Albert Hosmer Bowker (1919-2008)

A Remembrance by Ingram Olkin, Stanford University

NISS gratefully acknowledges permission of the IMS Bulletin,
in which this remembrance appeared originally, to use it here.

In so many ways Al Bowker was a man for all seasons. He was the inaugural Chair of the Statistics Department at Stanford from 1947 to 1963, and as Dean of Graduate Studies from 1960-1963 at which time he became Chancellor of the City University of New York. In 1977 returned to California as Chancellor of The University of California at Berkeley. In 1980 he was appointed as the first Assistant Secretary for Postsecondary Education in the newly formed US Department of Education. In 1981 he was founding Dean of the School of Public Affairs at the University of Maryland, and later became Executive Vice President. He returned to the City University of New York Research Foundation in 1986 where he served as Vice President for Planning.

Albert Bowker was born in Winchendon, Massachusetts on September 8, 1919, but grew up in Washington, D.C. His father was a scientist at the then National Bureau of Standards. His BS was in mathematics from MIT in 1941, after which he worked on several military projects, and later joined the Statistical Research Group (SRG) at Columbia. The SRG was a major statistical center during WWII and a statistical ‘Who’s Who’ that consisted of Abraham Wald, Churchill Eisenhart, Jimmie Savage, Milton Friedman, Georgle Stigler, Abe Girshick, Ken Arnold, Harold Freeman, Herbert Solomon, Ed Paulson, Millard Hastay, and Rollin Bennett.

Harold Hotelling was at Columbia and moved to Chapel Hill in 1946 to head the newly formed statistics department. Al studied multivariate analysis with Pao-Lu Hsu when Hsu visited Columbia, and when Hsu moved to Chapel Hill, Al followed. Hotelling suggested a thesis topic on asymptotic distributions. When Al completed the dissertation, which was submitted to Columbia, it was formally signed by Jack Wolfowitz. However, Wolfowitz was not a specialist in multivariate analysis, and actually Ted Anderson approved the thesis.

Al obtained the non-central distribution of Hotelling’s T² statistic using an invariance argument which is described in Section. 5.2.2. of Anderson’s book on multivariate analysis. Later, Al worked with Rosedith Sitgreaves on an asymptotic expansion for the distribution of a classification statistic. Rosedith was also a student at Columbia, and later on the faculty at Teachers College. She and Al were married in 1964, and when Al was Chancellor at Berkeley, Rosedith was on the faculty in the School of Education at Stanford.

It may already be clear that Al was a talented leader and developer. At Stanford he helped the mathematics department become an eminent department in applicable math (a term Al preferred to applied mathematics). He worked with the then Provost Fred Terman to bring the linear accelerator and the hospital (then in San Francisco) to Stanford. Al had the foresight to attract George Forsythe with the idea of starting a department of computer science, which may have been the first CS department in the country. A key feature in Al’s thinking was that a statistics department would not be large, and that for it to have a research effect in a university it would be wise to have joint appointments with other departments. Thus, over the years, the statistics faculty had joint appointments with psychology, economics, education, mathematics, linear accelerator, earth sciences, electrical engineering, operations research, and the medical school. As a result a rather small department has had great influence in the University.

When Al was chosen by Allen Wallis to be the Chair of the newly formed statistics department, he was in the process of completing his dissertation, but even at this young age he recognized talent, and managed to get Abe Girshick to join the department and David Blackwell to visit. Although Blackwell left shortly thereafter for Berkeley, Al was able to hire a stellar faculty. By 1956 the department had a formidable faculty consisting of Kenneth Arrow (joint with Economics), Herman Chernoff, Samuel Karlin (joint with Mathematics), Quinn McNemar (joint with Psychology), Charles Stein, Gerald Lieberman (joint with Engineering), Lincoln Moses (joint with the Medical School), and Emanuel Parzen. Shortly thereafter Rupert Miller, Vernon Johns, Herbert Solomon, and Herbert Scarf joined the department. These appointments were engineered by Al. He formed an Applied Math and Statistics Laboratory and obtained ONR support to sustain the Laboratory. Almost everyone in the two departments received some support form the Laboratory, and visitors from all over visited Stanford during those years.

In 1987 the idea of a National Institute of Statistical Sciences (NISS) was in its beginnings. If such an Institute was to be founded, a working committee needed to be created to move it along. Furthermore such a committee would need to negotiate with university administrations, which suggested that it be wise to have a senior statesman with name recognition as Chair. Al was an obvious choice. At this time he had completed his role as Chancellor at both CUNY and Berkeley, and I knew that he was somewhat at loose ends and would like to be involved in some project. I asked him if he would be willing to serve as Chair of a committee to try to form an Institute of Statistics. Al knew many of the Chancellors and Presidents of major universities, and these connections helped in generating an interest in founding an institute. Janet Norwood, formerly Commissioner of the Bureau of Labor Statistics was the other member. Thus we had two illustrious people, both committed to the furtherance of statistics and the development of an Institute. NISS did come into existence in 1990, and was located in Research Triangle Park in North Carolina. Al served as Chair of the Board. Much needed to be done, and Al guided the group with excellent counsel. His style was characteristic of the way he functioned in other roles. Al often bumbled along, letting everyone speak in a somewhat chaotic track, then at a certain point — and only he knew when that right point was reached, there appeared a motion of exactly what he wanted. Al told me that when he went into a meeting he knew what he wanted to come out at the end, and somehow managed to achieve this goal. Stories from his days as Chancellor confirm this. Thus, NISS moved along to fruition, and the profession owes him a hearty thanks for his efforts in NISS.

Al was concerned with what we now call “diversity” and “equity”. This is exhibited by two examples of his actions. The first occurred while Chancellor at CUNY. He expanded the University from four senior colleges and a few community colleges to 18 campuses. His policy of open enrollment was designed to permit all students to attend at least one of these colleges. This policy was controversial, but he believed that it was untenable for an all-white University to be located in the heart of Harlem without the community in attendance. A second act of his relates to women. Al was elected president of The Cosmos Club in Washington, which at the time was an all male club. At that time members (men) used the main entrance, and women were supposed to use a side entrance. One of Al’s first acts was to change the by-laws to allow women to become members.

The New York Times noted that Al “demonstrated extraordinary vision and coverage in promoting access and excellence in public higher education,” to which we can add that he was a promoter of statistics in every way that he could. He will be missed by all those who knew him, and by future students and faculty for his efforts in creating a foundation in statistics.