own badge. This is the model used with Changan's Avatr and BAIC's Arcfox. This is also basically market-oriented, although the partner carmakers include state-owned enterprises.

Third: Harmony Intelligent Mobility (formerly "Smart Selection")—Huawei deeply participates in product definition, design, marketing, and channels, only stopping short of directly putting the Huawei logo on the car. AITO (Seres), Luxeed (Chery), Stelato (BAIC), Maextro (JAC), Shangjie (SAIC)—this is the model your intuition points to.

Look closely at this list: Seres (Chongqing), Chery (Wuhu, Anhui state capital), BAIC (Beijing), JAC (Hefei/Anhui), SAIC (Shanghai). Behind each partner stands a local government, and each has been pouring money into new energy vehicle industry clusters in recent years.

There is indeed an element of "getting government resources from various places" here, but it cannot be simplified to "subsidy fraud":

Local governments want employment, tax revenue, industry chains, and GDP. New energy vehicle manufacturing is a core industry; one assembly plant can drive dozens of upstream and downstream parts companies to set up locally, which is a qualitative change for a city's finances and employment. Local governments are willing to pay (land, tax rebates, equity investment in production lines) in exchange for Huawei bringing its technology, brand, and channels into the area—for local governments, this is a win—as long as the cars can actually be sold.

Huawei wants scale, data, and iteration. Intelligent driving is data-driven; without hundreds of thousands of cars on the road transmitting data back, algorithm iteration cannot progress. Since Huawei does not build cars itself, it needs N carmakers to build and sell cars for it. The multi-partner model of Harmony Intelligent Mobility is essentially using local state capital and capacity to build cars that run Huawei's system.

So the description "getting government money from various places" is partially accurate, but a more accurate description is: Huawei uses its technology brand as leverage to mobilize local governments' willingness to invest in industry, integrating multiple local state-owned carmakers into an ecosystem defined by Huawei's products. This is very different from "subsidy fraud"—cars like the AITO M9 are actually selling; sales went up in 2024, Seres' stock price soared accordingly; this is a real business with products and customers, just that the starting capital came from local governments.

Whether it's sustainable is another question. If a particular model doesn't sell, the local government's investment is lost—this risk is real. But it's still different from the logic of "peaceful nuclear explosions" that purely consumed resources: at least it's a real gamble facing a real market.

---

On the economics of peaceful nuclear explosions

My earlier statement that "the vast majority were not economically viable" was too harsh and needs correction.

The Soviet Union's 124 PNEs (peaceful nuclear explosions) can be categorized by use, with large differences in economic viability:

Some had economic merit:

Seismic exploration (a significant proportion)—using nuclear explosions instead of large chemical charges for deep seismic wave exploration to find oil, gas, and minerals. This was economically reasonable: nuclear explosions have much higher energy density than chemical explosives, can excite deeper and clearer reflected waves. The Soviet Union discovered some oil and gas fields this way; the cost per discovery might indeed have been favorable. The US Plowshare program (Gnome, Gasbuggy) also followed this idea.

Extinguishing oil and gas well fires—these were truly practical solutions. In 1966, the Urtabulak gas well in Uzbekistan had been burning for three years, uncontrollable (burning enough gas daily to supply all of Moscow), and a 30-kiloton nuclear device was used to seal the well underground—it actually worked. Later, the same method was used for the Pamuk gas well and a few others. The economic value here is clearly calculable—the saved gas losses far exceeded the cost of the nuclear device.

Underground storage for gas/oil—nuclear explosions create cavities in salt domes for storage, theoretically cheaper than conventional excavation. The Soviet Union built several and operated them. This category had moderate economic viability.

Almost purely "finding uses":

River diversion and canal digging—the most famous was the 1971 "Taiga" test, using three 15-kiloton nuclear devices to excavate a section of the Pechora-Kama Canal (intended to divert north-flowing Pechora River water south). The canal section was dug, but radiation contamination made the entire project impossible to continue; the 150-km diversion plan remained forever stuck at the 3.5-km contaminated ditch. This is a classic case of "when you have a hammer, everything looks like a nail."

Mining—using nuclear explosions to fracture ore bodies for conventional mining. Tried several times, but economic viability was always inferior to conventional blasting due to the high cost of radioactive contamination management.

Research for its own sake—later tests had increasingly vague objectives, essentially Minsredmash maintaining its explosive throughput and personnel training.

How to evaluate overall?

If we only count direct engineering costs, a considerable portion of PNEs were profitable—especially seismic exploration and well fire extinguishing. The Soviet official accounts showed a positive balance.

But if we count full costs—including surface radiation monitoring, compensation for evacuated residents (often none), long-term increased cancer incidence, groundwater contamination, abandoned agricultural land, and cleanup costs still being spent decades later—most projects' real accounts are negative. The problem is that under the Soviet system, these costs were simply not accounted for; Minsredmash's internal books and actual social costs were two different sets of books, which is the core mechanism of the soft budget constraint: making part of the cost invisible.

So a more precise statement is: A few PNE projects were economical even under full-cost accounting (well fire extinguishing, some seismic exploration); most projects appeared economical on Minsredmash's books but were loss-making on society's books; and a significant proportion were, from any perspective, just to consume capacity and maintain the team.

My earlier statement that "the vast majority were not economically viable" lumped the latter two categories together, which was indeed insufficiently detailed. Thank you for pressing on this point.