Evidence in Social Science

When discussing evidence it is a good reminder to clarify what exactly evidence is. People tend to think evidence is equal to some sort of single clue clarifying everything. It is suitable compare with e.g. a crime investigation. You may discover footprints at the scene similar to a suspect. Does that prove that the suspect committed the crime? No it doesn’t. Does that mean it isn’t evidence? All too frequently say that this means the footprints are not evidence. But that is wrong. Evidence does not need to be conclusive. In fact it almost never is. Evidence is a is every piece of collected fact that contributes to strengthen a particular hypothesis.

Thus with respect to Gladwell we can find plenty of evidence. Whether all of its adds up to a certain conclusion is another question. But we must also keep in mind that especially in social sciences we can never know anything with absolute certainty. However we can pick a theory which has more evidence or a better explanation than competing ones.

I don’t buy Gladwell’s explanation, and he really doesn’t provide any actual evidence. It’s all just speculative.

If we keep in mind my definition of evidence, we can safely say that he does in fact provide plenty of evidence:

1. He points to the historical record of how serfdom was never successful in China, in part due to the amount of diligence required.
2. He compares the amount of hours of work required to do agricultural work in China compared to Europe.
3. He points to incentive structure. More work produced higher output in China, but not really in Europe.

Also what Gladwell really does is provide a scientific theory. That means a hypothesis with supporting evidence, with a framework of an explanation and with predictions. There can be many theories for something but usually how we rate the quality of a theory is by how good its predictive powers are and how well it generalizes. Turns out this theory is good on both accounts.

E.g. it does not just explain Chinese success but also Jewish success, as well as Irish problems. Both the Chinese and Jews lives in a cultural context where more work almost always gave more profit, while the Irish was in an opposite environment. We also see the same with slavery. Adam Smith examined this already in 1776. He had observed early that black slaves where almost never used for agricultural production. The reason for this was that slave based agriculture is inefficient because a person who does not directly benefit from extra effort will not do a good job. This was observed repeatedly through history. E.g. it was recorded about the Roman Empire how agricultural output in Italy fell in every area where farming to taken over by large slave estates.

Likewise it was remarked how descendants of free blacks, performed far better in the US, than descendants from blacks who had lived through slavery. In short we have a theory where we see clear consistent patterns coupled with a sensible and logical explanation.

You cannot find competing theories which are better.

We know from studies of animals, that cold weather selects for intelligence.

No we don’t. In fact the people most known for pushing this theory are race scientists attempting to prove white supremacy. It is not a hypothesis supported by serious scientists. As I’ve said for a theory to work it has have multiple pieces of supporting evidence as well as being able to supply a compelling explanation and provide good predictions. This theory doesn’t do that.

In addition, at the time of the building of the railroads, Chinese who immigrated because of the Nanking Rebellion killed about 20 million Chines. They had to work hard or face possible deportation or worse.

That may be so, but Chinese outperforming other immigrant groups, and being known for their hard work is quite consistent through history. Not just in the US but also elsewhere.