REAL RATE OF THE ESTONIAN KROON(1)

(1) I am grateful to Natalja Viilmann from Eesti Pank and Viktoria Trasanov from the State Statistical Office for calculations and information. The views expressed in this article do not necessarily coincide with the official views of Eesti Pank.

Contents


Introduction

The aim of the present article is to evaluate the level of the real rate of the Estonian kroon and its changes. The author is mainly interested in whether the real rate has gone up (or is going up) to such a degree that Estonia's domestic purchasing power has become (or could become) smaller than the purchasing power abroad.

In this connection we should point to another topical problem. Kravis (1986) stresses that the deviation of the real purchasing power from the nominal rate decreases with the development of the economy. The deviation is the biggest in developing countries and smaller in industrialized countries. In Estonia, where the economic development is lower, the real rate of the currency should be low. Actually, the author is mainly interested in whether the real rate of the kroon is higher than the level that suits Estonia's level of development or not. If the real rate is lower than indicated by the level of economic development, there are obviously no problems, but if the rate is higher than the level predicts this indicates that the real rate is too high at the moment (the real rate increases with economic development). Thus, we are interested in whether the level of the real rate is sufficient so that higher inflation that comes with development would not turn the kroon's actual purchasing power lower than the nominal value in the long run.

Our analysis evaluates the real rate of the kroon on three different levels. The 1993 rate has been calculated on the basis of actual data. The 1995 rate is based on the real rate index, and the 1999 rate is a forecast. The year 1999 has been chosen because by that time Estonia has presumably passed the period of high inflation and the annual rate of price rise has dropped below 10%. Would the real rate of the kroon reach such a level by that time that the real purchasing power of the kroon is lower than the nominal rate?

The article does not contain economic or political conclusions. The political interpretation of the changes in the real rate mainly has to do with the international competitiveness of the economy. (2) However, I would like to keep this topic apart from the topic of the present article because the real rate of the currency and competitiveness of a country's economy are in reality not so directly linked as it is often claimed. In case of Estonia, the growth of export has so far been determined by quite different factors, not the real rate of the kroon: the export getting more expensive has been accompanied by a constant growth in its volume. True, such development is typical of a transition economy but in countries with well developed market economies, the real rate of the currency and competitiveness of the economy are not always immediately connected. This is confirmed by both academic literature as well as practical analyses. In many cases it has turned out that despite the increase in the real rate of the currency the competitiveness of the economy of the respective country has actually increased (see for example Feldman's (1994) thorough analysis on Germany).

(2) See Clark - Bartolini - Bayoumi - Symansky (1994).

Real Exchange Rate

The real exchange rate of a currency is its equilibrium exchange rate. In modern economic literature the equilibrium exchange rate by definition is an exchange rate which guarantees both the domestic and international equilibrium for the economy. The latter means the simultaneous equilibrium of the money, trade and foreign markets. The equilibrium exchange rate need not manifest itself in short term. It needs a medium(3) or long-term period.(4)

(3) FEER or fundamental equilibrium exchange rate. See Williamson (1983).

(4) NATREX or natural real exchange rate. See Stein-Allen (1995).

An approximation of the real exchange rate can also be a simplified model of the above. According to this, the equilibrium exchange rate would be determined on the basis of the foreign sector (balance of payments) only.(5) The real rate will bring the balance of payments into equilibrium over a longer period.(6)

(5) See MacDonald (1988 and 1995) for example.

(6) Conditions that enable this have been defined by R. Nurkse (1945).

As we can see from the analyses based on the balance of payments (foreign sector) the real rate depends on the purchasing power parity (PPP) and (covered or uncovered) interest parity. By simplifying the analysis based on the equilibrium of the balance of payments (foreign sector) we can reach a model in which the real exchange rate is identical to the level of the purchasing power parity. The exchange rate derived from the purchasing power parity gives the respective currencies an equal (or parity) purchasing power in buying goods. The exchange rate derived from the purchasing power parity is called the purchasing power parity real rate (PPPRR).

In case of the PPPRR only the equilibrium determined by the price of goods and services matters. The circle of goods and services selected for calculating the PPP is another matter. The sensible alternatives here would be either exported and imported goods, or else, the goods sold and bought within the framework of the GDP or household consumption. If we limit ourselves to exported and imported goods we can show that the absolute PPP is only linked to the current account of the balance of payments and does not concern the money flows of the capital account. Hence we can conclude that simple models for evaluating the PPP are faulty in principle (MacDonald (1995)). The relationship between the PPP and the real rate is not so trivial if we link it to a wider selection of goods and services, those included in the GDP, for instance. The PPP exchange rate is widely known in this meaning.(7)

(7) The author of the PPP doctrine is the Swedish economist Gustav Cassell, who in 1918 characterised the exchange rate through the purchasing power of a national currency.

The PPP is connected with the law of one price. According to this law, the price of similar goods is the same in different countries. The similarity comes from arbitrage, that is, purchasing goods with the aim of reselling them at a higher price. The validity of the law of one price is proved by the price of homogeneous goods (coffee, oil, etc.) at international commodities exchanges.(8) In case of more complicated goods the application of the law of one price rule questionable. The concept of an absolute PPP, however, maintains that although every single commodity does not have the same price in different markets, the average price of the selection of concrete commodities is the same in different countries.

Under a more soft version, there need not be an absolute PPP but a relative PPP. The relative PPP is a reduced version of the absolute PPP according to which the price change in goods and services is the same for all countries.

(8) After adaptions to the time delay of delivery, etc.

Law of One Price and Purchasing Power Parity Real Rate

Unfortunately, there are a number of factors that restrict the validity of the law of one price and the manifestation of an absolute PPP.

The validity of the law is restricted by the existence of the sheltered sector of the economy. Since the sheltered sector does not compete with foreign suppliers the agents active in this sector have free hands to set the prices. Thus, the sheltered sector prices differ from country to country, which means that the law of one price does not apply to goods and services in the sheltered sector.(9) As the output of the sheltered sector is part of the GDP or other macroeconomic aggregate these, too, deviate from the law of one price.

(9) If the prices of production factors and production functions are identical from country to country then the law of one price does apply also to goods in the sheltered sector (Hard, 1995).

However, the prices in the open sector do not necessarily have to reflect the law of one price. The levelling out of prices between different countries is prevented by transport and information costs, trade restrictions (customs tariffs, quotas, etc.), as well as other state-introduced administrative instruments (taxes, for example).

Deviations from the law of one price are also caused by the rigidity of the prices that facilitates the shocks of the nominal exchange rate being transferred to the price level and the real rate.(10) The above impact is also connected with the different adaptation rate of finance and trade markets. Sudden changes on the financial market can cause a deviation of the nominal exchange rate from its long-term equilibrium state which is then reflected in the real rate. Deviations from the law of one price can also be explained by the different pricing of markets. The pricing to market theory of Krugman (1987) and Dornbusch (1987) maintains that the oligopolistic suppliers can ask a different price for the same good at different markets. Thus, for instance, the price of BMWs is different in the USA and Europe.

(10) See Isard (1977), Giovannini (1988), Engel (1993), whose empirical research testifies to relations between short-time deviations in the price level and changes in the nominal exchange rate.

The international deviation of prices by countries is also explained by a theory under which the price level of a country is a function of its level of development. We will have a closer look at this theory in connection with the empirical evaluation of the real rate of the kroon in the subsection of our article entitled "Real Rate and Level of Development".

Although the above factors speak against the law of one price and the absolute PPP derived from it, we still use the PPPRR in the real rate function in our article. This is partly because there is no alternative. There is neither information nor adequate models for determining other equilibrium exchange rates in Estonia.

Secondly, the meaning of the absolute PPP in long term is also unclear yet. The above factors undoubtedly prevent the law of one price to manifest itself in short term but at the same time they need not prevent the absolute PPP to become manifest in long term. Unfortunately, even empirical studies fail to provide us with answers about the validity or lack of validity of the PPP. According to Bartolini (1995) and Clark - Bartolini - Bayoumi - Symansky (1994), there are results that support this idea as well as results that refute it.

Thirdly, the PPPRR is, despite its shortcomings in the empirical treatment, still the most widely used indicator of the real rate. The PPPRR is used in the international comparisons of the European Union (EU), the Organisation of Economic Cooperation and Development (OECD) and the UN Development Programme. The PPPRR is popular also in national analyses.(11)

(11) The most notorious practical application of the PPPRR in determining the nominal rate took place in April 1925 when the British pound was returned to the gold standard. Based on the producer price index, the mistakenly quantified PPPRR caused the 12% overvaluation of the nominal rate of the pound (Hard 1995, p. 70).

Real Rate and Its Change Indicators

If we identify the real rate with the PPPRR, it also determines the indicator of the real rate in principle. The PPPRR is measured as the country's average price level compared to the average of another country or group of countries determined on the basis of a uniform classification of commodities.(12)

(12) See Comparison (1995, pp. 2...17) for methodology.

As a rule, the use of indicators of the economy's international competitiveness is recommended for characterizing changes in the real rate, for instance, indices based on the labour consumption per production unit (either in the entire economy, in industry, the open sector or the export production), the export-import deflators or value added deflators (consumer price index, producer price index), the ratio of the open and sheltered sector price indices, etc.(13)

(13) The pros and cons of different indicators have been discussed by Feldman (1994), Bartolini (1995), Marsh - Tokarick (1994), Lipschits - McDonald (1991).

Depending on the existence of data we have used the real effective exchange rate index (REER) below to follow changes in the real rate. The REER is an indicator that equals to the ratio of trade partners' inflation (usually determined by the consumer price indices but can also be defined on the basis of the producer price indices) and the domestic inflation, and the weighted geometric average of the indices of the nominal exchange rate.(14)

For Estonia, the REER has been calculated as an index of changing structure. The most recent normalized structure was the following: Finland 0.333, Russia 0.226, Sweden 0.121, Germany 0.108, Latvia 0.055, the Netherlands 0.044, Lithuania 0.042, Denmark 0.038, the USA 0.033. The REER is a link in the system of exchange rate indices used by the International Monetary Fund (IMF). The system also includes the nominal effective exchange rate (NEER) and the multilateral exchange rate model (MERM).(15)

(14) For the methods of calculating the REER see Franzén - Markowsky - Rosenberg (1980).

(15) For a methodological survey of the system, see Effective (1994).

The REER value that differs from 1 describes the change in the real exchange rate. If the REER is bigger than 1, the real rate of the national currency has increased against the currencies of the trading partners, but if the REER is below 1, it has become cheaper. According to the REER, the change in the real rate is caused either by the change in the nominal rate or due to the different rates of inflation of different countries.

Real Rate of the Estonian Kroon in 1993

Comparison with Austria

The 1993 level of the real rate of the kroon has been obtained from the results of the EU statistics organisation (Eurostat) and the OECD European Comparison Programme (see Trasanov (1996)). The aim of the programme was to compare the gross domestic product (GDP) and the prices and volumes of its components in different countries. The 1993 study also comprised Estonia. The results (unofficial so far) of the programme give us Estonia's per capita GDP (PC GDP) according to the purchasing power parity exchange rate and Estonia's level as compared to Austria on the basis of the average prices.(16)

(16) The study we refer to is the first and the only one so far where the real rate has been evaluated through direct calculations. However, this was not the first time an attempt was made to evaluate the real rate of the kroon. In earlier studies evaluations have been indirect. A typical method was to mix data for the Soviet Union and Estonia, that is, to proceed from the data of early 1990s and adjust them with later changes in the exchange rate, inflation and economic growth indicators. Unfortunately, results given in various sources proved to vary greatly. Thus, according to the Economist Intelligence Unit the real rate of the kroon was 0.171 in 1993. According to the Finnish Teollisen Yhteistyön Rahasto OY and its publication Finnfund News the real rate of the kroon in 1993 was 0.04...0.05. Evaluations differing that much made it impossible to judge the actual level of the real rate of the kroon.

The research data needed for analyzing the real rate have been given in Table 1. The data is presented as a comparison with Austria which served as a basis for comparison for Estonia and other Baltic countries in that project.

The price level given in the table reflects the average level of commodities (that is, goods and services, including services provided by the public sector) as compared to Austria. The price level equals to the ratio of per capita GDP calculated using the nominal and the real exchange rates. Besides differences in the price level between Austria and Estonia, this correlation also indicates the real rate of the Estonian kroon against the Austrian shilling based on the PPP.

If the real purchasing power of the kroon in Estonia were equal to its nominal rate the price levels would be equal in Estonia and Austria and commodities would cost the same in both countries. Actually, commodities in Estonia cost only a quarter of the price they have in Austria (according to the GDP). In other words, in Estonia you can buy four times more goods for a hundred kroons than for a hundred kroons exchanged into shillings in Austria. This means that the actual purchasing power of the kroon is four times stronger than its nominal value and that the Estonia's PPPRR is 0.25 against Austria (or the kroon against the shilling).

Average Real Rate of the Kroon

A comparison with Austria may not be the most suitable choice for characterizing our situation. This is because Austria is not a considerable trade partner for Estonia. The ideological background of the PPPRR is close economic links in the course of which arbitrage will level off the prices in different countries. Therefore, the level of the real rate and changes in it should be determined by our major trade partners abroad.

In order to consider Estonia's major trading partners we have to weigh together the bilateral PPPRR calculated for each partner in accordance to the share of the respective partner country in Estonia's total foreign trade. Of course there are no analyses where Estonia would be used as a comparison basis. But still, it can be calculated using indirect methods through the indices for Austria. The data in Table 2 has been calculated this way.

The choice of countries in the table corresponds to the REER, including all Estonia's major partners. The average real rate of the kroon has been calculated as the weighted average of the bilateral real rates.

We can see from the calculations that the PPPRR of the kroon was 0.54 as based on private consumption, and 0.59 as based on the GDP. Or in other words, the actual purchasing power of the kroon in 1993 was nearly two times higher than its nominal value.

The multilateral PPPRR is higher than indicated by the comparison with Austria. This is because the share of transition economies among Estonia's trade partners is large. These countries have basically the same price level as Estonia and so the average PPPRR is higher than against one or several industrialized countries.

Real Rate in 1995

In order to calculate the 1995 real rate we used the REER which indicated an increase from the 1993 average level to the average level of 1995. We will get the 1995 real rate by multiplying the 1993 PPPRR by the REER (see Table 3).

The modest value of the REER against the backdrop of our considerably higher inflation rate can seem somewhat surprising. At the first glance, the relationship between inflation and the real rate of the currency with a fixed exchange rate seems to be fairly simple: the considerable domestic inflation results in the increase of the real rate. In other words, according to the standard understanding the main factor for the real rate is the difference between the domestic and outside inflation. In case of a fixed nominal rate this does not affect the real exchange rate. It is also important to mention that a precondition for the above conclusions is the low rate of outside inflation. Thus, high inflation a priori means a serious increase in the real rate of the currency.

In Estonia's case, the increase in the real rate of the kroon is slower than inflation rate because of the impact of the transition economies. The current ratio of the consumer price index (CPI) and REER is caused by the fact that parallel to the real rate of the kroon increasing against industrial countries it is getting cheaper against countries with transition economies.

We can see from our calculations that despite the increase in the real rate the PPPRR of the kroon was 0.64 as based on private consumption and 0.70 as based on the GDP. Thus, the actual purchasing power of the kroon still remained higher than its nominal value, exceeding the latter nearly 1.5-fold.

Real Rate in 1999

We have chosen the end of 1999 as the horizon of the real rate prediction. By the turn of the century the period of high inflation has probably ended and the annual rate of price increase has fallen below 10%, which should not pose any danger to a fixed exchange rate of the kroon.(17) been used in calculating the worst variant of the increase in the real rate of the kroon. It is based on the presumption that there are no deflationary cycles throughout the period during which the decline in demand would bring the prices down. At the same time it should be mentioned that the impact of economic growth or activity on the price change is asymmetric. Clark - Laxton - Rose (1995), Laxton - Meredith - Rose (1994) and Bruno (1995) all agree that economic growth and increase in demand have a clearly inflationary impact. However, the decline in demand does not necessarily lead to the decrease in prices although generally the drop of demand does have a deflationary impact. Unfortunately, the impact of the decline in economic activity is smaller than the effect of an equal increase in economic growth. Hence the asymmetry. First of all we are interested in whether the real rate of the kroon increases to such a level where its real purchasing power becomes lower than the nominal rate during the period of high inflation. Or, whether high inflation rate poses a danger to the fixed nominal rate of the kroon. If this is not the case then there is no cause for fear that later on - in the conditions of low inflation - the increase in the real rate would put pressure on the nominal rate of the kroon.

(17) According to the pessimistic forecast of the Eesti Pank Macroeconomic Research Department, the CPI will be 25% in 1996, 17% in 1997, 12% in 1998 and 10% in 1999. This pessimistic forecast has

We have used the same method as for the 1995 PPPRR for predicting the real rate of the kroon. The 1999 real rate has been calculated by multiplying the 1995 PPPRR by the REER. This operation presupposes predicting the 1999 REER (against the corresponding 1995 figure). This has been calculated very simply, on the basis of the CPI and the REER ratios in 1993-1995.

The use of the ratio of earlier years means that we predict the continuation of the past tendencies also in the future. In the present case it would mean supposing that the real rate of the kroon will increase at a slower rate than consumer prices because the REER depends heavily on the processes taking place in transition economies.(18)

(18) Actually, the bilateral real rate of the kroon does not have to decrease against the currencies of other transition economies at the present rate. This creates an error in the prediction. The REER will increase at almost the inflation rate of Estonia, which will boost the level of the real rate. A prediction error decreases the level of the predicted real rate and increases the reserve of its increase.

On the other hand, the coefficient we have applied takes into account the 1993 fall in the exchange rate of the US dollar against the German mark. In the future, the dollar-mark relations will probably remain the same or the dollar will become more expensive against the mark. Other conditions being equal, our prediction thus overestimates the increase of the real rate of the kroon, due to the supposed decline in the dollar and mark nominal rate, and reduces the reserve of the real rate increase.

According to the above logic, by 1999 the real rate of the kroon will increase 1.18 times from the level of 1995. In such a case, the PPPRR will be 0.76 based on private consumption, and 0.82 based on the GDP. This means that the actual purchasing power of the kroon is 1.3...1.4 times higher than the nominal.

Real Rate and Level of Development

Deviations from the law of one price can be explained by a concept under which the price level of a country is a function of its level of development.(19) The higher the level of development of a country the higher its price level, or the real rate of the currency.

(19) The above concept is a development of the convergence theory. See for example Sala-i-Martini (1994), Ben-David (1994), Heliwell (1990), DeLong (1988).

The same can be seen from international experience: international comparison programmes indicate that the price level of a concrete country corresponds to its level of development. Therefore, the price level of developing countries is lower than that of industrial countries.(20)

(20) At the same time goods in the sheltered sector are cheaper than goods in the open sector (Kravis, 1982). This is due to the fact that the bulk of the sheltered sector commodities are services and these are, as a rule, more labour-consuming. Since the cost of labour in developing countries is (very) low, the price of services, too, is low as well as prices in the sheltered sector in general. According to Kravis (1986) the generally lower price level of the developing countries first and foremost derives from the cheap commodities of the sheltered sector.

The above theory also has a direct expansion into the sphere of the exchange rate. Kravis (1986) emphasizes that the extent of the PPPRR deviation from the nominal rate decreases simultaneously with the development of the economy. The deviation is the biggest in developing countries and the smallest in industrial countries.

For Estonia, which is a less developed country, the low PPPRR is only natural. How low, is a different matter. Is the real rate of the kroon higher than suitable for Estonia's level of development (DRR)? If the PPPRR is smaller than the DRR there are obviously no problems, but when the PPPRR is higher than the DRR, this would point to a too high real rate. Since development is accompanied by the increase in the real rate the actual purchasing power of the kroon may turn out to be below the nominal rate if the starting level is too high.

In determining the real rate that would suit the level of development (DRR) we proceeded from international experience. Such an approach in itself is not unique. The price level has been determined proceeding from the level of development by Kravis (1986) using data of 34 countries and ICP Phase III, by Hansson (1990) using data for Asian countries and Canada, and by Richards - Tersman (1995) using data for 73 countries and the ICP. However, we discarded these models as we regarded a suitable background most important for Estonia and the above models describe the price level relying on data of countries far removed from Estonia.

For us obviously the situation of Estonia's major trade partners is important (providing that Estonian economy is due to its openness similar to that of the partners). Since the leading position among Estonia's industrialized trading partners is occupied by the Nordic countries where the price level is higher than the international average Estonia's "appropriate" price level, too, is higher than the international average.

Unfortunately, we do not have any data that would allow us to calculate the exact correlation for our trade partners. Therefore, we have calculated the function reflecting this dependency on the basis of data for the European countries. The price and development levels have been expressed against the average of 12 European countries, or E12. As an indicator of the level of development we have used the PC GDP of a concrete country divided by the 12 countries' average PC GDP. For the exchange rate we have used the PPPRR.

We calculated the function on the basis of the 1993 data for 24 countries drawn from Comparison (1995). The values of the parameters of the function together with statistical criteria have been given in Table 4. The theoretical value of the function indicates the price level corresponding to the respective level of development (DRR).

Real Rate Against E12

In order to use the DRR we also have to present Estonia's price and development level in relation to E12. The ratio for 1993 is easy to calculate by using Austria as a bridge.(21) In Table 5 we bring the real rate of the kroon against E12 (denoted as PPPRR(E12)) based on both private consumption and the GDP, as well as their geometric average.

(21) Initial data have been taken from Comparison (1995).

Evaluation of later periods is more complicated. We calculated the 1995 PPPRR(E12) by multiplying the 1993 figure by the index of the effective real rate of the kroon against industrial countries (marked with REER(L)).(22) 5.43%; the USA - 4.44%. Of course, we should have used the index of the kroon's real rate against E12 but unfortunately we did not have enough data for this.

(22) The REER(L) formula is analogous to the above-mentioned REER formula, with country shares as follows: Finland - 49.75%; Sweden - 17.04%; Germany - 16.56%; the Netherlands - 6.78%; Denmark -

The same technique has been used to find the 1999 price level: the PPPRR(E12) has been multiplied by the REER(L) of the period of 1995-1999. The latter figure is a prediction. The prediction is based on the above-mentioned CPI forecast. The REER(L) has been forecast in a very primitive way, with the help of the CPI and the REER(L) ratio in 1993-1995.

In order to find the DRR, we have to determine Estonia's level of development against E12. The figure for 1993 can be calculated using the comparison with Austria. The calculations indicated that Estonia's level based on the PC GDP was 23% of the E12 average.

Comparison of the Real Rate with the DRR

Until 1995 the real rate had not exceeded the DRR. In this sense, the real rate of the kroon is devalued. In 1999 the real rate is higher than the DRR determined by the 1993 level of development.

When interpreting the latter fact we have to take into account two moments. The above comparison is based on the level of development for 1993. Hopefully, Estonian economy has been developing at a more rapid rate than the countries it has been compared to. This will increase the DRR.(23)

(23) We have tried to evaluate indirectly the increase in the real rate as determined by such an economic development. If in 1996-1998 the annual development rate were 3% quicker then by 1999 Estonia's PC GDP would be 26% against E12. The real rate values corresponding to this level of development have been given in Table 5, column DRR(1999).

However, the change in the level of development is not important - it is too small. What really matters is that the model describes the average relationship and it can be used to get a point value corresponding to the average. Actually, the point value is too strict a criterion. This can be seen from Figure 1 where the variation of an empirical value against the theoretical one is quite significant. Thus, it is rather difficult to explain why does the PPPRR of the kroon have to be located on the curve of the function or below it, whereas the real rate of the currencies of a number of other countries is above the function curve.

It would be more logical to use an interval evaluation as a criterion for analyzing the level of the real rate of the kroon. The best suited for this is the reliability interval that allows to include a likely deviation from the one based on the function.(24)

(24) The deviation stems from the impact of real factors that have not been included in the function. Such factors include, for example, a list of causes that restrict the law of one price.

Table 5 depicts the reliability interval of the functions at the 23 and 26% level of development. As we can see, the price level based on individual consumption fits in with the reliability interval also in 1999. In this sense the real rate of the kroon corresponds to the E12 average, that is, it is not more expensive than could be expected from the level of Estonia's economic development.

Unfortunately, the PPPRR(1999) based on the GDP is still higher than the maximum of the reliability interval. Still, it is difficult to make any definite conclusions from it. The ambiguity of the evaluation is, on the one hand, caused by the use of the CPI instead of a GDP deflator for calculating the real rate index. However, this approximate variable is somewhat misleading because the GDP deflator is smaller than the CPI. Therefore, the CPI is overestimating the Estonian inflation and the increase in the real rate of the kroon caused by it. Since the actual inflation is lower, the PPPRR(1999) should also be lower. Unfortunately it is impossible to predict the GDP deflator because we do not have any past time series for it.

One more factor serves to prove our point. In the European countries the price level of private consumption is higher than the price level of the GDP.(25) In Estonia the situation is the opposite - the price level of private consumption is below the price level of the GDP. Let us suppose that by the year 1999 Estonia will have a price structure similar to that of the European market economies.(26) In that case, by 1999 the price level of private consumption should be higher than the price level of the GDP. This, in turn, presupposes that prices of private consumption will have to increase slower than other components of the GDP. Thus, we reached a conclusion mentioned above that the use of the CPI leads to the overestimation of the GDP price level during the period in question.

(25) Therefore, the level of development calculated on the basis of the average private consumption is below the one calculated on the basis of the GDP.

(26) The supposition is based on a precondition that the reduction of the price increase rate in Estonia to the internationally acceptable level also means the normalization of the structure of relative prices. For Estonia's relative price changes (a variation of the price change) and the inflation rate see Sepp - Viilmann (1995).

Conclusions

From the calculations in the present article we could see that the real rate of the kroon determined by the purchasing power parity against our trade partners in 1993 was 0.54 based on private consumption and 0.59 based on the GDP. In other words, the real purchasing power of the kroon was approximately two times higher than the nominal one.

Despite the increase in the real rate of the kroon in 1995 it was 0.64 based on private consumption and 0.70 based on the GDP. Thus, the actual purchasing power of the kroon remained approximately 1.5 times higher than the nominal rate.

By the year 1999 the real rate of the kroon will increase 1.18 times due to the continuing inflation. Thus, in 1999 the real rate of the kroon will be 0.76 based on private consumption and 0.82 based on the GDP, that is, the actual purchasing power of the kroon will be approximately 1.3 times higher than the nominal rate.

It thus appears that throughout the period of high inflation the real rate of the kroon will remain less than 1 and the actual purchasing power of the kroon will be higher than the nominal rate. This means that there is no cause to worry about the pressure on the nominal rate of the kroon because of the increasing real rate. Understandably, this conclusion is valid on the condition that the predicted inflation rate and the increase in the real rate of the kroon prove to be correct.

International experience indicates that the price level of a country corresponds to its level of development. Therefore, the price level of developing countries is lower than that of industrial countries. Until 1995 the real rate of the kroon had not exceeded the level determined by Estonia's development (the latter has been estimated on the basis of the experience of other European countries). In this sense, too, the real rate of the kroon is devalued but by the year 1999 it should be higher than the indicator of Estonia's level of development.

Still, the price level calculated on the basis of private consumption fits into the reliability interval. In that sense the real rate of the kroon is not higher than allowed by Estonia's level of development.

Unfortunately, the real rate calculated on the basis of the GDP is higher than the maximum value of the reliability interval. However, it is hard to make any definite conclusions from this. The ambiguity is caused by the fact that instead of the GDP deflator the CPI has been used for calculating the real rate index. But this approximation variable is somewhat misleading: the GDP deflator is smaller than the CPI. Therefore, the CPI overestimates Estonia's inflation and hence also the increase in the real rate of the kroon. Since the actual inflation is lower the real rate based on the GDP should also be lower. Unfortunately, it is not possible to predict the GDP deflator because we do not have past time series for it.

In conclusion we can say that by the end of 1999 the real rate of the kroon fits into the interval derived from Estonia's level of development, on the condition that the inflation rate and the predicted increase in the real rate prove to be correct. The level of the 1999 real rate is sufficient to guarantee that the high inflation accompanying economic development would not make the actual purchasing power of the kroon smaller than its nominal rate later on.

Urmas Sepp

Literature