RICERCA ITALIANA     

New Keynesian dynamics in unionised labour markets and the design of EU institutions

Università degli Studi di Roma "La Sapienza"

Abstract

Our general interest is the optimal design of European institutions, with special attention to the interaction between the central bank and other public and private entities.
Three main findings have assumed relevance in the literature. First, the externalities associated with wage bargaining at the national level may call for international wage coordination. Second, the result that a conservative central banker increases welfare may no longer hold in a closed economy with several unions not concerned about inflation. Third, the relationship between trade unions and policymaking in the EU requires to pay attention to social pacts and corporatism. These findings, together with the relevance of externalities and of trade-offs among different targets existing within the EMU, suggest to develop our interest by addressing the coordination problems (and the associated optimal design of institutions) which arise in the interaction between fiscal authorities, monetary authorities and trade unions.
The aim of the project is to approach this topic reconsidering the results of the literature on the interaction between imperfect (goods and labour) markets and macroeconomic policymakers (LLMI) within the New Keynesian dynamic stochastic general equilibrium models with nominal rigidities (NK-DSGE models).
The Rome local Unit will be concerned with fiscal and monetary policies in a New-Keynesian framework where non-atomistic wage setters interact with the government and the central bank. The unit will investigate the effects of fiscal coordination, neo-corporatism and social pacts, within static and dynamic models merging the literature on non-atomistic wage setters and some elements of New Keynesian models. The unit will also build a model with endogenous fiscal policy and uncertainty on central bank's preferences. The analysis will finally be extended to an open economy.
The Milan local Unit will proceed along three lines. First, it will reconsider the symbiosis of monetary and fiscal policies in a monetary union with trade unions and will investigate the impact of price stickiness on the results obtained in LLMI models. Second, starting from overlapping-generations DSGE models it will study the effects of: a) trade-unions forward-looking behaviour; b) a broad set of tax instruments; c) monetary policy. This will allow us to investigate the interaction between debt, capital accumulation and wage setting behaviour. Third, the policy implications of the results for the EMU economy will be assessed.
The Siena Local Unit will perform two tasks. First, it will build a dynamic macro-model in order to demonstrate that the need for a conservative central banker may increase, rather than decline, with higher competition in goods market. On the empirical side, the Unit will estimate the degree of price rigidity (persistence) in OECD countries. Second, it will study monetary-fiscal policy interactions within a NK GEM employing the behavioural framework of the fiscal theory of price level (FTPL), at both the theoretical and the empirical level. The main goal is to show that this interaction does not depend only on supply vis-à-vis demand shocks, but also on the impact of interest rate variations on the secondary deficit and thus on the debt dynamics. <<<

Principal Investigator

Nicola ACOCELLA Università degli Studi di ROMA "La Sapienza"

Research Objectives

The general interest of our research group is the optimal design of European institutions, with special attention to the interaction between the central bank and other public and private entities. European institutions have been so far more stability oriented than growth oriented, but the European economy is now passing through a restructuring phase and we believe it is necessary to assess the efficiency and the effectiveness of its institutions, and to evaluate the possibility of producing welfare increasing changes in their design.
When many institutions co-exist, their performance can be measured by their collective ability to reach a set of targets more quickly and/or at lower cost than alternative combinations of institutions. This requirement is particularly relevant in the presence of externalities, of significant trade-offs among different targets and of coordination problems. It is hence particularly important in a monetary union such as the EMU, where a large number of different institutions interact and coordination problems arise among the fiscal policies of different countries, between monetary and fiscal policies, among the wage setters operating in different countries, between macroeconomic policies and wage setting. Many of these problems have so far been tackled in isolation, and some aspects have received little attention in the literature, leading to incorrect or politically-biased policy recommendations, or to ill-designed agreements, such as the Stability and Growth Pact.
Macroeconomic coordination is hence a crucial but not at all a trivial issue, as several standard propositions derived from two-country models may change with the introduction of additional countries and/or additional players. For example, as shown by recent contributions, the way the central bank is designed can make fiscal coordination among authorities subject to a tight budget constraint counterproductive or ineffective (Beetsma, Bovenberg, 1998; Buti, Sapir, 1998; Acocella, Di Bartolomeo, 2001; Uhlig, 2002; Beetsma, Debrun, Klaassen, 2001; Aarle et al., 2002).
Three findings have assumed special relevance in the recent literature.
First, the externalities associated with wage bargaining at the national level may ask for tight international wage coordination (see Cukierman, 2004 for a review). With national wage setting within a monetary union, the perception of the inflationary consequences of wage bargaining for the union economy is limited and unions may tend to be more aggressive than they would be if they could internalize such costs through international collective bargaining (Cukierman, Lippi, 2002; Coricelli et al., 2004).
Second, Rogoff's (1985) well-known result that a conservative central banker can increase social welfare may no longer hold in a closed economy with several unions which do not worry about inflation per se (see Coricelli, Cukierman, Dalmazzo, 2000 and 2004; Cukierman, Lippi, 2002; see also Cubitt, 1995; Gylfason, Lindbeck, 1994; Cukierman, Lippi, 1999; Allsopp, Artis, 2003; Hancké, Soskice, 2003; Cukierman, 2004). In this case, in fact, a tight monetary stance can reduce inflation and unemployment, thus increasing the "virtues"' of a conservative central banker. However, the national trade unions' ability to cooperate within a monetary union (i.e., the characterization of the labour market institutions) should be considered as an alternative (or complementary) institutional solution. (Acocella, Di Bartolomeo, 2004a)
Third, the relationship between trade unions and economic policymaking in the EU requires paying attention also to social pacts and corporatism, interpreted as arrangements involving bargaining, collaboration and accord between important organisations and institutions, especially between unions and governments (e.g. Cubitt, 1995; Burda, 1997). A common finding of the research in this field is that such cooperation may improve macroeconomic performance (Gylfason, Lindbeck, 1994; Burda, 1997; Acocella, Ciccarone, 1995), but there is no guarantee that unions would actually wish to implement cooperative solutions, as their welfare may not increases with corporatism (Acocella, Di Bartolomeo, Tirelli, 2004b). Moreover, it has been shown that in an institutional setting characterised by unionized labour markets (but different results can be produced by fiscal policy under different modelling assumptions, as in Beetsma, Bovenberg, 1998), several fiscal authorities and an independent central bank, uncoordinated fiscal retrenchments may be ineffective, as the potential output benefits from fiscal discipline are balanced by the higher real wages set by the unions (Acocella, Di Bartolomeo, Tirelli, 2005a; 2005b).
These findings suggest to develop our research by addressing first the coordination problems (and the associated optimal design of institutions) which arise, in the European contest, in the three-way interaction between fiscal authorities, monetary authorities and trade unions.
The aim of the project is to approach this topic, at both the theoretical and the applied levels, by bringing together the two lines of research on macroeconomic policymaking which have dominated the recent debate. More in particular, our goal is to reconsider the results obtained by the literature on the interaction between imperfect (goods and labour) markets and macroeconomic policymakers (LLMI) within the general framework represented by the New Keynesian dynamic stochastic general equilibrium models with nominal rigidities (NK-DSGE models), which presently propose themselves as the main theoretical frame of reference in the hands of scholars and central banks.
We believe that the two approaches can benefit from each other and that their "contamination" may produce particularly fruitful results.
On the one hand, NK-DSGE models consider three key aspects of the economy – the interaction between fiscal and monetary policymakers through demand-side channels; the explicit modelling of capital accumulation; a truly dynamic dimension – which are absent from the LLMI models, typically characterised by a Barro-Gordon-like static framework, where demand side channels are largely ignored and the capital stock is exogenous to wage and tax distortions. On the other hand, NK-DSGE models treat labour market distortions as exogenous.
Contributions to macroeconomic theory aside, we believe this approach may find coherent applications in our research on the optimal design of macroeconomic institutions in economies characterized by unionized labour markets and non competitive goods markets, such as the EMU.
In this context, the New Keynesian approach can also be fruitfully enriched with model uncertainty and expectations in the existing relation between the central bank, on the one side, and the fiscal authorities and the private sector of the economy, on the other side. The importance of information as a fundamental strategic variable in the hands of the central bank is in fact attracting increasing interest both in the theoretical literature and in that interested in the optimal design of institutions. This is motivated by the observation that private information and expectation formation (Honkapohia, Evans, 2001; and Faust, Svensson, 2002) represent key elements to understand the transmission mechanism of monetary policy (Smets, Wouters, 2002) and the interaction between monetary and fiscal policy (Benigno, Woodford, 2003). <<<

Timescale

24 months

National and international background

The foundations of our research approach are represented by two families of models: LLMI and NK-DSGE models.

1. LLMI
This strand of literature, which developed out of the Barro-Gordon (1983) problem, brought together the debates on the optimal degree of conservativeness in central banking and those on the hump-shaped relationship between wage moderation and the degree of centralisation of wage bargaining popularised by Calmfors and Driffill. As is well known, in the presence of competitive labour markets the degree of central bank conservativeness (CBC) affects inflation but not real variables (Rogoff, 1985), whereas, if labour markets are unionized, CBC affects both inflation and real variables like unemployment, output and real wages.
Hence, in the models dealing with unionised wage setting, both CBC and the degree of centralisation of wage bargaining explain the existing differences in the macroeconomic performance of the industrialised economies, with firms playing the background role of fixing employment by moving along the demand for labour function (e.g., Hall, Franzese, 1996; Cukierman, Lippi, 2002; Guzzo, Velasco, 1999). In this class of models, the non-neutrality of money is explained by the (direct or indirect) presence of inflation, or other objectives different from real wages and employment, in the utility functions of trade unions (Acocella, Ciccarone, 1997; Acocella, Di Bartolomeo, 2004b).
In Cukierman's and Lippi's (1999) multi-union model the relationship between degree of centralisation (number of independent unions) and macroeconomic performance is of the Calmfors-Driffill type only if unions are sufficiently inflation averse. If unions are not concerned with inflation, the relationship increases monotonically (decentralised arrangements are the best ones). More in general, if the central bank controls inflation and unions are inflation averse, conservative central banks fighting inflation stimulate high real wage demands and high unemployment. This is why, if there exists only a monopoly union in the economy and society dislikes both inflation and unemployment, a populist central bank (i.e., that cares only about unemployment) should be appointed (Guzzo, Velasco, 1999; Lawler, 2000; Lippi, 2003). By contrast, if inflation does not enter the unions' objective function, conservative central banks controlling the money supply may keep wage pressure at bay by raising the fear of unemployment, bringing about lower inflation and lower unemployment (Soskice, Iversen, 1998; 2000).
The economic set-up underlying the non-cooperative games between a single central bank and possibly several unions has progressively moved away from reduced forms of the AS-AD type to micro-founded general equilibrium economies with imperfections in the labour and, possibly, in the good markets (e.g., Coricelli, Cukierman, Dalmazzo, 2000, 2001; Cukierman, Lippi, 2002; for a survey see Cukierman, 2004). In this environment, even if unions do no care about inflation, monetary policy is effective. Moreover, for realistic values of the relative aversion of unions to inflation and to unemployment, higher CBC reduces the bargaining power of unions and leads to lower values of unemployment, real wages and inflation.
The investigation on the results brought about by the strategic interaction between a single bank and several trade unions has been progressively enriched by parallel investigations on the relation between the degree of centralisation of wage bargaining and national fiscal policy coordination (Fuest, Huber, 1999) and on the interaction between fiscal and monetary policies.
As for the latter, if labour market distortions are taken as exogenous, coordination of national fiscal policies within the EMU is undesirable (Beetsma, Bovenberg, 1998), since national fiscal authorities manage to anticipate the central bank's inflationary reaction to their tax/spending inclination but, being unable to fully internalize its effects on inflation expectations, also bring about a tax/spending bias. Things however change if strategic labour unions are introduced in the monetary/fiscal policy interaction: the inverse relation between labour market and tax distortions makes national fiscal policy coordination desirable within a monetary union, as well as cooperative arrangements like social pacts of various type (Acocella, Di Bartolomeo, Tirelli, 2005a; 2005b).
If the unions are also interested in public expenditure (union members care about specific components of public expenditures, such as pension funds, training schemes or unemployment benefits), cooperation between the fiscal policymaker and the unions becomes possible (which is consistent with European experience of the last decades); outcomes may vary from the traditional exchange between wage restraint and high public expenditure to opposite results, such as a reduction in tax distortions and an increase in real wages (Acocella, Di Bartolomeo, Tirelli, 2005b). In this context, attempts to reduce fiscal distortions by imposing institutional constraints (such as the Stability and Growth Pact) may be ineffective, as the potential output benefits from fiscal discipline can be offset (if unions perceive that there exists substitutability between tax and wage distortions) by higher real wages.
These developments suggest that the degrees of coordination of wage bargaining and of fiscal policies are fundamental institutional characteristics affecting, together with the actions of the single central bank, the performance of the EMU.

2. NK-DSGE models
From the late nineties the New Keynesian models propose themselves as fundamental tools for the analysis of macroeconomic policy and, more in particular, of monetary policy. These models seem to represent today the main theoretical framework employed by most Central banks and a major source of inspiration for both theoretical and empirical studies.
The core element of this approach is the introduction of distortions in the price and/or wage determination process into Dynamic Stochastic General Equilibrium (DSGE) models. A main innovation of this strand of models is a Phillips Curve augmented with forward looking expectations, derived from households' attempt to smooth consumption over time. This literature has widely developed and the current generation of DSGE models with nominal rigidities (Clarida, Galì, Gertler, 1999; Woodford, 2003) is able to tackle several important issues, both in a closed and in an open economy, including the relationship between wage and price distortions, and structural interactions between monetary policy instruments, output and inflation consistent with the persistence of inflation and output recorded by actual data.
Several issues have recently attracted increasing attention within the New Keynesian paradigm: on the one hand, the problems related to instability, multiple equilibria and uncertainty; on the other hand, the issue of fiscal and monetary policy interactions. Notwithstanding these efforts, and even though the New Keynesian approach is becoming the main available tool to analyse macroeconomic stabilization policies, many results are still unsatisfactory and call for further work. This is particularly true for the impact of fiscal policies, its relation with output and inflation (Blanchard, Perotti, 2001), and the effects of fiscal and monetary policy interactions. Finally, NK-DSGE models introduce monopolistic competition in the labour market and conceive labour as a differentiated factor entering a CES production function, but the mark-up over the competitive wage remains exogenous. This does not favour the analysis of the effects of macroeconomic policies on labour market equilibria.
Yet the monetary-fiscal policy mix may play a special role for the stability, uniqueness and existence of rational expectation equilibria, in the determination of the price level and in influencing the effects of debt accumulation on macroeconomic performances (Canzoneri, Diba, Cumby, 2001; Benigno, Woodford, 2003; Muscatelli, Tirelli and Trecroci, 2003). Depending on the assumptions which are made, it is in fact possible to generate different paths for the business cycles which are produced not only by supply and demand shocks, but also by the impact of the interest rate variations brought about by monetary policy on the dynamics of public debt: the policy of the central bank may make it impossible for the fiscal authorities to pursue at the same time the stabilization and the consolidation objectives.
Within the literature under examination, uncertainty has been investigated according to two approaches known as Bayesian and Knightian uncertainty. Within a general Bayesian framework, Brainard (1967) was the first to document that uncertainty about model parameters may lead to cautious policy. Although "Brainard conservatism principle" (Blinder, 1998) was found consistent with the characterization of decision making at the Federal Reserve, it has been recently challenged by a number of contributions fuelling passionate controversies (Ball, 1999; Sargent, 1999; Rudebusch, 2002). The alternative approach to uncertainty is robust control under Knightian uncertainty. Within this approach, aggressiveness seems to be an inherent feature of robust policy in a closed but not in an open economy (Giannoni, 2002; Söderström, 2002). <<<