<p>Nomenclature: Acronyms, full names and units.</p>

Models use additive country, sector and year fixed effects, with errors clustered by country–sector.</p> <p>Findings/conclusions</p><p>No hypothesis is supported. The coefficient on SME concentration is negative throughout but insignificant (β₁ = −0.034, p = 0.104), and the null is informative rather than inconclusive: the confidence interval [−0.074, + 0.007] excludes positive associations of consequence. Moderation by financing costs, tax exposure and enterprise maturity does not reach five per cent (p = 0.090, 0.690, 0.053).

One association is robust: sectors starting from higher intensity decarbonise faster (β₂ = +0.021, p < 0.001). A second, on tax exposure, proves an artefact of tax-base composition and of a shared denominator.</p> <p>Limitations/future research</p><p>Sectoral SME concentration is not a usable proxy for family-firm environmental behaviour: 58.7 per cent of its variance is sector membership, and correcting for measurement error leaves the estimate unchanged.

We read this as evidence about the limits of ecological inference in family business research, not as a refutation of the theory. Progress requires linking ownership registers to environmental records.</p></div>