Estimating Cross-Industry Cross-Country Interaction Models Using Benchmark Industry Characteristics

Abstract Cross-industry cross-country models are used to address a wide array of questions in economics. They do so by analysing how the economic performance of industries in different countries depends on an interaction effect between industry and country characteristics. As the relevant industry characteristics are unobservable in most countries, they are approximated by industry characteristics in a benchmark country. We show that this approach generally yields biased estimates of the industry–country interaction effect. The sign of the bias depends on whether or not technologically similar countries tend to be similar in other country characteristics. We propose an alternative estimation approach.