Note that there are different types of intelligences, e.g., Gardner's, and 'practical intelligence'
There are several possible reasons:
Test bias?
Genetics?
Environment?
Heredity is a factor within groups, accounting for perhaps 30-50% of variation (not the 80% argued by Jensen), but: there is no evidence that it accounts for differences between groups (e.g., blacks and whites).
Heritability is not inevitability (there is a very large reaction range)
Also flaws of 'reification' (believing
that there really is something called an IQ) and 'ranking' (assuming this
'thing' can be quantified in such a way that individuals and/or groups
can be ranked along the measure)
There are no
conclusive data demonstrating that group differences in IQ are genetically
related.
To learn more, I encourage reading of the APA release on intelligence. It can be found here.