150.4
A Scientific Controversy Across the Channel : Unemployment Theories, Sociology and the Rise of Mathematical Statistics in the Early 20th Century

Monday, July 14, 2014: 4:15 PM
Room: Booth 49
Oral Presentation
Christian TOPALOV , Centre Maurice Halbwachs, École Hautes Études Sciences Sociales, Paris, France
In the early years of the 20th century, on both sides of the Channel, social reformers were struggling to define what « unemployment » could be. Political economy, sociology and social statistics were mobilized in the process of shaping the novel category. Defining unemployment scientifically seemed to imply that the unemployed should be enumerated or otherwise transformed in solid figures. Various statistical devices were imagined in both countries for trying to solve that same puzzle, but quite different scientific languages were mobilized : William Beveridge used the language of political economy (1909), Max Lazard that of Durkheimian sociology (1909), and Arthur Bowley the new tools provided by mathematical statistics (1912).

An argument developed in professional journals between Lazard and Bowley, who doubted of the statistical techniques the other one was using : Lazard had calculated an occupational unemployment rate, Bowley an index-number of the variations in the volume of unemployment. In order to describe this controversy, one can use the notion of « national scientific styles » and observe how tools circulated between British statisticians and French ones, through the International Statistical Association and an intense exchange of literature. The limited reception of the correlation coefficient among French statisticians raises an interesting general issue of the conditions of the circulation of knowledge.

Accounting for the difference between Lazard’s and Bowley’s formalisations implies analyzing sociologically what they were designed for and the relations that were being established between scientists and the users of their knowledge in the respective labour administrations. In both countries, a close relationship between scientists and administrators conditioned the development of abstract statitistical forms that could be applied to formalizing a new social issue and contemplated public policies : in France unemployment insurance by industries, in Britain public works providing jobs when an unemployment crisis would be looming.