This paper presents an overview of selected bivariate count data regressions. I describe the properties of a standard bivariate Poisson distribution and discuss the main limitation of this model – its exclusion of a zero and negative correlation between two count variables. I then present two alternative approaches, the Poisson lognormal model and conditional Poisson model. I discuss a number of their merits and compare the properties of each of the two models that allow for a flexible correlation structure. Here I show that the practical applications of these models are not limited. Many examples demonstrate where one can use these models in various areas of economics.
JavaScript is turned off in your web browser. Turn it on to take full advantage of this site, then refresh the page.