*I analyze environmental policies that account for emissions damage due to fossil-fuel elec- tricity production and that foster investment in intermittent renewable capacities. I start with a setting of consumers on flat-rate tariff who cannot adapt their demand to inter- mittent electricity supply. I therefore define a constraint social welfare and study its decentralization in a competitive market with a Pigouvian tax. I find that the tax does not implement the first-best constraint allocations. With flexible consumers on state- contingent tariffs, I also find that the first-best allocations are unreachable with the tax. Similar observations are made in both settings of consumers if tradable emissions permits are implemented. Due to intermittency of renewables, I propose that these emissions pric- ing instruments be complemented with direct subsidies for renewables-based technologies such as the feed-in tariff. I am then closer to a second-best solution.

**What is the link between market competition and the equilibrium level of unemployment? This paper augments the standard Dixit-Stiglitz (1977) model with labor matching frictions and non-CES preferences. It connects the two well-known market distortions: the allocational distortion (too many firms producing too little) and the labor market distortion (firms posting too few vacancies). The first result describes efficiency in the labor market. I show that the unemployment level generated in the market equilibrium under monopolistic competition is always inefficiently high. Namely, the under-employment arises due to incomplete appropriability distortion, i.e. the fact that each firm internalizes only a fraction of the consumer surplus in its revenue. Second, I show that VES modeling is crucial to explain why increased competition in the product market can have a positive effect on the employment level. Only under CES preferences, the two markets are completely separable. The third result of the paper discusses the economic trade-off between product market efficiency and employment level. While correction of excessive firm entry by the means of licensing costs is an efficiency-enhancing policy on the product market, it is detrimental for employment on the labor market. Calibration of the model for the US economy shows that the predicted drop in employment due to the correction of Dixit-Stiglitz distortion is large enough to completely offset any benefits of such policy on the product market and to reduce the overall welfare.