For decades, economists have puzzled over what actually drives long-run economic growth. Once the contributions of capital (machines, buildings, infrastructure) and labour (hours worked) are accounted for, a large share of growth remains unexplained. Economists call this leftover portion “total factor productivity” and usually treat it as a stand-in for technological progress and innovation. Because it captures everything that isn’t explicitly measured, total factor productivity has been called “a measure of our ignorance.” This matters well beyond economics seminars: assumptions about how total factor productivity evolves are built into the long-run growth and climate models that inform EU energy and climate policy, including the scenarios used in major international assessments, such as the IPCC’s global climate and emissions scenarios. If those models treat future productivity growth as an unconstrained, ever-rising trend, they may be quietly assuming away physical limits that recent post-growth research argues need to be brought back into view.
In a recent study, published in the journal Energies, we test a more physically grounded explanation for productivity growth, rooted in the concept of exergy. Exergy measures the portion of energy that can do useful work: a barrel of oil, for instance, can power far more work than the same quantity of energy spread out as low-temperature waste heat. Economies convert energy in stages: final energy is what’s delivered to end users, such as the electricity reaching a factory or the fuel pumped into a car; useful energy is what remains after end-use devices (e.g., engines, boilers, light bulbs) convert that final energy into motion, heat, or light, with some inevitably lost to inefficiency along the way. The efficiency of this last conversion step, known as final-to-useful exergy efficiency, captures how well an economy’s machines and appliances turn the energy delivered to them into work that is genuinely useful.
Earlier research on Portugal (1960-2014) found a statistically robust long-run relationship between this efficiency and total factor productivity, with no need to assume any separate, unexplained trend driving productivity forward. We asked whether the same holds true in Austria, a mature, early-industrialised economy with a very different growth trajectory, using data from 1960 to 2019. To make sure any relationship we found was real and not a statistical accident, we tested 100 different model specifications and applied a deliberately strict statistical correction (a Bonferroni correction) that guards against false positives.
What we found
The relationship holds for Austria too: energy-conversion efficiency and total factor productivity move together over the long run, again without needing an additional unexplained trend. However, the strength of this relationship is weaker than in Portugal, which is consistent with, though not proof of, the idea that efficiency gains matter less for productivity as an economy matures. We also found that the relationship is not constant through time: it was significantly stronger before the mid-1970s oil crisis than after, although the long-run link between the two variables holds up even once this shift is considered. Finally, tests of the direction of causality show that productivity and efficiency are tied together in the long run, with productivity doing more of the adjusting between the two, but without clear evidence that one straightforwardly causes the other.
Why it matters for growth and climate models
These results add Austria to a still-short list of countries where this relationship has been rigorously tested, and strengthen the case that productivity growth is, at least in part, a physical phenomenon rooted in how efficiently economies use energy, rather than a purely abstract, unconstrained trend. This has two practical implications for policy-relevant modelling. First, because energy-efficiency gains are bounded by the laws of thermodynamics, continued reliance on efficiency-driven productivity growth implies there are physical limits to long-run growth if productivity depends materially on further such gains. Second, efficiency-driven productivity gains can trigger economy-wide rebound effects, where lower costs stimulate demand and partly offset the energy savings achieved. Integrated assessment models and other long-run growth and climate models commonly treat total factor productivity as exogenous and unconstrained; incorporating energy and exergy constraints into these models may improve the realism of long-term projections of growth, energy use, and emissions, relevant to the scenario work that underpins EU climate and energy policy.
Access the full article in Energies:
The full article may be cited as:
De Ketelaere, J., Santos, J., Domingos, T., 2026. Total factor productivity and final-to-useful exergy efficiency in Austria, 1960–2019: A cointegration analysis. Energies 19(17), 4026. https://doi.org/10.3390/en19174026