Searching for the Right Signals in a Market Economy to Prevent Crises and to Manage Them

6.1.1 Why Can Wrong Signals or Difficulties of

Signal Extraction Arise in a Market Economy?

The components of signals, which must be intended as pre­conditions for their effectiveness, are two: information pro­vision - together with the ability of agents to process it - and incentives (or disincentives) deriving from this provision.

The ability of private and public agents to perceive the right information for their conduct can be debated in a market economy. From this point of view, the issue emerges as to whether markets can send the right signals to policymakers and private agents and, if so, under what con­ditions. Symmetrically, what are the difficulties of accurate signal extraction in market economies?

One can say that information could be drawn by private agents from market trends and the constraints those trends impose on their conduct. However, there are a number of reasons why difficulties in signal extraction arise. Some refer to markets and political institutions in general. First, signals coming out of market trends can be noisy, depending on the existence of multiple equilibria. In this case, they could also direct agents towards a ‘bad’ equilibrium (in the case of equilibria that can be ranked in a univocal way). In addition, myopia of people and policymakers, populism and national specificities (including home bias) can lead agents to misinterpret information.¹ Short-sightedness is particularly acute with reference to financial market trends and certainly has been seen in the European Monetary Union (EMU; De Grauwe and Ji 2013). More generally, capital mar­kets can send signals, and policymakers and private agents may correctly interpret them.

Difficulties of private agents in discerning signals deriving from market trends can be remedied by proper supplemen­tary information supplied by policymakers (e.g. by announcements) and their policies. As to policies, excessive capital inflows and current account imbalances can be regu­lated by taxes or proper direct control. In the case of a current account imbalance due to competitiveness, wages could be lowered or raised according to the nature of the imbalance; correction also could be achieved by means of price and income policies. Imbalances in the current account not due to competitiveness could be faced by boosting or contracting aggregate demand.

Some authors are against similar proposals and raise objec­tions, especially addressed to using direct policy instru­ments. These are considered to be ‘naive and dangerous, because, by attempting to mimic through controls the out­come of market discipline, they are bound to confuse symp­toms with causes and direct the attention to policy tools that are entirely inappropriate as remedies against long-term structural deficiencies of market economies’ (European Economic Advisory Group 2011: 82, referring to the EMU).

6.1.2 Incentives, Signals, Moral Hazard

and Adverse Selection

As noted earlier, signals can induce proper actions by the private and public-sector agents only if some information is coupled with suitable incentives or disincentives. The case of a monetary union can help us to make the point. A number of requirements were prescribed for being admitted to the EMU. In the third and final phase of progress to the union, conditions of low inflation rates, low long-term interest rates, stable exchange rates and limits to public deficit and debt were required. Many countries were impressed by the pro­spective gains from being part of the EMU (the prize), and many private agents and the public sector behaved accordingly.

This outcome conflicts with what happened after admis­sion. The important signals of the balance of payments and the exchange rate were lost, not because of lack of informa­tion but because of a lack of proper, clear signals. For any country, relaxing the external constraint by means of free capital movements implied the possibility that a loss in competitiveness could be compensated by those movements. In fact, a rather high - but unsustainable - rate of growth in those countries could derive from the asset bubbles created out of easy credit conditions fed by capital that deficit coun­tries borrowed from abroad.

Constraints similar to those in existence before admission - such as the limits to public deficit and debt - were inherited by the EMU and are still in force for its member countries. However, they have often been violated. The 3 per cent limit on annual public-sector deficit has been ineffective after admission, since the incentives to comply are now lower than before admission, as shown by the case of violation of the rule by France, Germany and Portugal in 2003-4, not to mention later violations by others due mainly to the crisis and the inappropriate policies asked from member countries. This failure might have to do with the expected value of the benefit before entry compared to the cost after. Other factors that have probably acted at least in the case of France and Germany could be a calculation that they were in a position to avoid sanctions because of their dominant role in the EMU.

Other requirements, such as those implying sufficient competitiveness or limited current account imbalances, were not prescribed after admission.

Absence of clear signals and of a proper system of incen­tives or disincentives thus reduced the value of information and condemned the residual constraints to failure. To sum up from earlier, this absence could simply imply that some other signals let member countries think that the environ­ment in which they acted had lasting positive features, which induced public and private agents to believe that their current conduct was profitable to them and immune from risks, independently of their actions. In this case, sig­nals would have been noisy. Alternatively, agents were unable to correctly infer the consequences of their action or inaction. In any case, absence of signals and incentives would not imply moral hazard per se in the conduct of the agents. That would also require the existence of asymmetric information and conduct by agents that is irrational (detri­mental to the principal). Moral hazard then arises because an agent does not take into consideration the full consequences and responsibilities of his or her actions, thus acting less carefully than he or she otherwise would and leaving the principal to bear some negative effects.

Let us look at the different markets where moral hazard might have played a role. The main markets are those for goods, labour and financial assets. There is no immediate way of devising some kind of asymmetric information rele­vant for our issues in the first two markets. However, the change in institutions can act on the incentives usually exist­ing in those markets. Take the case of the Hartz IV reforms in Germany: the situation of public deficit induced the govern­ment to enact a series of reforms related to the labour market that had a positive impact on workers' incentive to accept a job and possibly also to change the terms of wage bargain­ing in a way that increased competitiveness and growth.

The bubbles that were created in deficit countries can act not only on the incentives of agents in general but also on the adverse selection of politicians and managers. As a matter of fact, they can add noise to the information available to voters and shareholders by making all the mistakes of politicians and managers ‘largely imperceptible', thus hiding mistakes and lowering the level and quality of both public and private accountability during the boom (Fernandez-Villaverde, Garicano and Santos 2013: 164).

Leaving issues of partisanship aside, when we introduce asymmetric information, separating bad from good politi­cians is very difficult, as the programme of future policy declared by each candidate before the election is always incomplete and may not correspond to his or her real inten­tions and choices. This is especially important for the case already made of some countries after their admission to the EMU. In fact, after 1999, the process of restructuring the economy of ‘peripheral' countries still had to be completed, but the prospects of continuing relatively high growth rates and the benefits from participation in the EMU were so diffuse. As a result, voters were more inclined to opt for candidates - even the less able and/or those having a special private interest in taking office - promising some relaxation of the tight policy experienced until then (Le Borgne and Lockwood 2002). Promises of soft budget con­straint to the ‘core' constituency were appealing and appeared credible and thus increased the probability of suc­cess at the polls or political survival of the ruling government (Robinson and Torvik 2009).

6.1.3 Signalling and Economic Policy after the Crisis: The Role of Rational Expectations

An additional issue arises when one assumes rational expec­tations (REs). On the one hand, this assumption takes for granted the ability of agents to correctly infer the outcomes of some facts or information. On the other, they are relevant for the incentive to cheat of policymakers. Since the work of Barro (1974), Sargent and Wallace (1975), Lucas (1976) and Kydland and Prescott (1977), forward-looking expectations have been regarded as placing severe limits on what policy­makers can achieve in a world of policy conflicts and for requiring strong policy commitments to get even that far. The Lucas critique and time inconsistency are often said to imply that announcements and loose commitments cannot be considered credible and will inevitably lead to policy failure. In the context of a recession, this would mean, in the immediate, difficulties in achieving a recovery path, creation of unsustainable debt or excessive costs for debt financing and future inflation caused by delaying the exit strategy or allowing the debt to be monetised.

These arguments, however, do not consider two possibili­ties. First, the theory we developed in Chapter 4 (in particu­lar, recall Section 4.5.5) has shown that REs do not impair policy controllability if certain conditions are met. In fact, even with REs, static controllability is guaranteed if the number of instruments equals that of targets. The true threat to controllability could arise when this condition cannot be satisfied in real life, e.g. when use of some instrument is prevented or constrained by institutions, such as for fiscal policy (see Section 5.1). Even so, dynamic controllability could be ensured if the time horizon is sufficiently long. An additional possibility to control the system is offered in the case where policymakers actively engage in managing expectations by policy announcements, along with direct policy interventions, for the express purpose of shifting the expectations path itself. In this case, REs can indeed help policymakers in succeeding in their attempts to control the economy.

One thus can clearly understand that this idea is also in the minds of the policymakers. In fact, the cases previously cited with reference to the US government, the Federal Reserve, the European Central Bank (ECB) and other important central banks have shown that policymakers do attempt to influence current economic conditions. They do so by outlining the future path of interest rates and announcing future monetary actions, new fiscal stimulus and credit guarantee packages, new bank regulations and future exit strategies and the instruments to be used. In all these cases they trust that REs by the private sector ensure the effectiveness of the announced policies.

The literature also has often used this idea in debates over the feasibility and desirability of trying to anchor inflation expectations for monetary policy changes or in arguments over the desirability of publishing interest-rate forecasts (Soderlind 1999; Woodford 2003, 2005; Blinder et al. 2008; Rudebusch and Williams 2008).

6.2