A LONDONER with a sudden urge for giant African land snails could do worse than head to the bustling marketplace in Brixton, a part of south London that is home to many people from Africa. A craving for freshwater fish in New Delhi is best met by haggling for the freshest rohu with fishmongers in Chittaranjan Park, the heart of the city’s fish-loving Bengali community.

Markets that cater to migrants, whether from a different part of the country or from far-flung corners of the globe, are not just great for gourmands. They are also testament to the fact that people often retain very strong preferences for the kinds of food they grew up eating. Just ask the expatriate Britons who flock to “Tea and Sympathy” in New York’s Greenwich Village for pots of Marmite, a yeast-based spread whose delights baffle other nationalities (and many of their own compatriots). Such nostalgia is the most obvious example of the influence exerted by loyalty to the brands of your youth. A new study by economists from the universities of Tilburg and Chicago* tracks the consumption patterns of American households over two years and finds striking evidence that such loyalty is widespread, deep and long-lasting.

Past research has shown that people are often willing to pay much more for a favoured brand than for seemingly identical alternatives. It is not always obvious why. For instance, people routinely express a strong liking for a brand that they are unable to tell apart from rivals in blind tests. And many studies have found that advertising alone cannot explain the strength of brand loyalty. So it seems plausible that a person’s past may play a role. Someone who spent their formative years in Massachusetts, say, may always hanker for Samuel Adams, a local beer.

The new study finds a clever way to test this idea. The researchers had data on the purchases of 238 kinds of packaged goods by 38,000 American families between 2006 and 2008. For each of the goods in question, the data allowed them to calculate the share of the most-preferred brand as a fraction of the purchases of the two leading brands. Different regions showed a lot of variation: there were clear local patterns in consumption, although the same brands were available everywhere.

But 16% of people studied were migrants: they had grown up in one state and moved to another. They had the same options, in terms of what was on offer and at what price, as everyone else in their adopted home. But although they consumed more local favourites than someone in their native state would have, they bought fewer local hits (and more of the favourites from back home) than a longtime resident. And this gap between the purchases of migrants and that of the locally born was quite stubborn: although it faded the longer a person lived in their new state, it still took 20 years to halve in magnitude. Even 50 years on, it was still large enough to show up in the data.

If this is generally true, it has important implications. For one thing, the benefits of being the first brand into a market could last longer than might be assumed. But David Atkin of Yale University suggests in another new paper** that the effects of habit formation in consumption may also lead economists to rethink the way they calculate the gains from trade. This is because opening up to trade is in some ways akin to migrating. It changes the composition and prices of the goods that are available to a person. In particular, it can raise the relative prices of the goods that a region or country has a comparative advantage in, such as crops that the country’s climate or soil favour. These are the things that would have been relatively cheap and common in a closed economy and therefore the things that people might have acquired a taste for. To the extent that such preferences persist, people will benefit less from the increased variety of goods and altered relative prices that trade brings about than they would do if habits were not a significant determinant of consumption.

Creatures of habit

Mr Atkin uses detailed data about people’s food choices in India, which covers a large number of climatic zones which have specialised in different crops. Despite being part of the same country, the preponderance of internal barriers to trade—including border taxes and a maze of state-level regulations—means that its regions are best thought of as being only partially open to trade. As a result, and as Mr Atkin’s data show, the foods that a region specialises in producing are indeed cheaper there.

What is more, people in that region show a distinct preference for these “regional specialities”. Tracking changes in what Indians eat reveals that, for every rupee spent on food (in a period when the national consumption of calories fell), people’s intake of calories declined most in regions where relative-price rises were more concentrated in local favourites. In theory, people should adjust their purchases of food when relative prices change so that they maximise their calorific intake. In practice, food habits meant that consumers kept buying the things they knew they liked even though they were relatively expensive.

Of course, people move between states in India just as they do in America. Mr Atkin’s analysis of their food choices suggests that Indian migrants, too, carry their food preferences with them. As a consequence, migrant families consume fewer calories per rupee of food expenditure than non-migrants do. Economists have shown that liberalising trade increases incomes by enabling specialisation. But because people are creatures of habit the effects of a given increase in income on people’s well-being will be smaller than if people’s tastes adjusted immediately. At least the world’s cities have more interesting markets as a result.