Experimental statistics

Industry: according to national accounts sections B+C+D+E, namely Mining and quarrying, Manufacturing, Electricity, gas, steam and air conditioning supply, Water supply; sewerage; waste management and remediation activities. The deepest breakdown of industry of present analysis is done according to TEÁOR ‘08 double-digit division breakdown, from 05 Mining of coal and lignite to 39 Remediation activities and other waste management services. We demark a total of 34 divisions (there is no division belonging to number 34 in the double-digit breakdown).

The definition of industry and its divisions corresponds with information disseminated by Eurostat. (Eurostat uses industry (except construction) for what we define as industry.)

Hungary uses in certain cases a narrower definition for Industry, e.g. interim industrial statistics considers mining, manufacturing and energy industry (B, C and D) sections as industry total, with an official short name as follows: industry except water and waste management.

Indicators in the analysis:

Output of industry and its divisions: totality of goods and services produced by a given economic unit for other units as well as for own final consumption.

Gross value added of industry and its divisions: output (at basic price) – intermediate consumption (at purchasers’ price) (Intermediate consumption: value of goods and services purchased during production in the accounting period from other producer units, used for creating new goods and services.)

Value chain length: proxy variable, ratio of gross value added to output.

Concentration analysis: concentration phenomenon may characterise the economic activity in two ways:

Absolute concentration: shows at how many units the sum of values of the phenomenon under observation is concentrated. Its maximum value is reached if the total sum of values concentrates at one single unit (e.g. there is only one region or division); its minimum value (total lack of absolute concentration) is non-definable, an extra actor could show up at any time.

Relative concentration: it shows how unequally the sum of values is distributed. Its maximum value is the same as in the case of absolute concentration (the total sum of values concentrates at one single unit), its minimum value (total lack of relative concentration) means the equal distribution of the sum of values.

The measure of absolute and relative concentration may change in one direction, but even in opposite direction (it is possible that the number of observations grows, thus the absolute concentration level decreases, while at the same time the sum of values distribution becomes more unequal, so the level of relative concentration increases).

Statistical indicators, analytical tools of concentration measurement:

The simplest indicator for the measurement of absolute concentration is n, that is, the number of observations (e.g. the number of divisions, territorial units).

The traditional analytical method for relative concentration is the Lorenz curve and the related Gini coefficient (or concentration ratio).

The Lorenz curve is a square-shaped figure in a Cartesian coordinate system, the horizontal axis showing cumulative relative frequency, the vertical axis showing cumulative relative sum of values. The diagonal represents equal distribution, a total lack of concentration. The farther the Lorenz curve from the diagonal, the more concentrated the phenomenon under observation, in relative terms.

The 0 value of the Gini coefficient (concentration coefficient) means equal distribution, 1 stands for total concentration. Its actual value may be quantified based on the Lorenz curve, the space below the Lorenz curve compared to the triangle representing total concentration.

Index measuring both absolute and relative concentration measure:

The Herfindahl-Hirschman index is one of the most frequently used concentration indicators, the square sum of the relative value amount. It measures between 1/n and 1, where the 1/n value stands for equal distribution, meaning the lack of relative concentration, the value of 1 stands for absolute concentration, 1 sum of values concentrated under observation. The Herfindahl-Hirschman index may be represented with the relative standard error and the number of observations, , where Z_i is the relative sum of values, n the number of observations, the relative standard error. The second half of the formula shows that the index has a greater value even if the relative concentration increases (the relative standard error measures it) and even if the number of observations decreases.

Dimensions, forms of concentration:

Territorial concentration: we quantify in our analysis based on NUTS3 regions for the whole EU. There are at present 1165 small regions in the EU on this level. Counties are Hungary’s NUTS3 regions. We quantify a given country’s territorial concentration of industrial production with a concentration coefficient calculated based on the NUTS3 regions of the country. The use of Gini coefficient is advisable in this case so that we only measure relative concentration. (the number of regions differs significantly, not necessarily meaning actual concentration differences, this is why the use of the Herfindahl index should be avoided.)

Division-based concentration (based on activities): analysis based on the activities, basically meaning concentration appearing as per sections or double-digit divisions.

Aggregate types:

Value aggregates (current price-based data) done as per Eurostat data in national currency or converted to EUR. Several ratios have been calculated based on value aggregates, for example proportion of industry within value added or the value-added ratio.

Volumes (fixed-price data): as well as their changes (a volume index characterising changes may be chain-based or fixed-base index).

Price indices are calculated either directly or indirectly by double deflation.

Temporal changes may be due to changes in original data, but if time series have a seasonal character, then data should be seasonally adjusted. The analysis presents calendar-day adjusted data, calendar-day adjustment meaning working-day and moving holiday adjustment.

Maps:

We present, beside traditional proportional, to scale maps so-called topological maps, too.

Topological maps are special, thematic maps where basic elements of original topology are shown, meaning that neighbouring territorial units are still neighbouring, the size of the territorial unit is, however, proportional with the socio-economic volume to be depicted. Present work applies the method by gross industrial value added modifying the size of the regions, while colouring shows the proportion of industry in the gross value added. We aimed to show where the regions producing the largest as well as the smallest industrial value added , as well as the most and least industrialised regions are situated. ScapeToad 1.1 software has been used for calculations.