Many goods markets are characterized by domestic products having a disproportionate market share. Taking the set of products available on a market as given, the main candidates for creating this "home bias" are preferences for home goods or trade cost reflected in higher prices of imports or weaker distribution networks for imported goods. We explore the contribution to home bias of these factors using a structural model of demand on very detailed data on wine sales in New Hampshire (brand per

more »

... eek per store over a one year period). Preferences rather than trade costs are the main force behind the home bias. Simulations where we confront New Hampshire consumers with an exogenous set of products (equally detailed data from Sweden) points to that the preferences for home goods fall far short of generating important home bias with this alternative choice set. Our findings reinforce recent work that stresses the extensive margin (new products and new suppliers) of trade.