The University of California at Berkeley has one of the strongest statistics departments in the world. It is of some interest, perhaps, to see how that department was established, and how it relates to the other parts of the university. The following quotations are taken from E. L. Lehmann's reminiscences [10]. Key sections for us include his comments on Jerzy Neyman.
Neyman was 44 when in 1938 he accepted the offer to start a statistics program in the Berkeley Mathematics Department.
A more ambitious development of courses and faculty to teach them had to wait until the end of the war but eventually led to the creation during the decade 1946-1956 of the Berkeley Statistics Department, which became one of Neyman's principal American achievements. For many years, it was the leading department of theoretical statistics in the country. Its curriculum sets the standards that were followed by many others; it trained hundreds of Ph.D. students from all over the world.
The year 1947 brought Neyman two great victories. From the beginning, he had envisaged an independent statistics department, separate from the mathematics department. ... Evans (Griffith C. Evans, chair of Math) adamantly opposed a separate department since he believed in a greater mathematics department that would include all mathematical sciences. A compromise was now reached that left statistics within the mathematics department but with a separate budget that no longer required Evans' approval, and with the right to make its own research appointments although Evans would still have a say on teaching appointments.
Neyman's wish for a completely independent Department of Statistics had to wait a few more years. When Evans retired in 1954, his successor Charles Morrey had no desire to keep a substantial group of statisticians in the Mathematics Department against their wishes and recommended the creation of a separate statistics department. Thus, in 1955 Neyman finally obtained his own department which consisted of seven tenured faculty members (), three tenure-track assistant professors () and several lecturers and visitors
The particular charge of inordinate expense in terms of number of students taught, although true at the time it was made, lost its validity as gradually the new department took over the teaching of all lower division statistics courses and as a result soon regularly taught statistics to almost 5000 students a year.There are clearly some interesting parallels between the situation at UCB then, and the situation at UCLA now. Of course the differences are at least as large as the similarities. Almost all major research universities now have substantial departments of statistics, there are many more undergraduates taking statistics courses, and the move of statistics away from mathematics has progressed considerably.