1
How does EMU membership affect the performance of
European countries on macroeconomic stability?
Huub Lubbers
*Master thesis
Supervisor: P. Milionis
Abstract
This paper investigates how the performance of European countries on macroeconomic stability is affected by EMU membership. As measures for macroeconomic stability inflation and growth volatility are considered. The difference in performance is assessed using panel data regressions using country and time fixed effects. The author finds no significant effect of EMU membership on macroeconomic stability for the average country. However, for countries with relatively closed economies, relatively low investments and or a relatively small government EMU membership lowers inflation and growth volatility.
Key words: EMU, Macroeconomic Stability, Panel Regressions
JEL Code: E31 E42 E52
2
I. Introduction
The European Monetary Union (EMU) is known as the most integrated economic union in the world. Starting from eleven member states in 1999 the Eurozone now consists of nineteen European countries. Several economic advantages of the EMU have driven the current member states on their decision to join. Cecchetti et al. (2002) mention two benefits. First, countries gain from increased credibility of the institutions underlying the EMU. As central banks of other member states have been successful in the past in achieving their targets, take for example the Bundesbank of Germany. Furthermore, member states enjoy the economies of scale of the integrated financial market. In addition, De Grauwe (2016) argues that by joining the EMU countries benefit from the eliminated foreign exchange risk with some of their European trade partners. Consequently, the increased price transparency between these countries could enhance competition which in turn fosters economic activity. However, the Eurozone does not only offer benefits for the member states. As the EMU is an incomplete monetary union there are fragilities to which member states are prone to. De Grauwe (2016) discusses two. For one thing, governments of EMU member states issue public debt in euros. As a result Eurozone countries have public debt outstanding in a currency over which they cannot exert control. Therefore, governments of EMU countries cannot guarantee repayment of bonds issued and can thereby be forced into default by speculation of investors. On the other hand, one could also argue that this feature of the EMU enhances fiscal discipline. Second, by joining the EMU a country loses the monetary policy tool as an instrument to steer the economy. Key interest rates as the marginal lending facility and the deposit facility rate are set by the European Central Bank (ECB) and are therefore removed from national control.
3 frameworks. Examples of these nationalist parties are the National Front in France, the Party for the Freedom in the Netherlands and the Five Star Movement in Italy. Up till now, none of these populist parties have delivered a president or prime minister yet, however the nationalistic movement is growing in Europe.
The research question will be investigated using panel regressions including 29 countries with time series data ranging from 1995 up to 2015. All regressions will include country and time fixed effects. First, the effects of entering the EMU are examined using a baseline specification. The baseline specification allows to examine whether being an EMU member has an effect on macroeconomic stability in general. Next, regressions including interaction terms investigate whether or not the effects on macroeconomic stability of being and EMU member are conditional on some country specific characteristics. The characteristics that will be considered are three indicators of financial intermediary development, the investment share of GDP, the government share of GDP and trade openness. Finally, a general specification will be considered allowing to check for the robustness of the previously found results.
For average countries, no significant effect of EMU membership has been found. However, the results do show a scope wherein being an EMU member is beneficial for macroeconomic stability. First, low investment countries tend to benefit from being in the EMU concerning both inflation and growth volatility. Second, relatively closed economies tend to be better off in the EMU regarding inflation volatility. Furthermore, countries with a relatively small government are found to experience lower growth volatility. Weaker results are found suggesting that entering the EMU is beneficial for countries with a large financial sector. Overall, the results boil to down the conclusion that countries which are relatively less developed compared to the EMU average are there to gain the most from entering the Eurozone.
4
II. Literature Review
Inflation Volatility
The academic debate on inflation volatility goes back to the Nobel lecture of Milton Friedman in 1977. During his lecture he argued that average inflation correlated positively with inflation volatility. Later empirical research has confirmed this argument as the correlation between inflation volatility and average inflation has been found to be robust (Aisen & Vega 2007). As volatile inflation in turn has been found to be negatively correlated with economic growth (Judson & Orphanides 1999) several researches have been done on factors lowering average inflation. Hence, the findings suggest a tripartite relation between average inflation, price stability and economic growth.
An influential paper in the field covering average inflation was the paper of Cukierman (1992). He found that a credible and independent central bank helps to achieve lower inflation. Moreover, a broad range of empirical research found a negative correlation between central bank independence and average inflation. Examples are Klomp & De Haan (2010), Alesina & Summers (1993) and Fischer (1995). Their key theoretical argument is that an independent central bank will not value short-run economic expansions over inflationary pressures in the long-run. Whereas politicians might favor a short-run expansion due to electoral pressures. Consequently, assuming rational expectations, the independence of the central bank will cause agents to lower their inflation expectations thereby decreasing actual inflation. In the literature this phenomenon is commonly referred to as the time inconsistency problem. In addition, Bodea & Hicks (2015) add that an independent central bank will be more successful in democratic countries. They argue that democracies tend to have better law enforcement and more government transparency causing an independent central bank to be more successful in maintaining the inflation target.
5 inflationary preferences. As a result, average inflation would be lower which in turn causes the volatility of inflation to be lower.
All the previous findings argue in favor of an inflation targeting independent central bank. This monetarist paradigm can be found back in the mandate of the ECB described in the Maastricht Treaty. The ECB has a single mandate stating that the primary target of the ECB is to keep inflation close but below to two percent. Therefore, member states of the EMU are expected to experience relatively low inflation levels resulting from this institutional framework.
Besides the importance of institutions there are also country specific characteristics that could affect inflation volatility. Bowdler & Malik (2005) have investigated the relation between trade openness and the volatility of inflation. They list two arguments proposing that trade openness reduces inflation volatility. Their first argument relates to seigniorage collection by governments of countries experiencing high fluctuations in economic activity. The costs of seigniorage collection result from increased volatility in the growth rate of the money base which in turn increases the volatility of the inflation rate. For open economies this cost of increased inflation volatility is higher when domestic firms face forceful competition from international competitors. Main issue is that their competitiveness falls when faced with higher fluctuations in nominal revenues resulting from the higher volatility of inflation. Important to note is that this argument does not hold for Eurozone countries as they cannot exert control over the monetary base of the Euro. However, this argument may be relevant for European countries outside of the Eurozone. Second, they argue that countries will be active in higher value added products as their trade terms are more open. This results in less volatile inflation as these products tend to have more stable terms of trade. On the other hand, increased openness also exposes countries to more foreign exchange risk. Especially countries operating with a flexible exchange rate are prone to foreign price fluctuations.
6 Roussea and Wachtel (2005) provide the argument that an inflationary environment depresses the development of the financial sector. High and volatile inflation lowers the incentives of financial intermediaries to be involved in long-term contracts. As a result capital formation and economic growth will be negatively affected as firms have a hard time in obtaining longer term funding. Therefore, the argument suggests a negative relation between inflation volatility and financial intermediary development.
Finally, Campolmi and Faia (2011) relate inflation volatility in EMU countries to labor market frictions. As unemployment benefits and other regulations are determined on national levels, it is possible for differences to arise between the countries member of the EMU. Important assumption made for this argument is that labor mobility is low. Hence, workers from for example Austria will not easily migrate to France even if labor market conditions are such that the Austrians workers should do so. Nevertheless, their research showed that more protection for the employed or less protection for the unemployed leads to higher wage differentials and therefore to increased inflation volatility. Reasoning behind this that firms tend to alter the wages of their employees more as laying employees off is more costly when employment protection is higher. The increased fluctuation in wages then causes an increase in the volatility of inflation.
Growth Volatility
7 Summers mentions is improved inventory management caused by developments in information technology. In more detail, the improved inventory management created the situation wherein firms were able to keep the production of durable goods stable over economic upswings and downturns. As firms were able to insulate their production levels from business cycle fluctuations economic growth could be more stable. Finally, good fortune could be an explanation for the great moderation as the frequency of large shocks was lower. For example, oil prices experienced less shocks during the period of the great moderation (Nakov & Pescatori 2010).
Similar to price volatility, growth volatility may also be influenced by country specific characteristics. For instance, financial intermediary development could be important for stable economic growth. Beck et al. (2006) have found that well developed financial intermediaries foster economic growth. However, they did not find a direct effect suggesting financial intermediary development reduces growth volatility. Nevertheless, the development of financial intermediaries may be of huge importance for Eurozone countries. Especially when asymmetric economic shocks occur over countries as argued by De Grauwe (2016). If this case occurs, the supranational central bank is paralyzed in taking policy measures as the shock does not affect the countries similar area-wide. Consequently, the financial sector becomes the main channel trough which funds can be redistributed over countries in order to smoothen the fluctuations in output.
8 measures in case of area-wide symmetric economic shocks which is the most likely case in a monetary union according to the European Commission view. In contrast, there is the Krugman view which argues that intensified trade between countries causes these countries to specialize in certain industries because of economies of scale. Hence, the Krugman view suggests that trade intensification fosters inter-industry trade. As a result, large adverse shocks will have asymmetric effects for the countries in consideration. As argued before, the supranational central bank is unable to impose measures in this case. Therefore, according to the Krugman view, Eurozone countries are expected to experience relatively unstable economic growth if the economies of these countries are more open.
Finally, a remark has to be made concerning the advocacy of inflation targeting caused by the findings of the literature previously discussed. An important assumption in the argument that lower inflation induces lower inflation volatility and that both in turn decrease growth volatility is that inflation and growth move parallel. Hence, the inflation targeting central bank faces demand shocks. However, as argued by Cecchetti & Ehrmann (1999), De Grauwe (2010) and Rogoff (1985) the inflation targeting central bank faces a more complicated decision in the case of a shock at the supply side of the economy. As with supply side shocks the price level and the output level move in opposite directions and therefore the policy maker faces a tradeoff. For instance, if a positive shock occurs at the supply side of the economy, this will result in a fall of the price level whereas output will rise. If the central bank acts solely on the lower price level, it will impose a monetary expansion in order to keep inflation at the target level. However, the central bank now fuels the growth boom caused initially by the supply shock. As a consequence, the economy now could overheat experiencing economic growth above natural levels resulting from the low interest rates set by the central bank. In the end, this excessive growth has to be adjusted for by an economic downturn or even a recession. Henceforth, in this case the central bank has, presumably unintentionally, enhanced the business cycle fluctuations causing an increase in the volatility of economic growth.
9 1 1. 2 0. 2 0. 6 0. 4 0. 8
In
fla
tio
n
v
o
la
tilit
y
1995 2000 2005 2010 2015Year
France Germany Italy Spain United KingdomFigure 1 Standard deviation of inflation
points in time fixed effects estimation may be applied since the EMU dummy variable will not be time invariant.
III. Data Description
For the empirical analysis several time series of macroeconomic variables are gathered. First the derivation of the data series for inflation and growth volatility will be discussed. Next, the main explanatory variables and other control variables will be described. The selection of countries to include is based upon data availability; all countries included are listed in appendix A.
10 0. 02 0. 06 0. 08 0. 04
G
ro
wt
h
v
o
la
tilit
y
1995 2000 2005 2010 2015Year
France Germany Italy Spain United KingdomFigure 2 Standard deviation of growth
period has been left out. The HICP data was obtained from Eurostat. Plots of the inflation volatility levels of France, Germany, Italy, Spain and the United Kingdom are illustrated in Figure 1.
The measure for growth volatility is derived in a similar way. The only difference compared to the measure of inflation volatility is that quarterly growth rates were used instead of monthly. The growth rates were derived from GDP data reported in current prices. Furthermore, the GDP data was obtained from Eurostat and is not calendar adjusted or seasonally adjusted. Figure 2 shows the volatility of growth for France, Germany, Italy, Spain and the United Kingdom.
11 Next, to investigate whether or not Eurozone countries perform differently compared to European non-Eurozone countries an EMU dummy variable is constructed. The EMU dummy variable takes a value of 1 if a country is a member of the Eurozone in that particular year and takes a value of 0 if a country is not a member of the EMU. As all entries by countries into the EMU have been on the first of January, there are no issues of countries being member of the EMU for only a part of a year.
Finally, in order to examine whether or not the effects of entering the Eurozone are heterogeneous over countries data series describing country specific characteristics are taken into account. First of all, three indicators of financial intermediary development are considered. In more detail, these indicators are commercial bank assets divided by commercial bank assets plus central bank assets (CA/TA), liquid liabilities divided by GDP (LL/GDP) and private credit divided by GDP (PC/GDP). The data series contain yearly data and are obtained from the International Financial Statistics database. Levine et al. (2010) discuss the properties of these indicators neatly. They argue that CA/TA measures how much of the allocation of savings is done by commercial banks instead of the central bank. Note that the lower the value of CA/TA is, the more dominant the central bank is in the financial market. Concerning LL/GDP they remark that this indicator may be subject to double counting as liquid liabilities equals currency plus demand and interest-bearing liabilities of banks and nonbank financial intermediaries. Therefore, interbank deposits are included in this measure. However, they still see LL/GDP as a good measure of financial sector size. Last, they argue that PC/GDP is their preferred indicator as PC/GDP solely measures credits issued to private sector parties. All three financial indicators are three year averaged in order to correspond with the measures of inflation and growth volatility.
12 representative to the measures of inflation and growth volatility. Table 1 reports the descriptive statistics of the variables used in the empirical analysis.
IV. Research Methodology
The effects of the EMU dummy are investigated using panel regressions including country and time fixed effects. The use of country fixed effects is warranted as countries cannot be sampled randomly. Time fixed effects are used to control for economic events happened during the sample period possibly affecting inflation and or growth volatility. For example the financial crisis in 2008 calls for the use of time fixed effects. Inflation and growth volatility will be examined using three specifications. The first specification will be quite basic including only baseline controls, the EMU dummy variable and one of the additional control variables CA/TA, LL/GDP, PC/GDP, investment share of GDP, government share of GDP or trade openness. This specification will therefore be referred to as the baseline specification.
Table 1 Descriptive statistics control variables
Variables Mean Minimum Maximum Std. Deviation
13 For inflation volatility, the baseline controls are average inflation and the natural logarithm of GDP per capita. These baseline controls ensure that the estimated coefficient of the EMU dummy variable will not be inflated due to high differences in inflation caused by differences in institutions. Hence, the baseline controls allow for a fair comparison between EMU countries and non-EMU countries. For growth volatility, the baseline controls are average growth and the natural logarithm of GDP per capita. Similarly, these baseline controls allow for a fair comparison of the countries. In the second specification heterogeneity of the effect of EMU membership over countries will be considered using interaction terms of the additional control variables also used in the baseline specification. Investigating the possible heterogeneity of the EMU dummy variable is of interest as the literature suggests several country specific characteristics affecting inflation and or growth volatility. Finally, the third specification will be the most general as it will include the share of government expenditures of GDP, the share of investments as a percentage of GDP and trade openness as controls. This general specification allows for a robustness check of the estimates of the previous estimations of the interaction terms and other control variables.
The first specifications will take the following mathematical expressions:
𝑉𝑉𝑖𝑖𝑖𝑖𝜋𝜋 = 𝛼𝛼1𝜋𝜋𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 + 𝛼𝛼2𝐿𝐿𝐿𝐿𝐿𝐿�𝑌𝑌𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎� + 𝛼𝛼3𝐷𝐷𝑖𝑖𝑖𝑖𝐸𝐸𝐸𝐸𝐸𝐸 + 𝛼𝛼4𝑋𝑋𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 + 𝜇𝜇𝑖𝑖 + 𝜏𝜏𝑖𝑖+ 𝜀𝜀𝑖𝑖𝑖𝑖 (1) 𝑉𝑉𝑖𝑖𝑖𝑖𝐺𝐺 = 𝛽𝛽1𝐺𝐺𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎+ 𝛽𝛽2𝐿𝐿𝐿𝐿𝐿𝐿�𝑌𝑌𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎� + 𝛽𝛽3𝐷𝐷𝑖𝑖𝑖𝑖𝐸𝐸𝐸𝐸𝐸𝐸 + 𝛽𝛽4𝑋𝑋𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎+ 𝜇𝜇𝑖𝑖 + 𝜏𝜏𝑖𝑖+ 𝜀𝜀𝑖𝑖𝑖𝑖 (2) Where in equation (1) and (2) 𝑌𝑌𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 is GDP per capita for country i in year t. D is the EMU dummy variable, 𝑋𝑋𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 represents the additional control variable used for country i
at time t, μ is the country fixed effect and τ is the time fixed effect. In equation (1) 𝑉𝑉𝑖𝑖𝑖𝑖𝜋𝜋 is the standard deviation of inflation in country i in year t and 𝜋𝜋𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 stands for the average level of inflation in country i in year t. Next, in equation (2) 𝑉𝑉𝑖𝑖𝑖𝑖𝐺𝐺 is the standard deviation of growth in country i in year t and 𝐺𝐺𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 is the average growth rate of country i in year t. Finally, α1,
α2, α3 and α4 are the coefficients to be estimated in equation (1) and β1, β2, β3 and β4 are the
14 Now, the second specifications, baseline plus interaction terms, will be an expansion of the baseline specifications with an interaction term of one of the previously listed control variables and will therefore have the following mathematical expressions:
𝑉𝑉𝑖𝑖𝑖𝑖𝜋𝜋 = 𝛾𝛾1𝜋𝜋𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎+ 𝛾𝛾2𝐿𝐿𝐿𝐿𝐿𝐿�𝑌𝑌𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎� + 𝛾𝛾3𝐷𝐷𝑖𝑖𝑖𝑖𝐸𝐸𝐸𝐸𝐸𝐸 + 𝛾𝛾4𝑋𝑋𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎+ 𝛾𝛾5𝐷𝐷𝑖𝑖𝑖𝑖𝐸𝐸𝐸𝐸𝐸𝐸∗ 𝑋𝑋𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 +
𝜇𝜇𝑖𝑖 + 𝜏𝜏𝑖𝑖+ 𝜀𝜀𝑖𝑖𝑖𝑖 (3)
𝑉𝑉𝑖𝑖𝑖𝑖𝐺𝐺 = 𝛿𝛿1𝐺𝐺𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎+ 𝛿𝛿2𝐿𝐿𝐿𝐿𝐿𝐿�𝑌𝑌𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎� + 𝛿𝛿3𝐷𝐷𝑖𝑖𝑖𝑖𝐸𝐸𝐸𝐸𝐸𝐸 + 𝛿𝛿4𝑋𝑋𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎+ 𝛿𝛿5𝐷𝐷𝑖𝑖𝑖𝑖𝐸𝐸𝐸𝐸𝐸𝐸 ∗ 𝑋𝑋𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎+
𝜇𝜇𝑖𝑖 + 𝜏𝜏𝑖𝑖+ 𝜀𝜀𝑖𝑖𝑖𝑖 (4)
Added in these specifications is the term and 𝐷𝐷𝑖𝑖𝑖𝑖𝐸𝐸𝐸𝐸𝐸𝐸 ∗ 𝑋𝑋𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 which represents the interaction term. By introducing the interaction term the effect of being in the Eurozone can be examined for different values of the variables describing country specific characteristics. Last, 𝛾𝛾1, 𝛾𝛾2, 𝛾𝛾3, 𝛾𝛾4 and 𝛾𝛾5 are the coefficients to be estimated in equation (3) and 𝛿𝛿1, 𝛿𝛿2, 𝛿𝛿3, 𝛿𝛿4and 𝛿𝛿5 are the coefficients to be estimated in equation (4).
The last specifications are the most general and include the share of investments of GDP, the share of government expenditures of GDP and trade openness as control variables in the specification for inflation volatility and growth volatility. The general specifications therefore have the following mathematical expressions.
𝑉𝑉𝑖𝑖𝑖𝑖𝜋𝜋 = 𝜃𝜃
1𝜋𝜋𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎+ 𝜃𝜃2𝐿𝐿𝐿𝐿𝐿𝐿�𝑌𝑌𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎� + 𝜃𝜃3𝐷𝐷𝑖𝑖𝑖𝑖𝐸𝐸𝐸𝐸𝐸𝐸 + 𝜃𝜃4𝑋𝑋𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎+ 𝜃𝜃5𝐼𝐼𝐼𝐼𝑉𝑉𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 + 𝜃𝜃6𝐺𝐺𝐺𝐺𝑉𝑉𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎+ 𝜃𝜃7𝑇𝑇𝐺𝐺𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎+ 𝜃𝜃8𝐷𝐷𝑖𝑖𝑖𝑖𝐸𝐸𝐸𝐸𝐸𝐸∗ 𝑋𝑋𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 + 𝜇𝜇𝑖𝑖 + 𝜏𝜏𝑖𝑖+ 𝜀𝜀𝑖𝑖𝑖𝑖 (5) 𝑉𝑉𝑖𝑖𝑖𝑖𝐺𝐺 = 𝜌𝜌1𝐺𝐺𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎+ 𝜌𝜌2𝐿𝐿𝐿𝐿𝐿𝐿�𝑌𝑌𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎� + 𝜌𝜌3𝐷𝐷𝑖𝑖𝑖𝑖𝐸𝐸𝐸𝐸𝐸𝐸 + 𝜌𝜌4𝑋𝑋𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 + 𝜌𝜌5𝐼𝐼𝐼𝐼𝑉𝑉𝑖𝑖𝑖𝑖𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 +
15 trade openness is considered in the interaction term. This holds for both equations (5) and (6). The coefficients to be estimated in equation (5) are 𝜃𝜃1, 𝜃𝜃2, 𝜃𝜃3, 𝜃𝜃4, 𝜃𝜃5, 𝜃𝜃6, 𝜃𝜃7 and 𝜃𝜃8. Finally, 𝜌𝜌1, 𝜌𝜌2, 𝜌𝜌3, 𝜌𝜌4, 𝜌𝜌5, 𝜌𝜌6, 𝜌𝜌7 and 𝜌𝜌8 are the coefficients to be estimated in equation (6).
Possible econometric issues
As the investment share of GDP may be endogenously related to inflation volatility the estimates of relative investment may be biased. Therefore, the quantity of the estimated coefficients of relative investment should be interpreted with caution. Furthermore, there may be some heteroskedasticity over the standard errors as for example high values of trade openness or low values of CA/TA are quite uncommon. Finally, there may be some multicollinearity issues as the explanatory variables could be correlated which could inflate the standard errors. However, when checking the correlations, no severely high correlation has been found. Therefore, the consequences of multicollinearity, if it is there, will be limited.
V. Results
Baseline specifications
16 estimate for trade openness is insignificant. Next, the estimates of the EMU dummy variable are insignificant in all six regressions. Therefore, one could conclude from these estimates that being in the Eurozone or not as a European country does not affect inflation volatility in general.
Table 2 Baseline specification inflation volatility
Variables 1 2 3 4 5 6 Inflation average 0.450*** (0.065) 0.480*** (0.063) 0.550*** (0.061) 0.566*** (0.063) 0.543*** (0.060) 0.562*** (0.060) Log of GDP per capita -0.511*** (0.050) -0.478*** (0.053) -0.421*** (0.047) -0.403*** (0.055) -0.529*** (0.071) -0.417*** (0.046) EMU -0.153 (0.028) 0.008 (0.030) 0.005 (0.028) -0.001 (0.027) 0.015 (0.029) -0.002 (0.028) CA/TA -0.004** (0.001) LL/GDP 0.000 (0.001) PC/GDP 0.000 (0.000) Investment % of GDP -0.145 (0.298) Government % of GDP -1.178** (0.569) Openness -0.001 (0.001) Observations 431 427 461 469 469 469 Countries 29 28 29 29 29 29 R-squared (within) 0.4372 0.3488 0.3774 0.3764 0.3823 0.3778 Notes:
- Standard errors reported in parenthesis
- Significance levels at which the null hypothesis is rejected: *** 1%, ** 5%, * 10%
17
Table 3 Baseline specification growth volatility
Variables 1 2 3 4 5 6 Growth average 0.374*** (0.103) 0.350*** (0.108) 0.382*** (0.101) 0.438*** (0.095) 0.310*** (0.090) 0.325*** (0.094) Log of GDP per capita -0.020*** (0.006) -0.017*** (0.006) -0.013** (0.005) 0.002 (0.006) -0.042*** (0.006) -0.013*** (0.005) EMU 0.000 (0.003) -0.002 (0.004) -0.001 (0.003) -0.003 (0.003) 0.002 (0.003) -0.002 (0.003) CA/TA 0.001*** (0.0001) LL/GDP 0.000 (0.000) PC/GDP 0.000 (0.000) Investment % of GDP -0.140*** (0.035) Government % of GDP -0.343*** (0.057) Openness 0.0001* (0.0000) Observations 480 474 512 520 520 520 Countries 29 28 29 29 29 29 R-squared (within) 0.1222 0.0935 0.0939 0.1233 0.1582 0.0988 Notes:
- Standard errors reported in parenthesis
- Significance levels at which the null hypothesis is rejected: *** 1%, ** 5%, * 10%
export-18 base broadness could be more informative in answering this question. However, this is beyond the scope of this paper.
Specifications of baseline plus interaction terms
Table 4 Baseline specification plus interaction terms for inflation volatility
Variables 1 2 3 4 5 6 Inflation average 0.453*** (0.065) 0.476*** (0.063) 0.545*** (0.061) 0.597*** (0.064) 0.541*** (0.060) 0.561*** (0.059) Log of GDP per capita -0.517*** (0.050) -0.488*** (0.054) -0.432*** (0.049) -0.349*** (0.058) -0.508*** (0.075) -0.410*** (0.046) EMU -1.744** (0.752) 0.077 (0.059) 0.037 (0.052) -0.281** (0.109) 0.099 (0.104) -0.156*** (0.051) CA/TA1 -0.004** (0.001) LL/GDP2 0.001 (0.001) PC/GDP3 0.000 (0.001) Investment % of GDP4 -0.744** (0.372) Government % of GDP5 -0.989 (0.612) Openness6 -0.001** (0.001) Interaction term 0.018** (0.007) -0.001 (0.001) 0.000 (0.000) 1.016*** (0.383) -0.427 (0.509) 0.001*** (0.000) Observations 431 427 461 469 469 469 Countries 29 28 29 29 29 29 R-squared (within) 0.4449 0.3519 0.3783 0.3867 0.3834 0.3962 Notes:
- Standard errors reported in parenthesis
- Significance levels at which the null hypothesis is rejected: *** 1%, ** 5%, * 10%
- The number 1 in CA/TA1 is there to clarify that the interaction term of CA/TA is reported in regression 1. Numbers added to other variables are to be interpreted similarly.
19 coefficient of the relative government size is now insignificant. Therefore, one should not draw strong conclusions on the estimates of Table 2 as the results do not seem to be robust. The EMU dummy variable shows a significant estimate in regressions 1, 4 and 6. Results on the interaction terms and the EMU dummy variable will be discussed later in the results section. Reason for this is that interpretation of these results is easier when one looks at graphed plots of the estimated marginal effects for given values of the control variables.
Table 5 Baseline specification plus interaction terms for growth volatility
Variables 1 2 3 4 5 6 Growth average 0.384*** (0.103) 0.374*** (0.108) 0.373*** (0.101) 0.441*** (0.095) 0.298*** (0.090) 0.333*** (0.094) Log of GDP per capita -0.020*** (0.006) -0.017*** (0.006) -0.014*** (0.005) 0.006 (0.007) -0.049*** (0.007) -0.013** (0.005) EMU -0.165* (0.099) 0.007 (0.006) 0.005 (0.006) -0.030** (0.013) -0.029** (0.012) 0.005 (0.006) CA/TA1 0.0007*** (0.0001) LL/GDP2 0.000 (0.000) PC/GDP3 0.0001* (0.000) Investment % of GDP4 -0.190*** (0.041) Government % of GDP5 -0.393*** (0.059) Openness6 0.0002** (0.0001) Interaction term 0.002* (0.001) -0.0001* (0.00005) 0.000 (0.000) 0.100** (0.045) 0.161*** (0.059) 0.000 (0.000) Observations 480 474 512 520 520 520 Countries 29 28 29 29 29 29 R-squared (within) 0.1279 0.1002 0.0972 0.1325 0.1714 0.1024 Notes:
- Standard errors reported in parenthesis
- Significance levels at which the null hypothesis is rejected: *** 1%, ** 5%, * 10%
- The number 1 in CA/TA1 is there to clarify that the interaction term of CA/TA is reported in regression 1. Numbers added to other variables are to be interpreted similarly.
20 savings allocation by commercial banks relative to the central bank and higher debt by private enterprises relative to GDP relates to higher growth volatility. Furthermore, the share of government expenditures of GDP is again estimated to be negative similar to the result in Table 3. Trade openness is again estimated to correlate positively with growth volatility similarly to the result in Table 3. Also, the investment share of GDP is estimated to be related negatively with growth volatility. Interesting finding is that the estimate for natural logarithm of GDP becomes insignificant if the investment share of GDP is included in the regression. The EMU dummy variable is estimated to be significant in regressions 1, 4 and 5. Again, the interaction terms and EMU dummy estimates will be further discussed later in the results section.
General Specifications
Table 6 General specification for inflation volatility
Variables 1 2 3 4 5 6 Inflation average 0.424*** (0.069) 0.470*** (0.067) 0.544*** (0.065) 0.591*** (0.064) 0.551*** (0.064) 0.562*** (0.062) Log of GDP per capita -0.690*** (0.082) -0.579*** (0.091) -0.527*** (0.080) -0.483*** (0.078) -0.498*** (0.081) -0.528*** (0.076) EMU -2.093*** (0.776) 0.111* (0.061) 0.056 (0.053) -0.336*** (0.110) 0.106 (0.104) -0.166*** (0.052) CA/TA1 -0.004** (0.002) LL/GDP2 0.001 (0.001) PC/GDP3 0.000 (0.000) Investment % of GDP4 0.181 (0.317) -0.025 (0.333) -0.050 (0.308) -0.791** (0.370) 0.057 (0.301) -0.293 (0.300) Government % of GDP5 -1.656*** (0.628) -1.067 (0.696) -1.03* (0.584) -1.452** (0.572) -0.942 (0.614) -1.551*** (0.568) Openness6 -0.0001 (0.001) -0.001** (0.001) -0.001 (0.001) -0.001 (0.001) -0.001 (0.001) -0.002*** (0.001) Interaction terms 0.021*** (0.008) -0.001** (0.001) 0.000 (0.000) 1.278*** (0.391) -0.477 (0.512) 0.002*** (0.000) Observations 431 427 461 469 469 469 Countries 29 28 29 29 29 29 R-squared (within) 0.4561 0.3615 0.3846 0.3994 0.3853 0.4082 Notes:
- Standard errors reported in parenthesis
21
- The number 1 in CA/TA1 is there to clarify that the interaction term of CA/TA is reported in regression 1. Numbers added to other variables should be interpreted similarly.
Table 7 General specification for growth volatility
Variables 1 2 3 4 5 6 Growth average 0.453*** (0.099) 0.366*** (0.108) 0.449**** (0.099) 0.395*** (0.091) 0.376*** (0.091) 0.392*** (0.092) Log of GDP per capita -0.035*** (0.008) -0.039*** (0.010) --0.024*** (0.008) -0.022*** (0.008) -0.032*** (0.008) -0.026*** (0.008) EMU -0.118 (0.095) 0.006 (0.006) 0.003 (0.006) -0.032*** (0.012) -0.033*** (0.012) 0.000 (0.006) CA/TA1 0.001*** (0.000) LL/GDP2 0.0001* (0.0000) PC/GDP3 0.000 (0.000) Investment % of GDP4 -0.155*** (0.037) -0.131*** (0.039) -0.150*** (0.034) -0.200*** (0.039) -0.143*** (0.033) -0.144*** (0.034) Government % of GDP5 -0.404*** (0.061) -0.506*** (0.071) -0.327*** (0.057) -0.354*** (0.056) -0.388*** (0.058) -0.334*** (0.056) Openness6 0.000 (0.000) 0.000 (0.000) 0.0002*** (0.000) 0.0001** (0.000) 0.0002*** (0.000) 0.0002** (0.000) Interaction term 0.001 (0.001) 0.000 (0.000) 0.000 (0.000) 0.121*** (0.044) 0.179*** (0.058) 0.000 (0.000) Observations 480 474 512 520 520 520 Countries 29 28 29 29 29 29 R-squared (within) 0.2433 0.2371 0.1995 0.2094 0.2127 0.1966 Notes:
- Standard errors reported in parenthesis
- Significance levels at which the null hypothesis is rejected: *** 1%, ** 5%, * 10%
- The number 1 in CA/TA1 is there to clarify that the interaction term of CA/TA is reported in regression 1. Numbers added to other variables should be interpreted similarly.
22 result will be discussed later in the results section. Second, in Table 7 similar estimates are shown for average growth and the natural logarithm of GDP compared to Tables 3 and 5. In Regression 4, now both the natural logarithm of GDP and the investment share of GDP show significant estimates. Therefore, the insignificant estimate for the natural logarithm in regression 4 of Table 5 should not be weighted to heavily. Furthermore, the share of government expenditures of GDP and the investment share of GDP are estimated to be significantly related with growth volatility in all six regressions. Finally, trade openness is estimated to be positively related with growth volatility in regressions 3, 4, 5 and 6. A significant negative estimate for the EMU dummy variable has been found in regressions 1, 4 and 6 of Table 6. In contrast, a significant positive estimate has been found in regression 2. Table 7 reports significant negative estimates of the EMU dummy variable in regressions 4 and 5. On the estimates of the interaction terms and EMU dummy variables will be further elaborated later in the results section.
EMU Interactions
23 -2 -1 .5 -1 0 -0 .5 E s ti m at e d m ar gi n al e ff ec t o f E M U 42 47 52 57 62 67 72 77 82 87 92 97 CA/TA Regression table 3 -2 -1 .5 -1 0 -0 .5 E s ti m at e d m ar gi n al e ff ec t o f E M U 42 47 52 57 62 67 72 77 82 87 92 97 CA/TA Regression table 5 Figure 3 Marginal effects of EMU on inflation volatility given CA/TA
0 -0 .6 -0 .2 -0 .4 0 .2 E s ti m at e d m ar gi n al e ff ec t 17 37 57 77 97 117 137 157 177 197 217 237 257 277 297 317 337 357 377 LL/GDP Regression table 3 -. 4 0 -0 .6 -0 .2 0 .2 E s ti m at e d m ar gi n al e ff ec t 17 37 57 77 97 117 137 157 177 197 217 237 257 277 297 317 337 357 377 LL/GDP Regression table 5 Figure 4 Marginal effects of EMU on inflation volatility given LL/GDP
effects of the EMU dummy variable for given values of CA/TA, PC/GDP and trade openness will not be shown as no significant estimates were found.
Figure 3 shows the estimated marginal effects of the EMU dummy variable on inflation volatility for given values of CA/TA. The plots show that as CA/TA decreases, Eurozone countries experience lower inflation volatility. Yet, as the average value of CA/TA is far to the right of the graph there is no significant marginal effect found for the average country. Moreover, the standard deviation of the distribution of CA/TA is relatively small causing the increasing range of the confidence intervals at the left of the plots. Hence, the cases for which the estimated marginal effect is the most pronounced are quite uncommon. Nevertheless, the results show that for countries where the central bank is relatively dominant in savings allocation, becoming a member of the Eurozone is beneficial.
24 0 -0 .4 0 .2 -0 .2 0 .4 E s ti m at e d m ar gi n al e ff ec t 0.07 0.1 0.13 0.16 0.19 0.22 0.25 0.28 0.31 0.34 0.37 0.4 0.43 0.46 Investment % of GDP Regression table 3 0 -0 .4 0 .2 0 .4 -0 .2 E s ti m at e d m ar gi n al e ff ec t 0.07 0.1 0.13 0.16 0.19 0.22 0.25 0.28 0.31 0.34 0.37 0.4 0.43 0.46 Investment % of GDP Regression table 5 Figure 5 Marginal effects of EMU on inflation volatility given the investment share of GDP
report a significant marginal effect for higher values of LL/GDP. This result suggests that EMU countries experience less inflation volatility as their financial sector becomes larger. This finding is in line with the theoretical predictions of Rousseau and Wachtel 2001. However, this conclusion is not supported by the estimations reported in Table 3. The estimated marginal effect is not significant at the average value of LL/GDP, therefore countries with an average size of the financial sector are not expected to experience more or less volatile inflation when being an EMU member.
The marginal effects of the EMU dummy variable on inflation volatility for given values of the investment share of GDP are displayed in Figure 5. Both plots show that the estimated marginal effect of the Eurozone dummy switches sign. For relatively low investment countries, the countries inside the EMU experience lower inflation volatility. Whereas for relatively high investment countries, countries inside the Eurozone experience higher inflation volatility. For countries with an average value the investment share of GDP the estimated marginal effect is estimated to be insignificant. One could argue that endogeneity plays a role here. The high level of investments could also be caused by the fact that prices are stable since stable prices raise incentives for agents to engage in long term contracts raising investment. Nevertheless, one can conclude from the results that the share of investments is an important indicator to consider for countries who are considering entering the Eurozone if inflation volatility is concerned.
25 0 -0 .2 0 .4 0 .2 0 .6 E s ti m at e d m ar gi n al e ff ec t 37 67 97 127 157 187 217 247 277 307 337 367 Trade openness Regression table 3 0 -0 .2 0 .2 0 .6 0 .4 E s ti m at e d m ar gi n al e ff ec t 37 67 97 127 157 187 217 247 277 307 337 367 Trade opennes Regression table 5 Figure 6 Marginal effects of EMU on inflation volatility given trade opennes
0 -0. 06 -0. 02 0. 02 -0. 04 E s ti m at e d m ar gi n al e ff ec t 17 37 57 77 97 117 137 157 177 197 217 237 257 277 297 317 337 357 377 LL/GDP Regression Table 4 0 -0. 06 -0. 02 0. 02 -0. 04 E s ti m at e d m ar gi n al e ff ec t 17 37 57 77 97 117 137 157 177 197 217 237 257 277 297 317 337 357 377 LL/GDP Regression Table 6 Figure 7 Marginal effects of EMU on growth volatility given LL/GDP
dummy variable switches sign. According to the regression results of Tables 3 and 5, countries in the EMU with relatively closed economies experience less volatile inflation compared to European countries outside the Eurozone. In contrast, Eurozone countries with relatively open economies experience more fluctuations in the inflation level than European countries not member of the EMU. Countries with an average level of trade openness are estimated not to be affected by EMU membership regarding inflation volatility. The results are contrary to the theoretical predictions of Bowdler & Malik (2005) arguing that more open economies will experience less volatile inflation. On the other hand, the results are in line with the argument that relatively open economies are more prone to foreign price fluctuations.
26 -. 0 4 -. 0 2 0 .0 2 .0 4 .0 6 E s ti m at e d m ar gi n al e ff ec t 0.07 0.1 0.13 0.16 0.19 0.22 0.25 0.28 0.31 0.34 0.37 Government % of GDP Regression Table 4 -. 0 4 -. 0 2 0 .0 2 .0 4 .0 6 E s ti m at ed m ar gi nal ef fec t 0.07 0.1 0.13 0.16 0.19 0.22 0.25 0.28 0.31 0.34 0.37 Government % of GDP Regression Table 6 Figure 8 Marginal effects of EMU on growth volatility given the government share of GDP
experience lower growth volatility compared to European countries outside of the monetary union. Yet, this result is not supported by the regression in Table 6 therefore one cannot conclude the existence of this relation with certainty. Both graphs show no significant marginal effect for the average country in the sample.
Next, Figure 8 displays the estimated marginal effects of the Eurozone dummy for given shares of government expenditures of GDP. Both plots show clear heterogeneity of the estimated marginal effects for given shares of government expenditures of GDP. In more detail, EMU countries with a relatively small government experience less volatile economic growth compared to other European non-Eurozone countries. On the other hand, Eurozone countries with a relatively large government tend to have more growth volatility than European countries not part of the monetary union. This could be due to the fact that large amounts public debt issued by governments in euros is a possible source of instability. For countries with an average share of government expenditures of GDP the estimated marginal effect is estimated to be insignificant.
27 0 -0. 04 0. 02 -0. 02 0. 04 E s ti m at e d m ar gi n al e ff ec t 0.07 0.1 0.13 0.16 0.19 0.22 0.25 0.28 0.31 0.34 0.37 0.4 0.43 0.46 Investment % of GDP Regression Table 4 0 -0. 04 0. 02 -0. 02 0. 04 E s ti m at e d m ar gi n al e ff ec t 0.07 0.1 0.13 0.16 0.19 0.22 0.25 0.28 0.31 0.34 0.37 0.4 0.43 0.46 Investment % of GDP Regression Table 6 Figure 9 Marginal effects of EMU on growth volatility given the investment share of GDP
these countries to keep the inflationary pressures under control which in turn stabilizes economic growth. For countries with shares of investment that can be considered average no significant marginal effect is found.
VI. Policy Recommendations
Based upon the empirical results it is possible to formulate policy recommendations from the perspective of countries contemplating to join the EMU. First of all, countries having issues with keeping inflation low caused by for instance a lack of central bank independence there are significant benefits concerning inflation volatility as suggested by the significant estimates on average inflation in Tables 2, 4 and 6. However, for countries already having a credible independent central bank able to keep inflation at the targeted level the question at hand becomes more complex.
28 high investment countries the results suggest that entering the EMU might not be a good decision as inflation volatility is detrimental for economic growth (Judson & Orphanides 1999),. The opposite holds for low investment countries with relatively closed economies, for these countries the results suggest that it is beneficial considering inflation volatility to become a member of the EMU.
Figures 6 and 7 displayed two relevant country characteristics when the issue as growth volatility is examined. First, countries with a relatively big government are worse of in the EMU regarding growth volatility as shown in Figure 6 and the estimated interaction terms of Tables 5 and 7. The contrary holds for countries with a relatively small government. One could argue that government debt becomes more risky when a country enters the Eurozone. In turn, the volatility of economic growth will increase for countries entering the Eurozone where a large share of GDP is accounted for by public activities. Second, the relative share of investments is an important country characteristic to keep in mind. Low investment countries could benefit from being in the Eurozone as low investment countries in the EMU tend to have lower growth volatility in comparison with similar European non-Eurozone countries.
The following policy implications may therefore be listed:
1. Open economies in the Eurozone experience increased inflation volatility. Based upon the theoretical predictions of Bowdler & Malik Eurozone countries should ascertain that their export base becomes or remains well diversified. Besides the effect that this lowers inflation volatility, a well-diversified export portfolio of EMU countries sets the framework for intra-industry trade which in turn could be beneficial concerning growth volatility as suggested by the European Commission view.
2. Eurozone countries with relatively large governments experience relatively unstable economic growth which may be caused by the issuance of public debt in euros. Possible solutions are the joint issuance of Eurobonds or moving from a monetary union to a budgetary union (De Grauwe 2016). However, this may introduce moral hazard problems.
3. Countries with relatively low investments, relatively closed economies and a relatively small government are advised to enter the EMU as for these countries the benefits on inflation volatility and growth volatility are clear from the results.
29 The question remains whether or not it is beneficial for the EMU as a whole that countries enter with the characteristics listed before. If that would be the case then entry would be beneficial for both sides. It would be an interesting yet tough question for future research.
VII. Conclusions
Overall, relatively low investment, closed economies and countries with a relatively small government benefit the most from the EMU both concerning inflation and growth volatility. Two of these characteristics are more common in economically less developed countries which are the relatively low investments and a closed economy. These countries also tend to have more difficulties in restraining average inflation and thereby inflation volatility. However, this also holds for countries with a relative large government who are worse of in the EMU. Nevertheless, one could argue that the Eurozone offers great benefits for countries which are economically less developed compared to the Eurozone countries from a perspective of macroeconomic stability. Especially if the EMU moves forward to a budgetary union, otherwise bad fiscal positions could raise problems for potential entrants.
However, for countries like for example the United Kingdom or Norway the question becomes more complex on whether or not entering the EMU would be beneficial for macroeconomic stability. On the one hand, one could interpret the results in a negative way stating that there are no benefits of entering the Eurozone according to the empirical results. On the other hand, one could interpret the outcomes in positively by stating that from a perspective of macroeconomic stability there are no costs of being in the EMU. Especially when noting that the EMU dummy variable was estimated to be significant and positive only once. Therefore, why should a country not enter as firms can benefit from diminished foreign exchange risk and the economies of scale which could foster economic growth.
30 Finally, looking at the future of the Eurozone, countries currently being member of the European Union but not a member of the EMU face the lowest bars to enter the EMU. Of these countries, Croatia and Bulgaria are most likely to enter given the results of this paper. Croatia is a relatively closed country and also has relatively low investments. Besides, the relative government size of Croatia is a bit above average but not severely large. Next, Bulgaria has a relatively low investment level making EMU membership beneficial concerning macroeconomic stability. Whereas the share of government expenditures of GDP and trade openness are slightly above average, but not severely high to have negative effects on macroeconomic stability if Bulgaria decides to enter to the EMU. Nonetheless, regardless of the decision of Croatia and Bulgaria to enter, the Eurozone is more likely to expand than to shrink in the future.
References
Aisen, A., & Vega, F. J. (2008). Political instability and inflation volatility. Public Choice,
135(3), 207-223.
Alesina, A., & Summers, L. H. (1993). Central bank independence and macroeconomic performance: some comparative evidence. Journal of Money, Credit and Banking, 25(2), 151-162.
Beck, T., Lundberg, M., & Majnoni, G. (2006). Financial intermediary development and growth volatility: Do intermediaries dampen or magnify shocks? Journal of International
Money and Finance, 25(7), 1146-1167.
Bodea, C., & Hicks, R. (2015). Price stability and central bank independence: Discipline, credibility and democratic institutions. International Organization, 69(1), 35-61.
Bowdler, C., & Malik, A. (2005). Openness and inflation volatility: Panel data evidence.
University of Oxford, (Working paper).
Campolmi, A., & Faia, E. (2011). Labor market institutions and inflation volatility in the euro area. Journal of Economic Dynamics and Control, 35(5), 793-812.
Cukierman, A., Web, S. B., & Neyapti, B. (1992). Measuring the independence of central banks and its effect on policy outcomes. The World Bank Economic Review, 6(3), 353-398. Cecchetti, S. G., & Ehrmann, M. (1999). Does inflation targeting increase output volatility? An international comparison of policymakers' preferences and outcomes. National bureau of
economic research, (No. w7426).
31
Clarida, R., Gali, J., & Gertler, M. (2000). Monetary policy rules and macroeconomic stability: Evidence and some theory. The Quarterly Journal of Economics, 115(1), 147-180. Fischer, S. (1995). Central-bank independence revisited. The American Economic Review,
85(2), 201-206.
Friedman, M. (1977). Nobel lecture: inflation and unemployment. Journal of Political
Economy, 85(3), 451-472.
De Grauwe, P. (2016). Economics of Monetary Union. Oxford University Press.
Giannone, D., Lenza, M., & Reichlin, L. (2008). Explaining the great moderation: It is not the shocks. Journal of the European Economic Association, 6(2‐3), 621-633.
Grüner, H. P. (2002). How much should central banks talk?: A new argument. Economics
Letters, 77(2), 195-198.
Haddad, M., Lim, J. J., & Saborowski, C. (2010). Trade openness reduces growth volatility when countries are well diversified, The World Bank, (No. w5222).
Judson, R., & Orphanides, A. (1999). Inflation, volatility and growth. International Finance,
2(1), 117-138.
Klomp, J., & de Haan, J. (2010). Central bank independence and inflation revisited. Public
Choice, 144(3), 445-457.
Levine, R., Loayza, N., & Beck, T. (2000). Financial intermediation and growth: Causality and causes. Journal of Monetary Economics, 46(1), 31-77.
Nakov, A., & Pescatori, A. (2010). Oil and the great moderation. The Economic Journal,
120(543), 131-156.
Rogoff, K. (1985). The optimal degree of commitment to an intermediate monetary target.
The Quarterly Journal of Economics, 100(4), 1169-1189.
Rose, J. (1991). Political uncertainty and macroeconomic performance. Economics Letters,
37(4), 377-381.
Rother, P. (2004). Fiscal policy and inflation volatility. European Central Bank, (No. w317). Summers, P. M. (2005). What caused the Great Moderation? Some cross-country evidence.
Economic Review-Federal Reserve Bank of Kansas City, 90(3), 5-32.
Rousseau, P. L., & Wachtel, P. (2001). Inflation, financial development and growth. In
Economic Theory, Dynamics and Markets, (pp. 309-324). Springer US.
32
Sarel, M. (1996). Nonlinear effects of inflation on economic growth. International Monetary
Fund Staff Papers, 43(1), 199-215.
Weber, C. S. (2016). Central bank transparency and inflation (volatility)–new evidence.
International Economics and Economic Policy, published online, 1-47. Appendix A
Table 8 Countries included and EMU entrance
Country Year of entrance Country Year of entrance
Austria 1999 Latvia 2014
Belgium 1999 Lithuania 2015
Bulgaria Not a member Luxembourg 1999
Croatia Not a member Malta 2008
Cyprus 2008 The Netherlands 1999
Czech Republic Not a member Norway Not a member
Denmark Not a member Poland Not a member
Estonia 2011 Portugal 1999
Finland 1999 Slovakia 2009
France 1999 Slovenia 2007
Germany 1999 Spain 1999
Greece 2001 Sweden Not a member
Hungary Not a member Switzerland Not a member
Ireland 1999 United Kingdom Not a member
