ABSTRACT
This speech was delivered at the Banquet and Award Ceremony June 17, 2018, during SAE 2018 -- an international conference on ‘Small Area Estimation and Other Topics of Current Interest in Surveys, Official Statistics, and General Statistics: A Celebration of Professor Danny Pfeffermann’s 75th Birthday.
I was told that I cannot speak for more than two hours, so I shall speak really fast. No, no, it will be shorter than a regular presentation. Actually, this is the first conference that I can recall in which I am not presenting any paper, so this is my first and last opportunity.
Thank you Partha, Li-Chun, Nikos, Sanjay and Natalie for suggesting that this conference will celebrate my birthday, and thank you Dongchu for taking the leading role in organizing it. Thanks to all the members of the organizing and scientific committees, looks like half of you are on one of them, and thank you all for attending this special celebration. This includes my core family of one wife, two daughters with husbands, and four grandchildren.
I was thinking what I should say at this special event and decided to review a few episodes of my professional career and discuss three important questions: how I became a Statistician, how I got involved in Small Area Estimation, and how I ended up at the Israel CBS. I cannot guarantee that everything I tell is also correct, but as Don Rubin said recently in a seminar at the Israel CBS, ‘memories don’t need to be correct, – they only need to be interesting’.
So how did I become a Statistician? I actually did my first statistical inference already at the age of 13, even before Jon Rao. My teacher gave us 20 cards with questions, and told us that each pupil will have to choose randomly one card and answer the questions in it. We were given 3 weeks to prepare, but my interest those days was solely in playing soccer, so I only managed to go over the first 10 cards. I was panicking on the day of the exam, fearing that I shall pick up one of the cards 11–20, but here came to my rescue my statistical talents. Since my surname starts with ‘P’, I was one of the last pupils to be called. I asked each child who already took the exam which card number he picked, and I quickly inferred that the cards were placed in two packs, cards 1–10 in the left-hand side and cards 11–20 in the right-hand side. With this vital information, I picked a card from the left side and got the highest mark in the class.
With this success, it looks like I had all the necessary qualifications to become a statistician, but what I really wanted to be was a high school Math teacher. So, after completing three years of army service I studied Math for my first degree. In Israel those days, you normally had to study two majors and it was then when I made my first smart move, deciding to study statistics as my second major instead of physics, which was the popular second choice. When I got my Bachelor degree I had to choose between Mathematics and Statistics for my Master degree, and it was then that I made my second smart move, opting for Statistics, but I didn’t give up my dream; I was a high school Math teacher for five years.
My next mission was to write a Ph.D. and I started looking for possibilities abroad, but it was then that the late Gad Nathan offered me a job at the Israel CBS. I couldn’t make up my mind whether to accept Gad’s offer, so I asked my mother for advice. She said: ‘join the CBS but tell Gad to supervise you for a Ph.D.’. What a brilliant idea. Why didn’t I think of it? So the next day I rushed into Gad’s office with my mother’s instruction. Gad choked for a while and then conceded, saying that he cannot argue with my mother. I was his first Ph.D. student. Gad was a great scholar and a wonderful man, and over the years we became colleagues and close friends. I wish that he could be with us today.
Upon completing my Ph.D. I moved to Southampton for my postdoc, and this is when my long lasting relationship with the University of Southampton started. I met Fred Smith who chose to be my informal mentor and another young Statistician by the name of Chris Skinner. Both Fred and Chris were supposed to be here with us but couldn’t come because of health issues. Wayne Fuller and Pedro Silva also couldn’t come because of back problems, and I wish them all quick and full recovery.
After Southampton, we moved to Baltimore for a few months where I worked with Richard Royall, and it is there where I met Ray Chambers for the first time. Ray later joined Southampton and one of his Ph.D. students there was Nikos, who is now the Head of the School of Economic, Social and Political Sciences. Well done Nikos. Many years later, Li-Chun joined us, so adding the names of Graham Kalton and Tim Holt, this more or less makes up the famous Southampton dynasty, later known as S3RI, of which I am the oldest non-retired person. I am probably the oldest non-retired person at the entire university. Incidentally, when I reached the age of 67, there was still mandatory retirement in the UK, but for some reason no one asked me to retire and it was then that I made my third smart move; I kept quiet and didn’t ask why I am not told to retire.
After Baltimore, I returned to Israel and started my academic career as a lecturer at the Hebrew University. A few years later, I spent my first sabbatical at RTI and on my first day they invited me for lunch during which, a woman by the name of Brenda Cox, which some of you will remember, told me in front of everyone: ‘we have bad experience with visitors, we hardly see them and they do nothing for us’. I promised to change that tradition but she didn’t seem to be convinced. Well, I stayed there for two years and two more summers and among other things, I managed to win a tender that earned RTI a lot of money. I thought that they underpaid me the first year so after this successful tender, I phoned the vice president of RTI and asked for a raise in my salary. The vice president was someone by the name of Dan Horwitz, also known for the Horwitz–Thompson estimator. Dan agreed that I deserve a raise and came up with a brilliant idea, almost as brilliant as the Horwitz–Thompson estimator; ‘we shall match your salary to what you earn in Israel’. Mind you, my monthly salary in Israel those days was about three times lower than what they already paid me during the first year. As you can see, Dan was also a very shrewd businessman.
While at RTI, I was promoted to senior lecturer and later to Associate Professor and Full professor. In case you didn’t notice, in Israel, you need to undergo no less than 4 torturous promotion processes before you can finally relax. I should say that I am still teaching at the HU and in Southampton, and for the last 25 years I consulted the BLS in Washington, until this year when a guy named Donald Trump decided to put on hold consulting contracts with Non-American citizens. Yesterday I heard that the BLS is making progress in renewing my contract, so who knows, maybe after signing a contract with his friend Kim Jong-Un, Trump will also find a way to sign a contract with me.
So now to the second question, how did I get involved in Small Area Estimation? Evidently, in my Ph.D. I worked on several SAE problems but I never heard of this concept in those days, so these three words that could make me really famous are not even mentioned in my dissertation. I heard about SAE only a few years later and even published a paper on this topic where I developed a Kosher Kalman Filter, but didn’t do much more on this until I suddenly received an email in 1999, inviting me to give the keynote address at the satellite conference on SAE in Riga. I have no idea who came up with this bizarre idea because as I said, my contributions to this topic at that time were close to nil, so I immediately composed a kind response, thanking for the invitation but declining the offer. I was about to hit the send key when I suddenly thought to myself; come on, you are invited to present a keynote address, who knows if you will ever be invited again to present a keynote address, go for it, which I did. I prepared a decent talk which I later published and even got commended for it for … being friendly. Another nice memory I have from that conference is that a few minutes before my talk a man, dressed in a suit and a tie came to me and introduced himself as the Israeli ambassador to Latvia. After the talk, he told me that the only part of my talk that he understood was the Greek letters beta and theta, which he heard before. He invited me for lunch at his home, so this was another very successful SAE conference.
Finally, to the last question, how did I end up as the National Statistician and Director of the Israel CBS? I already mentioned that I spent my first sabbatical at RTI. I actually spent most of my sabbaticals at National Statistical offices in countries such as Canada, Australia and Brazil, and as I told you already, I worked for the Israel CBS after my high school teaching career, so working for the ICBS again after retirement would not have been such a strange idea, except that I was much older than 67, the mandatory retirement age in Israel, when they were looking for a new director. As such, I didn’t pay any attention to the recruitment process, until one day my wife called me and asked if I read in the front page of a local newspaper that the search committee has troubles finding a new director. I was in Washington at the time. This got me curious so I emailed the head of the search committee a one-line message asking: Is there an age limit for this job? He immediately wrote back: ‘I have no idea’, I shall find out. Two weeks later he wrote again urging me to send my CV. This was it. Sara, every time you ask me why I don’t retire like all our friends, keep in mind that it is partly your fault.
And what about the job of National Statistician? Well, it was only 2 months later that I heard that such a position even existed, and I am the new person in charge of guiding the statistics produced in other Government offices. For some reason, this didn’t come up during my job interview. Incidentally, it took 2 years for the Israel parliament to change the title of the job from Government Statistician to National Statistician.
I am about to finish but I like to say one more thing. I owe much of my career to you, and I mean it. You were always very friendly and very supportive and I think that this is what best characterizes us. We argue on ‘minor issues’ such as model-based versus design-based, Bayesian versus Frequentist, administrative census Vs a big survey, but we are friends and encourage each other’s work. This is not always the case with other scientific communities, and we should be proud of it.
We are experiencing a new revolutionary era of huge technological developments with big data, machine learning and all the rest, and from time to time I ask myself how my professional life would have looked like if we celebrated today my 25th birthday and I was starting a career as a data scientist. Fortunately, I am 75 and I don’t have to start all over again to answer this question …
THANK YOU AGAIN, AND BLESS YOU ALL.