I read the entirety of The Bell Curve, and checked over some of the statistics behind the models, so I think I can claim to understand it.
The book makes the following arguments:
1) There is a significant positive correlation between IQ and economic success.
2) There is a significant positive correlation between measured IQ and
3) A substantial part of IQ is hereditary. 4) To the extent that social and ethnic groups represent distinct gene pools, they can have differences in average measured IQ.
All of these propositions were based on hard data, which was subjected to exhaustive statistical analysis to filter out other correlations. The IQ effect remains fairly large even after correcting for education level, social status, age, and all sorts of other factors.
Of course, the authors were immediately denounced by the American Left, which decrees that all people are identical and the only possible explanation for any difference in outcomes is discrimination. Nevertheless, their work remains the closest thing to scientific proof that I have ever seen in the social sciences.