The payment was one million rubles. Since the first of November it is one hundred thousand.

Saratov region's government made the change in a decree that, according to Kyiv Post, took effect that Sunday. Meduza dates the decision to late September, and Novaya Gazeta, which linked the document, gives it as decree No. 714-P of 24 September. Novaya Gazeta also reports that the family must file for the payment within six months of the registration of the soldier's death, and that the money is split in equal shares among the family members. The decree touches one sum only, the one-time payment the region makes to the family of a soldier killed in Ukraine, and it turns roughly $12,200 into roughly $1,200. No one in Saratov announced a policy of paying less for the dead. A line in a payment schedule changed, and a number went with it.

Moscow's part did not move. In its report on the regional cuts, Bloomberg describes the federal payment of five million rubles, about $61,000, as the main one and says it is unchanged for every family. That is the figure in the title. It is not the whole federal sum. Radio Free Europe/Radio Liberty puts the one-time federal total at 14.2 million rubles, about $175,000, of which five million is the bonus Putin decreed. The Belarusian newspaper Nasha Niva, reporting on a Bild analysis of messages from soldiers' families, gives more than thirteen million. On top of the federal sum, regions paid their own: one million in Saratov, three million in Kemerovo and Belgorod before they cut.

That is why the cut has to be sized before it is explained. Against a federal total of 14.2 million, a regional million is about seven percent of what a family can receive. A regional three million, added to the federal sum, is about seventeen percent of the combined amount. That is my arithmetic, from the RFE/RL total and the regional figures, and it fixes the scale of everything below: the layer being cut is the smaller layer. Any account that treats it as the whole payment overclaims.

A second document turned up within a week of the Saratov decree taking effect. On 7 November two hacker groups, the Belarusian Cyber Partisans and Silent Crow, said that records they took from Rosstat, Russia's statistics agency, show between 394,000 and 450,000 war-related deaths from February 2022 to August 2026. The Moscow Times, which gave the leak its fullest treatment, checked a random sample of fifty names against open-source records of the Russian dead and could not verify the whole dataset. Rosstat has acknowledged that its systems were attacked. I treat the numbers as a serious, unaudited claim that points the same way as other estimates. The two hacker counts and the British estimate do not measure the same thing: a register of certified deaths, a count of fallen soldiers and a tally that includes civilians and non-combat causes are different quantities, and I do not add them together. Britain's Ministry of Defence said in early September that more than half a million people had died fighting for Russia.

I read the two documents together because they draw one line from opposite sides. The leak is a register of deaths the state has written down. The regional decrees are a schedule of payments that starts when the state agrees a death has happened. The variable both turn on is not the size of the toll. It is the moment, and the word, under which a death is recorded.

Twelve Regions, Four Ways to Pay Less

The count depends on who is counting. Bloomberg's government source says more than ten regions have cut war-related benefits. Kyiv Post counts at least twelve that have cut, postponed or abolished death payments. bne IntelliNews, summarizing the leak, writes at least eleven since 2025. The direction is not in dispute, and the exact number matters less than the forms the cuts take, because they are not all the same cut.

The first form is the plain reduction. Saratov went from one million to one hundred thousand. In Ulyanovsk region the payment fell from one million to 500,000 in February and to 50,000 in April, according to Meduza. Kemerovo cut from three million to one million in February, and Omsk halved its payment the same month. Belgorod's governor, Alexander Shuvaev, cut three million to one million on 30 October, and Smolensk reduced one million to 700,000 in October. Bashkortostan ended up at 100,000. The sources disagree on where it started: Bloomberg's account runs from two million in September 2025, a twentyfold fall, while Meduza records a step from 500,000 to 100,000 in May and Caspianpost has one million, then 500,000, then 100,000. The sources also differ on the day Smolensk signed. I give both versions and rely on neither.

The second form is abolition. On 15 October the governor of Samara region, Vyacheslav Fedorishchev, signed a decree ending the regional payment of one million rubles, which Bloomberg says applies to families who file new applications. Orenburg region, according to Caspianpost, ended its regional payment for deaths that occur after 30 June.

The third form is delay. On 22 October Stavropol Krai postponed until 2027 the payments it had approved for November and December, Kyiv Post reports. Meduza adds that payments are supposed to arrive within five business days of a claim being approved, but authorities can hold them back when the regional labor ministry's account is short of money.

The fourth form is the one I want to separate from the others, because it changes which question the family faces. (I did not read the decrees themselves; every regional figure and rule below is as these outlets report it.) Bloomberg describes Samara's rule as applying to families who file new applications, and Caspianpost describes Orenburg's as applying to deaths after 30 June. I could not open either report in full, and other outlets that cover Orenburg (Ukrinform says the payment was abolished "as of July 1") do not say which date the rule turns on. So take the contrast as reported, not as read in the decrees. As reported, Samara attaches its rule to the day a family files, and Orenburg to the day of death. Both are ways of saying that the old amount stops at some date, and they disagree about which date. Kyiv Post reports that pro-war bloggers say Russian authorities often recognize a soldier's death only in 2026 even when he was killed a year or two earlier. That produces disputes over whether a family gets the higher sum that applied when the soldier died or the lower one introduced since, and Kyiv Post reports that Russia has no consistent legal precedent on the point. One of the pro-war bloggers, Anastasia Kashevarova, has urged relatives to go to court to create one.

A death has two dates, and a region can price either of them.

The Bidding War of 2022 and the Bill of 2026

When Russia launched its full-scale invasion, Bloomberg reports, regional authorities competed to offer soldiers the most generous incentives. The cuts mark the end of that race. The race had a purpose. The war needs fresh soldiers every month, and Western intelligence estimates put recruitment at 25,000 to 35,000 a month, according to bne IntelliNews, which also reports that Dmitry Medvedev, who oversees volunteer recruitment, said regions signed up about 200,000 contract soldiers in the first half of 2026. A researcher quoted by the Moscow Times, Maria Vyushkova, argues that the pull for men in poor regions is a combination: salaries far above local wages, generous benefits for the family if the soldier dies, and for some a way out of prison. Regional governments, the Moscow Times wrote in April, carry high social spending and help fund the recruitment tied to Moscow's war.

The regional treasuries that paid for the race are now short. Seventy-three of Russia's regions ended 2025 in deficit, up from 19 in 2021, according to a Re:Russia analysis of figures after federal transfers. Kyiv Post reports a combined regional deficit of about 1.5 trillion rubles last year and a Sberbank forecast of two trillion this year. A Polish institute that studies Russia, OSW, notes that part of the war's regional costs sits under headings like social policy and healthcare, and that much war spending is classified, which makes the full regional bill impossible to establish. The anonymous official who spoke to Bloomberg put it in two phrases: funding shortages, and insufficient compensation from the federal government.

The federal budget is going the other way. The draft for 2027 puts 17.1 trillion rubles into defense, 26 percent more than the previous plan for that year and the highest figure since the Soviet Union, the Moscow Times reports. Healthcare allocations fall by 5.5 percent, education by 5.4 percent and social programs by six. Moscow has chosen where its own money goes, and the regional budgets, by their own officials' account, are short. I found no document in which Moscow instructs the regions to cut, and I do not claim one exists.

The two levels of the payment now behave differently. Moscow wrote the floor in a presidential decree in 2022, and the floor is unchanged. The regions added the premium, in decrees of their own, at a time when their revenue looked secure. A premium of that kind is the thing a governor can change by signing a page.

The same pattern of timing shows at the federal end, on a smaller scale. On 30 October Meduza, citing the independent outlet Verstka, reported that the Defense Ministry had drafted an order moving service members' December pay to 11 to 21 January, the first such delay in at least eighteen years. The draft exempts members of the Joint Group of Forces fighting in Ukraine. It is a draft, and I do not read it as proof of anything beyond the state shifting when money moves. But two parts of one system were reported in one week as moving a date.

Moscow wrote the floor. The regions paid the premium. The premium is what a governor can change by signing a decree.

Kemerovo Region, Second Summer, One Word

Marina lives in Kemerovo region, in Siberia, and for two summers she has been trying to change one word in a Defense Ministry file. RFE/RL told her story in August. Her son is listed as absent without leave. She started a legal effort in the summer of 2025 to have his status changed to missing or killed, and by the summer of 2026, she told the broadcaster, it was the second year in which she had not heard her son tell her he was alive.

Her son is still listed as absent without leave.

The word matters for a reason RFE/RL spells out. A soldier declared absent without leave stops being paid, and his family loses the payments promised to relatives of the dead or the missing, as well as benefits such as preferential places for children at schools and universities. Women whose husbands have been declared missing, RFE/RL reports, are eligible for monthly payments equal to about $2,000, so a family's position changes with the category the file carries. An activist who advocates for soldiers' rights told RFE/RL that it is much simpler for a commander to declare a man absent without leave than to do what a missing-in-action finding requires: gathering documents, questioning fellow soldiers, notifying relatives. Participants in online groups for relatives of the missing told RFE/RL that notices of absence without leave surged in late 2025.

Another woman told the broadcaster that a commander admitted her husband had been on a combat mission with nowhere to run, and that he was still listed as absent without leave.

Nasha Niva's March report on the Bild analysis describes the same boundary from the other side. According to current rules, it writes, a death triggers more than thirteen million rubles from the federal budget, but payments are only possible after official confirmation of the death. Relatives told the messaging groups Bild analyzed that the paperwork often arrives a year late, and that some certificates say the cause of death is unknown or do not tie the death to combat. Independent Russian publications, Nasha Niva adds, have reported that it is common for the army to treat a soldier as missing when his body cannot be taken from the field, and that families then have to go to court to have him declared dead, because compensation is not paid without it.

None of this is a document from the ministry. It is what relatives, an activist and pro-war bloggers say, in public and to journalists, in 2026. I found no source that says the status is chosen to save money, and this piece does not claim it. What the accounts agree on is smaller and sturdier: a payment that starts at official confirmation makes the confirmation the gate.

The 450,100 Are the Ones the State Wrote Down

The hackers say they took the records from Rosstat's internal demography database, which holds information from death certificates registered between February 2022 and August 2026. Two figures came out of it. The Cyber Partisans counted 393,881 deaths, using a method that includes only military personnel or civilians who died in combat or from the lasting effects of the fighting. Silent Crow counted 450,100, adding civilians killed in strikes and in Ukraine's 2024 incursion into Kursk region, and by Kyiv Post's account some deaths of unspecified cause. The Moscow Times reports that more than 14,000 of the cases describe a soldier's death from non-military causes, such as illness, road accidents or suicide, and that more than 32,000 were deaths from wounds suffered at least a year earlier.

On the leaked figures the war has become the fourth most common cause of death in Russia since 2022. Kyiv Post, summarizing the data, says annual war-related deaths rose from about 19,000 in 2022 to more than 182,000 in 2025. The Moscow Times calculates 6.2 Russian deaths for each square kilometer gained. Janis Kluge, a Russia researcher at the German Institute for International and Security Affairs, called it the most significant leak of Russian casualty data, and told the Moscow Times that the war has never been more deadly than it is now. He also gave the economic scale: the dead are about half a percent of Russia's working-age population of 80 million, and the main significance, he said, is political.

The detail that matters for this piece is what the register leaves out. The Moscow Times notes that the data does not include soldiers missing in action, and that it excludes the 27,000 foreign fighters believed to have been recruited. A register built from death certificates contains the people whose deaths the state has certified, and no others. That is why it is a useful source and an incomplete one. It also means a man listed as absent without leave, like Marina's son, is in neither the leak nor the payment rolls.

The leak counts the dead the state has already agreed to count.

Where the dead come from also matters for who pays. Bne IntelliNews reports that Siberia and the Far East, home to about a sixth of Russia's population, supplied 26.7 percent of the dead by home address. The Moscow Times says the leak supports earlier research that Russia's ethnic republics have carried a disproportionate share. The families whose payments are being cut live mostly in regions that are poorer than Moscow and that carry the larger share of the war's dead.

Where the Two Lines Meet

Set the claims in four layers, because they sit at different distances from the documents. What was said: regional authorities describe the cuts as balancing the budget and removing duplication of other social support, as Bloomberg reports for Samara. What was known: no internal record of the decisions has reached me, and the one inside voice, Bloomberg's anonymous official, points to money. What happened: decrees, in a dozen regions, between January and November. What the evidence supports so far is narrower than a motive.

Four facts can be checked against reporting and fit together. Payments begin only after official confirmation of death, per Nasha Niva's account of the rules. In Saratov the clock for claiming also starts at registration: six months from the day the death is registered, per Novaya Gazeta. Some regions price the old amount against a date, the date of death in Orenburg and the date of application in Samara. And recognition of a death often lags the death by months or a year, according to relatives, pro-war bloggers and RFE/RL's reporting. Put those facts together and a consequence becomes possible without any assumption about intent: when a region lowers its sum, the date on which a death is recognized can decide which sum a family meets. A death in 2025 recognized in November 2026 can fall under the rule of November 2026.

I want to separate three claims here, because they stand at different distances from the evidence. First: an official status decides whether a family can claim at all. The reporting supports that well. Second: a delay in recognition delays the claim and the payment. Also well supported. Third: that delay can bring a lower legal amount into force. That is plausible under a rule like Samara's, which looks at the filing date. It is unverified for the other regions, whose transitional provisions I have not seen, and a decree can protect deaths that occurred before it. The third claim is the original part of this piece, and it is a hypothesis, not a finding.

How many families sit in that gap is a question I cannot answer, and I found no source that can. Whether the lag between a death and its registration saves the state money is, for that reason, an open question, and the piece leaves it open. The leaked dataset might settle part of it. It contains death certificates with causes, so a researcher with the full file could measure how long registration takes after death and how that has changed as regional payments fell. I have not seen that done.

The two documents also share a gap. The leak counts the dead on paper and leaves out the missing. The payment rolls pay the dead on paper and leave out the missing and the undetermined. Silent Crow's higher figure includes some deaths with unspecified cause, and relatives describe certificates that say the cause of death is unknown and block compensation. I cannot tell whether those are the same cases. The full dataset could show it.

That is the shape of the finding. The war's cost is being repriced at the edge of what the state has formally recognized, and the edge is where families are waiting.

The Strongest Case Against This Reading

The strongest counterargument does not dispute any of the facts above. It accepts the decrees, the leak and the delays, and says the conclusion reads too much into them. The regional layer is small, between seven and seventeen percent of what a family can receive, and the federal payments, which are most of the money, are intact. The cuts have an ordinary explanation: in 2022 regions bid against one another, and any such race overshoots. Samara's own account of its decision, balancing the budget and removing duplication, describes exactly that kind of correction. Documentation lags are an old feature of wars with mass casualties. Bodies are not recovered, certificates take time, and Nasha Niva itself reports the army's habit of treating unrecovered soldiers as missing. Recruitment has not visibly suffered: about 200,000 contract soldiers signed in the first half of 2026. Kluge puts the dead at half a percent of the working-age population, which does not threaten the economy.

This counterargument is structurally serious, and much of it is right. The reading offered here does not claim that Russia cannot pay for its war, that soldiers will stop signing, or that the army registers deaths late in order to save money. It claims something narrower: that the cost of the war's dead is being repriced where official recognition meets a regional budget, and that the date of recognition can become part of the price. The counterargument can explain why the regions cut. It does not explain why the amount a family receives should depend on the day a ministry signs a form, and under Samara's filing-date rule it appears to. Whether it does in practice, I have not shown. That is a question of law, and it is the first thing to check.

If regional rules across the board applied the lower amount only to deaths after the decree, with the date of death as the single test, then delayed recognition would cost families nothing and this mechanism would not apply. Orenburg's rule is of that kind. Samara's is not.

What Nobody Outside the Ministry Can Count

On the first of November the Saratov decree started to apply. According to the reporting, the regional payment is now 100,000 rubles where it was one million; I have not seen the decree's transitional rules, so I cannot say that every claim filed after that date falls under the lower amount. A family whose son was killed last year and recognized next month is not obviously covered by the old rule or the new one, and no court has said which.

I do not know how many such families there are. Neither does anyone I could reach. The register that might say, the Rosstat file, is in the hands of two hacker groups and an unverified spreadsheet. The register that pays, the regional labor ministries' accounts, is public only in the decrees. The register that decides, the one in which a man is written down as dead, missing or absent without leave, is public nowhere.

The state pays for the deaths it has counted, and what it counts is, in part, a question of what it must pay. I cannot show that the second half of that sentence is true, and the first half is only the arithmetic of the rules. But a payment that starts at confirmation turns the date of confirmation into a budget variable. In Kemerovo region, a mother was still waiting, that summer, for a word on a form to change.

Evidence Map

Facts, interpretations, forecasts, and disconfirming signals.

Core claim. Moscow's federal death payment is unchanged while more than ten Russian regions have cut, postponed or abolished their own top-ups, and the leaked Rosstat register counts only deaths the state has certified. Because payments start at official confirmation, and recognition often lags death, the date a death is recognized can become part of what a family is owed. This is a hypothesis. The available reporting describes a filing-date rule in Samara; how it applies in law remains to be checked against the decree. The regional layer is the smaller layer; the claim is about who bears the cost at the edge of recognition, not about the war's financing as a whole.

Evidence level. Facts (high, attributed): the Saratov decree and its effect from 1 November; the Samara decree of 15 October; the Stavropol postponement; the federal 5 million payment unchanged (Bloomberg); the 2027 defense figure. Facts (medium): the full list and amounts of regional cuts, where sources disagree on some starting figures and dates; the leak figures, which are unaudited. Interpretation (marked in the text): that the date of recognition can decide which sum a family meets. Open question (marked): whether registration lag saves the state money.

What would confirm this. A dataset analysis showing a rising lag between death and registration as regional payments fell; court rulings that apply the lower sum to deaths that occurred before a decree; further regions adopting application-date rules.

What would disprove this. Evidence that families with delayed recognition always receive the amount in force on the actual date of death; regional rules that use the date of death as the single test everywhere; a court precedent that protects the amount in force at the time of death. A stable registration lag would not disprove the mechanism, only the open question about whether delay saves money.

Watchlist. The number of regions that cut or delay payments; any court ruling on which sum applies; whether the Defense Ministry's draft on December pay is adopted; independent analysis of the full Rosstat file.

Sources. Opened on 9 November: Kyiv Post, Novaya Gazeta (Saratov), Ukrinform, the Russian edition of the Moscow Times, and Caspianpost's Smolensk report. Not opened in full: Bloomberg's 5 November report and RFE/RL's August report, which I rely on as reported by others. The decrees themselves were not read. Linked: the Kyiv Post report of 5 November on the regional cuts, Kyiv Post; Meduza's list of regional cuts of 1 October and its report on the Defense Ministry draft of 30 October; the Moscow Times on the Rosstat leak; and Nasha Niva on relatives' complaints. Where figures diverge, this has been noted.

Jerry van der Laan writes The Manifest Archive, Forensic Narrative Intelligence Writing on the systems beneath power, money, and history. He traces the structures beneath them.

The questions, the editorial judgments and the responsibility for every published claim are mine. AI assists with the research, the drafting and the source checks.