It is strange, indeed. I have been reviewing what seems to be the original study: http://www.pewinternet.org/2018/05/31/teens-social-media-tec.... It includes some detailed figures about these disparities, but they don't seem to provide insights nor opinions on which reasons could be leading to those differences.
A wild guess here: assuming that there is a correlation between higher-income and higher-education groups (quite likely), it could be that more educated individuals are more informed about privacy issues, being more wary of their data. The lower Facebook usage find in both groups could be motivated by that (eg recent scandals, like Cambridge Analytica). This guess would explain lower Facebook usage among people with higher education (directly) and also among those with higher income (indirectly due to the assumed correlation with education). That's my two cents, although just a guess.
A wild guess here: assuming that there is a correlation between higher-income and higher-education groups (quite likely), it could be that more educated individuals are more informed about privacy issues, being more wary of their data. The lower Facebook usage find in both groups could be motivated by that (eg recent scandals, like Cambridge Analytica). This guess would explain lower Facebook usage among people with higher education (directly) and also among those with higher income (indirectly due to the assumed correlation with education). That's my two cents, although just a guess.