Current human activity is degrading the environment and depleting biotic and abiotic resources at unheard-of rates, inducing global environmental change and jeopardising the development of humankind. The structure of human activity determines which resources are extracted, how they are transformed and where and how they are emitted back to the environment. Thus, the structure of human activity ultimately determines the human-Earth System interaction and human-induced environmental degradation. Several theories and empirical findings suggest that a cyclic structure would lower the resource requirements and emissions of the economic system, decoupling production and consumption from their environmental impacts. However, the cyclic structure has not been fully characterised nor related to the resource requirements or emission generation estimates of environmental impacts as calculated through models representing the physical structure of the economic system.

This thesis is interested in developing tools to analyse the physical structure of the economic system and, ultimately, to develop a method to identify its cyclic structure and relate it to the environmental impact induced by economic activity. Using this new knowledge, it might be possible to reduce the environmental impact of the economy by altering its physical structure.

In chapter 3, the different methods to calculate the emissions and resources associated to a given final demand of physical input-output tables are reviewed because they gather different results; it is argued that only two are valid. Surprisingly, these two methods reveal different physical structures; these are explored using a backward linkage analysis and their differences explained. It is found that only one method is appropriate to analyse the physical structure of the economic system and this method is in fact a new input-output model capable of tracing by-products as final outputs. Also, since traditional input-output structural analyses provide aggregate measures, a visual representation of input-output tables enabling researchers to perform disaggregated structural analyses and identify intersectoral patterns is developed.

In chapter 4, a method to derive the full cyclic structure of the economic system is developed using network analysis within the Input-Output framework; it identifies the intersectoral cycles and the resources and emissions associated to cycling. It is shown that cyclic flows maximise the system throughput but lower the resource efficiency of the system vis-à-vis the system outputs. It is demonstrated that 1) the complete structure is composed of a cyclic-acyclic and a direct-indirect sub-structure, challenging the common understanding of the functioning of the structure, and 2) cycling is composed of pre-consumer cycling, post-consumer cycling, re-cycling and trans-cycling.

In chapter 5, a set of indicators are developed to capture the weight and emissions associated to each sub-structure and the sub-structures are related to the economy's resource efficiency and emissions.

In chapter 6, it is illustrated how to use the concepts, indicators and methods developed in previous chapters to identify strategies to improve the resource efficiency of the economy by altering its structure.

Finally, in chapter 7, it is suggested to refine the definition of recycling to integrate the different systemic effects of pre-consumer and post-consumer cycling and it is argued that the ideal structure of a circular, close-loop economy should minimise its pre-consumer cycling in favour of more efficient acyclic flows while maximising its post-consumer cycling.