Selection bias and China trade

1 minute read

* Note: this is a backdated post from an email I sent to friends, lightly edited for personal content and clarity

I’m not very familiar with Scott Sumner, but he wrote something interesting recently about “How do you fight against selection bias as you consume information about the world?”

I found the following interesting:

“Read material on both sides of the ideological spectrum, indeed on many different sides.” and “Read extremely smart bloggers, not people you agree with. Read people who annoy you.”

With the caveat that the sources are logical and well thought out. There’s a lot of trash. e.g. an argument for why raising minimum wage could be a bad idea might be worth reading. A book on how the earth is flat is…probably not

“Read about a wide range of topics”

I agree, I think trying to remain curious is a good trait

“Travel is another good source of information. If you travel to China and speak with the people you meet, it might give you a very different view of the country than what you get reading about China in the US media.”

I wish I could do this more

Pettis is also one of the best writers on China I’ve come across. He did a recap of what options are available to China now. It’s a very long read, especially if you dig into his accompanying prior linked articles, but logical and well written. Below are some highlights:

If the trade situation causes a contraction in China’s current account surplus, by definition that’s a reduction in the savings-investment gap. i.e. either investment share of GDP rises, or savings share declines, or some combo of the two

China has only four options available to them. 1) Raise investment, 2) reduce savings by letting unemployment rise, 3) reduce savings by allowing debt to rise, 4) reduce savings by boosting household consumption

Understanding that this is an economic fact rather than opinion was a major revelation for me in revising my framework on China.

Lastly, this paper seems to have an interesting finding: “We found that delay discounting is more predictive of income than age, ethnicity, or height” . I didn’t read the paper in full but the intro / conclusions might be worth looking at

Telegram