Vital Failures and the Second Pillar: The Ineffectiveness of Policy Action

In this section, I deal with the second type of vital objections to the discipline, i.e. those concerning the effectiveness of economic policy.^[26] These objections complement the cri­tiques implicitly dealing with the effectiveness of policy action in favour of citizens that can be derived from the personal preferences of policymakers but, contrary to them, had a long and more composite origin and have been settled less easily.

These objections are essentially due to the introduction of expectations, initially adaptive expectations, with respect to monetary policy (see Section 2.5.1) and then, in a more for­ceful and general way (i.e. referred to all policy actions), due to the assumption of REs in a context similar to that of Lucas (1976) (see Section 2.5.2). These objections - while contri­buting to the general progress of economic thought in various ways^[27] - negated the effectiveness of policy action even in the absence of agency issues and led to assertions about the low or negative value of fiscal multipliers and the need to avoid coordination of fiscal policy across countries (see Sections 2.5.3 and 2.5.4). In their weakest form, they con­strained public policy into a Nessus shirt by prescribing policy action only under a rather rigid set of rules to avoid time inconsistency and suboptimal outcomes. Stronger ver­sions of this kind of criticism denied any active role to public policy when in conflict with the conduct of private agents (see Section 2.6).

These problems have long been unaddressed, thus contri­buting in a decisive way to the decline of economic policy as an autonomous discipline. To some extent, the very practical application of the discipline in its heyday could have dis­tracted its founders and followers from further research and innovation, as was certainly the case for the Stockholm school and possibly also for the Oslo school.

2.5.1 The Role of Monetary Policy with Adaptive Expectations

In the second half of the 1960s, independent analyses by Edmund Phelps (1967) and Milton Friedman (1968) led to the conclusion that monetary policy aimed at higher employ­ment and income is ineffective in the long run, when there is no trade-off between unemployment and inflation (the long- run Phillips curve is vertical). Thus, any monetary expansion that attempts to reduce the ‘market’ rate of unemployment below the ‘natural’ one is doomed to failure and can only cause inflation. Monetary policy should pursue a target of monetary stability rather than trying to influence real vari­ables. This explains why the European Central Bank (ECB) has been assigned a monetary target (price stability) instead of a dual or multiple mandate, as in the case of the Federal Reserve. To be sure, price stability would not be the proper objective in this line of analysis because the Friedman (1969) rule must be obeyed, which should ensure (at least in a completely flexible price context) a zero nominal interest rate and a deflation rate equal to the real interest rate on safe assets. The decision taken by the ECB to choose an inflation target less than, but close to, 2 per cent - then a positive one - is only apparently contrary to the Friedman rule, as in reality prices are not completely flexible and hedonic prices may be of some importance.

The issue of the independence of central banks seems, however, more remote from Friedman’s thought. In his opinion attributing to the central bank an independent status is a second-best option, for both political and economic reasons. As far as the latter are concerned, rules are prefer­able to independence, as the main objectives of monetary policy are to avoid money itself being a factor in major dis­turbances in the system and to offer a stable background for the economy. It is true that Friedman’s objections to inde­pendence refer to a situation where the central bank is given ‘a good deal of separate power’, whereas some central banks, like the ECB, are governed by a rule (i.e.

We therefore must refer to the theoretical foundations of the rules governing central banks and their targets other than those of Friedman, even if Friedman’s (1962) argument contains the seeds of further thought justifying an indepen­dent and conservative central bank. This anticipation of further developments can be linked to one of the arguments Friedman makes use of, in particular, when, in the case of an independent central banker, he is a critic of the uncertain personality of those in charge of monetary control, who may or may not give any assurance of steady and firm conduct. This argument is a prelude to the assertion of the virtues of commitment and of a conservative central banker, which are linked either to the passage from backward- to forward­looking expectations or to political economy arguments - or to both.

2.5.2 Rational Expectations, the Neutrality Proposition and the Need for a Conservative Central Bank

Introduction of REs led, first, to a statement of the ineffective­ness of monetary policy that was even more forceful than that stated by Friedman (Sargent and Wallace 1975). Similarly, with rational forward-looking expectations, fiscal policy was considered to be ineffective as an instrument for managing income levels (Barro 1974), a result that will be considered into detail shortly. A proposition of policy neutrality or policy ‘invariance’ was thus stated concerning the most important macroeconomic policy instruments. From a more general point of view, Lucas (1976) showed that if the private sector has REs, it can fool any attempt by either the central bank or the government to pursue a given target for any real variable through the use of any instrument (generalised policy neutral­ity).

In the same vein, any promise by governments that is time inconsistent is deemed not to be credible by private agents having forward-looking expectations (Kydland and Prescott 1977). This result can be avoided by self-restraint of the policymaker, whose temptation to cheat is balanced by a fear that he or she might lose his or her reputation and no longer be able to act effectively in the case of repeated interactions with the private sector. However, in a world filled with uncertainty, signals are more difficult to interpret, and then the best practical solution to the problem of time inconsistency is that the policymaker credibly commits to some rule (Barro and Gordon 1983).

2.5.3 Ineffectiveness of Fiscal Policy: Low Value of Multipliers

In the vision of the policymakers of the 1980s and 1990s, which reflected to a large extent the economic theories that had been introduced since the end of the 1960s, problems do not come from markets, which should indeed be freed of any obstacle (or at least of regulations and obstacles deriving from the government action). They come from the discretionary action of public agents, as in each time period these tend to pursue targets that are unattainable in the presence of private agents having either backward- or forward-looking expecta­tions. In attempting to obtain their objectives, governments are fooled by the private sector, and a suboptimal outcome results. Discretionary monetary and fiscal policies are ineffec­tive with respect to real variables, and the first best desired by public agents can never be obtained. Complying with some kinds of rules can at least ensure a second-best outcome.

Setting rules to constrain monetary and fiscal policy (e.g. monetary rules, limits to budget deficits and public debts) has multiple theoretical roots. First of all, by referring to political economy contributions, one can have an analytical justification for the assumption underlying Barro and Gordon's (1983) model, according to which the government's desired unemployment rate is lower than the natural one.

The ineffectiveness of government action due to ultra­rationality (Barro 1974) or REs, the low values of multi­pliers, time inconsistency and other factors move in the same direction. Finally, within the realm of public action, the potentially negative influence of coordinated fiscal pol­icy on the price level and the capacity of monetary counter­action by an independent and conservative central bank would justify the absence of fiscal coordination in currency unions and application of the principle of subsidiarity to this matter.

Of special interest are two issues raised in the literature that have inspired recent fiscal policy attitudes towards the crisis in some countries, e.g. in Europe: on the one hand, a widespread belief in the existence of a limit beyond which an increase in public debt would have negative consequences on growth (e.g. Reinhart and Rogoff 2010; Checherita- Westphal and Rother 2010; Kumar and Woo 2010; and, long before these contributions, Modigliani 1961; Diamond 1965; Saint-Paul 1992); on the other hand, the assertion of very low (in the limit, null) or even negative spending and tax multipliers (Giavazzi and Pagano 1990).

In the following sections, we deal only with two of the above-mentioned arguments in favour of policy ineffective­ness, i.e. the low value of multipliers and the negative effects of fiscal policy coordination. At this point, I must note that one reason for policy ineffectiveness really derived precisely from the space acquired by markets - specifically financial markets - over time. This was largely an effect of both globalisation and the stimulus given to these markets by relaxation of certain constraints that had been set at Bretton Woods under the influence of Keynes' pressure in order to leave some room for autonomous national policies (Carabelli and Cedrini 2014).

2.5.4 Inefficiency and Negative SpiH-Overs from

Coordinating Fiscal Policyin a Currency Union?

The neoclassical and New Keynesian approach that incorpo­rates some sort of Barro-Ricardo (consumption-smoothing) effect - thus asserting a low value of multipliers - tends to suggest fiscal policy ineffectiveness for expansionary pur­poses and, in the case of a crisis, the need for fiscal contraction. The familiar investment saving-liquidity pre­ference money supply (IS-LM) fixed-price textbook model teaches us that an increase in public spending leads to a greater than proportional increase in private consumption and output because of the Keynesian multiplier. Looking more closely at the text, the effect is not so simple. For instance, crowding out may reduce (or even offset) the effects of government expenditure, since the higher interest rates may discourage private investors.

Assuming fixed prices, as in the IS-LM model, all changes in aggregate demand are satisfied by a passive aggregate supply. However, this kind of analysis can be extended to explicit aggregate demand and supply functions to show the effects of policy changes on wages, prices and employment, as well as on real performance and financial conditions. Often the induced changes in wages and prices reduce, if not negate, the impact of expansionary fiscal policies. The effects of government action also can be offset by the central bank, when the monetary authorities do not accommodate the fiscal expansion.

As mentioned earlier, rational agents lead to a different result: increases in public spending financed by debt may lead to a reduction in consumption (a non-Keynesian effect) and possibly in output. The effect on consumers derives from their prediction that the increase in public debt should be repaid through higher taxes in the future.

In the real business cycle realm with price flexibility, the dynamics of fiscal policy and its effects are determined by aggregate supply. If prices are fully flexible and non- distortionary taxes are assumed, a change in public con­sumption does not affect capital or labour productivity but again influences the economy only by its wealth effect. Increases in public expenditure lower permanent income because rational agents expect future taxes. They therefore reduce private consumption, and real wages fall. Private investment then increases, offsetting part of the fall in con­sumption (Baxter and King 1993).^[28]

The inclusion of price stickiness is not enough to avoid the effects described by Baxter and King (1993), which survive into the canonical New Keynesian dynamic stochastic gen­eral equilibrium (DSGE) model. Here the fall in consumption and the increase in labour supply add to labour demand. Then, if the monetary policy is not too active in responding to the production level, real wages can rise (instead of fall­ing), but not by enough to arrest the fall in consumption and thus the non-Keynesian effects of fiscal policy (Goodfriend and King 1997; Linneman and Shabert 2003).

Even in these neoclassical and New Keynesian models, separable utility, deep habits consumption, rule-of-thumb consumers and spending reversals could restore positive and significant Keynesian-like effects of public spending increases on output (Hebous 2010). In the absence of such mechanisms, some kind of Barro-Ricardo effect not only would imply the ineffectiveness of Keynesian policies but also would suggest the need for fiscal consolidation, under the form either of a reduction in expenditures or of a rise in taxes. This suggestion would be strengthened considering also the negative long-term effect of debt on growth. A positive effect of government expenditure cuts both from a short- and a long-run perspective derived from some empirical research. As mentioned earlier, this was the con­clusion of Giavazzi and Pagano (1990), who explained the positive effects on consumption of the cuts of the 1980s in Danish and Irish public expenditures as deriving from households’ expectations of permanent cuts in the level of the government budget. Along similar lines followed Alesina and Perotti (1995, 1997) and Giavazzi and Pagano (1996).

This fin de siecle credo of low or negative values of multi­pliers was certainly not in favour of traditional Keynesian fiscal action and imbalances and can be thought of as influ­encing the draft of the European Monetary Union (EMU) institutional setup (Acocella 2014a). In addition, it has also inspired the idea of an expansionary fiscal consolidation (the doctrine of ‘expansionary austerity') that has been at the basis of exit policies from the crisis - and the related idea of ‘gain without pain'⁷ from fiscal consolidation, or the so- called German view (Acocella 2015).

From this perspective, it is not strange that in Europe active fiscal policy had been put in plaster by the Stability and Growth Pact (SGP), and (more recently) the ‘fiscal compact' has been agreed on. But these limitations on discretionary fiscal policy cannot be fully understood without considering open economies explicitly, in the context of other European institutions, which I do below.

Theoretical models of open economies are of specific interest to me. In this context, the impact of budget policies on the real exchange rate plays an important role in determining the size of the multiplier effect, as this could be increased by real exchange rate depreciation. Also, other effects must be taken into account in an open economy, such as the existence of incomplete inter­national financial markets (Kollman 2010) and the possibility of a home bias in consumption (Ravn, Schmitt-Grohe and Uribe 2007), as both increase the expansionary impact of public expenditure. In an open economy context, positive spill-over effects operating via trade also have a special interest. Beetsma, Giuliodori and Klaassen (2006, 2008) and Beetsma and Giuliodori (2011) explore the international spill-overs from fiscal policy shocks in Europe. A fiscal expansion stimulates domestic activity, which leads to more foreign exports and, hence, higher foreign output. Erceg, Gust and Lopez Salido

I use the expression of Perotti, who confuted the idea (Perotti 2013), as we will see later.

(2007) and Spilimbergo et al. (2009) argue in fact that fiscal coordination increases multiplier effects.

Some of these effects are scarcely relevant in the case of fixed exchange rates, as in the EMU. In fact, changes in the real exchange rate are possible only to the extent to which the price level can be lowered in the country with an expansion­ary fiscal policy, which contradicts what one should expect to happen in a monetary union. Also, the home bias is limited in the European Union as far as the effect of national protection­ist policies is concerned, as both trade and non-trade barriers were drastically lifted. In fact, the income multiplier is reduced by the high value of the propensity to import from other EU countries. This high propensity, instead, while hav­ing a negative impact on expansionary fiscal action in one country only, would per se support a coordinated fiscal action.

This conclusion misses interactions between fiscal and monetary policies, which have an impact on the nature and value of fiscal multipliers and spill-overs. In a monetary union such as the EMU, assigning monetary authorities the primary target of price stability implies a negative spill-over of fiscal policy; in fact, any expansionary fiscal action by one country has an impact on the union's price level and thus calls for a deflationary intervention by the ECB. Beetsma and Bovenberg (1998), Beetsma and Uhlig (1999) and Beetsma, Debrun and Klaassen (2001), while using different modelling approaches, all find negative effects on income from fully coordinated fiscal expansion due to the deflationary reaction of the central bank.

According to Beetsma and Bovenberg (1998), in a monetary union such as the EMU, time inconsistency provides the rationale for a conservative central bank and against the coor­dination of national fiscal policies. In fact, in their analysis, the system suffers from both a spending bias and an inflation bias and thus faces a trade-off between them. By adjusting either monetary or fiscal institutions, not both, only a subop- timal outcome can result. Monetary unification enhances the strategic position of the monetary authority and introduces a disciplinary effect on governments. Fiscal coordination would eliminate this disciplinary effect and worsen the stra­tegic position of the central bank. The need for introducing subsidiarity in fiscal policymaking is thus asserted.

The only problem left is whether the existence of a com­mitted central bank alone and national fiscal authorities can avoid the negative effects on price stability of free-riding by the latter or whether other institutions are needed to comple­ment the type of central bank that has been chosen. Beetsma and Uhlig (1999) claim that a pact like the Stability and Growth Pact (SGP) can reduce the negative spill-overs arising from political distortions that can be exacerbated in a mone­tary union. Beetsma and Uhlig (1999) give two possible justifications for constraining the action of national fiscal pol­icy. One refers to a country, closed or open, that wants to draft a fiscal constitutional rule to tie the hands of its own govern­ment (based on the arguments developed in Section 2.3). The other lies in the existence, in a monetary union, of nega­tive spill-overs deriving from a country's budgetary policy and accumulation of debt on the common inflation rate.

The same problem - i.e. sufficiency of a committed central bank for ensuring price stability - has been investigated from another point of view. The ‘unpleasant monetarist arithmetic' of Sargent and Wallace (1981) held first the view of the insuffi­ciency of a monetary policy rule for price stability due to REs. Given these kinds of expectations, bond financing of public expenditures and tight money could give rise to immediate inflation. Along similar lines, Woodford (1996) applied the fiscal theory of the price level to the case of the EMU. In the absence of fiscal self-discipline by governments, he found that the theory supported introduction of limits to public deficit and debt as a way to complement the monetary rule chosen by the common central bank - or even to set up a precondition for this bank to be charged with maintaining price stability.

A final justification for setting limits to national fiscal policy in the context of a common monetary system was suggested by Casella (1989): a country's fiscal deficit has negative spill-overs on the interest rates and bond prices of the area and should then be limited. In order to eliminate the tendency of governments to create inflation, the inflation bias, that may be present in the constituency and the govern­ment, many possible rules have been suggested.

The school of public choice suggested introduction, in particular, of constitutional rules to be decided by following a unanimity procedure. Quasi-rational individuals could agree to limit the temptation to draw short-run benefits and agree on such subjects as balanced-budget rules, limits to governmental growth and transfers (Brennan and Buchanan 1980; Buchanan, Brennan 1981). One thus could explain both the reason why some rules adopted by the EU - such as the SGP - were required to be constitutionally grounded (Inman 1996) and the recent provision of the EU fiscal com­pact according to which constitutional rules constraining discretionary fiscal policy should be passed.

2.5.5 Summary

Then, in the 1970s and 1980s, the rationales were laid for advocating rules setting constraints to discretionary fiscal policy. These were

1. Political economy considerations about the attempts of governments to force the unemployment rate below the natural one;

2. Time inconsistency; and

3. Ineffectiveness of fiscal action, with possibly negative multipliers and effects of accumulated public debt on growth.

The SGP and the fiscal compact in the European Union and the ceilings set to the size of federal debt in the United States^[29] were the legal transposition of such statements, reflecting the idea that the true problem was of ensuring that no harm could derive from fiscal policy. In the United States, these limits were tempered by a non-conservative central bank.

In the EMU, other rules tended to prevent

1. Negative spill-overs on the real interest rates abroad; and

2. Negative spill-overs on the price level, aggravated by policy coordination and the ensuing monetary policy counter-reaction (Hughes Hallett and Acocella 2016b).

A specific consequence of the theoretical developments in the 1970s and 1980s was the introduction of independent central banks and authorities. Again, a wide array of institu­tions were implemented in the EMU as a consequence of this. I deal with this topic in Section 2.6.

2.6