[karpierz]

Jamaicans are even more prone to be poorer than Americans.

Would your retort be "but Jamaicans who've won multiple gold medals in sprinting are wealthier than most Americans?"

People who smoke are more likely to get lung cancer.

Would your retort be "13 year olds who smoke have less incidents of lung cancer than the non-smoking population?"

If you subsample enough, you'll find outliers. That doesn't invalidate general trends/statistics.

[tomp]

No, but it does introduce other variables.

E.g. age is important for lung cancer as well as smoking. Only when controlling for age do you get correct estimate about the danger of smoking.

What other variables does "Nigerian immigrants" introduce besides skin color and culture? The latter is eliminated by most US sociologists that keep screaming "it's not about culture!" so only race remains...

[InGoodFaith]

Isn't the fact that they immigrated likely bias towards the higher skilled echelon that managed to go through the visa process?

When you say "outperform most domestic US groups", are you normalising for the same skillset and education levels?

Perhaps a source to accompany your statement will help shed some light.

[karpierz]

Don't forget to control for:

- Wealth upon entry (If all Nigerian immigrants come from millionaire families, then it doesn't matter if they're underpaid; they'll be far wealthier than most Americans)

- Location (If all Nigerian immigrants live in New York City, then the effect size of being underpaid can be masked by the higher income (and cost of living) in New York City)

- Profession/Job Title (if all Nigerian immigrants are neurosurgeons, then it doesn't matter if they're underpaid; they'll be way above average income)

- Age (if all Nigerian immigrants are older, then they'll have had more time to generate skills/wealth, which could mask the effect of being underpaid)

- Health (if all Nigerian immigrants are healthy, then the effect of being underpaid could be masked by the fact that many Americans are less productive due to health issues)

There are obviously more confounders, but doing an analysis using these would be a good place to start to actually answer the question.

[hn_throwaway_99]

That's not at all how I interpreted the parent comment.

I read it as the types of hardships and barriers that, say, a US descendant of slaves faces can be very different from a Nigerian immigrant, and that's one problem with lumping everyone with the same skin color together as "Black" when calculating diversity metrics.

[tharne]

This reminds me of an individual from South America I worked with many years ago. Some well-meaning individuals were asking him how the company could reach out to its Hispanic employees better. He looks at them and wryly says, "Well to start, you could stop thinking of us as 'Hispanic' since it encompasses people from 3 different continents, a dozen different countries and doesn't mean much."