It is with great sadness that I report that Sir Clive Granger passed away last night. He had been a wonderful colleague and good friend.
Clive always had a particular passion for identifying what was predictable in economic relationships. He emphasized the importance of sorting out which variables are helpful for purposes of forecasting others as one of the first steps in understanding underlying causal relationships. That philosophy came to be a regular feature in thousands of economic studies and seminars, as scholars would routinely report investigations of “Granger-causality.”
Clive also discovered, in a paper with Paul Newbold published in 1974, the phenomenon of spurious regression. The pair demonstrated by Monte Carlo simulation that if a researcher fails to take account of the underlying dynamics, a regression of one trending variable on another can produce what look like marvelously significant t-statistics, even though the reality could be that the two series are completely independent of each other.
The key contribution for which Clive was awarded the Nobel Memorial Prize in Economic Sciences in 2003 was his development of the concept of cointegration, a scheme for recognizing and understanding stable relationships in the midst of an otherwise constantly changing economic environment.
Perhaps above all else, Clive was a crucial leader in developing an approach to economic research that is grounded in predictive ability and empirical relations rather than models that theorists find elegant. His influence was an important balancing force for several generations of economists.
I remember him fondly as an amazingly creative individual, always brimming with new ideas, new questions, and new insights. My association with Clive has been a real personal treasure for my career at U.C. San Diego.
UPDATE: Here are some tributes to Clive from around the world: