Good statistics are accurate, timely, consistent and comparable. Only then can they be the unbiased reality check needed for responsible decision-making. The ECB blog looks back at past mistakes and what Europe has learned from them.
We are often asked: how independent are European statistics, really? Following recent international headlines, this question has gained new momentum. To answer it, let me take you back to Europe's own "deficit drama" in statistics, explain how it was resolved, and illustrate how the governance we now have in place is a rock-solid safeguard for accountability, markets and policymaking.
From deficit revisions to stronger statistics
The credibility of European statistics came under scrutiny in 2009 when Greece revised its government deficit statistics sharply upwards. This was not just a fiscal matter - it highlighted the broader challenges involved in ensuring trust in official numbers.
"Στέργει γὰρ οὐδεὶς ἄγγελον κακῶν ἐπῶν"(Nobody likes the man who brings bad news), Antigone 276-277, Sophocles
The European response was substantial: to maintain trust, European legislation gave Eurostat - the statistical office of the Union - and the national statistical offices stronger verification powers in 2010. Eurostat also received mandatory access to government accounts, the right to carry out on-site inspections, and new laws strengthened sanctions for misreporting. Today, there is full transparency concerning the real magnitude of government spending - or, as the saying goes, we now call a spade a spade.
However, subsequent legal proceedings initiated in 2010 by the Greek judicial authorities against Andreas Georgiou, then head of the Hellenic Statistical Authority (ELSTAT), drew wide international attention. He was accused of allegedly conspiring to artificially inflate the 2009 government deficit figures while these numbers were validated by Eurostat. These legal proceedings were widely considered to have a substantial political context.[1] This case was one reason for the reinforcement of statistical independence through the regulation on the European Statistical System partnership in 2015.[2]
The new law not only clarified Eurostat's coordination role, but also explicitly strengthened the call for national statistical authorities and their heads to be professionally independent. Notably, it brought in safeguards to ensure that appointments and dismissals are transparent and based only on professional criteria.
Two systems, one culture of quality
Before discussing more lessons on good statistics, let's take a look at the architecture of public statistics in Europe. European official statistics are produced by two systems: the European Statistical System (ESS), composed of Eurostat and the national statistical offices, and the European System of Central Banks (ESCB), made up of the ECB and the national central banks. Central banks often compile balance of payments statistics, financial accounts and government debt data. Statistical offices compile figures on inflation, GDP and the government deficit.
These two systems work under different laws and structures. Yet cooperation runs deep - from working groups to high-level committees - and the methods, processes and results are highly transparent. To guarantee trust, the ESCB follows its own Public Commitment on European Statistics[3] , which is equivalent to the Code of Practice adopted by the ESS[4]
. Both ensure that statisticians - be it in the central banks or statistical institutes - work under identical principles of independence, soundness and quality.
Figure 1
Two statistical systems working closely together in multilateral settings such as the Committee on Monetary, Financial and Balance of Payments Statistics (CMFB)
Source: ECB.
Certainly, our governance is complex, multilateral and sometimes feels intrusive. This, however, is a strength. Our thinking is exposed to many stakeholders early on. So, any attempt at manipulation would not only be unlawful - it would be practically impossible. A bit like trying to make fake honey in front of an entire beekeeping club - too many people would notice. And once exposed, your reputation would be damaged beyond repair.
HICP: a solid, independent benchmark
But why are we so keen to have reliable statistics in the first place? To showcase this, let's take a closer look at one of the cornerstones of statistics, where figures which are essential for monetary policy are compiled outside of the bank. The HICP (Harmonised Index of Consumer Prices) is the official measure of inflation in the euro area, showing how the average prices of goods and services bought by households change over time. It is compiled by Eurostat and the national statistical offices. The diverse sample behind the HICP takes into account what we buy, how much, and where we do that, be it in a supermarket or an online shop. The index reflects realistic consumption patterns estimated from national accounts, household budget surveys and transaction data. To ensure its accuracy, these weights are updated annually. To reflect what consumers actually pay, the HICP also includes taxes and discounts. For this, each month millions of individual prices are collected to calculate the EU HICP from all Member States, thereby ensuring a robust and representative outcome. It is also one of our most timely indicators: we already have the euro area (and EU) flash estimate for the reference month on the last day of the month or shortly thereafter. This provides policymakers with crucial real-time information which is fundamental for their decisions.
Though essential for its policy, the ECB (ESCB) does not calculate the HICP. That is the exclusive responsibility of the ESS - and rightly so. We use the data as a core reference indicator; we are not the calculator. This separation gives assurance to markets, the public and policymakers that the methodology behind the inflation measure guiding monetary policy is not influenced by the central bank.
Chart 1
Development of the Harmonised Index of Consumer Prices in the euro area
Left-hand scale: HICP monthly index with reference baseline 2015=100; right-hand scale: year-on-year rate of change in percentage.
Source: Eurostat.
For another tangible example, consider the €STR, the euro short-term rate, which replaced EONIA as part of the global reforms that followed the LIBOR manipulation scandals. The LIBOR case showed how fragile benchmark interest rates can be, as they often rely only on price quotes and not on actual transactions or a sufficient number of transactions. The global response of public authorities to the LIBOR manipulation led to a shift towards more robust rates, based on real trades with stronger methodology and governance. In the euro area the newly developed €STR benchmark rate replaced EONIA.
EONIA was an overnight transaction-based lending rate administered by the European Money Markets Institute (EMMI), a private benchmark provider. With the application of the EU Benchmarks Regulation[5] , it no longer complied with the new standards - mainly because of the lack of sufficient underlying transactions and the high concentration of contributions. The ESCB therefore developed the €STR, which reflects the wholesale euro unsecured overnight borrowing costs of banks located in the euro area, and entrusted its production to the ESCB's statistical function, building on its solid reputation.
Since 2 October 2019 the €STR has been published Monday through Friday at 08:00, based on transactions from the previous day. It is a major achievement, building on our long-term investments in daily micro data collections. The daily collection of money market data is more than a technical exercise: it supports the analysis of euro money markets and monetary policy transmission, provides early signals of potential fragmentation and reveals market expectations. This again is crucial information for those who decide on monetary policy.
Chart 2
Evolution of €STR rates since its introduction
Source: ECB.
What makes for good statistics?
As the HICP and the €STR show, good statistics are accurate, timely, consistent and comparable. And, above all, they can be put to use where they are needed. For this, statistics need to be reliable.
The independence of statisticians and statistical offices from direct political or market pressure is one precondition for good statistics, and in the case of central banks, this independence is firmly safeguarded too. Equally important are transparent and robust statistical processes, which allow for sources and methods to be scrutinised, as illustrated by the ECB's annual methodology reviews for the €STR. Finally, sound quality management ensures that errors are corrected, and improvements are made when necessary. This is why in the above-mentioned reviews we maintain a dedicated transparency page on errors.[6] See the €STR Transparency on errors page on the ECB's website.
Statistics are a pillar of accountability
Independent statistics are not an academic concern. They are a pillar of accountability and effective policy. Without appropriate governance to ensure their relevance and quality, markets and people cannot trust the figures on which their choices depend.
Thanks to the rules we have in place - from the Treaty to the Code of Practice - and thanks to our deeply rooted culture of multilateral cooperation, policymakers, market participants and journalists can be confident that European statistics are compiled to the highest standards, free from interference or meddling. That trust is indispensable for our work and for Europe's success.
The views expressed in each blog entry are those of the author(s) and do not necessarily represent the views of the European Central Bank and the Eurosystem.
