Trade matters. There is indisputable evidence across countries that firms that participate in international markets are more productive. But literature is less clear as to why. What are the mechanisms driving the trade-productivity premia observed in the data? To investigate this, we use an innovative dataset originally developed by Wales et al (2018), which combines administrative data on trade in goods with survey data on firm labour productivity. We extend this dataset to include trade in services and updated trade in goods information between 2017-2022. This dataset, covering the 2005-2022 period, allows us to build the most comprehensive picture of British traders and to, not only estimate the most up-to-date relationship between trade and productivity but also to disentangle the role of self-selection of productive firms into exporting, from other causal impacts of trade on productivity via, for instance, technological upgrading or learning-by-doing.
We find that firms in Great Britain that exported or imported goods or services were 35% more productive than their non trading counterparts, after controlling for differences in industry, age, employment and foreign ownership status. However, consistent with the literature, we find that most of the productivity gap is due to selection, ie. it is more productive firms that choose to enter international markets. Using trade in goods data for exporters born after 2005, we find that goods exporters-to-be are 59% more productive than firms that never export even before the time of first export. Using a differences-in-differences design among goods exporters-to-be only, where conditional on observable characteristics the timing of trade is assumed to be exogenous, we find that there are no significant differences in productivity before exporting, but labour productivity of exporters increases 11.9 % compared to exporters-to-be, in the first year of exporters’ first entry in international markets. There is significant heterogeneity at the intensive margin, as our results show that these productivity benefits accrue only to large exporters.
While the estimated impact takes effect too quickly to be attributable to technological upgrading, our estimates suggest goods exporters increase capital investment and ICT, 5 years after entering international markets. These estimates are relevant for policymakers aiming to use trade policy as a lever to boost UK growth. Future extensions of the work should investigate the dynamic trade behaviour of the firm, integrating in the analysis changes in the product and country mix over time, leverage administrative data on trade in services and finally, take advantage of employer employee information to understand the impact of trade on workers.
