Growth turns green after a threshold in OECD economies, 33-year panel study finds

After more than three decades of data spanning 35 OECD countries, a new panel econometric study reports that economic growth eventually begins to ease environmental pressure in the world’s advanced economies, but only once income passes a turning point, and only if that growth is paired with a decisive shift toward renewable energy. The study, published in Humanities and Social Sciences Communications by Ferhat Özbay, Mehmet Ali Demir and Bilgehan Tekin, analyzed annual data from 1990 to 2022 and tested three influential frameworks in environmental economics: the Environmental Kuznets Curve, the Load Capacity Curve, and the Pollution Haven Hypothesis.

The Environmental Kuznets Curve hypothesis, first proposed by Grossman and Krueger in 1991, holds that environmental degradation rises with income in the early stages of development but falls once a country grows rich enough. The Load Capacity Curve applies a similar logic to the load capacity factor, defined as biocapacity divided by ecological footprint, which measures whether a nation’s ecosystems can supply more than they are depleted. In this framework, the load capacity factor first declines as economies industrialize and then recovers, tracing a U-shaped path. The Pollution Haven Hypothesis, by contrast, predicts that stringent environmental rules in rich countries push polluting industries to relocate to economies with weaker regulation.

OECD countries form a critical test case for these ideas. The study notes that the 35 nations examined account for more than 60 percent of global GDP, cover roughly 20 percent of the world’s land area in population terms, and yet still draw approximately 85 percent of their energy consumption from fossil fuels, with only about 15 percent from renewables. The researchers calculate that these countries alone are responsible for nearly one-third of the global ecosystem’s pollution stemming from fossil fuel energy use, which makes their trajectory decisive for international sustainability goals and for the net-zero target of the Paris Agreement.

Methodologically, the study is unusually careful about the statistical quirks of interconnected economies. The authors first verified that the variables follow a normal distribution, then confirmed strong cross-sectional dependence using the Pesaran CD test, meaning shocks and outcomes in one OECD country are correlated with those in its neighbors through trade, capital flows and shared global factors. Slope homogeneity tests rejected the idea that every country behaves identically. Unit root testing with the second-generation CIPS approach showed that carbon dioxide emissions and GDP are stationary only in first differences, prompting the use of Pedroni and Westerlund cointegration tests, which identified long-run relationships among the variables, strongly for the load capacity model and partially for the emissions model.

For estimation, the researchers relied on two estimators robust to these dependencies: Panel-Corrected Standard Errors, which allow for heteroskedasticity and contemporaneous correlation across panels, and Driscoll-Kraay standard errors, which correct for general forms of spatial and temporal dependence. The core result is unambiguous. In the base models, a 1 percent increase in GDP reduces the load capacity factor by roughly 0.45 to 0.86 percent and raises carbon emissions by about 0.52 to 1.72 percent, but once the income turning point is passed, a further 1 percent of income growth increases carrying capacity by 0.02 to 0.04 percent and trims emissions by 0.02 to 0.09 percent. This confirms both the inverted U-shaped EKC and the U-shaped LCC for the OECD panel.

Renewable energy consumption emerges as the strongest environmental lever in the analysis. A 1 percent increase in renewable energy consumption raises the load capacity factor by 0.06 to 0.068 percent and reduces carbon emissions by 0.088 to 0.113 percent. The authors argue that this effect can pull the entire curve forward, accelerating the structural transformation in which energy-intensive technologies are replaced by more efficient ones. These findings align with prior studies reporting that solar, wind and geothermal expansion improves environmental quality, though the broader literature remains divided, with some country-level analyses of Brazil, Mexico and Malaysia finding weak or even negative effects where renewable capacity is insufficient or inefficient.

Globalization tells a darker story in this dataset. Across both the emissions and load capacity models, a rise in the KOF globalization index reduces ecological carrying capacity by approximately 0.47 to 0.51 percent while accelerating emissions growth. The authors attribute this to increased consumption of natural resources, transport-related emissions and polluting industries, aggravated by weak environmental taxation and concessions granted to attract investment. This contrasts with earlier studies in which social or financial globalization improved environmental quality in South Africa, Brazil and Mexico, underscoring that the globalization-environment relationship is far from settled and appears to depend heavily on governance quality and accompanying policy.

Population density produced one of the study’s most striking asymmetries. Denser populations significantly weaken the load capacity factor, with a 1 percent increase reducing carrying capacity by roughly 0.47 to 0.55 percent as ecological footprints expand and biocapacity is overused. Yet population density showed no statistically significant effect on carbon dioxide emissions. The authors interpret this to mean that consumption habits and infrastructure, not the sheer number of people per square kilometer, are the dominant drivers of emissions in wealthy economies, a finding that shifts the policy focus away from demographic control and toward sustainable cities, resource efficiency and green technology.

The Pollution Haven Hypothesis received a split verdict. Foreign direct investment showed no significant effect on the load capacity factor, leading the authors to reject the hypothesis within the LCC framework. Under the EKC framework, however, FDI became statistically significant for emissions under the Driscoll-Kraay estimator, a result the authors describe as weak confirmation of the pollution haven pattern in the emissions dimension. Financial development fared worse still in nominal terms, with highly significant coefficients suggesting harm to environmental quality under PCSE, but those coefficients lost significance entirely under Driscoll-Kraay correction. The authors conclude that the apparent causal link between financial development and environmental outcomes is largely fueled by cross-sectional forces such as global macroeconomic shocks, international capital movements and regional trade spillovers rather than a robust direct effect.

The study’s authors caution that their results are bounded in scope. Generalizing to low- and middle-income countries, where environmental pressures and institutional contexts differ sharply, is limited, and proxy indicators such as the load capacity factor capture only part of concepts like natural carrying capacity and environmental stress. Causal claims remain difficult given potential omitted variables and reverse causality. Still, their policy implications are concrete: prioritize renewable energy adoption and grid modernization, phase out fossil fuel subsidies with predictable planning, embed environmental clauses and supply chain transparency in trade agreements, and channel financial flows toward green technologies. Emissions alone, the study argues, cannot steer the global economy toward sustainability, because a resilient future requires managing the full carrying capacity of natural systems alongside the atmosphere.

Subject of Research: Testing the Environmental Kuznets Curve, Load Capacity Curve and Pollution Haven Hypotheses in OECD countries from 1990 to 2022

Article Title: Environmental stress and well-being in OECD countries: testing EKC, LCC, and PHH models

Article References: Özbay, F., Demir, M. A., & Tekin, B. (2026). Environmental stress and well-being in OECD countries: testing EKC, LCC, and PHH models. Humanities and Social Sciences Communications, 13(1), Article 1871. https://doi.org/10.1057/s41599-026-09146-z

Image Credits: AI Generated

DOI: 10.1057/s41599-026-09146-z

Keywords: OECD countries, Environmental Kuznets Curve, Load Capacity Curve, Pollution Haven Hypothesis, renewable energy, carbon dioxide emissions, globalization, financial development, foreign direct investment, panel econometrics, biocapacity, ecological footprint