The only thing I want to say here is that folks should be careful interpreting this data, as it’s only descriptive (as opposed to inferential). There’s no information on whether these differences are statistically significant, and even if they are what the causes and drivers of them are. I think it’s safe to say that the average Carleton or Swarthmore student has more precipitating factors (initial desire, motivation, knowledge about PhD-level careers) and probably academic achievement/prowess than the average Beloit or Willamette student, but that says nothing about the outliers - for example, whether the same sufficiently motivated, superstar student would have better chances attending Carleton than Beloit.
This becomes important when students are comparing scholarship offers. For example, let’s say you’ve got a superstar student who has already started doing high school research and has a paper published and one in preparation. She’s got a full ride scholarship to Whitman and her family is full pay at Carleton. Would she necessarily have better chances going to get a PhD if she chose Carleton? No, that’s committing the ecological fallacy in statistics. A student who has a publication in high school could probably go practically anywhere and still come out on top.
The other thing these lists don’t do is stratify. Note that Spelman and Morehouse, both historically black colleges, are towards the middle of the list. Also note, though, that in other places analyses show that black science and engineering PhDs disproportionately come from HBCUs, and Spelman and Morehouse rank very highly when you take that into account (and also into account that very large proportions of their student bodies are low-income going by Pell recipients).
Would it be better for an aspiring PhD who is African American to go to Colgate or Morehouse, or Kenyon or Spelman? There’s no answer to that - it’s going to be very individual and depend a lot on the student. The same is likely true of many other minority groups - women, low-income students, Latinos, etc. (I can say that for my part, going to an HBCU - Spelman, in fact - was instrumental in my decision to get a PhD in a STEM, albeit not “hard” science, field. The fact that I had lots of black female professors in front of me in my classrooms, and a black female president and black female deans and black female alumni who were getting PhDs…it left a tremendous impression on me that could not have been replicated at a predominantly white school. I’m not saying I wouldn’t have gotten a PhD if I had gone anywhere else; I’m just saying it made a huge difference.)
Also, some colleges may either simply attract students who are less likely to want PhDs or foster a spirit in kids to go pursue other avenues. Note that Scripps, Claremont McKenna, and Pitzer, all elite LACs, are near the bottom of this list. Is there something about the atmosphere at the Claremont Colleges that encourages entering careers other than science & engineering research? Maybe! Would that mean that an aspiring PhD should attend Seattle Pacific instead of Scripps, Whittier instead of Pitzer, or St. Lawrence instead of Claremont McKenna? Eh, probably not.