Publications
Partners
Highly influential global estimates of illicit financial flows (IFFs) are complete nonsense designed to overstate private sector complicity in underdevelopment, economists contracted by the Chamber of Mines declared this week.
These estimates include the 2015 report from Thabo Mbeki’s High-Level Panel formed by the AU and UN Economic Commission for Africa. It had popularised the accusation that Africa loses $50bn (R658bn) per year to IFFs – mostly in the form of trade misinvoicing by companies.
It also includes the ongoing work of US nongovernmental organisation Global Financial Intergrity (GFI), which has been pivotal to catapulting the idea of these multibillion dollar IFFs into development orthodoxy.
More specifically, it includes a controversial study from the UN Conference on Trade and Development (Unctad) last year that claimed South Africa’s gold was almost entirely “smuggled” out of the country and that significant amounts of platinum and iron exports were also undeclared.
Eunomix has succeeded in providing innocuous reasons for Unctad’s findings that mostly involve data quirks and errors.
City Press had, after the Unctad report, independently reported on some of these obvious errors.
Eunomix, however, strays far beyond these specific rebuttals.
‘An ideological dimension’
The whole concept of measuring trade misinvoicing is built on ideologically driven “pseudoscience”, “mumbo-jumbo” definitions and weak research, claims Eunomix CEO, Claude Baissac.
Baissac and the chamber are quick to say that they do not dispute that IFFs are real.
“No one disputes that there are large IFFs, but it does not show up as trade misinvoicing. Transfer pricing happens, we know that, but that has not been proven [by Unctad’s research].”
“I suspect they will say South Africa is a special case, but for the rest of the countries, it is true ... our theory remains.”
For Baissac, these estimates represent an attempt to understate the importance of corruption – and overstate the evils of the private sector.
“There clearly is an ideological dimension to this. It seems to me GFI has found a pseudoscientific methodology that supposedly proves that IFFs do exist on a massive scale – and lo and behold, it is mostly related to private sector.
“It allows the AU to say corruption is only 5% of it. We question the motives of this methodology.”
The mistakes
City Press at the time reported how the massive estimate for gold “smuggling” actually stemmed from a simple data error.
This followed an in-house recreation of Unctad’s work unrelated to the chamber.
The UN’s world trade database used for the study, Comtrade, does not register “monetary gold”, which is what most South African gold exports were classified as up to 2010. The re-export of gold from Ghana, Tanzania and Mali that gets refined at Rand Refinery seems to explain away the remaining discrepancies well and is documented by the SA Revenue Service (Sars).
Much of the platinum “misinvoicing” also had a straightforward explanation – the Comtrade database is missing two years’ data.
These problems were immediately apparent before the chamber started rebutting the claims.
Eunomix has come up with an explanation for the iron ore estimate as well, but SARS refutes this. Eunomix claims that iron ore exporters started reporting their exports on a cost-insurance-freight basis in 2011 instead of the normal free-on-board basis. This would explain data that looks like misinvoicing.
SARS however denies this in an email to City Press.
“SARS was never consulted nor requested to provide data for the study. SARS therefore refutes the claim by Eunomix.”
SARS however added that it had raised “concerns” with Unctad and agrees with Eunomix that the report was “theoretically, factually and empirically incorrect”.
Now what?
The chamber wants a formal retraction of the report, but Unctad this week tweeted that it stood by its “landmark report”.
These estimates have evolved to become fairly complicated calculations.
At heart, however, the premise is simply that country A’s declared exports get compared to its trading partners’ declared imports.
The difference is presumed to be misinvoicing if it is larger than a certain margin, usually 10%, to accommodate the fact that transport costs usually get reflected in import statistics.
With most of the South African estimates debunked, questions can be raised about estiamtes involving the rest of the world.
“In the process, we came to a conclusion that there is limited value in trying to do a counterfactual penny for penny,” said Baissac.
“We have done that ... but it is actually quite meaningless because the real problem is the methodology. It just does not work. In the case of South Africa we have demonstrated it is false.”
Accusations and Explanations
Consultancy Eunomix this week released its final report, commissioned by the chamber, to rebut an Unctad study claiming to have discovered enormous trade misinvoicing by South African mines.
It was conducted by University of Massachusetts economics professor Leonce Ndikumana and was first released in July last year.
It found that South African mines had, between 2000 and 2014, underinvoiced a staggering $78bn in gold exports, $24bn in platinum exports and over $4bn in iron ore exports.
The theory is that these “undeclared” exports amount to capital flight and tax evasion.
Read Fin24's top stories trending on Twitter:
