The database comprises several variables related to the organizational characteristics of agencies on the one hand, and to the formal policymaking process on the other. Questions included the year of establishment of the agency and whether it was operative on 31 December 2010, together with fundamental organizational and institutional variables. These refer to the current legal document governing the regulatory agency and the specific regulatory policy competencies assigned to it, including the capacity to establish prices, to determine market entries and exits, to elaborate norms, to supervise, to implement sanctions, and to intervene in conflict resolution. Moreover, the database aimed to capture whether these competencies are the sole responsibility of the agency, or whether they are shared with other public actors, namely, other public agencies, the executive or the parliament. Finally, we established who is responsible for final regulatory decisions and whether this power lies exclusively with the agency head, the agency board, or neither.
In terms of the main actors responsible for the agency's regulatory decisions, data included the provisions and requirements determining the appointment, renewal and dismissal of agency heads on the one hand, and of boards and board members on the other. The specific functions of boards and their composition were also integrated into the database. In addition to some general questions on the size of the organization, the data also captured specific aspects of the internal organization and the finances of the agency. These referred to the formal internal distribution of responsibilities among board members and the identification of who is responsible for the elaboration of the budget proposal, how this is distributed, and who is in charge of personnel policies and practices.
To examine the accountability profiles of regulatory agencies, several questions identified the particular mechanisms devised by each agency. On the one hand, we looked for upward accountability mechanisms and distinguished to whom the agency was accountable, namely, the particular sector ministry, the executive collectively, and the parliament. Furthermore, the database captured the content of the specific obligations of the agency towards each one of these principals. We distinguished between one-sided (providing information only) and two-sided (approval of annual report or involving multiple controls) accountability mechanisms. Additionally, we looked for downward accountability mechanisms intended to allow the participation of agencies' audiences. These included a number of one-sided downward accountability mechanisms, namely, making available minutes of board meetings, agency resolutions, and annual reports to the public; and twosided mechanisms, including public hearings, open consultations, consumers' office and advisory council. Figure 2.1 lists all these mechanisms according to their popularity among agencies.
By constructing an Accountability Cube, we can further observe the percentage distribution of regulatory agencies based on their accountability characteristics. As to the information dimension, we can measure the distribution of one-sided accountability mechanisms, either upwards or downwards, to see to what extent the regulatory agency provides an account of its behaviour towards its fora. The second phase refers to the discussion dimension, where our focus is on the existing two-sided downward mechanisms as they allow the forum to ask for further information, formulate new questions and provide an assessment of the information provided during the previous phase. Finally, and in terms of the consequences dimension, we can observe and measure the distribution of upward two-sided mechanisms that also allow the passing of sanctions, either positive or negative.
As Figure 2.2 shows, agencies relying on both high information and high discussion accountability mechanisms represent only about 10 per cent of the total, while agencies displaying both low information and discussion mechanisms constitute more than 50 per cent of the cases. Such distribution already suggests that the dominant pattern of agencies' accountability is more closely related to conventional models of public bureaucracies than to innovative frameworks. However, it also confirms the existence of large variations, and to some extent the presence of alternative forms of accountability in some of these agencies. Obviously, our classification does not capture all existing mechanisms, especially informal downward mechanisms related to the dialogue and consequences dimensions. Nevertheless, through the Accountability Cube we expect to capture an initial distribution of how agencies perform their accountability duties.