The per­ils and pit­falls of a mod­ern-day statis­ti­cian

The Star Early Edition - - BUSINESS REPORT - Pali Le­hohla

GALILEO Galilei (1564-1642), is an Ital­ian who im­mersed him­self in ob­ser­va­tion to ad­vance and trans­form knowl­edge from pos­tu­lates to the­o­rems. He was a sci­en­tist of im­mense rigour who, al­beit a Chris­tian, would not be in­tim­i­dated by priestly threats that too of­ten stood in the path of sci­en­tific en­quiry.

House ar­rest and im­pris­on­ment would not stop him from ar­tic­u­lat­ing a sci­en­tific ob­ser­va­tion that the earth re­volves around the sun and ro­tates on its axis. “E pur si mouve” which trans­lates into “Yet it moves”.

This was a pa­pal pun­ish­ment against what was seen to be heresy. It was only to­wards the clo­sure of the 20th cen­tury that a pub­lic pa­pal apol­ogy to pun­ish­ing Galileo’s “e pur si mouve” was made.

In the pre­vi­ous col­umn I raised the emer­gent per­ils con­fronting a mod­ern statis­ti­cian.

These dan­gers are not ac­ci­den­tal, but emerge as sta­tis­tics comes of age in sys­tems of ac­count­abil­ity.

This is es­pe­cially as democ­racy takes hold and is sup­ported largely by the di­men­sion of trans­parency, which is brought about by an ex­plo­sion of in­for­ma­tion tech­nol­ogy and the power of own hand-held de­vices.

Sys­tems of ac­count­abil­ity were first ar­tic­u­lated by the ju­di­ciary. This is well cap­tured by An­dreas Ge­or­giou, a for­mer Greek statis­ti­cian whose un­end­ing pros­e­cu­tion by suc­ces­sive Greek au­thor­i­ties since 2010 is a mat­ter of pub­lic knowl­edge.

And it is a mat­ter of pe­cu­liar cu­rios­ity demon­strat­ing the Galileian per­ils that con­front a chief statis­ti­cian in mod­ern sys­tems of govern­ment four cen­turies later. He makes ref­er­ence to Mon­tesquieu’s state­ment in 1748 on the in­de­pen­dence of the ju­di­ciary and ar­gues that this prin­ci­ple must ap­ply to of­fi­cial sta­tis­tics.

In his pa­per de­liv­ered at the 61st World Con­gress of Sta­tis­tics ti­tled The pro­duc­tion of of­fi­cial sta­tis­tics needs to be a sep­a­rate branch of govern­ment he ar­gues that “the in­de­pen­dence of of­fi­cial sta­tis­tics has to be real and not ap­par­ent merely” con­sis­tent with Mon­te­quieu’s state­ment.

My in­formed view on mat­ters of of­fi­cial sta­tis­tics is that sys­tems of hu­man en­deav­our deal­ing with ac­count­abil­ity have evolved.

With each mi­lieu they ad­vance an arm of the state into the realm of Mon­tesquieu’s in­de­pen­dents. The first to ma­ture and get early recog­ni­tion was the ju­di­ciary through in­deed the Mon­tesquieu prin­ci­ple of in­de­pen­dence.

Fol­low­ing closely on the ju­di­ciary was the leg­is­la­ture as an in­de­pen­dent arm of the state, largely driven by the ad­vent of par­lia­men­tary democ­ra­cies that started hold­ing their own in many a ju­ris­dic­tion.

More re­cently in the 21st cen­tury, supreme au­dits claimed their in­de­pen­dent space as a cru­cial part of sys­tems of ac­count­abil­ity that have to do in the first in­stance with fi­nan­cial pru­dence.

So un­sur­pris­ingly so, as the sec­ond decade of the 21st cen­tury comes to a close, largely driven by the ad­vent of in­for­ma­tion tech­nol­ogy and ubiq­ui­tous hand-held de­vices as well as the Sus­tain­able Devel­op­ment Goals, of­fi­cial sta­tis­tics has in­creased its vis­i­bil­ity and its in­trin­sic value of a con­duit of trust. In this re­gard they have been cat­a­pulted to the high ta­ble and thus knock­ing in this fa­mous space of Mon­tesquieu in­de­pen­dents.


An­dreas Ge­or­giou is a for­mer pres­i­dent of the Hel­lenic Sta­tis­ti­cal Au­thor­ity, serv­ing from 2010-2015. He is an MIT grad­u­ate. He was re­cruited from the IMF to be the pres­i­dent and at the end of his ten­ure, he be­came a vis­it­ing lec­turer at Amherst Col­lege in the US and a con­sul­tant in of­fi­cial sta­tis­tics.

His first task that put him in trou­ble was to re­vise the Greek deficit of 2009. The level of deficit was re­vised to 15.4 per­cent from 13.6 per­cent in 2010.

These find­ings, which were un­der­taken ap­ply­ing Euro­stat prac­tices and con­firmed by the IMF, re­vealed that Greece had historically un­der­es­ti­mated its deficit.

The govern­ment was not pleased, and Ge­or­giou has since faced charges of breach of faith by his govern­ment, with mem­bers of his board serv­ing as wit­nesses in this case where he and two oth­ers are accused of ar­ti­fi­cially in­flat­ing the deficit.

If con­victed Ge­or­giou faces up to five years im­pris­on­ment. To his sur­prise and ours as prac­ti­tion­ers of sta­tis­tics, Ge­or­giou’s find­ings fol­lowed the EU reg­u­la­tions and stan­dards of com­pi­la­tion.

“The sta­tis­tics we have been accused of ‘çook­ing’ are the ones that did not re­ceive any reser­va­tions from Euro­stat”, says Ge­or­giou.

On De­cem­ber 6, 2016, Ge­or­giou was found in­no­cent of all charges prof­fered by a First In­stance Court, but the unan­i­mous ac­quit­tal by a three-judge panel was an­nulled by another pros­e­cu­tor and Ge­or­giou faced a dou­ble jeop­ardy trial last week on July 18 and the hear­ing will re­sume on July 31.

Ge­or­giou has been there be­fore, as in Septem­ber 2015, he was first ac­quit­ted by the Ap­peals Court Coun­cil of any wrong­do­ing, but the pros­e­cu­tor in the Supreme Court de­cided to an­nul the ac­quit­tal on Au­gust 1, 2016.

Over the past six years Ge­or­giou, whose daugh­ter was seven when the court case started, has been fac­ing trial on mat­ters of of­fi­cial sta­tis­tics, mak­ing of­fi­cial sta­tis­tics a high risk ca­reer.

When I served as a vice-pres­i­dent of the In­ter­na­tional Sta­tis­tics In­sti­tute from 2009 to 2013, our board wrote to the Greek par­lia­ment in 2013 protest­ing the per­se­cu­tion of Ge­or­giou.

In March 2012 at the UN Sta­tis­tics Com­mis­sion Sem­i­nar on “Vi­o­la­tions of the UN Fun­da­men­tal Prin­ci­ples”, I was part of the panel. In my pre­sen­ta­tion on the list of of­fend­ers by omis­sion or com­mis­sion were Canada – 2010 Cen­sus long sam­ple de­sign non-ran­dom; Ar­gentina fal­si­fi­ca­tion of the Con­sumer Price In­dex; Greece fal­si­fi­ca­tion of deficit num­bers; In­done­sia – house­hold pri­vacy vi­o­la­tions and South Africa – in­flat­ing the Con­sumer Price In­dex (CPI) over fif­teen months from 2002.

By the way, I was al­ready then the Statis­ti­cian-Gen­eral and the in­flated num­bers were un­der my watch.

What were the coun­tries’ re­ac­tions?

The Ar­gen­tinian re­ac­tion was most vir­u­lent, as they called me a liar and walked in protest out of the pre­sen­ta­tion where there were in ex­cess of 300 par­tic­i­pants.

Their am­bas­sador ac­costed the South African Per­ma­nent Mis­sion and they pe­ti­tioned the direc­tor of the UN Sta­tis­tics di­vi­sion as well. South African am­bas­sador Baso Sanqu sum­moned me on the same evening for an ex­pla­na­tion.

I also re­ported this in­stance to Min­is­ter Trevor Manuel as soon as I ar­rived back home. My re­sponse to all these au­thor­i­ties was un­der­stood to be im­pec­ca­bly pro­fes­sional, fair and firm.

The Canadian statis­ti­cian re­sponded as well and ar­gued that the cen­sus still re­mained sci­en­tific and my re­tort was that a vol­un­teer sam­ple can­not be sci­en­tific.

House ar­rest and im­pris­on­ment would not stop him from ar­tic­u­lat­ing a sci­en­tific ob­ser­va­tion that the earth re­volves around the sun and ro­tates on its axis.


What hap­pened in these ju­ris­dic­tions since these vi­o­la­tions?

In South Africa I ad­mit­ted to er­ror and no amount of apolo­gies could ever be enough given the im­pact of the er­ror, but I in­deed re­main gra­cious to the govern­ment and peo­ples of South Africa for their mag­na­nim­ity and un­der­stand­ing that er­rors do and can oc­cur.

I led a gru­elling re-en­gi­neer­ing of the eco­nomic, so­cial and pop­u­la­tion sta­tis­tics and the mod­erni­sa­tion pro­gramme of our sta­tis­tics sys­tem and cor­rected the 2002 er­ror in the CPI.

That has re­mained a big blot and a con­stant re­minder in my ca­reer as a statis­ti­cian and as a statis­ti­cian-gen­eral.

In Canada, the 2016 Cen­sus has rein­tro­duced a manda­tory long form and there is ma­jor leg­isla­tive re­form that was oc­ca­sioned by govern­ment ac­tion of an­nulling or tem­per­ing with the sam­ple de­sign im­ple­men­ta­tion of the long cen­sus form.

The Ar­gen­tini­ans have re­mained very ap­pre­cia­tive of my un­re­lent­ing stance of cri­tiquing them dur­ing their er­rant days. They are now re­form­ing their sta­tis­tics and I have been in­vited to view their work and progress they have made.

The for­mer Greek statis­ti­cian Ge­or­giou re­mains my in­tel­lec­tual com­rade as we weave through the per­ils of of­fi­cial sta­tis­tics and trust that Greece will see the light of day. In­done­sia re­mains stuck with benev­o­lent vi­o­la­tion of pri­vacy and the risks re­main real.

Like Galileo, the sci­en­tist who faced the per­ils of re­li­gion, chief statis­ti­cians face the per­ils of pol­i­tics. Will the Mon­tesquieu prin­ci­ple of in­de­pen­dence save them? The South African Sta­tis­tics Act 6 of 1999 is now un­der leg­isla­tive re­view.

Dr Pali Le­hohla is South Africa Statis­ti­cian-Gen­eral and Head of Sta­tis­tics South Africa.


This un­dated file im­age shows an etch­ing of as­tronomer Galileo Galilei. Galileo had to fight re­li­gious bias, statis­ti­cians of­ten have to fight po­lit­i­cal bias, says the au­thor.

Newspapers in English

Newspapers from South Africa

© PressReader. All rights reserved.