Feenstra and Ma (2008) develop a monopolistic competition model where firms
choose their optimal product scope by balancing the profits from a new variety
against the costs of “cannibalizing” sales of existing varieties. While more productive
firms always have a higher market share, there is no monotonic relationship between
firms’ productivity level and their choices of product scope. In the model having a
higher market share means that firms are hurt more by the “cannibalization effect”.
Therefore, the incentive to add more products weakens as productivity rises. This
leads to Lemma 3 in Feenstra and Ma (2008): There is an inverted U-shaped
relationship between firms’ productivities and the range of varieties they choose to
produce. This empirical note takes this Lemma to the data for firms from German
manufacturing industries. Empirical evidence is in line with the results from the
theoretical model.