This raises the good point that some groups are more subject than others to being stereotypically conflated with specific subcultures within their group.
To take an example from a minority group I happen to belong to, namely Jews: About 12% of American Jews and about 17% of Israeli Jews are Haredi or so-called “ultra-Orthodox”. Haredi culture tends to valorize studying Torah as a lifework for men, and also encourages large families, which often limits career opportunities for women. They also participate less in secular formal education, meaning that their math, science, etc. skillset tends to be more limited.
Consequently, disproportionately high numbers of Haredi families rely on various forms of government assistance for low-income families. In Israel over one-quarter of all income in the Haredi community consists of government benefits. In the US, for example, the Satmar Hasidic village of Kiryas Joel, NY, which houses nearly 1 in 200 of US Jews, is notoriously one of the poorest cities in the entire US, and correspondingly reliant on government benefits like Medicaid.
Yet there aren’t any popular stereotypes that “Jews are more likely to need government handouts” and “Jews have a higher poverty rate than some other minorities”. The subculture of Haredi customs isn’t being conflated with the group “Jews” as a whole. (Some other Jewish subcultures, particularly those of highly educated urban economic-elite Jews, do get stereotypically conflated with “Jews” as a whole, as DemonTree’s Nobel-Prize stereotyping illustrates. But Jewish “welfare dependency” subcultures don’t.)
Black American populations, by contrast, include some subcultures such as urban gangs in which crime rates are very high, and those subcultures do get conflated with their larger group in popular stereotyping.
When the “average American” of stereotype thinking sees a Haredi family getting food stamps, they don’t link that to stereotypes of “Jews are more likely to be on welfare”. But when they see a Black gang member dealing drugs, they do link that to stereotypes of “Blacks are more likely to commit crimes”.
And when you consider that there are about 48 million Black Americans in the US and about 1.5 million gang members of whom about one-third are Black, it’s clear that the percentage of Black Americans in the “gang criminal” subculture is substantially less than the percentage of Jewish Americans in the “strict Orthodox benefits recipient” subculture.
Which just re-confirms how unreliable and biased it is to extrapolate from specific subcultures to the propensities of an entire group.