Centrality of Government to Innovation in Australia: A Social Networking Analysis*

Abstract

Using a social networking approach, this study explores how government innovation collaborations, or lack thereof, might be behind the tepid productivity and innovation performance of Australian businesses. I use patent data to construct the network. The government has been very central to this network. However, since 2000, the centrality of government has declined. Using exponential random graph models, I can show that the drop in government centrality is primarily due to a narrowing down of government's focus to a smaller set of technologies and a drop in business collaborations, somewhat balanced by increasing foreign collaborations. This shift appears to be associated with changes in business innovation in Australia, where I detect a drop in the scope of patents lodged by businesses in tandem with declining government centrality. The lessons learned have important implications for policies aimed at enhancing basic innovation and promoting long-term productivity growth.