Sustainable development across Central and Eastern Europe: key factors driving the economic growth of the countries/Darnusis vystymasis Centrineje ir Rytu Europoje: pagrindiniai ekonominio augimo aspektai.

Lapinskiene, Giedre ; Tvaronaviciene, Manuela

1. Introduction

Sustainable development is the leading concept of our days embracing economic, social and environmental dimensions (Brundtland 1987). The measuring and management of this process is a difficult task because the concept varies, depending on the changing conditions of life. The measurement of social and economic development of a country is a complex phenomenon, which is described by a set of criteria (Podvezko 2008). Many international institutions presented the assessment systems of indicators to measure the sustainability around the world. Despite the main pillars, classifications and sets of indicators differ across various institutions. The problems of analytical analysis based on these indicators were considered by a number of scholars (Grybaite, Tvaronaviciene 2008; Tvaronaviciene et al. 2008).

For many years, GDP has been presented as the main variable showing the level of economic development for a particular country, but it could not reflect the welfare of the country. It is important to ensure that all indicators of sustainable development should change positively. However, there are some indicators having the strongest impact on the development level of every region. Finding the most important variables could simplify the process of monitoring and help to determine areas and policy for future development.

It is assumed that the most important sustainable development indicators are those which are most closely connected with economic growth (expressed as GDP growth).

The aim of this paper is to identify the set of key sustainable development indicators (from Eurostat database) having the strongest impact on the growth of the whole region of Central and Eastern Europe. Correlation and regression analyses are used for estimating the effect of variables. The countries development level will be evaluated based on the calculated regression equation in the context of Central and Eastern Europe.

2. Theoretical background

A great number of economists have tried to understand the economic processes and create models which could help to manage the growth of economy since ancient times. Adam Smith with the 'Wealth of Nation' saw the realization of the economies of large-scale production as an important source of growing national prosperity can be considered to be a predecessor of growth theories (Greenwald 1994). Jumping from classical scholars (Smith, Malthus, Ricardo) to the theories of neoclassical economists (Harrod and Domar, Hicks, Solow), it is seen that growth theory economists have tried to define a systematic frame for the equilibrium paths of the economy. Solow, the best-known neoclassical scholar, presented a model where economic growth was stimulated by changing the constant capital output ratio by a richer standard of the technology in the equilibrium model (Solow 1988). J. A. Schumpeter, with his business cycle theories based on innovation, and John von Neumann, with mathematical theories of economic growth, as well as many other researchers made a valuable contribution to the development of fundamental macro economy theories. The latest theories of macro economy are associated with the intense work on growth theory in the late 1980s and 1990s known as endogenous growth theory. The early contributions here were by Romer and Lucas. Paul Romer (1994) emphasized that economic growth is an endogenous outcome of an economic system, not the results of forces that impinge from outside. R. Lucas (2003) argued that there were economic gains from providing people with better incentives to work and to save, not from better fine-tuning of spending flows. Not going into theoretical considerations about the factors driving economic growth, the paper concentrates on the factual interplay between economic growth and sustainable development indicators in order to find the main variables determining the economic growth in Central and Eastern Europe.

3. The process of selecting sustainable development indicators impacting GDP

The analysed sample is Central and Eastern European countries, including Bulgaria, the Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovakia and Slovenia. All of them, except Bulgaria and Romania, joined the EU in 2004 (the latter two countries--in 2007). In recent years, these countries have demonstrated the high growth rates. Therefore, to ensure further growth, sustainable development policy should be pursued.

The analysis has four phases. Firstly, all variables from the Eurostat sustainable development database were reviewed and only those satisfying the conditions were chosen for analysis: i.e. Lithuanian data is available, the same data sets cover more than one country, data are gathered annually in the period from 1998 without intervals, the variables are statistically measured.

Secondly, according to the aim of the paper, correlation analysis was used as a statistical method to define the relationship between GDP and the sustainable development indicators considered. Correlation coefficients were calculated using the log change of sustainable development indicators and the log change of GDP (in constant prices) for Lithuanian data. The results of calculations were obtained using MS Excel. They are presented in Appendix 1.

Thirdly, the correlation results were evaluated for statistical significance and economic logic. In many sources, the correlation coefficient of |0,30| is described as a minimal level for the relationship to be valid, but this is only true for large data samples (higher than 50 items). For a small data sample (as in this work), the significance of the correlation coefficient can be determined by using standard distribution calculated by Student t test (Mason et al. 1999). Alternatively, a simple formula to determine the approximate critical value of the correlation coefficient at 0.05 level of significance was introduced (Walsh 2008): 2/[square root of n], where n is the number of data items.

Accordingly, the calculated threshold for the correlation coefficient in the presented data sample is |0,63|. Hence, the following criterion is used to shortlist indicators: at least one country in the Baltic states (the Baltic states are taken as countries developing in a similar way) should have a coefficient of more than |0,63|; the relationship between the variables should follow the with economic logic. Only the indicators satisfying the criteria defined were chosen. They are presented in Appendix 2. Correlation results are displayed in Fig. 1.

[FIGURE 1 OMITTED]

Four of these indicators have not met up the correlation criteria defined, but they are included based on economic logic. The correlation between the employment rate by highest level of education attained and GDP shows positive trends--the better results could lead to higher GDP. As the Lithuanian strategy is aimed at creating knowledge society, this indicator could be very important, therefore, it is included in further analysis. It is evident that higher rate of labour productivity per hour worked should result in higher GDP. The calculated correlation coefficient is below the threshold set. However, it is still demonstrating positive trends. Real effective exchange rate (REER) can be used to assess the competitiveness of the state's currency. It should be noted that national currency in all Baltic states is historically pegged to base currency (USD, SDR, EUR), while policy of exchange rate of other Central and Eastern European countries is different. Therefore, in the Baltic states, the relationship is not so straightforward, though for other countries it might be economically important. Economic logic should indicate negative correlation, implying that the increase of competitiveness (decrease in REER) causes GDP growth in Eurozone countries. Energy intensity shows the amount of energy needed to produce one unit of economic output. The calculated correlation coefficient is below the threshold set, but in sustainable development context, it is important to monitor the impact of this variable on the development of Eastern European countries. The results of correlation analysis between GDP and sustainable development variables define the key indicators, which have a strong impact on the economic growth in the Baltic states.

At the fourth stage of this investigation the most valuable set of variables from Appendix 2 is found using the regression analysis. The statistical approach allows us to forecast the dependable variable (GDP) using independent variables. This process is called the regression analysis. The multiple regression case extends the equation to include additional independent variables (Cekanavicius, Murauskas 2002). A general formula for pooled data regression analysis with fixed effect estimation is as follows:

[Y.sub.it] = [alpha] + [X.sub.it] [[beta].sub.it] + [[epsilon].sub.it] where

[Y.sub.it] is a dependent variable for country i at time t;

[X.sub.it] is an independent variable for country i at time t;

i is the country's number;

t is time period;

[alpha] is a fixed coefficient;

[beta] is regression coefficient;

[[epsilon].sub.it] is estimation error coefficient.

A general formula is widely used in various economic calculations, e.g. the New Global Competitiveness index by World Economic Forum calculates weights based on the regression of the pooled data set on country GDP per capita (Martin et al. 2008). The data on ten Eastern European countries (Bulgaria, Czech Republic, Estonia, Hungary Latvia, Lithuania, Poland, Romania, Slovakia and Slovenia) for all indicators from Appendix 2 for the period from 1998 to 2007 is entered into Eviews software database (2000). After several estimations of regression coefficients, minimal data set was defined. The calculated result of the analysis is presented in the standard statistical form produced by software (Table 1).

As shown in the above table when any coefficient (column two) is statistically close to zero it means that the variable associated with this coefficient is not important in determining the dependent variable. After some basic analysis, the coefficients, which can be equal to zero with probability higher than 5%, were excluded from further calculations. Probability of a coefficient statistically not different from zero is calculated by the software and shown in the column "Prob." in Table 1. Energy intensity of the economy, household expenditure per inhabitant, growth rate of labour productivity per hour worked, unemployment rate and gross inland energy consumption are all statistically significant. R-squared for the estimated equation is 0.73 (the coefficient of determination is the percentage of the variation expressed by the equation).

Therefore, the regression analysis shows that only five out of twenty one indicators, which passed the correlation analysis, were chosen based on regression analysis. These indicators are seen to be most important, strongly influencing the economic growth in the Central and Eastern Europe region.

4. Specifying the main determinants' impact on economic growth in Central and Eastern European countries

The identified indicators are incorporated into a regression equitation in order to test the model and determine the impact of the above on a particular country. A general multiple regression formula may be rewritten for the specified estimated model using five chosen sustainable development indicators and estimated coefficients to calculate the simulated GDP growth. A specific formula and methods of calculating GDP growth in Lithuania for the year 2007 are presented below:

dlog(GDPLT,2007) = [alpha]+[beta]1,LT*dlog(D1LT) + [beta]2,LT* dlog(D2,LT) + [beta]3,LT*D3,LT + [beta]4,LT*dlog(D4,LT) + [beta]5,LT*dlog(D5,LT);

dlog(GDP) = 0.021--0.234*--0.096 + 0.223*0.110 + 0.001*5.700--0.054*--0.264 + 0.261*--0.022 = 0.082.

The result is close to actual GDP numbers (0.085). Similarly, the data on the remaining countries were calculated and presented in Table 2. It should be noted that the results in the table are derived directly from the formula, (e.g. GDP change is presented as log difference), therefore, the actual GDP change in percentage should be recalculated based on the results presented. In the next graph, the recalculated GDP changes (in percent) are given.

As the model results are close to actual data (see Fig. 2) (the discrepancies are not discussed in this paper), it could be useful to consider them in detail for further analysis of the specific aspects of economic development in Central and Eastern Europe. The impact of particular separate indicators on GDP growth in the considered countries of Central and Eastern Europe is shown in Fig. 3.

[FIGURE 2 OMITTED]

[FIGURE 3 OMITTED]

Comparing Lithuanian position to that of other Central and Eastern European countries, it can be observed that the impact of the chosen indicators is very similar in the Baltic states. The household expenditure is the main determinant in Lithuania, similar to Romania, Estonia and Latvia. Household expenditure refers to any spending made by a person living alone or by a group of people living together in shared accommodation and with common domestic expenses (Eurostat). The strong impact of household expenditure on economic growth shows about its significance for economic development. Hence, policymakers should spare no effort to increase household disposable income (e.g. in recession it may be appropriate to cut income taxes, assuming that the resulting increases in disposable income will raise household spending, thereby reducing the severity of recession) (Johnson et al. 2005). It is clear that spending policies should be weighted against government ability to borrow, as well as overall health of public finances.

The impact of energy intensity is in the second place, being similar to that observed in Latvia, Estonia and Slovakia. Energy intensity shows the amount of energy needed to produce one unit of economic output. A lower coefficient value indicates higher energy efficiency. While differences in energy intensity levels can be attributed to such factors as geography, wealth, culture, natural endowment and economic structure, their movement over time reflects the combined effects of efficiency improvements, structural changes in the economy, changes in energy-using activities and types of fuel substitution (Joskow 2003).

Unemployment rate and labour productivity indicators relate to labour market. In Lithuania, the impact of unemployment is similar to that in Estonia, Poland, Czech Republic and Bulgaria. Unemployment rates represent unemployed people as a percentage of the labour force. The labour force is the total number of employed and unemployed people (Eurostat). Unemployment is the classical index of macro economy and the governments of many states take much effort to reduce this variable. The impact of labour productivity is very similar in all countries (except Poland). GDP has to grow based on labour productivity because it means the effective usage of resources. The ultimate goal of a well-functioning labour market is high and growing labour productivity, which, in turn, translates into higher wages and salaries for workers (Fraser Forum 2004). Therefore, the significance of labour market can be clearly seen and its monitoring is very important.

Gross inland energy consumption has a small negative impact in Lithuania similar to the situation in Estonia, Hungary and Slovakia. Gross inland energy consumption shows the usage of various energy sources (oil, gas, renewable, etc.). The growth of energy consumption is a result of rapid economic growth, creating larger demand which is caused by the increase in investment levels, population, and trade in energy. High energy consumption leads to environmental degradation (Chousa et al. 2008). Gross energy consumption is related to energy efficiency, and a positive impact of the latter can be an offsetting factor of lower total energy consumption. This shows a positive trend towards sustainable use of energy.

Household expenditure, energy intensity and labour productivity are economic indicators having the strongest impact in the whole sample. Two remaining indicators, i.e. unemployment and gross inland energy consumption, belong to social and environmental groups. Their effect is smaller in countries considered. The results are in good agreement with economic logic, proving that Central and Eastern European countries are developing in a similar way. As shown by the data, the Baltic states have nearly the same main determinants of economic growth and are on the same path of development. The economically logical results prove the validity of analysis. Hence, the areas considered should be monitored more closely in order to achieve higher economic growth.

5. Conclusions

The suggested hypothesis implies that, in a particular region, there might be a set of sustainable development indicators reflecting the factors strongly influencing GDP. Sustainable development indicators from Eurostat database for Central and Eastern Europe region were taken as a dimension for detailed consideration. Not every variable from the Eurostat sustainable development database satisfied the defined primary conditions (starting that Lithuanian data is available; the same data sets cover more than one country; the data gathered annually in the period from 1998 without intervals; the variables are statistically measured). To test the hypothesis, correlation and regression analyses were used. However, this approach has some drawbacks: the data cover only nine years but this period is statistically small, most of the variables analysed have the trend of the rapid growth, making the correlation results artificially high, while the data itself may be hard to measure. Nevertheless, the analysis performed can identify the trend of development.

Only five out of the analysed variables were found to be the most significant in the region of Central and Eastern Europe. They are energy intensity of the economy, household expenditure per inhabitant, growth rate of labour productivity per hour worked, unemployment rate and gross inland energy consumption. Household expenditure, energy intensity, labour productivity are indicators from the economic group, with their impact being the strongest in all the countries considered. It is compliance with economic logic stating that economic indicators are most significant at the transition stage of development. Two remaining indicators, i.e. unemployment and gross inland energy consumption, belong to social and environmental groups and their impact is less visible in the countries analysed. Only the increasing economic power allows the states to invest in social and environmental development. The results of the study show that Central and Eastern European countries are developing in a very similar way. It particularly applies to the Baltic states which have nearly the same significant determinants of economic growth and are on the same path of development. Despite the regions' similarities, every country has to establish its own competitiveness level and find its own opportunities to win its share in the global market. Only the country which is able to pursue sustainability policy and get the economic benefits from this can catch up with Western Europe much faster. For example, in stimulating household expenditure, the importance to spend rationally citizens' savings should be emphasized. The state policy of energy saving should be implemented by every economic agent. The effectiveness of using labour force should be sought by all enterprises, companies and institutions. However, sustainability philosophy obliges us to estimate every driving factor for long-term consequences.

Received 09 February 2009; accepted 02 July 2009

Iteikta 2009-02-09; priimta 2009-07-02

References

Brundtland, G. 1987. Our Common Future. The World Commission on Environment and Development. Oxford: Oxford University Press. 416 p

Cekanavicius, V.; Murauskas, G. 2002. Statistika ir jos taikymai II. Vilnius: TEV.

Chousa, J. P.; Tamazian, A.; Chaitanya, K. V. 2008. Rapid Economic Growth at the Cost of Environment Degradation? Panel Data Evidence from BRIC Economies. William Davidson Institute Working Paper No. 908.

EUROSTAT. Available from Internet: <http://ep peurostat.ec.europa.eu/>.

Eviews 4 Users guide. 2000. Irvine CA: Quantitative Micro Software, LLC.

Fraser Forum. 2004. Labour Market Performance, 13-23.

Greenwald, D. 1994. The McGraw-Hill Encyclopedia of Economics. McGraw-Hill, Inc.

Grybaite, V.; Tvaronaviciene, M. 2008. Estimation of sustainable development: germination on institutional level, Journal of Business Economics and Management 9(4): 327- 334. doi:10.3846/1611-1699.2008.9.327-334

Johnson, D.; Jonathan, A. P.; Souleles, N. S. 2005. Household Expenditure and the Income Tax Rebates of 2001. Bureau of Labor Statistics. Princeton University, Department of Economics and University of Pennsylvania, Finance Department.

Joskow, P. L. 2003. Energy policies and their consequences after 25 years, Energy Journal 24(4): 17-49.

Lucas, R. E. 2003. Macroeconomic priorities, American Economic Review 93(1): 1-14. doi:10.1257/000282803321455133

Martin, X.; Blanke, J.; Hanouz, M. D.; Geiger, T.; Mia, I.; Paua, F. 2008. The Global Competitiveness Index: Prioritizing the Economic Policy Agenda. Word Economic Forum, The global competitiveness report 2008-2009.

Mason, R. D.; Lind, D. A.; Marchal, W. G. 1999. Statistical Techniques in Business and Economics. Tenth edition. Irwin McGraw-Hill.

Podvezko, V. 2008. Sudetingu dydziu kompleksinis vertinimas, Verslas: teorija ir praktika [Business: Theory and Practice] 9(3): 160-168.

Romer, P. M. 1994. The origins of endogenous growth, Journal of Economic Perspectives 8(1): 3-22.

Solow, R. M. 1988. Growth theory and after, American Economic Review 78(3): 307-317.

Tvaronaviciene, M.; Ginevicius, R.; Grybaite, V. 2008. Baltijos saliu issivystymo palyginimas: praktiniai kompleksinio poziurio taikymo aspektai, Verslas: teorija ir praktika [Business: Theory and Practice] 9(1): 51-64.

Walsh, D. A. 2008. Simple Rule of Thumb for Statistically Significant Correlation. Available from Internet: <http://frank.mtsu.edu/~dwalsh/436/CORRSIG.pdf>.

doi: 10.3846/1648-0627.2009.10.204-213

Giedre Lapinskiene. Master of Management and Business Administration, Vilnius Gediminas Technical University. Research interests: sustainable development, indicators of sustainable development, economic growth.

Manuela Tvaronaviciene. Doctor, Professor, Department of Enterprise Economics and Management, Vilnius Gediminas Technical University. Research interests: economic development, foreign direct investment, business environment.

Giedre Lapinskiene (1), Manuela Tvaronaviciene (2) Vilnius Gediminas Technical University, Sauletekio al. 11, LT-10223 Vilnius, Lithuania E-mails: (1) [email protected]; (2) [email protected]

Giedre Lapinskiene (1), Manuela Tvaronaviciene (2) Vilniaus Gedimino technikos universitetas, Sauletekio al. 11, LT-10223 Vilnius, Lietuva El. pastas: (1) [email protected]; (2) [email protected]

Appendix 1. The correlation between GDP and sustainable development
indices' changes in the Baltic states

 Lithuania Estonia Latvia

Labour productivity per person 0.44 -0.06 0.15
 employed--GDP in Purchasing
 Power Standards (PPS) per
 person employed relative to
 EU-27 (EU-27 = 100)
Total investment--% of GDP (#) 0.67 0.79 0.75
Public investment--% of GDP 0.13 0.29 0.36
Business investment--% of GDP (#) 0.70 0.79 -0.09
Dispersion of regional GDP per -0.28 0.32 -0.52
 inhabitant--in % of the
 national GDP per inhabitant
Net national income--% of GDP 0.26 -0.25 0.33
Gross household saving--% of -0.26 0.05 -0.09
 gross household disposable
 income
Labour productivity per hour 0.25 na na
 worked--% change over previous
 year
Total R&D expenditure--% of GDP 0.36 -0.36 0.59
Energy intensity of the 0.36 -0.31 -0.55
 economy--kgoe per 1 000 euro
Total employment rate--% (#) 0.43 0.70 0.65
Employment rate, by highest 0.17 0.40 0.62
 level of education attained--%
 of age group 25-64 years (#)
Electricity consumption by 0.05 -0.08 0.49
 households--1 000 toe
Electricity consumption by 0.05 -0.08 0.49
 households--1 000 toe
Household expenditure per 0.69 0.91 0.85
 inhabitant, by category--Volume
 index (1995 = 100) (#)
Total long-term unemployment 0.01 -0.48 -0.50
 rate--%
Lifelong learning--% 0.26 -0.23 na
Public expenditure on -0.16 -0.54 -0.16
 education--Percent of GDP
Early school-leavers--% -0.24 -0.42 na
Employment rate of older workers 0.12 0.56 0.32
Net migration, including 0.41 0.02 0.28
 corrections--persons
Incidence of salmonellosis--new 0.37 0.22 -0.19
 cases per 100 000 persons
Death rate due to chronic 0.00 0.48 0.55
 diseases--per 100 000 persons
Total greenhouse gas 0.82 0.35 0.69
 emissions--index base
 year = 100 (#)
Renewables in gross inland -0.69 -0.05 -0.11
 energy consumption--% (#)
Energy dependency--% -0.10 -0.21 -0.06
Implicit tax rate on -0.33 0.10 -0.21
 energy--Ratio of energy tax
 revenues to final energy
 consumption, deflated
Electricity generated from -0.06 0.09 0.07
 renewable sources--% of gross
 energy consumption
Energy consumption of transport, 0.77 0.44 0.54
 by mode--1 000 toe (#)
Shares of environmental taxes in -0.31 0.14 -0.28
 total tax revenues--%
Shares of labour taxes in total -0.71 -0.82 -0.55
 tax revenues--% (#)
Growth rate of labour 0.40 0.29 0.38
 productivity per hour worked--%
 change over previous year
Real effective exchange 0.07 -0.05 0.52
 rate--index 1999 = 100
Employment rate, by gender--% (#) 0.43 0.70 0.65
Unemployment rate, by -0.41 -0.58 -0.70
 gender--% (#)
Municipal waste generated--kg per 0.70 0.06 0.52
 capita (#)
Municipal waste treatment, by 0.90 -0.01 0.41
 type of treatment method--kg
 per capita
Emissions of acidifying 0.45 0.23 0.48
 substances, by source
 sector--1 000 tonnes acid
 equivalents
Emissions of ozone precursors, 0.62 0.35 0.59
 by source sector--1 000 tonnes
 ozone-forming potential
Emissions of particulate matter 0.58 0.04 0.51
 by source sector--1 000 tonnes
 particulate-forming potential
Final energy consumption, by 0.80 0.62 0.62
 sector--1 000 toe (#)
Persons with low educational -0.06 0.31 0.20
 attainment, by age group--%
Employment rate of older 0.12 0.56 0.32
 workers--%
Life expectancy at age 65, by -0.25 -0.10 na
 gender--years
Total fertility rate--number of 0.17 -0.06 -0.07
 children per woman
General government debt--General -0.83 -0.50 -0.56
 government consolidated gross
 debt as a percentage of GDP (#)
Suicide death rate, by age -0.60 -0.69 0.19
 group--Total--crude death rate
 per 100 000 persons (#)
Greenhouse gas emissions--index 0.82 0.35 0.69
 base year = 100 (#)
Greenhouse gas emissions by 0.84 0.35 0.68
 sector--million tonnes
 [CO.sup.2] equivalent (#)
Greenhouse gas emissions 0.19 -0.03 0.12
 intensity of energy
 consumption--index 2000 = 100
Gross inland energy consumption, 0.64 0.26 0.57
 by fuel--1 000 tonnes of oil
 equivalent (#)
Implicit tax rate on -0.33 0.10 -0.21
 energy--Euros per tonne of oil
 equivalent
Modal split of freight 0.51 -0.09 -0.45
 transport--% in total inland
 freight tonne-km
Volume of freight -0.36 -0.53 -0.13
 transport--Index 2000 = 100
Greenhouse gas emissions from 0.77 0.62 0.50
 transport--1 000 tonnes of
 [CO.sup.2] equivalent (#)
People killed in road 0.61 0.47 -0.57
 accidents--Number of killed
 people
Emissions of ozone precursors 0.60 0.29 -0.51
 from transport--i 000 tonnes of
 ozone-forming potential
Emissions of particulate matter 0.65 0.60 0.41
 from transport--1 000
 tonnes (#)
Forest trees damaged by 0.76 -0.04 0.10
 defoliation--%
Shares of environmental and -0.31 0.14 -0.28
 labour taxes in total tax
 revenues--%

Chosen for regression analysis are indicated with (#).

* na-data not available

Appendix 2. The correlation between GDP and main sustainable
development indices' changes in the Baltic states

Indicator Lithuania Latvia Estonia

Total investment--% of GDP 0.67 0.79 0.75
Business investment--% of GDP 0.70 0.79 -0.09
Total employment rate--% 0.43 0.70 0.65
Employment rate, by highest level 0.i7 0.40 0.62
 of education attained--% of age
 group 25-64 years
Unemployment rate -0.4i -0.58 -0.70
Growth rate of labour productivity 0.40 0.29 0.38
 per hour worked--% change over
 previous year
Real effective exchange 0.07 -0.05 0.52
 rate--index 1999 = 100
Energy intensity of the 0.36 -0.3i -0.55
 economy--kgoe per 1 000 euro
Municipal waste generated--kg per 0.70 0.06 0.52
 capita
Final energy consumption--1 000 toe 0.80 0.62 0.62
Household expenditure per 0.69 0.9i 0.85
 inhabitant--Volume index
 (1995 = 100)
Shares of labour taxes in total -0.7i -0.82 -0.55
 tax revenues--%
General government debt--General -0.83 -0.50 -0.56
 government consolidated gross
 debt as a percentage of GDP
Suicide death rate, by age -0.60 -0.69 0.19
 group--Total--crude death rate
 per 100 000 persons
Greenhouse gas emissions--index 0.82 0.35 0.69
 base year = 100
Gross inland energy 0.64 0.26 0.57
 consumption--1 000 tonnes of
 oil equivalent
Forest trees damaged by 0.76 -0.04 0.10
 defoliation--%
Energy consumption of 0.77 0.44 0.54
 transport--1 000 toe
Greenhouse gas emissions from 0.77 0.62 0.50
 transport--1 000 tonnes of
 [CO.sup.2] equivalent
Emissions of particulate matter 0.65 0.60 0.41
 from transport--1 000 tonnes
Renewables in gross inland energy -0.69 -0.05 -0.11
 consumption--%

Giedre LAPINSKIENE. Master of Management and Business Administration,
Vilnius Gediminas Technical University. Research interests: sustainable
development, indicators of sustainable development, economic growth.

Manuela TVARONAVICIENE. Doctor, Professor, Department of Enterprise
Economics and Management, Vilnius Gediminas Technical University.
Research interests: economic development, foreign direct investment,
business environment.

Table 1. The regression analysis of the chosen sustainable
development indicators

Dependent Variable: Real GDP
Method: Pooled Least Squares
Date: 04/17/09 Time: 12:15
Sample (adjusted): 1998-2007

Included observations: 10 after adjustments
Cross-sections included: 10
Total pool (balanced) observations: 100

Variable Coefficient

C 0.021135
Energy intensity of the economy -0.233946
Household expenditure per 0.222965
 inhabitant
Growth rate of labour 0.000994
 productivity per hour worked
Unemployment rate -0.054495
Gross inland energy consumption 0.260752

R-squared 0.726159
Adjusted R-squared 0.711593
S.E. of regression 0.013317
Sum squared resid 0.016669
Log likelihood 293.0744
Durbin-Watson stat 1.789463

Variable Std. Error t-Statistic

C 0.002878 7.343504
Energy intensity of the economy 0.041922 -5.580505
Household expenditure per 0.042994 5.185947
 inhabitant
Growth rate of labour 0.000428 2.321334
 productivity per hour worked
Unemployment rate 0.009843 -5.536396
Gross inland energy consumption 0.044452 5.865883

R-squared Mean dependent var
Adjusted R-squared S.D. dependent var
S.E. of regression Akaike info criterion
Sum squared resid Schwarz criterion
Log likelihood F-statistic
Durbin-Watson stat Prob(F-statistic)

Variable Prob.

C 0.0000
Energy intensity of the economy 0.0000
Household expenditure per 0.0000
 inhabitant
Growth rate of labour 0.0224
 productivity per hour worked
Unemployment rate 0.0000
Gross inland energy consumption 0.0000

R-squared 0.051056
Adjusted R-squared 0.024796
S.E. of regression -5.741488
Sum squared resid -5.585178
Log likelihood 49.85287
Durbin-Watson stat 0.000000

Table 2. Comparison of estimated and actual GDP changes (based on the
data for 2007)

 Country

Indicator
(changes), D Bulgaria Czech Estonia Hungary

 Real GDP 0.060 0.058 0.061 0.011
 C
1 Energy intensity -0.033 -0.041 -0.131 -0.046
2 Household 0.084 0.054 0.122 0.032
 expenditure
3 Labour 2.800 3.900 5.700 1.300
 productivity per
 hour
4 Unemployment -0.266 -0.306 -0.227 -0.013
 rate
5 Gross inland 0.028 0.020 -0.025 -0.008
 energy
 consumption
 Model estimation 0.072 0.069 0.091 0.039
 for Real GDP

 Country

Indicator
(changes), D Latvia Lithuania Poland Romania

 Real GDP 0.095 0.085 0.064 0.061
 C
1 Energy intensity -0.086 -0.096 -0.015 -0.034
2 Household 0.205 0.110 0.048 0.128
 expenditure
3 Labour 6.300 5.700 -15.500 6.200
 productivity per
 hour
4 Unemployment -0.125 -0.264 -0.370 -0.132
 rate
5 Gross inland 0.029 -0.022 0.046 0.041
 energy
 consumption
 Model estimation 0.108 0.082 0.052 0.082
 for Real GDP

 Country

Indicator Coefficent,
(changes), D Slovenia Slovakia [beta]

 Real GDP 0.065 0.099
 C 0.021
1 Energy intensity -0.050 -0.094 -0.234
2 Household 0.024 0.061 0.223
 expenditure
3 Labour 4.000 6.400 0.001
 productivity per
 hour
4 Unemployment -0.203 -0.188 -0.054
 rate
5 Gross inland 0.006 -0.012 0.261
 energy
 consumption
 Model estimation 0.055 0.070
 for Real GDP