Therefore, this paper contributes to the literature which seeks to understand if and when lobbying activities can help improve the social welfare. To the best of our knowledge, this literature is not much developed, because most researches adopt the positive point of view. We briefly review below a selection of articles that shed light on this issue, with varying domains of application.

Becker (1983) determines the political equilibrium resulting from the competition among pressure groups for political influence. He shows that the pressure groups have an incentive to reduce the deadweight costs of policies. Gradstein (1993) compares the performance of the private and public provisions of local public goods. Both allocation procedures create inefficiencies, the former being undermined by free riding incentives, the latter entailing socially wasteful costs of lobbying. Gradstein (1993) finds that public provision performs better when there is a small number of large jurisdictions. Aidt (1998) derives the equilibrium environmental policy in a common agency model of politics (Grossman and Helpman, 1994), where the government seeks to attract campaign contributions from lobby groups in order to maximize the likelihood of being reelected. Aidt (1998) shows that the equilibrium environmental policy is the (socially optimal) Pigouvian tax if all agents have their interests represented by an interest group. Recently, Aidt and Hwang (2014) have generalized the argument to the case of cross-national externalities and foreign lobbying. They show that the political equilibrium is socially efficient if all social groups are organized and all governments are equally corrupt. Finally, Graichen et al. (2001) analyze a Tullock contest, where an incumbent monopolistic utility and environmentalists confront for the contract for the supply of electricity. At the equilibrium, they show that the political contest induces the monopolist to use a more environmentally friendly technology and to reduce his output.

Tullock (1967) is the first to alert that the real resources and efforts that lobbies expend to influence regulatory decisions are socially wasted. Since then, a large body of literature has focused on evaluating the magnitude of the social loss due to rent-seeking activities (Krueger, 1974). The initial framework designed to investigate it considered symmetric rent-seekers simultaneously determining their efforts under complete information (Tullock, 1980). Using stylized contest success functions (lottery and all-pay auction), Tullock (1980) and Hillman and Samet (1987) find that the rent dissipation is either smaller than or equal to the rent at stake. This result has been extensively challenged by the literature and can be considered as robust. Hillman and Katz (1984) and Alcalde and Dahm (2010) confirm it with more general contest success functions. Hillman and Riley (1989), Baik (1994) and Nti (1999) obtain that rent dissipation reduces if the rent-seekers are asymmetric, in the sense that they put different valuations on the prize and/or differ in their ability to win the contest. Leininger (1993) and Morgan (2003) extend the property of under-dissipation of the rent considering sequential timings of the contest game. Exploring a variety of information structures, Wärneryd (2003) and Malueg and Yates (2004) show that the intensity of rent-seeking does not increase in settings with incomplete information. Finally, Katz et al (1990) and Nitzan (1991) obtain that the free-riding incentives within pressure groups also limit rent dissipation. ¹¹Recent surveys are Hillman (2013), Konrad (2007) and Long (2014).