This paper attempts to determine whether conditions
amenable to successful selective interventions to capture
cross-industry externalities are likely to be fulfilled in practice.
Three criteria are proposed for good candidates for industrial
promotion: that they have strong interindustry links to the rest
of the economy, that they lead the rest of the economy in a
causal sense, and that they be characterized by a high share of
industry-specific innovations in output growth.
According to these criteria, likely candidates for successful
intervention are identified in the Korean data, but these were not
the industries that were in fact promoted.