RICERCA ITALIANA     

Research program

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

University Co-ordinator

Università degli Studi di SIENA - ECONOMIA POLITICA - SIENA(SI)

Research Unit Leader

Fabrizio CORICELLI

Description

The research program of the Siena Local Unit covers two main areas, both linked to the objective of bridging the models of interaction between policy-makers and non-atomistic actors with dynamic models used as guide for policy-making. The first is the relationship between price level inertia and persistence of shocks with different market structures and labour market institutions, on one side, and policy rules on the other. The second is the test of the fiscal theory of the price level. In line with the general project, the empirical test of the FTPL will also focus on possible linkages between the monetary-fiscal policy mix and the wage dynamics.

1. Price level inertia: the role of market structures and policy rules

So far, the relationship between central bank conservatism and degree of competition in markets (both goods and labour), has been analyzed in a static framework, as if the main relevance of such interaction would be in the long run. This is perhaps the main reason why this literature has largely remained separated from the standard macroeconomic modelling widely used for guiding macroeconomic policy. In this project we take a different approach and analyze the interaction between degree of conservatism of the central bank and degree of imperfections in goods and labor markets in the context of a typical dynamic New Keynesian dynamic macro-model (Woodford, 2003), that emphasizes the short and medium run horizon for assessing effects of policies, rather than steady states. Indeed, it is not easy to justify that a relevant trade-off between output and inflation applies to the long run.
In contrast to the findings of the static models, we would like to study whether the need for a conservative central banker might increase, rather than decline, with higher competition in goods market. Indeed, from the New Keynesian literature (see Woodford, 2003), we know that higher competition in goods markets make firms more concerned about their relative prices. Following aggregate shocks, firms may thus adjust more slowly in more competitive markets. Moreover, we plan to study the role of centralization of wage bargaining on macroeconomic outcomes. The interest for these issues is that in a dynamic setting, the role of the characteristics of the central banker has to be assessed in connection with the dynamic properties of equilibria.
These include the speed of adjustment to steady states or, in other words, the degree of persistence of shocks. Such persistence crucially depends both on the characteristics of goods and labor markets and the policy rules of the central bank.
Thus, the first objective of this area of study of the Siena Unit is to analyse a model in which it is possible to highlight the inertia in the price level and the persistence of shocks in connection with: (i) degree of competition in goods markets; (ii) different labor market and bargaining structures; (iii) different policy rules.
In addition, when one considers the dynamics of both prices and wages in a Calvo-type model (Calvo, 1983)), a policy rule aimed at stabilizing consumer-price inflation and the output gap is not optimal because the dynamics of wages have also an impact on the output gap. As shown by Woodford (2003) and discussed as well by Blanchard and Giavazzi (2003), the central bank should also take into account explicitly the behaviour of wage inflation. We plan to study whether higher centralization of wage bargaining may reduce the negative effects of a policy geared to the control of consumer-price inflation, that is the dominant real life approach.
The research program includes also the empirical analysis of the issues discussed above. We plan to estimate empirically the relative role of market structure and policy rules on the degree of inertia in prices and thus persistence of shocks. Such work includes the comparative analysis of EU members, old and new. The analysis of new EU members permits to consider a wider range not only of market structure and wage bargaining arrangements but also of policy rules. Several new EU members, indeed, have until recently adopted some form of accommodative policy, based on the use of exchange rate policies driven by a real exchange rate targeting.


2. Inflation and fiscal policy
The second research area deals with the empirical test of the Fiscal Theory of the Price Level (FTPL); in line with the general project, this empirical test will be also concerned with possible links between the monetary and fiscal policy mix and wage dynamics.
We will investigate the explanatory power of the FTPL, in connection with the complex strategic interaction which occurred during the European Monetary System and now takes place in the institutional setting of the European Monetary Union (EMU), whereby many fiscal authorities are present. Dixit and Lambertini (2001) have maintained that the monetary and fiscal policy coordination can be waived, provided that both authorities share the same objectives for the inflation rate and the output level. Yet, even in this case, the output-inflation trade-off could be differently pursued, because of different weights for inflation and output gaps in their respective loss functions, or different forecasts of the inflation and output growth rates. After possible output and inflation volatility, each authority might implement interventions aimed to counteract the other authority's stance. In case of a negative supply shock, a tighter monetary policy should be compensated by an accommodating fiscal policy. In presence of a negative demand shock, a lower interest rate implemented by the central bank should be compensated by an offsetting fiscal stance. The question is that in a monetary union with many fiscal authorities coordination may be difficult to achieve: for a cooperation between the two authorities to be viable and monetary policy turning to an expansionary stance, fiscal stances must become tighter in all countries.
Further questions are to be tackled on the econometric side. Econometric tests of the FTPL presents ambiguous results (Afonso, 2002; Ballabriga and Martinez-Mongay, 2003; Cochrane, 1999; Favero and Monacelli, 2005; see also Wren-Lewis, 2001). It has been argued that a positive relationship between fiscal consolidation and increasing future surpluses finds a plain explanation in the Sargent-Wallace framework (surpluses pay back the public debt), while a correlation showing the negative sign would require a cumbersome explanation (Canzoneri et al., 2001). A major problem in econometric tests is the difficulty to establish the right causality link between surpluses and the debt dynamics. Furthermore, many authors maintain that the FTPL is untestable, as the public sector budget constraint is a relationship consisting of equilibrium data in which no causality can be detected (Christiano and Fitzgerald, 2000).
Our empirical testing of the FTPL will focus on the monetary-fiscal policy mix of the last forty-five years. When fiscal authorities act as Stackelberg leader, take into account the anticipated reactions of the monetary authorities. A fiscal expansion determines the monetary-fiscal policy mix in the sense that is operated under the expectation of a decision of the central bank to increase the interest rates. The fiscal stance then endogenises a future increase of both the short-term and the long-term interest rate, and even exchange rate variations. Contrary to Dixit and Lambertini (2001), to achieve the long-term consolidation of the stock of public debt, the coordination between monetary and fiscal policy is far from being irrelevant.

It can be observed that the switch to fiscal surpluses of the second half of the ‘80s was followed in the EMS countries by a rise in inflation, which allowed the real value of the debt stock to diminish. On the one hand, this could be a clue that a restrictive monetary policy accompanying a fiscal consolidation may determine a wealth effect on public debt (due to the rise in the interest payments) causing a jump in the price level. On the other hand, the opposite interpretation is also possible. Favero and Monacelli (2005) find that in the US a period of "fiscal dominance" (as an effect of the mismatch between the weight put on inflation in the Taylor equation of monetary authorities and the weight on output gap attributed by the fiscal authority) ended in 1987, when fiscal policy switched to consolidation. Since monetary policy started to be restrictive after 1980, a real interest rate higher than the growth rate might have caused the public budget constraint to become more binding, and the "monetary dominance" to take over.

The part of the Siena project dedicated to the empirical test of the FTPL will also focus on possible linkages between the monetary-fiscal policy mix and wage dynamics. According to the standard macroeconomic framework, to achieve monetary stability three conditions are needed: 1) an appropriate reaction function of the central bank; 2) a rule stating that public budgets must be "close to balance or in surplus"; 3) labour market flexibility, that is a bargaining between the firms and the unions such that the wage dynamics equalize on the average the productivity dynamics. We will then focus on the impact of fiscal sustainability – as analysed by the FTPL – on the macroeconomic equilibrium in Europe. In particular, we will try to investigate to what extent macroeconomic instability has to be traced to the lack of compatibility between fiscal policy and labour market institutions.

Granted that the central bank reaction function consists of a Taylor rule aimed to keep the inflation rate close to the target and the output close to potential, a trade-off can be envisaged between the behaviour of fiscal authorities and the functioning of the labour market. During a recession, as an alternative to an expansionary fiscal policy target, the absorption of excess unemployment with respect to its natural rate requires wage contracts not to fully exploit productivity increases. During an expansion, to curb inflationary tensions a fiscal surplus is needed. Alternatively, this can be obtained through a real appreciation of the exchange rate, arising from an increase in wages and prices relative to the other countries of a monetary union. In a monetary union with coordination among fiscal policies consisting of a "balanced budget rule", differently from the single-country case, the interest rate adjustment is unable to balance possible country-by-country divergences between savings and investment. Therefore, in countries where the inflation and output gaps are different from the average but single values both for the interest rate and the nominal exchange rate (like in the EMU), the macroeconomic equilibrium in each country depends on the labour market flexibility allowing for appropriate variations in the relative competitiveness. In other words, a further adjustment is in order, through an adjustment in each country's balance of payments and a variation in each country's real exchange rate, that is in the wage level relative to productivity.
In countries where the level of the common interest rate fixed by the central bank is too low (high) with respect to the width of inflation and output gaps, the wage or the productivity dynamics has to be reduced (accelerated). In the case of a fall in aggregate demand, whenever monetary policy fails in counteracting a downturn, the output slowdown keeps going on. If the commitment to convergence to a "close to balance" budget has not been fulfilled, the limit constraining the deficit / GDP ratio starts biting soon, and an excessive pressure is put on the labour market as the last resort for the overall macroeconomic adjustment.
We will follow two lines of empirical research on the behaviour of fiscal authorities. In the first one, we are interested in finding empirical evidence for the Fiscal Theory of the Price Level in European countries. On the basis of this evidence, in the second we are interested in analyzing the interplay between monetary and fiscal policy in achieving the two goals of stabilization - given the degree of flexibility of the labor market in each country - and fiscal sustainability.
2.A. The FTPL will be tested for the EU countries by using the methodology proposed by Canzoneri, Diba and Cumby (2001). In particular, we study the interaction between liabilities (the sum of public debt and monetary base) and budget surplus by means of impulse response functions. Nine EU countries are considered – Belgium, Finland, France, Germany, Italy, Netherlands, Spain, Sweden, UK – belonging to the EMU or that have participated to the ERM, for the period 1960-2002. In a Ricardian regime, the surplus at time t pays off some of the debt, and liabilities at t+1 fall. In an NonRicardian regime, there are three possibilities. First, an innovation in surplus is not correlated with the subsequent surpluses. In the case we are considering, liabilities at time t+1 should not be affected by the innovation in surplus at time t. Second, an innovation in surplus at time t is positively correlated with future surpluses and discount factors. In this case, liabilities at time t+1 should rise. In either of these cases, it should in principle be possible to differentiate between R and NR regimes. For example, the impulse response function from a VAR in surplus and liabilities would tell us how liabilities respond to an innovation in surplus. If liabilities fall, we have an R regime; if it does not, we have an NR regime.
Differently from other cases of "fiscal sustainability" (Ricciuti, 2004), preliminary results show that the series are characterized by unit roots and are co-integrated. This issue was neglected by Canzoneri, Diba and Cumby (2001), which found weak evidence in favor of stationarity. For this reason, the analysis of the impulse part of the response functions is realized by the estimate of Cointegrated VAR. Estimation of confidence bands will be obtained via Hall percentile intervals find through bootstrapping. A further step in the study is devoted to the analysis of the stability obtained through the VAR. Indeed, the VAR model assumes that there are no breaks in the relationship that is analyzed. We will develop this issue through the analysis of recursive residuals.

2.B. The second research line analyses the monetary-fiscal policy mix. Taylor (2000b) argues that a fiscal rule can be specified for the U.S. by simply relating the measure of the fiscal stance to the deviation of output from its equilibrium level. He finds evidence of a countercyclical pattern of systematic fiscal policy. Bohn (1998) argues that a century of U.S. data reveals a positive correlation between the government surplus to GDP ratio and the government debt to GDP ratio. A fiscal Taylor rule is used to examine the fiscal policy reaction with respect to the debt accumulation and the macroeconomic performance. The first step will be to estimate "Taylor fiscal rules" for a number of European countries. Since these rules do not explicitly allow for a reaction to the evolution of the government debt, the second step is to include in the estimates monetary policy measures, taking into account possible endogeneity of the variables involved. Quarterly data (1970(1)-2004(1)) for 4 EU economies (France, Germany, Italy, UK) are used. In order to bring into the picture the trade-off between the orientation of the national fiscal stance and the degree of flexibility of the labor market, also structural breaks will be adopted. The third step will be to investigate whether the real interest rate has been too high with the effect of lowering the GDP dynamics, via both smaller investments in the private sector, and government spending restraint in order to achieve fiscal sustainability.