Outsourcing versus Vertical Integration: A Dynamic Model of Industry Equilibrium.

Why do supply relations vary across industries and across firms within industries? Recent evidence by Hortaçsu and Syverson (2009) shows that vertically integrated producers are more productive, their size distribution dominates (in first order stochastic dominance sense) the size distribution of not vertically integrated manufacturers and there is assortative matching of upstream and downstream plants by productivity and size. Besides vertical integration (VI) and procurement of inputs from spot-exchange markets, there exist other forms of supply relations. Networks are intermediate levels of organization between vertical integration and markets. Kranton and Meinhart (2000) link different supply relations to differences in the dispersion of buyer's idiosyncratic demand shocks. The current paper presents a dynamic stochastic model of an industry with heterogeneous firms that interact as buyers and sellers and endogenizes the vertical integration and outsourcing of firms. In the model economy, an industrial structure emerges as the result of optimal investment decisions that firms undertake under uncertainty. In order to produce, firms either vertically integrate by investing in dedicated assets to procure their own inputs, or they can invest in links to external sellers from which they get specialized inputs, or alternatively, buy their standardized inputs by transacting with suppliers in the market. This environment also includes search frictions that induce a hold-up problem to the manufacturers. The main results are: 1) an industry with higher manufacturer's idiosyncratic uncertainty shows higher outsourcing/VI ratio, 2) VI firms are larger and more efficient on average and there is assortative matching of manufacturers and suppliers, 3) downstream firms with the same productivity levels can differ in their vertical structure, and those that are vertically integrated can end up disintegrated or remain integrated.