Signals, Transparency and Forward Guidance

6.2.1 Transparency As a Precondition for the

Effectiveness of Signals

In the preceding section we introduced a discussion of sig­nals. The problem there was whether institutions and poli­cies can lay down the right signals or system of signals for a proper conduct of both private agents and the policy­makers, in particular, to prevent the start of a crisis.

The information provided by signals relates to the trans­parency of actions and goals of the various agents. By the term ‘transparency’ we mean visibility of the information referred to one’s action (especially policymakers) and its inferability, i.e. the potential for other agents to understand its implications (Michener and Bersch 2013). These aspects - or preconditions for transparency - have numerous require­ments, in particular, the issuance of proper signals by some agents and the ability of other agents to rationally deduce their implications.

Reciprocal transparency of signals issued by private and public agents would be useful. However, on the one hand, the private agents’ conduct is easier to foresee. On the other, even if - and to the extent to which - it is not, in a market system policymakers cannot require transparency on the side of private agents for most of their choices, even if excep­tions are diffuse, such as in the case of antitrust legislation or tax compliance and entitlement to subsidies.

As for public agents, transparency has been referred in particular to monetary policy, for a number of reasons, espe­cially in connection with independence of central banks, which raises the additional issue of their accountability. Transparency can refer to at least four different aspects of monetary policy: objectives and strategies, motives behind certain policy decisions, economic outlook of the economy and future monetary policy decisions (Blinder et al. 2008).^[115] Actually, until not long ago, monetary policy was conceived as an art, almost an arcane matter. After the 1960s, the mys­tique of central banks gradually disappeared to full disclo­sure of the various objects of transparency along an increasing trend.

The literature on the effects of transparency about the action of the central bank trying to influence inflation and output has grown in the last fifteen years, parallel to its growing implementation. The conclusions derived have changed through time. In the 1980s, it was common to ask for a high degree of opacity, mostly under the influence of the argument that the ‘surprise effect' was a necessary element. Rather recently, Blinder et al. (2008) have concluded that higher transparency positively affects macroeconomic out­comes. Indeed, higher transparency confers greater credibil­ity to the actions of the central bank, as this can explain to the public the details of the actions and how they can affect the economy. It influences inferability and expectations, thus enabling agents in the private sector to plan their decisions (especially those that need long-term projections, such as investment) and to coordinate their plans. In fact, they can base their plans on firm and reliable expectations about future short-term interest rates, as the long-term rate - which influences investment - depends on them.

To be more specific, up to the crisis, the literature on transparency predicated an optimal degree of transpar­ency intermediate between the two extremes of complete opacity and complete transparency, with various accents on the two extremes according to the models used.

In more recent work, Demertzis and Hughes Hallett (2015) have concluded that transparency does not much affect macroeconomic averages but lowers inflation variability (the opposite may occur for output, which justifies a limited degree of ambiguity). Another argument for this position is offered by Friedman (2015: 16), who is still scep­tical about the ability of central banks to supply the optimal degree of transparency, since ‘the market will inevitably want to know more than what policymakers can possibly disclose'.

The orientation in favour of some ambiguity has chan­ged to a large extent as an effect of the practical imple­mentation of forward guidance. In fact, as far as theoretical analysis is concerned, REs may even reinforce the policymaker's ability to pursue his or her targets when he or she provides suitable information about his or her future actions through forward guidance, the object of the next subsection. Before passing to it, we must add that one aspect of the optimal level of transparency has specifically to do with its inferability. It has been shown that there is a trade-off between the level of information provided and its possibility of being acquired and used by the public, due to the risk of overwhelming communication and the inability of the private sector to choose among the various signals and correctly process the right ones.

6.2.2 Transparency and Forward Guidance

Forward guidance has introduced novelties into the topic of transparency. In fact, transparency has risen to prominence recently with specific reference to the targets of central banks and their future actions, in an attempt to help the economies recover from the crisis or stabilise them in a crisis. The issue is whether this is limited to the current situation of zero lower bound (ZLB), which requires some kind of engagement for future policy (Williams 2013).

In the abstract case where not only the conditions for dynamic controllability are satisfied but also perfect knowl­edge of both the model and of the policymaker’s preferences is ensured, forward-looking expectations actually enhance the ability of the policymaker to reach his or her targets and guarantees controllability in a shorter time. In this case, knowl­edge of the policymaker’s preferred target values becomes crucial to form ‘correct’ and thus ‘useful’ expectations by the private sector. This knowledge can be spread directly by pro­viding information about the policy targets or indirectly via other announcements that offer information about future poli­cies. According to this view, in the case where expectations reinforce policy actions, a good communication system becomes an essential prerequisite for good policy.

In order to be able to consider policy announcements as an instrument of economic policy for supplementing or extend­ing the impact of conventional policy instruments, some circumstances must be clarified. In fact, even if real policy interventions show the multiplicity of practical cases where intentions of future policy are announced, until now, the literature has failed to identify (1) the conditions under which the expectations deriving from policy announcements (and also real policy interventions) can be managed, (2) their effect on the scope for policymaking (as distinct from the possibility of managing expectations) and (3) how and when unconventional policy instruments will be necessary.

Only after giving a satisfactory answer to these questions can the conditions under which policy announcements may be used as policy tools be stated. If these conditions are met, announcements can be added as new policy tools to those already in existence. This can allow policymakers to satisfy conditions for controlling the economic system even in the absence of other instruments, in particular, fiscal ones, thus getting rid of the conditions under which an impasse devel­ops that impairs policy control.

If, by contrast, announcements are ineffective, REs could make the whole policymaking action ineffective, unless other solutions of the kind indicated in Chapter 5 can be found. In particular, in order to overcome a crisis situation - such as the one that has recently developed - mutually supporting combinations of fiscal and monetary policy would be needed.

Forward guidance should be seen from the perspective of transparency and signals on policy action. From this per­spective, policy announcements, if communicated properly, can be used to supplement or extend the impact of conven­tional policy instruments. Specifically, within a general RE framework, policy invariance can only arise in specific cases of parameter values. In all other cases, policy announce­ments can be used by policymakers to help steer economic behaviour, and as a result, certain targets can be reached in reduced time. Thus, expectations typically enhance the power to control the economy over time. The rationale for this result can be understood by using the concept of con­trollability from the theory of economic policy and its dynamic extensions (see Chapter 4). Put differently, if a policymaker can achieve any desired vector of targets given exogenous expectations, then he or she will also be able to do so with endogenous expectations, i.e. with expec­tations generated by its announcements. If nothing else, the policymaker can attempt to steer the endogenous expecta­tions to achieve specific targets in a shorter time.

To make use of this property of REs, however, another ingredient should be added. The policymakers must be able to communicate, in a clear and effective manner, the intent and purpose of their policies and how exactly those policies can be expected to produce the desired result. This will be necessary to convince the private sector that the promised policy measures will in fact be undertaken when they become due and that the intended outcomes can reasonably be expected to be achieved. Otherwise, there are no grounds to suppose that the private sector would shift, or anchor, its expectations in a way that adds to the policymakers’ ability to reach their desired goals, such as recovery, smooth exit, debt reduction and low inflation.

The crucial element is to reaffirm the targets and why the policies chosen can be expected to reach them. Woodford (2003) observed that policy trade-offs will be eased when expectations fall in line. And Libich (2009: 685) underlines that ^,[s]everal recent empirical papers contributed to this debate by showing that in countries with an explicit (i.e., legislated) inflation target expectations are better anchored’.

Astute policymakers will therefore realise that good com­munication lies at the heart of the policy problem if they want to reach their policy targets in the early periods or at lower cost. This is a fact that has not been lost by central bank policymakers in their attempts to control or anchor private­sector expectations of future inflation in such a way as to make interest-rate policies more effective (Woodford 2005; Blinder et al. 2008; Rudebusch and Williams 2008).

Thus, forward-looking expectations are a powerful mechanism, in combination with the chosen policy values, for influencing the natural dynamics of the economy. The proposition on announcement-based controllability sta­ted earlier (Chapter 5) shows how communication and policy announcements can be exploited to supplement and extend

the impact of conventional policy instruments. This gives a formal justification for using policies designed to manage expectations, such as publishing interest-rate forecasts, or a future exit strategy, or the details of a programme of quan­titative easing (QE), as well as defining the circumstances in which expectations cannot be anchored or steered (known as cases of policy impotence, instances in which it would be wiser for central banks not to act). Recent research on the effects of US unconventional policies on the economy show that an important factor has been the gradual change in policy expectations by the public, which thought that the pace of recovery would be much faster, anticipating a rapid rebound of the economy and cancellation of the policies. The novelty of the situation could have contributed to this (Engen, Laubach and Reifschneider 2015).

A key implication of forward guidance is that the quality and credibility of communication by policymakers are key issues (Acocella, Di Bartolomeo and Hughes Hallett 2013). Examining the conditions that permit effective signalling and commitment is important.^[116] But equally important is to recognise that there is a large class of problems for which effective signalling and commitment are neither necessary nor relevant. However, clear communication is still required in practice for a large class of policy problems in order to allow the agents to check the consistency of the announced policies and target values with their own information. This is relevant from a number of points of view: in particular, for ensuring time consistency and to allow for policymakers’ accountabil­ity. The latter issue is dealt with in the next subsection.

6.2.3 Rules, Incentives, Degree of Centralisation for the Accountability of Policymakers

This subsection is organised as follows: (1) We first discuss the issue of constraints on the actions of policymakers as a way to ensure certain results; (2) we show that they can be insufficient to get such results and that it is advisable to follow the other path to avoid moral hazard, i.e. the one of incentives (and punishments); (3) we then deal with the nature of rewards, whether economic or of a different kind; (4) within non-economic incentives we discuss the interac­tions between formal incentives and social norms, the accountability of policymakers and the relevance of partisan­ship; and (5) we finally discuss the issue of the optimal organisation and accountability of the government from the point of view of its degree of centralisation.

As to the incentives deriving from constraints imposed on the action of policymakers, one could think that institutions should establish a regime of fiscal and monetary discipline and reforms, as these would enhance efficiency and reduce moral hazard by politicians. Strict policy rules as well as market liberalisation, privatisation and so on are often thought to constrain private agents to act in an efficient way and public agents not to pursue their personal or partisan interests. In the EMU, a regime of discipline was enforced both in the way to the common currency (to constrain the countries with higher inflation and public deficit and/or debt, for which fulfilling the Maastricht requirements was more problematic) and afterwards (as the ECB had to estab­lish a reputation of conservativeness and the Stability and Growth Pact (SGP) was in effect). Also in the United States and certain other countries, public debt and/or deficit caps are set. Market liberalisation and similar constraining poli­cies have often inspired international institutions, e.g. the IMF and the World Bank (as an expression of the Washington Consensus) when providing lending or aid.

The literature on the usefulness of constraints to public action is on the whole rather sceptical about the effectiveness of a disciplinary regime. In Coricelli, Cukierman and Dalmazzo (2006), a stricter monetary policy has positive effects on both inflation and unemployment, as it imposes a discipline on trade unions. Fatas and Mihov (2006) find that, in the United States, of the two possible effects of bud­getary restrictions - reducing policy discretion and limiting the fiscal responsiveness to shocks - the first effect prevails, and fiscal policy restrictions lead to less volatility.

On the opposite side, according to Dalmazzo, ‘commit­ment to price-stability may allow governments to persist in “bad” fiscal policies and tolerance for low competition, as governments can trade part of the social gains deriving from it for distortionary taxation, redistribution, patronage and the like' (Dalmazzo 2014: 4). Hence, a more conservative central bank tends to produce the effect of a rise in the tax rate, thus questioning the desirability of this type of mone­tary authority claimed by Coricelli, Cukierman and Dalmazzo (2006). In addition, monetary discipline reduces market deregulation. Acemoglu et al. (2008) have presented a model in which reforms of the kind advocated by the Washington Consensus can be detrimental. In their opinion, such reforms induce politicians to adopt other instruments for furthering their redistributive actions, patronage and so on, thus originating a kind of ‘seesaw’ effect. These results cast doubts on the validity of the argument in favour of resorting to an external constraint in the form of a conservative central bank or other constraint in order to reform countries characterised by lax fiscal policies and scarcely competitive goods markets. As a conclusion on this point, discipline per se is not a way to get efficient conduct from policymakers and private agents if rules and signals do not contain proper incentives.

A problem to some extent related to that of disciplinary rules but indeed more general is that of obedience to rules. According to Lawrence Write, reported by Vanberg (2014: 219), ‘if everyone knows that the rule of law will be followed, such that nobody will be bailed out, the incen­tive for imprudence disappears... [and at least] there won't be system-wide mal-incentives producing an epi­demic of imprudence'. In order to discourage imprudence, there should also be rules to follow in case of emergency situations.

The problem arising from this position is that rules cannot foresee all possible contingencies. To face situations not included in the rule, some discretionary attitude must be necessary. An alternative to this way of avoiding malincen- tives could be that of providing positive incentives, which might lead to better results.

Then let us consider incentives rather than ‘disincentives'. Tirole (1994) underlined that economic theory has devel­oped a theory of organisations that has been applied for devising effective incentives in private institutions. This theory is based on three ingredients, i.e. adverse selection, moral hazard and incomplete contracting. Difficulties arise in governments making use of the tools that have been devised for firms, as the field of public organisations is rather different. In fact, both the type of maximand and the number of specific targets are different. The unity of goals in private organisations (maximisation of profits) must be replaced by multiple targets for the government, and maximisation is not required or is difficult or impossible to pursue. The existence of multiple principals, such as when realisation of some goals is entrusted to different departments, can also compli­cate the effectiveness of control. Finally, asymmetric infor­mation can compound troubles in the case of public organisations, because of the possible capture of public offi­cials by lobbies and interest groups. In any case, incentives also can be laid down for public institutions.

Then discussion is open on the proper ways to lay down incentives for effective policymaking. To do this, there are institutional (constitutional) rules that can provide an incen­tive to policymakers not to act in their interest and to be respectful of their constituencies. However, sometimes these rules are double edged and can have undesired effects, which implies that they should be designed with extreme accuracy.

As to the issue of the kinds of incentives to implement, the literature on the economics of organisations and employee compensation, which mainly refers to the organisation of private firms, provides a foil both for economic and for other, non-economic motivations (Prendergast 1999). Drawing on works from the corporate culture literature (e.g. Kreps 1990), we can refer to four kinds of incentives: work inputs; formal incentives such as pay, bonuses and so on; career incentives; and non-economic kinds of incentives. Work inputs refer mainly to employees. This is the simplest way of checking whether workers comply with their duties, and extra work is usually given an incentive. The ineffectiveness of such inputs is largely demonstrated, and we disregard them. We concentrate here on non­economic incentives and leave the issue of career incentives to later.

Akerlof and Kranton (2005) emphasise the pitfalls of monetary rewards and punishments to motivate workers to work in a firm's interest. They show that identity can act as a substitute for monetary incentives. Also, according to other authors, pecuniary incentives can crowd out work­ers' motivations to cooperate in the workplace (see e.g. Prendergast 2008; Berdud, Cabases and Nieto Vazquez 2014). More generally, Benabou and Tirole (2006) show that people's actions reflect a mix of altruistic motivation, material self-interest and social or self-image concerns. This mix varies across individuals and situations. Sometimes, altering the weights of the three components changes the outcome, thus leading to pro-social (or antiso­cial) behaviour.

A parallel issue involves the assessment of government performance. This is an awkward object of analysis because, in a world of partisan politics, performance can be differently judged according to the very different preferences of the electorate. In addition, partisan governments can issue sig­nals apparently contradictory to their real preferences in order to gain a larger support (Chatagny 2015).

An analysis of how the accountability of politicians affects public choices is necessary. Political accountability is, in fact, a fundamental feature for evaluating democracy and the performance of governments (Powell 2000). Different institutional arrangements provide differential incentives for politicians to behave in a correct way and to respect their engagements and promises. However, accountability requires a number of conditions to be satisfied for fulfilling this role and, if these are not met, entails some drawbacks.^[117] Along this route, Maskin and Tirole (2004) set out a two- period model. In each period, a homogeneous electorate must decide between two possible actions. The electorate believes one action to be optimal and can decide itself (direct democracy) or delegate some official, who will judge which action is optimal. Maskin and Tirole conclude that account­ability has both benefits (removal of officials whose conduct appears to be non-congruent with the electorate’s will) and drawbacks (as it encourages cheating by officials or pander­ing). It is desirable when (1) the electorate is informed about the optimal action, (2) acquiring information about the decision is not costly, which is not the case with technical decisions and (3) feedback about the quality of decisions is fast, as the electorate can easily understand the optimality of the politician’s decision.

Partly along this route, one can ask whether rules devising the possibility of reappointment can serve as a way to reduce moral hazard and have policymakers acting in the ‘public interest'. Again, the issue returns to information and the cost and time for acquiring it in order to refute false claims. In fact, Lambert-Mogiliansky (2015) shows that reappoint­ment of public officials based on ‘cheap talk' complaints and their public defence of previous actions can be very useful.

The real issue is whether these conditions can be met in practice. From the point of view of transparency and costs to citizens for the accountability of policymakers, one of the most important issues, raised especially in more recent times, is distrust of citizens. A system for increasing account­ability and confidence and ‘bring[ing] government back to the people' might consist in a larger degree of decentralisa­tion of policymaking. We discuss this topic below and in Chapter 7.

Decentralisation and federalism - as opposed to centra­lised governance - have numerous supporters and detractors and can have mixed outcomes. The theoretical literature on federalism and decentralisation is vast and concerns in par­ticular (1) competition among sub-national governments, (2) fiscal federalism, (3) veto players and points, (4) accountabil­ity and (5) the size of government. We discuss here only the aspects related to accountability. Other issues will be dis­cussed in Chapter 7.

Decentralisation was initially supported on the argument that it can help in the taking of better decisions and ensure allocation efficiency. Local governments would be more responsive and accountable because of their proximity to interested citizens and their better information than the national government about local conditions and preferences (Hayek 1945). On this, Oates (1972) added that in the absence of spill-overs from one jurisdiction to the others - which would call for centralisation - fiscal responsibilities should be decentralised, as local governments are responsive to preference heterogeneity and needs (Oates' theorem of decentralisation). An argument pointing in the same direction stated that public goods should be provided by the jurisdiction of the minimum geographic area necessary for internalising the benefits and costs of such provision. The local government’s malfunctioning would be subject to the threat posed by citizens’ ‘voting with their feet’ due to competition among jurisdictions (Tiebout 1956; Oates 1972). A variant of this argument is that federalism can ensure better information - due to decentralisation - and competi­tion among jurisdictions, thus providing a credible commit­ment of the engagements taken at the lower levels of government (Qian and Weingast 1997).

However, decentralised solutions would ensure self­interested decision-making while disregarding the existence of positive or negative externalities and interrelationships between issues in the various jurisdictions. Apart from this issue, a regime ensuring these characteristics ‘does not come about spontaneously, but depends crucially upon a number of minimum political and social conditions’, such as an open, fair political system with binding rules, low barriers to accountability and, thus, transparency, social cohesion and organisation, with ‘central government as [a] neutral administrator and referee’ (Faguet 1997: 16).

Public-choice theorists insist on the necessity of binding rules in decentralised settings. Rules should limit politi­cians’ discretionary power and force them to serve the public interest (Buchanan 1995; Inman and Rubinfeld 1997). Rules, however, are not a ‘panacea’ for solving issues of account­ability, policy effectiveness and trust in government.^[118] Decentralisation works better if there is local democracy and self-governance, which are easier in a community that is homogeneous from economic and social points of view (Azfar et al. 1999), an issue that we will discuss further in Chapter 7. Dissemination of information is also relevant, since information heterogeneity favours centralisation to reduce rent extraction, while preferences hetereogeneity pushes towards decentralisation (Boffa, Piolatto and Ponzetto 2016).

From an empirical point of view, results have been mixed. First, lobbies and vested interests can ‘capture’ local govern­ment bodies and impair competition among local constitu­encies (Shah 2008). In addition, in some cases, poor analysis of the characteristics of local communities, inexperience and the weak capacity of local government (especially in less developed countries), as well as an excess of controls by central governments, together with the multiplication of veto players, have impaired the results of decentralisation. In order to succeed, decentralisation should take place in a manner that adds to the accountability of local govern­ments, e.g. by increasing representation of women and other vulnerable groups and allowing for recalls of elected officials from public office (Ylmaz et al. 2008). However, as mentioned earlier, decentralisation would ensure the self­interested decision-making of communities at the cost of disregarding the existence of positive or negative externalities.^[119] It should then be avoided when these are relevant, unless a federal government can provide effective solutions to account for such interrelationships. We will also return to this issue in Chapter 7.

Quite different conclusions are reached by authors advo­cating anarchist solutions, who hold that a society could be governed by free and equal people directly or through their delegates (in a representative democracy) as an outcome of a convergence of interests (Candela 2014).

6.3