Ivan Fellegi

I cannot believe that I am writing an article in memory of Leslie Kish. Just a few months ago I wrote a partly humorous little speech on the occasion of his 90th birthday celebration. I jokingly asked why we are making such a fuss about a 90th birthday-after all the Queen mother just celebrated her 100th. I emphasized that that was something. He laughed heartily, with the well known “Kish twinkle” in his eye. I was struck once again by the extent to which he remained fun-loving, vibrant, insightful, in fact young in all aspects of behaviour-even if somewhat limited in his mobility. He told me about his forthcoming partial knee replacement operation and confided that his doctor told him that he will either undergo this operation, or he will need to use a walker to get around. Of course, a walker was not to be contemplated: he needed to have his full mobility. And mobility, at 90, meant not just getting around at home but traveling around the world several times a year. He died due to post-operative complications, having fought for several weeks with his usual indomitable courage.

In my mind the most characteristic feature of his life was his incessant giving. One of his last acts of giving was to inspire his friends and colleagues to establish the Leslie Kish International Fellows Fund to help students from developing countries obtain training in population sampling.

Leslie was born in 1910 in Poprad, then part of the Austro-Hungarian Empire, now in Slovakia. He used to relate how, at various times throughout history, Poprad belonged to five different countries-an appropriate symbol of his life motivated by a love of people from all parts of the world. In 1925, his parents decided to migrate to the United States-together with hundreds of thousands of other Hungarians. As the great Hungarian poet Attila Jozsef put it: “one and a half million of our people staggered out to America.”

Soon after their arrival Leslie’s father died. The remaining family of mother and four children had to decide whether to stay in the United States. They did, but that meant that the two oldest children, including Leslie, who was then 16 years old, would have to work in order to help the others.

Leslie continued his schooling in the evening. By 1937 he was within a year of completing his undergraduate studies. But this 27-year-old was once again ready to sacrifice himself in order to help improve the world.

He interrupted his studies in order help fight the fascists in Spain as a member of the International Brigade. His love of things Spanish, and of people oppressed, stayed with him forever.

At the end of the Spanish Civil War in 1939 he returned to the United States and completed his studies at City College of New York and received a degree in mathematics. He moved to Washington D.C., where he was fortunate to become a member of pioneering groups, first at the Bureau of the Census and then at the Department of Agriculture.

He interrupted his career again, to volunteer for service in World War II. In 1947, he finally moved to the University of Michigan at Ann Arbor where, together with a small band of enthusiasts, he helped found the Institute for Social Research. He said later that he never worked as hard as he did in those early years obtaining his master’s and PhD while working full time but also finding time to teach.

In statistics, he gave us several superb books. These include the pioneering “Survey Sampling” which became not just a bible of the field (like the original one, a source of lofty inspiration), but also a day to day tool of practice. Indeed, much of the world’s statistical system has embedded in it the hundreds of pearls of practical wisdom of “Survey Sampling.” In 1988 (when Leslie was a young 78) came “Statistical Design for Research” which integrated and organized a lifetime of acquired statistical wisdom. Before, in between, and after came a stream of articles, lectures and talks. He, sometimes working with others, introduced the concepts into our thinking and the words into our language of design effects; he was among the first to explore the issue of inference from complex samples and developed the innovation now known as balanced repeated replication (actually with Marty Frankel); was among the pioneers of studying response errors; became the apostle of rolling samples and censuses; pioneered controlled selection; formulated the concept of multipurpose designs; did some of the early work on small area estimation; and so on. But important as these works are, I think just as crucial were some of his other contributions.

He was one of very few people whose early applied work made sampling respectable and admired. In addition to having been one of the founders of what became the Institute for Survey Research at Ann Arbor, he taught generations of statisticians, both American and foreign ones through the legendary Summer Program for Foreign Statisticians. After his formal retirement he continued to do so through lectures in the Summer Program; through decades of editing or contributing to one or another of the Question and Answer columns of the Survey Statistician; and through numerous lectures and consulting assignments. At international meetings I used to “bump into” his past students and current friends. One no longer “bumps into” them, because they have become completely ubiquitous: I wonder how many better known foreign samplers there are who were not at some point Leslie’s students. And I do not want to forget about two of my favourites among his many contributions. His years of faithful service to Statistics Canada as a founding member of our Advisory Committee on Statistical Methods; and his ASA presidential address of 1977 (published in JASA in March 1978)—the best address that any president of ASA gave in my living memory.

He gave a lot to ISI as well. During the early ’70s Leslie and I were members of a small organizing committee which brought about the International Association of Survey Statisticians. I can fairly say that 98 per cent of the work was done by the two of us. Later, he served as a member of Council, as vice president, and finally president-elect and president. In effect, Leslie was at the centre of IASS for almost 15 years; from 1971 to 1985. He was also a member of two ISI ad hoc committees: on Statistical Development in Developing Countries and on ISI’s Possible Role in Assessing the Risks of Nuclear War.

For his accomplishments he received world wide recognition. Of his dozens of awards I will just single out a few: he received an honorary doctorate from the University of Bologna on the occasion of its 900th anniversary; the Samuel Wilks Medal, which is ASA’s highest recognition; the Henry Russell lectureship, which is the highest recognition of University of Michigan; the title Honorary Fellow of the ISI which I regard as a kind of Nobel prize in statistics; and perhaps the most personally meaningful for him: a slew of the highest possible recognitions from Hungary (honorary doctorate from the largest university in Budapest, honorary membership in the Hungarian Academy of Sciences and the Officer’s Cross of the Order of the Merit).

Over and above what he gave us in statistics, he gave us the phenomenon known as “Leslie Kish, a force of nature”: the Spanish Civil War fighter, the philosopher of all things statistical, the ever young agitator for human rights, raconteur, avid reader, author of the best annual Christmas letters, loving husband and father, and lifelong friend to hundreds, perhaps thousands.

When I spoke at his 90th birthday celebration, I ended by saying that I was hoping to be present at Leslie’s really big anniversary-the one the Queen Mother had just passed. And that was not just a joke: he was so full of life, it was not only quite possible to contemplate him living to be a hundred, but rather it was impossible to think about the opposite. Unfortunately, he did pass away. His final act of giving was to donate his body to medical research. Wouldn’t it be fitting if the resulting work gave us some insight into the human wonder that was Leslie Kish?

This article originally appeared in the International Statistical Institute (ISI) Newsletter, Volume 25, No. 73. It is reprinted with permission from ISI and the author.