More on Oil Import/Export statistics

In response to yesterday's post, several commenters noted that there's an apparent discrepancy between the EIA numbers for oil import/exports, and the numbers in the BP Statistical Review of World Energy.  The two sets of series are above, but in the BP case, imports and exports are identical to the barrel, suggesting they are deliberately harmonized, rather than being independent estimates.

The reason for the discrepancy becomes fairly clear if we look at oil exports from out of the United States.  For 2007-2009, EIA has those at 27kbd, 29kbd, and 44kbd - very small numbers, reflecting the fact that the US is a major importer of crude oil, not an exporter.  Presumably, there are a few wells somewhere that are easier to connect to a pipeline in Canada or Mexico than in the US, but that's about it.

However, the BP spreadsheet shows US exports for the same years as 1439, 1967, and 1916kbd.  These are much larger numbers, and strongly suggest that the BP numbers include trade in oil products, not just crude oil itself.  As Gregor McDonald noted recently, US exports of refined products are increasing recently as US consumers are having to conserve, and thus are not using all the products of American refineries, some of which are now being reexported elsewhere.

So presumably the BP numbers overall are higher because they include export of oil products, rather than just of crude oil.  In some cases, this may be double counting (for example oil coming into the US and being reexported as refined diesel), but in other cases not (for example, Saudi Arabia in recent years has exported 1.2-1.4mbd of refined products, and those exports would be counted in the BP line and not the EIA line above (which was only for crude-and-condensate exports/imports).

At any rate, whichever way you measure it, all of these series show declines in the last couple of years.