Publication
The network industries have been undergoing a process of reform, where liberalisation is one of the main features. This thesis studies the effect of liberalisation, with and without competition, on the incumbent's incentives to innovate. A model of incumbent network operator is developed and analysed when the incumbent is a monopolist, as well as when it faces an entrant. The objectives of the incumbent are specified in a general manner to allow for revenue, profit, and/or welfare maximisation. The marginal cost of the incumbent is assumed to depend upon the investment in new technologies and processes. A strictly convex and decreasing cost function is assumed. The incumbent maximises its objective function with respect to prices and to investment in innovation. The entrant is assumed to maximise profits with respect to prices. The incumbent's incentives to innovate under monopoly and duopoly are compared. One of the main results is that the difference between the investment in innovation under monopoly and under duopoly is determined by the incumbent's elasticity of demand under monopoly as well as by the incumbent's market share and elasticity of demand under duopoly. For certain values of these variables it exists an interval where duopoly provides more incentives to innovate than monopoly. The market share of the incumbent has a non-linear relationship with the investment in innovation under duopoly. Until a certain point an increase in the incumbent's market share creates more incentives to innovate under duopoly and from that point on the contrary happens. A decrease (in absolute value) in the incumbent's elasticity of demand has a negative effect on the incentives to innovate under both market structures. Another major result is that the incentives to innovate increase when the incumbent places greater weight on social welfare. The effect of liberalisation and competition on innovation in the postal sector is empirically assessed. The impact of the quantity supplied and of some control variables is also analysed. An original dataset is put together to perform the analysis. It includes data for seventeen European countries, over ten years. Innovation is measured using an innovation index, the accumulated number of innovations (both based on the results of a survey developed for this purpose), and labour productivity. A liberalisation index is built in order to measure the percentage of liberalised market. The econometric analysis performed, where several models were estimated by GLS and using PW-PCSE, shows that: (1) market liberalisation has a positive effect on innovation, and (2) an increase in the market share of the competitors stimulates the incumbent's investment in innovation, at least until the market share of the competitors reaches a certain threshold. Letter volume and GDP per capita are also significant and have a positive relationship with innovation. In general, the models estimated have a high explanatory power. The economet