|
on Technology and Industrial Dynamics |
| By: | Dany Bahar (Center for International Development at Harvard University); Shreyas Gadgin Matha; Ricardo Hausmann (Harvard's Growth Lab); Santiago Segovia (Harvard's Growth Lab) |
| Abstract: | Japan remains one of the world’s most technologically sophisticated economies, yet its labor productivity has been stagnant for more than two decades. This paper investigates the apparent contradiction between Japan’s high R&D intensity and its weak productivity performance by examining the allocation, composition, and effectiveness of innovation across industries. Using industry-level data from the OECD, patent-level data linked across technology and industry classifications, and a set of nine technological taxonomies, we document that Japan disproportionately concentrates R&D in mid-technology manufacturing sectors—such as motor vehicles, electrical equipment, and chemicals—that generate relatively low productivity spillovers. High-technology sectors, including ICT, pharmaceuticals, scientific R&D, and advanced digital services, receive a significantly smaller share of investment and exhibit much higher productivity contributions in other countries. We further show that Japan’s indirect, tax-based system of R&D support reinforces this equilibrium by favoring large incumbents and under-supporting SMEs. We conclude by assessing the potential of Japan’s new 17-sector strategy to reorient the innovation system toward frontier technologies. |
| Keywords: | Japan, innovation, industrial policy |
| Date: | 2026–05 |
| URL: | https://d.repec.org/n?u=RePEc:glh:wpfacu:269 |
| By: | Myungkoo Song (Korea Institute for Industrial Economics and Trade) |
| Abstract: | This study assesses the evolving innovation climate of South Korea’s manufacturing industry by examining the relationship between perceived innovation barriers and innovation activities. Using firm-level microdata from the Korean Innovation Survey covering the period from 2020 to 2024, the analysis reveals a post-pandemic paradox: a heightened perception of innovation barriers has coexisted with a significant increase in overall innovation activity.<p> However, this aggregate trend masks a critical divergence based on technological intensity. For firms in the low-technology sector, the results support a “revealed barrier” effect, in which heightened barrier perception stems from persistent engagement in innovation.<p> Conversely, for firms in the high-technology sector, barriers act as genuine “Deterrents, ” with financial and market constraints significantly reducing innovation success rates. These findings suggest that a one-size-fits-all policy approach is insufficient. The study concludes that innovation policy must be bifurcated: focusing on de-risking financial and market uncertainties for high-technology sectors, while addressing operational and capability frictions for low-technology sectors. |
| Keywords: | innovation; innovation activities; barriers to innovation; innovation policy; technology; technology adoption; technology policy; research and development; R&D |
| JEL: | L60 O31 O32 O38 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:ris:kieter:022557 |
| By: | Andrea Bacchiocchi (Department of Economics, Society and Politics, University of Urbino Carlo Bo); Germana Giombini (Department of Economics, Society and Politics, University of Urbino Carlo Bo); Ludovica Segneri (Department of Economics, Society and Politics, University of Urbino Carlo Bo); Francesco Venturini (Department of Economics, Society and Politics, University of Urbino Carlo Bo) |
| Abstract: | Does financial risk affect the firm decision to develop a new technology? We study this issue in the context of the take-off of Artificial Intelligence (AI). Using data on 28, 000 Italian firms (2012–2019) matched with patent records, we find that companies handling higher cash-flow volatility are significantly more likely to innovate in AI. The role of financial risk is weaker for relatively more mature technologies, suggesting that firms more subject to financial uncertainty are more willing to undertake innovation in high-uncertainty, high-reward domains and drive frontier technological change. |
| Keywords: | Artificial Intelligence, Financial risk, Cash-flow volatility, Technological uncertainty. |
| JEL: | O31 G32 L25 C23 |
| Date: | 2026–05 |
| URL: | https://d.repec.org/n?u=RePEc:anc:wmofir:199 |
| By: | Pierre-Alexandre BALLAND; Valentina DI GIROLAMO; Florence BENOIT; Julien RAVET; Alexandr HOBZA |
| Abstract: | Fragmentation in the European Union's R&I system is increasingly acknowledged as a major hindrance to its performance. However, theoretical frameworks and empirical evidence remain scarce. This paper introduces a novel complexity-based approach to analyse the competitiveness costs of R&I fragmentation, focusing specifically on hub connectivity as a key metric. Using a comprehensive dataset combining patent records (OECD REGPAT) and scientific publications (OpenAlex) from 2000 to 2023, we examine R&I networks across multiple spatial levels and technological domains. Our analysis identifies three critical findings. First, the European R&I system is much more fragmented compared to the US, with major European hubs showing notably weaker interconnectivity than their US counterparts. Second, we demonstrate that hub connectivity becomes particularly crucial for complex technologies and scientific fields. Third, we find that there is an efficiency gap between the US and Europe in all domains, but it is most pronounced in complex ones, resulting in a substantial competitive disadvantage in strategic sectors. These findings have significant implications for European innovation policy and suggest the urgent need for targeted interventions to enhance cross- regional R&I collaboration in complex technological and scientific domains. |
| Keywords: | Research Fragmentation, Economic Complexity, Inter-Regional Connectivity, Innovation Networks, EU Competitiveness |
| Date: | 2026–05 |
| URL: | https://d.repec.org/n?u=RePEc:egu:wpaper:2608 |
| By: | Enghin Atalay; Ali Hortacsu; Nicole Kimmel; Chad Syverson |
| Abstract: | How do investments in innovation translate into future productivity growth? Empirically answering this question is challenging. R&D spending is an observed input into the innovation process, but mapping it to productivity growth requires assumptions about the depreciation of R&D capital, gestation lags, and how well such expenditures capture true innovative effort (Hall, 2007). Patents, an alternative measure, capture successful innovations but vary widely in novelty (Kelly et al., 2021) and economic value (Kogan et al., 2017). Firms may forgo patenting to preserve secrecy, while others patent strategically to protect existing products even when their underlying innovations are marginal. |
| Keywords: | productivity measurement; innovation |
| JEL: | D24 E31 O31 O47 |
| Date: | 2026–04–20 |
| URL: | https://d.repec.org/n?u=RePEc:fip:fedpwp:103326 |
| By: | Minsung Kang (Korea Institute for Industrial Economics and Trade) |
| Abstract: | This paper analyzes R&D investment and Digital Transformation (DX) technology utilization at South Korean manufacturing and service enterprises during the period from 2017 to 2023 using data extracted from the Survey of Business Activities by the Ministry of Data and Statistics.<p> Overall, R&D activity levels are higher, more consistent, and more intensive in the manufacturing sector than in the service sector, but even as service firms are less likely to engage in R&D, they have pursued digital transformation (DX) at a higher rate, albeit with greater variability and significant size-based polarization.<p> DX contributes more to competitiveness in services than in manufacturing, and small firms lag significantly in both sectors.<p> Based on the results of the analysis, the paper recommends expanding customized integrated support packages, building a full-cycle DX support system with strong human capital development, and shifting the regulatory sandbox toward a more flexible, ex post regulation framework to accelerate adoption and reduce disparities. |
| Keywords: | innovation; innovation activities; research and development; R&D; corporate R&D; R&D investment; digital transformation; DX |
| JEL: | O31 O32 O33 O38 L60 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:ris:kieter:022561 |
| By: | Leanne Cass; Federico Fabio Frattini; Misato Sato; Aurelien Saussay; Francesco Vona |
| Abstract: | There is growing evidence that green jobs have higher skill requirements, but whether they offer sufficient wage incentives to encourage workers to acquire those skills remains unclear. We study the green wage premium and its drivers to isolate the average return to green tasksusing online job vacancy (OJV) data for EU countries over the period 2018-2023. We develop a transparent LLM-based approach to classify job vacancies as green when they list at least one green task. Green jobs pay a premium of 5.5% relative to comparable postings within the same occupation, and this estimate is little changed when controlling for nonmonetary job attributes making these jobs more attractive. Roughly half of this premium is explained by firm fixed effects, consistent with an important role for firm rents. An Oaxaca-Blinder decomposition shows that the higher skill complexity explains a further one tenth of the premium, leaving a residual return to green tasks of around 2%. The green wage premium is higher outside the manufacturing sector, and for low-carbon roles. |
| Keywords: | green wage premium, skill gaps, green tasks, LLM |
| Date: | 2026–05–27 |
| URL: | https://d.repec.org/n?u=RePEc:cep:cepdps:dp2185 |
| By: | Erik Hellberg (Uppsala History of Inequality and Labor Lab (UHILL) and Department of Economic History, Uppsala University); Jakob Molinder (Uppsala History of Inequality and Labor Lab (UHILL), Department of Economic History, Uppsala University, and Department of Economic History, Lund University) |
| Abstract: | Ever since Smith and Marx there has been a debate about the impact of industrialization on the content of work. While shifts in the occupational structure associated with industrialization are well charted, we lack systematic evidence on the tasks workers actually performed during this period of economic change. We assemble a large corpus of job advertisements from Swedish newspapers from the 1860s to the 1900s and use a large language model (LLM) to extract task statements. We then categorize tasks along two dimensions: manual vs. cognitive and routine vs. non-routine. We first document a large decline in non-routine manual work, which was replaced in fairly equal proportions by analytical and routine cognitive and manual tasks, so that work overall did become more routine, in line with Marx's pessimistic prediction. The shift had more to do, however, with the decline of low-paid service activities like cooking and cleaning than with the disappearance of artisanal methods of production. At the same time, the economy also became more cognitive, relying more on human capital than physical labor. Routine jobs tended to be in the middle to upper half of the pay scale, meaning that structural change led to a "hollowing in" of the occupational structure and to inclusive job growth. We also find pronounced gender segregation: even among unskilled farm workers, men's and women's task sets were largely distinct. We also show that many occupations spanned multiple economic sectors, with implications for how we measure and interpret structural change. |
| Keywords: | labor markets, tasks, industrialization, Sweden, economic history |
| JEL: | N33 N34 J24 C81 |
| Date: | 2026–05 |
| URL: | https://d.repec.org/n?u=RePEc:hes:wpaper:0301 |
| By: | Björn Thor Arnarson; Magnus Tolum Buus; Andreas Moxnes; Jakob Roland Munch; Chong Xiang |
| Abstract: | We study how firm growth reorganizes the division of labor across firms in global value chains. Using a novel dataset linking cross-border firm-to-firm transactions to matched employer–employee data, we show that demand shocks increase trade between firms while reducing occupational similarity, implying greater specialization. We develop and estimate a model of task outsourcing in which firms expand by reallocating tasks to suppliers. The model matches the data and implies endogenous scale economies. Eliminating outsourcing reduces average labor productivity by 25 percent and increases input costs by 10 percent, highlighting the central role of specialization in shaping firm performance. |
| Keywords: | supply chains, global value chains, outsourcing, scale economies, labor specialization, labor productivity, production networks |
| JEL: | F10 F12 F16 D24 L11 L25 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:ces:ceswps:_12710 |
| By: | Fernandez Macias Enrique (European Commission - JRC) |
| Abstract: | This paper presents a conceptual framework for analysing the impact of digital technologies on work and employment, developed on the basis of nearly a decade of research by the Employment team of the Joint Research Centre of the European Commission. The framework combines three elements. First, a historical situation of the digital revolution within the Schumpeterian theory of long-wave technological change, which also helps locate recent advances in AI within a longer cycle of innovation and socio-economic transformation. Second, a discussion of three structural peculiarities of the digital economy (the hyperabundance of usable information, the centralised flexibility facilitated by algorithmic control, and the platformisation of economic activity), each of which is being intensified and extended by current developments in AI. Third, an analysis of the three main vectors (digitisation, automation and platformisation) through which the digital revolution is transforming work and employment. Drawing on our empirical findings, we argue that the main impact of the digital revolution has been on the modes of coordination and control of work rather than on employment levels: digitisation has driven a significant intensification of work and erosion of worker autonomy; automation has had small and generally positive employment effects; and platformisation is extending from digital labour platforms to regular work environments, a tendency that the rapid integration of AI is likely to accelerate. The central policy challenge, we argue, is institutional: digital technologies have transformed work without a commensurate transformation of the institutional framework that should govern them. |
| Date: | 2026–05 |
| URL: | https://d.repec.org/n?u=RePEc:ipt:laedte:202604 |
| By: | Xian Jiang; Hannah Rubinton |
| Abstract: | Custom software is distinct from other types of capital in that it is non-rival—once a firm makes an investment in custom software, it can be used simultaneously across its many establishments. Using confidential U.S. Census data, we document that while firms with more establishments are more likely to invest in custom software, they spend less on it as a share of total capital expenditure. We explain these empirical patterns by developing a model that incorporates the non-rivalry of custom software. In the model, firms choose whether to adopt custom software, the intensity of their investment, and their scope, balancing the cost of managing multiple establishments with the increasing returns to scope from the nonrivalrous custom software investment. Using the calibrated model, we assess the extent to which the decline in the rental rate of custom software over the past 40 years can account for a number of macroeconomic trends, including increases in firm scope and concentration. |
| Keywords: | Technology adoption, Non-rivalry, Concentration, Firm scope |
| JEL: | D24 E22 O33 |
| Date: | 2026–04 |
| URL: | https://d.repec.org/n?u=RePEc:cen:wpaper:26-28 |
| By: | Christos Makridis; Christos A. Makridis |
| Abstract: | AI is transforming work, but workers’ responses to these technologies depend not only on exposure to AI, but also on how organizations, especially managers, oversee the transition. Using longitudinal data from the Gallup Workforce Panel from 2023-2026, I examine whether managers and workplace practices shape employees’ fears that AI will eliminate their jobs. Across survey waves, roughly 3-4 percent of workers say their job is very likely to be eliminated within five years because of new technology, automation, robots, or artificial intelligence, while about 14-19 percent say it is somewhat or very likely. Concern is substantially higher among frequent AI users. Stronger workplace practices are associated with lower displacement fear: a one-standard-deviation increase in workplace quality is associated with 13-24 percent lower odds of reporting greater displacement risk, and workers reporting the highest level of organizational wellbeing support are 6-6.8 percentage points less likely to say their job is somewhat or very likely to be displaced in cross-sectional specifications. Frequent AI use is positively associated with perceived displacement risk, with estimates ranging from about 3-12.9 percentage points across the main specifications and reaching 6 percentage points in the most saturated respect model. However, this association is weaker in higher-quality workplace environments: among workers reporting the highest level of organizational wellbeing support, the frequent-AI-use premium is reduced by up to 9.0 percentage points. In short, managers play a central role in shaping how workers interpret AI adoption. |
| Keywords: | artificial intelligence, job displacement, managers, workplace practices, worker expectations, technological change, organizational capital |
| JEL: | D23 J24 J28 M12 O33 |
| Date: | 2026 |
| URL: | https://d.repec.org/n?u=RePEc:ces:ceswps:_12678 |
| By: | Mattia Leoni (Department of Economics, Roma Tre University); Stefano Di Bucchianico (Department of Economics and Statistics, University of Salerno and CELPE) |
| Abstract: | This paper investigates the distributive effects of climate change mitigation policies across a panel of 29 OECD countries over the period 1990-2020.We estimate the dynamic impact of a one-standard-deviation increase in the level of the OECD Environmental Policy Stringency (EPS) Index on income inequality indicators from the World Income Inequality Database. To address potential endogeneity in policy adoption, we employ an Instrumental Variables Local Projections (IV-LPs) framework. The paper contributes to the literature on the macroeconomic and distributive consequences of climate action in two main ways. First, building on evidence regarding the role of institutions in enabling effective environmental action, we propose a novel instrument based on global climate-related speeches by central banks, combined with country-specific institutional characteristics. Second, the results show that the distributive effects of environmental policy are highly heterogeneous across instrument types. Market-based climate policies generate regressive, although short-lived, effects. By contrast, Technology-support policies, namely public research and development (R&D) expenditure and renewable energy support, generate progressive redistributive effects and expansionary macroeconomic effects on real GDP and investment, although at the cost of upward pressure on inflation. On the labour-market side, the transmission channels operate through long-term improvements in employment and women’s labour-market participation. Uncovering these heterogeneous effects is essential for designing policies that can secure a “just” socio-economic transition. |
| Keywords: | Climate Change; Public R&D Policy; Income Inequality; Just Transition; IV Local Projections |
| JEL: | C33 D63 E25 Q58 |
| Date: | 2026–05–26 |
| URL: | https://d.repec.org/n?u=RePEc:sal:celpdp:022586 |
| By: | Marie Sciaccitano (Université Côte d'Azur, CNRS, GREDEG, France); Lionel Nesta (Université Côte d'Azur, CNRS, GREDEG, France; OFCE, Sciences Po Paris, France; SKEMA Business School, France) |
| Abstract: | We examine how income growth reshapes the composition of household consumption between green and non-green goods across countries. Using harmonized data on Environmental Goods and Services (EGS) for 138 countries over 1995–2015, we measure green consumption as expenditure on goods and services explicitly designed to prevent or reduce environmental degradation and estimate a non-linear environmental Engel curve. Green consumption behaves as a luxury at low income levels but progressively transitions toward a necessity as income rises. We further show that income elasticities of green consumption are systematically higher than those of non-green consumption, revealing income-driven composition effects in the consumption basket. These non-homothetic demand patterns imply that income growth systematically shifts household expenditure toward or away from goods intended to reduce environmental pressures, with important policy implications. Applying our estimates to a stylized Climate Fund redistribution, we show that accounting for heterogeneous income elasticities changes its predicted outcome. Relative to the homothetic benchmark with unit elasticities, the response of global green consumption remains limited, whereas the increase in global non-green consumption is substantially larger. |
| Keywords: | Measurement of green consumption, income elasticities, environmental Engel curve, luxury & necessity goods and services, climate policy |
| JEL: | E01 E21 Q5 |
| Date: | 2026–05 |
| URL: | https://d.repec.org/n?u=RePEc:gre:wpaper:2026-15 |