This study uses original French firm-level data to estimate the causal impact of green innovations (i.e. made to protect the environment) on job quality and inequalities across workers. Doing so, our study lies at the frontier between the fields of socio-economy of labour and ecological economics. Research question The need for an ecological transition of our economies – to contain global warming and preserve biodiversity – is now widely recognized. This burgeoning transition has major implications for employment and work. While it can be an opportunity to create new jobs and skills, it also carries risks of increased inequalities on the labour market. The academic literature in economics mainly evaluates the ecological transition according to its impact on economic well-being or growth while its effects on employment are less documented (Perrier and Quirion, 2017). Moreover, existing studies generally adopt a forecasting approach and focus primarily on the quantity of jobs potentially created or destroyed. The relative quality of these jobs and the characteristics of workers who occupy them is hardly addressed (ILO, 2018). We propose a firm-level approach of the relationship between the ecological transition on one side and labour on the other side. Indeed, firms are key players in the ecological transition, in particular through their effort to innovate (creating new products or processes) so as to protect the environment or the biodiversity. However, research on the effects of their environmental innovations (green investments) on employment is still scarce and does not provide clear-cut results (Elliott et al., 2021). Our objective will therefore be to identify firms engaged in environmental innovations/investments, and then to observe the extent to which this impacts the quantity and quality of jobs over several years, as well as inequalities across workers according to their gender, age and skill-level. Do we find that green innovations are more likely to increase employment for higher occupations as it was proved for innovation in general in France (Duhautois et al., 2022)? Or do we rather observe a polarization effect (Autor et al., 2003, Goos et al., 2014) when it comes to environmental innovations? Our project will thus bring new insights on the potential inequalities that may arise following environmental innovations, in terms of skill-level (proxied by occupation) but also in terms of gender and age. Do environmental innovations have a greater impact (positive or negative) on men’s or women’s employment? Do younger workers benefit more from environmental innovations? Data and methods To identify and measure environmental innovations and investments, we will use two distinct surveys, the Community Innovation Survey (CIS) on one side and the Antipol survey (Enquête sur les investissements dans l'industrie pour protéger l'environnement) on the other side, while information on employment and some dimensions of job quality will be found in the administrative database Base Tous Salariés (BTS). We will build two different datasets: one matching CIS with BTS and the other matching Antipol with BTS. The CIS survey includes information on environmental innovations (Horbach and Rennings, 2013) and ANTIPOL on green investments. We will thus be able to test the sensitivity of our results on these two databases, depending on the proxy we use to measure innovations or investments related to environment. More precisely, CIS includes 18,000 firms with 10 or more employees in the market sector, excluding agriculture. The survey is conducted every two years and some editions, in particular the 2020 edition, include a section entirely devoted to environmental innovations. This section reports whether or not firms have introduced, in the past three years (2018 to 2020), innovations that provide specific environmental benefits: the list of environmental benefits includes the reduction of CO2 emissions, replacement of fossil fuels by renewable energies, the reduction in waste and water consumption, etc. Companies are also asked about the factors that prompted them to implement these environmental innovations (regulations, direct cost of pollution, reputation, etc.). The Antipol survey aims to inform, on a yearly basis, on the amount and nature of investments engaged to protect the environment (defined as the purchase of buildings, land, machinery or equipment to treat, measure, control or limit the pollution generated by the economic activity of the workplace). These investments are decomposed into different environmental benefits with items quite similar to the CIS survey. The survey covers all manufacturing industries but not agriculture nor services. The (representative) sample includes about 11,000 establishments with 20 employees or more. Regarding employment, we will use the Base Tous Salariés (BTS). The BTS are administrative data collected every year on the basis of establishments’ compulsory declarations. These include information at the firm, workplace and job level in the private sector on employment and wages. Information can be disaggregated by occupations (except information on contract types), gender and age, allowing to shed light on the issue of workers’ heterogeneity and potential inequalities induced by these environmental innovations/investments. Finally, we will use fiscal data (FARE-FICUS) to obtain standard accounting information, used as control variables in our regressions. Based on our two matched datasets, we will perform multivariate regressions aiming to estimate causal relationships between environmental innovations (using CIS) or environmental investments (using Antipol) on job quantity, some dimensions of job quality and inequalities across workers, depending on their gender, occupation and age. To do so, we will rely on a two-step empirical strategy (already implemented by Duhautois et al., 2022 for technological innovations based on CIS). In a first step, we will use a Propensity Score Matching (PSM) model to obtain a sample of innovative and non-innovative firms/workplaces as similar as possible on observables. Then, we will use a difference-in-differences approach, to observe the evolution of employment before and after environmental innovations, as compared to the evolution in non-innovative organizations – therefore controlling for unobservable heterogeneity across organizations. Policy implications Our focus on (labor market) inequalities and on the heterogeneity of the effect of ecological transition across workers’ situations echoes the literature on “just transition”. Indeed, women appear to be under-represented in so-called “green jobs” while the necessity of skills’ upgrade in the ecological transition may raise problems for low-skilled and ageing workers. Our findings will help to better understand these potential sources of inequalities, in terms of gender, skills and age. As such, it opens to a discussion on policy recommendations aiming at tempering induced inequalities and/or exclusion as well as encouraging virtuous practices. These questions are of paramount importance in a context of rapid transformations in the world of work combined with the urgency of an ecological transition (OECD, 2023).