WORKING PAPER SERIES 6. Michal Franta, Jan Libich and Petr Stehlík: Tracking Monetary-Fiscal Interactions Across Time and Space - PDF

Description
WORKING PAPER SERIES 6 Michal Franta, Jan Libich and Petr Stehlík: Tracking Monetary-Fiscal Interactions Across Time and Space 1 WORKING PAPER SERIES Tracking Monetary-Fiscal Interactions Across Time

Please download to get full document.

View again

of 30
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information
Category:

Religious & Philosophical

Publish on:

Views: 19 | Pages: 30

Extension: PDF | Download: 0

Share
Transcript
WORKING PAPER SERIES 6 Michal Franta, Jan Libich and Petr Stehlík: Tracking Monetary-Fiscal Interactions Across Time and Space 1 WORKING PAPER SERIES Tracking Monetary-Fiscal Interactions Across Time and Space Michal Franta Jan Libich Petr Stehlík 6/1 CNB WORKING PAPER SERIES The Working Paper Series of the Czech National Bank (CNB) is intended to disseminate the results of the CNB s research projects as well as the other research activities of both the staff of the CNB and collaborating outside contributors, including invited speakers. The Series aims to present original research contributions relevant to central banks. It is refereed internationally. The referee process is managed by the CNB Research Department. The working papers are circulated to stimulate discussion. The views expressed are those of the authors and do not necessarily reflect the official views of the CNB. Distributed by the Czech National Bank. Available at Reviewed by: Tomáš Adam (Czech National Bank) Jacopo Cimadomo (European Central Bank) Nora Traum (North Carolina State University) Project Coordinator: Jan Babecký Czech National Bank, June 1 Michal Franta, Jan Libich, Petr Stehlík Tracking Monetary-Fiscal Interactions Across Time and Space Michal Franta, Jan Libich and Petr Stehlík* Abstract The fiscal position of many countries is worrying and getting worse. Should formally independent central bankers be concerned about observed fiscal excesses spilling over to monetary policy and jeopardizing price stability? To provide some insights, this paper tracks the interactions between fiscal and monetary policies in the data across time and space. It makes three main contributions. The first one is methodological: we combine two recent econometric procedures time-varying parameter vector autoregression with sign restrictions identification and discuss the advantages of this approach. The second contribution is positive: we show how monetary-fiscal interactions and other macroeconomic variables have changed over time in six industrial countries (Australia, Canada, Japan, Switzerland, the UK, and the U.S.). The third contribution is normative: the paper highlights the role of the institutional design of each policy on the outcomes of both policies. Specifically, it first offers some evidence that an explicit long-term commitment of monetary policy (a legislated numerical target for average inflation) gives the central bank stronger grounds for not accommodating debt-financed fiscal shocks. Our second set of (albeit weaker) results then indicates that this threat of a policy tug-ofwar may improve the government s incentives and fiscal outcomes reducing the probability of both a fiscal crisis and unpleasant monetarist arithmetic. JEL Codes: C1, E61. Keywords: Fiscal gap, monetary-fiscal interactions, sign restrictions, time-varying parameters VAR, unpleasant monetarist arithmetic. * Michal Franta, Czech National Bank ( Jan Libich, La Trobe University ( ); Petr Stehlík, University of West Bohemia ( ). We would like to thank Tomáš Adam, Jacopo Cimadomo, Adrian Pagan, Nora Traum, the audience at our 1 American Economic Association meetings session, and CNB seminar participants for useful comments. Financial support from the Australian Research Council (DP879638) and the Czech Science Foundation (GA4/1/146) is gratefully acknowledged. The views expressed here are those of the authors and not necessarily those of the affiliated institutions. Michal Franta, Jan Libich and Petr Stehlík Nontechnical Summary In the aftermath of the global financial crisis, several high-income countries have faced a substantial amount of fiscal stress. Most others are likely to follow in their footsteps within a decade or two, primarily due to the demographic trends of aging populations and ballooning per capita health care costs. That such a fiscal trend is undesirable and threatens economic prosperity is uncontroversial. For example, the 11 Global Risks Barometer by the World Economic Forum ranks fiscal crises as the number 1 risk in terms of the perceived financial losses (out of 37 economic, geopolitical, societal, environmental, and technological risks), perceived as very likely to occur in the next ten years. It is, however, an open question whether such fiscal stress may affect the outcomes of monetary policy, and if so, how. The unpleasant monetarist arithmetic first warned of the possible inflationary consequences of fiscal excesses. The fiscal theory of the price level identified a slightly different channel through which fiscal stress can spill over to monetary policy and jeopardize the goal of price stability. To contribute to the debate, this paper tracks the interactions between fiscal and monetary policies in the data across time and space, namely, from 198 in six advanced countries: Australia, Canada, Japan, Switzerland, the UK, and the U.S. We do so using a novel empirical approach based on a combination of two recent econometric procedures time-varying parameter vector autoregression with sign restrictions identification. Our analysis highlights the role of the institutional design of monetary policy on the behavior of the central bank when faced with excessive fiscal policy. Specifically, we demonstrate that legislating a strong monetary commitment in the form of a numerical inflation target substantially changed the bank s interest rate responses to debt-financed government spending shocks. These shocks were no longer accommodated, and in fact they were offset by higher interest rates. Intuitively, a committed central bank engaged in a tug-of-war with the government in its pursuit of low inflation. Importantly, we show that this altered the government s incentives, as marked improvements in fiscal policy towards sustainability were observed 1 3 years after the adoption of explicit inflation targeting. The implied policy conclusion is therefore that a strong commitment of monetary policy in the long term, reduced the threat of undesirably high inflation as well as increased the chances of the necessary fiscal reform. Tracking Monetary-Fiscal Interactions Across Time and Space 3 1. Introduction Many countries have been experiencing substantial fiscal stress. The responses to the global financial crisis combined with a large structural gap between government expenditures and revenues have led to rapidly growing debt-to-gdp ratios, which are forecasted to deteriorate much further due to aging populations. 1 These cyclical and structural fiscal policy developments have given rise to a new wave of discussions on whether such fiscal stress affects the conduct of monetary policy, and if so, how. Does it (eventually) spill over and lead to sub-optimally high inflation as many observers fear? Or is formal central bank independence sufficient to shelter monetary policy from such fiscal spillovers? To provide some answers, this paper uses a novel empirical framework to track fiscalmonetary interactions over time in six major countries. The fact that monetary and fiscal policies are inter-related is widely accepted. Both policies jointly affect a number of economic variables, including private agents expectations, and these in turn affect the payoffs of central bankers and government officials. In addition to the obvious channels (such as the crowding-out effect or inflationary pressures arising from excessive government spending), the seminal work of Sargent and Wallace (1981) and Leeper (1991) identified two avenues through which fiscal excesses may spill over to monetary policy. When fiscal policymakers are unable or unwilling to balance their budgets, both the unpleasant monetarist arithmetic and the fiscal theory of the price level eventually imply undesirable departures from price stability. Our game theoretic work Libich et al. (1) analyzed such strategic monetary-fiscal interactions (the policy game of chicken) and identified two main institutional variables at play. The likelihood of inflationary fiscal spillovers into monetary policy was found to decrease with the degree of long-term monetary commitment (the explicitness of the inflation target) and to increase with the degree of fiscal rigidity (the size of the fiscal gap). These variables are graphically depicted for high-income countries in Figure A1 of Appendix A, which is accompanied by a discussion of the underlying intuition. The likelihood of unpleasant monetary arithmetic was found to diminish with the monetary commitment to fiscal rigidity ratio, i.e., it is lowest in Australia and New Zealand, and highest in the United States and Japan. The presented paper attempts to assess these theoretic predictions using a novel econometric approach. We use vector autoregressions (VARs) with time-varying parameters (TVP) as introduced in Primiceri (5) and Cogley and Sargent (5). The flexibility of this approach 1 See, for example, IMF (9), which reports the net present value of the impact of aging-related spending on fiscal deficits to be in the order of hundreds of percent of GDP for advanced countries (and on average over the G countries about ten times higher than the effect of the global financial crisis). Specifically for the United States, Batini, Callegari, and Guerreiro (11) provide a recent estimate of the fiscal gap (unfunded liabilities) arguing that: 'a full elimination of the fiscal and generational imbalances would require all taxes to go up and all transfers to be cut immediately and permanently by 35 percent' (italics in the original). TVP-VARs have been used by many studies, mainly to analyze monetary policy transmission (e.g. Canova et al., 7; Benati and Surico, 8). But there have also been applications to fiscal policy (Kirchner et al., 1; Pereira and Lopes, 1), financial issues (Eickmeier et al., 11), exchange rate dynamics (Mumtaz and Sunder-Plassmann, 1), oil price shock transmission (e.g. Baumeister and Peersman, 8), and yield curve dynamics (Bianchi et al., 9). 4 Michal Franta, Jan Libich and Petr Stehlík enables us to examine medium to long-term changes in policy behavior over and above the shortrun stabilization issues explored in fixed-parameter VARs. Given the dire long-term fiscal projections, we believe that this broadened focus is warranted. It must, however, be acknowledged that the use of TVP-VARs requires a reduced number of endogenous variables and lags to keep the set of parameters manageable. In comparison with standard approaches featuring structural breaks, the TVP-VAR framework allows for structural policy changes to be gradual and differ in their timing across the two policies. As such, an analysis based on TVP-VARs can be superior to an analysis based on data sub-samples. 3 We use the framework to contrast the differences in monetary policy responses to debt-financed government spending shocks in three early inflation-targeting countries (Australia, Canada, and the United Kingdom) before and after adoption of the regime, and compare them to those in countries without a legislated numerical inflation target (Japan, Switzerland, and the United States). The methodological contribution of this paper, discussed in detail in the next section, is an extension of the TVP-VAR framework using an identification of fiscal shocks based on a combination of sign, magnitude, and contemporaneous restrictions. So far, only Kirchner et al. (1) and Pereira and Lopes (1) have employed the TVP-VAR framework to assess the effect of fiscal policy shocks. Kirchner et al. (1) focus on the euro area using the traditional recursive assumption (e.g. as in Fatás and Mihov, 1) to identify government spending shocks. Pereira and Lopes (1) examine the United States and identify the tax-net-of-transfers shock and the spending shock along the lines of Blanchard and Perotti (), who exploit institutional information on taxes and transfers to separate automatic movements of fiscal variables from fiscal shocks. While the identification approach in Kirchner et al. (1) and Pereira and Lopes (1) based on the assumption of lagged reactions among endogenous variables is suitable in some contexts, it may be too restrictive for the analysis of monetary-fiscal interactions. This is because it implies that either the monetary authority does not react contemporaneously to fiscal shocks, or the fiscal authority neglects contemporary movements in monetary policy. Intuitively, such specification may implicitly impose unrealistic timing assumptions about the interaction between the monetary and fiscal authorities. As the game theoretic examination of monetary-fiscal interactions dating back to Sargent and Wallace (1981) suggests, the exact timing of policy moves is a crucial determinant of the outcomes of both policies. Similarly, Caldara and Kamps (8) show that different identification approaches can lead to qualitatively different results in terms of monetary policy responses to government spending shocks. Therefore, an additional advantage of using the sign restrictions framework in the policy context is that no timing assumptions on the monetary-fiscal interaction need to be imposed. On the other hand, sign restrictions are a weak identification approach in terms of there being many structural models that correspond to the estimated reduced-form model and satisfy the signs imposed on the 3 It is well established that many advanced countries have experienced structural breaks in monetary and fiscal policy, with their policy regimes changing over time see, for example, Davig and Leeper (1) and Clarida, Gali, and Gertler (1998). Fiscal policy analyses based on sub-samples can be found in Pappa (1), Perotti (7), and Blanchard and Perotti (). Tracking Monetary-Fiscal Interactions Across Time and Space 5 impulse responses (Fry and Pagan, 11). We mitigate this potential problem by adding a set of contemporaneous and magnitude restrictions. Our analysis offers several insights regarding the monetary-fiscal interaction: how it changed over time, how it differed across countries, and how the institutional design of the policies may explain the changes and differences. In particular, it is shown that in the inflation-targeting countries considered, the degree of monetary policy accommodation of debt-financed fiscal shocks indeed decreased after the adoption of a numerical inflation target. In contrast, in the non-targeters the degree of accommodation over the same period did not change much, or, most notably in the United States, increased. Importantly, the inflation-targeting countries have not only improved their monetary outcomes. With a delay of 1 3 years after the adoption of the regime, their fiscal outcomes started improving as well, and remained in good shape until at least the recent crisis. These findings are consistent with the game theoretic predictions of Libich et al. (1) that a long-term monetary commitment may help the central bank discipline governments (induce fiscal reforms) through a credible threat of a policy tug-of-war. This may explain the negative correlation in Figure A1: institutional reforms increasing long-term monetary commitment (moving a country to the right) may also induce a reduction in fiscal rigidity (a movement down). However, one needs to be careful in drawing conclusions about causality between stronger monetary commitment and improved fiscal outcomes our evidence is limited to correlation. The tentative policy recommendation is therefore as follows: to get an upper hand in the policy game of chicken, central banks should try to commit as explicitly as possible to their long-term inflation objective. 4 The fact that the Federal Open Market Committee has subscribed more explicitly to the % long-term inflation target is consistent with our recommendation. The committee s justification also seems to point to the channels examined in our paper: Communicating this inflation goal clearly to the public helps keep longer-term inflation expectations firmly anchored. Our analysis offers additional results, most importantly regarding the size of output and private consumption multipliers and how these evolved over time. Due to space constraints we will cover these results in detail in a separate paper.. Identification Three approaches to the identification of fiscal policy shocks have been established in the literature. First, the event-study approach (Ramey and Shapiro, 1998) focuses on describing the effects of an unexpected increase in government defense spending. Second, the structural VAR approach (Blanchard and Perotti, ) draws on the assumption of a lagged reaction of fiscal variables to changes in economic conditions. Third, the identification scheme based on sign restrictions, developed originally for the analysis of monetary policy shocks, has been applied to fiscal policy (Mountford and Uhlig, 9; Pappa, 9; Canova and Pappa, 7). Recently, the 4 It should be stressed, however, that since the target is specified as a long-term objective achievable on average over the business cycle, it does not seem to reduce short-run policy stabilization flexibility: for recent evidence see e.g. Kuttner and Posen (11) and for theoretic modeling see Libich (11). 6 Michal Franta, Jan Libich and Petr Stehlík sign restrictions identification approach has been enriched by additional identifying assumptions based on, for example, cointegration (Dungey and Fry, 9) and magnitude restrictions (Hur, 11). Our identification procedure complements sign restrictions with magnitude and contemporaneous restrictions, building on Franta (11). Our focus is on the identification of a debt-financed government spending shock. Government spending is defined as government consumption and investment, i.e., total expenditures excluding government transfers. Similarly to Canova and Pappa (7), Pappa (9), and Dungey and Fry (9) we assume that a positive debt-financed government spending shock increases: (i) government spending for four quarters, (ii) government debt for four quarters, and (iii) output for two quarters. The length of the imposed sign restrictions is related to some aspects of the data, which we discuss in Section 4. As shown in Pappa (9) such restrictions, at least on impact, are consistent with standard structural models of both the real business cycle and the New Keynesian tradition, and they do not result from productivity, labor supply or monetary shocks. 5 A rise in output and government debt can, however, also be brought about by a tax cut and/or an increase in transfers. Therefore, to filter out the effects of government transfer and tax shocks, we impose a magnitude restriction that an identified debt-financed spending shock does not increase government debt by more than the amount of government spending. 6 The situation where tax cuts imply an increase of tax revenues cannot be distinguished from a government spending shock within our identification framework, but such a scenario is arguably unlikely. Next, to capture the fact that government purchases do not react much to the business cycle, we impose a zero contemporaneous restriction on the effect of a business cycle shock on government spending. This is reminiscent of the recursive identification of shocks when government spending is ordered before GDP. Nevertheless, we do not restrict the contemporaneous feedback between government debt and output to allow for the effect of automatic stabilizers on the fiscal variables (taxes/debt). The contemporaneous restriction on the relationship between output and government spending enables us to distinguish between a generic business cycle shock (Mountford and Uhlig, 9) and fiscal shocks. As shown by Wouters (5) a higher number of shocks identified implies greater reliability of the sign identification procedure. Finally, let us stress that we do not impose any restriction on the interest rate because it is our main variable of interest, summarizing the responses of monetary policy to debt-finan
Related Search
Similar documents
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks