Independence and accountability vis-a-vis political and social or economic actors
RAs can be independent from and accountable to two types of subjects: political actors on the one hand, and economic actors on the other. Political actors are the 'principals' of RAs, those who delegate specific tasks to them and give them more or less autonomy and discretion in carrying out their work. Needless to say, political actors are the only ones who can delegate in the first place and who can extend, modify or revert the decision to delegate. 'Economic actors' can be both the regulatees - those whose activity is scrutinized by RAs - and other social actors who have an (economic) interest in the decisions taken by the agency - namely, associations of consumers, professionals and workers.
With regard to politics, independence is considered important because it helps build confidence in the state's regulatory stability. National and international investors are reassured by the presence of a regulator that cannot be influenced by the governments that alternate in office (see Majone 1996, 1997). If a country suffers from a lack of credibility, independent RAs can offer valuable help in the 'building up' of such credibility. However, the positive impact of independence is not merely symbolic: increasing independence also 'increases the agency's ability to incorporate knowledge of policy consequences into its decisions' (Bawn 1995: p. 66), thus yielding better regulatory performance. Typical instruments to create independence from politics are the length of appointment of agency members, the impossibility to dismiss them, the impossibility to renew the appointment and so forth.
On the other hand, politicians may also find it useful to hold RAs accountable for how they implement their tasks. To be effective, delegation should entail some form of control over the delegated person or body. This requirement derives from a more normative argument, according to which there must be continuity in the 'chain of delegation' that links voters to MPs, and MPs to bureaucrats (Strom 2000; Strom et al. 2003). Delegation is efficient if the principal manages to take advantage of the expertise and better information of the agent in order to improve policy enforcement. At the same time, if the principal has no means to prevent the agent shirking or colluding with the regula- tees, delegation can cease to be a net gain for the collectivity. Politicians can hold an agency accountable by asking it to present regular reports on its activity, by controlling its budget and by threatening to restrict its discretion.
However, an agency could be very independent from politics but at the same time collude with the economic and social actors it has to regulate. Political independence does not avoid capture. Indeed, it could make it even more difficult to prevent. This is why it is in the interest not only of politicians, but also of the collectivity, that forms of collusion between regulators and regulatees (and any stakeholder in general) are discouraged. Several means can be employed in this respect: requiring members to be experts in the field with high reputation; demanding that members of RAs must not have worked for regulated firms and must not work for them after they cease to work for the agency; and giving the agency the possibility to retain a part of the fines it levies on the regulatees in order to prevent conflicts of interest between the two.
Conversely, there do not seem to be advantages in making RAs accountable to regulatees, stakeholders and other social actors. When investigating regulated firms or when taking regulatory decisions, agencies must certainly ensure the greatest transparency - on the one hand, because in certain instances their functions are para-judiciary, and therefore they are bound to impartiality and fairness; and on the other, because their regulation decisions may reallocate costs and benefits among the regulatees, thus yielding an implicit redistributive effect. However, this commitment to transparency must not turn into accountability, as we have defined it in this chapter. This means that an agency must be completely autonomous from the firms when taking its decisions. It must face no consequences whatsoever for its rulings in favour of or against some economic actors.
This recapitulation of the different types of independence and accountability helps us identify the advantages and drawbacks of each. In the next section, I will employ these assumptions about politicians' preferences and the effects of independence and accountability to derive a decision-making model.