Abstract
Michael Hills died at the age of 86 on 7 January 2021 of COVID-19 after a short period of ill health. He touched many in the Stata and broader statistical community, even those who knew him only slightly or who met him through his writings or his software. Behind a slightly reserved exterior, Michael was friendly, funny, frank, a great explainer and a great contributor to statistical research, and very much his own man, a maverick in the best sense. He loved simple direct solutions and waged small wars against confusion and pretentiousness. Many of his Stata commands, most written jointly with David Clayton, have long since been folded back into official Stata in some way and so have enduring impact.
With her gracious permission, we are honored to carry a tribute to Michael by his daughter Claire, as well as a semitechnical appreciation.
Michael (and his twin brother) were born in Hertford, north of London, in 1934, into a world that just five years later was to be plunged into war. His father worked as a clerk for Islington Council; his mother hadn’t known that she was expecting twins and had only two names planned, so they got just the one each. Independent from the start, as a very small child, Michael set off down a steep hill on his elder brother’s bike. He fell off and broke his arm. It was set poorly and he was never able to straighten it; with his usual unflinching style, he described it recently on a medical form as a “deformed left arm”.
The twins had a freedom in childhood that’s hard to imagine today, running around bombsites looking for shrapnel, making bonfires, playing at being Emil and Gus from the children’s book Emil and the Detectives; for Gus the nickname was to last for life. They were eventually evacuated but not to strangers: a family friend in the countryside offered to take them in. He was a professional artist and we have drawings by him of Michael—and of Michael’s boots. But while our father’s wartime upbringing probably fostered his strongly independent character, it was a disaster for his education. He failed his 11-plus exam: partly, he thought, because he didn’t know his alphabet—something that he never truly learned even later in life, as he couldn’t quite see the point.
At this point, Michael had a stroke of good luck. Another family friend, a teacher, recognized his potential and somehow managed to sneak him into East Barnet Grammar School. Here he met another pair of twins who were to become lifelong friends: Robin and John Stace. Seven years later, he won an open scholarship to read mathematics at Magdalen College, Oxford—and was cited by those arguing against the 11-plus system.
I don’t think Oxford was a huge success for Michael. He didn’t like the snobbery and felt that the system was very unsuitable for teaching mathematics. He played a lot of tennis but managed to emerge with a respectable degree and eventually started work as a schoolteacher. Posts included one in Barbados—with a typical sense of adventure, he went out on a banana boat. He loved the weather and the challenge of teaching—and was amused by the colleague who put so much rum in the coffee he made it cold. Michael always loved to travel—particularly anywhere hot—and his career afforded him many opportunities; over the years he seems to have visited everywhere from America to New Zealand and made some wonderful friendships along the way, particularly the Breslow family from Seattle.
Back in London in the late 1950s, he was invited to a party by John Stace. He couldn’t find the address and almost gave up, but luckily he didn’t because at the party was Fiona, who had met John studying English at Oxford.
What drew our parents together? Obviously a shared sense of humor, a love of literature, and an interest in intellectual debate of all kinds, but perhaps also a determined unconventionality. This wasn’t always easy to live with—those with a sense of interior design might question why there was white plastic garden furniture in the living room, while anyone trying to help with computer problems in later life might wonder if it was really necessary to shun the conveniences offered by Microsoft in favor of a Linux operating system—but neither Michael nor Fiona ever did much merely because it was what other people did.
They married in 1961 and had three children in fairly short succession, moving from a rented flat in Hampstead to the house in Highgate that was to be their home for more than 50 years. Michael loved babies but those early years must have been tough, as he was working full-time—at first as a systems analyst for IBM and a little later as a lecturer at the London School of Hygiene and Tropical Medicine—as well as studying for a diploma and then a PhD at nights. Nonetheless, I remember him as a constantly engaged and loving parent, helping me to learn the recorder, teaching us mathematics with the help of a character called Mr. Decimal Point, taking us to concerts and art galleries, and driving us to music lessons around north London. On my 10th birthday, the two of us went for a walk and discussed very earnestly what it felt like to be “in double figures”.
He was also funny and full of enthusiasms, whether it was the home microscope for which he created slides (including a head louse, scientifically labeled “Louse: M. Hills”) or the sudden decision to launch the blow-up plastic furniture in one holiday house in France onto a nearby lake—my brother and I paddling an armchair each, he and my sister on the sofa—as my mother looked on aghast from the bank. As a younger man, he cycled everywhere—from Highgate to South Kensington and back every day when he worked as Head of Biometrics at the Natural History Museum (a Civil Service post), and even around Houston, Texas (where he spent a summer teaching), much to the amazement of local residents. Later, back pain forced him to swap the cycle for a motorbike—impatient with traffic jams, he loved the freedom it afforded him.
As the groaning shelves in his house bear witness, books were hugely important to him—novels, autobiographies and history as well as mathematics, and he was always keen to recommend things to read—Flaubert, Nabokov, Anthony Powell’s A Dance to the Music of Time.… At the Natural History Museum, rather fed up with the job, he would spend lunch times in his office lying on the floor beneath his desk to ease his back, reading Proust. In the end he decided that the Civil Service was not for him (though it provided him with plenty of material for blackly funny stories, such as the tale of the scientist who went mad and ran amok with a barometer, shattering glass cases full of specimens until he was restrained). Instead he worked for a couple of years at the Open University, motorcycling up to Milton Keynes when necessary, and then returned to the School of Hygiene in 1985, describing the move with characteristic modesty and humor as “climbing down the ladder of success”.
In fact, it was a great success, because, as his children and grandchildren know, Michael was an inspired and inspiring teacher. (Nick is now Professor of Computational Engineering at the University of Surrey, while three out of his four grandchildren continue to study maths.) We have been so touched by the testimonies of colleagues and students from both the School of Hygiene and a variety of statistical courses held around the world, particularly the ones led with David Clayton and organized by Franco Merletti and Rodolfo Saracci in Florence. Michael leaves two textbooks, one coauthored with David and the other with Bianca De Stavola. Even after retirement, he was teaching abroad into his late seventies; characteristic photographs show him happily sitting in the sun with a book or a newspaper, or chatting to colleagues, sharing jokes and gossip, and enjoying a glass of red wine.
As Anthony Powell once said, “[G]rowing old is like being increasingly penalized for a crime you haven’t committed.” Michael bore it with fortitude, continuing to read novels and think about current affairs, learning to cook, and looking after Fiona with great love and compassion until she died in 2015, after more than 50 years of marriage. He missed her enormously, and also his twin brother, Gus, who died last year, but he was still reading and watching formidably highbrow television until almost the end of last year; we recognized how very ill he had become in December, even before he caught COVID, by the fact that he didn’t have a book on the go. It’s a grim irony for an epidemiologist to die during an epidemic, but we are grateful at any rate that such an independent man continued independent until the very last.
Claire Hills
Michael worked at the London School for Hygiene and Tropical Medicine (LSHTM), first between 1962 and 1971 and then from 1985 to 1996, when he retired. In between, he worked at the Natural History Museum, as Head of Biometrics and Computing, and then at the Open University. Many of those who met him during his second spell at the School, students as well as (junior and senior) staff, were fundamentally shaped by attending his lectures and from teaching and collaborating with him.
Michael was a wonderful teacher, able to reduce statistical and methodological concepts to their essence and then to communicate them with exceptional clarity. He often mentioned his years as a school maths teacher as the source for his unique style. He used to say that a lecture should be like a Chinese meal: you always leave the table feeling hungry for more. Indeed, his book with David Clayton (Clayton and Hills 1993) is a typical example of this. Every line there is succinct, and yet it encourages you to think more deeply and find out more about where the concepts take you.
The book with David Clayton is a classic reference for epidemiologists, especially in Europe. Like no other textbook before, it offered an accessible but formal link between statistical modeling and estimation and the goals of epidemiological research. It is not surprising, therefore, that Michael’s contribution to LSHTM’s teaching program lives on today in much of the teaching in the Statistical Methods in Epidemiology module and in the Advanced Course in Epidemiological Analysis short course.
The same must be said for the European Educational Programme in Epidemiology, held in Florence every summer, where he taught from its inception in 1991 until only a few years ago. There, he influenced scores of young researchers from across Europe and beyond, who—year after year—would surround him at every break with questions on the likelihood principle and who would give him the warmest farewells at the end of the course. It has to be said that there were also other advantages that brought Michael to teach in Florence.
Beyond his teaching, Michael made several valuable and enduring contributions to the literature. The references below list his two textbooks already mentioned (that with Bianca De Stavola having morphed through various previous editions); a concise but interestingly different monograph (Hills 1974); and a capricious selection of papers likely to be of interest to a wide cross-section of readers. Beyond several much appreciated papers are some small gems, with original ideas still deserving exploration decades after publication. Beyond this selection lie many papers with specific applications to several areas of medicine, to systematics, and to physical anthropology.
Often a teacher with Stata, and a teacher of Stata, Michael wrote several commands, usually with David Clayton, that mostly have been folded back into official Stata. His influence can be seen directly in official commands
Michael was one of the most self-effacing and modest people one could imagine. This went along with a rigorous and deep intellectual integrity. Michael avoided all pomp and pretension, both in himself and in others. Despite his quiet and somewhat retiring character, he would make it clear when he thought ideas lacked rigor (or lectures were not up to standard!). Michael was very dear to many: there is no doubt that his legacy will continue and the memory of his kindness will stay with us.