This dissertation investigates the process of economic growth with heterogeneous agents from a multi-level perspective. Building upon Kaldorian and Evolutionary principles, growth is defined as a path-dependent and complex phenomenon, which requires structural variation and the interplay between demand and supply at distinct analytical levels. Two concomitant and dependent 'layers' of this process are emphasised: the supply-led 'intra-sectoral development trajectory' and the demand-led 'inter-sectoral development trajectory'. The key element in the first is the firm size, which is shown to have a non-linear influence on the process of technological change. The second layer is shown to depend on the growth of income and patterns of production and consumption reflected on the inter-sectoral composition and level of 'sophistication' of the productive structure. The key to understand divergent growth trajectories lies in the interaction between these layers and the contradictory effects imposed at each analytical level both by demand (top-down) and supply (bottom-up). The approach is both theoretical and empirical and the analysis reveals important stylised facts of growth at the firm, sector and country levels. The text is structured in four sections comprising 9 chapters. Section I introduces the theoretical foundations of the work and the limitations of Evolutionary and Kaldorian schools to explain the multi-level 'allocation problem'. Section II presents the databases and empirically assesses the influence of the (re)allocation of labour on growth at each analytical level. Section III investigates the foundations of the process of micro-meso and macro process of development. The final section proposes a unified theoretical framework to connect the multi-level evidence. The analysis reinforces the interplay between demand and supply in growth trajectories, prompting a number of original policy implications.

A high priority in order to combat climate change is disposal of waste. In low-income countries, a large portion of biomass residues generated in the forestry, agricultural and industrial sectors could be usable, instead of being seen as waste. For instance, it could be converted into biochar, which is proven to have many environmental benefits. In Rwanda, the agricultural sector employs 80% of the population and accounts for 35% of GDP. This sector, together with later refinement of crops and forestry production, cause large amounts of residue that many times is considered as waste. In this report, a literature study was conducted to evaluate possible biochar production from agricultural and wood wastes in Rwanda. Characteristics that determine if a biomass could be suitable for a biochar production were identified as C, H, O, N, S, hemicellulose, cellulose, lignin, ash and moisture content, residue-to-product ratio, and low heating value. These characteristics were assessed for the chosen Rwandan agricultural and wood wastes, by compiling values from published reports. The result shows that there are large volumes of residues that have potential for biochar production instead of being seen as waste in Rwanda. Biochar production from these wastes could enable environmental benefits for Rwanda, although further investigation of each single biomass could be needed in order to see if it is practically, technically and financially possible to do in reality. / För att bekämpa klimatförändringen är avfallshantering en hög prioritet. I låginkomstländer kan en stor andel av biomassarester som genereras i skogsbruk, jordbruks- och industrisektorer vara användbara, istället för att ses som avfall. Till exempel skulle det kunna omvandlas till biokol, som har visats sig ha många miljömässiga fördelar. I Rwanda arbetar 80% av befolkningen inom jordbrukssektorn och den står för 35% av BNP. Denna sektor, tillsammans med förädling av grödor och skogsbruksproduktion, orsakar stora mängder rester som många gånger betraktas som avfall. I denna rapport genomfördes en litteraturstudie för att utvärdera möjlig produktion av biokol från jordbruks- och träavfall i Rwanda. Egenskaper som avgör om en biomassa kan vara lämplig för en biokolsproduktion identifierades som C-, H-, O-, N-, S-, hemicellulosa-, cellulosa-, lignin-, ask- och fukthalt, samt andel avfall som uppstår i förhållande till färdig produkt och värmevärde. Dessa egenskaper utvärderades för det valda jordbruks- och träavfallet genom att sammanställa värden från publicerade rapporter. Resultatet visar att det finns stora volymer rester som har potential för biokolsproduktion istället för att ses som avfall i Rwanda. En biokolsproduktion från dessa avfall skulle kunna ge miljömässiga fördelar för Rwanda, även om ytterligare undersökning av varje enskild biomassa skulle behövas för att se om det är praktiskt, tekniskt och ekonomiskt möjligt att genomföra i verkligheten.