FACTA UNIVERSITATIS Series: Economics and Organization Vol. 15, N o 2, 2018, pp. 125 - 133 https://doi.org/10.22190/FUEO1802125J © 2018 by University of Niš, Serbia | Creative Commons Licence: CC BY-NC-ND Original Scientific Paper MONETARY VERSUS FISCAL DOMINANCE IN EMERGING EUROPEAN ECONOMIES 1 UDC 336.02 338.23:336.74(4) Gordana Jevđović 1 , Ivan Milenković 2 1 Guarantee Fund of AP Vojvodina, Serbia, 2 University of Novi Sad, Faculty of Economics, Subotica Abstract. The conventional macroeconomic paradigm is that monetary policy provides the nominal anchor for inflation expectations and that fiscal policy is disciplined in implementing credible and timely revenue-expenditure measures when debt rises, in order to ensure sustainability. In this scenario, monetary policy is active, whereas fiscal policy is passive, which is referred to as monetary dominance. However, the proponents of the Fiscal Theory of the Price Level emphasize that another regime may be possible – the one of fiscal dominance. In this setup, primary balance follows some arbitrary path, not necessarily compatible with the evolution of government debt, and monetary policy is faced with limited room for maneuver, as it has no option but to adjust to fiscal developments. Following these theoretical foundations, the aim of this paper is to empirically ascertain the prevailing policy regime (monetary versus fiscal dominance) in five emerging European economies (Hungary, Romania, Bulgaria, Serbia, and Macedonia). In line with expectations, results overwhelmingly suggest that monetary policy may have been subordinated to fiscal policy over the period of analysis in all economies under scrutiny and that fiscally-led regime prevailed. Key words: Fiscal Theory of the Price Level, fiscal dominance, monetary dominance, emerging European economies JEL Classification: E62, H62 Received February 27, 2018 / Accepted April 12, 2018 Corresponding author: Gordana Jevđović Guarantee Fund of AP Vojvodina, Novi Sad, Serbia E-mail: gordana.jevdjovic@gmail.com 126 G. JEVĐOVIĆ, I. MILENKOVIĆ INTRODUCTION The 2007/2008 financial crisis and the accompanying Great Recession have brought an abrupt shift in the attitude towards fiscal policy among both researchers and policymakers. Although often neglected as a macroeconomic policy tool, in the wake of the recent crisis it was once again proven that fiscal policy too can have powerful and effective transmission channels that affect aggregate demand. Arestis (2011, p. 144) argues that it was precisely fiscal policy that “saved the world from the second Great Depression and produced only the Great Recession”. Yet, this stabilization came at a price. Government measures to rescue the financial sector, provide safety nets for workers and the unemployed and stimulate the economy devastated public finances all over the world. In the aftermath of the Great Recession, and in some cases even after fiscal consolidation actions, many countries are still faced with substantial fiscal deficits and sizeable debt burdens. In light of these considerations, concerns have emerged about the broader macroeconomic consequences of fiscal developments. Often embedded within such concerns is the issue whether or not fiscal trajectories can jeopardize monetary policy objectives. Namely, under the so-called fiscally dominant regime central bank independence may not be sufficient to anchor inflation expectations. In particular, price stability cannot be guaranteed without a sustainable fiscal position of the government. Taking into account the fact that fiscal dominance is more likely to be a problem in developing economies than in advanced economies (Montiel, 2011; Zoli, 2005), the idea of this paper is to search for evidence of such dominance in five emerging European economies. The rest of this paper is organized as follows: Section 1 outlines an overview of the main theoretical arguments concerning the issue of monetary versus fiscal dominance nexus. Section 2 introduces the empirical methodology and presents our dataset. Section 3 lays out results of empirical investigation with accompanying discussion, while the last section offers some concluding remarks. 1. THE RELEVANCE OF DIFFERENT REGIMES - THEORETICAL BACKGROUND Traditional macroeconomic analysis is based on the fact that fiscal policy makers determine the level of the primary budget balance in order to ensure fiscal sustainability regardless of the price level, whereas monetary authorities are expected to set the desired price level without any constraints. Under these circumstances, the prices are determined exogenously, in the money market (for instance, following the quantitative theory of money) as opposed to the primary balance which adjusts endogenously in line with the intertemporal budget constraint. The aforementioned scenario is referred to in the literature as Ricardian or Monetary Dominant (hereinafter: MD) regime. However, as Bajo-Rubio et al. (2009, 2014) highlights, a new trend of economic thought has emerged in the 1990s - Fiscal Theory of the Price Level (hereinafter: FTPL). Sims (1994), Leeper (1991) and Cochrane (2001), have laid the FTPL groundwork and the theory was further developed for instance by Leeper (2010, 2013), Cochrane (2011) and many others. The proponents of FTPL challenge the aforementioned conventional assumptions and imply that fiscal authorities may be able to set primary balances that follow some arbitrary process, not necessarily compatible with sustainability. In such a case, the path of the primary budget balance becomes exogenously determined, while the prices now adjust endogenously Monetary versus Fiscal Dominance in Emerging European Economies 127 in order to maintain sustainability. Although monetary policy may still be able to control inflation, it is certainly far less powerful than it would be under MD regime. In fact, as observed by Sargent & Wallace (1981), the only thing that monetary authorities can control in this scenario is the timing of inflation. These propositions represent the essence of the Fiscal Theory of the Price Level and the regime based on them is called non-Ricardian or Fiscal Dominant (hereinafter: FD). The issue of whether a particular economy operates in a monetary or fiscally-led regime is determined by the behavior of monetary and fiscal authorities. In fact, as explained by Komulainen & Pirttila (2000) there are two equilibrium conditions that are critical for price level determination. The first one is well known and refers to quantitative theory of money: (1) where Mt stands for money used in the period t, Y represents the income (GDP), V is the velocity of money and Pt is the price level. The second condition, according to the same authors (Komulainen & Pirttila, 2000) is: ⁄ ∑ ( ) (2) where Wt denotes government liabilities that themselves represent the sum of the stock of net interest bearing liabilities of the government (Bt) and Mt (i.e. ), β is the discount factor and the sum of sj and τj represent the expected government primary surplus which includes both primary surplus itself (s) and the central bank transfers (seigniorage, τ). The first equation (1) is evidently the money demand function while the second equation (2) expresses the present value of the government intertemporal budget constraint. Equilibrium requires both of these equations to hold, and since we are looking at two equations with one unknown (P), the following equilibrium condition can be derived: ⁄ ∑ ⁄ (3) Decisions of economic policy makers influence the formation of public debt, primary surplus and money (Bt, sj, Mt). Consequently, fiscal (Bt, sj) and monetary (Mt) policy must be coordinated in order to achieve the desired price level. But, as emphasized by Komulainen & Pirttila (2000), two special cases of such coordination can exist. In the first scenario, fiscal policy makers react first and independently set the level of debt and deficit (Bt, sj) which requires prices to be determined from the equation that refers to the current value of the intertemporal budget constraint (Eq. 2). In this case, fiscal variables determine the price level and monetary variables adjust. On the other hand, if monetary authorities act first and determine the monetary policy variables independently, then the only thing left for fiscal policy makers is to adjust variables Bt and st to this (monetary determined) price level. Evidently, in the first case, there exists a fiscally-led regime, while the second case is referred to as the monetary dominance. Finally, it should be stated that the issue of monetary versus the fiscal dominance is closely related to the indirect way of fiscal sustainability analysis by means of Bohn’s (1998) fiscal reaction function framework (see details below). However, it is essential to emphasize that in the equilibrium, fiscal sustainability can exist in both MD and FD regime. The crucial difference is how this sustainability has been achieved, or in the words of Sargent & Wallace (1981) “who has imposed discipline on whom?”. Specifically, in the monetary dominant regime, the primary budget balance is set in order to ensure fiscal sustainability, independently 128 G. JEVĐOVIĆ, I. MILENKOVIĆ of the price level. In this case fiscal policy is subordinated to monetary policy, or as Leeper (1991) puts it - monetary policy is active whereas fiscal policy is reactive. In contrast,, in FD regime fiscal policy makers tend to set the primary budget balance exogenously to the level of government debt, and monetary policy must accommodate such fiscal path, even at the cost of excessive monetary tightening. In this scenario fiscal policy dominates over monetary policy and may undermine its goals. Furthermore, in case of fiscal dominance tighter monetary policy today can cause higher inflation in the future (or even immediately if rational agents anticipate this) – a scenario to which Sargent & Wallace (1981) refer to as the unpleasant monetarist arithmetic. 2. EMPIRICAL FRAMEWORK AND DATA According to Bajo-Rubio et al. (2014) in the game theory, the solution for the monetary versus fiscal dominance nexus would be given by the leader–follower model. In particular, it would be determined by the matter of fact which policymaker has moved first - the central bank or the fiscal authority. However, in economic practice, distinction between the two regimes is a pure empirical question centered around ways to examine whether fiscal authorities tend to set an endogenous or exogenous path for primary deficits. In addressing this issue, we follow the so-called backward-looking approach proposed by Bohn (1998) who searches for a systematic relationship between the primary surplus and one period lagged level of government debt. Yet, instead of estimating fiscal reaction function per se (Eq. 4) we are more interested in the direction of causality between the two variables in the system: (4) where PB denotes primary balance (surplus/deficit) scaled by GDP, D represents government debt (scaled by GDP), a and b are parameters and et stands for a white noise error. In a series of papers Bohn (1998, 2005, 2007) emphasizes that a positive and significant estimate of parameter b indicates fiscal sustainability as well as prevalence of monetary dominant regime over the period of analysis. However, the proponents of FTPL argue that under certain circumstances the unbiased estimate b > 0 can also indicate fiscal dominance. Accordingly, although the question of monetary versus fiscal dominance is closely related to the indirect way of fiscal sustainability investigation by means of Bohn’s (1998) fiscal reaction function, that framework may not be able to give a definite answer to the prevalence of a certain policy regime. Hence, following the recent literature on this issue (Bajo-Rubio et al. (2009, 2014), Mackiewicz-Lyziak (2015), Afonso (2017), etc.), in order to distinguish between the two regimes, we resort to Granger causality tests between primary balance and public debt. In this sense, unidirectional causality running from primary surplus to government debt would suggest fiscal dominance, while unidirectional causality running the other way around may be an indication of MD regime. The above-mentioned causal relationships are addressed by applying Toda & Yamamoto (1995) Granger non-causality test. Theoretically simple and calculationally straightforward TY causality procedure involves a Modified WALD (MWALD) test in an augmented VAR(k+dmax) model, where k refers to the optimal lag length (i.e. the correct VAR order) and dmax stands for maximal order of integration of underlying time series. The general specification of the augmented (k + dmax) VAR model for two variables is as follows: Monetary versus Fiscal Dominance in Emerging European Economies 129 (5) (6) where αj, βj, γj and δj are parameters of the model, k is the optimal number of lags in the original VAR model, dmax is the maximum order of integration of the series in the system and εt and ηt are two independent white noise errors. Our investigation is focused on five emerging European economies, namely: Hungary, Romania, Bulgaria, Serbia and Macedonia. Quarterly data on primary balance and government consolidated gross debt, both in percent of GDP (variables PB and D, respectively) cover the period from 2005Q1 to 2016Q4 and come from official sources (Eurostat in case of EU countries and national Ministries of Finance in case of Serbia and Macedonia). We use general government data for all economies, with the only exception of Macedonia in case of which similar to Trenovski & Tashevska (2015), in the absence of such data we resort to central government statistics instead. Prior to econometric modeling all the data series have been seasonally adjusted using TRAMO/SEATS method. 3. RESULTS AND DISCUSSION As the first step of the analysis, we investigate the order of integration of the primary surplus and government debt variables. To this end, we use common unit root tests such as standard Augmented Dickey Fuller (ADF) test as well as more robust DF-GLS test (Elliott et al., 1996). Complementary, we also rely on Kwiatkowski-Phillips-Schmidt- Shin (KPSS) stationarity test (Kwiatkowski et al., 1992). The null hypothesis under ADF and DF-GLS test is one of a unit root, while KPSS test has stationarity as null. As highlighted by Hatemi-J (2002) from statistical point of view, not rejecting the null hypothesis does not necessarily imply accepting it. Hence, a combination of unit root and stationarity tests seems like a reasonable procedure, since the null hypothesis in one test is the alternative hypothesis in the other. Finally, it should be noted that lag lengths in cases of ADF and DF-GLS test are determined using sequential procedure of Ng & Perron (1995), with max lag set to 6. The bandwidth for the KPSS stationarity test is based on the Newey–West estimator using the Bartlett kernel function. Table 1 Unit root and stationarity tests (variables in levels) Country Variable ADF DF-GLS KPSS Hungary PB -2.952 (1) -2.910 (1) 0.118 (4) D -1.145 (1) -1.279 (1) 0.180 (6) Romania PB -2.395 (1) -2.457 (1) 0.153 (5) D -1.309 (1) -1.145 (1) 0.117 (5) Bulgaria PB -1.860 (1) -1.975 (1) 0.193 (2) D -2.073 (1) -0.755 (1) 0.226 (5) Serbia PB -0.155 (3) -0.636 (3) 0.231 (5) D -2.798 (1) -1.407 (1) 0.185 (6) Macedonia PB -2.229 (1) -2.359 (1) 0.162 (4) D -1.813 (2) -1.673 (3) 0.204 (5) Notes: Constant and trend are included as deterministic components. At 5% significance level critical values are: -3.515 for the ADF test, -3.190 for DF-GLS test and 0.146 for the KPSS test. Lags are shown in parentheses. Source: Authors 130 G. JEVĐOVIĆ, I. MILENKOVIĆ As can be observed on Table 1 the underlying variables in levels seem to exhibit a non- stationary kind of behavior over the period of analysis. However, after first-order differencing all time series become stationary (Table 2). These findings suggest that both PB and D may be considered integrated of order one, i.e. I(1) in all economies of our sample. Table 2 Unit root and stationarity tests (variables in first differences) Country Variable ADF DF-GLS KPSS Hungary PB -8.089 (0) -7.469 (0) 0.239 (3) D -7.430 (0) -7.513 (0) 0.322 (5) Romania PB -10.886 (0) -10.233 (0) 0.077 (4) D -4.551 (0) -3.699 (0) 0.212 (4) Bulgaria PB -7.421 (0) -7.435 (0) 0.090 (1) D -3.898 (0) -2.958 (0) 0.781 (4) Serbia PB -3.312 (2) -1.748 (2) 0.657 (4) D -3.499 (0) -3.538 (0 0.463 (5) Macedonia PB -9.263 (0) -9.099 (0) 0.082 (3) D -3.452 (1) 2.098 (2) 0.453 (4) Notes: Only constant is included as deterministic component. At 5% significance level critical values are: -2.930 for ADF test, -1.948 for DF-GLS test and 0.463 for KPSS test. Lags are shown in parentheses. Source: Authors At this point it is important to emphasize that we acknowledge the fact that traditional unit root tests may not allow us to precisely differentiate between formally integrated and stationary but very persistent time series, especially keeping in mind our relatively small sample size (see also Lame et al., 2014). This further corroborates our choice of strategy for Granger causality tests, namely Toda & Yamamoto (1995) approach that is applicable irrespective of integration and cointegration properties exhibited by data. Nevertheless, the choice for the maximum order of integration (dmax) in all cases remains one (dmax=1). In line with the empirical methodology outlined earlier, the next step is to determine the appropriate lag length (k) for country-specific bivariate VAR models. For this purpose usual information criteria are employed (Akaike info criterion - AIC and Schwartz Bayesian information criterion - SIC). However, in some cases additional lags are included in order to account for autocorrelation issues. This yielded dynamically stable, well-specified VAR models, namely VAR(2) for Serbia and Macedonia, VAR(3) for Hungary and Bulgaria and VAR(4) for Romania, that represent an adequate base for undertaking the Toda & Yamamoto (1995) Granger causality testing. The results of the Toda & Yamamoto (1995) Granger non-causality tests for variables under scrutiny are presented in Table 3. As can be seen on this table, the results reflect a rather homogeneous pattern in all emerging economies under investigation. Not surprisingly, gathered empirical evidence indicates Granger causality that runs from primary surplus to government debt, while causality that goes in the opposite direction has not been found. This suggests fiscally-led regime in Romania and Serbia (at 5% significance level) as well as in Hungary and Bulgaria (at 10% level). In case of Macedonia no definite conclusion can be drawn at usual significance levels, although some (albeit weak) evidence found at the 15% significance level also confirms fiscal dominance in this country, which is in line with the earlier findings of Trenovski & Tashevska (2015). Monetary versus Fiscal Dominance in Emerging European Economies 131 Table 3 Toda & Yamamoto Granger non-causality tests Null Hypothesis k+dmax Chi-square p-value Hungary D does not Granger cause PB 3+1 1.307 0.728 PB does not Granger cause D 6.868 0.076 Romania D does not Granger cause PB 4+1 7.044 0.134 PB does not Granger cause D 18.854 0.001 Bulgaria D does not Granger cause PB 3+1 0.801 0.849 PB does not Granger cause D 6.512 0.089 Serbia D does not Granger cause PB 2+1 2.009 0.366 PB does not Granger cause D 6.510 0.039 Macedonia D does not Granger cause PB 2+1 0.296 0.863 PB does not Granger cause D 4.020 0.134 Source: Authors CONCLUDING REMARKS Following theoretical foundations of the Fiscal Theory of the Price Level, the aim of this paper was to shed some light on the issue of monetary versus fiscal dominance in five emerging European economies (Hungary, Romania, Bulgaria, Serbia, and Macedonia). Results overwhelmingly indicate that the non-Ricardian (FD) regime prevailed, which further suggests that fiscal (not monetary) policy may have been the nominal anchor for inflation expectations in economies under scrutiny. These findings strongly emphasize the need for prudent fiscal policies in all of the examined countries, even though some of them are reporting solid fiscal performance after the full effect of post-crisis consolidation measures. Previous considerations are of vital importance especially in those economies that adopted the inflation targeting regime (Hungary, Romania and Serbia) having in mind that it is well documented that fiscal dominance reduces the ability of monetary authority to effectively set policy to achieve its own objectives. Taking into account all of the above, a general conclusion would be that credible fiscal policy is absolutely essential not only in terms of strengthening the capacity of public finances and further reduction of deficit pressures, but also in a broader economic context that refers to the efficiency of monetary policy strategies, the success of the implemented exchange rate regimes and consequently the preservation of aggregate macroeconomic stability. REFERENCES Afonso, A. & Jalles, J.T. (2017). Euro area time‐varying fiscal sustainability. International Journal of Finance & Economics, 22 (3), 244-254. Arestis, P. (2011). Fiscal policy is still an effective instrument of macroeconomic policy. Panoeconomicus, 58 (2), 143- 156. Bajo-Rubio, O., Díaz-Roldan & C., Esteve, V. (2009). Deficit sustainability and inflation in EMU: An analysis from the fiscal theory of the price level. European Journal of Political Economy, 25 (4), 525–539. 132 G. JEVĐOVIĆ, I. MILENKOVIĆ Bajo-Rubio, O., Díaz-Roldan, C. & Esteve, V. (2014). Deficit sustainability, and monetary versus fiscal dominance: The case of Spain, 1850–2000. Journal of Policy Modeling, 36 (5), 924–937. Bohn, H. (1998). The behavior of U.S. public debt and deficits. Quarterly Journal of Economic, 113 (3), 949– 963. Bohn, H. (2005). The sustainability of fiscal policy in the United States (CESifo Working Paper, 1446), Center for Economic Studies and Ifo Institute Bohn, H. (2007). Are Stationarity and Cointegration Restrictions Really Necessary for the Intertemporal Budget Constraint? Journal of Monetary Economics, 54 (7), 1837-1847. Canzoneri, M., Cumby, R. & Diba, B. (2001). Is the price level determined by the needs of fiscal solvency? American Economic Review, 91 (5), 1221–1238. Cochrane, J. (2011). Understanding policy in the great recession: some unpleasant fiscal arithmetic. European Economic Review, 55 (1), 2–30. Cochrane, J. (2001). Long-term debt and optimal policy in the fiscal theory of the price level. Econometrica, 69 (1), 69–116. Elliott, G., Rothenberg, T.J. & Stock J.H. (1996). Efficient tests for an autoregressive unit root. Econometrica, 64 (4), 813–836. Hatemi-J, A. (2002). Fiscal policy in Sweden: effects of EMU criteria convergence. Economic Modelling, 19 (1), 121-136. Komulainen, T. & Pirttila, J. (2000). Fiscal Explanations for Inflation: Any Evidence from Transition Economies? (BOFIT Discussion Paper, 11), Helsinki: Bank of Finland, Institute for Economies in Transition. Kwiatkowski, D., Phillips, P., Schmidt, P. & Shin, Y. (1992). Testing the Null Hypothesis of Stationarity against the Alternative of a Unit Root: How Sure Are We that Economic Time Series Have a Unit Root? Journal of Econometrics, 54 (1-3), 159-178. Lame, G., Lequien, M. & Pionnier, P. (2014). Interpretation and limits of sustainability tests in public finance. Applied Economics, 46 (6), 616-628. Leeper, E. (1991). Equilibria under “active‟ and “passive‟ monetary and fiscal policies. Journal of Monetary Economics, 27 (1), 129–147. Leeper, E. (2010). Monetary Science, Fiscal Alchemy (NBER Working Paper 16510), Cambridge: National Bureau Of Economic Research Leeper, E.M., & Walker, T.B. (2013). Perceptions and Misperceptions of Fiscal Inflation. In: Alesina, A. & Giavazzi, F. (Eds), Fiscal Policy after the Financial Crisis (pp. 255 - 299), University of Chicago Press Mackiewicz-Lyziak, J. (2015). Fiscal Sustainability in CEE Countries – the Case of the Czech Republic, Hungary and Poland. Equilibrium. Quarterly Journal of Economics and Economic Policy, 10 (2), 53-71. Montiel, P.J. (2011), Macroeconomics in emerging markets, Cambridge: Cambridge University Press. Ng, S. & Perron, P. (1995). Unit Root Tests in ARMA Models with Data Dependent Methods for the Selection of the Truncation Lag. Journal of the American Statistical Association, 90 (429), 268-281. Sargent, T. & Wallace, N. (1981). Some unpleasant monetary arithmetic. Federal Reserve Bank of Minneapolis Quarterly Review, 5 (3), 1-17. Sims, C.A. (1994). A simple model for study of the determination of the price level and the interaction of monetary and fiscal policy. Economic Theory, 4 (3), 381–399. Toda, H. & Yamamoto, T. (1995). Statistical inference in vector autoregressions with possibly integrated processes. Journal of Econometrics, 66 (1-2), 225-250. Trenovski, B. & Tashevska, B. (2015). Fiscal or monetary dominance in a small, open economy with fixed exchange rate – the case of the Republic of Macedonia, Proceedings of Rijeka Faculty of Economics: Journal of Economics and Business, 33 (1), 125-145. Uribe, M. (2016). Is the monetarist arithmetic unpleasant? (NBER Working Paper, 22866), Cambridge: National Bureau Of Economic Research Zoli, E. (2005). How does fiscal policy affect monetary policy in emerging market countries? (BIS Working Paper, 174), Switzerland: Bank for International Settlements Monetary versus Fiscal Dominance in Emerging European Economies 133 MONETARNA VERSUS FISKALNA DOMINACIJA U EVROPSKIM EKONOMIJAMA U RAZVOJU Konvencionalno makroekonomsko stanovište podrazumeva da je monetarna politika ključna za usidravanje inflacionih očekivanja, dok je fiskalna politika disciplinovana u sprovođenju kredibilnih i blagovremenih prihodno-rashodnih mera u slučaju rasta javno duga, a kako bi osigurala održivost. U ovom scenariju monetarna politika je aktivna, dok je fiskalna politika pasivna, što se u literaturi označava kao monetarna dominacija. Međutim, zagovornici fiskalne teorije nivoa cena naglašavaju da može postojati još jedan režim – režim fiskalne dominacije. U ovoj postavci, primarni budžetski saldo prati neki arbitraran put, koji nije nužno kompatibilan sa evolucijom javnog duga, a monetarna politika se suočava sa ograničenim prostorom za manevar budući da na raspolaganju nema mnogo opcija i u krajnjoj instanci je prinuđena da se prilagodi fiskalnim kretanjima. Prateći opisan teorijski okvir, cilj ovog rada je da empirijski utvrdi preovlađujući režim monetarna versus fiskalna dominacija u pet evropskih ekonomija u razvoju Mađarska, Rumunija, Bugarska, Srbija i Makedonija . U skladu sa očekivanjima, rezultati ukazuju na to da je monetarna politika tokom perioda analize bila potčinjena fiskalnoj u svim analiziranim ekonomijama, odnosno da je preovladavao režim fiskalne dominacije. Ključne reči: fiskalna teorija nivoa cena, fiskalna dominacija, monetarna dominacija, emergentne evropske ekonomije