The foreseeable outcomes of the transition towards low-carbon economies are multiple and affect in different ways policy-makers, world regions, firms and consumers. It has long been acknowledged that at the core of this transition stand environmental innovations which are developed to enhance the long-term sustainability of economic growth.

The main pillars of this study are two. First, environmental challenges are different, and so are the responses that are needed to tackle them. The main consequence of this is that the current focus on green technology as a homogeneous block of undifferentiated entities is misleading.

Second, the adaptation of production and distribution systems is ultimately carried through by human labour and analysing the transition to environmentally sustainable societies requires a thorough understanding of how work activities are designed, implemented and changed to accommodate new policy imperatives and/or new technological opportunities.

Empirical evidence on either of these two pillars is scant or fragmented. The present thesis seeks to fill these gaps through the development of a database on green innovations, of a measure of the life cycle of green technologies, and of the corresponding explorations to scrutinize the relation between green technology production, the territories’ characteristics and skills’ base of labour market over time and space.

The dataset is created identifying green patent in PATSTAT 2016a database using ENV-TECH classification (OECD, 2016) and geolocalizing their inventors. The result is a database covering green innovation worldwide from the 19th century to 2015, even if the period studied is smaller: 1970-2010. This dataset permits a first overview of green technologies evolution over time and space, where we can see differences in terms of country evolution and among technologies in terms of complexity, maybe related with the presence of an heteregeneous body of emerging and mature technologies.

To explore further this hypothesis, we develop a methodology to measure technology life cycle stages, and we apply it to understand the patterns of evolution of green technology production at country level. We find that capabilities are more important than wealth to diversify in green innovation, and mature green technologies are positively associated with specialization.

We continue the exploration of the relation between local capabilities, life cycle and patent activity in US federal states where we discover that green innovation is more associated than innovation in general with the recombination of distant knowledge, especially in early phases of the life cycle.

Finally, we investigate at US commuting zones level the effects of public procurement on green innovation, taking into account local capabilities again, but using labour market skills instead of knowledge recombination variety. We find that green public procurement has a positive and significant effect, in particular in territories with an important share of abstract skills in labour population.