VoxEU Column International trade

There goes gravity: How eBay reduces trade costs

Geographic distance continues to encumber international trade despite advances in transportation and communication technologies. This column shows that eBay, an online market, reduces the effect of distance on trade by 65%, mainly by reducing information frictions. As consumers 'put their money where their mouse is', welfare gains are largest where they are most needed, i.e. in remote countries with bad institutions.

In the 1990s, many commentators believed that with advances in transportation and communication technologies, geographic distance between countries would soon no longer encumber international transactions. Frances Cairncross (1997) famously predicted the “death of distance”. But despite some anecdotal evidence in support of the death-of-distance hypothesis (e.g. Friedman 2005), a large number of academic papers have established that distance has been thriving, rather than dying (e.g. Jacks 2009, Disdier and Head 2008).

Chaney (2011) seems to have provided the last nail in the coffin of the death-of-distance hypothesis. He argues that the persistent effect of distance on international trade flows can be explained by the need for direct interactions between trading partners resulting from information frictions, rather than shipping costs. But this would suggest that all the technological advances of the last decades have failed to reduce information frictions.

In a new CEPR Discussion Paper (Lendle et al. 2012), we breathe new life into the death-of-distance hypothesis. We argue that the right place to look for the death of distance is in online markets which, contrary to offline markets, make full use of those technologies that can effectively reduce information frictions. Indeed, online platforms, such as eBay or Amazon.com, no longer require exporters to make multiple phone calls, send faxes, write emails, and attend trade fairs and networking events in order to make business. And while importers still incur some search costs, these are brought down to a simple mouse gesture.

The heart of our paper is a dataset on cross-border transactions conducted on eBay, the world's largest online marketplace. It covers all eBay transactions between 62 countries during 2004-2007 (representative of 92% of total world trade), and disaggregated into 40 product categories. This dataset allows us to compare the impact of distance on international trade online and offline. Keeping in mind that the comparison of online and offline trade is no small feat (see Hortacsu et al. 2009), we match eBay product categories to product descriptions from the six-digit level HS classification to build comparable baskets of goods to avoid comparing online apples with offline oranges.

The death of distance online is reality

Prima facie evidence (Figure1) indicates that the relationship between trade flows and distance is indeed more flat-sloped on eBay (top panel) than offline (bottom panel).

Figure 1. The effect of distance on trade: Online vs. offline

To identify as precisely as possible the effect of distance, we use a gravity framework, controlling for other standard gravity trade costs such as the absence of a common language, a common legal system, a common border, a colonial history, or a free trade agreement. We find that distance matters 65% less online than offline.1

How eBay shrinks the world

Traditionally, economists have attributed the effect of distance to shipping costs. However, it seems unlikely that a reduction in shipping costs is driving online efficiencies, since eBay trade is typically single-item shipments to end-consumers, rather than bulk shipments that benefit from scale economies. Rather, we hypothesise that eBay reduces information frictions that are captured by the distance coefficient. We show that the distance-effect reduction is indeed largest for exporting countries which are relatively unknown to consumers (as measured by Google search results), and less trustworthy (as measured by corruption indices). This confirms that eBay shrinks the world even more for exporters from obscure and unknown places.

We also show that the distance-effect reduction is largest for information-intensive products, as measured by elasticities of substitution (Broda and Weinstein 2006). As a further test, we look at whether distance is reduced by the eBay seller-rating mechanism, which may increase buyer trust in exporters. As predicted, we find that distance matters significantly less for sellers with higher ratings.

All in all, this confirms our hypothesis that eBay reduces information frictions and establishes trust between market participants that would not otherwise have traded.

Different demographics online and offline could also be driving the different distance effects. To isolate the technology effect we show that even within a subset of countries where eBay and offline traders are likely to be most similar, i.e. in countries with low inequalities, where internet penetration is high, and from which it is easy to export, eBay still significantly reduces the distance effect.

How much would countries gain from a flatter world?

We conclude with an estimate of the gains from trade brought about by internet technologies. We use the approach proposed by Arkolakis et al. (2012) to calculate the welfare gains that would result from a drop in offline information frictions to the online level. We find that in the average country, real income would increase by 29%. Quite straightforwardly, the gains would be highest in countries remote from the rest of the world (Figure 2). It is important to note that these welfare gains would result from a reduction in the distance effect only, and not from taking all trade on eBay, where shipping costs may be higher. Still, this is quite promising in terms of the potential of technology to render world trade more efficient.

Figure 2. Welfare gains from world flattening

Conclusions

The death-of-distance hypothesis is alive – at least online. On eBay, the distance effect on trade is 65% smaller than for comparable offline transactions. We show this is because technology helps overcome asymmetric information problems often due to government failures. With a simple click, online shoppers can put their money where their mouse is and benefit from lower prices.

If offline information frictions were reduced to the level prevailing on eBay, the trade gains would translate into large increases in real income. These gains would be largest where they are most needed – in remote countries with bad institutions. E-commerce can thus be seen as a veritable development tool, as remote sellers from unstable countries integrate into world markets.

References

Arkolakis, Costas, Arnaud Costinot and Andres Rodriguez-Clare (2012), "New Trade Models, Same Old Gains?", American Economic Review, forthcoming.

Broda, Christian and David Weinstein (2006), "Globalization and the Gains from Variety", The Quarterly Journal of Economics 121(2), 541-585.

Cairncross, Frances (1997), The Death of Distance, Cambridge: Harvard Business School Press.

Chaney, Thomas (2011), "The Gravity Equation in International Trade: an Explanation", mimeo

Disdier, Anne-Celia and Keith Head (2008), "The Puzzling Persistence of the Distance Effect on Bilateral Trade", Review of Economics and Statistics 90(1), 37-48.

Friedman, Thomas L. (2005), The world is flat, Farrar, Straus & Giroux.

Hortacsu, Ali, A. Martinez Jerez, and Jason Douglas (2009). "The geography of trade in online transactions: Evidence from eBay and MercadoLibre", American Economic Journal: Microeconomics 1(1), 53-74.

Jacks, David (2009) “Revisiting the death of distance”, VoxEU.org, 12 Sep.

Lendle, Andreas, Marcelo Olarrega, Simon Schropp and Pierre-Louis Vézina (2012), “There goes gravity: How eBay reduces trade costs”, CEPR Discussion Paper 9094.


1 This difference in distance coefficients is statistically significant at the 99% level, robust to using OLS or Poisson pseudo-maximum likelihood estimations as well as to various aggregations of goods online and offline, including using disaggregated data.

1,373 Reads