Policy Volatility and the Temporal Orientation of Innovation

• In Event: Governing Clean Technology Diffusion Across Society: From Firm Innovation to Household Adoption

Abstract

Innovation is widely recognized as essential for addressing large-scale societal challenges such as climate change, energy transition, and industrial decarbonization (Mazzucato, 2018). Yet, despite the central role of time in shaping processes of technological change for sustainability transitions, existing innovation scholarship has paid limited attention to the temporal dimension of innovation. Extant green innovation literature distinguishes between innovation types, such as product versus process innovation (Demirel and Kesidou, 2011), fundamental versus applied (Popp et al., 2010), and radical versus incremental innovation (Dechezleprêtre et al., 2014). While these distinctions are important, and prior work highlights how the focus of inventive activity evolves over a technology’s lifecycle (Huenteler et al., 2016), they do not adequately capture whether innovation is oriented toward current market needs or toward anticipated future market needs. This distinction is particularly consequential for long-horizon challenges like climate change, where addressing the problem requires technological changes that extend beyond current market demands.

This paper introduces the concept of long-term innovation (Doblinger et al., 2022) into the study of policy environment and firms’ strategic behavior. We examine how policy volatility influences firms’ allocation of innovative effort between long-term and short-term innovation in the clean energy sector. Empirically, we focus on the wind energy industry across OECD countries, a setting characterized by high capital intensity, long investment horizons, and strong dependence on policy support. This makes it particularly well-suited for analyzing how changing policy environments shape firms’ innovation strategies.

Our dependent variable is the orientation of innovation activity. Following Doblinger and colleagues (2022), we apply an LDA topic modelling algorithm to the text of patent titles and abstracts to generate a set of latent topics, which we further validate and interpret manually. We use these topics in conjunction with additional keyword searches to classify patents into long-term and short-term innovation categories. This method allows us to distinguish innovations aimed at more immediate technological refinement, such as improving blade materials, from those associated with broader future-oriented technological development, such as developing components for floating wind turbine platforms. We then assess how firms respond to volatility across five major clean energy policy domains: carbon trading schemes, renewable energy certificate or trading systems, carbon taxes, public research and development expenditures, and price support mechanisms for solar and wind technologies through feed-in tariffs. We construct four measures to capture policy volatility: policy adjustment intensity, instability relative to historical trends, frequency of policy change, and unexpected policy movement.

We hypothesize that policy volatility is positively associated with long-term innovation. Rather than uniformly suppressing innovation, policy volatility may reshape firms’ innovation portfolios. In volatile environments, firms are likely to reduce short-term innovation while reallocating resources toward long-term innovation to build strategic flexibility and secure future advantages. This perspective highlights the importance of temporal orientation in understanding how policy volatility shapes firm strategy in clean energy transitions.

Authors

• Myriam Gregoire-Zawilski, University of Ottawa
  Presenting Author

• Xue Gao, Florida State University
  Non-Presenting Co-Author

• ZIYAO ZHU, Florida State University
  Non-Presenting Co-Author