Upon returning to Harvard in 1945, now married to the reader and writer of all those V-letters, I chose, almost casually, to go on with economics.During that year I was also working on my Ph.D. thesis, an exploratory attempt to model changes in the size distribution of wage income using interacting Markoff processes for employment-unemployment and wage rates.So, in 1949-50, I spent a fellowship year at Columbia University, in the lectures of Abraham Wald, Jacob Wolfowitz and T.W. Anderson, along with my fellow student and friend, Jack Kiefer.He was also responsible for my introduction to empirical work: as his research assistant I produced the first set of capital-coefficients for the input-output model.The thesis was awarded the Wells Prize at Harvard, which offered publication in book form and $500 (in 1951 prices!) upon completion.Somehow - the memory is lost - I became interested in statistics and probabilistic models.But I never returned to that work and the thesis remains unpublished (and the check uncashed).By a piece of good luck, Wassily Leontief became my teacher, guide and friend.In those days, the teaching of statistics at Harvard was, to put it kindly, eccentric.I learned a lot from Frederick Mosteller, whose appointment was in the Department of Social Relations.Eventually he advised me to study more intensively in a place where that was possible.When I reread the thesis, however, I thought that I could do it better.