MA EITEI (2008/2009)
EU Trade Policy in Relation to Central Europe and Its Impact on Local Market: the Case of the Czech Republic
By
Jan Posko č il
University of Economics Prague, the Czech Republic
Supervisor: prof. Giuseppe Celi (UB)
Date: 30
thOctober, 2009
2
Preface
The accession of the Czech Republic to the European Union on 1st May, 2004 deeply influenced the Czech political, economic and social environment. This dissertation is a reflection of it trying to look back and evaluating this reality.
I would like to thank very much my supervisor, prof. Giuseppe Celi, for his valuable advice, support and helpful comments.
I would also like to thank my brother for his beneficial ideas and the rest of my family for their support.
3
Abstract
This dissertation investigates the impact of EU Trade Policy and EU membership on new members which joined the European Communities on 1st May, 2004. The focus is on the Czech Republic and aim is to approach the issue also from point of view as seen by potential customer. The main analytical part tests the impact of import penetration on Czech price levels and price level convergence of food segment when compared to the EU27 average.
4
EU Trade Policy in Relation to Central Europe and Its Impact on Local Market: the Case of the Czech Republic
Outline
1. Introduction 2. Background and research strategy
2.1. The Czech Republic: EU accession negotiation and membership 2.2. Tariffs and other trade barriers
2.2.1. Impact of tariffs on prices
2.2.2. Other trade barriers and trade policies
2.3. Main changes and effects of joining the EU from the perspective of the Czech market
2.4. Literature and other aspects of the analysis
2.5. Specification of the analysis in order to single out the most relevant issues
3. Analytical part
3.1. Descriptive and comparative trend statistics analysis
4. Formalized analysis
4.1.Research Question
4.2.Data and software details 4.3.Segment selection
4.4. Indicator selection 4.5. Limits of the analysis
4.6. Empirical estimation (economic theory and price evolution) 4.7. Hypothesis
4.8. Indicators construction 4.9. Methods used
4.10. Results and interpretation
4.11. Synthesis and interpretation of results from comparative trend/descriptive statistic with formalized econometrical analysis
5. Conclusion
6. Sources
5
1. Introduction
It is not unusual that international matters and things happening in the EU get broad national news coverage and keep attention from both political elites and residents of individual countries. Even in present time in connection to the Lisboan treaty analysts and politicians argue about the benefits of the treaty and the EU as such. Since the same discussions were going on during the Czech candidacy and subsequent accession to the EU, it would be interesting to actually take a look at the arguments and doubts and compare them to the real situation afterwards. That is precisely the topic of this dissertation which will try to combine laic or popular arguments and opinions and put them against analytical and empirical
examination of the same issues.
To specify it more concretely this dissertation will take a look at the process of the EU integration¹ and with focus on the Czech Republic (CR) investigate the effect of the EU trade policy and the EU market as such on the Czech economy around its accession time to the European Union. The aim is to take a point of view from the customer’s perspective since that was the main concern of both public and part of academic society and see what the impact was. Since there is variety of options for the research areas and methods and virtually none is able to measure the impact in its full complexity, the focus will be on price evolution which seems to be very illustrative and most publicly concerned as well.
2. Background and research strategy
2.1. The Czech Republic: EU accession negotiations and membership
After the Velvet revolution in the year 1989 the Czech Republic quickly started to develop cooperation with the EU (respectively the European communities) and with the disintegration of Comecon the trade flows rapidly changed the direction from the former Soviet Union towards the new EU business partners, in particular Germany in the case of the Czech Republic. It was then already on 23rd January, 1996 when the Czech EU membership application was submitted to the Commission and by 31st March, 1998 the official pre-entry negotiations were launched.
¹ New countries entered EU in 2004: the Czech Republic, Cyprus, Estonia, Hungry, Latvia, Lithuania, Malta, Poland, the Slovak Republic and Slovenia. Bulgaria and Romania entered EU in 2007.
6
Table 1 demonstrates the major Czech trade partners and change in the international trade flows in period after division of the Czech Republic and Slovakia in 1993 until the year 2000.
The biggest share of the total foreign trade of the Czech Republic represented export and import to/from Germany most of all due to the logistics reasons/geographical closeness of both states and due to the huge trade potential of the German market.
The second biggest market for the Czech Republic became Slovakia. As the former part of Ex-Czechoslovakia, it was very deeply economically and socially connected with the Czech Republic.
The export to the Russian Federation had a decreasing tendency in the years 1993 – 2000. The Czech companies started to be more oriented on the European markets due to the economic reasons (limited entry barriers, more profitable business). On the other hand as the Czech Republic was historically depended on the deliveries of raw materials from Russia, the import from this country had increasing trend.
Table 2.1. The most important trade partners of the CR, trade volumes in mil. CZK in the years 1993-2000 Countries Imports
Exports
1993 1994 1995 1996 1997 1998 1999 2000
Austria Imports 32 756 40 147 45 981 47 959 52 384 54 356 55 676 61 332 Exports 25 427 32 915 37 323 38 954 46 059 53 189 59 434 66 956 France Imports 14 446 18 002 26 897 33 295 35 869 39 137 47 915 61 643 Exports 8 367 11 279 15 072 17 314 20 744 27 053 35 550 45 085 Germany Imports 124 102 149 446 209 668 241 526 275 351 314 402 331 889 400 538 Exports 122 406 156 439 209 579 218 874 256 006 320 780 381 198 453 521
Hungary Imports 5 181 5 151 5 783 7 608 11 282 12 779 15 709 19 894
Exports 8 669 10 639 9 997 10 641 13 522 16 096 16 246 21 011 Italy Imports 19 197 24 099 35 485 45 088 47 253 48 596 52 624 64 194 Exports 20 269 19 060 21 112 19 666 26 235 31 458 33 307 42 388 Netherlands Imports 11 242 13 525 17 562 18 665 21 779 22 212 23 434 29 018 Exports 9 580 11 171 13 610 12 370 13 867 18 827 22 199 25 779 Poland Imports 9 998 12 888 18 079 22 015 27 830 31 277 35 016 44 332 Exports 11 947 15 899 25 497 32 632 41 160 47 821 50 756 60 898 Russian
Federation Imports 37 307 36 382 49 778 54 384 58 723 50 950 48 146 80 237 Exports 17 306 15 784 16 726 18 840 24 319 20 986 13 186 14 915 Slovakia Imports 67 746 65 841 78 424 73 061 72 514 66 761 60 893 74 582 Exports 83 200 67 800 79 480 84 800 91 790 89 254 75 329 86 056 United Kingdom Imports 13 288 17 732 25 197 29 815 33 524 35 204 37 742 51 339 Exports 14 811 15 355 18 116 15 031 21 862 28 551 30 493 48 096 United States Of
America Imports 12 249 15 804 22 585 25 375 32 701 34 703 38 496 52 541 Exports 7 193 9 570 10 941 12 807 17 321 18 233 21 497 31 578 Source: Czech Statistical Office (CSU)
7
In my dissertation work some of the most important milestones in the EU enlargement process and the impact of the EU Trade Policy on the Czech market are described for more complex view of the analysis.
From an economic perspective one of those milestones was certainly foundation of
the Central European Free Trade Agreement (CEFTA)² in the year1992 in Krakow. It was a free trade agreement and its purpose was apart from the traditional goal of reduction of trade barriers and promotion of trade activities also something that could be called a test lab.
The member countries were supposed to learn to cooperate, harmonize their trade rules and gradually decrease the trade barriers within this block. One could say it was in a way
miniature of the European Union from the economic perspective and therefore its purpose was also to support the less developed countries and let the members states help each other and learn the functioning of such organization. Already in CEFTA there were more protective countries (Poland, Hungary) facing the more liberal ones like the Czech Republic cooperating together in order to achieve reduction of trade barriers and trade promotion. It is important to realize this because it shaped the economic activity during integration process and influenced the economic reality under which central and eastern European countries were entering the EU. While many trade barriers from mainly industrial production were reduced to zero, numerous products in mostly agriculture segments were protected almost to the last moment.
To be more illustrative and specific, it was for example cheaper to invest into frozen-food industry in Poland or Hungary rather than in the Czech Republic since the investor knew they would be able to export back to the Czech Republic easily unlike the other way around where they would face high toll. The main point is, nevertheless, that already by the time of Central and Eastern European countries (CEECs) accession into EU, many of the tariffs and other trade barriers have been in fact removed. (WTO Trade Policy Reviews)
It was already in the year 1995 when the Czech Republic received the so called White Book which was a document guiding the Central and East European countries (CEEC) on their way to EU membership. The economic transition and preparations were therefore gradual over a decade and this also influenced price evolution before the EU enlargement itself. Needless to say, the process of transition to market economy had far reaching impact for CEECs as well.
__________________
² CEFTA members in the years 1992-2004: the Czech Republic, Hungary, Poland, Slovakia, Slovenia, Romania, Bulgaria, Croatia
8
Graph 2.1. The Czech Republic trade with the EU in the years 1989 - 2008, mil.CZK
Source: CSU
Graph 1 demonstrates the trends of the Czech Republic trade with EU countries from the year 1989 (e.g. the year when the Communist Party lost its leading role in the Czech Republic and the democratic state system was established) to the year 2008.
The turnover of the Czech Republic had in this time period increasing trend with exception of the years 2001-2 and the year 2008. The positive impact of the accession of the Czech
Republic into the EU is demonstrated by rapid increase of the Czech Republic turnover with EU members and positive trend of foreign trade balance. The increasing trend of the turnover was interrupted again in the year 2008 due the worldwide crises.
As was noted above, the Czech Republic had in general quite liberal and therefore lower tariff system and being a small country it was of course less complicated as well. Therefore, the main change brought by the 2004 EU accession was the abolition of CEFTA and the Czech trade and customs tariff rules and accepting the new trade regime following the new EU trade policy rules and its customs tariffs. When talking about the trade within the European Union, this meant removing most of the remaining barriers for trade with goods and some services.
On the other side, it also meant accepting new rules for the trade with non-EU members. This was important for the Czech customers since some goods fell under new stricter trade rules and could become more expensive. For the Czech Republic probably the most painful example of this was the quota system on imports of aluminum ore used heavily in the Czech heavy industry. Unlike Hungary which received exemption on aluminum ore imports from
-500 000 0 500 000 1 000 000 1 500 000 2 000 000 2 500 000 3 000 000 3 500 000 4 000 000 4 500 000
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
Turnover Export Import Balance
9
Russia for instance, the Czech Republic did not managed to negotiate something similar.
Nevertheless, the common EU toll was also due to Czech effort reduced to 3% (from 4, 5%) and should reach zero eventually. (WTO, Trade Policy Reviews) Other exceptions were mainly from the agriculture sphere for example the notoriously known products such as bananas, sugar or diary products. Finally, even after the EU accession some areas within EU market remained regulated. Among those were for example: buying of agriculture land and other real estates in the Czech Republic (and most of the other CEECs). Later on specific segment of products will be chosen to see the issue from a perspective where the actual price effect is more easily traceable.
2.2. Tariffs and other trade barriers 2.2.1. Impact of tariffs on price
Tariff overview of the Czech Republic prior to the EU entry is showed in the following table 2.
The average level of tariffs of the below mentioned product categories was between 4-5 % in the years 1996 – 2003 and it shows the decreasing trend (UNCTAD).
Table 2.2. The Czech Republic tariff averages overview
YEAR 1996 1999 2002 2003
MARKET PRODUCT CAT.
Czech Republic
Manufactured goods, ores and metals 6.15 5.2 4.59 4.59
Ores and metals 1.77 2.33 1.23 1.23
Manufactured goods 6.38 5.35 4.76 4.76
Chemical products 5.25 4.42 4.18 4.18
Machinery and transport equipment 5.96 5.88 4.06 4.06
Other manufactured goods 6.94 5.4 5.39 5.39
Source: UNCTAD
When one takes into consideration mainly the tariffs, the actual price effect of the changes is hard to be traced. Reason for that is simple; most of the tariffs were between 4% - 5% (see the table 2 above). Due to the relatively low level of tariffs valid in the Czech Republic before the accession of the Czech Republic into EU there was only small direct effect on the prices from accepting new EU trade rules as such there. Change of 4% - 5% is a change which can be relatively easily out weighted by local market conditions or could be captured by large chain stores in their provisions due to the increased competition.
For the food resellers the margins are usually around 30%, therefore, the 4-5 % change can be absorbed or passed to producers. (Szczyrba 2009) From this example it is quite clear that
10
there is little sense in tracking price changes to tariffs only except to some specific cases.
Instead, this paper will take a look at the price indices in general and see if there was the feared price effect compared to the previous price evolution trend and other countries prices evolution at the time. This simple method used in the first analytical comparative trend statistics section can quickly show whether there was some noticeable price effect immediately after the accession and in the time period which followed.
2.2.2. Other trade barriers and trade policies
Apart from tariffs there are also other trade barriers which are less easily measurable. The EU frequently uses quota system for imports of the products. These quotas are subsequently distributed to importers/individual states. In certain cases special documents, certificates and test of products are required from importers. These documents have to be issued by
specialized laboratories or organizations and it is very costly and time consuming to obtain them. Production and export subsidies influence the price as well, since the exporters might choose to export their production rather than selling it at home. The barriers which are the most hard to measure are mainly different technical barriers and requirements of origin for tariff calculation. Measurement of tariffs and non-tariff barriers is simply quite complicated and Anderson–Neary Trade Restrictiveness Index (TRI) is mostly used for this. (Bach 2001) For the purpose of this dissertation there is, however, little benefit in developing the index.
The expected effect is rather limited and the thesis focus is on other aspects of price evolution.
2.3. Main changes and effects of joining the EU from the perspective of the Czech market
The main change from the perspective of the Czech market was clearly the fact that the Czech economy became part of the currently biggest economy in the world. (CIA fact book, 2009) This means direct access to the huge market for both producers and consumers and at the same time it means greater competition. According to economic theory and European Commission this should eventually lead to greater choice and lowering of prices. There should also be gradual price convergence until the price differences reflect only the specific characteristics of the individual markets (Balassa-Samuelson effect, transportation cost, differences in product quality). There are therefore two aspects here. For trade within EU and with EU partners there is free flow of goods and most services. For trade with countries outside the European Communities the Czech Republic falls under the EU trade policy rules and this could mean change of structure of the international trade with those countries.
11
Economic theory would therefore suggest trade diversion from the non-EU countries towards the EU partners which have no trade barriers. Another aspect is greater labor mobility and greater inflow of FDI. For foreign producers (investors) the fact that the Czech Republic is in the EU means that they can sell from there to the whole EU market without additional
restrictions. There are simply many aspects and not all can be properly estimated. The goal of this paper is therefore not finding complex causal nexus but rather testing if there is some relationship between price evolution, changes in trade rules and level of economic integration of the Czech Republic into the European market.
2.4. Literature and other aspects of the analysis
Literature dealing with the new EU member states and their accession to the EU is quite broad. Nevertheless, when one wants to search for more specific price analysis the number of papers is surprisingly very limited. Author dealing with this specific subject is for instance Gabor Pellenyi. (2008) In his paper he analyses panel dataset of prices in the European manufacturing industries. He finds small negative, but significant effect of industry import penetration on prices. Unfortunately, for the new EU members this significance does not hold around the date of entry suggesting some other effects playing more important role. His suggested explanation for that is that countercyclical movements of markups might have played a larger role in explaining manufacturing price levels evolution on those markets.
There are also other authors writing about price evolution in the different EU countries such as Wolfgang Pollan (1996) for instance. In his paper “The Effects of Austria's Accession to the EU on Consumer Prices” he found that prices in Austria dropped and especially in food segment those lower prices were soon passed to the customers. The problem with using the analogy is that Austria, when it entered into the EU, was in a different situation compared to the Central and East European countries. Austria, unlike the Central and East European ex- communist states, was a country with developed economic and social democratic system.
Another key paper from the problematic is “Price Convergence in the Enlarged Internal Market” written by Dreger et al. (2007). This thesis partially builds on his ideas and develops some of them more specifically for the case of Czech Republic. In his results he concluded finding certain degree of price convergence, however, with very slow pace with half-lives of approximately ten years. Another important paper was written by Chen, Imbs and Scott. The paper is called “The Dynamics of Trade and Competition” and in it they also find small but significant price-reducing effect of import penetration for the manufacturing sector. This study is conducted for Western Europe and period of 1989 to 1999 so it unfortunately cannot
12
be directly compared to the analysis tested in this dissertation. In summary, there are many papers dealing with price convergence and evolution. Nonetheless, for the case of EU enlargement in 2004 there is as far as to my knowledge no other study which would use similar approach as was used in this thesis. There are also practically no recent analysis for the price evolution after the EU enlargement for this region.
Even among the CEECs there are major differences and specifics so one always needs to distinguish individual countries. As for the price evolution and effects there are also number of Czech authors writing about it. Unfortunately, they are from ex-ante perspective or are covering earlier time period only. For international reader another difficulty comes with the fact that they are published only in the Czech language many times as well. Regardless, they are very useful for a proper analysis since they bring local insight into the problem. Finally, when dealing with the sources one too encounters some issues with integrity of the data and comparability prior and after the accession. Even the mere fact of currency variations makes the analysis less clear complicating the process of estimating the economic effect since they may be hidden or delayed.
For an illustration the following graph 2 shows variation of the Czech export of services and products in comparison to the real effective exchange rate between CZK and EUR in the years 2001-2007 in %.
Graph 2.2. Red line represents exports of goods and services, blue line represents the real effective exchange rate
Source: CSU
¹ The key export markets for the Czech Republic are the EU countries. Germany is the significant one.
This is the reason why export of the products and services is most of all realized in EUR.
13 Comments to the graph 2:
The foreign trade from the Czech Republic is in most cases realized in EUR¹. The exchange rate between EUR and local currency influences the foreign trade significantly.
Each appreciation of the local currency (CZK) increases the price of products exported from the Czech Republic and decreases the export of them. Therefore the lines in the graph 2 change their directions at the same moment and they have opposite trends.
2.5. Specification of the analysis in order to single out the most relevant issues In the process of searching for the best way to address the issue of economic effect estimation one encounters following problems. The first question is in fact how to see and show the impact on the local market. At first I wanted to analyze impact on the Czech customer before and after the accession and changes which come with the EU membership and deeper
integration of the Czech Republic into the EU economy. Defining customer is quite
complicated so I had to choose a specific area where to show the changes and effects brought by the EU membership. The most logical path was to take a look at the price levels before and after the EU accession and as well attempt to trace price development to specific changes in trade rules during the pre-entry harmonization process and after the 1st of May 2004 and subsequent gradual integration of the Czech Republic into the common market. In the first part of my analytical section I am using general price indices and show the development of the prices in relation to the EU and other countries taking into account the general state of economic changes in the Czech Republic.
The first problem with this approach comes directly from its excessively general nature.
Looking closer at the products one will finds that each one has its own evolution and quite specific factors which influence the price of it. The initial presumption that price changes can be mostly traced down to the evolution of tariffs and joining the common EU tariff and trade rules system proved to be premature. As the data show, the truth is that entering into EU trade zone and accepting its rules had in fact much less significant impact then one would expect, at least from the perspective of prices.
In the case of industrial products the tariff burden was due to preferential agreements with EU very low both prior and after the accession and even in the case of agriculture products and services the change was surprisingly less significant than one would expect. Clearly there are exceptions to this and they would be therefore interesting for closer investigation. In the case
14
of the Czech Republic those areas include mainly agriculture products such as sugar, diary products, wine and import duties on aluminum ore as was already mentioned. As for most of the other products there are further effects at work that influence the prices and they cannot be fully controlled for. Looking at the evolution of prices or price indices of certain segment for that matter is interesting and illustrative but the truth is that one cannot ever be sure what the whole story behind the price changes is. That is the reason why I do not claim that the price change occurred because of some specific event. Instead I search for connection between different economic indicators and compare it to other countries to investigate their price evolution during the same period considered. Nevertheless, since both common EU tariffs and trade rules do not have such high significance in price evolution as expected there
immediately comes the other side of the coin. The Czech Republic joining of the EU meant that we entered the biggest world market. In theory, the greater the market is the greater the competition and the lower the prices should be. On the other side, economic theory also suggests gradual convergence of prices so in the end the price difference should more or less represent only the specifics of the markets plus transportation costs. In the next section I will investigate the price evolution and use the comparative trend and descriptive statistical analysis to estimate the situation and price effect. I will then use time series model to test it and see whether an econometrical dependency can be found using formalized analysis.
15
3. Analytical part
3.1. Descriptive and comparative trend statistics analysis
Firstly the descriptive statistics is used to see the price development. Generally speaking if there would be a price effect connected to the Czech Republic’s accession to the EU and its trade policy rules even looking at a simple price evolution matrix or graph should easily unfold whether this is true or not. On the other side, just the price evolution does not have the power to explain what the reason behind the price changes is. Even in the simplest case one should consider the general state of the economy and control for other possible effects which would have the power to influence the price evolution in some major way. However, the purpose of this dissertation is not to develop a general equilibrium model but instead combine descriptive statistics with trend comparison between regions and then investigate whether it is possible to find certain causal relationship between the indicators which were expected to play a role in the price development.
For the reasons of high volatility and changes in exchange rate (gradual appreciation over time) between Czech koruna and both Euro and the Dollar I consider prices denoted in euro and use the Harmonized Indices of Consumer Prices (HICP) for more precise comparison between the countries. Firstly, a graph showing comparative price indexes evolution is shown.
The EU15, EU27 and the Czech Republic price indexes are denoted in the same measurement units so one can clearly see the differences between prices and how did the prices evolve over the time period of 1999 – 2008. Already from this is easily visible that prices of both EU15 and EU27 grew comparatively faster then prices in the Czech Republic which is interesting.
The price index lines are separated into period until 2004 and period after the year 2004 when ten new EU member states joined the European Communities. This allows the comparison of trends for those two periods and trends between the Czech Republic and EU. While the trends of EU15 and EU27 remained very similar before and after the accession which is to be
expected, the general trend of Czech Republic changed with prices growing faster for the period after the EU accession in 2004. Nevertheless, as seen from the equations representing trends for EU15 and CR the price divergence continues with slower pace only. The short time period clearly does not allow for any confident conclusions but it representatively shows the situation for the time period considered.
16
Graph 3.1. comparative price indexes evolution of CR, EU15 and EU27
Source: EUROSTAT, generated by MS Excel Note: lineární = linear
Using the similar trend comparison for the food segment in order to complement the formal analysis later on one can see that the general outlook is quite similar. The difference is in the greater variability of prices. There is also increase in period starting in mid 2007 which is much larger then increase in the EU index. Again, Czech economical situation and changes in political orientation bringing certain economic reforms play their role here.
y = 0.192x + 87.95
y = 0.173x + 89.20
y = 0.155x + 45.26 y = 0.105x + 50.08
40 50 60 70 80 90 100 110 120
1999M01 1999M06 1999M11 2000M04 2000M09 2001M02 2001M07 2001M12 2002M05 2002M10 2003M03 2003M08 2004M01 2004M06 2004M11 2005M04 2005M09 2006M02 2006M07 2006M12 2007M05 2007M10 2008M03 2008M08 2009M01 2009M06
EU27 after EU27 before EA15 after EA15 before CZ after CZ before
Lineární (EU27 after) Lineární (EU27 before) Lineární (EA15 after) Lineární (EA15 before) Lineární (CZ after) Lineární (CZ before)
17
Graph 3.2. Comparative food price indexes evolution of CR, EU15 and EU27
Source: EUROSTAT, generated by MS Excel Note: lineární = linear
For direct measurement of price convergence between new EU member states and selected old EU members as well it is more illustrative to demonstrate price convergence to the changing EU27 price level directly. On the Gretl generated graph 3.3. based on the data from formal analysis part the general trend of slow convergence is apparent for most of the
countries. The Czech Republic is a specific case in this sense. Very probably this has a lot to do with the local market situation with great competition among large groups of retail chain stores, the local market specifics which are mentioned in the formal analysis part. Graph itself is very intuitive and probably the only other interesting part worth pointing out is the strong seasonal effect in the case of Greece which is very specific compared to all the other countries in this sense.
y = 0.256x + 88.17
y = 0.258x + 92.2
y = 0.169x + 54.53 y = 0.023x + 63.37
0 20 40 60 80 100 120 140
2000M01 2000M07 2001M01 2001M07 2002M01 2002M07 2003M01 2003M07 2004M01 2004M07 2005M01 2005M07 2006M01 2006M07 2007M01 2007M07 2008M01 2008M07 2009M01 2009M07
EU27 after EU27 before EA15 after EA15 before CZ after CZ before
Lineární (EU27 after) Lineární (EU27 before) Lineární (EA15 after) Lineární (EA15 before) Lineární (CZ after) Lineární (CZ before)
18
Graph 3.3. General Price index evolution of selected EU member states (EU27=100)
Source: EUROSTAT (generated by Gretl)
Comments to the graph 3.3. General Price index evolution of selected EU member states:
Price indexes´ convergence - value “100” is the price level of EU27.
States:
BG – Belgium, CZ - the Czech Republic, DK – Denmark, EE – Estonia, IE – Ireland.
GR – Greece, LV – Latvia, LT – Lithuania, HU – Hungary, AT – Austria, PL – Poland, PT – Portugal, RO – Romania, SI – Slovenia, SK - Slovakia
0 20 40 60 80 100 120 140 160
2000 2002 2004 2006 2008
BG__TotalPriceI CZ__TotalPriceI DK__TotalPriceI EE__TotalPriceI IE__TotalPriceI GR__TotalPriceI LV__TotalPriceI LT__TotalPriceI HU__TotalPriceI AT__TotalPriceI PL__TotalPriceI PT__TotalPriceI RO__TotalPriceI SI__TotalPriceI SK__TotalPriceI
19
The last thing which needs to be mentioned here is the general evolution of Czech GDP and international trade with imports and exports growth. Without much need for explaining it is clear that GDP growth was larger then the growth of international trade and as we will see in the next section this makes a difference in the formal analysis and overall results
interpretation. Other graphs regarding the food price convergence between EU15 and CR as trends of simple difference and ratios as well are included in the appendix.
Graph 3.4. Czech Republic GDP and international trade evolution (in millions of Euro)
Source: EUROSTAT, generated by MS Excel 0
2000 4000 6000 8000 10000 12000 14000
1999M01 1999M10 2000M07 2001M04 2002M01 2002M10 2003M07 2004M04 2005M01 2005M10 2006M07 2007M04 2008M01 2008M10
GDP before [Meur]
GDP after [Meur]
Export before [Meur]
Export after [Meur]
Import before [Meur]
Import after [Meur]
20
4. Formalized analysis
4.1. Research question
In the last formalized analysis part of the dissertation empirical test is performed in order to investigate price evolution and its explanatory factor related to the accession into EU. Unlike in the case of previous enlargements new EU member states were transitional economies and therefore both price convergence towards higher EU price level and pro-competition effects lowering prices are to be expected at the same time. For this reason the price level change is not taken directly but instead it is considered in a form of price convergence, price ratio between the Czech Republic and EU27. On the other side, explanatory variable tested is the role of increasing international trade with EU members connected to EU membership and EU trade policies. International trade measure is specified as an import penetration in the chosen food segment. This is related not to GDP created by food directly but instead to apparent consumption in this segment calculated by food segment contribution to overall GDP plus food imports minus the exports. It is expected that positive increase of food import
penetration explanatory variable should bring positive impact on the price convergence measure. In other words, the greater is the food segment import ratio to the apparent consumption, the smaller the difference between Czech and EU27 prices should be.
4.2. Data and software details
Data used for the dissertation originate from various sources.
• The data for the main econometric part were obtained from EUROSTAT due to the fact that the statistics are more easily comparable in between different countries and regions.
• In addition to it data from the Czech Statistical Office and the Czech Statistical Database of International trade are also used.
• Mainly for the overview statistics there are also data obtained from statistical yearbooks available in the Czech statistical office library.
• Interactive collection Global Market Information Database GMID was also consulted.
• For the purpose of trade policies WTO trade policy reviews and tariff database was used.
• Last but not least the OECD statistical database and IMF databases were used in the process of searching for the best suitable data.
21
• For purely Czech regional data from the Czech ministry of Foreign affairs, Czech trade and INCOMA GfK s.r.o. (Ltd) retail market research were supplementing for otherwise unavailable statistics.
For the reasons of ensuring the same data collection methodology and measurement units, indicators and statistics used for calculations in the main econometrical part come from EUROSTAT. More concretely those are Harmonized Indices of Consumer Prices (HICP) for individual countries and EU27, Comparative price level indices (with EU27=100), data on inflation evolution, international imports and exports and GDP data details. Due to
availability issue and also in order to avoid possible distortive effect such as the Czech economical downturn in 1998 or financial crisis effectively bringing distortions from the years 2008 - 2009 already, the monthly data are extracted for the period of year 1999 until the year 2008.
Data are not seasonally adjusted until they are integrated into the GRETL model and used for the analysis. Lastly, the data extracted are for most of the new EU member states and also some old EU members of similar sizes or economic properties for comparison. Only the specific ones are used for the dissertation output directly but they are all employed for the general insurance of trend and methodology appropriateness and overall better overview of the problematic.
Software used is mainly Microsoft Excel for the more basic analysis and data preparation.
For the data testing and modeling itself Gretl (version 1.7.9.) econometric software is utilized.
For graphical and data outputs both of those programs are used in supplement with official charts from primary statistical sources.
4.3. Segment selection
Both the overall general price indices and more specific food segment are employed in the analysis. Narrower segment selection is chosen for the reason of greater comprehensibility and also in order to avoid the extreme complexity of general price index. Food segment selection itself comes from the experience of other authors such as Wolfgang Pollan (1996) analyzing similar problematic for other countries and is also chosen for its high relevance to the price evolution problematic as viewed by general public. It is to be expected that in the non-durable goods segment for immediate use (such as food) and also due to the large
competition on many markets in this area the price adjustment will be passed on to consumer
without excessive lags between the investigated explanator
change of price index. The food prices are adjusting quite quickly and the segmentation of the analyzed Czech food retail market is in general (Prague being the only partial exception) very limited compared to most of the o
price. This holds for majority of population across different income groups. ( shopping monitor 2009) When mentioning
should be also taken into consideration. That is the high level of competition among retail sellers and their significant dominance over smaller mostly individual suppliers. This can be traced to the increasing role of large retail chain store groups
price competition. For an illustration
sellers in the Czech Republic and their growing market share. Proxy (market strength of large retail groups)
detailed and very expensive commercial data supported be extended specialist analysis the potential of this market proxy
disturbances of food price evolution is dealt with by using p and filter when necessary.
Graph 4.1. Market share in sales of 10 top retailing chain store groups with non Republic
Source: INCOMA GfK, s.r.o.
0 10 20 30 40 50 60 70
Sales of 10 top retailing chain store groups with non
22
without excessive lags between the investigated explanatory variable change and the actual change of price index. The food prices are adjusting quite quickly and the segmentation of the analyzed Czech food retail market is in general (Prague being the only partial exception) very limited compared to most of the other developed EU states with the main emphasis being on price. This holds for majority of population across different income groups. (
) When mentioning the specifics of Czech market another factor sideration. That is the high level of competition among retail sellers and their significant dominance over smaller mostly individual suppliers. This can be
ole of large retail chain store groups and their influence
ce competition. For an illustration I enclose graph (4.1.) of growth of the 10 top retail sellers in the Czech Republic and their growing market share. Proxy for market conditions
of large retail groups) could be added into the analysis as well
detailed and very expensive commercial data supported be extended specialist analysis the would be limited. Finally, the disadvantage of seasonal
disturbances of food price evolution is dealt with by using price index convergence indicator
ales of 10 top retailing chain store groups with non-durable goods in
Sales of 10 top retailing chain store groups with non-durable goods in CR
Market share %
y variable change and the actual change of price index. The food prices are adjusting quite quickly and the segmentation of the analyzed Czech food retail market is in general (Prague being the only partial exception) very ther developed EU states with the main emphasis being on price. This holds for majority of population across different income groups. (Incoma,
the specifics of Czech market another factor sideration. That is the high level of competition among retail sellers and their significant dominance over smaller mostly individual suppliers. This can be
and their influence on greater of growth of the 10 top retail
for market conditions well but without detailed and very expensive commercial data supported be extended specialist analysis the
he disadvantage of seasonal
rice index convergence indicator
durable goods in the Czech
23 4.4. Indicator selection
The two key indicators employed in the analysis itself are selected in order to keep in mind the specifics of the issue which is under testing and the regional specifics as well. Unlike in the case of previous EU enlargements the new accessing countries exert not only vast economic differences but they are also tested shortly after their transition into the regular market economy. Even though the transition could be considered as a finished for the most part in the case of the time horizon tested (1999-2008), it still cannot be disregarded fully.
This issue will be also discussed in the methodology part as well. For those reasons the price evolution dependent variable chosen is not only the general and food segment price level evolution but instead the comparative price convergence of individual countries towards the EU27 average. This ensures that both possible price effects, so called catching-up effect and the pro-competition effects, working generally in opposition to each other will be taken into consideration. As an explanatory variable the indicator showing the general EU import penetration compared to overall apparent consumption and food segment import penetration are put into use. Reasoning for the selection and its limitations are also briefly discussed in the methodology part. GDP measure is adjusted to show the apparent consumption consisting of GDP plus added general or specific segment imports minus outgoing exports for their better explanatory power of the actual consumption within the Czech Republic.
Table 4.1: Food price evolution of selected EU member states ( general price level convergence in graph 3.3.) Comparative food prices overview - all countries
country Indicator mean minimum maximum standard deviation
Belgium FoodPriceIndex_BG 54.443 44.799 67.773 6.1229
Czech Rep. FoodPriceIndex_CZ 66.772 61.573 72.726 2.826
Denmark FoodPriceIndex_DK 142.77 137.12 152.01 3.7276
Estonia FoodPriceIndex_EE 72.975 66.8 84.754 4.9709
Ireland FoodPriceIndex_IE 126.42 116.05 132.84 3.766
Greece FoodPriceIndex_GR 93.578 88.507 100.14 2.2154
Latvia FoodPriceIndex_LV 63.75 53.769 85.638 10.224
Lithuania FoodPriceIndex_LT 64.547 56.59 75.547 5.1701
Hungary FoodPriceIndex_HU 74.222 64.342 92.511 7.4896
Austria FoodPriceIndex_AT 112.94 109.65 118.58 2.3051
Poland FoodPriceIndex_PL 62.399 57.993 65.872 1.8469
Portugal FoodPriceIndex_PT 88.219 80.05 92.262 2.6671
Romania FoodPriceIndex_RO 63.064 41.385 72.754 8.6337
Slovenia FoodPriceIndex_SI 88.693 80.857 94.482 3.4617
Slovakia FoodPriceIndex_SK 62.82 58.938 65.37 1.3387
Source: EUROSTAT data file, data tested in Gretl
24 4.5. Limits of the analysis
As was already noted above any estimation or model, regardless of its complexity, which will be used for analyzing the impact on prices will be essentially suffering by the limited time span of data available. It is also a little controversial to test a wide range of data across various countries and products since each is individually different and would therefore need specific treatment. This constitutes the main dilemma of the dissertation. If one wants to answer the question of the impact of EU trade policy on the Czech Republic while it still has some relevance and recency there will always be only limited available data for traditional
verification. In this dissertation I am well aware of that because that is essentially the reason why there are as far as to my knowledge no similar ex-post studies of price evolution in new EU member states yet. That also clearly predestines the way this thesis approaches the analysis and interprets the results. In simple words, one could say that this analysis should be interpreted more as an early inspection into the issue instead of a typical complex empirical proof in that sense. Its purpose is to attempt to answer the underlying question using
combination of descriptive and comparative trend statistics supported by regular formalized econometrics, regardless of its limited applicability. In fact it would be indeed interesting to conduct the same experiment also later on when a more reasonable number of years will be available and see if the results would exhibit similar features. The unfortunate reality is, although, that the crisis will be quite disruptive for studies in near time horizon as well.
4.6. Empirical estimation (economic theory and price evolution)
Regular economic theory point of view is that greater integration and bigger market cause greater trade flows and more competitive environment. This should subsequently bring greater effectiveness and lower prices in the end. Import duties and other trade barriers abolishment in integrated EU market under the common EU trade policy could be compared to reduction of transportation costs for example. This is the case for the trade within the EU market, as for the trade with non-EU members the situation would be clearly different.
Nevertheless, for majority of products in tested segment EU27 is by far the predominant business partner. Under imperfect competition point of view the foreign competitors should cause price reduction by reducing their competitor’s mark-ups. With firms viewed as being heterogeneous, on the other hand, competitors would be driving the least efficient producers out of the market, therefore improving the average productivity and reducing prices on the market. Where economies of scale are present bigger market clearly enables their better usage as well. (Pellenyi 2009) One could argue that greater concentration of production can lead to
25
higher prices eventually but this would be most probably the case for a medium to long term analysis therefore this paper does not take it into consideration.
In the case of recent EU enlargement, however, this argumentation cannot be taken as such since there are large differences between old EU member states and the new member states in both price levels and overall structures of their economy. On top of that, the Czech Republic and most of the other new member countries except Poland are rather smaller countries so specialization into different sectors of economy plays important role as well. Estimation of all of the above mentioned effects would be at least very complicated, if not impossible. That is the reason why is this dissertation not trying to prove exact causal relationship but instead takes a look at the price evolution and investigate if there can be a relationship between its price convergence and the level of import penetration established.
In the case of recent EU enlargement there are basically two large effects at work here both influencing the price levels in different direction. If one would like to use terminology of Dreger (2007) those would be called the pro-competition effect and catching–up effect.
The first one reflects the traditional higher competition effect with its price lowering effect as was discussed in the first paragraph. The second one, catching-up effect, reflects the increased inflation in the new member states economies bringing gradual price convergence between the old and new member states. Bearing this in mind, this paper does not take only the general price levels as such, since they would be appropriate rather for countries of similar economic state or phase. To reflect the difference between economies, index showing the price
differentials (ratios) between EU27 and the Czech Republic price levels is used. The
hypothesis to be tested is that indeed it was the increasing trade with EU member states due to joining the common EU trade policy that was and is influencing price harmonization and convergence with its speed. Both general indicators and specific segments were tested but focus is only on the results for the food segment. For the specific food segment one needs to have comparable segment price levels and also measure indicating the size of imports from EU compared to the size of the local Czech market. For greater comprehensibility this
indicator will be called import penetration even though it does not fully overlap with indicator used by Pellenyi in his paper “Price Effect of the Internal Market” for instance. As one can see from the table 4.2., the food import penetration measure exhibits issues of imperfect comparability between segment categories available for indicators. For the analysis purposes, nevertheless, it is the evolution that matters and that is predicative enough.
26 4.7. Hypothesis
The hypothesis under testing is that explanatory variable import penetration should exert some degree of positive impact on the dependent variable, price convergence between the Czech Republic and EU27 price levels. Import penetration is always measured against the apparent consumption in the segment and price levels are for food segment as well. The chosen significance level for results acceptance is 5%.
Table 4.2. Main data statistics for the Czech Republic
Source: EUROSTAT data file, used in the Gretl modeling
4.8. Indicators construction
Dependent variable, price convergence measure is calculated by combining two indicators.
Firstly, the Comparative price level (CPL) indices for EU27 and the Czech Republic are taken. This shows the proportion of Czech prices to EU27 prices each period. EU27 prices work as a benchmark in this sense. The relationship between EU27 and Czech prices each year does not have the time dimension which would show changes over time then. For that the index needs to be combined with the index of price evolution in the specific food segment.
The final indicator is afterwards already capable to show the changes of prices and ratios over time between EU27 and CR. In the beginning it was showed that the volatility of exchange rates was great therefore the prices indexes used are denoted for the same currency.
Data overview - Czech Republic (1999-2008)
Indicator mean minimum maximum standard deviation
CZ ExTotal_V_GDPTotal_CZ 0.21396 0.14062 0.31101 0.038033
CZ ExFood_V_GDPTotal_CZ 0.03671 0.019803 0.058484 0.010451
CZ ExFood_V_GDPFood_CZ 1.2541 0.76644 1.7457 0.21427
CZ ImFood_V_GDPTotal_CZ 0.026388 0.014162 0.043423 0.0071578
CZ ImTotal_V_GDPTotal_CZ 0.17778 0.12939 0.25519 0.028351
CZ ImFood_V_GDPFood_CZ 0.9052 0.54063 1.2558 0.14968
Appar.Cons.Food (mil. Euro) 141.03 53.5 277.7 47.517
GDP_food segment (mil. Euro) 213.4 154.8 335.67 36.807
Food_Market (Appar.Cons.) 141.03 55.5 277.49 45.492
CZ FoodPriceIndex_CZ 67.06 63.398 72.726 2.7295
CZ TotalPriceIndex_CZ 58.257 56.656 60.355 0.84436
Food/GDP year change 0.6095 -28.936 51.16 15.666
27
The regressor used is slightly more complicated to get since more things need to be combined.
Firstly apparent consumption is calculated by adding GDP plus imports minus exports for both overall economy and specific food segment. Then EU import to apparent consumption ratio is taken for both overall economy and food specific segment. This basically shows how much the general and segment import weights compared to apparent consumption indicator each year. In order to single out trade flows from EU27, imports from EU member countries only are taken while for apparent consumption imports and exports for all countries are used.
Combining those statistical data and using correct measurement units with comparable currency we have the final indicator. This indicator will be for simplicity denoted import penetration from now on and it basically shows the role that EU imports played in the general economy and specific segment tested. Nonetheless, because import penetration explanatory variable is considered as a ratio between Czech Republic and EU27 price levels, only one form of their possible relationship is explored. For this reason, import penetration measure is in addition tested separated into its two components: import size and size of the market.
4.9. Methods used OLS
Firstly, the data are preliminary tested using the simple ordinary least square (OLS) regression model. Without using any additional filter are the results not convincing. (Appendix table 5.1.) For the Czech Republic versus EU27 I got positive coefficient 1,66 for the import penetration in the food segment. The p-value suggests results are still within our significance level. Interestingly enough, one can see a high positive relationship between dependent variable and the explanatory import penetration variable. In every case, OLS is used only as an initial test here since it is usually not suitable for time series application and can exert seemingly good results especially when considering t-statistics and r-square results even though this is not true. Testing normality of residuals and for heteroscedasticity showed the results are not appropriate. Nevertheless, since the import penetration variable was used here and time element observed is limited it certainly bears some informative value for further testing as well.
28 Time series
For more advanced analysis which enables us to take into consideration the time dimension while testing price convergence between the Czech Republic and EU27 and its explanatory variable as well, the time series ARIMA model is implemented. Apart from the desirability of time dimension effects the choice of model comes also mainly from the requirements on the analysis. There are lags to be expected between the explanatory variable, import penetration change and impact on the price convergence and those lags may vary across time as well. By choosing ARIMA with option of first and possibly second degree differentials not only the effect but also the change and the change of speed of change over time can be seen.
Table 4.3. Comparison of model parameters
Source: EUROSTAT data file, model generated by GRETL
Above is the table with results for different parameter settings for illustration. Since we are investigating trend evolution rather then the cyclical component of the data moving average side of the ARIMA model is less useful. The second degree differential option does not help the model’s explanatory capabilities so it seems longer period of time with better properties would be needed or simply there is only a little effect of change of speed to be traced. When I subsequently test ARIMA with the best properties setting AR order to one and difference order to one as well the results from the model are significant and show very small positive coefficient of 0,231 within our chosen significant level. (Results appendix table 5.2.)
Num. Model Dependent variable Independent variables 1 AR order Difference MA order HQC BIC AIC 1 ARIMA CZ_FoodPriceIn Import food / agriculture market 0 0 0 527.641 532.441 524.367 2 ARIMA CZ_FoodPriceIn Import food / agriculture market 1 0 0 193.778 200.177 189.412 3 ARIMA CZ_FoodPriceIn Import food / agriculture market 2 0 0 178.072 186.072 172.615 4 ARIMA CZ_FoodPriceIn Import food / agriculture market 0 1 0 181.24 186.024 177.978 5 ARIMA CZ_FoodPriceIn Import food / agriculture market 1 1 0 167.839 174.217 163.489 6 ARIMA CZ_FoodPriceIn Import food / agriculture market 2 1 0 169.188 177.161 163.75 7 ARIMA CZ_FoodPriceIn Import food / agriculture market 0 2 0 203.274 208.042 200.024 8 ARIMA CZ_FoodPriceIn Import food / agriculture market 1 2 0 200.1 206.457 195.766 9 ARIMA CZ_FoodPriceIn Import food / agriculture market 2 2 0 200.659 208.605 195.241 10 ARIMA CZ_FoodPriceIn Import food / agriculture market 0 0 1 413.057 419.456 408.691 11 ARIMA CZ_FoodPriceIn Import food / agriculture market 1 0 1 179.196 187.196 173.739 12 ARIMA CZ_FoodPriceIn Import food / agriculture market 2 0 1 180.292 189.891 173.743 13 ARIMA CZ_FoodPriceIn Import food / agriculture market 0 1 1 168.265 174.644 163.915 14 ARIMA CZ_FoodPriceIn Import food / agriculture market 1 1 1 170.038 178.011 164.6 15 ARIMA CZ_FoodPriceIn Import food / agriculture market 2 1 1 170.716 180.284 164.191 16 ARIMA CZ_FoodPriceIn Import food / agriculture market 0 2 1 190.995 197.352 186.661 17 ARIMA CZ_FoodPriceIn Import food / agriculture market 1 2 1 178.365 186.311 172.947 18 ARIMA CZ_FoodPriceIn Import food / agriculture market 2 2 1 179.902 189.437 173.4
