Scientists Beware: Shooting the Messenger

05/30/2013 04:29 pm ET | Updated Jul 30, 2013

By Joseph B. Kadane

While every scientist faces the possibility of being the bearer of bad news, statisticians are especially vulnerable. At present, two countries -- Greece and Argentina -- are bringing charges against statisticians who are reporting truthful, yet unpleasant, data.

Greece, of course, has undergone a terrible economic contraction brought on by the austerity demanded by the European Union (EU) and the International Monetary Fund (IMF) as the price of preventing an economic collapse. A key number in these negotiations is the fiscal deficit.

As a result of years of reporting numbers that later were found to be inaccurate (i.e., wrong and possibly doctored), the Hellenic Statistical Authority (ELSTAT) was reorganized and a new president, Andreas Georgiou, a veteran IMF statistician, was appointed. The new team re-examined the Greek fiscal deficit.

In accounting for the deficit of 2009, they took into account the losses of companies wholly owned by the government, since surpluses of these companies are regarded as income to governments under EU rules. This acceptable action resulted in felony prosecution of Georgiou and two of his managers by Greek prosecutors, who charged them with "false attestation to the detriment of the state" and "violation of duty."

International statistical bodies have given strong support to Georgiou and his colleagues. Walter Radermacher, the director general of Eurostat, the EU's statistical office, wrote the following on behalf of EU heads and European Free Trade Area national statistical institutes:

"... We are concerned that political debates surrounding judicial action taken against the head of the Greek statistical office (ELSTAT) and the calling into question of the validity of data, which have repeatedly passed the stringent quality checks applied by Eurostat to ensure full compliance with European law, seem to disregard commonly agreed European procedures. ... We are encouraged that, in Greece, many important steps have been taken in recent years to reinforce the credibility of official data, such as the new statistics law establishing ELSTAT as an independent statistical institute or the signing by the Greek government of a Commitment on Confidence in European statistics and its subsequent endorsement by the Greek Parliament. We are confident that Greek authorities will do everything necessary to protect the independence of ELSTAT, its management and its staff from political and all other interference and support ELSTAT in ensuring that statistics in Greece are produced in line with the principles of the European Statistics Code of Practice."

Since it is not a governmental body, the International Statistical Institute, was blunter:

"[We] express [our] profound disappointment and serious concern over the prosecution of Mr. Andreas Georgiou, president of the Hellenic Statistical Authority (ELSTAT) and two of his ELSTAT managers. The charges against Mr. Georgiou and two of his managers of exaggerating the estimates of Greek government deficit and debt for the year 2009 are fanciful and not consistent with the facts. ... We believe there is absolutely no merit to the charges brought against Mr. Andreas Georgiou and his two ELSTAT colleagues."

Similarly, the United Kingdom's Royal Statistical Society wrote:

"On the basis of the information available to us, we have deep concerns about the validity of the case against Mr. Georgiou. We wish to urge that all is done to ensure that he is given a scrupulously fair hearing.

"Furthermore, it will be vital that those working within the Greek statistical service can be confident that their work to deliver data of integrity can be done according to their professional duty and in line with the United Nations' principles. They should be in no doubt that they do so free of any interference or pressure, implied or real. Therefore, we hope suitable assurance will be provided to them."

Meanwhile, the situation in Argentina is somewhat different.

In 2007, the government fired Graciella Bevaqua and other statisticians who collected its price statistics and inflation estimates. Since that time, large and disturbing -- even shocking -- discrepancies have developed between the official government inflation estimates (roughly 10 percent per year) and privately generated estimates released by Bevacqua and others (roughly 25 percent per year).

Rather than publishing and defending its methodology for measuring inflation, the Argentine government chose to attack the messengers. In doing so, it instituted administrative fines and pursued criminal charges against Bevaqua and others.

The official explanation for imposing the fines is that the government was responding to a complaint from wholesale fruit and vegetable vendors who claimed their sales decreased. The government asserted the reason for this decline in demand was the publication of the private inflation estimates by the statisticians.

It is hard to understand how a change in the demand for fruits and vegetables could result in a claim against the statisticians. While the idea that announcements by private statisticians can move markets is flattering, it is much more likely that the actual increase in the prices of produce, which the inflation rate reflects, led to the decrease in demand. The fines imposed on the statisticians are substantial, on the order of $100,000 to $150,000 (U.S.). They have yet to be adjudicated by a court.

The criminal charges are even more serious. These are based on the proposition that the announcement of the private inflation estimates endangered Argentine national security.

The first trial of Bevaqua and economist Nicolas Salvatore based on this flawed theory of liability resulted in an acquittal. It is not known whether the government will appeal. Additionally, there are still criminal charges pending against several other statisticians and economists.

The position of the government differs in these cases. In Argentina, the government is bringing the charges. In Greece, the accused are part of the government and are defended by it. Nonetheless, the effect on the individuals accused is similar.

Statisticians and other scientists must be aware that, as participants in the public space, they may face similar unpleasant incidents, however unjustified these may be. As Benjamin Franklin reportedly remarked when the Declaration of Independence was signed, "We must all hang together, or assuredly we shall all hang separately."

Joseph B. Kadane is Leonard J. Savage University Professor of Statistics and Social Sciences, Emeritus, at Carnegie Mellon University, and chair of the American Statistical Association's Committee on Scientific Freedom and Human Rights.