Multiproduct firms account for a large fraction of economic activity and actively change their product mix. This dissertation consists of two chapters studying the product dynamics of multiproduct firms.

In Chapter 1, I investigate changes in product scope, the number of products that a firm offers, over the business cycle and decompose the impact of such changes on aggregate output. I use the Nielsen Retail Scanner data of U.S. consumer goods purchases for 2007-2014. I find that firm product scope is an important margin of adjustment. In the recession, firm product scope decreases on average and the decreases are heterogeneous across firms. Such product scope changes affect aggregate consumption and output by changing the total number of products and by affecting firms' markups. These impacts are shown in a model featuring heterogeneous multiproduct firms, oligopolistic competition and free firm entry. Firm and aggregate outcomes vary in different states of economic activity. In a recession state, lower average product scope implies a lower number of product varieties, which disincentivizes consumption. Additionally, since the most productive firms have higher market shares, as the data suggests, they charge higher markups as oligopolistic competitors. The average markup goes up and further decreases consumption.

In Chapter 2, I explore the within-firm product dynamics beyond the product scope changes. The measures of interest include within-firm product entry, exit, net entry and reallocation. The net entry is related to the product scope changes (and is expected to be procyclical), and the reallocation is an important channel for endogenous growth. To investigate the cyclicality of these measures and locate the firm-level factors that affect the within-firm product dynamics, I combine the Nielsen Retail Scanner data with the Compustat fundamental data for the listed firms, by matching the firm names. The listed Nielsen firms have lower net entry and reallocation rates in the recession years. The regressions of the net entry and reallocation rates on the unemployment rate reveal that the net entry is significantly procyclical on average and the reallocation is not. Moreover, firms' financial constraint condition and R&D significantly affect the reallocation, but not the net entry.