Leviathan as an inventor: An extended agency model of state-owned versus private firm invention in emerging and developed economies

We examine the relative merits of state (SOE) and private (POE) ownership on firm invention output in emerging as well as developed economies. Firm SOE affords a vehicle by which governments throughout the world encourage risky investment in new technological paths. Yet, from a traditional agency-theoretical standpoint, SOEs are plagued by low-powered incentives and dysfunctional involvement (e.g., wasteful budgets and political meddling), which reduces SOE performance, relative to POEs. To examine this paradox, we build an extended agency model which, rather than overlooking agency cost drivers, entertains a more balanced view, in which these so-called agency "liabilities" in fact benefit SOEs, particularly in the sphere of inventions. We argue that, given relatively higher levels of managerial autonomy, SOEs may outperform POEs in some types of inventive output. We further propose that key boundary conditions operate to modify this base effect, in that contexts lacking political constraints - a condition found in most emerging economies - increase the risk of political meddling, reducing managerial autonomy so that any SOE inventive advantages wane significantly. We also hypothesize that this negative effect of weak institutions is attenuated in high invention productivity sectors. Empirically, we track the frequency, pioneerism, and impact of patent inventions of 521 SOEs and matched POEs for 16 years, across 43 countries and 22 industries. We find empirical support for our hypotheses based on a robust set of fixed-effects regression analyses, aided by a battery of checks to assess self-selection and omitted variable bias.