Media Briefings


  • Published Date: December 2012

The most innovative and productive firms send their exports to the world’s richest countries. That is the central finding of research by Rosario Crinò and Paolo Epifani, published in the December 2012 issue of the Economic Journal. Their analysis of a representative sample of Italian manufacturing firms suggests that what firms produce – and how they produce it – is closely related to where they sell it.

This has important implications, particularly for manufacturers in poorer countries, for whom quality upgrading may be a prerequisite for effective access to richer countries’ markets. Moreover, the research suggests, trade liberalisation between richer and poorer countries may have a limited impact on rich countries’ industrial structure.

The study documents some new and perhaps surprising facts about the pattern of firms’ exports across destinations. In particular, analysing a reliable dataset used in other studies, it finds a strong and robust negative correlation between firms’ productivity and their share of total exports to low-income destinations. This fact seems at odds with the common wisdom, positing that only the most productive firms are profitable enough to break into harder-to-reach destinations.

The researchers argue that this and other empirical regularities can arise from the interplay between cross-firm heterogeneity in product quality and cross-country heterogeneity in quality consumption. Specifically, they conjecture that more productive firms tend to concentrate their sales in high-income markets because they produce higher-quality products for which relative demand is higher in high-income destinations. This conjecture is strongly supported by the data.

The researchers start by providing extensive evidence of a negative cross-firm correlation between productivity (however measured) and the export share to low-income destinations. The negative correlation holds strong independent of sample size and is not affected by outliers, estimation method and specification details.

Next, they formulate a general model that clarifies the main insight behind their interpretation of the evidence, and estimate its parameters structurally. The crucial assumptions for the results are that consumers choose quality consumption based on their income and firms choose product quality based on their productivity.

Consistent with the theory, estimated parameters imply the preference for quality to be monotonically increasing in per capita income of the destinations. On average, the preference for quality is almost three times higher in high-income destinations. In the richest destination (North America), it is roughly 20 times as high as in the poorest destination (Africa).

Moreover, with the estimated set of parameters, the model matches the data well and correctly predicts a negative correlation between exporters’ productivity and their export share to low-income destinations.

Importantly, the model also predicts a strong negative correlation between firms’ innovation activities and their export share to low-income destinations. The study finds that a number of proxies for innovation activities (for example, R&D intensity) are strongly negatively correlated with the export share to low-income destinations.

Finally, the model predicts the correlations between the export share to an individual destination and productivity, R&D intensity or product quality to be all increasing in the destination’s per capita income. The study finds strong support for these predictions.

The results have implications for North-South trade liberalisation, which may have a limited impact on rich countries’ industrial structure because of the trade-reducing effect of preference heterogeneity on the one hand and firm heterogeneity in product quality on the other.

The researchers comment:

‘Although in recent years economists have dramatically improved our understanding of firms’ export behaviour, the determinants of the popularity of foreign destinations from the standpoint of domestic exporters are not yet fully understood.

‘We hope that by showing how firms’ exports may depend on the interplay between productivity, product quality and quality consumption, our contribution can shed light on this important issue.’


Notes for editors: Productivity, Quality and Export Behaviour by Rosario Crinò and Paolo Epifani is published in the December 2012 issue of the Economic Journal.

Rosario Crinò is at CEMFI in Madrid. Paolo Epifani is at Bocconi University in Milan.

For further information: contact Romesh Vaitilingam on +44-7768-661095 (email:; or Paolo Epifani via email: