Ricardo A. López

This paper uses plant-level data from Chile to examine the determinants of importing intermediate inputs paying special attention to the role of importing spillovers. The results show that plants that pay higher wages and plants with previous experience importing intermediate inputs are more likely to import inputs. This study also finds a positive correlation between the number of importers in the same region, regardless of their industry affiliation, and the probability of importing intermediate inputs. This suggests that importing spillovers may be important, at least in the case of Chilean manufacturing.

We use disaggregated data on Chilean plants, and the Chilean input–output table to examine the impact of agglomeration spillovers on total factor productivity (TFP). In common with previous studies, we find evidence for intraindustry spillovers, but none for general cross‐industry spillovers. This picture changes when we take vertical industry relations into account. There are important productivity spillovers from plants in upstream industries. Interestingly, a similar effect cannot be found from plants in downstream industries. The number of plants in these sectors has no effect on plant‐level TFP, just as the number of plants in other industries that are neither important upstream suppliers nor downstream customers also has no effect.

Recent studies find that exporters are more productive than non‐exporters and that entry into exporting does not increase firms’ productivity. Thus, firms self‐select into foreign markets. This paper examines productivity before entry into exporting. Using Chilean plant‐level data, we find that productivity and investment increase before plants begin to export. Moreover, productivity of entrants to exporting, but not that of non‐exporters and exporters, increases in response to increases in foreign income, before entry but not after that. The results suggest that the productivity advantage of future exporters may be the result of firms increasing their productivity in order to export.

Abstract.  Recent empirical evidence documents the superior characteristics of exporters relative to non‐exporters. Three explanations for this phenomenon have been proposed: self‐selection; learning‐by‐exporting; and conscious self‐selection. We test these three hypotheses using plant‐level data from Chile. We find that plants that enter international markets show superior initial performance compared with non‐exporters, consistent with self‐selection; we observe increases in productivity after plants begin to export, which is consistent with learning‐by‐exporting. We also find strong evidence supporting the idea that self‐selection is a conscious process by which plants increase productivity with the purpose of becoming exporters. JEL classification: F14; O54; D21