“Noise with incorrect data and based on untrue assumptions” characterizes, in an official statement issued by the Hellenic Statistical Authority, the article by the professor of economics and former minister, Tassos Giannitsis, which was published in the press under the title “Hints for Greek Statistics in 2025 – Questions about the National Accounts and the real course of our economy”.

According to ELSTAT, Mr Giannitsis’ analysis “ignores the European framework, the estimation cycle, the role of stocks and quality controls. The ELSTAT applies uniform, binding methods, transmits and publishes data after validation, and is systematically audited in depth, as it is for all Member States. Public debate on the economy is necessary and useful when it is based on data. Not on unchecked suspicions, innuendo, and conspiracy.”

In addition to the technical responses detailed below, the ELSTAT statement points out and deplores that:

– Mr. Tasos Gianitsis has served as a minister in periods when Greece’s GDP was compiled in the same European framework of principles and methods, and did not receive the level of recognition it has today. It is, therefore, at least odd that he would choose to challenge the institutional architecture of the national accounts for the first time in the year 2025, that is, at a time when for the first time Greece has not a single reservation in its GDP figures and the external recognition of its authority by its peers in the European Union and internationally is unquestionable.

– Even more surprising is the fact that he chose to comment on such a serious technical issue with an apparent lack of clear knowledge of the system, sources and estimation cycle, without even asking ELSTAT for clarification before publishing the article. “Criticism is welcome and is a means of improvement, but the prerequisite for criticism is that it is based on true technical assumptions. Otherwise, it is either an error due to ignorance or a deliberate attempt at defamation,” it stressed.

Specifically, in the extensive statement issued by the ELSTAT, it responds to the “untrue assumptions and erroneous conclusions of the article” as follows:

“On 13/09/2025, the newspaper “Ta Nea” published an article by Tasos Giannitsis entitled “Hints for Greek Statistics in 2025 – Questions about the National Accounts and the real course of our economy.”

In this article, Tasos Giannitsis supports erroneous conclusions, which he bases on untrue assumptions.
Specifically:

According to the following statements.
A. How is Greece’s GDP calculated, and what are the reserves

.1.1.
1) Calculation of Gross National Product The calculation of Gross National Product (GNP) in all EU Member States is governed by the European System of Accounts (ESA 2010), which sets out uniform and binding rules.

The fundamental definition of GDP is that it expresses the net value of output added to the economy over a period of time (quarter or year). Because GDP, through the production method, captures what is produced within an economy, it follows as a natural consequence that there are two other methods: the expenditure method, which captures the uses of the goods and services produced (consumption, investment, net exports, inventories), and the income method, which captures the income generated by this production process.

Quarterly GDP is compiled and determined in most Member States through the production method, rather than the expenditure or income method. This means that the result is obtained by combining the Gross Value Added (GVA) of all sectors of the economy plus taxes on products, minus subsidies on products. The choice of the output method as the dominant method for quarterly estimates is due to the availability and completeness of statistical sources every quarter for the sizes of the economy.
This ensures consistency, validity, and comparability of the data, even if some expenditure variables (such as stocks) are not available every quarter at the time of compilation. The expenditure side in the quarterly national accounts is balanced by the output approach using changes in stocks as a residual element.

The above practice is followed at the European level by the majority of countries.

The expenditure side of the accounts is supplemented with all the figures during the final estimation of annual GDP in October of the following year, when the sources required for this type of calculation become available for the first time, such as Business Structure Statistics, Family Budget Survey, annual Balance of Payments and Foreign Trade data, annual General Government data, etc..
Therefore:

● Quarterly estimates are compiled using the production method, which does not require the calculation of consumption, capital formation, and net exports/imports, and are released approximately 65 days after the reference quarter, after validation by Eurostat.

● Annual estimates are compiled twice a year: the initial estimate, also using the output method, in March of the following year of the reference year, and the final estimate in October of the following year, when expenditure data for the year become available for the first time, using the Supply and Use tables by product method.

2) The “stocks” figure in the National Accounts

The concept of stocks in national accounts is standard and is provided for by ESA 2010. Stocks include:

1. materials and supplies
2. work in progress, such as
● growing crops
● trees and livestock at maturity
● unfinished construction
● unfinished other fixed assets, e.g., ships and floating oil rigs;
● unfinished research for legal or consulting purposes;
● incomplete film productions;
● unfinished computer programs.

3. finished products
4. goods for resale.

As can be seen from the above, when calculating the quarterly GDP and the first annual estimate, stocks are not directly counted, nor is the change in stocks counted as a physical quantity. Instead, total output in the economy is determined from available data, and final consumption expenditure, investment, and net exports are estimated.
Stocks, and hence their change, are then calculated residually, i.e., they are the arithmetic remainder of the subtraction from output of the estimates of final consumption expenditure, investment, and net exports. Stocks, in other words, temporarily compensate for the divergence between supply and demand.

When the data are finalised, i.e., in the final GDP estimate in October next year, inventories are broken down by product.

Therefore, in the calculation of quarterly GDP, it is by definition impossible to affect other variables through the “manipulation” of stocks, since the stocks themselves are obtained as a numerical balance.

They can be used to calculate the value of the stock as an absolute figure.

B. What are the untrue assumptions and erroneous conclusions of the article:

1) GDP is calculated by the expenditure method, and stocks directly affect the size of GDP. By increasing the size of stocks, GDP is also increased.

The main implicit claim of the article is that the ELSTAT increased reserves are attributed to the growth rate, because, in the author’s opinion, this does not follow from the other measures.

This, i.e., an artificial increase in the growth rate through manipulation of inventories, would possibly only be possible if inventories were calculated as a stand-alone measure under the expenditure method in quarterly GDP. But this is not the case. GDP is estimated directly by the method of production from existing sources and takes a specific value. Stocks do not enter into the calculation of this value, but are the arithmetic balance between the production method and the expenditure method in quarterly GDP.

2) The quarterly GDP estimates for the first quarter of 2025 and the initial estimate of annual GDP 2024 that have been published show a sharp and unexplained increase in inventories, and in particular in industrial manufacturing products in stock, which are by the article mainly those that can be stockpiled.

As explained, on the one hand, inventories are not only industrial products that have not been disposed of, and on the other hand, in the quarterly GDP 2025 and the first estimate of the annual GDP 2024, which the article comments on, the “inventories” figure does not reflect products produced and not yet disposed of on the market, but the statistical difference between the method of production and the method of expenditure.

3) Stocks in the rest of the eurozone countries are almost zero.

Not true. Interested parties can refer to the data for the rest of the countries in the whole EU and see the stocks of each country in detail.

4) Reserves in Greece in 2024 jumped to 4.1% of GDP compared to previous years.

This figure seems to have been derived incorrectly as follows: the level of reserves is derived by subtracting from Gross Capital Formation (P.5) at constant prices the fixed capital formation component (P.51G).

This approach is incorrect, as the change in inventories (P.52) does not make sense to deflate, since it is a difference and not an actual flow of output or expenditure. The variable “change in inventories” shows the change in the value of inventories between two points in time, not the physical quantity of goods stored or consumed. Stocks may include goods acquired or produced at different times and therefore at different prices. There is no fixed ‘base price’ that can be applied uniformly. And this variable may take negative values when stocks decrease during the period.

In such cases, deflation loses all meaning, since there is no “real product” that has declined, but merely a decline in the book value of inventories. The percentages resulting from this deduction are therefore misleading and do not reflect the true size of inventories. It should be noted that changes in stocks, because they are expressed exclusively in current prices, can show considerable fluctuations from year to year, not only because of quantitative changes in stocks, but mainly because of developments in the general price level, especially in periods of significant inflationary pressures.

In any case, numerically isolating a residual element of quarterly expenditure and projecting it as an ‘anomaly’ overlooks the accounting architecture of the system. In the first annual estimate and in the quarterly estimates, GDP is “locked” on the output side. Expenditure balances out with stock changes, absorbing heterochronism and source availability mismatches. The second annual estimate is based on annual sources and uses supply-demand tables by product. The figure reproduced in the paper, in addition to being scientifically incorrect, is inherently incomparable with mature annual estimates.

C. What are the production conditions of Greece’s GDP

1) Transmission and validation

GDP data are transmitted to Eurostat within the specified deadlines of the ESA 2010 transmission schedule and are published by ELSTAT and Eurostat after technical validation. There has been no case of non-validation of Greek GDP data since November 2010 to date.
There has been no case of a previous update of the Eurostat data since the last update in November 2010.

2) Systematic and thorough checks

The GDP compiled by ELSTAT is further subject to systematic and thorough checks by Eurostat, as part of the GNI Verification cycle. These checks concern the sources, methodology, and results, and are a prerequisite for the certification of the data used in the calculation of the EU’s own resources and the allocation of EU funds.

In the context of the verification cycle carried out in all EU Member States in accordance with Regulation (EU) 2019/516 on the harmonisation of Gross National Income at market prices, Greek data, after the completion of the major revision of the National Accounts in September 2024, are consistent, without interruption, for the whole time series since 1995.

3) Output over time

At the 2014 National Accounts benchmark revision, with a base/reference year of 2010:

ELSTAT adopted the European System of Accounts ESA 2010, replacing the older ESA 1995, in accordance with Regulation (EU) 549/2013, applicable from September of the same year, and proceeded to revise the base year and compile the 1995-2013 time series of national accounts.

Among the main Methodological changes mentioned are:
1. Changes due to the implementation of ESA 2010

The main changes are the following.
● Accounting for Research and Development (R&D) as a fixed asset
● Calculation of military equipment as Gross Fixed Capital Formation

● Classification of institutional units into public and private sectors using updated ESA 2010 criteria

2. Other methodological changes:

Other changes to the structure of the public sector:
● New estimate of the capital stock and consumption of fixed capital for the whole economy and for the general government sector

● Calculation of own account software production

In the 2020 national accounts benchmark revision, with a base/reference year of 2015:

During the work, the following were incorporated:

● Use of administrative tax data from 2017, in conjunction with enterprise structure surveys, to improve coverage of the sources used to compile national accounts.

● Compilation of the institutional sector figures of the MFIs using administrative tax data on an annual basis.
● Integration of the updated statistical register of enterprises of ELSTAT for the years 2011 and onwards using administrative and tax data. Until 2017, the outdated statistical register of enterprises of the year 2010 was used for the compilation of GDP. The receipt of administrative data from the tax authorities has made it possible to update the statistical business register annually. At the same time, the statistical register of Non-Profit Institutions Serving Households (NPIs) for the base year 2015 was updated using tax data.

● Improvement of the methodology for estimating financial intermediation services measured indirectly (FISIM). Specifically, data on loan and deposit balances and corresponding interest by institutional sector of the economy were used to estimate FISIM. At the same time, the calculation of the internal and external benchmark interest rate has been aligned with the requirements of ESA 2010 (paragraphs 14.09 and 14.10).

● Update of sources and methods of calculation of variables for agriculture, forestry, and fishing (NACE Rev.2 division A).

● Improvement of the methodology for estimating stock changes in the annual compilation of GDP. Stock changes of the different categories (finished goods, materials and supplies, work in progress, goods for resale) have been aligned with those defined in ESA 2010. At the same time, holding gains and losses have been excluded from the changes in stocks in all sectors of the economy.

● Revision of the Greek shipping activity according to the new model of the Bank of Greece. The model estimates maritime transport based on data drawn from international trade databases and administrative sources.
● Incorporation of updated statistics on international trade in goods based on the principle of economic ownership. ELSTAT made new estimates on imports/arrivals, exports/shipments of ships and aircraft based on the principle of change of economic ownership in accordance with the provisions of EU Regulations 96/2010 and 113/2010.

In the major (benchmark) revision with a new base year of 2020 and a new reference year of 2021 in September 2024:

During the work on the revision, the following were incorporated in particular:

During the work of the benchmarking process, the main elements of the work were incorporated into the main elements.
● Methodological improvements in the context of the cycle of verification of Gross National Income (GNI) methods and sources:

⮚ Introduction of improvements for housing services

Updated all years’ values from Family Budget Survey data for each year and updated the housing stock based on the 2021 Census of Housing Population data. This improvement has had a significant impact on the time series.
⮚ Introduction of improvements relating to the insurance sector

The methodology was updated to be ESA 2010 compliant, and data from a new source, the Association of Insurance Companies, was used.

⮚ Production for own account
A markup for producers of tradable goods in all sectors of economic activity was included in Gross Value of Production.

⮚ Updated estimates for illicit activities from drugs, alcohol, and tobacco

All the best available data sources, as well as data from official sources, were used in the context of refining the estimates.
⮚ Wages in kind – Recording of daily allowances
Data were updated so that the daily allowances received by employees in all sectors of the economy for business travel are recorded in wages with a corresponding deduction of this amount from intermediate consumption. A revaluation of wages and salaries in kind was also carried out, as there were sectors of the economy where no amounts were recorded in this transaction.
⮚ Updating the statistical register of Non-Profit Institutions Serving Households (NPIs) and reassessing the figures.
⮚ Incorporation of the results of the Population – Housing Census, year 2021, on employment and housing stock.

At the same time, integration was made:

Incorporation of the following elements.
● The results of the Business Structure Survey, with a reference year of 2021, for all sectors of economic activity.
● the Statistical Register of Enterprises for the year 2021,
● the Statistical Register of Enterprises for the year 2021,

for all sectors of the economy,

for the year 2021.
● the updated estimates of household final consumption based on the Family Budget Survey (FBS),
● updated data on Non-Profit Institutions Serving Households (NPIs) from administrative sources,
● updated general government data,
● the updated Foreign Trade and Balance of Payments data,
● updated data for the sectors of Agriculture, Forestry, and Fisheries,
● updated short-term indicators data,

● updated employment data.

During this period, the completion of the lifting by Eurostat of a total of 21 country-specific reservations and 5 cross-conditionality reservations was also achieved.

D. Concluding remarks

The first estimate of GDP 2024 was published on 7 March 2025 and, like all previous estimates, has been validated by Eurostat. It is an initial estimate and is based on the sum of quarterly results. The final estimate is scheduled to be published on 16 October 2025.

The best available sources of output and expenditure are used to produce GDP. ELSTAT follows international practice and applies the Eurostat1 and OECD manuals. For the compilation of Greece’s quarterly GDP, users can also refer to the ”QNA inventory based on ESA 2010 methodology, posted on the ELSTAT portal, which describes in detail the production methodology. The time series are available to any interested party on the ELSTAT website, and any serious analyst can check and reproduce the data, documented on the methodology applied.

The coincidence of GDP size with the two methods of production and expenditure could only be observed at an ideal theoretical level, as it presupposes the coincidence of timing and completeness of the respective sources.

Tasos Giannitsis has served as a minister in periods when Greece’s GDP was compiled in the same European framework of principles and methods, and did not receive the level of recognition it has today. It is, therefore, at least odd that he should choose to question the institutional architecture of the national accounts for the first time in the year 2025, that is, at a time when for the first time Greece has not a single reservation in its GDP figures and the external recognition of its authority by its peers in the European Union and internationally is unquestionable.

Even more surprising is the fact that he chose to comment on such a serious technical issue with an apparent lack of clear knowledge of the system, sources, and estimation cycle, without even asking ELSTAT for clarification before publishing the article. Criticism is welcome and is a means of improvement, but the prerequisite for criticism is that it must be based on true technical assumptions. Otherwise, it is either an error due to ignorance or a deliberate attempt at defamation.

Finally, the fact that in the year 2025, fifteen years after its establishment, conspiracy theories about the production of Greek statistics by ELSTAT are being circulated with such levity is, at best, sad. The personal indictment of every one of the ELSTAT people working on the production of GDP and their stigmatization would be unremarkable as another attempt at mud-slinging if it were not also so dangerous for the country. It is distressing that ELSTAT is recognized abroad by its peers for its credibility and validity, and at home, there is an attempt to revive a historical trauma of the country on the back of the agency that restored it.

In conclusion

The publication creates noise with incorrect facts and on the basis of untrue assumptions. It misunderstands the European framework, the estimation cycle, and the role of stocks and quality controls. ELSTAT applies uniform, binding methods, transmits and publishes data after validation, and is systematically checked in depth, as is the case for all Member States. Public debate on the economy is necessary and useful when it is based on data. Not on unchecked suspicions, innuendo and conspiracy theories.”

The ELSTAT statement concludes.