Karl G. Karsten

In the 1910s charts and graphs had become a prominent subject in the theoretical discussion of methodology in the social sciences,[8] and Karsten was one of the first to publish a book solely on this topic. Karsten (1923) explained:

In and since the [First World] War the use and developments of charts has been almost phenomenal — so large, indeed, that one able economist who is interested in such things think that we as a country have gone chart-mad. But the development has not been confined to this country, and it has a very solid base in practical utility. There is little question that the chart represents a genuine saving in time and mental effort.[9]

In 1923 he published his most known work "Charts and graphs: An introduction to graphic methods in the control and analysis of statistics." This work was especially focussed on statistics, and years ahead of its time.[10]

It the 1910s statistics and graphs were introduced in production control and management in the US and in Britain.[11] Karsten's work, over 700 page and 500 illustrations, was designed for educational purposes and not as handbook for businessmen. In Britain it inspired Thomas Gerald Rose (1886-1963), who wrote since the 1930s on higher control in management, management accounting and business charts.[12]

In his 1924 "Theory of Quadrature in Economics" Karsten "showed a detailed understanding of how some economic time series could be generated as partial sums of others and that there could, therefore, be a pair of series where one was the cumulation of the other."[13] The basic idea of this quadrature was described by Karsten:

Two forces are said by Mr. Edge to be in quadrature when they trace curves such that the fluctuations of one of the curves correspond to the fluctuations of a curve of the integration or cumulation of the data of the other curve. When a cause-and-effect relation exists between such two phenomena, the first may then be said to be cumulatively affected by the second.[14]

An furthermore:

The quadrature method therefore applies cumulation to obtain a series representing the integral function, and differencing to obtain a series of data representing the derivative.[14]

Karsten suggested in this theory, that "the aim of correlation analysis should ultimately be the recovery of causal connections."[15] The next year Jan Tinbergen confirmed this claim of Karsten in an ongoing discussion.[15]

Boumans (2004) recalls, that "Karsten ... had touched 'not without merit' on the problem of causal relations, had shown the existence of cumulative relations between the three Harvard barometer indices, which he interpreted as causal relationships. In the first place, he found, using correlation analysis, that the cumulative values of the Harvard B-index parallel those of the Harvard A- index, with a lag of 3 ... Second, he found the empirical relationship that the C-index was a cumulative of bot the A and B indices ... Thus according to Karsten (1926:417), the B-index was the 'generating force' of the three; the other two indices depended upon, and were derived from, changes in the business index."[15]

Kasten had founded Karsten Statistical Laboratory in New Haven, Connecticut, specifically to develop so called "barometers", to theoretically forecasts business conditions. Lhabitant (2011) explained that these forecasts included "barometers of volume of trade, of building activity, of interest rates, of the wholesale price level, of indices of certain industries, of railroad stocks, of public utility stocks, of steel stocks, of oil stocks, of automobile stocks, and of store stocks."[16]

His 1931 Scientific forecasting was based on Karsten's experience as stock market analyst, and was written as a handbook for stockbrokers. This work on scientific forecasting focussed on the methods and application to practical business and to stock market operations, and relied on the techniques of the economic barometers.

According to Keuzenkamp (1987) "Karsten was fairly optimistic about his abilities to make prognoses, but, as he had to admit , 'at the present time this investigation can not account for economic changes which are wholly due to the element of mob-, crowd- or herd-psychology'.[17] In general, however, Karsten's conclusion is that these psychological factors are relatively unimportant, which leads him to find reason to state that his results support "economic determinism". There is only one exception: the stock market. Of course it is not surprising that Karsten was moderate with claims about stock market prediction, so shortly after the generally unexpected crash of 1929."[18]