There are a number of competing explanations for the proliferation of barter in Russia. Most firm managers maintain the view that barter is explained by the liquidity squeeze due to tight monetary policy. In contrast, Russian government officials often claim that managers use barter to avoid paying taxes in full. Outside investors often claim that managers use barter to divert profits, protect their jobs, and delay restructuring.

In an attempt to clarify this debate, we argue that discussion of barter in Russia is incomplete without taking into account the role of market structure. Indeed, the anecdotal evidence suggests that natural monopolies are the most likely of all Russian firms to engage in barter.

Building a model of barter as a means of price discrimination, we find that even though barter is inefficient, it can emerge in equilibrium if markets are sufficiently concentrated. In fact, the amount of barter increases with concentration. The intuition is straightforward. Since monopolies, in equilibrium, will usually underproduce relative to the social optimum, sellers may be interested in an additional channel of sales, even if that channel is costly.

In order to test the model, we built a data set that matches a survey of managers on the degree of barter in their firms with firm-level data from official statistics. The empirical analysis supports our model: barter positively and significantly depends on market concentration.

Our result raises a legitimate question. If barter is explained by high concentration of market power, why is it observed in Russia but in virtually no other economy? One answer to this question would be that Russian markets are more concentrated than markets elsewhere. This claim is well accepted by the general public and policymakers, but it is not supported by the data.

Our model may offer another explanation. For the same level of concentration there may be two stable equilibria: one with barter and one without barter. Therefore there may be a path dependence. In 1995 a liquidity shock threw the economy into a high barter state. Since that time, price flexibility should have restored an equilibrium level of real money balances. The real money supply, however, remains two to three times lower than before the liquidity shock, supporting the multiple equilibria hypothesis.

The multiple equilibria argument is rather common in modern literature on transition and development. It constitutes the essence of the so-called "post-Washington consensus" that is gradually replacing the Washington consensus on economic transition. The post-Washington consensus states that institutions matter a great deal for economic transition and may fail to emerge spontaneously. Government should intervene to promote good institutions, otherwise the economy will find itself in a low-level equilibrium.

What our model suggests is not simply a restatement of the fact that Russia is in a low-level equilibrium. We show that at some level of competition the barter equilibrium disappears and industry jumps to the no-barter equilibrium. This argument has important policy implications. In order to reduce barter, government should promote competition. Moreover, even if competition policy may have had only a small effect on barter so far, the government should not give up. Our model (along with empirical analysis) suggests that barter may fall dramatically when a certain threshold level of competition is achieved.