Working Paper: NBER ID: w14668
Authors: Arnaud Costinot; Lindsay Oldenski; James E. Rauch
Abstract: What determines the boundary of multinational firms? According to Williamson (1975), a potential rationale for vertical integration is to facilitate adaptation in a world where uncertainty is resolved over time. This paper offers the first empirical analysis of the impact of adaptation on the boundary of multinational firms. To do so, we first develop a ranking of sectors in terms of their "routineness" by merging two sets of data: (i) ratings of occupations by their intensities in "problem solving" from the U.S. Department of Labor's Occupational Information Network; and (ii) U.S. employment shares of occupations by sectors from the Bureau of Labor Statistics Occupational Employment Statistics. Using U.S. Census trade data, we then demonstrate that, in line with adaptation theories of the firm, the share of intrafirm trade tends to be higher in less routine sectors. This result is robust to inclusion of other variables known to influence the U.S. intrafirm import share such as capital intensity, R&D intensity, relationship specificity, intermediation and productivity dispersion. Our most conservative estimate suggests that a one standard deviation decrease in average routineness raises the share of intrafirm imports by 0.26 standard deviations, or an additional 7% of import value that is intrafirm.
Keywords: multinational firms; adaptation; intrafirm trade; routineness
JEL Codes: F23; L14
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| less routine sectors (R59) | higher share of intrafirm trade (F12) |
| routineness (L23) | boundary of multinational firms (F23) |
| average routineness decrease (C69) | share of intrafirm imports (F23) |