I don’t understand this graph

February 21st, 2017 at 7:58 am

I’ll have a lot more to say about this later, but I don’t understand this figure from the front page of my WSJ this AM.

Source: Wall St. Journal

The idea, which has some merit but is easily abused, is that the Trump admin wants to refine the way we measure trade flows–imports and exports–by accounting for re-exports, goods made elsewhere that pass through our country en route to somewhere else.

There’s a cogent argument that these goods should not be treated as exports, since they were not made here, and thus did not involve the economic activity of regular exports (other than some port/warehousing activities, I guess).

But here’s the thing, and the reason I find the figure above confusing, though I may well be missing something and will read up on this ASAP: if you’re going to disregard the exports, then you must also disregard the imports. That is, when a good is re-exported, I see a rationale for subtracting its value from US exports. But, as a statistician emphasizes in the article, that same value must also be subtracted from US imports.

That won’t change the overall trade balance, but it will change our bilateral balances with our individual trading partners. From what I’ve seen, for example, our trade balance with Mexico will worsen, but our trade balance with some Asian countries might improve (implying imports from Asia that pass through the US on their way to Mexico).

So, a) I don’t get why the WSJ figure appears to show a larger overall trade deficit, and b) you just have to be really nervous when team Trump starts fooling around with this sort of thing.

More to come…

Print Friendly

5 comments in reply to "I don’t understand this graph"

  1. pgl says:

    “if you’re going to disregard the exports, then you must also disregard the imports. That is, when a good is re-exported, I see a rationale for subtracting its value from US exports. But, as a statistician emphasizes in the article, that same value must also be subtracted from US imports.”

    Granted but I think this is missing something:

    “That won’t change the overall trade balance”

    If the price of re-exports exceeds the price of imports which is often the case under many multinational transfer pricing schemes, then taking out these values will lower the trade balance. Not that I am saying that transfer pricing is right as it may not be. Which is the point of my Econospeak post.


    • thibault says:

      Well, if the price of re exports exceeds the price of imports, it means that you have added value somewhere in the process (be it real value added or artificial value added through the game of transfer prices) and i see no reason why this exported added value should be excluded from the trade balance… No? This proposal still looks like a non sense to me ! (=i share Jared s reaction 100%…)


      • pgl says:

        “i see no reason why this exported added value should be excluded from the trade balance”.

        Well said. I’m wondering when Brad Setser weighs in on this as he covers this important issue very thoroughly.


  2. Brad Setser says:

    generally speaking, transfer pricing is now used to keep profits out of the US — not to move them to a US sub or parent. So would be surprised if there is a major transfer pricing effect.

    As i understand it netting out both reexports and corresponding imports (on a bilateral basis) is hard for two reasons:

    1) some reexports are used goods, so they may have been imported many years ago (used A320s, used cell phones, etc)
    2) the folks who collect the data think that some of the reexports are showing up on the import side as goods for domestic consumption, so there isn’t a clean adjustment — see https://www.usitc.gov/publications/332/tradestatsnote.pdf

    basically goods for reexport clear customs in US rather than staying in bonded warehouses (e.g. legally offshore) and show up in the “standard” import data

    my solution (rough)is to just subtract exports from both exports and imports if you want to adjust — though this misses the first bit (e.g. some of this year’s reexports are prior years imports)


  3. pgl says:

    Brad – of course you are right. My Econospeak example was based on the alleged value of the Ford designs. Many US multinationals either lowball this value or parks the IP in places like Switzerland or Ireland. Of course the incentives there are to declare a high transfer price differential which would this difference would be yuuuge. And yes – it does show up in the balance of payment and national income accounting.


Leave a Reply

Your email address will not be published.