Saturday, March 17, 2012

University dropout rates by field

The German "Bildungsbericht 2010" (report on education), has a table on the dropout rates by field that I thought would be of interest for some of you. The table is from this PDF, page 297. These are the dropout rates in percent at German universities, I've ordered them by decreasing percentage.

Physics, Geoscience 36
Mechanical engineering 34
Electrical engineering 33
Linguistic and Cultural science 32
Computer science 32
Mathematics 31
Chemistry 31
Economics 27
Science of education, sport 20
Architecture 16
Biology 15
Geography 15
Art 12
Social sciences 10
Law 9
Agriculture, forestry and nutrition science 7
Pharmacy 6
Human medicine 5
Dental and veterinary medicine 3

The data is for the year of 2006, it is averaged over several years in which the students might have started and includes only German students.

I was surprised that the rate in mathematics is not higher. Sitting in a classroom that was a little emptier every week, it certainly seemed higher to me (in 1995). Though my estimate would have included those who had taken mathematics as a minor field of study, so the rate was probably indeed higher. I have no idea what's wrong with linguistic. And I would have expected the dropout rate in medicine to be vastly higher.

For some fields there is data for men and women separately. In almost all cases where there is data, the dropout rate among women is lower than among men. The one exception is medicine, where women drop out in a higher fraction.

It would be interesting to compare these rates to countries in which there are tuition fees to see if it makes a large difference in commitment. If somebody has a reference, please let me know in the comments.


  1. considering how much you read and the diversity of your sources, do you have a special app to keep all those little pieces of information together or even better, have a tool to connect all those pieces? (one can dream ;))

  2. "Look to the left, look to the right. One of you will not make it." Reject good enough early, so they can retarget. Room and resources are insufficient for best.

    September 1969, Michigan State University, first term majors organic chemistry, 1200+ students enrolled. June 1973, 14 BS/Chem conferred. 14! Then... diversity.

    "Are you the very best at what you do?"
    "Yes Sir!"

  3. 36 to 31 % difference within science/techno"logy" (silly word) is not really sigificant.
    But the 9 % in law ist significant and, totally wrong. "Staatsexamen" is the big thing in the end, and this exam is outside universities. in subjects where most of the studends look for some job as teachers (state employment) things are similar. So, as ever, do not believe in statistics You didn't fake Yourself :=(

  4. Lots of American stats. here:

    but only gross rates - not by field of study, as far as I can tell.

  5. Hi Tom,

    Yeah, I have a special app to keep all the little pieces together, it sits between my ears ;o) More seriously, this blog is part of my archiving system. It works imperfectly though in that even I sometimes can't find a post that I know I wrote. There's a Firefox app which works very well for archiving pretty much everything online (forgot what it's called). I tried that for some while and it worked quite well, except that, well, I don't use Firefox, so I don't use it anymore. By and large, I get along, and the pressure to use something more sophisticated isn't large enough. I also use my browsing history. I find that the time order in which I've come across information works very well for me to keep track of it. Best,


  6. Medizin hat einen Numerus Clausu. Physik nicht. Ich denke das sollte das untere Ende der Liste hinreichend erklären....

  7. Hi Frank,

    Ja, hat denn Jura nicht auch einen NC, zumindest regional? Ich meine, mich an so was erinnern zu koennen. Aber Du hast recht, NC ist vermutlich korreliert mit der Abbruchquote. MfG,


  8. Hi Arun,

    Thanks for the link, though I couldn't find a useful table on the website. Best,


  9. Are these people who leave University, or does it include people who change field of study? When I taught mechanical engineering, the story was that our high dropout rate was the business college's gain.

  10. From: Ars Technica

    One finding that did separate engineering from other major fields of study was apparent in the percentage of graduates who had started college in their eventual major; that's an indication of how many switched into a major partway through their collegiate career. Only half of those with social science degrees started in that field, 60 percent of physical science majors did, but a full 93 percent of engineers began their academic career in engineering.

  11. Hi CIP,

    That's all those who start studying X and don't finish studying X, so it includes those who change fields. Best,


  12. Hi Bee,

    Not much here I find too surprising, although I would have wished the dropout rate in law were much higher such that more might be familiar with natural laws as to be be affected and so guided as opposed to the unnatural kind:-)



  13. Hi Bee,

    As I said, only gross rates are given, for instance,

    9.3a National Six-Year Graduation Rates of Bachelor's Degree-Seeking Students, 2002-08

    which I take to mean people who finish their 4-year bachelors in 4 to 6 years.

    Public - 55.3% graduation rate
    Private(Not-for-profit) 65.1%
    Private(For profit) 23.5%

    If you allow for 8 years to do the degree, the numbers go upt a bit

    Public - 58.3%
    Private(Not-for-profit) 66.4%
    Private(For-Profit) 37.8%

  14. Nonlocality is not a problem in dense aether model, in which energy is transfered both with transverse waves of light, both with longitudinal gravitational waves, which manifest like CMBR noise because they're superluminal.

  15. This comment has been removed by the author.



Comment moderation on this blog is turned on.
Submitted comments will only appear after manual approval, which can take up to 24 hours.
Comments posted as "Unknown" go straight to junk.