Maximum likelihood estimates of the

Maximum likelihood estimates of the fiscal policy rule are presented in Table 5. We assume that fiscal policy follows two
states as in monetary policy and the states can be characterized as ‘active’ and ‘passive’ fiscal policy regimes. Empirical
results in Table 5 confirm the presence of two regimes in fiscal policy for all countries except for Poland and the Slovak
Republic. While the estimated coefficient of the lagged debt to GDP ratio is negative or statistically insignificant in the first
state (this result implies an active fiscal policy regime), the estimated coefficient of lagged debt to GDP ratio is positive and
statistically significant in the second state (and hence the second state is a passive fiscal policy regime) for all countries
except for the Poland and the Slovak Republic.
The estimated coefficients for the government expenditure to GDP ratio are positive and statistically significant in the
active fiscal regime for all countries. These results imply that an increase in government expenditure to GDP ratio raises
tax revenue in the active fiscal regime. Moreover, the relationship between the government expenditure-GDP ratio and
8The Owyang and Ramey (2004) model has an inflation target that involves a trade-off between inflation and unemployment, captured by a preference
parameter that follows an independent two-state Markov process reflecting periodic shifts in the natural rate of unemployment. In the ‘‘dove regime’’ the
central bank accommodates increases in the natural rate, and a ‘‘hawk regime’’ where there is less accommodation.
Table 3
Note: The figures in parentheses give the standard errors of coefficients. r (st) gives the standard error of regression for the regimes. pii indicate regime transition probabilities. d is the mean duration of regimes. P-v2 indicates the Portmanteau serial correlation test, N-v2 indicates the normality test and H-v2 indicates the heteroskedasticity test of the residuals (for more details on these tests, see Krolzig (1997)).
***Indicate statistical significance at the 1% level, respectively.
**Indicate statistical significance at the 5% level, respectively.
*Indicate statistical significance at the 10% level, respectively.
Table 4
Table 5
Note: The figures in parentheses give the standard errors of coefficients. r (st) gives the standard error of regression for the regimes. pii indicate regime
transition probabilities. d is the mean duration of regimes. P-v2 indicates the Portmanteau serial correlation test, N-v2 indicates the normality test and H-v2
indicates the heteroskedasticity test of the residuals (for more details on these tests, see Krolzig (1997)).
***Indicate statistical significance at the 1% level, respectively.
**Indicate statistical significance at the 5% level, respectively.
*Indicate statistical significance at the 10% level, respectively.
the tax revenue-GDP ratio is not statistically significant in the passive fiscal regime for the Czech Republic, Estonia and Slove-
nia. On the other hand, the estimated coefficient of the government expenditure -GDP ratio is positive and statistically sig-
nificant in the passive fiscal regime for Hungary, Poland and the Slovak Republic.
As Leeper (1991) emphasized, monetary and fiscal policy must be consistent to sustain the policy rule; as such, regime
switches between fiscal and monetary rule should be synchronized. Note that monetary policy in general in the Slovak
Republic is consistent with fiscal policy results. Even though we find both active and passive monetary policy in the Slovak
Republic, the passive monetary regime seems to be short lived (about three years). Hence one can conclude that active fiscal
policy over the sample is likely to have required active monetary policy to be short lived in the Slovak Republic.
The estimated transition probabilities in Table 5 show that the active fiscal policy regime is more persistent than the pas-
sive fiscal policy regime for all countries except for the Czech Republic. On the other hand, the passive fiscal policy regime
seems to be more persistent for the Czech Republic. The mean duration of an active fiscal policy regime varies between 1.94
(in Slovenia) and 13.00 (in Poland) quarters. Also, the passive fiscal policy regime lasts between 1.41 and 9.00 quarters. Note
that an active fiscal policy where tax revenues fall in response to increases in government debt is not necessarily unsustain-
able since the intertemporal budget constraint can still hold if the monetary authority ‘‘acts passively.’’ Monetary authority
acting passively will allow the price level to adjust appropriately so as to equate the value of outstanding government debt to
the discounted present value of future expected primary surpluses (and this is consistent with the fiscal theory of the price
Please cite this article in press as: Cevik, E.I., et al. Monetary and fiscal policy interactions: Evidence from emerging European economies.
Journal of Comparative Economics (2014), http://dx.doi.org/10.1016/j.jce.2014.05.001
Fig. 1. The smoothed transition probabilities for the first regime.
level). Debt sustainability hence requires looking at the interactions of monetary and fiscal policy and discerning policy
mixes that allow for such revenue and/or price adjustments.
We present smoothed transition probabilities for the first regime (active monetary and fiscal policy regime) obtained
from the monetary and fiscal policy rule models (equations 7 and 9 above) in Fig. 1. The smoothed transition probabilities
in Fig. 1 present a clear picture regarding the timing of regime switches of monetary and fiscal policies in each country.
According to results in Fig. 1, countries seem to have followed an active monetary policy regime at beginning of sample
which implies monetary authorities in all these countries conducted an active policy on the eve of the transition.
To investigate the policy mix and monetary fiscal policy interactions, we calculate the joint transition matrix in Eq. (11)
and the results on the timing of joint monetary-fiscal regimes are illustrated in Figs. 2–7.9
The results in Fig. 2 show that monetary policy was generally passive over the sample period. Although the policy mix
was indeterminate (passive monetary and fiscal policy) on the eve of transition, it turned into the Fiscal Theory (passive
monetary and active fiscal policy) at the end of 1996. There seems to be two periods where the policy mix is explosive using
the Leeper terminology and these periods are correlated with the crisis in the Czech Republic and the global financial crisis.
More specifically, monetary policy was always passive after 2002 except for 2009 in the Czech Republic. On the other hand,
we observe an active fiscal policy regime in the 2003–2004 and 2008–2009 periods. Note that both monetary and fiscal pol-
icies were active one in 2009 due to the global financial crisis. Finally, regime switches in monetary and fiscal policy were not
well coordinated in the Czech Republic in the 2000s.
The policy mix for Estonia in Fig. 3 shows that policy mix switched between Fiscal Theory and indeterminacy in the 2000s
in Estonia. According to transition probabilities, there was only a period in which both monetary and fiscal policy were active
which may be related to Russian crisis in 1998. After the Russian crisis, fiscal policy turned passive and policy mix was
9
We consider 50% as the threshold level for the smoothed transition probabilities to determine active and passive policy regimes in Figs. 2–7.
Please cite this article in press as: Cevik, E.I., et al. Monetary and fiscal policy interactions: Evidence from emerging European economies.
Journal of Comparative Economics (2014), http://dx.doi.org/10.1016/j.jce.2014.05.001

10
E.I. Cevik et al. / Journal of Comparative Economics xxx (2014) xxx–xxx
Fig. 2. Estimated monetary and fiscal regimes for the Czech Republic. Note: AM indicates active monetary regime, PM indicates passive monetary regime,
AF indicates active fiscal regime and PF indicates passive fiscal regime.
Fig. 3. Estimated monetary and fiscal regimes for Estonia. Note: AM indicates active monetary regime, PM indicates passive monetary regime, AF indicates
active fiscal regime and PF indicates passive fiscal regime.
Fig. 4. Estimated Monetary and Fiscal Regimes for Hungary. Note: AM indicates active monetary regime, PM indicates passive monetary regime, AF
indicates active fiscal regime and PF indicates passive fiscal regime.
Ricardian until 2000 in Estonia. Although fiscal policy was passive specifically for the periods of 2007–2008, it switched to
active one due to global financial crisis.
The policy mix seems to have alternated mainly between three policy combinations (Indeterminacy, Fiscal Theory and
Ricardian) in Hungary as can be seen in Fig. 4. Note that, monetary and fiscal policy do not seem to have been well
coordinated in Hungary over the sample. For example, Hungary seems to have followed active monetary and passive fiscal
regime on the eve of transition, after which both monetary and fiscal policy switched to a passive one (indeterminacy mix).
Please cite this article in press as: Cevik, E.I., et al. Monetary and fiscal policy interactions: Evidence from emerging European economies.
Journal of Comparative Economics (2014), http://dx.doi.org/10.1016/j.jce.2014.05.001

E.I. Cevik et al. / Journal of Comparative Economics xxx (2014) xxx–xxx
11
Thereafter the policy mix was consistent with the Fiscal Theory of the Price Level between 1999 and 2001. It was only a brief
period where the policy mix was ‘‘explosive’’ in 2004. Fiscal policy was generally active after 2007 where the policy mix
turned into Fiscal Theory of the Price Level in Hungary.
The results in Fig. 5 show that the policy mix is consistent with the Fiscal Theory (monetary policy was passive and fiscal
policy was active) at the beginning of the sample period