REGIONAL DEVELOPMENT AND ITS MEASUREMENT IN VISEGRAD GROUP COUNTRIES

The aim of the paper is to measure regional development and construct an index for the Visegrad Group countries at NUTS 2 level. This index, called the Regional Development Index the RDI is created as an extension of the Human Development Index in order to obtain a better composed index at regional level. Twelve socio-economic indicators are selected for this purpose: three economic indicators, three educational indicators, three health variables and three indicators of the standard of living which create four dimensions. These variables are tested for their reliability through the pairwise correlation and the min-max method is used for the construction of the index. The data are compared between 2008 and 2013 and the assumption about worsening the situation in regions after the crisis is set. The results show that the values of the RDI improved in nearly all regions (with the exception of Prague in the Czech Republic and Közép-Magyarország in Hungary) in the monitored years. The assumption that regional development was negatively influenced by economic crisis has not been confirmed.


INTRODUCTION
Regional or national development is usually measured by indicators such as domestic product, national income, or alternatively by the Human Development Index and the Index of Sustainable Economic Welfare or Gross National Happiness ( Van den Bergh, 2009). While getting the data and setting of the above mentioned measurements at national level is not too complicated, the problem arises at the level of a region. Therefore, it is necessary to modify the measurements at this level and to find suitable socio-economic indicators which should contain adequate information. As Rovan & Sambt (2003) claim, the socio-economic issue among regions should be of primary interest to economists as well as politicians and their differences should be maintained within the sustainable limits for the sake of the welfare of the country as a whole. The analysis of these indicators may serve as the basis for development policy at the regional level. The major distinction in most cases is the fact that regions are open spatial entities (in contrast to countries), while the competence of a region may normally be superseded by the nations (Nijkamp & Abreu, 2009).
The aim of this paper is to construct a regional index for selected members of the European Union by the most often used measurement of human development, the HDI, and to compare this regional index in a period before and after the economic crisis. The countries of the Visegrad Group (hereafter V4) at the NUTS 2 level have been chosen for the analysis. This group includes 35 regions in the Czech Republic, Hungary, Poland and Slovakia -eight in the Czech Republic, seven in Hungary, sixteen in Poland and four in Slovakia. The data were obtained and compared for the years 2008 (before the crisis) and 2013 (after crisis). The assumptions that regional development was negatively influenced by economic crisis and the value of the RDI worsened between 2008 and 2013 were defined.
The Regional Development Index (hereafter the RDI) will be created as an extension of the HDI in order to obtain a better composed index at regional level. Thirteen socio-economic indicators will be selected for this purpose: three economic indicators (GDP per capita, R&D expenditure and unemployment), three educational indicators (tertiary educated population, people in lifelong learning and young people neither employed nor educated), three health variables (life expectancy at birth, health personnel -number of doctors and infant mortality) and four indicators of the standard of living (stock of vehicles i.e. -passenger cars, nights spent at tourist accommodation establishments, victims by accidents -killed and municipal waste). These variables will be tested for their reliability through the pairwise correlation and the RDI will be constructed using the min-max method after the selection of the correlated indicators.
Even though this paper is not the first attempt to study the development of (not only) the above mentioned socio-economic indicators, it differs from the existing studies in using more complex concepts of the given issue in the min-max method at the regional NUTS 2 level.
The paper is organized as follows: the second section presents a brief literature review, the third section describes the model and methodology used in the paper. Section four discusses the results in detail and the fifth section concludes the paper.

THEORETICAL BACKGROUND
The selection of indicators used in this paper was inspired by many studies in which the authors confirmed linkages among some indicators.
Firstly, the implementation of the GDP indicator was influenced by Sen´s opinion (Sen, 1999) who considered the income (product) as a primary mean to achieve human development. The relationship between economic growth and unemployment is very well known according to Okun´s law. A further indicator, research and development expenditure (R&D) and its increase is very important for increasing competitiveness (Bočková, 2013). Nevima & Kiszová (2011) claim that gross domestic expenditures on research and development are the sources for further economic growth. According to Hudec & Prochádzková (2015), the innovative capacity of a region can be considered as its ability to produce and commercialize innovations to drive a long-term economic growth and wealth creation. They examined the regions of the Visegrad countries by considering R&D expenditures by the concept of the knowledge production function (Cobb-Douglas type). The result was that not the capital regions are the most innovative ones, because several Polish regions (Lodzkie and Malopolskie) and Czech regions (Střední Morava and Jihovýchod) belong to the most efficient regions. Similar results were obtained by Kozuń-Cieślak (2016) who used the methods of the composite indicators and the DEA method.
Secondly, higher education and lifelong learning contribute to economic development as well. Florida, Mellander & Stolarick (2008) assert that human capital and the creative class affect regional development through different channels. Whilst the creative class outperforms conventional educational attainment for regional labour productivity, conventional human capital does better for regional income. Positive relation between tertiary graduates and economic growth in Visegrad countries was found in Verner & Chudarkova (2013) as well.
The adult education systems (lifelong learning) currently in place tend to reinforce existing economic disparities, with greater frequency of re-skilling and up-skilling by more educated adults, with higher income levels (WEF, 2017).
Thirdly, Michaud & van Soest (2008) claim that in many industrialized countries there is a positive association between health and wealth and population; health tends to rise with the country's level of economic development (Semyonov et al. 2013). In addition, health improvements tend to reduce the mortality rates of infants (Bloom & Canning, 2003). Anand & Bärnighausen (2004) argue that a strong relevance between health personnel and infant mortality exists in more than 80 countries.
Fourthly, Riley (2002) examined the influence of population growth, increased urbanization and economic development on the rapid growth of motor vehicles in China. Medlock & Soligo (2002) did a research on the effect of economic development on the demand (numbers) of the motor vehicles in 28 countries and developed a model of the relationship between economic development and per capita private car ownership. A practical example of relationship between economic development and an amount of vehicles is obvious with Toyota (Toyota, 2017): as the Japanese economy expanded (15% in the period of [1955][1956][1957][1958][1959][1960][1961][1962][1963][1964][1965][1966][1967][1968][1969][1970], the demand for passenger cars in particular grew rapidly, and the sales volume achieved an average annual growth rate of 32 percent. According to Tuan (2011), the gross regional product per capita in Thailand Provinces might have strongly exponential effects on car ownership. Shafik (1994) found out that increasing income indicates the waste generation deterioration. Eugenio-Martin et al. (2004) state that tourism provides two positive effects on economy: on one hand, an increase in production and income; on the other hand, as the tourism sector is labour intensive, it causes an increase in employment. It has certainly exerted a very important economic, productive, and cultural influence (Pérez and Nadal, 2005). Similarly, tourism plays an important role in solving economic and social problems, providing more jobs, initiating the employment growth of economically active population and increasing the welfare of a nation, and at the same time it has a stimulating effect on the development of many related fields of the economy -it contributes to socio-economic development (Gabdrakhmanov & Rubtsov, 2014). Borowy (2013) was dealing with road traffic injuries using the discourse analysis. He explored how development has been (re-)negotiated through the discourse of these injuries and vice versa. Gebru (2017) found that a road traffic accident is a human security threat with multifaceted effects on the economy of households and the national economies of states, especially in the developing countries. It affects the national economy of countries and households directly or indirectly because it causes a loss of the economically active population. According to Agbeboh & Osarumwense (2013) accidents cause heavy costs to society especially in case of a loss of able bodied men and women who would have been involved in productive economic activities as a loss of intellectuals, a loss of resources to government and families, a loss to insurance companies and a damage to properties. Road traffic injuries and deaths are a growing public health problems worldwide. Banthia et al. (2006) have shown that road traffic injuries are major causes of death and disability globally, with a disproportionate number occurring in developing countries.
The Human Development Index is primarily a nation level indicator, estimated for a country as a whole (Basu & Basu, 2005), but due to its general nature it cannot be applied by all economies in general. Therefore, many countries have introduced their own modified indexes in order to reflect their local circumstances better (Pagliani, 2010, or Gaye & Jha, 2010. Gnesi et al. (2010) have published the Index of the Regional Quality of Development Some authors analysed human development at regional level using cluster analysis, as in the case of China between 1982(Yang & Hu, 2008  The above mentioned authors (Nevima & Majerova, 2016) applied factor analysis of human development within the same group of countries as well. Their assumption that the most important factor of human development is economic level, measured by gross domestic product per capita, was not confirmed and was found that the most important role is played by another factor -life-long learning. This finding confirms that education of population is a very important variable of regional as well as national significance.
The closest research to the topic of this paper was done by Hardeman & Dijkstra (2014) who developed a composite indicator which was capable to measure patterns and trends in human development across the EU region in 2012. They chose (only) six reliable indicators out of 22 -healthy life expectancy, infant mortality, NEET, general tertiary education, net disposable income and employment rate, using the min-max model.

OBJECTIVES AND METHODS
This paper investigates the impact of economic crisis on regional development of all regions in the Visegrad Group countries. At first, a sample and variables (as a model) are described and selected through pairwise correlation. Subsequently, they are used in the Regional Development Index by the min-max method (according to UNDP, 2016).

Model
The economic geography of Europe is characterised by wide levels of a number of socioeconomic variables that are both a cause and a response to differences in growth and levels of income per capita (Fingleton, 2003). As it has already been mentioned, the Visegrad Group countries (V4) at the NUTS 2 level are analysed. There are 35 regions at this level -eight in the Czech Republic, seven in Hungary, sixteen in Poland and four in Slovakia. The list of the regions in our sample is shown in Table 1. As it has been noted, the annual data were obtained from the Eurostat regional database (Eurostat, 2017), which contains data for NUTS 1 to 3 regions. Not all the data are available for all regions of the EU and for each level of classification, so the selection of indicators was rather limited. For the purpose of this paper, thirteen regional socio-economic variables were chosen, the units, codes and relations of which can be found in Table 2 (Eurostat, 2014). The higher values of GDP per capita are associated with higher levels of development.
Intramural the implicit gross domestic product (GDP) for R&D statistics, although it is recognized that they reflect the opportunity cost of the resources devoted to R&D rather than the "real" amounts involved. For the purpose of this paper, relative indicator per capita in PPS was used.
This indicator is positive for the regional development.
An indicator of Unemployment (UNP) by sex, and age in NUTS 2 regions represents all inhabitants aged 25 or over and is expressed as a percentage of active inhabitants in the age of 25-64 years. This age level was chosen to complement the age group used in the indicator NET, i. e. the age group between 15 and 24. This indicator (its high level) has negative effects on regional development, representing a social problem connected with negative effects on economic activities.
The share of Tertiary educated people (TEE) in the productive age population of the region is connected with the ability of people (and regions) to reflect the needs of knowledge of economy, and it also reflects the level of human development.
Lifelong learning (LLL) as the percentage of the regional population participating in education and training encompasses all learning activities undertaken throughout life (after the end of initial education) with the aim of improving knowledge, skills and competences, within personal, civic, social or employment-related perspectives (Eurostat, 2017). Due to lifelong learning people extend their possibilities for increasing their incomes, well-being and development. These indicators´ higher values are associated with higher levels of development.
The indicator Young people neither employed nor in education or training (NET) corresponds to the percentage of the total population of a given age group (15-24) that is not employed and not involved in further education or training. The age group was selected to complement the age range used for UNP to eliminate too high correlation or autocorrelation.
This variable has a negative effect on development.
The The variable Nights spent at tourist accommodation establishments (NST) is calculated as total nights per thousand inhabitants spent by a guest, resident or a non-resident in a region.
1 Passenger car is presented by road motor vehicle, other than a moped or a motorcycle, intended for the carriage of passengers and designed to seat no more than nine persons (including the driver). Included are: passenger cars, includes micro cars (needing no permit to be driven), vans designed and used primarily for transport of passengers, taxis, hire cars (provided that they have fewer than ten seats), ambulances and motor homes. Excluded are light goods road vehicles, as well as motor-coaches and buses, and mini-buses/mini-coaches (Eurostat, 2017).
Tourist establishments are hotels and similar accommodation, holiday and other short-stay accommodation, camping grounds, recreational vehicle parks and trailer parks. As it has been mentioned, tourism (in this research the capacity utilization of tourist facilities) contributes to the development of a region.
The quantity of waste reflects the differences in economic wealth among regionswealthier regions usually generate more municipal waste and have a negative impact not only on environment but on development as well. In this paper, Municipal waste (WST) expresses the total waste per inhabitant in tons and it consists of waste collected by the municipal authorities, or directly by the private sector (business or private non-profit institutions). The bulk of the waste stream originates from households, though similar wastes from sources such as commerce, offices, public institutions and selected municipal services are included as well.
It also contains bulky waste, but excludes waste from municipal sewage networks and municipal construction and demolition waste (Eurostat, 2017).
The last but not least variable is the Victims of accidents (VOA) per million inhabitants of the region. For the purpose of our paper the persons killed (any person killed immediately or dying as a result of an injury accident 2 , with the exception of terrorist acts and suicides), were selected, due to no possibility of their further positive contribution to enhance regional development (through consumption, higher education or lifelong learning etc.). This variable is chosen as a factor with a negative influence.

Methodology
The purpose of this paper is to construct a composite index of regional development using the min-max model (UNDP, 2016). A majority of the previous studies were devoted to analyses of other methods or a narrower range of this index, but a more comprehensive analysis is made in this paper.
Concerning the RDI index, not only the same method as the HDI construction was chosen (with minor deviations, see below), but also the same principle of its creation, i.e. -the component indicators should be assigned the same weight and divided into the relevant dimensions with a positive or negative influence on development. Suitability of selected indicators, weight and impact, was tested through the pairwise correlation analysis, namely 2 Injury accident is any accident involving at least one road vehicle in motion on a public road or private road to which the public has a right of access, resulting in at least one injured or killed person. It includes collisions between road vehicles; between road vehicles and pedestrians; between road vehicles and animals or fixed obstacles and with one road vehicle alone. Included are collisions between road and rail vehicles. Multi-vehicle collisions are counted as only one accident provided that any successive collisions happen within a very short time period. Injury accidents exclude accidents incurring only material damage (Eurostat, 2017).
the Pearson correlation coefficient (UNDP, 2015or Halásková & Mikušová Meričková, 2017. When using equal weights, it may happen that -by combining variables highly correlated (above ±0.90) -an element of double counting may be introduced into the index. In response to this problem the indicators are tested for statistical correlation -and then only those indicators are chosen which report a low degree of correlation (but more than ±0.30 3 ) or adjusting weights correspondingly, e.g. giving less weight to correlated indicators (OECD, 2002). The results of the pairwise correlation can be seen in Table 3, when both years (2008 and 2013) were tested. As shown in the above table, the indicator of waste (WST) reported a very low value (less than ±0.30) in both monitored years, so it does not correlate with any other indicators and has been excluded for further analysis. Even though the indicator VOA showed lower values of correlation in more than half of the cases, it has not been excluded from further analysis (because of higher value of the rest of indicators). Then, 12 indicators with the same weight were left, and four dimensions were created from these indicators, each with three indicators, two positive and one negative, according to the results of correlation (see Table 4).
and index for variables with negative direction (2) where V index is the respective value of the 12 component indicators, V real is a real value, V min is a minimum value and V max is a maximum value. The values of the sub-indexes EC, ED, HE and SL are calculated as the arithmetic means of the three component values of the (3) The principle of the Regional Development Index is calculated as the geometric mean of all the above indices, as shown in (4) (4) The

CONCLUSION
The aim of the paper was to construct a regional index for the Visegrad Group countries at NUTS 2 level. 35 regions of the Czech Republic, Hungary, Poland and Slovakia were chosen for this purpose. The data were obtained for the years 2008 (before the crisis) and 2013 (after the crisis).
The Regional Development Index, RDI, was built as an extension of the Human Development Index in order to obtain a more complex index at regional level. Therefore, twelve socio-economic indicators were selected: three economic indicators (GDP per capita,