Thread by @Brad_Setser

An obvious question given today's weak manufacturing survey data is how much of the weakness in U.S. manufacturing stems from China's retaliatory tariffs.

The answer is less than you would think. US exports of manufactures to China haven't wildly underperformed Chinese imports

To be sure, US exports to China have underperformed overall US exports (but that's true for most countries, China's imports of manufactures have been weak).

And yes, chips are obviously part of the story

(2/x)

But US exports of manufactures to China are only about 1/10th of total US exports of manufactures, so mechanically a 2-3% fall in US exports to the rest of the world delivers a bigger negative impulse than a 15% fall in exports to China

3/x

of course, this is backwards looking (last data point is July) and lags the most recent intensification of the trade war. And the trade war indirectly impacts trade with the rest of the world (and US demand) as well.

4/x

One other complexity -- the biggest falls in US exports of manufactures to China have come from cars and planes. Cars is mostly tariffs (tho China's weak demand doesn't help). But the fall in aircraft exports is a self-inflicted wound (Boeing narrow bodies)

5/x

It is implicit in the previous graph but it is easier to see if you look at a trailing 12m sum -- the big fall off in US exports to China has been on the commodity side, not the manufacturing side.

6/x

There are of course backward linkages. Fewer soybean/ lumber exports to China = less demand for US made agricultural and forestry equipment.

But it also is an analytical error to ignore the drag from China's rising trade surplus/ falling imports on the global economy