A dynamic model of vertical integration for the American pulp and paper industry

The focus of this research is to learn about the factors that influence the decision of a manufacturing firm to vertically integrate into the production of its input. The American paper industry has a feature that makes it particularly suitable for this purpose: over the years paper mills of apparently similar characteristics have made different decisions with regards to their integration status. This work draws on the insight that there must be some unobserved mill characteristic that drives the decision process for a mill. Mills ’ choices of whether to exit the industry, and with regards to their integration status when they choose to stay in operation, depend on their productivity. This generates selection and simultaneity biases in a reduced form estimation. In order to deal with these issues, I propose a dynamic model in the spirit of Olley and Pakes (1996). This approach not only takes care of the estimation biases, but also allows me to learn about the unobserved characteristics of the firms in my data, and to use them to determine which firms vertically integrate and which firms do not. In addition, the model I propose allows me to learn about how vertical integration affects productivit