Using a United States Patent and Trademark Office dataset with name-based gender attribution for inventors, we examine gender differences in US patent outcomes. On average, patent teams with at least one woman inventor not only have fewer citations (as documented in recent studies), but are also less likely to have related patents granted in Europe and Japan (triadic grants) and less likely to be renewed. Patent assignees account for a large share of the gender gap, but disparities remain even after controlling for technology, application year, patent examiner and assignee fixed effects. However, within organizations, we find no significant gender differences in outcomes for sole-inventor patents. Instead, we find that mixed-gender teams led by men have, on average, weaker outcomes than all-men teams. A small gap in citations and triadic grants remains even when we control for the identity of the first inventor. We discuss potential mechanisms and implications of these findings.
The European Union seeks to reduce wage-productivity and technological disparities between more developed and less developed European countries through a range of policy initiatives, including Horizon Europe. Despite these efforts, regional-level evidence on the relationship between wage pressure and business R&D expenditure remains limited and inconclusive. This study contributes to the literature by examining the relationship between wage pressure and business R&D expenditure across 214 NUTS-2 regions in 28 Member States. The empirical results indicate an inverted U-shaped (nonlinear) relationship between wage pressure and business R&D expenditure, particularly in laggard regions. In particular, moderate wage pressure stimulates business R&D investment, whereas excessive wage pressure has a negative effect in laggard regions. In contrast, in high-productivity regions, higher wage pressure is associated with increased business R&D expenditure. Further, public R&D investment undertaken by government and higher-education institutions complements the private-sector innovation, as reflected in higher levels of business R&D expenditure. To identify long-run effects, this study employs a system GMM estimator with lagged dependent variables alongside a dynamic common correlated effects estimator, using an unbalanced panel dataset spanning the period 2000-2016. From a policy perspective, moderate (high) wage pressure may serve as a useful mechanism to stimulate technological innovation in laggard (advanced) regions.
In the era of the digital economy, artificial intelligence (AI) is reshaping the landscape of corporate production and income distribution. Based on data from China A-share listed companies from 2010 to 2024, this study empirically examines how AI reshapes labor's share of income and its underlying mechanisms. The findings reveal that the application of AI has significantly increased the share of labor income in enterprises, with this effect being particularly pronounced in labor-intensive and high-tech firms. Mechanistic analysis indicates that AI primarily empowers labor factors and enhances their income share through three pathways: improving total factor productivity, boosting innovation efficiency, and optimizing employee skill structures. This study provides new empirical evidence for understanding the income distribution effects of AI at the micro-level of enterprises.
The research university is one of the longest-surviving institutions in the modern world, having repeatedly adapted to changing technological, political, and cultural environments. This essay examines the challenges and opportunities that the emerging twenty-first-century communications environment poses for research universities and the institutional responses those conditions may call for. It surveys the fragile equilibrium that universities must maintain among their multiple missions - formal teaching, open-science research, advanced training, independent expertise, and custodianship of scholarly knowledge - as they confront pressures to commercialize intellectual property, to compete for dwindling public resources, and to engage with industrial and governmental partners both domestic and global. Drawing a careful distinction between the logics of 'open science' and of proprietary R&D, the essay poses a series of open questions about the future role of universities as nodes in a global research network, the evolving nature of graduate and continuing professional education, the preservation of academic independence, and the stewardship of digital repositories and teaching materials.
This study investigates the relationship between Artificial Intelligence (AI) and innovation inputs in Spanish manufacturing firms. While AI is increasingly recognized as a driver of productivity and economic growth, its role in shaping firms’ innovation strategies remains underexplored. Using firm-level data, our analysis focuses on whether AI complements innovation inputs - specifically R&D and Embodied Technological Change (ETC) - and whether AI can be considered as a Method of Invention, able to trigger subsequent innovation investments. Results show a positive association between AI adoption and both internal R&D and ETC, in a static and a dynamic framework. Furthermore, empirical evidence also highlights heterogeneity, with important peculiarities affecting large vs small firms and high-tech vs low-tech companies. These findings suggest that AI may act as both a complement and a catalyst, depending on firm characteristics.
Why do technological revolutions fail to appear in productivity statistics? This paper develops a game-theoretic model in which firms with heterogeneous absorptive capacity make technology adoption decisions under competitive pressure. In equilibrium, high-capability firms adopt while low-capability firms either wait or adopt at a net loss due to competitive pressure. Aggregation across firm types generates a macro-level productivity paradox: substantial micro-level technological change produces negligible aggregate productivity gains. We provide micro-foundations for the Solow Paradox and derive three results: (i) a separating equilibrium in which only frontier firms adopt, generating the paradox as a composition effect; (ii) a pooling equilibrium in which competitive pressure forces unprofitable adoption; and (iii) multiple equilibria with coordination failure, where the Pareto-dominant equilibrium may not be selected. A welfare analysis shows that the social planner's optimal policy depends critically on whether non-adoption reflects coordination failure or genuine capability deficit. A formal decomposition shows that this composition channel is independent of, and additive to, the measurement channel identified in the directed technological change literature. We further show that workforce aging structurally expands the parameter space where the paradox persists, generating testable predictions for economies undergoing rapid demographic transitions.
Radical innovation (RI) is vital for overcoming technological barriers and enhancing core competitiveness. Public data openness (PDO) reshapes corporate innovation by improving information accessibility and fostering multi-agent collaboration. While prior research has examined its impact on general innovation, evidence on RI remains limited. Using Chinese government-led PDO policies as a quasi-natural experiment, we employ a staggered difference-in-differences model to examine the impact of PDO on RI. Our results reveal that PDO significantly promotes RI by facilitating heterogeneous knowledge acquisition and cross-boundary collaboration. Meanwhile, PDO has a stronger driving effect on RI for non-state-owned firms, technology-sensitive industries, and digitally advanced regions. Furthermore, data quality and utilization are crucial for realizing the promotional effects of PDO. These findings advance our understanding of PDO and RI, offering critical insights for the design of innovation policy.
The release of ChatGPT by OpenAI, a generative pre-trained transformer (GPT), marked a significant leap in the capabilities of large language models (LLMs). This study investigates whether the release triggered spillover effects that catalysed further innovation in artificial intelligence (AI). Using a Difference-in-Difference event study, we examine trends in GPT/LLM-related patent filings submitted to both the US Patent and Trademark Office (USPTO) and the European Patent Office (EPO). This is conducted over a two-year window surrounding the launch. To isolate these effects, we compare these filings against a control group of unrelated computing patents. Our results show a marked increase in related filings in the US, with a less pronounced effect observed in the EU. We attribute this divergence to institutional differences in patent eligibility criteria, procedural complexity and the regulatory environment. The findings support the view that a jurisdiction's approach to patent law influences its ability to formalise and capture innovation. We suggest that policy reforms aimed at improving procedural efficiency and reducing barriers to patentability may enhance the EU's AI development landscape.
This study explores whether user entrepreneurship - the commercialization of innovations triggered by user experience - influences the initial funding of start-ups. Drawing on signaling theory, we conceptualize user experience as an unobservable underlying quality - demand-side human capital - communicated to external capital providers through observable, costly actions. However, the credibility of this signal depends critically on the type of user experience. Using original survey data from Japan, we show that firms founded by professional user entrepreneurs are more likely to raise external capital than firms founded by non-user entrepreneurs, while end-user entrepreneurs do not differ significantly from non-user entrepreneurs in this regard. Specifically, such firms tend to raise initial funding through external equity financing rather than debt. We argue that professional user experience sends a high-commitment signal to investors, reducing information asymmetry more effectively than other types of user experience. The findings suggest that merely being innovative is insufficient unless backed by a credible signal of quality based on the entrepreneur's human capital endowment.
This study investigates how trade openness (TOPEN) influences firm s' productivity dynamics in the manufacturing sector by analyzing variations in total factor productivity (TFP), efficiency change (EFFCH) and technological change (TECHCH) across economies with different degrees of high and low TOPEN economies. Using firm s' data from 2011 to 2022, the study applies the Malmquist Productivity Index to capture intertemporal productivity changes and integrates macro factors through Ordinary Least Squares (OLS), Fixed Effects Models (FEM) and the Generalized Method of Moments (GMM) to ensure robustness and address potential endogeneity. The DEA results reveal substantial heterogeneity in productivity performance between high- and low-TOPEN economies. Firms operating in high TOPEN economies exhibit significantly stronger productivity growth, with TFP, TECHCH and EFFCH increasing by 17%, 18.83% and 96%, respectively, compared to firms in less TOPEN economies. These findings demonstrate a pronounced productivity premium associated with TOPEN and underscore its role in accelerating technological upgrading and efficiency improvements at firms. Further findings confirm that TOPEN exerts a positive and statistically significant effect on both TECHCH and EFFCH, with effects markedly greater in highly TOPEN economies.
We extend existing data sets for domestic and foreign private and public R&D stocks, as well as labour-augmenting technical change data based on CES production functions. We cover slightly more periods and many more countries, now 44, up from 17, using the perpetual inventory method. We consider panel unit root issues for a large sample of 41 countries and two smaller sub-samples, with 21 rich and 20 emerging economies. Autoregressive regressions show negative time trends in the growth rates of labour-augmenting technical change, domestic private and domestic and foreign public R&D stocks, indicating the growth slowdown in the data period; foreign private R&D has a positive time trend. Residual-based panel cointegration tests, with cross-sectional dependence, support only triples of variables for the set of 21 countries with long data series. Residuals stem from estimations using DOLS, FMOLS, and PMG-ARDL. Coefficients of variation , calculated across countries for each year, show no or temporary signs of (in)equality trends for growth rates of private and public R&D and productivity.
Environmental regulation and Public Procurement for Innovation are two important instruments for delivering effective public policies in the field of biodiversity. They both seek to influence the direction of innovation activity in the business sector towards socially desirable objectives. This study investigates the relationship between environmental regulation and Public Procurement for Innovation in the field of biodiversity, using a novel dataset of 3,831 tenders retrieved from the Tenders Electronic Daily platform and combined with country-level data from the OECD and the World Bank. Our results reveal a negative association between environmental regulation and the use of public procurement for innovation related to biodiversity. This suggests that environmental regulation tends to be used as a substitute for demand-driven innovation policies such as Public Procurement for Innovation.
The globalization of economic activities increasingly compels firms to protect their technologies beyond domestic borders. While the rapid surge in international patent filings has drawn considerable attention from academics and policymakers, the underlying mechanisms dictating the spatial direction of these cross-border patent flows remain heavily debated. Relying on a structural framework of international patenting, this study first provides a theoretical decomposition to isolate the core determinants of global patent applications. We then empirically investigate the effect of bilateral trade on international patent filings utilizing an extensive panel dataset of 135 & times; 186 country pairs from 1995 to 2024. Our robust empirical results demonstrate that cross-border trade structurally drives international patent filings. Furthermore, we uncover profound heterogeneous dynamics: while trade and patenting act as complements in North-South and South-South relationships, they manifest as strategic substitutes among advanced North-North country pairs. Ultimately, these findings demystify the 'market-seeking' motives behind global intellectual property protection and significantly advance our understanding of the spatial diffusion of technological innovation.
Industrial robots have transformed manufacturing production methods and labor structures. This study examines how robot integration affects employment in A-share listed manufacturing firms from 2014 to 2023, focusing on government intervention. Key findings reveal that automation reduces the overall workforce size but increases the proportion of high-skilled labor, suppressing low-skilled jobs. These employment outcomes vary by industry, region, and ownership. Specifically, labor-intensive sectors, eastern regions, and private enterprises experience the most significant impacts. Furthermore, automation indirectly boosts employment by increasing productivity and firm scale. Fiscal incentives and industrial support policies effectively mitigate automation's adverse employment effects, whereas tax exemptions show minimal impact. Ultimately, targeted legislative action can manage these shifts, balancing human-machine collaboration with improved employment quality.
Previous studies on the use of secondary data - or data reuse - found that researchers confront considerable (un)foreseen challenges related to finding, accessing, decoding, and recontextualizing secondary data. However, why some researchers persist in this endeavor while others do not is unclear. We develop a plausible theoretical explanatory model - the data-reuse mechanism of this phenomenon. This model captures the relational aspects of the different elements involved in the process, i.e. scientific horizon and reward system, researcher's cognitive frameworks and relational structures, the properties and relational structure of data, specific conditions, and time. We argue that it is the interaction among these elements that ultimately allows the researcher to decide about data reuse. Guided by this model, we analyze the data reuse process in ten molecular/computational biology and epidemiology case studies. The results of our analysis show how the intertwined synchronic and diachronic relations among these elements add considerable uncertainty to the reuse process. It is precisely this uncertainty combined with researchers' tenacity, scientific horizon, and flexibility in changing this horizon that enable data reuse. Our findings demonstrate also that while some data properties such as open access may facilitate data reuse, they do not fully determine whether data are reused.
This study examines how competition style (quantity versus price competition) influences environmental corporate social responsibility (ECSR) strategy and social welfare. Previous research on ECSR assumed that firms invest in reducing pollutant emissions, whereas we observe that firms sometimes invest not only in reducing emissions but also in the marginal generation of pollutants, a type of investment that has been overlooked. We employ three assumptions: (a) firms invest in reducing the marginal generation of pollutants, under which pollutant emissions can decrease even if they increase the production of a good; (b) firms care only about their pollutant emissions and invest in reducing marginal pollutant generation from their production activities; and a quadratic environmental damage function. Findings show that, under quantity competition, firms do not choose a positive degree of ECSR, whereas under price competition, they set a positive degree of ECSR. However, when comparing total environmental damage under quantity and price competition scenarios, the latter is found to cause more damage. Assuming low marginal investment cost, social welfare is higher under price competition than under quantity competition only when product differentiation is moderate; otherwise, social welfare is greater under quantity competition.
Globalization has profoundly shaped Latin American economic policies, with Foreign Direct Investment (FDI) playing a central role in technology transfer strategies. This paper examines whether FDI generates innovation spillovers among Chilean firms, using firm-level data from the National Innovation Survey (2017-2023) combined with sectoral FDI flows from the Central Bank of Chile. We apply the Crepon-Duguet-Mairesse (CDM) model to correct for selection bias and distinguish between horizontal (industry-level) and vertical (supplier and client linkages) spillover channels. The analysis also examines heterogeneous effects across economic sectors and by firms' absorptive capacity. We find no evidence of innovation spillovers in primary and service sectors. However, significant positive spillovers emerge in manufacturing, where higher FDI presence increases the probability of local firms implementing new products or processes. These findings suggest that spillovers are sector-specific and depend on the technological intensity and linkage potential of FDI activities. The results have important implications for designing targeted FDI policies in developing economies.
Existing research has not sufficiently explored the mechanisms underlying how Artificial Intelligence (AI) investment affects corporate markup rates. Against the backdrop of enterprises pursuing high-quality development, the causal pathways by which AI enhances markup rates require urgent clarification. This study empirically examines the impact of AI investment levels on firm markup rates using panel data from Chinese A-share listed companies from 2010 to 2023. Benchmark regression results reveal that AI investment significantly and positively promotes firm markup rates. Mediating effect analysis indicates that AI investment enhances markup rates through three pathways: inventory management efficiency, capital utilization efficiency, and marketing efficiency. Heterogeneity tests reveal that this effect is stronger among enterprises in central and western China, non-competitive industries, and labor-intensive enterprises. Further analysis indicates that the degree of managerial myopia and the intensity of government subsidies negatively moderate this relationship, suggesting that internal governance and external policies may distort the returns on AI investment. This study provides rigorous empirical insights to strengthen economic resilience and sustainable development in emerging economies.
