Social accountability and transparency tools in practice
We have selected Spain and Italy based on two criteria: similar administrative traditions and differing degrees of structural independence between their IRAs in the telecommunications sector. The Spanish authority, CMT, is generally thought to be less independent than the Italian one (Gilardi 2008; Nesti and Righettini 2012). The Spanish CMT, established in 1996, can be described as a semi-independent regulator (Jordana and Sancho 2005), as the government holds a considerable degree of regulatory power, whereas the Italian central government has delegated significant regulatory competencies to the Italian AGCOM, 'established in 1997 to replace the previously existing Guarantor for Radio and TV broadcasting and Press' (Nesti and Righettini 2012: p. 22).
Another important difference between the two IRAs concerns the issue of convergence: AGCOM is responsible for the audiovisual sector, whereas CMT is not. In 2010, an attempted reform of the Spanish audiovisual regulators failed; since then the Spanish government has definitively given up on this idea for economic reasons related to the global crisis and has distributed the authority's audiovisual functions among the Comision del Mercados y Competencia and the Ministry of Industry. Moreover, in October 2013 even the CMT was merged with the new multi-utility regulator CNMC (Comision National de los Mercados y la Competencia) agency in order to reduce state expenditure. 
In order to analyse the several ways in which transparency and accountability are embedded in institutional arrangements, we have focused on the relationships between the authorities and their policy constituencies, namely the operators and their final users/consumers. These relationships will be considered with respect to different kinds of transparency policies that allow exchanges of cognitive resources between IRAs and stakeholders:
- 3. Regulatory Impact Analysis (RIA);
- 4. Other participation enhancement mechanisms.
Consultation can be defined as a set of techniques used by administrators to begin an exchange of information and feedback with stakeholders on a particular issue (Schrefler 2012). But this does not always ensure the inclusion of stakeholders' positions in the final decision (Bishop and Davis 2001). Essentially, the purpose of consultations is to gather information and opinions or to provide feedback on regulatory acts, therefore it is considered an input-oriented transparency tool. In this respect, stakeholders are involved through contributing their opinions and views, which increases the information available to the authority. Access to information has a significant impact on the promotion of better regulation, since it recognizes the capacity of regulated industries and stakeholders to discuss and generate policy options and helps to increase the authority's knowledge of potential conflicts at the implementation stage (Radaelli 1995, 1999; Schrefler 2012). The involvement of regulated subjects and stakeholders can increase the transparency and legitimacy of the regulatory process by filling the so- called democratic deficit of IRAs, which are not elected (Noveck 2004), and by ensuring IRA policy decisions are socially accountable.
Benchmarking is a way of learning and controlling performance 'using a loose form of community-based control' that requires two conditions: well-defined targets and individual participants who are willing to provide the necessary information (Scott 2001: p. 16; de la Porte et al. 2001). It serves as a process of comparing providers based on their technical performance, content, products, services and practices in order to focus on the industry's best practices and to facilitate their diffusion. Dimensions typically measured are quality, fairness and cost of services or products. Through the diffusion of benchmarking results we can learn how IRA targets, such as competitiveness and consumer protection, have been reached. Benchmarking can be defined as an indicator of transparency, as it provides useful information to competitors and users; it is an output-oriented transparency tool as it is a way of assessing and controlling operator performance (products and services) throughout their collaboration.
Regulatory Impact Analysis (RIA) focuses on better or smarter regulation and provides a selection of regulatory options based on discussions and the prediction of potential conflicts and effects due to specific interventions. RIA evaluates costs and benefits of regulation in order to estimate the advantages and disadvantages derived from new rules and policies (Radaelli and Meuwese 2012; Flori 2011). Used ex ante, RIA intervenes during the rule-making process and seeks to ensure that all options and possible effects are taken into account. RIA can be considered an indicator of transparency as it provides the most complete information on the options. It also depends on the recipients' involvement and provides the authority's reactions, observations, acceptance and implementation of recipients' proposals; therefore, it can be considered an input-oriented transparency tool.
Finally, other participatory mechanisms may assume different forms such as meetings, panels, interviews, focus groups, message boards and technical boards. These techniques, defined as closed consultations, involve a limited and selected number of stakeholders: firms, associations, operators and so on (Raiola 2010). Closed consultations may also be considered input-oriented transparency tools, although they are somewhat more limited in scope. All the data, gathered from 2005 to 2012 and used in this chapter, are derived from primary sources.2