Competition between trading institutions like eBay and many ''business-to-business'' (B2B) platforms can produce outcomes for buyers and sellers that are worse than having a single monopoly provider of ''match-making'' services. That is the central conclusion of research by Carlos Alós-Ferrer, Georg Kirchsteiger and Markus Walzl, published in the March 2010 Economic Journal.
In their study, designers of trading institutions decide on the rules of their platforms while buyers and sellers learn which platform to use. The analysis shows that platform providers have a strong incentive to introduce trading rules biased against either buyers or sellers whenever they face competition from other providers.
These findings suggest that, despite economists'' general enthusiasm for the benefits of competition, matchmaking as offered by trading platforms is a service where increased competition might actually reduce market efficiency. The researchers conclude that competition policy for such markets needs to be designed with utmost care.
Competition is good, say economists. It results in lower prices, which directly benefits consumers, and provides appropriate incentives for innovation on the side of the producers.
That is, as long as the market fits the consumer-producer microeconomic paradigm. The last few decades have seen an increase in the importance of a qualitatively different market structure, where one should more properly speak of sellers and buyers. These do not interact directly with each other but through a trading institution, for example, a shop, an agency or an online platform.
The rise of the internet has dramatically increased the importance of such trading platforms. Internet marketplaces like eBay readily come to mind, but in terms of revenue many highly relevant examples do not involve consumers at all, as is the case with B2B platforms.
In B2B platforms, large quantities of relatively standardised goods are traded on a daily basis – for example, vitamins, acids, sheet metal or other base products of the chemical, pharmaceuticals or the construction industry. In such markets, there is a third party involved in the trade: the trading institution itself, which generates costs and revenues and is an economic activity in its own right.
These trading institutions offer a service: matchmaking, the organisation of trade, etc. Does the old ''competition is good'' saying still apply to these markets? Should we hope for more competition between trade institutions as well? Does the emergence of these institutions enhance competition and allocative efficiency and thereby represent an indisputable benefit for society?
Alas, that seems not to be the case. Competition between trade institutions can very well reduce efficiency. The reason for this is that the market for trading platforms – and thereby the optimal business strategy of platform providers – is substantially different from ordinary commodity or service markets.
A successful trading institution not only has to secure its share of the revenues from trade: it also needs to attract customers. These two objectives are often at odds.
Consider, for example, a trading platform that requests a certain fraction of trade revenues as a price for its services. Increasing this fraction will raise the profit made with each transaction between buyers and sellers, but it will also make the platform less attractive for customers.
To mitigate this conflict, platform providers not only attempt to maximise their share of the pie, they also carefully design the mechanism governing trade at their institution (the rules). Different trade mechanisms may well have different effects on the two sides of the market.
A posted offer market, for example, generates prices that are biased in favour of sellers while a posted bid market favours buyers. But a biased trading rule that favours sellers not only makes this platform relatively more attractive for the sellers, it also prevents market clearing (that is, not every buyer who has a willingness to pay above the unit costs of production actually purchases the good) and thereby results in inefficient, less competitive outcomes within the platform.
In the model in this study, market designers actively decide on the characteristics of their respective platforms while buyers and sellers learn which platform to use. The analysis shows that platform providers have a strong incentive to introduce biased trading rules whenever they face competition from other platform providers.
A monopolistic trading platform would never favour one market side, as sellers and buyers lack the opportunity to trade elsewhere and the only objective of such a platform is to maximise its revenue. In contrast, two competing platforms have strong incentives to opt for biased trading rules.
If, for example, sellers would make zero profits with an unbiased trading rule (perhaps due to perfect competition among sellers who produce with scale-independent unit costs), competing platforms would bend their price-determining rules in favour of the sellers to offer them at least some part of the surplus and thereby attract all sellers from rival platforms.
Hence, competition leads trading platforms to introduce biased prices at which markets do not clear and thereby results in less competitive trading outcomes and reduced efficiency. In this sense, match-making as offered by trading platforms is a service where increased competition might reduce market efficiency. Therefore, competition policy for such markets needs to be designed with utmost care.
''On the Evolution of Market Institutions: The Platform Design Paradox'' by Carlos Alós-Ferrer, Georg Kirchsteiger and Markus Walzl is published in the March 2010 issue of the Economic Journal.
