Wallonia's CAN trade continued

A couple of questions have been posed regarding the graph I developed for Wallonia’s trade share with Canada. Here are some of my explanations and further thoughts on the issue.

One remark that was made related to the fact that these figures would be biased because Flanders has Antwerp. Antwerp, like Rotterdam, are known for what trade-economists call their “entrepôt” activities. These two cities import lots of goods because of their ports, store them, and then re-export them again to other parts of Europe, i.e. the hinterland, which I indeed allude to in the piece.

This bias is true, but only to some extent. The NBB source that separates regional trade between Flanders and Wallonia on the one hand and Canada on the other presents data at a level where these imports and re-exports of non-residential entities are excluded, i.e. pure transit trade appears to be out. 

Transit trade of goods that arrive in Flanders’s main port are therefore as it seems left out. The difference is a quarter. Although I doubt all transit trade is excluded from here, it’s nonetheless hard to imagine that this change is solely made up by non-residential expats. A big chunk must come from “Antwerp”.

However, these clean regional trade numbers do not match with trade figures from Eurostat / OECD, which is needed to sort out the two Belgian regions’ EU trade share with Canada. They provide data including these entrepôt activities.

Now, since Eurostat / OECD figures includes this transit trade one must take a consistent share of Belgian’s total trade as part of the EU with Canada because other countries such as the Netherlands also have transit trade. Belgian’s share in total EU-Canada trade is 7.55 percent, which includes both imports and exports.

Going back to the NBB data, Belgian’s regional division of trade is 1.38 percent imports and 9.5 percent exports for Wallonia against Flanders. The latter percentage was also shown in Politico’s bar chart, which actually got me started. I used similar way of coming to a regional total trade division by summing up imports and exports, which gives a figure of 4.4 percent, but added Brussels and rounded up to 6 percent.

With these numbers at hand, one can sort out the rough back-of-the-envelope calculation which I give in the pie chart. Simple rough computations, yet indicative with the available information at hand. Of course, all this computed with real trade data from all these sources.

Although Wallonia’s part could be bigger by being even more generous, I don’t think that would significantly differ the main pattern. My guess is that Wallonia just trades much less precisely for reasons that is has not a port, even after correcting for Antwerp’s pure transit re-exports.

Today’s trade is intermediate input trade with unfinished goods that cross borders multiple times around the globe. It is also called global supply or value chain trade. In total, almost 60 percent of total trade is intermediate. In each stage of the production line the input is imported and processed by adding value and then re-exported again to finally arrive at the point of destination.

Flanders has a port, but also many industries that import intermediate goods, process them, and quickly export them again. Value is therefore added, which should not be disregarded. On the contrary, the global value chain literature has made a huge turn-around in trade studies by precisely emphasizing this type of trade, and warns us not to stay focused on finished goods trade.

Hence, these questions touch upon a more fundamental issue. Leaving out pure entrepôt activities is understandable as I have done to the extent possible, but one should not forget that precisely because Flanders has a port it trades so much more in other types of goods as well, not only pure transit goods.

On yet another level these questions also hint to the fact that trade figures are in “gross” terms, which also plays a role for Wallonia’s dwarfed numbers. Gross figures count the total value of the good twice: when it comes in, and when it goes out at greater values.

If anything, this constitutes the real distortion and the recent value-added approach that the trade literature has been taking is a good start to correct this interpretation. Unfortunately, however, developments in this area are still far away from sorting out regional value-added trade and for most recent years.

Turning to the figure itself, the purpose of the graph is to negate the fact that the Wallonia government claims it is a matter of trade, i.e. an economic issue. On the contrary, it exports 4 times more to Canada than it imports from it.

Moreover, opposition to trade agreement often revolves around industries that would rather not like to see greater levels of competing goods coming into their country. However, it’s hard to think that the 1.38 percent of Wallonia’s import share from Canada within Belgium comes along with that strong efforts of lobbying.

Therefore, other reasons must be at play which was also pointed out by Paul de Grauwe. Domestic politics is an obvious candidate, but I would also add the role of ideas. 

An often-heard worry in Wallonia is that with trade agreements in place big multinationals are coming and therefore can form a threat. Even though big foreign companies bring a lot of good to the local economy, these arguments should nonetheless be taken seriously.

Yet my last check was that the Commission did take these concerns into account, particularly regarding the ISDS. So again, my guess is that it all comes down to domestic politics.