Nicolas Véron

Integrity of Official Statistics Under Threat

This op-ed, jointly written with Ted Truman, was published last month by the reference Greek daily Kathimerini, in print as well as online in Greek and English. It was reposted this week by the Peterson Institute and Bruegel. 

Andreas Georgiou, former head of the Hellenic Statistical Authority (ELSTAT), has become the public face of the global threat to the integrity of official statistics and statisticians. The facts of the case are well known. The International Statistical Institute, the American Statistical Association and multiple regional statistical bodies have decried the threat in their expressions of support for Georgiou. A recent petition collected over a thousand signatures, including from nine Nobel laureates in economics.

Georgiou has been convicted for “violation of duty” for not submitting ELSTAT’s final figures on the 2009 budget deficit for approval by a politically appointed board. Given what is known of the views held by that board’s members, seeking its approval would surely have led to misleading figures. Instead, he followed the European Statistics Code of Practice, which assigns “sole responsibility” to heads of national statistical institutes (NSIs), such as ELSTAT, for their data production and dissemination. The appeals court conviction, upheld by the Supreme Court, ignored the fact that the European Statistics Code of Practice had been explicitly endorsed under Greek law before Georgiou took up his post. The irony is that successive Greek governments, while claiming to stand aside from the prosecutions of Georgiou, have used the same figures and statistical methodologies to justify financial support and now debt relief from their European partners. Unfortunately, European leaders have largely refrained from bringing pressure on Greek politicians to stop damaging the reputation of Greece and the European Union.

Reliable official statistics are essential building blocks of sound economic policy. Budget projections inevitably contain a combination of revenue and expenditure estimates and a dose of politics. Budget outcomes, as recorded in official statistics, however, are different. They must be accurate and above suspicion of political manipulation. They are crucial to the transparency and democratic accountability of governments. They are critical to the functioning of market activity, saving and investment, and thus to growth, employment and prosperity. Concerns about political manipulation of statistics are especially relevant for the euro area, where joint economic policies rely critically on the accuracy of national statistics on economic growth, budget outcomes, and sovereign debt, which are aggregated (but not compiled) by Eurostat, a department of the European Commission.

Moreover, prudent lenders in international financial markets rely on the integrity of national statistics to make investment decisions; less confidence in that integrity translates into higher borrowing interest rates for the government as well as the private sector. The case of past inflation statistics in Argentina illustrates the damage that politically motivated statistical manipulation can do to a country’s international reputation and access to markets, as well as to its economic prospects more broadly.

Georgiou’s case exposes the inadequacy of the existing safeguards in Europe’s policy on official statistics. The EU system is failing the most basic test of its ability to deter political pressures on official statistical production, with potential chilling effects on official statisticians in other member-states and more widely in countries around the world. For example, the process for compiling the all-important decennial census in the United States is in danger of political manipulation by the addition of a question about citizenship in the 2020 census, which could discourage non-citizens from participating. Everywhere, a major source of potential political pressure on statistical agencies is via constraints on their budgetary funding, which is controlled by national parliaments.

The multiple criminal and civil prosecutions against Georgiou have already damaged Greece’s international reputation by raising doubts about its sustained capacity and willingness to produce reliable official statistics. Greece will feel consequences if it seeks to return to international markets without reversing this damage. More broadly, Greece will not achieve the long-lasting growth, stability and prosperity that it wants without establishing trust in its official statistics, which cannot be achieved as long as Georgiou’s persecution continues.

The damage extends beyond Greece. EU leaders and institutions have underestimated how much the case of Georgiou is harming the reputation of the euro-area policy framework, which relies heavily on economic statistics from its member-states. The persecution of Andreas Georgiou must end. And Greece and the EU must ensure that similar cases never happen in the future.

Germany's Savings Banks Are Uniquely Intertwined with Local Politics

This post, co-authored with Jonas Markgraf, was published on July 17 by the the Peterson Institute and on July 18 by Bruegel. 

The roughly 400 savings banks in Germany, known as Sparkassen (singular Sparkasse),  represent about 15 percent of the country’s domestic banking assets. They have unusually direct links with local political communities, which are widely known in Germany but not well understood in the rest of Europe. This post describes these structural links and presents a new dataset about them.

The significance of disparate practices within the euro area banking systems has risen with the advent of banking union, or the pooling of supervisory policy with a central role for the European Central Bank (ECB) in force since November 2014. Even though the banking union is incomplete, and almost all Sparkassen are “less significant institutions” (LSIs) subject to only indirect supervision by the ECB, the banking union implies greater integration and interdependence among all banks in the euro area than in the past, when supervision and crisis resolution were strictly national competencies. Thus, differing banking structures are now relevant to the entire euro area.

The idiosyncratic structures of the Sparkassen derive from German financial history. Deutsche Bundesbank data indicate that Sparkassen represent €1.2 trillion out of Germany’s total domestic banking assets of €7.8 trillion as of April 2018, not counting other public-sector banks (such as the regional banks or Landesbanken) with which the Sparkassen also have structural links.

This analysis is based on a unique dataset constructed from individual Sparkassen annual reports. All the underlying figures are publicly available but are not easily aggregated. The dataset is based on individual data on the composition of Sparkassen (supervisory) boards (Verwaltungsrat) in 8 out of 16 German federal states, namely Bavaria, Brandenburg, Hesse, North Rhine-Westphalia, Rhineland-Palatinate, Saarland, Schleswig-Holstein, and Thuringia. (The sample may be extended to all states in future research.) The eight states in the sample represent 60 percent of Germany’s population (2016), 61 percent of its GDP (2017), and 278 Sparkassen out of a national total of 413, or 67 percent of all Sparkassen and 64 percent of all Sparkassen assets (DSGV 2015). There is no reason to assume that the patterns in the eight states outside the sample should be markedly different from the ones observed within it.

This post focuses on the Sparkassen relationships with elected politicians. Three findings stand out and do not appear to have been specifically documented before.

1. Elected Politicians Chair Most Sparkassen.

Sparkassen are noncommercial public-sector entities, whose controlling “owners” (Träger) are municipalities and/or counties, even though these local governments hold no liability for Sparkassen losses as has been established by EU competition policy jurisprudence. (All Sparkassen, however, are members of elaborate formal risk-sharing frameworks at the state and federal levels, including a shared deposit insurance scheme and an institutional protection scheme.) Accordingly, the politicians sitting on Sparkassen boards tend to be those elected at the county or municipality level and not federal- or state-level politicians, with few exceptions.

The definition of “politician” adopted for this analysis is highly restrictive, namely a currently serving elected official at the county (rural Landkreis or Kreis, urban Stadtkreis or kreisfreie Stadt) or municipality level (Gemeinde). The elected chief officer of a county is a Landrat (rural) or Oberbürgermeister (urban), and the chief officer of a municipality is a Bürgermeister. This nomenclature omits elected members of local government councils (at the municipality or county level) who are not council heads. It also omits former elected officials and other individuals who may strongly identify with political parties and interests. Thus, the methodology tends to underestimate the extent of the links between local political communities and Sparkassen boards.

Under these definitions, politicians represent a minority of Sparkassen board members but the overwhelming majority of board chairs, as summarized in table 1. In five of the eight states in the sample, all Sparkassen chairs are politicians.

Table 1: Presence of Politicians among Sparkassen Boards and Chairs

Source: annual reports, authors’ calculations. All data are as of January 2015.

These observations are fairly constant over time. Over the decade 2006 to 2015, the number of Sparkassen in the eight states declined slightly, from 293 to 278; the proportion of politicians among board members increased slightly, from 16 to 18 percent; and their proportion among board chairs was stable, fluctuating between 82 and 84 percent.

2. Most County Heads Are Involved in Sparkassen.

Conversely, an overwhelming majority of county heads (Landräte and Oberbürgermeister) chair a Sparkasse, as summarized in table2. There has been a lot of consolidation (mergers) of municipalities in recent decades, but less so of counties, which partly explains the high observed ratios.

Table 2: Presence of Sparkassen Board Chairs among County Heads

Source: annual reports, authors’ calculations. All data as of January 2015.

3. Many Politicians Derive a Tenth or More of Their Personal Income from a Sparkasse Position.

Only North Rhine-Westphalia appears to disclose board fee amounts paid to every board member. For each Sparkasse in that state, disclosures of the chair’s annual fees for 2015 can be added to an estimate of that same individual’s 2015 salary as a local government official (“government salary”), available from the official salary bracket, to provide an estimate of aggregate income (thus omitting any additional sources of income though). Only those Sparkassen whose boards are chaired by politicians are included in the sample. Table3 shows the results, with Sparkassen ranked by decreasing order of total assets. Unsurprisingly, the amount of chair’s fees is correlated to the volume of a Sparkasse’s total assets, explaining much of the observed variation. On average, using these definitions, politicians in North Rhine-Westphalia who chair a Sparkasse’s board receive 12 percent of their individual income from Sparkassen board fees.

It is hard to assess the extent to which this finding may be idiosyncratic to North Rhine-Westphalia. Individual board fee disclosures in other states would be needed, of course, to extend this analysis to the whole of Germany.

Table 3: Compensation of Sparkassen Chairs (North Rhine-Westphalia)

Source: DSGV (Sparkassen assets and employee numbers), Bundesanzeiger (income data). All data as of January 2015.

Euro Area Comparative Perspective

These three stylized findings appear highly specific to Germany’s Sparkassen. Leaving aside promotional and public development banks, which do not compete directly with commercial banks (e.g., Spain’s ICO, Finland’s Municipal Finance, France’s Caisse des Dépôts and SFIL, or Germany’s own promotional banks at the national [KfW] and state levels), bank board membership of politicians (as defined here) appears to be virtually nonexistent in other EU member states. In some of the member states, such board memberships were significant in the past, though rarely as significant as currently in Germany, but have nearly or entirely disappeared as the banking sector was reformed and/or restructured. In Germany itself, politicians may sit on boards of banks other than Sparkassen, including some of the cooperative banks and especially Landesbanken, but less frequently than observed in Sparkassen.

Our respective research on banking structures and specific recent enquiries supports this claim. For example, in Italy and Spain, some politicians may sit on the boards of some of the local political foundations that are shareholders in some of the banks, in certain cases with a controlling stake, but not on the banks’ boards themselves, with few exceptions that are all insignificant from a national banking system perspective. In France, regional savings banks, now integrated in the BPCE Group, may have one board member who represents the local government and is a politician, but only one and never the chair. In Austria, the savings banks (also called Sparkassen) share some attributes with those in Germany but have been consolidated in the Erste Group, whose governance is essentially commercial and whose board does not include politicians. In the Netherlands, there are no politicians even on the boards of the two national promotional banks, BNG and Nederlandse Waterschapsbank, let alone other banks. This is not to say that there are no links between political and banking communities in these countries but that such links do not take the form of politicians holding a board chairmanship or even, in the overwhelming majority of cases, any board membership.

The implications of these findings are beyond the scope of this post. A future research agenda could be developed to complete the sample, extend the comparison to systems and countries beyond the euro area, and explore the resulting policy challenges and tradeoffs.

 

Jonas Markgraf is a PhD student and research associate in State, Risk, and Society at the Hertie School of Governance (Berlin). Starting September 2018 he will be a postdoctoral research fellow at the University of Oxford. The research reported in this post was partly funded by the German Federal Ministry of Education and Research (Research Grant No. 01UF1508). Nicolas Véron is senior fellow at Bruegel and the Peterson Institute for International Economics.