Anomie or imitation? The Werther effect of celebrity suicides on suicide rates in 34 OECD countries, 1960-2014

Anomie and imitation have been prominent mechanisms explaining the Werther effect, i.e., the effect of celebrity suicides on a general population's suicide rate. This study presents a new approach to empirically disentangle both mechanisms. Imitation theory suggests that celebrities act as role models, and that the Werther effect is triggered by the status of the celebrity in question. Anomie theory, on the other hand, suggests that the Werther effect is triggered by the unexpectedness of the event. To this end, we empirically compare the effects of celebrity suicides with the effects of celebrities who died unexpectedly from causes other than suicide (accidents, illnesses, alcohol abuse). Based on language and page-link data from 3,855 Wikipedia pages of 495 celebrities who committed suicide between 1960 and 2014, we measure the status a celebrity has in a particular country and calculate the potential country-specific imitation effect of their suicide. In the same manner, we measure status effects of celebrities who died unexpectedly from accidents, illnesses, or alcohol abuse to reflect anomie-related effects. We use these measures in a time-series cross-sectional dataset for 34 OECD countries to assess their effects on a country's overall annual suicide rate. Fixed-effects analyses reveal that country-specific status effects of celebrity suicides lead to significant increases in overall suicide rates, while anomie-related, unexpected celebrity deaths show no effects. The findings remain robust across a number of alternative specifications, such as controlling for further anomic factors at the macro level (divorce or unemployment rate, for instance). We conclude that the results support the imitation mechanism as an essential social explanation for the Werther effect.