Abstract
This paper analyses the optimal environmental policy design in situations where the regulator - hereafter a government - can not strongly commit to announcements about future tax and subsidy levels. The motivation is that long-run perspectives of environmental policies often face short run concerns. One consistent illustration in Europe is related to the french government which cancelled the carbon tax increase for year 2019 following recent demonstrations of the yellow vest movement. Other examples include the case of the australian government abolishing the carbon tax in 2014, or the spanish government abruptly cancelling the renewable energy subsidies in 2012.I specifically consider environmental policies which aim at supporting the transition from the use of dirty technologies to clean technologies by subsidizing innovation. The interplay between innovation and environmental policies has been extensively addressed.1 However, a large share of the literature abstracts from the issue of commitment. In most papers, the analysis consists in comparing the optimal policy and a business-as-usual scenario (see Bosetti et al., 2009 ; Edenhoffer et al., 2006 ; Popp, 2006). Yet several authors point out that the government lack of commitment may lead to inefficient environmental innovation (see Wirl, 2013 ; Montero, 2011). The question thus arises: if a government can not strongly commit to announcements about future tax and subsidy levels, is there an efficient policy design? And, if so, how does it differ from the case of strong commitment?