Tyler Cowen and Arnold Kling both point a paper by Philip Babcock and Mindy Marks on the declining number of hours students spend studying per week.
After accounting for framing, we observe statistically significant declines in study time of about 8 hours per week between 1961 and 1981, about 2 hours per week between 1988 and 2004, and about 10 hours per week between 1961 and 2003.
The out of class weekly study time for full time college students decreased from 24 hours in 1961 to 14 hours in 2003. The surprising thing about this study is that the bulk of the decline occurred between 1961 and 1981 instead of between 1988 and 2004 during the time when personal computers became really useful for the majority students.
The paper is pretty thorough and finds the decreased study time occurring across different majors and at all prestige levels. However, two possible theories for the change in study habits are incorrectly dismissed: increased female student population and more students working part time to pay for college. These potential causes are dismissed because female students study more than male students, and the hours spent studying by full time students also decreased. The key aspect here is that both of these phenomena give students the ability to study less and still maintain their relative ranking within the college. The top quartile student only has to study a little more than his classmates. If the median students now have better things to do than study (dating, a necessary part time job) then the top quartile students can still easily retain their class ranking despite a reduced study time.
Another explanation is just that the culture of the baby-boomers was not that interested in studying, and the trend was self reinforcing for the relative ranking reason mentioned above.
On a tangential note, if students are graduating while working 10 hours less a week then catching up to the knowledge frontier probably isn't a good explanation of why younger scientists are less productive.