Key points are not available for this paper at this time.
We study rare events data, binary dependent variables with dozens to thousands of times fewer ones (events, such as wars, vetoes, cases of political activism, or epidemiological infections) than zeros (
King et al. (2003) studied this question.