I would like to draw the attention to the precise wording around the statistics:

""" In 1970, just 1 in 5 U.S.-born PhD graduates in economics had a parent with a graduate degree. Now? Two-thirds of them do, [...] """

These numbers (1/5, 2/3) refer to US-born PhD graduates (in economics).

But the number of US-born PhD students is very small. The vast majority of PhD students are immigrants.