Title: Theory of collective opinion shifts: from smooth trends to abrupt swings

Abstract: We unveil collective effects induced by imitation and social pressure by analyzing data from three different sources: birth rates, sales of cell phones and the drop of applause in concert halls. We interpret our results within the framework of the Random Field Ising Model, which is a threshold model for collective decisions accounting both for agent heterogeneity and social imitation. Changes of opinion can occur either abruptly or continuously, depending on the importance of herding effects. The main prediction of the model is a scaling relation between the height h of the speed of variation peak and its width $w$ of the form h ~ w^{-kappa}, with kappa = 2/3 for well connected populations. Our three sets of data are compatible with such a prediction, with kappa ~ 0.62 for birth rates, kappa ~ 0.71 for cell phones and kappa ~ 0.64 for clapping. In this last case, we in fact observe that some clapping samples end discontinuously (w=0), as predicted by the model for strong enough imitation.