The Colleges Where PhD’s Get Their Start

What schools produce the most undergraduates who end up heading off to graduate school?

The subject came up yesterday because a friend of mine was telling me about a brilliant teenager who wants to eventually get a PhD in physics. The student lives in California, but the mom wants him to apply to schools in the Midwest where she grew up.

I asked my friend if the teenager had checked out Lawrence University.


I realize that might be your reaction, but here’s the thing – many of the schools that are feeder institutions for the nation’s PhD programs are liberal arts colleges. While most liberal arts colleges are not well-known among families with teenagers, these institutions — and their reputations — are very well known to graduate schools. Lawrence University, a liberal arts college Appleton, WI, for instance, happens to be 10th on the list among all four-year colleges and universities that produce, per capita, the most physics PhDs. (See chart below.)

PhD Feeder Schools

I bet most families assume that attending a public flagship university or a nationally-known private research university is the best ticket to graduate school. If you look at the following lists of the most successful PhD feeder schools for different majors, you will see a somewhat different story. Not a single public university makes any of the lists. The entire Cal State system, however, is considered the No. 1 producer of humanities PhD’s.

I thought you’d be interested in the list of undergraduate institutions that, per capita, produce the most PhD’s. I pulled the names off the website of  Reed College, a liberal arts college in Portland, OR, that is understandably proud of its impressive record of turning out undergrads who end up earning PhD’s. The statistics cover the years 1997 to 2006 and come from the National Science Foundation and the federal government’s education database.

On the list of schools that have the most undergrads who ultimately earn a PhD  in all disciplines, Reed comes in No. 3 behind Cal Tech and Harvey Mudd College.

As you’ll see in the first column below, seven of the 10 schools, whose graduates earn the most PhD’s by the percentage of students are liberal arts colleges.


Bottom Line:

Students can increase their odds of being accepted to graduate school if they earn their bachelor’s degree at a liberal arts college. On a per capita basis, for instance, liberal arts colleges produce twice as many students who earn a PhD in science than other institutions. That makes sense since students have more opportunities to work closely with their professors, they can actually learn more due to small classes and receive glowing recommendations from their teachers, many of which, graduated from prestigious graduate programs.

What’s even more remarkable about the prominence of liberal arts colleges on the science lists is this: many students major in other disciplines at liberal colleges while students who attend schools like Cal Tech and MIT overwhelmingly expect to pursue careers in the sciences and engineering.

I’ve attached a lengthy essay by Thomas R. Cech, a Grinnell College grad, Nobel Laureate and chemistry professor at the University of Colorado, who wrote a fascinating essay contrasting the science experience for undergrads at colleges versus universities. Cech believes science majors at liberal arts colleges enjoy an advantage over undergrads who attend universities. Here are Cech’s thoughts:  Science at Liberal Arts Colleges: A Better Education?

Learn More!lab

The No. 1 way to cut the cost of college is to become an educated consumer. You can learn how by attending my popular online course, The College Cost Lab. I’ll be relaunching the course in June and if you’d like to be notified when I have more details, please click here!



Let's Connect

Leave a Reply

  1. Pingback: How Important Is Publishing Before Grad School? - academical. | academical.

  2. From Thomas Cech’s article, Chem/Bio Chem Prof at UC Boulder

    Only about 8 percent of students
    who attend four-year colleges or universities are enrolled in
    baccalaureate colleges (a category that includes national lib-
    eral arts colleges).
    Among the students who obtain Ph.D.’s in
    science, 17 percent received their undergraduate degree at a
    baccalaureate college.
    Thus, these colleges are about twice as
    productive as the average institution in training eventual Ph.D.’s.
    On the other hand, these same schools trained only 4 percent of
    the eventual Ph.D.’s in engineering, so their productivity is half
    the average in that field. This is unsurprising, as few liberal arts
    colleges have engineering programs.
    A more detailed view is provided by considering students
    trained by the top national liberal arts colleges. The institutions
    listed alphabetically in table 1 are representative of the best in
    the United States. Examination of table 1 indicates that most of
    the nation’s top colleges educated one to three hundred of the
    students who obtained Ph.D.’s during the five-year period from
    1991–1995. These numbers put several of the liberal arts col-
    leges in the top hundred of all institutions in Ph.D. production
    (see “Rank” in table 1). However, most of the institutions
    ranking in the top hundred are research universities with typi-
    cal enrollments of twenty to thirty thousand students, whereas
    the liberal arts colleges typically enroll thirteen to twenty-six
    hundred, roughly tenfold fewer. Thus, to compare relative Ph.D.
    productivity of institutions of different size, the ratio of Ph.D.’s
    per hundred enrolled has been calculated. Note that this ratio
    is approximately equal to the percentage of baccalaureate de-
    gree recipients from the college who eventually obtain a Ph.D.
    in science or engineering. (Because it integrates five years, it
    would exactly equal the percentage if one-fifth of a college’s
    total enrollment graduated in any given year; considering attri-
    tion and the number of students who take more than four years
    to graduate, this is a reasonable approximation.) Thus, most of
    the top liberal arts colleges see between 5 percent and 18
    percent of their graduates going on to obtain a Ph.D. in science
    or engineering (table 1, last column). Considering that their
    graduates majored in English, history, art, and other humani-
    ties disciplines as well as in science, this represents an astound-
    ing percentage.

  3. D, You assume that everyone choosing whether to get a PhD. is doing so based on an economic analysis. No one I know made the decision that way. Instead, most PhDs are in it for the love of learning and the joy of life in academia. If anything, this mostly indicates that liberal arts colleges are better, for good or bad, at instilling in students a love of continued learning over other priorities.

    Any undergrad who is even remotely aware knows full well the salary a PhD earns. After all, the faculty at liberal arts colleges are PhDs, and their salaries are in no way hidden from the students. Perhaps those PhDs at teaching oriented schools are happier, and so more likely to encourage undergraduates to follow in their steps.

    1. Kas,
      Yes you are exactly right. Undergraduates from liberal arts schools only have interactions with professors at their schools, not graduate students and postdocs because these colleges do not have them. The liberal arts student sees that these professors enjoy teaching, have modest but reasonable salaries, and decent lifestyles. They work 9-6, teach, and wear sandals to work everyday. These liberal arts professors have a good job and are fairly happy.

      In contrast, the student from a research 1 university has contact with professors, graduate students, and postdocs. These graduate students and postdocs tell the undergraduate that their chances of actually getting a tenure track job are around 14% (this is for the life sciences, it is probably worse in humanities) by the time they are in their late 30’s. So the student from a research 1 university sees that obtaining a PhD is a bad investment. Not necessarily because the pay is low, but rather because it is very unlikely that they will ever get a tenure track job. Most of the PhD students will end up switching careers in their mid thirties. At age 35, they will not have earned any money, not started a family, not bought a house, and they will have to wear real shoes to work.

      My top ten STEM graduate department apparently had a hard time recruiting American students from top notch universities. Instead they had to rely on naive students from liberal arts colleges and foreigners. Once the liberal arts student gets there they realize that they will never get that tenure track job; because they never saw how graduate departments worked at their colleges.

        1. Lynn, Nice article!
          I also like the Washington post article from last week, Why a phd doesn’t equal a job.
          I am glad to see the popular press finally picking up the real story with phd programs.

      1. The last numbers to come about English, a discipline that is among the worst in terms of job prospects, has about 50% ending up on a tenure line within 5 years of PhD receipt. The job market is bad across the board, but something about that 14% figure seems off – is it that lots of life sciences PhDs willingly take private sector jobs?

  4. The true reason that more liberal arts schools produce more per capita PhD students is the fact that liberal arts students are not exposed to PhD programs during undergraduate (because their schools do not have them).

    When your at a research1 institution as an undergraduate you get first hand exposure to labs and PhD programs. These students realize pretty quickly (from this direct exposure to PhD students and postdocs) that getting a PhD is a poor investment of your time. For example, in the life sciences you can spend 12 years in training (PhD+postdoc) and then have a hard time getting a job that pays $45,000. Consequently, the students from these universities either go to professional school or enter the workforce directly. These are better options.

    Students from liberal arts colleges, however, never get exposure to the total scam that PhD programs are. These poor suckers end up being feeder schools to the PhD programs that cannot get better quality students from larger universities. To liberal arts students getting a PhD is still a good idea because no one ever tells them otherwise.

    The data from this study appears to suggest that liberal arts schools give good education because they produce PhD students (and maybe they are….thats not the point). But if you realize that getting a PhD is a bad investment then the interpretation of this data is very misleading.

    “Why Doing a PhD Is Often A Waste Of Time” (The Economist)
    “Graduate school in the humanities: Just don’t go” (Chronicle of Higher Education)
    “Fix the PhD” (Nature)

  5. or… are we going back to get PhDs because, once in the workforce, we find ourselves working with people who were by far the smartest students at their easy school (the kind that offers multiple-choice exams) & who feel threatened by us? Or, perhaps they interpret our well-honed critical thinking skills & ability to think outside of the box as “attitude issues,” and we realize that we lack the moxie and the charisma on our own to overcome that hurdle.

    Then we retreat back into academia, get a little respite from loan payments, imagine the world once more as a meritocracy, and put off career heartbreaks until we’ve amassed the same amount of debt as a small country?

  6. Aren’t we going on to get PhDs because we get out into the work force and find that our liberal arts degrees don’t equate to jobs that pay a living wage unless we want to work in sales or finance?

  7. I’m a big fan of liberal arts colleges, but some caution needs to be taken in using “per capita Phd.” rankings. In short, per capita Phd production can be somewhat deceptive.

    For instance, according to the American Institute of Physics, which has tracked graduation numbers in physics and phd origins since the 1930’s, notes that in 2010:

    Lawrence college had 9 seniors majoring in physics in 2010
    Beloit college had 8 seniors majoring in physics in 2010

    A 25% rate of students going to get a phd when there are only 9 or so students, means that just 2 students go on to earn a Phd.

    Meanwhile, UC Berkeley had 121 seniors majoring in physics, UC Santa Cruz had 114 physics majors, U of Illinois U-C, 103.

    An 15% rate from a school with 121 or so students, means 18 students are going on to get doctorates — that means UC Berkeley is sending more than almost 10 TIMES as many students on to phd programs.

    In fact, AIP research also has shown that in pure numbers, the LARGEST undergraduate source of phd recipients in phd came from research universities.

    This doesn’t take away from the fact that students can – and do – get a very good undergraduate education in physics from small liberal arts colleges. But, it is incorrect to assume or imply that students attending research universities for physics are somehow all doomed to getting an inferior education in physics compared to someone who chooses a liberal arts college.

    Generalizations like that are misleading. Remember: there is a tremendous amount of variation between the quality of physics programs at both research universities AND liberal arts colleges. There are research universities with superb physics programs and liberal arts colleges with terrible physics programs.

    The question is not liberal arts vs. research university, but rather: what is the quality of the individual program at the individual institution?

    When it comes to college planning, there is no “one size fits all” choice that will apply to every student or every college/university. For some students, doing physics at a particular arge public research university will be the best choice. For other students, doing physics at a particular liberal arts college is the best choice. It really depends on the individual college/university, the individual physics program AND, most importantly, the individual student.

    Instead of jumping to conclusions such as “liberal arts is better” or “research university is better,” students interested in physics should be counseled to first focus on learning about what makes a good undergraduate program in physics. Good undergraduate physics programs share characteristics, whether they are part of a small liberal arts college or a large state university. PhD. production is not the ONLY measure of the quality of an undergraduate physics department.

    Once you know what makes a good undergrad physics program (or biology program, or studio art program, or English program, etc….), you will be armed to ask the RIGHT questions of individual colleges and universities in order to make an informed decision, rather than relying on meaningless “rankings” that tell you little about whether a particular college or university is really a place that you, personally, can be happy and successful.

    If anyone is interested in comparing some hard data about the size of physics programs, the data can be found here on the AIP website, along with other good information on what to look for in choosing a physics program:

    1. Good post. I am a huge fan of SLACs and would indeed send my children to them if appropriate, but I wouldn’t draw many hard and fast conclusions from the given data. These schools have different missions and draw different kinds of undergraduate students.

  8. “Students can increase their odds of being accepted to graduate school if they earn their bachelor’s degree at a liberal arts college. On a per capita basis, for instance, liberal arts colleges produce twice as many students who earn a PhD in science than other institutions.”

    This reasoning has – and always will be – absurd. I attend one of the touted liberal arts institutions that ranks extremely highly in per capita PhDs (I believe we’re the second-highest liberal arts college in the country in per capita PhD rates), and I can say that part of the reason we produce doctorates is because the students here came in with a predisposition for research and less interest (comparatively) in careers after college.

    Presupposing that per capita PhD rates _NECESSARILY_ correspond to better preparation or “odds” for acceptance to a graduate school is not only statistically questionable, but shows a lack of the oft-touted “critical thinking” skills institutions like mine sell themselves as promoting and fostering.

    Of course liberal arts colleges will dominate PhD _PER CAPITA_ numbers when non-liberal arts institutions often offer vocational majors and have tens of thousands of students compared to a few thousand at most. Acceptance rates on average for liberal arts applicants vs. non-liberal arts applicants to graduate school at equally prestigious institutions (although finding ‘equally’ would be difficult) could be taken as indicator of quality of undergraduate preparation, but per capita rates alone are meaningless.

    Articles like these, which I see floating around all of the time, only serve to make me shake my head in disappointment at the apparently well-educated individuals who are often so critical of other forms of education and who still misuse and misrepresent statistics. The education my college offers is terrific, and many students here who are seniors (including myself) have been accepted to doctoral programs already, but that does not equate to what you wrote.

  9. Lynn, I have tried to find this list in order to see institutions beyond the top ten, and I cannot find it. Do you have a link to it? Thanks.

  10. You might want to make some kind of correction for the academic quality of students at these institutions. Amherst, for example, has a slightly higher percentage of top percentile students than UMich.

  11. Very interesting article. Even though my high school kids have no idea what they want to study, let alone getting a Ph.D. degree in a particular field, this is an important article for me. I would hope there are more qualitative measures beyond the no. of Ph.D.s per capita that’s shown here. Dr. Chec’s article is very interesting, but was at Grinnell from ’66 to ’70, when the college-going population, the US and world economy, as well as the society in general were, well, more than 4 decades different from what they are today.

    Qualitative comparison between the liberal arts colleges and larger university is important and so is a deeper look at statistics like those quoted here. This is because a Ph.D. doing research can become Nobel Prize winners and another Ph.D. teaching required Physics courses to uninterested for-profit college students might have 1/3 the income of a MS degreed engineer and little in the way of satisfaction.

    I myself have a Ph.D. degree in engineering from 20 years ago and would say that certain engineering and biotech fields really require Ph.D. as a starting point, getting that degree in general is not really linked to success at the workplace. As people talk about the 80-20 rule (where 20% of the people do 80% of the work), those truly exceptional Ph.D.s create new field and advance human knowledge (and sometimes they get financially rewarded), but the rest of us usually are also runs who become more logical and methodical and gained more knowledge when we do our Ph.D. work but really don’t make waves ourselves.