Abstract
The author presents data from a survey of newspapers on the incidence of the dismissal of employees during industrial disputes and raises the question that studies of what managers actually do are likely to produce different results compared to those produced from questionnaires addressed to managers.