SOCIETY
Do Russian Official Statistics Give An Idea Of What’s Going On In The Country’s Economy?
September 20, 2023
Tatiana Rybakova
Journalist and writer
Tatiana Rybakova writes about omissions and distortions in Russian official economic statistics and asks economists for their views on what problems this creates for analyzing the state of the economy.
Rosstat, the Federal Service for State Statistics. Source: Wiki Commons
There is a joke in Russia: “there are lies, there are big lies, there are statistics and then there is Rosstat.” This well reflects the attitude toward the data put out by government agencies. However, lately even observers accustomed to all this have grown worried. After the start of the war, Russian agencies began closing some of the data and providing some in an aggregated form, while the accuracy of what remains in the public domain is increasingly being questioned.
Higher and higher
In August, Rosstat released revised data on industrial growth for 2022. It turned out that industrial output, which had peaked after the start of the war, not only did not decline year-over-year but managed to grow 0.6%, while a fall of 8.0-10.0% had been initially expected.
Manufacturing also did pretty well last year: instead of the 1.3% decline initially penciled in by Rosstat in April 2023, growth of 0.3% was announced in August. Even the extractive industries, where Russian gas saw critical changes and oil and coal flows were redirected with difficulty from west to east, improved on the 0.8% growth announced in April, with the final estimate for 2022 coming in at 1.3% growth.
Revising estimates is a common thing. It happens with statistical indicators all over the world, as numbers arrive after things are calculated, errors are found, etc. But Rosstat’s revisions are notable for the fact that besides the numbers being adjusted, the calculation methodology regularly changes (for example, industrial data for 2022 and January-June 2023 are being calculated using a new methodology in line with Rosstat regulations dated August 18, 2020).
Still, the biggest suspicion is that the revision always goes in the direction desired by the authorities. “Another resounding revision of statistics from Rosstat,” writes the MMI Telegram channel, co-founded by the current director of the Central Bank’s Monetary Policy Department, Kirill Tremasov. It notes with irony: “It seems that without the help of Rosstat it would be hard to return to a growth trajectory.”
Higher and higher
In August, Rosstat released revised data on industrial growth for 2022. It turned out that industrial output, which had peaked after the start of the war, not only did not decline year-over-year but managed to grow 0.6%, while a fall of 8.0-10.0% had been initially expected.
Manufacturing also did pretty well last year: instead of the 1.3% decline initially penciled in by Rosstat in April 2023, growth of 0.3% was announced in August. Even the extractive industries, where Russian gas saw critical changes and oil and coal flows were redirected with difficulty from west to east, improved on the 0.8% growth announced in April, with the final estimate for 2022 coming in at 1.3% growth.
Revising estimates is a common thing. It happens with statistical indicators all over the world, as numbers arrive after things are calculated, errors are found, etc. But Rosstat’s revisions are notable for the fact that besides the numbers being adjusted, the calculation methodology regularly changes (for example, industrial data for 2022 and January-June 2023 are being calculated using a new methodology in line with Rosstat regulations dated August 18, 2020).
Still, the biggest suspicion is that the revision always goes in the direction desired by the authorities. “Another resounding revision of statistics from Rosstat,” writes the MMI Telegram channel, co-founded by the current director of the Central Bank’s Monetary Policy Department, Kirill Tremasov. It notes with irony: “It seems that without the help of Rosstat it would be hard to return to a growth trajectory.”
“Rosstat’s help is needed not only to communicate that the economy is growing again, but also to convince citizens that their well-being is also improving.”
Thus, Rosstat managed, with the help of a bias in the sample of enterprises, to overestimate the growth of real wages by about two percentage points, as the Kholodny Raschet Telegram channel of Bloomberg Economics economist Alexander Isakov calls attention to. According to Rosstat, real wages increased 10.5% year-over-year for June 2023, while their accumulated monthly growth, according to Rosstat, amounted to only 8.8% for June 2022-June 2023.
The slight discrepancy in relative figures should not mislead. For example, Rosstat revised its estimate of the number of poor people in the country: in 2022 it was not 10.5% of the population, but 9.8%, the lowest level ever observed. Meanwhile, as the media project Esli Byt Tochnym (“to be precise”) calculated, thanks to Rosstat more than 3 million people dropped out of the poor category (and therefore as potential recipients of social support. This is because Rosstat changed its calculation methodology: previously (the previous methodology was also very biased toward underestimating poverty) the number of poor people would have been 12.0%. And if the European approach were used, whereby poverty is determined by the number of citizens with an income below the median, then the figure in Russia would rise to 36.5%, or 53 million people, with more than half of the residents of every fourth region in the country considered poor and more than two thirds in seven regions.
‘Do something nasty – make the bosses happy’
This is how one interviewed expert commented on the actions of Rosstat. However, the objections do not end with Rosstat; the Central Bank and the Ministry of Finance have also surprised with their inflation data, and the Ministry of Economic Development with its GDP forecasts. The data on Russia’s gold and foreign exchange reserves has become a closely guarded secret, and this in particular prevents experts from understanding what is behind the current weakening of the ruble – a banal shortage of convertible currencies or less global and shorter-term reasons.
Allowing companies to withhold full financial statements not only hurts Russian investors, who are left guessing about the real state of the companies whose shares they buy, but also raises serious doubts about whether the Russian authorities have a real understanding of the state of the national economy.
In the minds of the population and the authorities, a picture is being crafted of an economy “rising from its knees,” growing despite all the machinations of the West, notes economist Sergei Petrov (name changed). “The very ethics of distorting statistics is very philosophically similar to… the Soviet era: back then there was a whole hierarchy of who could distort what and by how much” he notes.
A part of the distortions is rational: this concerns statistics on real disposable incomes and in turn the level of poverty. “It affects the level of federal budget spending on social support,” explains Petrov. Of course, from the standpoint of tightening the budget, which in the first half of the year saw a massive deficit, this looks rational, but it is hardly rational from the standpoint of preserving and growing human capital.
Still, for experts analyzing the Russian economy, omissions are even worse than distortions. “In the last year and a half, the main problem is not the worsening quality of statistics, but the inaccessibility of a significant part of the data. For example, in terms of foreign trade and the balance of payments, the Central Bank does not report indicators at all or publishes only a small part of them. There are problems with the data on the budget, with customs statistics, etc.,” the well-known economist explained on condition of anonymity.
Who is lying the most?
The slight discrepancy in relative figures should not mislead. For example, Rosstat revised its estimate of the number of poor people in the country: in 2022 it was not 10.5% of the population, but 9.8%, the lowest level ever observed. Meanwhile, as the media project Esli Byt Tochnym (“to be precise”) calculated, thanks to Rosstat more than 3 million people dropped out of the poor category (and therefore as potential recipients of social support. This is because Rosstat changed its calculation methodology: previously (the previous methodology was also very biased toward underestimating poverty) the number of poor people would have been 12.0%. And if the European approach were used, whereby poverty is determined by the number of citizens with an income below the median, then the figure in Russia would rise to 36.5%, or 53 million people, with more than half of the residents of every fourth region in the country considered poor and more than two thirds in seven regions.
‘Do something nasty – make the bosses happy’
This is how one interviewed expert commented on the actions of Rosstat. However, the objections do not end with Rosstat; the Central Bank and the Ministry of Finance have also surprised with their inflation data, and the Ministry of Economic Development with its GDP forecasts. The data on Russia’s gold and foreign exchange reserves has become a closely guarded secret, and this in particular prevents experts from understanding what is behind the current weakening of the ruble – a banal shortage of convertible currencies or less global and shorter-term reasons.
Allowing companies to withhold full financial statements not only hurts Russian investors, who are left guessing about the real state of the companies whose shares they buy, but also raises serious doubts about whether the Russian authorities have a real understanding of the state of the national economy.
In the minds of the population and the authorities, a picture is being crafted of an economy “rising from its knees,” growing despite all the machinations of the West, notes economist Sergei Petrov (name changed). “The very ethics of distorting statistics is very philosophically similar to… the Soviet era: back then there was a whole hierarchy of who could distort what and by how much” he notes.
A part of the distortions is rational: this concerns statistics on real disposable incomes and in turn the level of poverty. “It affects the level of federal budget spending on social support,” explains Petrov. Of course, from the standpoint of tightening the budget, which in the first half of the year saw a massive deficit, this looks rational, but it is hardly rational from the standpoint of preserving and growing human capital.
Still, for experts analyzing the Russian economy, omissions are even worse than distortions. “In the last year and a half, the main problem is not the worsening quality of statistics, but the inaccessibility of a significant part of the data. For example, in terms of foreign trade and the balance of payments, the Central Bank does not report indicators at all or publishes only a small part of them. There are problems with the data on the budget, with customs statistics, etc.,” the well-known economist explained on condition of anonymity.
Who is lying the most?
“Experts surveyed are unanimous – Rosstat data is the least reliable.”
“The hierarchy of reliability of Russian statistics is very simple: the Central Bank has the least distortions, the Ministry of Finance is in second place, then everyone else. The reasons for this hierarchy are also simple: the Central Bank and the Ministry of Finance are responsible for [ensuring] stability, and they need reliable information to perform their own functions.
Everyone else is much more dependent on the [political] situation and, as a result, are simply forced to make distortions to please their superiors,” says Petrov. In his opinion, the macroeconomic data from the Central Bank and the Ministry of Finance can still be trusted. “For now,” he adds.
However, he says, it is hardly possible to directly compare the quality of statistics from different sources. “For example, Rosstat or the Central Bank obtain most of their estimates by adding up sample data, while the Ministry of Finance gives exact figures for spending, expenditures, government borrowing, etc.,” he explains. Whereas in past years economists had more doubts about Rosstat’s GDP estimates, the fewer objections we see now are not attributable to an improvement in the quality of the agency’s statistics, but to the fact that the current dearth of data simply does not allow economists to check for discrepancies between different indicators, Petrov says.
Overall, distorting statistics is not news – for example, China was (and is) often accused of it. Therefore, methods to indirectly verify them have existed for a long time. “You can check the consistency of data at different levels. For example, economic activity is closely related to the volume of goods transported, GDP growth rates can be compared with production dynamics in key sectors of the economy, retail sales with real incomes, etc. Experts often perform such comparisons if figures are in doubt – precisely because statistics are important for economists, like test results are for a doctor, as if he, say, prescribes treatment based on incorrect biopsy results, the results can be tragic,” says the well-known economist.
Of course, it is always rough estimates, even in peacetime, admits University of Chicago economist Konstantin Sonin. But now, with the war, notes Petrov, there is such a major restructuring of the economy that indirect methods are becoming almost useless. “The hypertrophying of the military-industrial complex has greatly reduced the multiplier effect of government spending (meaning the increase in government spending applies to a significantly limited number of sectors), as direct spending on the army and military-industrial complex is virtually useless for the economy as a whole. This is very similar to the late USSR: there, too, part of GDP was useless for the broad economy,” he explains.
Everyone else is much more dependent on the [political] situation and, as a result, are simply forced to make distortions to please their superiors,” says Petrov. In his opinion, the macroeconomic data from the Central Bank and the Ministry of Finance can still be trusted. “For now,” he adds.
However, he says, it is hardly possible to directly compare the quality of statistics from different sources. “For example, Rosstat or the Central Bank obtain most of their estimates by adding up sample data, while the Ministry of Finance gives exact figures for spending, expenditures, government borrowing, etc.,” he explains. Whereas in past years economists had more doubts about Rosstat’s GDP estimates, the fewer objections we see now are not attributable to an improvement in the quality of the agency’s statistics, but to the fact that the current dearth of data simply does not allow economists to check for discrepancies between different indicators, Petrov says.
Overall, distorting statistics is not news – for example, China was (and is) often accused of it. Therefore, methods to indirectly verify them have existed for a long time. “You can check the consistency of data at different levels. For example, economic activity is closely related to the volume of goods transported, GDP growth rates can be compared with production dynamics in key sectors of the economy, retail sales with real incomes, etc. Experts often perform such comparisons if figures are in doubt – precisely because statistics are important for economists, like test results are for a doctor, as if he, say, prescribes treatment based on incorrect biopsy results, the results can be tragic,” says the well-known economist.
Of course, it is always rough estimates, even in peacetime, admits University of Chicago economist Konstantin Sonin. But now, with the war, notes Petrov, there is such a major restructuring of the economy that indirect methods are becoming almost useless. “The hypertrophying of the military-industrial complex has greatly reduced the multiplier effect of government spending (meaning the increase in government spending applies to a significantly limited number of sectors), as direct spending on the army and military-industrial complex is virtually useless for the economy as a whole. This is very similar to the late USSR: there, too, part of GDP was useless for the broad economy,” he explains.
A line to a government documentation center, Moscow. Source: VK
Everyone in the same boat
According to Sonin, the matter is not only – and not so much – in the distorting of statistics, but in what is happening broadly with the Russian economy. “When people talk about statistics, they don’t know what they’re talking about. Because when we estimate, for example, GDP, the statistics are only a small, unimportant part, while the important part is the model into which the data is inputted. Plugged into another model, this data or coefficients will give a completely different picture. If the model is wrong – even if you were given all the data, completely and accurately – the macroeconomic indicators put into a bad model will still give an inaccurate result,” he explains.
There is no bad model now, Sonin believes, but the Russian economy is currently being restructured so quickly that the previous model applied to the current situation is completely inadequate and not calibrated. No one knows which model is correct. “And when, for example, [First Deputy Prime Minister] Andrei Belousov gives data based on the model he uses, they are no different from if I or, for example, Sergei Aleksashenko say, ‘[it is off] 5-10%.’ But more competent economists, like Sergei Guriev, will put together that everything is approximate,” says Sonin.
“For example, I understand that if you put a tank into the previous GDP model,” continues Sonin, “then this tank gives a very large increase in GDP – employment in the defense industry increases, investments – the GDP indicators rise.
According to Sonin, the matter is not only – and not so much – in the distorting of statistics, but in what is happening broadly with the Russian economy. “When people talk about statistics, they don’t know what they’re talking about. Because when we estimate, for example, GDP, the statistics are only a small, unimportant part, while the important part is the model into which the data is inputted. Plugged into another model, this data or coefficients will give a completely different picture. If the model is wrong – even if you were given all the data, completely and accurately – the macroeconomic indicators put into a bad model will still give an inaccurate result,” he explains.
There is no bad model now, Sonin believes, but the Russian economy is currently being restructured so quickly that the previous model applied to the current situation is completely inadequate and not calibrated. No one knows which model is correct. “And when, for example, [First Deputy Prime Minister] Andrei Belousov gives data based on the model he uses, they are no different from if I or, for example, Sergei Aleksashenko say, ‘[it is off] 5-10%.’ But more competent economists, like Sergei Guriev, will put together that everything is approximate,” says Sonin.
“For example, I understand that if you put a tank into the previous GDP model,” continues Sonin, “then this tank gives a very large increase in GDP – employment in the defense industry increases, investments – the GDP indicators rise.
“Nevertheless, this means that GDP ceases to be an indicator of people’s well-being, because a tank does not increase the standard of living.”
Therefore, Sonin believes, the statistics have deteriorated not in the sense that something has been classified, but in the fact that the model needs to be recalibrated to understand what unemployment, inflation and economic growth is now. “How do you calibrate this if everything happened in one year, while, if you do it right, the data needs to be looked at for a long period? This happens during all crises – it’s just that the crisis now in Russia is absolutely epic,” he concludes.
The problem of the quality of statistics and the accuracy of the model in which these statistics are inputted is not a matter of theoretical debate among economists – it is, first and foremost, the ability of government entities to soberly assess reality, make high-quality forecasts and make the right decisions. Thus, Sonin recalls that throughout the 1980s, until the very beginning of the collapse of the Soviet economy in 1991, most economists believed that there would be no economic decline from a transition to a market economy. “It was not that the structural model was wrong, the conceptual model was. And if you have the wrong model, then all the conclusions will be wrong,” explains Sonin.
But perhaps the Russian authorities have more complete data that allows them to see the undistorted picture? Rumors often circulate about “secret” statistics going to the Kremlin, either from the relevant agencies or the Federal Protective Service (FSO). Alas, says Petrov, these rumors are groundless: the information that the government has, of course, is more complete than that available to experts, but otherwise the government uses the same estimates. “Neither Rosstat nor any other agency builds multiple versions of statistical indicators – an ‘accurate’ one for the government and a ‘distorted’ one for others. Thus, in terms of the quality of statistics, everyone – the government, experts and business – are in the same boat; everyone is interested in the accuracy and reliability of all information,” he concludes.
The fact that the government must make decisions based on embellished data is extremely worrying. Suffice it to recall that the Kremlin made the decision to invade Ukraine based on embellished information from Russian agents there. The habit of pleasing your boss can be deadly sometimes.
The problem of the quality of statistics and the accuracy of the model in which these statistics are inputted is not a matter of theoretical debate among economists – it is, first and foremost, the ability of government entities to soberly assess reality, make high-quality forecasts and make the right decisions. Thus, Sonin recalls that throughout the 1980s, until the very beginning of the collapse of the Soviet economy in 1991, most economists believed that there would be no economic decline from a transition to a market economy. “It was not that the structural model was wrong, the conceptual model was. And if you have the wrong model, then all the conclusions will be wrong,” explains Sonin.
But perhaps the Russian authorities have more complete data that allows them to see the undistorted picture? Rumors often circulate about “secret” statistics going to the Kremlin, either from the relevant agencies or the Federal Protective Service (FSO). Alas, says Petrov, these rumors are groundless: the information that the government has, of course, is more complete than that available to experts, but otherwise the government uses the same estimates. “Neither Rosstat nor any other agency builds multiple versions of statistical indicators – an ‘accurate’ one for the government and a ‘distorted’ one for others. Thus, in terms of the quality of statistics, everyone – the government, experts and business – are in the same boat; everyone is interested in the accuracy and reliability of all information,” he concludes.
The fact that the government must make decisions based on embellished data is extremely worrying. Suffice it to recall that the Kremlin made the decision to invade Ukraine based on embellished information from Russian agents there. The habit of pleasing your boss can be deadly sometimes.
Share this article
Read More
