ArgumentWhen the New Deal administration attained power in the United States, it was confronted with two different problems that could be linked to one another. On the one hand, there was a huge problem of unemployment, affecting everybody including the white-collar workers. And, on the other hand, the administration suffered from a very serious lack of data to illuminate its politics. One idea that came out of this situation was to use the abundant unemployed white-collar workers as enumerators of statistical (...) studies. This paper describes this experiment, shows how it paradoxically affected the professionalization of statistics, and explains why it did not affect expert democracy despite its Deweysian participationist aspect. (shrink)