Literature review

Innovation analysis and game theory: a review

Andrzej Baniak
Central European University, Budapest, Hungary; Wroclaw University of Economics,Wroclaw, Poland

Igor Dubina
Altai State University, Department of Economics Information Systems, Barnaul, Russian Federation

PP: 178 - 191

Abstract

In this paper we review game theory and innovation related papers published in the last 25 years. We divide our review into three levels of innovation games: 1) intra-organizational games, which are played within the firm and main players are an innovator, a project manager and\or resource administrator; 2) inter-organizational games, where main players are a firm and its competitors, partners and customers; 3) meta-organizational games, where main players are a social planner and an innovative entrepreneur. Compared with the previous periods, one can observe a significant increase of the number of papers, which use game theory models to analyze the innovation processes. Moreover, there is a tendency to move from abstract and formal models towards practically applicable and empirically tested studies.

Keywords

Innovation, R&D, game theory, competition, licensing, patents

Article Text

Game theory, which can be described as a mathematical modeling of strategic interaction among independent agents, recently become one of the most powerful analytical tools in economics, especially in microeconomics and industrial organization. In particular game theory, among other applications, offers a way to formulate predictions, delivers prescriptions and recommendations for decision makers, helps to develop and implement efficient strategies.

The purpose of this paper is to put together the main contributions of applied game theory over the last twenty five years to a better understanding of innovation process.  We believe there is a strong need for such a review at least for two reasons. First, the previous reviews on game-theoretic analysis of innovation, namely Beath, et al. (1989) and Reinganum (1984), were published more than twenty years ago. Second, the number of publications directly related to game-theoretic analysis of innovation has significantly increased in the last decades.

In this paper we attempted to identify the main trends of game theory application to innovation analysis. We tried to prove a concerted framework within which different approaches and models coherently demonstrate the mainstream in the current innovation studies based on game theory principles and methods. Although our survey is mostly concentrated on the relevant literature of the last two decades, we consider necessary to cite some authors, whose works published earlier created a base for the subsequent research.

This paper surveys works on game theory and innovation which were published in the period of 1985-2010. For this survey, we researched full-text databases, primarily EBSCO. We focused on EBSCO since this comprehensive and multi-disciplinary database is a great searching tool for such a poly-disciplinary topic as innovation. Moreover, EBSO has more than 7,100 full-text periodicals, abstracts from more than 11,200 journals and nearly 12,000 publications including monographs, reports and conference proceedings; this database covers a wider spectrum of publications than any other database.

The papers on theoretic-game analysis published in the period of 1970-1980s were reviewed in (Beath, et al., 1989; Reinganum, 1984). These reviews demonstrate that there were just a few dozens of publications directly related to game-theoretic analysis of innovation published before 1985. As our analysis shows, the number of such publications has exponentially increased for the last 25 years. According to the EBSCO research databases, there were only 6 papers referred to both game theory and innovation and published in 1960s, 20 in 1970s, 10 in 1980s, 29 in 1990s, and 105 in 2000s.

However, there is still a very narrow overlapping area of game theory and innovation analysis. For example, EBSCO registered only 170 papers containing the words relevant to "game theory" and "innovation" and published in the period of 1962-2009 and 71 papers containing the words relevant to "game theory" and "R&D".

To get more relevant works, we looked for papers with the words "game theory" and "innovations" in their title, abstract, or keyword section and got only 55 papers registered in the EBSCO for "game theory" and "innovation"/"R&D" indexes, while there are 21850 for "game theory", 665581 for "innovation", and 82191 for "R&D". This result can be roughly interpreted that 0,25% of papers on game theory consider innovation, and less than 0,01% of innovation research use game theory methods and tools. A similar, although not the same, proportion is observed in other databases (for instance, the Web of Knowledge has 17 papers referred to both game theory and innovation by topic, 1028 papers on game theory and 9612 papers on innovation). It is a rather surprising result, knowing that innovation is an extremely popular research topic, and game theory is a very important methodological part of the modern economic analysis.

The main directions of game-theoretic analysis of innovation can be divided into three levels of innovation interactions, or three types of  innovation games: 1) intra-organizational game, which is played within a firm; in this game main players are: an innovator (idea generator, project initiator or implementer, etc.), a project manager and\or resource administrator; 2) inter-organizational game, where the main players are: a firm, its competitors, its partners (e.g., venture inventors, distributors, suppliers, outsourcing allies, research centers, etc.), clients (customers), and sometimes a patent or other intellectual property holder; 3) meta-organizational game, where the main players are: a social planner (innovation policy maker, government, a social or government institution, e.g. a research foundation) and an aggregate innovative entrepreneur. The structure of all levels and their relations are depicted in Figure 1.

The main objective of analysis on the intra-organizational level is 1) to find and substantiate optimal variants of organizational and economic interaction of innovation process participants (inventor, innovation entrepreneur, project manager, investor, project implementers, etc.); 2) to secure a high level of creative activity of idea generators and project implementers; 3) to effectively stimulate knowledge sharing in the firm, etc.

The existing literature mostly concentrates on the inter-organizational and meta-organizational levels. On the level of inter-organizational interaction, game theory models help to define optimal competition and cooperation strategies, in particular to determine optimal R&D expenditure, optimal mechanism of financial relationships for innovation participants (royalty, fixed payment, innovation project profit or revenue sharing amount, etc.), the best time to introduce a new product into a market (time to launch innovation), to develop the most profitable licensing agreement, to select the right project or make a right choice of investment alternatives or innovation strategy, etc.

On the macro-level innovation games describe the interaction between an innovator and a social planner of innovation policy (e.g., government). The main questions are: What is the optimal configuration of intellectual property rights (IPR) policy? What are the welfare implications of licensing? Which industry and market structure provides the highest incentive to innovate? Game-theoretic approach helps to construct the optimal government policy with respect to innovation activity, to maximize the capitalized value of net social surplus, to enhance technological innovation and international competitiveness, etc.

The summary of our survey in terms of the problems modeled and model types used is given in the table 1.

It is important to note that a models initially developed for innovation analysis on one level can be often adapted to the analysis on another level (model transfer). For example, an intra-organizational knowledge sharing agent-based model (Yang and Wu, 2008) can be also adapted for the analysis of inter-organizational interactions in an innovation consortium.

The reviewed literature concentrates mostly on non-cooperative games, however cooperative game models are also used, mostly to solve the problem of optimal sharing of innovation project outcomes among the member of innovation alliances. In the most of cases the specificity of innovation process is often "hidden" in a model. It is typically assumed that an innovation process results in low-cost or quality improvement technology, etc. Hence researchers often go from latent and non-measurable variables (e.g. R&D effort) to some aggregate variables (e.g. R&D costs depending on R&D effort).

In the next sections of the paper, we consider more specific fields in which game theory approaches are applied to innovation analysis. In section 2 we discuss intra-organizational games. Next two sections are devoted to innovation competition; in particular in section 4 we discuss models with different innovative strategies in competitive environment. Section 5 presents game theoretic models of licensing. Section 6 is devoted to the games with deployment of innovations. In section 7 we discuss the papers about innovation consortia. Section 8 is devoted to policy problems related to innovations. Finally section 9 concludes.

View references

Navigate: Archives | Contents | Top of page

Search

Translate

Related Publications from eContent Management