The study, conducted by Olimjon Gaybullaev and Henry Obaga Were of the University of Pécs in Hungary, is notable for how it dissects the timing of productivity effects. Using an augmented autoregressive distributed lag model combined with frequency-domain Granger causality analysis, the authors were able to separate the average long-run relationships between innovation, energy and productivity from the short-term dynamics that conventional methods tend to blur. The result is a nuanced picture in which green patents and R&D temporarily depress measured productivity before delivering substantial gains, while renewable energy works in precisely the opposite pattern.

The choice of Sweden as a test case is deliberate. The country ranks second in the 2024 Global Innovation Index, invests more than 3.6 percent of its GDP in research and development, and pursues a national target of net-zero carbon emissions by 2045. Its renewable energy share of final consumption stands at roughly 66 percent, among the highest of any advanced economy. Programmes such as Produktion2030 have woven Industry 4.0 technologies, circular economy principles and environmental innovation into the industrial fabric, making Sweden an unusually clean laboratory for asking whether green transformation and economic competitiveness can genuinely coexist.

The econometric strategy was tailored to the data. Because annual observations over 30 years produce a small sample, the researchers relied on the augmented ARDL bounds testing framework, which performs reliably with as few as 30 to 40 observations and accommodates variables with mixed integration orders. Unit root testing with the augmented Dickey-Fuller and Zivot-Andrews procedures, the latter allowing for structural breaks, revealed that productivity and green patent intensity behave as trending, cumulative processes, while R&D intensity and renewable energy shares show mean-reverting tendencies shaped by policy interventions. A significant structural break around 2008, corresponding to the global financial crisis, was captured with a dummy variable to prevent biased estimates.

The long-run results are striking. A 1 percent increase in green technology innovation, measured by the share of environment-related patents among all domestic inventions, is associated with an approximately 0.37 percent rise in total factor productivity. R&D expenditure carries an even larger elasticity of roughly 0.96 percent, underscoring how Sweden’s sustained research investment builds the absorptive capacity that allows new knowledge to be commercialised and diffused. Renewable energy consumption, by contrast, showed a negative and statistically insignificant long-run coefficient of about minus 0.24, a result the authors attribute to the fact that Sweden’s green transition is already mature, so the easy efficiency gains have largely been banked.

The short-run story inverts that picture. In the immediate term, both green innovation and R&D spending significantly reduce measured productivity, with coefficients of minus 0.278 and minus 1.844 respectively. The explanation lies in adjustment costs: retooling production lines, training workers, waiting out patenting delays and absorbing organisational restructuring all impose sunk costs before new technologies spread through the economy. In a mature, fiercely competitive market like Sweden, firms may also face saturation effects that slow diffusion further. Renewable energy adoption, however, delivered a positive short-run coefficient of 0.712, reflecting immediate efficiency gains from lower energy costs, cleaner production processes and reduced operational disruption.

The frequency-domain causality analysis adds a crucial layer that average elasticities cannot capture. Green technology innovation and total factor productivity Granger-cause each other across short, medium and long frequency bands, indicating a mutually reinforcing loop in which innovation drives productivity and productivity feeds back into further innovation. R&D shows a similarly bidirectional, long-run-dominated relationship. Renewable energy displays unidirectional causality flowing to productivity at long- and medium-term frequencies, even though the average long-run ARDL coefficient is insignificant. In other words, the productivity contribution of renewable energy is real but gradual, materialising through structural and technological adjustments in the energy system rather than through immediate output gains.

These patterns map closely onto two foundational economic theories. The Porter hypothesis holds that well-designed environmental regulation stimulates innovation that ultimately offsets compliance costs and enhances competitiveness, and the Swedish evidence supports both its weak and strong versions: regulation-driven green innovation exists, and it produces net productivity gains over time. Endogenous growth theory, which holds that purposeful investment in knowledge and technology generates spillovers and increasing returns, explains the delayed returns to R&D. The short-run productivity dip observed for innovation is exactly what these frameworks predict, as resources flow into knowledge creation before the returns arrive.

The policy implications are considerable. The authors argue that renewable energy policies should not be judged solely on environmental outcomes or short-run productivity metrics, since their economic value unfolds over extended horizons through infrastructure adaptation and capital-intensive adjustment. Investments in smart grids and energy storage could accelerate the integration of renewables with existing production systems. More broadly, the study makes the case for coordinated frameworks that jointly promote environmental technologies, R&D investment and renewable energy deployment, rather than treating each lever in isolation. Sweden could also deepen its position within the Nordic innovation ecosystem, where recent evidence shows that climate technologies and cross-border spillovers deliver the strongest green productivity gains precisely in Sweden.

The study’s authors acknowledge limitations. The single-country focus limits generalisability to economies at earlier stages of the green transition, and the 30-year sample may lack the statistical power to detect some long-run effects, which could partly explain the insignificant renewable energy coefficient despite the persistent causal signal in the frequency domain. The linear ARDL framework also cannot capture asymmetric or nonlinear dynamics, and aggregate national data mask sectoral heterogeneity. Future work using nonlinear models, alternative cointegration estimators or firm-level data could sharpen the picture. What the current evidence establishes, however, is that in one of the world’s most innovative economies, going green has not come at the expense of competitiveness. It has, on the evidence of three decades of Swedish data, helped pay for it.

Subject of Research: The effects of green technology innovation, R&D expenditure, and renewable energy consumption on Sweden's total factor productivity and economic competitiveness from 1993 to 2022.

Article Title: The role of green technology innovation, R&D, and renewable energy in Sweden’s economic competitiveness

Article References: Gaybullaev, O., & Were, H. O. (2026). The role of green technology innovation, R&D, and renewable energy in Sweden’s economic competitiveness. Clean Technologies and Environmental Policy, 28(10), Article 258. https://doi.org/10.1007/s10098-026-03605-6

Image Credits: AI Generated

DOI: 10.1007/s10098-026-03605-6

Keywords: green technology innovation, total factor productivity, research and development, renewable energy, Sweden, economic competitiveness, A-ARDL model, frequency-domain Granger causality, Porter hypothesis, endogenous growth theory, renewable energy transition, environmental patents