The Future Lies Ahead

In preparation for ASA’s 175 anniversary celebration in 2014, column “175”—written by members of the ASA’s 175th Anniversary Steering Committee and other ASA members— chronicled the theme chosen for the celebration. This is the last column from the series.

Contributing Editor
Stephen M. Stigler taught at the University of Wisconsin-Madison before moving to The University of Chicago. He was editor of JASA: Theory and Methods from 1978–1981 and has published several books, including The History of Statistics and Statistics on the Table.

Our 175th year is coming to an end, and with that come two temptations. One is to look forward to the 200th anniversary, 25 years from now. The other is to look backward at the past 175 years and assess what we have accomplished. The shorter view, looking forward, is the one I will take here, and it should be the easier of the two, but it is not—at least if some semblance of accuracy is expected.

Fifty years ago at our 125th anniversary, our president, Allen Wallis, was asked to predict an even more distant future, and he told a story. At The University of Chicago in 1952, he had been on a panel assembled to celebrate the 10th anniversary of the first controlled nuclear reaction. The panelists—Enrico Fermi was another member—were asked to predict when the generation of electricity through atomic energy would become competitive. Wallis, as a true statistician, approached this question (for which there was no usable data) honestly. “My prediction is that none of the predictions made here will bear any relation at all to what is going to happen.” When he repeated that statement in 1964, he had already been proven correct in several cases.

The pace of change in 2014 is much faster than in 1964, and looking forward is naturally more difficult. There is one compensating factor, however. The increase in publication rate and rapid change in format make it increasingly unlikely that anyone in 2039 will ever see these remarks. And this alleviates the usual worry a scholar might have, namely being proven wrong. And so encouraged, I offer a few unusually precise predictions. I predict (with unwarranted confidence) that in 2039:

• The ASA will no longer issue any printed publications; instead, it will host “in the cloud” a set of continuously changing databases, each linked to a set of also continuously changing analyses, constantly changing in rank order via a refined page rank algorithm.
• The annual meeting will become a purely social event, the speed of knowledge transmission having made the presentation of new results nearly impossible.
• D. R. Cox’s 1972 article on proportional hazards regression will remain the most highly cited item in statistics, the number of citations by then having passed 1 million.
• Statistics textbooks will survive, but only barely, driven primarily by the demand for their use from professors who received tenure before 2020.
• The ASA will be thriving and enjoying a new membership base as statistics increases its presence in the K–12 curriculum and the world realizes we are the discipline best able to cope with information surfeit.

I plan to save this column and trumpet it loudly in 25 years, but only if my predictions should turn out to be accurate. Meanwhile, the anniversary celebration is over—time to get back to work!