The definition of the relevant antitrust market is a critical first step in a competition economics investigation. For certain products, publicly available data in the form of the large sample household consumption expenditure survey can be used to implement small but significant and non-transitory increase in price tests for a range of consumer goods at different levels of geographic aggregation. Demand elasticities for a set of consumption goods are estimated through use of the Heckman sample selection model, which corrects for sample selection biases and specification errors that appear in survey data. Among other things, the conditions under which biased elasticity estimates lead to overly wide or overly narrow market definition are also illustrated.
