Russia’s demographic secrets: What the Rosstat leak reveals about the war

The leak of Rosstat’s database, which was announced on 1 November, has provided access to one of the most valuable datasets on Russia’s population to appear since the start of the war. This data makes it possible not only to establish a more accurate picture of the scale of Russia’s human losses, but also to assess other demographic trends that the Kremlin has systematically attempted to conceal throughout this period. NEST Centre has obtained the Rosstat database and is preparing a series of publications examining how the war against Ukraine has affected Russia’s demographic situation.
Classified documents
Demographic policy has been one of Vladimir Putin’s main failures during his time in power. Over the past two and a half decades, Russia’s pro-natalist policies have produced modest results, and the state has failed to reverse negative trends in any area of demographic behaviour. Russia has one of the world’s highest divorce rates, one of the largest gender gaps in life expectancy persists, and the birth rate has fallen almost to its lowest level since the early 2000s. During Vladimir Putin’s rule, the country’s population has declined from 146.9 million at the beginning of 2000 to approximately 143.7 million at the beginning of 2025, excluding annexed Crimea.
With the start of the full-scale invasion, the demographic situation deteriorated, and the state’s main priority in the field of statistics became the concealment of the most sensitive data. According to the think tank CEDAR, federal agencies removed almost 600 datasets from their open data sections during the first two and a half years of the war. Demographic data was the first to be affected, as it provided the clearest means of assessing the initial consequences of the war. The state did everything it could to make it impossible to determine mortality levels, even through indirect indicators.
However, the secrecy surrounding this data was lifted after the hacker groups Cyber Partisans and Silent Crow hacked into Rosstat’s electronic systems. They gained access to a unique dataset that, even before the war, had been available only to a limited circle of specialists.
The consequences of the war
The main feature of the Rosstat database attracting researchers’ attention is the detailed information on deaths among Russians between 2021 and August 2026. In the first days after the leak, estimates of military fatalities resulting from the invasion of Ukraine ranged from 393,881 to 450,000. The difference reflects narrow and broad interpretations of the cause-of-death codes. The narrower estimate included International Classification of Diseases codes Y36 (‘Operations of war’) and Y89.1 (‘Sequelae of war operations’), as well as textual references linking deaths to combat operations. The broader estimate also included non-combat deaths among military personnel and records that did not specify a cause of death.1
The published data shows how the composition of fatalities has changed. Among men aged 18–29, the war accounted for around two-thirds of all deaths in the first half of 2025. Among men aged 50–59, the share of war-related deaths rose from less than 1 per cent in the first months of the war to more than 16 per cent in the first half of 2025. In Buryatia, Tuva, and Zabaykalsky Krai, the war accounted for around one-third of deaths among men aged 18–49 between the start of the invasion and mid-2024, compared with around 10 per cent in Moscow and St Petersburg. Access to the detailed database, provided that complete information is available, will also make it possible to assess other characteristics, such as the educational background and ethnicity of those who died, and to indirectly assess the war’s impact on the socio-economic situation in smaller communities. It will be particularly important to analyse the effect of the war on indigenous minority peoples.
Although the published figures and their classification show signs of haste and minor shortcomings, it is already evident that the war against Ukraine has become Russia’s largest and deadliest armed conflict since the Second World War. The number of fatalities is many times greater than the combined losses in all subsequent wars, including the Soviet war in Afghanistan and the two Chechen wars. For Russia’s demographic history, this is a catastrophe with no precedent other than the Second World War.
Bureaucratic chaos
An initial assessment makes it evident that Rosstat itself does not have complete information on all those killed in the war. First, the database does not contain up-to-date information on those missing in action. Second, registration delays are increasing: in the second half of 2023, the median time taken to register a war-related death was around six weeks, but by 2025 it had exceeded four months. Third, the database probably does not fully account for foreign nationals who fought on Russia’s side: if their deaths are not registered by Russian civil registry offices, they may be recorded only in the Ministry of Defence’s internal records.
The data published on the occupied territories is of particular interest. On 12 November 2026, Cyber Partisans released a series of documents concerning these territories, including population figures and official letters from the heads of Rosstat’s territorial offices, prepared at the request of government authorities and intended for internal use only. According to Rosstat’s expert estimates, the total population of Donetsk, Luhansk, and Kherson Oblasts as of 1 January 2025 was 4.1 million. However, even this figure is inflated. Its unusually sharp increase compared with the previous population estimate is particularly notable and reflects a change in methodology: the estimated population of Donetsk increased by 180,000 in one year, while that of Luhansk Oblast rose by 78,300.
These documents confirm that the Russian statistical authorities are unable to collect data directly in the occupied territories and, in some cases, have to rely on expert estimates both of the number of people who have left and of the remaining population. In this respect, they are doing the same work as analysts and researchers who estimate these figures using indirect indicators.
Overall, Rosstat still has no clear understanding of how to organise data collection in these territories. They are excluded entirely from the official population estimates for Russia. Rosstat’s internal data included in the leak contains some information on these territories, but it is limited, and no system of regular data collection has been established.
The published data provides initial insights into Russian demographics and reveals the dual nature of the statistical system: on the one hand, rigorous data collection continues; on the other, the administration of the occupied territories is chaotic, with no clear understanding of how to collect data there. However, the greatest value lies in the data which has yet to be processed.
The extensive mortality dataset will allow researchers and journalists to partially answer the question of how many people have been killed in the war. It will not provide an exact figure, and the actual number is likely to be higher than the one which it indicates. However, the database also opens up opportunities to examine other questions: how has the war affected Russians’ overall quality of life? How much has mortality from other causes increased? How severely has quality of life deteriorated among working-age Russians? Additional access to data about births will also make it possible to examine what has happened to birth rates in the territories most affected by the war, and whether some people have indeed decided against having children while the war continues.