This paper studies the effects of tax incentives on firms’ choices between different types of innovation. We explore how the introduction of front-end (e.g., R&D tax credits) and back-end (e.g., patent boxes) tax incentives alter a firm’s optimal choice across product, process, or mixed innovation using a simple theoretical framework.

We then provide empirical evidence on the heterogeneous effects of these tax instruments across innovation types using European Patent Office (EPO) data. Our empirical results suggest that the introduction of patent box regimes can lead to an increase in mixed patents applications at the expense of product patents. These findings help explain the growing prevalence of mixed patents observed over time and highlight that fiscal incentives can unintentionally distort firms’ innovation portfolios- encouraging certain innovation types while may induce underinvestment in others.

Tax incentive is recognized as essential for promoting innovation and sustainable economic growth. Yet literature and policy debates have almost exclusively concentrated on the generosity of tax benefit, neglecting other policy dimensions. This paper exploits China’s reform allowing firms to claim the R&D expenditure for additional deduction while prepaying corporate income tax as a quasi-experiment to examine whether advancing tax incentives enhances innovation efficiency. We find that simply moving the timing of tax incentive around 6 months ahead would significantly raise firms’ R&D efficiency. The effect is more salient for large, high-tech, and more financially constrained firms. Mechanism analysis reveals that the reform lifts not only R&D intensity but also the share of researchers’ remuneration and R&D expenditure per researcher, indicating that the efficiency gain is mainly driven by stronger incentives for the research workforce. These findings draw important policy lessons for designing an optimal tax incentive to achieve innovation.

Open innovation—where firms leverage external knowledge and collaborative partnerships to achieve technological breakthroughs—has become critical amid rising market uncertainty. The VAT refund policy achieves tax neutrality by alleviating improper tax encroachment on corporate cash flow, creating a neutral institutional environment for collaborative innovation in new energy enterprises. Utilizing the 2018 VAT refund policy reform as a quasi-natural experiment, we select A-share listed companies in China's new energy sector. Employing difference-in-differences methodology, we examine how tax neutrality impacts open innovation through supply chain transmission (SCT) mechanisms. Results show the policy significantly enhances open innovation by improving cash flow, promoting information sharing, and optimizing risk-sharing mechanisms. Policy effectiveness is moderated by supply chain context: digitization and bargaining power amplify positive effects, while supply chain risks and excessive financialization dampen impacts. This research enriches the literature on tax neutrality and SCT mechanisms, informing innovation policy optimization.

We study how R&D spillovers propagate through buyer–supplier networks, exploiting a major reform of Japan's R&D tax credit system in 2003. The reform replaced the incremental credit for large firms with a volume-based scheme, reducing the marginal cost of R&D for firms with eligible expenditure below a ceiling but not for those above it or for SMEs. Using difference-in-differences, we find the reform increased R&D expenditure, innovative output and sales of treated firms. We find evidence of positive forward spillovers to downstream firms: the reform raised productivity of firms with a greater share of treated suppliers. Conversely, we find no evidence of backward spillovers to upstream firms. Treated firms also reallocated R&D from overseas affiliates to Japan while expanding affiliate employment, suggesting that headquarters' innovation enabled scaling up of foreign production operations.