4 Comments

I was recently reading a post over at Construction Physics that did a similar analysis with regards to construction productivity. The stuff that buildings are made of is generally large and heavy, so transportation costs can be substantial. Unlike most industries, there are benefits to local production even if this limits the scale of production. Sometimes the gradient is so extreme that it makes sense to do manufacturing on site as opposed to at some efficient central or regional facility. (for example, concrete skyscraper floors)

Expand full comment

>>This failure of trade to grow between the 1910s and the 1980s is, you might say, the dog that didn’t bark. <<

What I see from the chart is that WWII was a huge, huge event, and following it global trade has been on a meteoric rise, and has done so continuously for almost 80 years now.

So better to say that trade dropped catastrophically in the 1940s and despite very rapid growth, there are limits to how fast that growth can be under even the best of circumstances (e.g., oceanic containers are awesome, but you can only redesign ports, create a container shipping fleet and build a manufacturing base in China so fast).

So deglobalization requires the most destructive war in history, but other than that it's been a pretty steady upward process for approaching two centuries now? Do I have that right?

Expand full comment

He did say that: "That is, there was a surge in protectionism in the interwar years, then a gradual process of liberalization — mainly through trade agreements — after World War II."

It is the interwar slump that was responsible for the rapid decline in trade. Post-war liberalization was the engine to restart trade growth. What I am not sure about is what counts as exports these days. It has been suggested that much of the US-China trade is just US companies shipping components to China for assembly and addition of parts and reimportation into the US. Is the shipment out of the US counted as exports, and shouldn't they be counterbalanced by the final reimported goods if they are?

The problem I have with Brad's model is that it assumes perfect substitutability between domestic and imported goods. This is a gross simplification. It may have applied when economies produced lots of commodities, like steel, but hardly in a world of highly differentiated goods sold via marketing of features and brand quality.

Also, I don't know whether Ricardian (comparative advantage) trade still exists, but if it does, it is about the relative costs of goods in countries that drive trade, irrespective of absolute costs. Does it all wash out in some averaging, or does this undermine the model entirely?

Expand full comment

Of course it is "a gross simplification". That's why it is called "A toy model".

And I think that the point is relevant to comparative advantage. The reason is that, if there are costs involved in trade, then there is an added wrinkle to comparative advantage.

For example (also oversimplified), if some state can produce widgets at cost X, and in my state it costs X + a, then trading for widgets makes sense - unless the trade itself (transport, etc.) makes the end cost for traded goods X + 2a.

The Ricardian model (if I understand it correctly) suggests that it is always (?) reasonable to trade for the goods where one is least productive - because you can then expend your efforts where they are more productive. But this model assumes that trade has no costs. But if the trade itself imposes costs, then it can be the case that these costs more than outweigh the opportunity costs of less-productive producing.

(This is too short and possibly unclear, but other things demand attention.)

Expand full comment