Technological progress is one of the main driving forces of economic development and affects all sectors of production in terms of intersectoral spillovers, changes of relative prices and income distributions. Models in the literature on technical change do not take all these multi-sectoral interdependencies into account and focus on particular issues. The dissertation at hand attempts to narrow this gap by investigating innovations in large-scale multi-sectoral economies, both theoretically and empirically.The thesis develops a multi-sectoral, constant returns to scale Cobb-Douglas model and implements product, process, organizational, and social innovations in this framework. The main analytical findings are: (i) The wage-profit-distribution of any Cobb-Douglas economy is given by a simple hyperbola and depends negatively on the elasticities of capital inputs. (ii) Hicks-neutral technological progress is unlikely, but possible in multi-sectoral approaches. Harrod- and Solow-neutral change is generally ruled out. (iii) Ex ante, sectoral innovations can neither be classified as capital- nor labour-augmenting due to intersectoral spillovers. Hence, there are significant differences with respect to conventional one-sectoral models that are widely used in economics.The most important conclusions of the empirical analysis of ten different countries in the period 1995 to 2009 are: (i) Technological change does not cause a convergence of actual wages and profit rates between countries, but the underlying wage-profit-curves are becoming more similar. (ii) On average, a share of about 60% of total wage growth is driven by capital-augmenting innovations and a share of 40% is imputed to total factor productivity increases. Country results differ and geographical connectedness supports similar developments. (iii) Technological key parameters of the ten countries under consideration are strongly correlated, especially for sectors with tradable commodities.