for a stationary state. Since the economy does not conform to
this condition one would suspect that there is indeed no reason
why cross section and time series should give the same results.
Again,more generally, economists have been inclined to
interpret regressions based on cross section data as
expressions of economic relations, possibly of cause and
effect. These interpretations seem to neglect that the data in
question reflect conditions belonging to different periods of
time. Perhaps this mety be understood more easily if one were to
think of the age distribution of a population which evidently
depends on fertility and morbidity over a series of years past.
Moreover,simple regressions reflect, or are influenced, by more
than one economic relation. As I show in the last of the
following papers the regression income-wealth reflects
influences from one to the other,in both directions.Generally
speaking, the fact that our economy has grown and has been
subject to all kinds of chops and changes must distort more or
less the pattern of the cross section data.
I dwell on this subject because it forms the recurrent theme of
the following papers. They purport to show the influence which
the past has wrought on the present, and how past growth is
expressed in the patterns and formations of the present which
bear the imprints of events gone by. As I discovered ex post I
-sun in a sense following the philosophy of Rene Thom with his
stress on morphology,on the study of forms and their evolution.
My studies suggest that past growth is closely linked to a
certain pattern of statistical distribution,namely the Pareto
distribution. In a stationary economy,so at least one of my