What the sources say
In 2025, the Trump administration signed an executive order criticising the Smithsonian Institution’s exhibition framing race as a social construct — calling it “harmful ideology.” This brought a scientific question into sharp political relief: is race a biological reality, or a social and historical construction? The answer matters for medicine, policy, and culture. This page examines what the science actually shows.
of human DNA is identical across all people. The Human Genome Project (2003) provided the definitive finding.
of all human genetic variation exists within populations, not between them. More variation within any “race” than between races (Lewontin, 1972; confirmed repeatedly)
racial categories that correspond to discrete genetic clusters when mapping global human variation — variation is clinal (gradual and continuous), not stepped
number of “continental ancestry clusters” typically identified by algorithms — but this number changes depending on parameters, and does not match social racial categories
Discrete, bounded biological categories corresponding to socially defined races. The claim that “white,” “Black,” “Asian” etc. are scientifically valid taxonomic categories with clear genetic boundaries. This is what the Human Genome Project, modern population genetics, and the scientific consensus consistently refute.
Real, measurable differences in allele frequencies between human populations. Genetic ancestry can be estimated with high accuracy. Some health conditions are more prevalent in certain ancestry groups. These are scientifically valid — but they do not map onto racial categories as socially defined.
Statistical groupings of people who share common ancestors. Real and medically relevant. But: they are continuous, not discrete; they vary by how many clusters you ask the algorithm to find; and they do not correspond reliably to socially defined races.