There is a statistic measurement in population genetics known as Sewall Wright's population structure statistic (AKA the F-statistic) which calculates the heterozygosity of different populations in a species. That is, based on a large set of genetic samples, it can tell you what degree of individual variance is attributed to between-group factors, rather than within-group and species-wide factors. The F-statistic for humans is actually extremely low, generally between 10 and 15%. As the vast majority of DNA has no actual function in the phenotype, limiting the calculation to nucleotyde polymorphs (the most common form of mutations) that actually could have an effect on development resulted in an even smaller F-value. Essentially, this means that the functional differences between different "races" are statistically insignificant when measured against overall human variation. There certainly is not enough net genetic difference to conclude differences in behaviour or intelligence, which are extremely complex traits requiring interactions between hundreds if not thousands of different genes.
(Interesting side fact: a single population of chimpanzees has more genetic diversity than multiple human populations scattered across different continents. This is most likely because humanity suffered from a major population bottleneck in relatively-recent evolutionary history before migrating to other areas of the planet.)

https://www.ncbi.nlm.nih.gov/pubmed/19687804
https://www.ncbi.nlm.nih.gov/pubmed/3293728
https://www.ncbi.nlm.nih.gov/pubmed/14655871
https://www.ncbi.nlm.nih.gov/pubmed/23185452
https://www.sciencedaily.com/releases/2012/03/120302101706.htm

Now, you're inevitably going to try and move the goal-posts from genetics to something else, like sociological factors (probably because you can't actually be bothered to read any of my sources), so since I'm never going to talk to you ever again I'll do my best to cover that as well, although biology is more of my particular specialization.
I forget if it was you or someone else who kept bringing up The Bell Curve, but the following information is directed at the neo-racialist "data" Arthur Jensen tried to perpetuate. There are certain studies that attempt to establish racial differences with "IQ tests". To begin with, by 2018, IQ tests are seen as outdated and unscientific for measuring any precise difference between different groups (although it can still be accepted as a measurement of cognitive growth or decline within an individual). More than differences in intelligence, IQ tests measures familiarity with the specific test-taking process (biasing it toward those from specific forms of academic training). Rather than identifying differences in intelligence, these IQ scores reflect differences in education (and not even positive or negative differences, just structural differences from the system the test is based on) between difference groups.
The Flynn Effect (the gradual increase in IQ scores over time) reflects this weakness of the test. As education becomes more standardized, not only are students overall scoring higher (as schooling more closely reflects IQ testing procedures) but differences in testing between groups are becoming less and less significant.
On the other hand, while there is no valid evidence for any differences in ability between "racial" groups, there is a whole lot of evidence that the perception of race can bias the behaviour of others. One common form of this is "audit" studies, where two groups are prepared that are equivalent in every way except for apparent race. Despite having exactly the same ability, when presented to supposedly neutral qualifiers such as hiring managers or recruitment offices at universities, it has been shown time and time again that certain "racial" groups are valued less than others. How can you try to claim "reality is racist" when there's literally evidence that racial prejudice, on a large scale, is independent of any real ability?

https://www.journals.uchicago.edu/doi/10.1086/262014
https://onlinelibrary.wiley.com/doi/abs/10.1111/0019-8676.00131
http://psycnet.apa.org/record/1987-17534-001?doi=1
https://academic.oup.com/sf/article-abstract/93/4/1451/2332119?redirectedFrom=fulltext