I'm more interested in the original claim that there were no (zero, nada, < epsilon) conservatives at 18F.
Particularly if the tent is broadened to include traditional pre-Trumpian conservatives.
Zero, as claimed, seems unlikely.
"Should" seems prescriptive.
Within a standard deviation of the mean general distribution for the US Federal service would be probable by the Central Limit Theorem.