IS GDP ‘GEN­ERAL DATA POVERTY’?

CityPress - - Busi­ness - DE­WALD VAN RENS­BURG de­wald.vrens­burg@city­press.co.za

Are most of the world’s key eco­nomic data wrong?

UK-based think-tank World Eco­nomics last week re­leased a new Data Qual­ity In­dex (DQI) try­ing to show that, the­o­ret­i­cally, most of the world’s al­limpor­tant GDP num­bers are wrong.

The DQI goes fur­ther, by rank­ing 154 coun­tries in terms of the “qual­ity” of the GDP num­bers they pro­duce (see side­bar).

African coun­tries clut­ter the bot­tom of this rank­ing ta­ble, fol­lowed closely by South Amer­i­can coun­tries.

The broad lessons from the in­dex are that con­di­tions the­o­ret­i­cally tied to bad eco­nomic data ex­ist ev­ery­where – ex­cept in North Amer­ica and Western Europe.

The DQI very point­edly gives China a ter­ri­ble score, far lower than many African coun­tries.

Like the sta­tis­tics be­ing ques­tioned, how­ever, the think-tank’s in­dex also re­lies on a cou­ple of largely untested pre­sump­tions.

Morten Jer­ven, a Nor­way-based econ­o­mist who courted con­tro­versy on the African con­ti­nent with his 2013 book Poor Num­bers, is dis­mis­sive of the ex­er­cise.

“It adds very lit­tle value,” he told City Press of the new DQI this week.

Jer­ven said he had been ap­proached to con­trib­ute to the in­dex but found the pre­dic­tive pow­ers of its com­po­nents un­re­li­able.

Jer­ven’s book doc­u­mented the in­ca­pac­ity of many African states’ sta­tis­ti­cal of­fices.

He also chal­lenged the blind ac­cep­tance of eco­nomic data used to eval­u­ate de­vel­op­ment poli­cies or to pro­claim that Africa is “ris­ing”.

The Nor­we­gian went on to face ex­treme crit­i­cism from African state statis­ti­cians, es­pe­cially Stats SA’s statis­ti­cian-gen­eral Pali Le­hohla, and he was heck­led when he ad­dressed a sta­tis­ti­cal con­fer­ence in Gaborone, Botswana’s cap­i­tal, in 2014.

Talk­ing to City Press this week, he said the DQI “bor­rows bi­ases from other data sources”.

One-fifth of each coun­try’s score is com­posed of the score they re­ceived from the non­govern­men­tal or­gan­i­sa­tion Trans­parency In­ter­na­tional’s an­nual Cor­rup­tion Per­cep­tions In­dex.

As the the­ory goes, a cor­rupt gov­ern­ment is more likely to cook the books.

The cor­rup­tion scores, how­ever, re­ally just re­flect the prej­u­dices of the busi­ness ex­ec­u­tives sur­veyed, claims Jer­ven.

“Us­ing sec­ond-hand data to eval­u­ate sec­ond­hand data is not go­ing to get you very far.”

Another fifth of the score de­pends on the of­fi­cial GDP per per­son in the coun­try.

Apart from the cir­cu­lar­ity of judg­ing the qual­ity of GDP data by ac­cept­ing it at face value, this mea­sure also as­sumes that a lower GDP per per­son au­to­mat­i­cally means fewer re­sources for its na­tional sta­tis­ti­cal of­fice.

Densely pop­u­lated China gets a low score on this count, de­spite be­ing ar­guably the largest econ­omy on earth.

Us­ing the DQI method, the eco­nomic su­per­power’s abil­ity to fund sta­tis­ti­cal work is pre­sumed to be on par with that of South Africa and Costa Rica.

Another less con­tentious com­po­nent is the age of the coun­try’s base year for cal­cu­lat­ing GDP.

The po­ten­tially huge dis­crep­an­cies be­tween of­fi­cial eco­nomic mea­sures and re­al­ity – be­cause of out­dated base years – have been dra­mat­i­cally il­lus­trated by re­cent GDP re­bas­ing ex­er­cises.

Ghana re­vised its GDP up­wards by 60% in 2010; Nige­ria did the same, by 90% in 2014.

Ac­cord­ing to Jer­ven, how­ever, the age of a coun­try’s base year gen­er­ally has rather low pre­dic­tive value, aside from spec­tac­u­lar ex­cep­tions.

He adds that re­bas­ing can drag GDP num­bers in the other di­rec­tion too.

For in­stance, Zam­bia had pre­vi­ously re­based its GDP from 1970 to 1977. The for­mer was a boom year for cop­per. The re­sult was a far less im­pres­sive set of data.

The rea­son GDP growth is gauged from a base year is to avoid the need to com­pletely sur­vey the econ­omy ev­ery year. Con­se­quently, that year’s prices will de­ter­mine the weight­ing of dif­fer­ent sec­tors in sub­se­quent years, even if prices change con­sid­er­ably. It also means new sec­tors will be cap­tured prop­erly only when a re­bas­ing is done.

The­o­ret­i­cally, bad GDP data have mon­u­men­tal knock-on ef­fects.

It is more or less the uni­ver­sal mea­sure of gov­ern­ments’ eco­nomic suc­cess, and poli­cies are rou­tinely judged by what they do to GDP.

The world’s coun­tries are di­vided into de­vel­op­men­tal cat­e­gories prin­ci­pally on the ba­sis of GDP per capita, which de­ter­mines el­i­gi­bil­ity for trade pref­er­ences and aid.

GDP can be the ba­sis of claims in the in­ter­na­tional arena, such as Brazil’s push for more power in in­sti­tu­tions such as the In­ter­na­tional Mone­tary Fund on the ba­sis of its GDP be­ing larger than that of Euro­pean mem­bers, which none­the­less have more votes.

GDP is also rou­tinely in­voked as a sym­bolic proxy for coun­tries’ rel­a­tive power and im­por­tance – for ex­am­ple, the con­test be­tween the US and China to be the largest econ­omy in the world, and the same con­test in sub-Sa­ha­ran Africa be­tween Nige­ria and South Africa.

Newspapers in English

Newspapers from South Africa

© PressReader. All rights reserved.