The consumption correlation puzzle refers to the concept that consumption is much less correlated across countries than output. In a complete market world in which agents are able to exchange every good with other agents without transaction costs, you may expect that domestic consumption growth doesn’t depend too much on country-specific output.

Therefore, the theory suggests that consumption should be much more correlated across countries than output. But empirical studies indicate that consumption is much less correlated across countries than output. In other words, there isn’t much international risk sharing in terms of consumption.

Explanations of the consumption correlation puzzles emphasize a variety of market imperfections and, therefore, transaction costs. For example, empirical evidence indicates that financial markets more effectively promote risk sharing within a country than between countries. Therefore, financial markets are able to smooth consumption among the U.S. states at a much higher degree than among developed countries.

In this context, the term smoothing refers to counterbalancing the increase or decrease in household consumption by exporting or importing consumption goods, respectively.

The existence of nontraded goods can also break the link between prices and quantities and make it harder for trade to smooth household consumption. Suppose that households consume goods that cannot be traded — for example, services like haircuts. If a positive technology shock increases the supply of services, households cannot smooth their consumption of services by exporting haircuts to other countries.