Numbers that lie: How Russia manipulates official statistics

Manipulating mortality

In Russia, the deaths of ordinary citizens are recorded quite meticulously, allowing for a reasonable level of confidence in determining the monthly number of deaths (and births) among Russians. Rosstat publishes this data based on documentation from civil registry offices (ZAGS), with each death certificate detailing the cause of death. This forms the foundation of state statistics.

Based on this information, the effectiveness of programs targeting specific causes of death (such as diseases, road accidents, alcohol and drug overdoses, etc.) is assessed. However, in Russia, a system has been developed over the years where, if necessary, the recorded cause of death on the death certificate does not reflect the actual illness or reason that led to the person's demise.

One of the primary motivations behind such actions is Putin's “May decrees” of 2012. These decrees aimed to significantly reduce the mortality rate by 2018 for tuberculosis, cardiovascular and oncological diseases, as well as decrease the number of deaths from road accidents and among infants. These objectives were achieved. For instance, the mortality rate from tuberculosis experienced a sharp decline, but concurrently, the number of fatal outcomes among people with HIV increased. The “May decrees” did not address the issue of HIV, so the Ministry of Health continues to feign ignorance about the existence of this epidemic in Russia.

Moreover, as revealed by an investigation conducted by RBC, elderly individuals who had been previously passing away due to heart attacks, strokes, and cancer, started dying simply of “old age” after 2012. Additionally, Russian citizens abruptly stopped committing suicide, with deaths now being attributed to vague “undefined” external causes, as highlighted in the study conducted by Novaya Gazeta.

“The statistics are manipulated because the goal is not to obtain the most complete picture of what is happening, but to assess how officials are performing at the local level. For example, after the initial May decrees, which set the goal of reducing mortality from cardiovascular diseases, our respondents often mentioned that the recorded cause of death was incorrect,” says Olga Molyarenko, Associate Professor at the Higher School of Economics (HSE) and Academic Director of the Public Administration program. She conducted field research on the work of municipal bodies in Russia for many years and kept diaries during her expeditions.

“According to the medical statistician, the most common cause of death is cardiovascular diseases (strokes) and cancer. Many young people die from liver cirrhosis, hepatitis, and anemia. When asked about 'inaccurate' data, she explained that their section initially submitted real data to Nizhny Novgorod ('we don't want to adjust the numbers'), but there were instances when orders came down, and they had to change the data. <…> For example, there is a plan to reduce mortality from cardiovascular diseases. If a person died from a stroke but also had diabetes, the cause of death would be listed as diabetes. Instead of diabetes, it could be asthma. <…> The section initially provided accurate data, but then an order came down [to meet the norm]. 'Moscow demands the numbers, and we present them.'“

In 2020, Russia faced an unprecedented mortality rate from COVID-19. However, the authorities took a long time to acknowledge that many of the measures taken were ineffective. Excess mortality, which refers to the number of individuals who would not have died in 2020-2022 if not for the pandemic, turned out to be one of the world’s highest in Russia. Official statistics on coronavirus could not account for a significant portion of this excess mortality. According to calculations by MediaZona, as of June 2022, excess mortality in Russia amounted to nearly 1.2 million people. The Economist estimates it at 1.3 million people as of April 2023.

“The manipulation of statistics during the pandemic is closely linked to the falsification of death causes in previous years. If there hadn't been such falsification before the pandemic, there wouldn't have been a prepared ground for such easy, shameless, and blunt lies during COVID-19,” notes independent demographer Alexey Raksha. He previously worked at Rosstat but was dismissed from the agency in the middle of the epidemic following his Facebook posts about the government downplaying the numbers of COVID-19 cases and deaths.

A glaring example of falsification is the infamous “Sobyanin shelf,” which occurred in August 2020. During that time, Moscow officially reported a daily death toll from COVID-19 ranging from 650 to 700 infected individuals and no more. Remarkably, this mortality rate coincided exactly with the prediction made by Moscow Mayor Sergey Sobyanin. The Novaya Gazeta has documented similar cases, and experts well-versed in statistics consider these “shelves” or “corridors” as clear signs of data manipulation. “Such patterns are statistically impossible, epidemiologically implausible, and devoid of any logical sense,” Alexey Raksha explains.

“If such falsifications remain undetected, the perception of the situation drifts further away from reality. As a result, the authorities are deceived into believing that the situation is improving and that they have full control over all processes. This illusion is reinforced by the positive statistical responses attributed to the decisions being implemented,” Olga Molyarenko points out.

Manipulating inflation

To calculate one of the key indicators in the economy, inflation (or the consumer price index, CPI), Rosstat determines which goods and services Russians purchase most frequently. For example, these could be flour, candy, canned meat, pasta, doctor visits, rent, gasoline, or utilities. Then, the agency measures the prices of these goods and services in all regions of the country. The price of each item is multiplied by its estimated consumption (i.e., the household expenditure share), and all these values are added up to obtain the resulting sum. Inflation is the difference between the sums obtained from month to month or from year to year.

The inflation rate is critically important for budget calculations, as explained by Alexander Isakov, an economist at Bloomberg Economics, in an interview with The Insider:

“The Consumer Price Index (CPI) directly affects the government's obligations to pay pensions. According to the budget for 2023, the Social Security Fund of Russia will allocate approximately 9.5 trillion rubles for pensions, and pensions are indexed annually strictly above the inflation rate. This means that an additional 1% increase in prices in 2023 will require a budget expenditure increase of approximately 95 billion rubles in 2024.”

Payments on government bonds are indexed depending on the CPI, and the Central Bank uses it in their forecasts (e.g. when determining the key interest rate). Businesses also rely on inflation when planning their expenses.

Meanwhile, the calculated index always turns out to be significantly lower than the inflation experienced by the citizens of Russia. For example, in 2022, the official inflation rate was 13.8%, but according to a survey conducted by inFOM, Russians felt it at a level of 17-22%. Moreover, every seventh Russian is certain that prices have gone up more than 30% over the year.

This could be explained by the fact that Russian citizens observe their own “personal inflation,” which reflects the rising costs of their individual market baskets, and everyone's basket is different. However, according to surveys commissioned by the Central Bank, 24% of the country’s households have enough money only for food, while another 8% have even less. Food prices experienced the fastest increase in 2021-2022, which is why inflation for many Russians was significantly higher than the official rate— their consumption basket lacked most of the other goods that did not rise in price as quickly.

The list of goods used to calculate the CPI consists of over 500 items, and it changes quite frequently. The government includes the most frequently purchased goods and services in the index. To determine what Russians buy and how often (based on which purchases are assigned coefficients in the basket), Rosstat conducts extensive household surveys.

However, the basket includes goods that are known to increase in price much slower than others, and many of them are purchased only once in a lifetime or never at all. For example, it may include a coffin, an engagement ring, a trip to the UAE, a fur coat, long-distance phone calls billed per minute, as well as flash drives that rapidly decrease in price from year to year.

Additionally, it is assumed that Russians spend less than 2% of their income on rent, so the cost of rent is not heavily accounted for in the CPI. The cost of living in one's own home is not considered at all—this is one of the significant differences between Rosstat's calculations and those of Western statistical agencies. “Rosstat tracks only expenses that are directly paid by the consumer. Therefore, the share of housing expenses is underestimated—only direct rental expenses are taken into account. Some countries consider imputed rent—thus, they adjust upward the expenses related to living in one's own home,” explains Alexander Isakov. According to him, Rosstat is conducting experiments to evaluate this aspect, and it is possible that the CPI calculation formula will evolve accordingly.

Thus, even though the government does not falsify the inflation rate, it is still not entirely relevant for a large part of the country's population.

Another trick used by Rosstat to understate inflation is to track prices and expenditures only in the largest cities. In each region, 2-4 cities are selected where the agency conducts price measurements.

“Ignoring the expenses of rural residents does not comply with international recommendations <International Conference of Labor Statisticians, IMF, OECD, Eurostat, World Bank — The Insider> and may lead to inflation understatement. Rosstat justifies this approach, firstly, by stating that rural residents in Russia travel to cities for shopping. Secondly, by claiming that expanding the coverage area would require a significant increase in Rosstat's budget, which the government is unlikely to approve,” analysts from the Center for Macroeconomic Research at Sberbank wrote back in 2017.

Since then, the problem has not been resolved, despite experts' proposals to pick up prices from cash registers. “Currently, Rosstat independently tracks most prices by visiting retail outlets. Some countries use more sales data from points-of-sales in retail chains, engage in scraping of prices from websites <automated website data collection — The Insider>, or purchase data from third parties,” Alexander Isakov explains.

An alternative view on inflation in Russia is provided by the research center Romir. Sociologists survey approximately 40,000 people in 220 cities across Russia every month. The Romir consumer basket includes 156 goods as well as utilities services. The latest publicly available data is from November 2022. At that time, sociologists calculated that compared to November of the pre-war year, inflation reached 41%. Meanwhile, Rosstat calculated inflation for November 2022 at only 15% year on year.

Manipulating poverty

In March 2023, Rosstat recorded the lowest level of poverty in its entire observation history. However, real disposable incomes and wages (adjusted for inflation) have declined.

The poverty rate according to the official definition in 2023 is the share of the population living below the poverty line. Thus, for it to decrease, either people must earn more (at the poverty line level) or the country's population must increase. The latter is unlikely since Russia's birth rate has been declining for 8 years, and the adult population is decreasing due to tens of thousands killed, hundreds of thousands mobilized, and hundreds of thousands leaving the country.

In order to reduce the number of poor people amid the economic crisis, the government constantly applies new formulas to determine who is considered poor and who is not. In 2020, Putin instructed to halve the poverty level by 2030. Real incomes of Russians have been declining for years, so amendments were made to the subsistence minimum law to achieve the set goal. These amendments detached the subsistence minimum from food prices (the main source of inflation for the poor). This helped reduce the poverty level, but it was not enough. Until 2025, a “transitional period” was declared, during which each region has been setting its own subsistence minimum using coefficients. However, already in the first year of using the new calculation methodology, on paper the number of poor people decreased by 1.7 million, despite high inflation.

However, this was still not enough, so at the end of 2021, the government introduced another indicator called the “poverty line.” It is now used to calculate the number of poor people in the country. The poverty line is calculated based on the cost of the minimum consumer market basket in the fourth quarter of 2020, multiplied by accumulated inflation. For regions, this indicator is determined only once a year. The Insider calculated the discrepancy between the “poverty line” and the subsistence minimum determined through the food basket and food inflation. The poverty line is always lower than the subsistence minimum, and if in 2022 the number of poor people were calculated using the old method, there would be at least 20 million poor people rather than 14.3 million (a record low value). This would be comparable to 2015, the first year when Russia faced unprecedented sanctions at that time.

Manipulating unemployment and wages

Deputy Prime Minister Tatyana Golikova, responsible for social policy and the labor market in the government, reported that in January 2023, the unemployment rate in Russia reached a historic low of 3.6%. However, this does not mean that everything is fine in the labor market in Russia. To achieve such figures, the government simply applies the necessary formula again.

The figure reported by Golikova represents the ratio of registered unemployed individuals to the total number of working-age population. Therefore, for this figure to be low, it is necessary to have as few people as possible registered with the employment service, while the labor force participation rate should increase. The number of potential workers has been slowly decreasing in 2020-2021 due to COVID-19, and also in 2022-2023 due to the war in Ukraine, which claims the lives of working-age men.

What about registrations with the employment service? Their numbers are also decreasing, and not only due to the reasons mentioned above. For example, the authorities do not take into account individuals for whom it makes no sense to travel to the employment center: regular visits there often cost more than the benefits received. Due to COVID-19, the government allowed remote applications for being listed as unemployed, but visiting the employment center is still required - first to submit the original documents, and then 1-2 times a month to confirm that suitable work has not yet been found. Olga Molyarenko explains that in many regions, one has to spend several hundred rubles on travel to the district center to register with the employment exchange. The futility of such a trip was conveyed to her by the head of one of the rural settlements in the Vologda region.

“Traveling to *** [the district center] costs 350 rubles ($4). 350, and you need to go twice a month. That's 1,400 ($15.5). We also need to eat. And what they give to the unemployed is 900 rubles ($10). What's the point of going twice a month? Things are really bad for us here.”

Many people rely on the income of their relatives or spouses and see no reason to register as unemployed, partly because they can only find low-paying jobs in their own city. Women who resign before giving birth and do not plan to return to work are not considered unemployed either. According to statistics, there were about 10% of such individuals in Russia in 2022.

Nor does the government take into account part-time employment in its statistics. Individuals are considered employed if they work for at least one hour per week. Consequently, despite reporting a decline in wages, Russia did not witness a significant rise in unemployment during crises. However, according to data from Rosstat, part-time employment and temporary layoffs increased notably in various sectors in 2022, particularly in tourism, trade, and aviation. For instance, the number of aviation industry employees on temporary layoffs surged by 2,000%, while those working part-time at the request of their employers rose by 780 times compared to 2021 (from a few dozen to over 10,000 people). In the trade sector, the number of employees on temporary layoffs was five times higher than before the war. This trend can be attributed in part to certain companies that departed Russia but continued to provide partial salaries to their employees, including Inditex (brands like Zara, Bershka, Pull & Bear), McDonald's, Carlsberg, Renault, and Levi's.

In 2022, 4.1 million people were partially employed according to agreements with their employers, while another 187,000 had their working hours reduced without such agreements.

Pavel Zhuravlev, a professor at the Department of Labor Economics and Personnel Management at Plekhanov Russian University of Economics, told The Insider that in the West, part-time workers are often included when determining the unemployment rate, and calculations are based on the number of applications for benefits.

Manipulating crime rate

In January 2023, Prosecutor General Igor Krasnov halted the publication of crime statistics on the specialized portal crimestat.ru. It appears that the nearly decade-long history of open crime statistics in Russia has come to an end. This is a significant loss for researchers, but the published data hardly reflected the true state of crime in Russia, as they contained both omissions and falsified figures.

When the Prosecutor General's Office began publishing open information about registered crimes in 2013, the agency was already aware of at least 3 million cases of concealed crimes annually, approximately the same number as the recorded crimes. Criminology experts classify such cases as latent crime, and it was precisely this type of crime that the Prosecutor General's Office aimed to combat through systematic data collection and the publication of open data. As noted by researchers of law enforcement practices in Russia, Maria Shklyaruk and Dmitry Skugarevsky, latent crime accounts for about 10% of all criminal activity.

In addition, there is also natural latent crime, where law enforcement agencies are unaware of the incident and, therefore, do not register it. For example, Russians often do not report telephone fraud to the police, which has become one of the most common types of crime in the past five years. According to a survey conducted by the Institute for Law Enforcement Problems, after theft, robbery, or assault, approximately two-thirds of Russians turn to law enforcement agencies, but in the case of telephone fraud, only about 40% do so. And if the attempt was unsuccessful, only 10% of victims file a police report.

In Russia, police officers are assessed based on key performance indicators (KPIs) that include the number of crimes registered and solved. Although the Ministry of Internal Affairs (MVD) issued an order based on the Medvedev law “On the Police” to theoretically abolish the “quota system,” quotas still exist in a modified form. Specifically, there is a negative evaluation for an increase in the number of registered serious crimes, while a high rate of identifying suspects and sending cases to court is considered positive. Additionally, law enforcement agencies view an increase in the number of acquittals as a negative factor. This may explain why the chances of receiving an acquittal in Russian courts, as calculated by Novaya Gazeta Europe, are as low as 0.15%.

Furthermore, the impact of the “quota system” is evident in the statistics released by the Prosecutor General's Office. There are instances where certain rare criminal offenses experience a sudden and significant increase. These occurrences are statistically hard to explain, but they can be readily understood as a result of the police being directed to improve the clearance rate for specific offenses. This was observed, for instance, in crimes falling under Article 327 of the Criminal Code of the Russian Federation (related to document forgery) and in raids targeting illegal migrants.

To reduce the negative indicators in their performance reports, police officers can intentionally reduce the number of crime reports they accept. This is especially true for serious crimes or cases that would be difficult to solve. Researchers on law enforcement practices in Russia, Maria Shklyaruk and Dmitry Skugarevsky, have written about this. In their work, they quote an anonymous prosecutor from one of the country's regions:

“How can you make this figure <the clearance rate— The Insider> rise or fall? Well, the first thing is entry control. You go to a police department (sighs) — damn, it's a separate matter — you go to a police department, and they tell you... In the worst case, it's simply closed. Well, it's closed, and that's it. You can't reach them. The officer might be drunk, sleeping, or whatever. They just look to see who’s there. Oh, the hell with them. No one’s opening the door. Well, that’s true for remote rural areas. So, if you were robbed on the street and you run to the police for hell as fast as you can, but the door's closed. That's the simplest scenario. Recording and registration — the simplest scenario. It's a disciplinary violation in the field of recording and registration. The second scenario. They finally open the door. Oh! There's no operative. The operative on duty should receive you. But he’s not there. He’ll be there in the morning or whatever. They make you wait until you give up and say, 'Forget it, I'm going home.' Right? That's the second scenario. The third scenario: you are persistent and wait for the operative. He starts talking to you, starts bullshitting you in different ways. 'I'll write everything down in my little notebook. We'll find out. We'll inform you immediately.' But they won't take any action.”

However, the reason for the inadequate collection of crime data is not only due to “quotas” and intentionally unaccepted crime reports. Due to the reform of the MVD that took place in 2010-2011, the staffing levels of law enforcement agencies in Russia have been reduced. While there used to be an average of 15 local police officers collecting primary information per municipal district in 2010-2011, now there are only three, as explained by Olga Molyarenko. Therefore, they physically may not be able to keep track of recording all crimes.

Furthermore, there is also the issue of police officers manipulating the classification of crimes. Criminologist Alexey Knorre, from the University of Pennsylvania and the Institute for Problems of Law Enforcement, has shed light on such practices. He analyzed crime records by extracting data on the weight of seized heroin and marijuana. The analysis revealed that the quantity of these substances found on defendants was not random but intentionally exceeded the thresholds for “large” or “significant” amounts as defined by the Criminal Code. This suggests that drug dealers and users of heroin or hashish, prior to being apprehended, would always possess precisely the amount of drugs that would result in a more severe charge, while rarely having an amount that would lead to a less severe charge.

Such distribution of confiscated drug weights is statistically unlikely. Therefore, it cannot be ruled out that drugs were planted on the accused, either in reality or on paper, to reach the desired weight. “I do not know of any plausible explanation, unrelated to police corruption, as to why heroin seized from drug users more frequently weighs between 2.5 and 3 grams, and less frequently between 2 and 2.5 grams—since it is after 2.5 grams that the classification of the crime becomes more severe,” commented Knorre on the research.

When crimes are not registered or are selectively recorded, it distorts the overall picture of criminal activity in the country. This leads to the government making incorrect decisions regarding the management of law enforcement agencies, hinders the police from accurately assessing the crime situation in cities or districts, and leaves citizens more vulnerable and distrustful.

The author wishes to thank the experts who chose to remain anonymous for their contribution to the preparation of the material.