Publications
Partners
Matthew Salomon, senior economist at Global Financial Integrity (GFI) in Washington, DC, told City Press that Eunomix’s work does not invalidate their research.
GFI is the foremost global proponent of mirror-trade analysis for identifying illicit financial flows (IFFs) and its estimates are widely accepted by the activist community.
Without disputing the empirical findings – that misinvoicing from South Africa is massively inflated when researchers use global databases such as Comtrade – he said the point was to make a “rough and illustrative global estimate”.
According to Salomon, the problem with South African gold-export data had cropped up in GFI’s own research before the UN Conference on Trade and Development (Unctad) or the chamber raised it.
In a report in 2015, GFI identified South Africa as the seventh worst country in the world in terms of its IFF outflows and claimed that about $21 billion per year was lost this way, mostly through misinvoicing.
This year, in a new report in May, GFI dialled back the South African estimate.
In its latest report on global IFFs, GFI has effectively demoted South Africa out of the list of countries whose detailed bilateral trade data can be trusted.
GFI’s new approach to South Africa would, however, still include most of the mistakes that have come to light in Unctad’s work, making it almost certainly still a vast overestimate of IFFs.
Salomon, however, lashes out at Eunomix for what he says is the unwarranted belief that one country’s data disprove the whole method and all the global estimates.
He calls Eunomix’ treatment of the methodology “deficient” because it conflates the method as a whole with specific uses of it.
“Statistical analyses using a variety of methodologies are always worth pursuing.
“The Eunomix report purports to have debunked the mirror-trade methodology,” he told City Press.
“I hope it is clear that GFI had no role in the Unctad research under review. Second ... this methodology has a long and distinguished pedigree in economics.”
Eunomix accuses GFI and others of using the huge “misinvoicing” estimates to present the private sector as a prime culprit in underdevelopment, but according to Salomon, the primacy of misinvoicing is not the “crucial determination” Eunomix makes it out to be.
“Sadly, the integrity of what Eunomix presents as disinterested analyses is undermined by such arguments,” said Salomon.
What happened in South Africa is precisely what should happen, he suggested.
“Mirror-trade analysis ... proceeds by a kind of process of elimination: After accounting for plausible and legitimate reasons for discrepancies in trade reports as much as possible, the researcher examines what’s left of the gaps to see if the remainder can be related to misinvoicing in a reasonable way. How this proceeds from there depends on the nature of the research question and the data available.
“Because research using Comtrade is increasingly undertaken worldwide, the complexities and quirks are now more widely known than has been true in the past, so a growing awareness of UN-Comtrade nuances and quirks benefit all,” said Salomon.
“I don’t believe the wholesale outlawing of particular methodologies is any bit less preposterous than it sounds.”
Read Fin24's top stories trending on Twitter:
