Can Russian data be trusted? A hazard map of official statistics
After the outbreak of the full-scale war in Ukraine, many academics, experts, and journalists found themselves cut off from their usual sources of information about what is happening in Russia. The country's authorities have classified dozens important statistical indicators, and those that remain publicly accessible are increasingly suspected of being unreliable. Russian official data had previously raised questions, but now statistics has become a key part of "information warfare". We have selected 30 important indicators from different areas and assessed them for manipulation.
This study has been conducted by the Cedar (Center for data and research on Russia) team for the Known Unknowns project with the support of the Boris Nemtsov Foundation and the Ideas for Russia initiative. The sources used include monitoring data from the To Be Precise project.
One of the problems created by Russia’s full-scale invasion of Ukraine in 2022 has been the increasingly limited availability and reliability of official Russian statistics. There had been certain challenges concerning the use of official data before, but it was after the outbreak of war that a number of new restrictions were introduced, affecting both availability and quality of government statistics.
For example, Russian authorities have restricted access to official data on foreign trade. Oil and natural gas production figures do not get published anymore. Over the past two years, at least 35 Russian government bodies have closed certain statistical indicators to the public. Almost 500 datasets have disappeared from the official websites of federal agencies alone. In some cases, entire statistical platforms stopped functioning.
Officials tend to justify restrictions on data availability by the need to protect the Russian economy from sanctions. Some experts, however, lean toward other explanations. One reason for the restrictions might be the desire to gloss over the impact that the war has had on Russian society and the state’s resilience.
For researchers and journalists working on Russia-related topics, the increasingly limited availability and reliability of data has become a serious issue. Experts have said as much in the recently published report of the Known Unknowns project, produced by the Ideas for Russia initiative with support from the Boris Nemtsov Foundation for Freedom. The report identified reduced reliability and availability of statistical data as one of the main challenges facing people studying and reporting on Russia since 2022.
This is a particularly acute issue because after the full-scale invasion of Ukraine official statistics have often become the only source of Russia-related data for many scholars, analysts, and journalists who had previously been able to retrieve information through other means. Many of these people can no longer work in Russia, collect data directly, or interact with Russians in person. Opportunities to obtain information about Russia have now either disappeared altogether or been significantly reduced.
Any government data should be treated with caution, even more so in authoritarian regimes. There are many reasons for this. Some distortions in official figures arise naturally — for instance due to data collection difficulties. Other data is distorted intentionally, in order to meet KPIs on paper or compete for bureaucratic resources. Further challenges may occur at the interpretation stage, when an indicator is claimed to reflect something that it does not.
However, it is also important to avoid the idea that all Russian official data is always falsified and that the country has become impenetrable for an outside observer. Russia, unlike North Korea or the Soviet Union, does not seek to close itself off from the outside world completely (at least not yet). Despite wartime censorship, Russia still has plenty of high-quality open data — more than some other countries that are generally more democratic. What matters is how to properly use this data.
The present guide to the do’s and don’ts of official Russian statistics is designed to help with this task. It contains information that will help researchers, journalists, and experts dealing with Russia-related topics to better navigate those state-provided statistical indicators which are still available. It thus raises awareness of distortion risks for particular indicators and helps understand which data can or cannot be trusted, to what extent and for what reasons.
For the first edition of the map, we selected 30 indicators that are used to assess the demographic, social, and economic situation in modern Russia. For each of these indicators, we conducted an expert review, taking into account various aspects, from the completeness of the methodology and the quality of primary data to manipulation susceptibility and the relationship to resource allocation.
The assessment results were then validated by at least one, and more often several, experts on the subject matter.
For each indicator, we noted ‘red flags’ — potential issues that require special attention. These include:
Sharp changes in methodology;
Discrepant data from different sources;
Abnormal jumps in regional data;
Data inaccessibility or publishing issues;
Confirmed cases of indicator manipulation;
Poor record keeping or low quality of primary data;
Sloppy data interpretation.
Based on the results of our study, we assigned reliability scores to all indicators. There are three levels of reliability:
Green — for reliable indicators,
Yellow — for indicators that can be trusted under certain conditions,
Red — for unreliable indicators.
Study results
Red indicators
Twelve indicators on our list received a ‘red’ score. This generally means that an indicator is either highly susceptible to manipulation that makes it unsuitable for direct use, reflects reality extremely poorly, or that the necessary data is no longer published.
This group includes the minimum subsistence rate, poverty rate, migration balance, census results, foreign trade volume, causes of death, crime rate, reforestation, income and property of civil servants, prison population, Russia’s losses in the war with Ukraine, and the number of mobilised soldiers.
The four economic indicators in this category are foreign trade, civil servants’ income, minimum subsistence and poverty rates. For the first two, publication of data virtually ceased after the war began. For the other two, we found dramatic changes in methodology that have led to deliberate underreporting.
For example, the poverty rate is several percentage points lower when calculated using more recent methodology. The reduction of poverty is stated as a high priority for the Russian government and is therefore closely monitored. Since the outbreak of war, the authorities have reported unprecedentedly low poverty rates twice. This stems partly from heavy investment of state funds into the economy and partly from underreporting through methodology manipulation.
Three ‘red’ indicators are demographic: migration balance, census results, and causes of death.
For migration, the accuracy is adversely affected by data collection issues, namely vague criteria which make it impossible to estimate migration accurately. The other two demographic indicators are highly susceptible to manipulation: experts say that the most recent census covered a record low number of Russians, while the way causes of death are reported has been affected by regional and federal authorities’ political interests.
Two further ‘red’ indicators are the crime rate and prison population. In the first case, the use of crime solvency rates as the main indicator for assessing police performance encourages law enforcement to distort the data, which results in unreliable statistics for certain offences. As for the prison population, the indicator itself is well accounted for, but Russia’s Federal Penitentiary Service has stopped publishing it.
Restricting access to data is not merely a characteristic of law enforcement and the military. Civil agencies have also embraced this strategy, as evidenced by the Ministry of Health, among others. This undermines the credibility of statistics, even if the data itself is reliable. Since the invasion of Ukraine, Russian government bodies have generally become more closed to the public. Even traditionally more open agencies, like the Statistics Service, are seeing a decrease in public debate and discussions among experts.
Reforestation and deforestation is also on the ‘red’ list. Here, the issue is not with data unavailability or manipulation, but rather with the indicator itself being purely formal and thus unsuitable for any adequate assessment. For example, it does not account for the species of trees, but includes the natural overgrowth of clearcuts. Other environmental indicators also present certain accounting issues, but this generally does not make them completely unsuitable for analysis.
The two final ‘red’ indicators come from the military: Russia’s losses in the war with Ukraine and the number of mobilised soldiers. These data are extremely unreliable and closed to the public. Russian losses aren’t officially disclosed at all, and any statements about them may be considered a criminal offence by Russian law.
Green indicators
Seven indicators have been marked ‘green’. Their limitations tend to be of a technical nature, are easy to pinpoint, and do not significantly affect the use of data. Almost all indicators in this group involve relatively simple facts recorded in administrative databases (e.g. births or court sentences) and are not subject to strong external pressures.
Indicators in this group include fertility rates, infant mortality, number of court sentences and court cases, number of cancer patients, and RLMS-HSE data.
Minor problems may arise if different data sources are used to arrive at the total figure (e.g. cancer patient population). Regional distortions may also occur: for instance, “birth tourism”, when women from smaller towns and villages travel to Moscow to give birth, results in an inflation of the fertility rate in the capital.
The seventh indicator in this group is the Russia Longitudinal Monitoring Survey conducted by the Higher School of Economics, a large and internationally recognised longitudinal survey of income, consumption, and health of households. It too has limitations (e.g. only presenting data at the nationwide level but not at the regional level), but these limitations are objective in nature. Moreover, the indicator does not directly draw on government statistics.
Yellow indicators
Eleven indicators have been marked ‘yellow’. This means that their statistics can be used, but there are significant drawbacks that need to be kept in mind. Indicators in this group are well suited for certain purposes, but are not always used appropriately.
This group includes income, unemployment rate, GDP, GRP, regional budgets, consumer price index, population size, abortion rate, number of people with HIV, air pollution, and municipal solid waste.
For example, GDP data are likely not highly susceptible to direct manipulation, despite being regularly revised upwards. However, both the GDP calculation methodology and structure of the Russian wartime economy mean that the indicator does not necessarily reflect the general standard of living or the health of Russia’s economy.
Researchers risk incorrect interpretation when looking at unemployment rates, which are at an unprecedented low largely due to labour shortages and specific features of the Russian labour market. Such indicators look favourable at first glance, but are in fact evidence of imbalances in the Russian economy even if the data seems reliable and has not been deliberately distorted.
Another indicator that falls into this group is the number of people living with HIV (PLHIV). Its main limitation is the discrepancy in data from different sources. The Ministry of Health and Rospotrebnadzor have published indicator values that differ by 200,000 to 300,000 people. In addition, there have been problems with record-keeping in recent years, leading to vulnerable groups being undercounted. However, HIV data cannot be classified as completely unreliable as its limitations are not critical.
The reliability of the GDP figures provided by Rosstat depends on how the indicator is used. Due to certain methodological features, the current official GDP values are growing largely due to the spending in the military-industrial complex and the build-up of reserves, which does not necessarily correlate with improved standard of living and a healthy economy.
The methodology is in line with international standards, and obsolete data is regularly updated. Rosstat’s production-based calculation methodology accounts for the production of goods and services but underestimates their usage.
GDP calculations cannot be independently verified since Rosstat does not publish calculation and adjustment details, and not all primary data is available.
Additionally, there are various indirect signs of economic activity that may indicate that official economic data has been overestimated both before and especially after the outbreak of the war, but this evidence isn’t definitive.
The reliability of the GRP as provided by Rosstat depends on the purpose of use. The calculation methodology is in line with international standards and can be used for interregional comparisons or to study the dynamics of individual regions.
However, due to sharp changes in methodology the data before and after 2016 cannot be compared, and the indicator is not suitable for operational monitoring due to long delays in data publication.
One must also take into account its limitations in assessing regional economies: underestimation of interregional activity, overestimation of the GRP in metropolitan regions, and the failure to account for public services and the field of finance.
3. Foreign trade volume
Indicator
Foreign trade volume
Field
Economics
Owner(s)
Federal Statistics Service (Rosstat), Central Bank of Russia, Federal Customs Service of Russia (FCS)
The Federal Customs Service (FCS) collects high-quality detailed data, but restricted access to it in 2022. Nevertheless, some researchers have access to detailed (commodity-by-commodity) data from Russia’s customs statistics through backchannels.
The commodity-by-commodity volume of foreign trade with many countries (e.g. EU countries and the USA) can be estimated relatively accurately from the open mirror data of Russia’s trading partners, as well as from international trade statistics aggregators (such as the UN or the IMF). These stats are less detailed than FCS data, lacking details by region, for instance. Moreover, there are always discrepancies in mirror data due to the differences in customs regulations.
About a third of such mirror data exists in a “grey zone”, making it impossible to obtain a complete picture of Russia’s foreign trade. This includes countries and regions (mostly in Latin America, Asia, and Africa) that have less developed systems of reporting to international organisations.
4. Regional budgets
Indicator
Regional budgets
Field
Economics
Owner(s)
Regional financial authorities, Federal Treasury, Ministry of Finance, Federal Statistics Service (Rosstat)
Links
Reports on the execution of consolidated regional budgets on the website of the Federal Treasury
Appendix to the “Finances of Russia” compendium on Rosstat’s website
Summary reports on the execution of consolidated regional budgets on the website of the Ministry of Finance
Rating
Somewhat reliable
Red flags
Data unavailability
Executive summary
The present indicator accurately reflects the budgetary obligations of Russia’s regions. These commitments are fixed in regional budget laws, so they are not subject to distortion.
However, there is no detailed information on further spending and financial flows in the public domain: i.e. who receives funding and what exactly the money is spent on. This information is only available in aggregated form because regional public funds are transferred to distributing bodies that are not obliged to publish expenditure data.
In April 2022, the Federal Treasury stopped publishing monthly reports on the execution of regional budgets. Summary information continues to be published on the website of the Ministry of Finance, but it contains only aggregated data.
The main drawback of the present indicator is its opaque calculation methodology. Descriptions of how data from dozens of different sources is combined are not publicly available. For instance, we do not know exactly how Rosstat estimates the volume of “other cash receipts”[1], which in some regions accounts for over a third of the total per capita income.
The values published by Rosstat are 1.3-2 times higher than income estimates obtained from sample surveys. This discrepancy arises because Rosstat bases its calculations on reports of companies, government agencies, and banks rather than population surveys. This discrepancy is observed in many countries, but in Russia the lack of a transparent methodology means that there is virtually no expert discussion on how the indicator is calculated. There are almost no studies on this topic.
When using the present indicator, one should pay attention to whether data of the 2021 population census, which changed population estimates both nationally and regionally, were used in calculating the indicator’s value.
6. Minimum subsistence rate
Indicator
Minimum subsistence rate
Field
Economics
Issue
Poverty
Owner(s)
Federal Statistics Service (Rosstat)
Links
Population section, Inequality and Poverty subsection
Calculations of the minimum subsistence rate have become less transparent since 2021 when it became redefined through the share of median per capita income (44.2%). Previously, the indicator was based on commodity bundle prices, which could be estimated independently using Rosstat data on prices of goods. The new methodology uses median income, which can only be calculated by the authorities, meaning the indicator has become uncoupled from the minimum standard of living or income.
Moreover, in 2021 poverty rate calculations stopped using the minimum subsistence rate and instead moved to use the poverty line, which is set separately. These changes in the methodology mean that the values of the present indicator cannot be compared across different years.
In 2023 and 2024, the minimum subsistence rate has been set by the federal budget law.
7. Poverty rate
Indicator
Number of people whose income falls below the poverty line (this indicator coincided with the minimum subsistence rate before 2021)
Field
Economics
Issue
Poverty
Owner(s)
Federal Statistics Service (Rosstat)
Links
Inequality and Poverty section on the Rosstat website, Poverty Line subsection
The indicator can be manipulated and artificially improved if the state simply lowers the official poverty line. The official poverty line in Russia falls in the central part of the income distribution, which means that shifting it even by a few hundred roubles will significantly change the poverty rate. This has led to expert criticism of the indicator.
The indicator’s values for different years are not comparable. Since 1992, the method by which the poverty line is calculated has changed several times. In 2021, the poverty rate was decoupled from the minimum subsistence rate and is now calculated using the poverty line. This new method has lowered the poverty rate, putting it at a record low 9.8% in 2022. However, if calculated using the previous methodology, the figure would be 11.9%.
The indicator reflects only the extent of absolute monetary poverty and does not take into account a person’s own perception (subjective poverty) or the actual level of access to benefits. All that matters within this indicator is whether the average per capita income is above or below the poverty line set by the state.
The “Consumer price indices for goods and services” section at Rosstat’s Statistical Data Showcase
“Consumer price index (December to December)”, “Consumer price indices for goods and services” in the Unified Interdepartmental Statistical Information System (EMISS)
“Prices” annual compendium, which collects information on price changes over a long period of time, including the level and dynamics of prices between 1960 and 1990 in the USSR
- Monthly data on the CPI published in the compendium “The Socio-Economic Situation in Russia”
Monthly reports with information on individual groups of goods and services
Description and characteristics of representative goods whose prices are monitored by Rosstat
Structure of consumer expenditures by goods and services for CPI calculations
Rating
Somewhat reliable
Red flags
Methodological flaws
Executive summary
The methodology of the indicator is generally in line with international standards. However, both Russian and foreign experts highlight several shortcomings. Firstly, since 2013 Rosstat has evaluated the structure of household consumption, which is the basis for the CPI, for two previous years as opposed to one year. This means that during crises, when prices change rapidly, the CPI may lag far behind real price levels. Secondly, the calculation of the CPI does not take into account the quality of goods. Thirdly, the CPI does not take into account changes in the cost of housing (imputed rent). This may lead to a gap between official statistics and people’s perception of inflation.
As in other countries, the Russian CPI may diverge strongly from the inflation rate for certain socio-economic groups as well as from the subjective perception of inflation due to differences in the structure of consumption.
In 2022 and 2023, the gap between the dynamics of average consumer prices and consumer price indices grew. For example, while the average prices of electronics and some other goods were rising, their consumer price index was falling. These discrepancies can be explained by the way Rosstat substitutes goods when monitoring prices.
9. Unemployment rate
Indicator
Unemployment rate
Field
Economics
Owner(s)
Federal Statistics Service (Rosstat)
Links
Labour Resources, Employment and Unemployment section, Employment and Unemployment subsection
Statistical compendium “Labour Force, Employment, and Unemployment in Russia”
Monthly report “The Socio-Economic Situation in Russia”.
Data from macroeconomic surveys of the Bank of Russia
Quarterly statistical bulletin “Results of Sample Surveys of the Labour Force”
Rating
Somewhat reliable
Red flags
Sloppy interpretation
Executive summary
The indicator correctly reflects the unemployment rate as defined by the International Labour Organization (total unemployment rate). However, this indicator should not be confused with registered unemployment, i.e. the number of citizens who have filed for unemployment. In Russia, many job seekers do not file for unemployment due to low unemployment benefits. As a result, registered unemployment is 3-4 times lower than the total unemployment rate.
When interpreting the values of the present indicator, one should take into account the specific features of the Russian labour market, which reacts to negative shocks not so much by increasing unemployment as by reducing wages and switching some workers to part-time employment.
Rosstat’s estimates of the unemployment rate have relatively low accuracy[2] because it only surveys 77,000 people each month, while the results are distributed to the workforce as a whole (75 mln people).
At the same time, Rosstat data aligns well with the dynamics of job search queries. The assessment of alternative indicators of workforce underutilisation (e.g. counting those who are not ready to start work immediately as unemployed) shows that their dynamics does not differ much from the standard indicator.
One should take caution when referencing absolute unemployment figures for 2023 and 2024, because Rosstat now uses data from the 2021 census, which showed an increase in the population, to extend the results of the sample survey to the whole country. Some experts believe that the census overestimated Russia’s population by several million people.
10. Income, expenditure and property of civil servants
A presidential decree and a new law adopted after the invasion of Ukraine suspended the publishing of civil servants’ asset declarations, prohibited their publication of declarations of members of parliament, and exempted war veterans from submitting such declarations. Instead, only anonymised data is published and officials can disclose information about themselves at will. Up-to-date data is virtually inaccessible (although the state still collects it), and the published data for the pre-war years has rather historical value.
Anti-corruption declarations are fairly accurate in reflecting the registered income, expenses, and property of civil servants and their close relatives. However, those who wish to evade proper reporting may resort to offshore schemes or register property as belonging to distant relatives, acquaintances, or third parties. Declarations do not provide an accurate assessment of the value of property.
Healthcare
11. PLHIV
Indicator
Number of people with HIV (PLHIV)
Field
Healthcare
Issue
HIV
Owner(s)
The Federal Service for the Oversight of Consumer Protection and Welfare (Rospotrebnadzor), Ministry of Health (Minzdrav)
The Ministry of Health does not publish its data separately, but it is available upon request and in media reports
Rating
Somewhat reliable
Red flags
Data unavailability
Data collection issues
Discrepant data from different sources
Executive summary
The Ministry of Health and Rospotrebnadzor publish discrepant values of PLHIV, differing by 200,000-300,000 people. At the end of 2022, Rospotrebnadzor reported 1.17 million people with HIV, while the Ministry of Health reported 890,000. In public statements, officials may rely on data from bothagencies.
The indicator shows the registered number of people living with confirmed HIV. However, for epidemiological purposes it is important to also know the estimated number of people with HIV — those whose diagnosis has not been confirmed. Rospotrebnadzor estimates there are about 400,000 cases in addition to the registered ones (i.e. approximately 1.5 million people living with HIV).
Both Rospotrebnadzor and the Ministry of Health have their own database of people living with HIV. The HIV/AIDS Prevention and Management Research Centre publishes aggregate data annually. General data from the Ministry of Health is available only at specialised conferences or on the websites of some (but not all) regional AIDS centres, and the Ministry does not provide data by region on request. Therefore, researchers have to choose which of the two versions to rely on when calculating UNAIDS treatment and prevention cascades and other indicators.
In addition, the number of people living with HIV among vulnerable groups is expected to be undercounted due to changes in the populations that tested for HIV each year.
12. Malignant neoplasm rate
Indicator
Number of patients diagnosed with malignant neoplasms
Field
Healthcare
Issue
Malignant neoplasms
Owner(s)
Ministry of Health (Minzdrav)
Links
Data collections are available on different websites.
The P. Herzen Moscow Oncology Research Institute, for instance, publishes data by region.
Rating
Reliable
Red flags
Data collection issues
Executive summary
The methodology for the calculation of the present indicator is generally in line with the methodology of the International Agency for Research on Cancer and provides a reliable evaluation of the malignant neoplasm issue. Russia includes more types of malignant neoplasms in its calculation than is standard international practice, but this does not interfere with data comparisons (since the indicator can be recalculated excluding certain ICD codes). Data on the number of people with malignant neoplasms is published yearly and sorted by region.
Key problems: the indicator may skew upwards since data on deaths cannot always be updated promptly; patients with haematological disorders may be undercounted; the primary data that enters regional cancer registers may be of poor quality, affecting further indicator quality.
Current records: Demography section, Population Size and Composition subsection
Rating
Somewhat reliable
Red flags
Discrepant data from different sources
Regional distortions
Executive summary
The main source of data is the All-Russian Population Census, which was most recently conducted in 2021. In between censuses, the Federal Statistics Service (Rosstat) annually updates population estimates using current records — data on births, deaths, and migration provided by other government agencies.
Population census data may diverge from such current records. For example, the 2021 census showed there were 1.5 million more people in Russia than the current records did. Rosstat corrects current records data for previous periods using the latest census data.
There are significant discrepancies in the breakdown of individual sex and age groups. The extensive use of administrative sources in the last census resulted in a severe undercount of children under 12 months of age and undercount of children under 15 years of age. The problem is also evident in other age groups — for example, the number of men and women aged 20-24 years is significantly higher in the census data than in current records.
One of the main issues with the present indicator is that it is deliberately overestimated in some regions. For example, in Ingushetia, the last census counted 511,000 inhabitants, although some demographers believe that there are only about 330,000 of them. In Dagestan, over 500,000 residents exist only on paper, demographer Alexei Raksha claims. The 2010 census may have overestimated Russia’s population by 2.3 mln people.
14. Migration balance
Indicator
Migration balance
Field
Demographics
Owner(s)
Federal Statistics Service (Rosstat)
Links
Demography section, Migration subsection on Rosstat’s website
In 2011, there was a change in the calculation methodology, which began including people who had been registered in Russia for over nine months. Previously, the indicator only included migrants registered in the country for over 12 months. This makes it difficult to compare the indicator’s values before and after 2011.
The present indicator is not always up to date. For example, some people who left Russia in 2021 were included in the 2022 statistics due to the automatic renewal of expired documents.
The indicator does not include people without official registration, which makes it unsuitable for studying illegal migration.
Furthermore, the indicator does not exclusively reflect the population of foreign nationals. Russian citizens who have returned to the country are also included in the statistics of international migration, their share accounting for 20% to 40% of total migration growth depending on the year.
Rosstat uses data from the national population census to calculate the present indicator.
The methods of data collection are transparent. Although Rosstat expanded ethnicity categories before the last two censuses (2010 and 2020), this did not significantly affect the population’s ethnic composition. The indicator is suitable for comparative analyses.
One should keep in mind, however, that ethnicity is self-identified and may be influenced by external circumstances and become highly politicised in certain regions. Examples include the republics of Tatarstan and Bashkortostan, where ethnicity being a politicised issue leads to spikes in the number of people identifying as Tatar and Bashkir in large censuses, but not in microcensuses.
The indicator’s main problem comes from the challenges of data collection during the census, especially in large cities, where people more often refuse to open the door to census enumerators. This means that the quality of data for cities with over a million inhabitants is not very high.
16. Abortion rate
Indicator
Abortion rate
Field
Demographics
Owner(s)
Federal Statistics Service (Rosstat), Ministry of Health
Links
Rosstat: Health Care section, Morbidity subsection
Reports by the Ministry of Health: “Main indicators of maternal and child health and the work of the child protection and obstetrics service in Russia”
Rating
Somewhat reliable
Red flags
Methodological shortcomings
Executive summary
Russia is one of the few countries that keep official abortion statistics. Rosstat collects data on terminated pregnancies in private clinics, while state medical institutions report to the Ministry of Health. Unlike in many other countries, Russian abortion statistics include miscarriages, vacuum aspirations, and illegal abortions.
The criteria for what constitutes an abortion have changed several times. For instance, the expansion of the criteria for miscarriages in 2012 led to an increase in the number of spontaneous abortions in the overall statistics.
It is difficult to compare the indicator’s values cross-regionally due to abortions being undercounted in certain regions (such as Moscow, Ingushetia, and Chechnya).
17. Population census
Indicator
Population census
Field
Demographics
Owner(s)
Federal Statistics Service (Rosstat)
Links
Population census data on Rosstat’s official website
Samples with microdata (anonymised data on individuals and households) from the 2002 and 2010 censuses on the IPUMS website
Rating
Unreliable
Red flags
Discrepant data from different sources
Indicator manipulation
Data collection flaws
Executive summary
The quality of primary data collected by Rosstat during the 2021 population census is low. Independent surveys have shown that over 40% of citizens did not take part in it, meaning census enumerators had to enter data about these people by themselves, using administrative sources. Demographers have estimated that this could be the case for at least 25 mln people. Consequently, the 2021 census saw a sharp increase in the proportion of non-responses, since the entries made using administrative sources only indicate sex and age. For example, 17% of Russians over the age of 15 didn’t provide their education level. In 2010, it was only 2.9%, and in 2002 the figure was as low as 1.12%.
There is evidence that the 2021 census overestimated the population in certain regions (such as the North Caucasus and Sevastopol). For example, in Ingushetia, the last census produced a figure of 511,000, although some demographers believe that the republic only has about 330,000 inhabitants. In Dagestan, over 500,000 residents exist only on paper, demographer Alexei Raksha claims.
The data published by Rosstat is highly generalised. The only indicator available for breakdown by individual settlements, municipalities, and districts is population size. All other data is available only at the federal or regional level.
18. Total fertility rate
Indicator
Total fertility rate
Field
Demographics
Owner(s)
Federal Statistics Service (Rosstat)
Links
Demography section, Natural Population Dynamics subsection
This indicator is generally reliable, although it does have certain drawbacks. Its values tend to be undercounted in the regions of the North Caucasus. This is due to the fact that fertility data does not agree with the population data in the Caucasian republics, while the indicator of the average order of births shows the highest intensity of births in Russia.
Given the prevalence of birth tourism in Russia, the indicator may imprecisely reflect the situation in large cities. For example, the value is underestimated in the Saint Petersburg region and overestimated in Moscow, since many women travel to the capital to give birth.
Births are recorded not only by parental residence, but also by maternity hospital. Since 2016, many maternity hospitals have been consolidated and perinatal centres have been created. This led to a decrease of fertility rates in rural municipalities and their rise in larger towns and cities.
19. Infant death rate
Indicator
Infant death rate
Field
Demographics
Owner(s)
Federal Statistics Service (Rosstat), Ministry of Health
Links
Demography section, Natural Population Dynamics subsection
Rosstat’s Demographic Yearbook, section “Data on healthcare for pregnant women, women in labour, and new mothers”
Rating
Reliable
Red flags
Sharp change in methodology
Executive summary
After Russia started using WHO criteria for live births in 2012, experts have evaluated the indicator as reliable. Russia has a well-established system for registering infant mortality, and the data is comparable with international statistics.
Before 2012, babies that died within seven days of birth could be reported as stillbirths, and the weight of newborns in the 1,000-1,499 grammes category could be undercounted, both of which led to the underreporting of infant mortality. In 2012, the weight threshold for a live birth was moved from 1,000 grammes to 500 grammes. Furthermore, there was a change in the criteria for “late abortion”, leading to many late-term pregnancy terminations being reclassified as extremely premature births, meaning that if the infant died it would be included in mortality statistics instead of abortion stats. Thus, indicator values before and after 2012 cannot be compared.
20. Mortality by cause of death
Indicator
Mortality by cause of death
Field
Demographics
Owner(s)
Federal Statistics Service (Rosstat)
Links
Demography section, Natural Population Dynamics subsection on Rosstat’s official website
Russian Fertility and Mortality Database of the New Economic School’s Centre for Demographic Research
Rating
Unreliable
Red flags
Indicator manipulation
Data collection flaws
Executive summary
The present indicator is among the most subject to control and distortion. The manipulation is often political in nature, as was the case with COVID mortality statistics. The main mechanism of distortion is the doctor's choice to favour a particular cause of death when coding it in order to meet the targets set by the government.
For example, during the state fight against circulatory system diseases, some of these deaths were shifted to the “old age” category, which led to the underestimation of cardiovascular deaths and inflated deaths from old age.
Experts note extremely low-quality mortality statistics in 2019 to 2021 for the Sakhalin region due to the high share of unspecified external causes. Such data was also low-quality in most regions of the Far East, Western Siberia, Southern Russia, and the North Caucasus. The quality of statistics was particularly good in the Penza Region, as well as in the regions of the Russian North-West, the Urals, and Southern Siberia.
A concise nomenclature of causes of death based on ICD-10 was introduced in 1999; before that, the 1981 Soviet nomenclature of causes of death had been in use. It has been revised several times. From 1999 to 2005 it had 236 entries. In 2006, it was revised and grew to 239 entries. In 2011, it listed 295 causes of death. The current Rosstat nomenclature lists 306 causes of death (while the ICD includes thousands of codes), making any detailed analysis by cause of death impossible.
The principles of coding may also differ depending on the type of medical statistics: hospital statistics only record a patient’s main disease, while outpatient and polyclinic statistics record all other diagnosed diseases as well, excluding complications of the main disease. Government statistics only record and analyse the immediate cause of death.
Courts and prisons
21. Prison population
Indicator
Prison population
Field
Courts and prisons
Owner(s)
Federal Statistics Service (Rosstat), Federal Penitentiary Service (FPS), The Court Department of the Russian Supreme Court
Links
Brief description of the Russian penitentiary system
Open data of the Federal Penitentiary Service ( most datasets are currently unavailable on the official FPS website but can be found on the To Be Exact platform)
Rosstat’s compendium “Social Status and Standard of Living in the Russian Population”
Rating
Unreliable
Red flags
Data unavailability
Executive summary
The main issue with this indicator is the inaccessibility of source data and the haphazard, non-transparent publishing process which may distort the data.
A presidential decree from 2011 obliges the Federal Penitentiary Service (FPS) to publish statistics concerning its activities. However, the decree does not specify exactly what kind of information must be made available, how often, and to what extent. Moreover, any formal requirements and regulations to this extent are absent from the public domain. There are two different bodies responsible for collecting and publishing such data — the Research Institute of Information Technologies of the Federal Penitentiary Service and, presumably, the press service of the Federal Penitentiary Service for the Moscow Region.
For the most part, the FPS does not document datasets. Neither does it specify whether the figures published are absolute or averages for the reported period. In different versions of the datasets, the values for the same time period may differ. There are discrepancies between the total size of the prison population when calculated using different characteristics. For example, in 2020 the totals in the three main sections (term of imprisonment, number of prisoners and age) differed, even though they should have been the same. The difference, however, is usually insignificant (10-30 people).
In late 2022, the FPS stopped publishing monthly updates. The most recent report dates from January 2023 and only lists the total number of persons held in penal institutions without further detail.
Rosstat publishes FPS-provided data every two years. Rosstat’s publications are more consistent and detailed, including the number of inmates by type of penal colony and their gender composition.
FPS and Rosstat data differ slightly due to the dates of data collection and the level of aggregation. The FPS most likely publishes the average headcount for the last month of the year, while Rosstat publishes the absolute headcount at the end of the year.
22. Sentencing rate
Indicator
Sentencing rate
Field
Courts and prisons
Owner(s)
The Court Department of the Russian Supreme Court and its regional divisions
Datasets of the Judicial Department in the Unified Interdepartmental Statistical Information System (EMISS)
Rating
Reliable
Red flags
Data collection issues
Executive summary
Criminal record statistics includes a group of indicators that describe in great detail the number of persons convicted under particular articles, types of punishment, socio-demographic characteristics of convicted persons, and grounds for exemption from liability.
The present indicator is relatively reliable, although some of the data may be distorted during both collection and aggregation. One of its main drawbacks is that the source data (statistical observation forms) cannot be verified using alternative sources.
Statistical observation forms change regularly, which complicates analysis over time, and a number of data are deleted retroactively. For example, the section “Crimes against military service” disappeared from the report on the number of convicted persons under all offences of the Russian criminal code sometime after 2022. At the court level, data may be deliberately distorted to improve statistics, for example, by underreporting the number of amended sentences of an individual judge. The data published through the Unified Interdepartmental Statistical Information System (EMISS) is not broken down by regions or types of sentences, meaning it is poorly representative of the field.
23. Crime rate
Indicator
Crime rate
Field
Criminal statistics
Owner(s)
Ministry of the Interior (MVD), Prosecutor General’s Office, Federal Statistics Service (Rosstat)
Data from the General Prosecutor's Office on investigative work and number of crimes reported
“Social Status and Standard of Living of the Population of Russia” (Rosstat compendium).
Rating
Unreliable
Red flags
Indicator manipulation
Sloppy interpretation
Executive summary
Indicators in this group are highly susceptible to manipulation. Russia’s current crime control system indirectly encourages MVD employees to distort data. Since the data they provide is used primarily to assess their efficiency, it is in their interest to improve the data by any means, including fraudulent ones. For the state, the key indicator of efficiency is the crime solvency rate. The easiest way to increase its value is to refuse to initiate proceedings in cases that will be difficult to bring to trial. Thus, the statistics often severely undercounts cybercrime and phone fraud as well as minor crimes for which the statute of limitations has expired. Conversely, if a case is easy to bring to trial, the police will find a way to initiate proceedings.
Violent crime statistics are the most reliable, partly because these offences are also recorded by healthcare providers. However, victimisation surveys show that many victims of violent crimes do not go to the police: according to a 2021 survey by the Institute for Law Enforcement Problems, only 61% of assault victims had filed a report, 19% managed to get a criminal case opened, and only 8% got as far as a trial. Up to 70% of unidentified corpses with signs of violent death are not autopsied. The number of registered offences remains 5 to 6 times lower than the number of reported offences and 10-15 times lower than the number of reports filed to the authorities.
Between 2010 and 2012, the Ministry of the Interior said it would abandon its current system and instead assess efficiency through public monitoring, as a 2011 law mandated. But in reality, the solvency rate has remained the main criterion, tied to 9 out of 27 indicators of police performance assessment.
The data is also distorted due to the specifics of data collection. According to the Regulation on crime data recording, compound, repeated, and ongoing crimes as well as crimes committed against several people are counted as episodes of a single case. That is, if three murders are seen as belonging to a single case, they will be recorded as one murder, not three.
The Russian methodology for calculating the homicide rate differs from international standards: the MVD includes attempted murders in the murder category, but not the infliction of grievous bodily harm resulting in death.
The methodology of how investigators fill in reports may also have a negative impact on the quality of the data. Investigators are obliged to fill in boxes with additional characteristics of the crime, but may perceive this as an unnecessary bureaucratic burden and do it sloppily, which affects data quality.
Official statistics do not reflect the level of latent crime, i.e. cases where victims do not report to the police.
24. Court activity
Indicator
Court activity (Number of court cases)
Field
Courts and prisons
Owner(s)
The Court Department of the Russian Supreme Court and its regional divisions
Links
Summary statistics on the work of general jurisdiction federal courts and magistrates on the Judicial Department website
Data of regional judicial departments
Datasets of the Judicial Department in the Unified Interdepartmental Statistical Information System (EMISS)
The present indicator is generally reliable, though some data may be distorted accidentally or deliberately to improve reporting.
The indicator includes a group of statistics that detail the judicial process:
Case details: number of cases reviewed, returned, terminated, types of sentences imposed, etc.
Specific features: timeframes, composition of the panel, jury participation, use of conference calls, etc.
Pre-trial procedures: preventive measures, seizure of property, search warrant requests
Post-trial procedures: remissions of penalty, clemency applications, commutation of punishment, probation, reimbursement of costs.
The indicators separately reflect criminal, administrative, and civil proceedings. Commercial court data is also published separately.
The data can be inconsistent as the individual figures do not always align with the total value. This may point to unintentional or deliberate distortion. There are no alternative sources to verify such data.
Data for this indicator is excessively aggregated. For example, the datasets published by the Judicial Department generally summarise all cases handled at a certain judicial level.
Statistical forms are much more informative, but even here a number of indicators are excessively aggregated and not broken down by offence. This includes the numbers of appeals, petitions, and complaints. The number of satisfied parole or pretrial detention petitions is available, for instance, but not the particular charges.
Moreover, regional data isn’t available in full since some regional offices of the Judicial Department publish statistical forms in a smaller volume or for an incomplete time period, while others provide only summary reports.
War
25. Russian losses in the war with Ukraine
Indicator
Russian losses in the war with Ukraine
Field
Armed conflicts
Owner(s)
Russian Ministry of Defence, government officials, independent media
Links
Rating
Unreliable
Red flags
Data unavailability
Indicator manipulation
Executive summary
The present indicator cannot be reliable because the state does not publish official statistics, and attempts to publicly discuss Russian losses in the war are punishable by law as “spreading false information about the Russian armed forces”.
However, we can distinguish several types of sources for data on deaths and injuries among servicemen involved in the war on the territory of Ukraine:
1. Statements by Russian and Ukrainian defence ministries and officials. Such data are published irregularly and are the least reliable due to their bias.
2. International assessments by third-country officials or foreign media based on their own sources. These assessments correlate with alternatively sourced data better than official statements in Russia and Ukraine.
3. Independent media evaluations. Russian media that are not state-funded bear the legal and economic risks of publishing data on Russian losses, which increases the credibility of their estimates. Journalists generally publish a transparent methodology, though it is often limited by data collection issues: media rely on open sources, potentially underestimating the number of cases.
Figures from different types of sources vary drastically. For example, the most conservative estimate based on obituaries and reports of military deaths in official sources and social media ( conducted by Mediazona and BBC Russian Service) found 44,654 deaths. A study using data from the register of inheritance cases (conducted by Meduza, Mediazona and Dmitry Kobak) confirmed the deaths of 75,000 men by the end of 2023. Reuters, citing an anonymous source in US intelligence, cited 315,000 killed and wounded.
26. Number of mobilised soldiers by region
Indicator
Number of mobilised soldiers by region (excluding the Lugansk and Donetsk ‘People’s Republics’)
According to official statements of the Russian Ministry of Defence, the partial mobilisation of 2022 was supposed to encompass 300,000 people. When its completion was announced in late October 2022 (the mobilisation order was never revoked), official estimates ranged from 302,000 to 318,000 people drafted into the army.
Independent estimates
Mediazona based its initial estimates on the number of expedited marriages (men who received a summons were allowed to marry without the one-month waiting period) and found that 492,000 people had been drafted. After the methodology was refined, the estimate rose to 527,000 men aged 18-49 mobilised between September and October 2022.
The Insider’s alternative estimate based on search query trends puts the figure for the same time period at 300,000-400,000.
iStories used local government statements and data from regional journalists and found that at least 213,000 people had been drafted in 53 regions in the first two weeks of mobilisation.
In 2023, additional estimates using data on ‘frontline’ government benefits identified 100,000 new servicemen. This category, however, also includes contract soldiers and volunteer fighters. Between September 2022 and September 2023, 250,000 servicemen were granted payment breaks on their loans.
The present indicator has a high potential for manipulation by officials since the mobilisation has proven to be a politically sensitive topic, largely due to its coercive nature.
Environment
27. Air pollution volume from stationary sources
Indicator
Air pollution volume from stationary sources
Field
Environment
Owner(s)
Before 2018: Federal Statistics Service (Rosstat)
After 2018: Federal Supervisory Natural Resources Management Service (Rosprirodnadzor)
Links
Data from 2018 are available on the Rosprirodnadzor website
Data for 1992 to 2017 are accessible through the Unified Interdepartmental Statistical Information System (EMISS)
Rating
Somewhat reliable
Red flags
Data collection flaws
Sloppy interpretation
Executive summary
This indicator typically features low-quality data, since companies benefit from underreporting pollution rates to avoid higher emission charges. There may also be unintentional errors due to the complex calculation methodology.
Reduction of emissions is one of the goals of the national project “Environment” and is among criteria for assessing the performance of regional executive authorities. All this makes the indicator highly susceptible to intentional and unintentional distortions.
The indicator more or less accurately reflects the impact of industry on climate change through greenhouse gas emissions, but poorly describes the issue of air pollution. When analysing air pollution, one must take into account not only the gross volume of emissions, but also the geographical location of the entities responsible for them as well as monitoring data on concentrations of harmful substances. A more or less reliable picture can be obtained only when all existing data (as well as independent monitoring data) are used.
28. Reforestation & deforestation
Indicator
Ratio of reforestation/afforestation areas to deforestation areas
The indicator does not accurately reflect the state of the field due to severe methodological flaws.
The indicator does not account for the species of cut and planted forests, but only for the total reforested/deforested area. However, deforestation often affects valuable species, while less valuable species are used in reforesting. In some regions, reforestation also takes into account “natural regeneration”, which does not require significant human involvement and differs little from the natural overgrowth of clearcuts (at the first stage mainly with low-value species). Forests can regenerate on their own without human involvement or investment. Artificial reforestation reduces regeneration timeframes and increases wood quality, which the present indicator does not account for.
Moreover, reforestation and afforestation are long-term processes that require maintenance for at least 20 years. The indicator only takes into account the initial stage of reforestation and afforestation. Without proper maintenance, reforestation usually does not produce significant results, and the current levels of maintenance are subpar.
The areas of dead and felled forests do not take into account certain types of felling (e.g. for the construction of linear structures or geological surveying of mineral resources). The estimates of forest areas lost to wildfires in 2021 made by the Ministry of Natural Resources and the Russian Academy of Sciences differ by a factor of 90.
Furthermore, the indicator does not account for the important functions of old-growth forests (climate regulation, biodiversity conservation, soil carbon emissions that occur when such forests are cut down, etc.), which are not reforested.
All these factors make the present indicator highl research-main__block-content_redy unreliable.
29. Municipal solid waste
Indicator
Volume of municipal solid waste (MSW)
Field
Environment
Owner(s)
Since 2019: Federal Supervisory Natural Resources Management Service (Rosprirodnadzor)
Rosprirodnadzor, which is responsible for aggregation of primary data, does not verify its reliability or make it publicly available. The data is provided primarily by MSW management companies, whose income directly depends on the amount of waste generated.
Expert calculations carried out at pilot sites have shown that some regional MSW management companies have overestimated waste volumes by over 100%. There is no weight control at the overwhelming majority of container sites and waste management facilities.
The presence of distortions is indirectly confirmed by noticeable jumps in time and between reporting units, as well as by significant discrepancies between Rosprirodnadzor data and that from regional MSW management. The statistics does not take into account waste that ended up in unauthorised landfills or was reused or recycled in projects unconnected to regional MSW management facilities.
Methodological problems also include counting the use of waste for energy production (in particular through incineration) as a type of recycling. Producing recycled goods and obtaining energy through incineration are fundamentally different processes that should not be combined in one indicator due to the negative environmental impact of incineration, which releases dust, heavy metals such as mercury and lead, and toxic organic compounds such as dioxins and sulphur dioxide into the atmosphere.
MSW only accounts for about 5% of total waste volume, meaning MSW-related indicators do not fully reflect the issue of waste in any given region.
Other
30. RLMS-HSE
Survey
Russia Longitudinal Monitoring Survey of HSE (RLMS-HSE)
Information on the website of the University of North Carolina
Rating
Reliable
Red flags
Sharp changes in methodology
Executive summary
Since 1992, Russia’s Higher School of Economics, the demographic journal Demoscope, and the University of North Carolina at Chapel Hill have monitored income, consumption, health, and mood of individuals and households in Russia. The sampling model is in line with international best practices, and the survey’s duration is nearly unprecedented, spanning over 30 years.
Studies based on RLMS-HSE indicators have been published in leading international journals.
RLMS indicators have several limitations. Firstly, the data collected is representative at the national level, but not at the level of individual regions. Secondly, the sample has shifted over 30 years of data collection, so post-stratification weights must be used to restore representativeness.
Sometimes data on non-standardised questionnaire blocks is published with a considerable delay. For example, data on 261 variables for the year 2019 was only published in late 2023.
Contact
to request data
We and selected third parties use cookies or similar technologies for technical purposes and, with your consent, for other purposes. Use the “Accept all” button to consent. Use the “Decline all” button to continue without accepting.
