In the epilogue to his The System of Liberty: Themes in Classical Liberalism--an excellent book that anybody seriously interested in classical liberal ideas should read and ponder--George H Smith takes a snapshot of the very different reasons to which we can trace the decline of liberalism in the 20th century. This is of course a most complex issue, but the interesting thing in Smith's approach is that it is more "introflexed" than the ones usually taken. That is, Smith seems interested to challenge classical liberals on their own failures, besides noting that objective conditions (i.e., the emergence of mass politics) conspired against their ideas.
In particular, he raises a point of great interest and truly eccentric, at least from the perspective of mainstream social science. Among other, Smith writes,
The widespread appeal of liberal ideas during the 18th century was indebted to its stress on developing a comprehensive 'science of man', as David Hume, Adam Smith, and their contemporaries called it. (...) There was a general agreement among Enlightenment liberals, including those who stressed the moral foundations of liberalism, that freedom could not be defended adequately from a single, narrow point of view and that an interdisciplinary approach was essential to the survival of liberalism.
This is of course at odds with the contemporary approach to scholarship, social sciences included. To make his point, George Smith quotes the inspiring inaugural address delivered to the University of St Andrews in 1867 by John Stuart Mill.
As Mill wrote:
Every department of knowledge becomes so loaded with details, that one who endeavours to know it with minute accuracy, must confine himself to a smaller and smaller portion of the whole extent: every science and art must be cut up into subdivisions, until each man's portion, the district which he thoroughly knows, bears about the same ratio to the whole range of useful knowledge that the art of putting on a pin's head does to the field of human industry. (...) Experience proves that there is no one study or pursuit, which, practised to the exclusion of all others, does not narrow and pervert the mind; breeding in it a class of prejudices special to that pursuit, besides a general prejudice, common to all narrow specialities, against large views, from an incapacity to take in and appreciate the grounds of them.
Now, the fact that specialization can have side-effects is not an argument against specialization per se --which increases productivity in the field of scholarship no less than in other human endeavours. However, I find Smith's point very intriguing. Perhaps, having a passion for human freedom requires a high degree of curiosity. Probably, the pressure towards ever narrower specialization decreases the time available to fulfill that curiosity. Furthermore, it might well be that intellectual specialization is by itself breeding a more social engineering-oriented approach. Increased knowledge of a particular field may bring about more anxious speculations about its future and thus breed impatience with the workings of spontaneous orders that seldom arrive on schedule. I don't know if specialization has played a role, in distancing intellectuals from classical liberal ideas. However, it seems to me a hypothesis worth considering - and hopefully, falsifying.