In over 100 social time-series, logistic trend lines have been fitted or suggested. In 30 of these it has been proved that the observed trends are not due to mere random variation. Various other mathematical formulas have been applied to S-shaped trends; but, where rigorous tests of relative closeness of fir have been applied, the logistic trend, or a series of logistic cycles, has usually proved superior. The Gompertz curve has been preferred by Kuznets for trends in production of certain metals and by Davis for trends in patents and inventions. Since mere random variation has been ruled out for 30 logistic curves fitted to social trends, the empirical "explanation" is merely a confession of ignorance as to what the natural laws involved actually mean.
No takes yet. Share an insight, caveat, or question.
Hornell Hart (1945) studied this question.