1) They focus more on potential revenue loss (from trade tariffs, VAT and CIT) rather than capital flight, acknowledging that the calculations show apparent flows in both directions
-
-
Show this thread
-
(although... I'm not sure how the VAT calculation works? If you evade import VAT unless the importer is the final consumer someone still pays the VAT?... also why does import overinvoicing reduce CIT, but import underinvoicing does not increase it?)
Show this thread -
2) They are based on detailed COMTRADE data - commodity by commodity and using import-export country pairs, rather than balance of payments.
Show this thread -
This allows the data problem of transit trade to be isolated and removed from the data (these are problems i've written about before https://hiyamaya.wordpress.com/2016/07/20/misinvoicing-or-misunderstanding/ …). GFI calls these 'lost' and 'orphaned' trades
Show this thread -
3) They look at individual pairs of data and consider whether the mismatch is plausible -- implausibly large mismatches in volumes more likely due to reporting error than misinvoicing are down-weighted.pic.twitter.com/ghWXnuo94Y
Show this thread -
4) They break the results down by commodity and by trade partner, so you can see the story the data is telling. In Kenya, the big numbers relate to imports of worn clothing, cereals, electrical equipment, vehicles and fuel.pic.twitter.com/1qxXSLIOy4
Show this thread -
But, as they highlight, these are not the commodities with big apparent misinvoicing price gaps.pic.twitter.com/5t53kEmhTQ
Show this thread -
Similarly the countries that are the biggest contributors to the absolute numbers - China, India, Pakistan, UK do not necessarily have the biggest apparent misinvoicing gaps (apart from Pakistan)pic.twitter.com/Cj1D8azVuH
Show this thread -
5) There is a whole discussion about transport & insurance markups -- should it be 14%, 10%, 6% ? GFI use a model which takes into account different products, countries etc.. which averages at 9% for Kenya
Show this thread -
Many of the major areas of apparent misinvoicing (vehicles, electrical equipment, fuel) are underinvoiced by 1-3% ..... this seems like it could be within the range of uncertainty about transport costs
Show this thread -
Used clothing and cereals have larger misinvoicing margins. There is a nifty graphic which shows absolute and relative size of the apparent tariff losses by product and countrypic.twitter.com/AW3Yd3laZ1
Show this thread -
This seems to me like a useful risk assessment tool. These are specific red flags to follow up on. Are there innocent explanations for the patterns in the data or is there something dodgy in these specific trades? (like whats up with used clothing exports from UK to Kenya?)
Show this thread -
I'd like to see what the potential losses in different categories look like with different modelling assumptions on the transport costs -- how robust are they? But all in all a lot of improvement in the methodology.
Show this thread -
Its not quite so clear in the Nigeria case - again they focus detailed analysis on import underinvoicing (absolute numbers driven by small margins on large volumes of vehicle imports). But import underinvoicing is not the biggest category in Nigeria...what is the export story?pic.twitter.com/t2gX52TCWc
Show this thread -
Kudos to
@GFI_Tweets for working to address the methodological issues and improve the estimates (they could publish the underlying data and calculations too!)Show this thread
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.