The Application of Tax on Means of Transport in Local Tax Policy in Poland. Evidence of City Counties

Tax on means of transport is one of the own revenues of city counties in Poland. They may affect its construction within the scope defined by the legislator. The authors of the article examined and evaluated the volume, stability and diversification of the significance of revenues achieved by city counties on account of tax on means of transport and discussed the significance of certain factors determining their level. Among these factors the local tax policy deserves a special attention. Therefore, the authors examined the scale of application of instruments of the local tax policy as well as their practical effectiveness reflected by the budget revenue growth in the subsequent periods. The following methods were used: descriptive statistics, correlation analysis, classical Pearsons independence test or Spearmans rho correlation test. The article also presents econometric models illustrating the effects of tax policy in relation to particular categories of tax payers of tax on means of transport.


A b st r a c t
Tax on means of transport is one of the own revenues of city counties in Poland. They may affect its construction within the scope defined by the legislator. The authors of the article examined and evaluated the volume, stability and diversification of the significance of revenues achieved by city counties on account of tax on means of transport and discussed the significance of certain factors determining their level. Among these factors the local tax policy deserves a special attention. Therefore, the authors examined the scale of application of instruments of the local tax policy as well as their practical effectiveness reflected by the budget revenue growth in the subsequent periods. The following methods were used: descriptive statistics, correlation analysis, classical Pearsons independence test or Spearmans rho correlation test. The article also presents econometric models illustrating the effects of tax policy in relation to particular categories of tax payers of tax on means of transport.

Introduction
The tax on means of transport (referred to as TMT) is a revenue of communes (and consequently also city counties) included in these entities' own revenues. It is an example of taxation on movable property used in some areas of business pursuit. It belongs to the group of levies which local governments may use to follow their own tax policies as their competences in this area mean the right to reduce the rates of the tax in question and to apply other tax preferences. The role it plays as a source of revenue of local entities is determined on the one hand by the scale of application of commune competence and on the other by the scope of tax authority and other regulations, included in the Act on tax and local charges, as well as a number of factors connected with local determinants including, for example, the infrastructure, commune transport closeness and accessibility, the quality of traffic routes, customer closeness and easiness of implementation of business functions. The aforementioned factors, in the first place the scale of independently pursued tax policy by the local authorities as well as local determinants diversify the level of TMT revenues generated by particular entities and its fiscal significance in the whole country.
The article is aimed at the examination and evaluation of the volume, stability and diversification of the significance of revenues achieved by city counties on account of tax on means of transport and also at the indication of the significance of certain factors determining their level. It was assumed that a special attention should be paid to the local tax policy among all those factors. In order to measure the impact of city counties TMT tax policy on the volume of TMT revenues generated by them, the analysis covers a fairly long period of 2007-2014. The article makes use of the Rb-27s report data from all city counties.
The research of city county transport tax policy accounts for different relations, i.e. fiscal diversification of effects of tax policy with regard to all tax payers, and also separately with regard to natural and legal persons. In order to achieve the goals of the article several research methods were used, first of all descriptive statistics (structure and dynamics indicators), also correlation analysis, classical Pearsons independence test, or Spearmans rho correlation. The article also presents econometric models illustrating the effects of tax policy in relation to particular categories of tax payers of tax on means of transport.
The choice of the subject of the present article results primarily from the fact that in Poland the empirical research on local tax policy, its diversification and effects is extremely rare1. Primarily, there is a real shortage of research in the area of corporate movable property tax. Therefore, we focus our attention in this article exclusively on the tax on means of transport, which is interesting not due to the scale of revenues (the tax efficiency is lower than in the case of property tax), but to a relatively big tax base mobility, which is decisively bigger than in the case of other wealth taxes. It is the only tax, from among local taxes in Poland, in the case of which we deal with the tax base mobility in the literal meaning of the word.

Legal Regulations and Economic C ontent o f Tax on M eans o f Transport in th e C ontext o f Tax A u th o rity o f Com m unes
Tax on means of transport is regulated in the Act on local taxes and charges2. It is connected with the possession and use of means of transport specified by the legislator3. This tax is, then, a typical example of direct wealth taxation. However, there are different opinions according to which tax on means of transport should be classified as an income tax. The income character of TMT, as the followers of this stance claim, results from the economic character of means of transport, the subject scope of the tax and a normative construction of the tax rates assigned to particular vehicles4. When discussing in this section of the article the rights in the area of tax authority given to communes, a competent tax institution with regard to TMT should be pointed to. With regard to the matters of tax on means of transport, a competent tax authority is the one whose competence covers the tax payers registered office (in relation to legal persons or possibly entities with no legal personality) or the place of residence (criterion for natural persons)5. In the case of a multiple-plant enterprise, the tax should be paid in favour of the authority whose competence covers the plant's location, i.e. to the account of the commune on whose area the plant possessing the vehicles is registered.
The tax base is the number of transport means owned by the tax payer or regis tered in his name. The assumption that the tax base is uniform for the whole tax is a certain simplification as in the case of this tax the kind of the means of transport is significant (a lorry, a tractor unit, a ballast tractor, trailers and semi-trailers or buses) and its characteristics (maximum permissible weight, the number of axles, kind of suspension or the number of seats). An important element of the TMT subject scope, affecting also the height of tax rates, is the maximum permissible weight6. In the case of lorries it should not pose any difficulty as it results from the registration card. However, setting the maximum permissible weight of a vehicle unit includes the curb weight of the tractor unit and the permissible trailer weight. It should be potentially largest trailer or semi-trailer the tractor can pull. On the other hand when setting the maximum permissible weight of a trailer and semi-trailer together with the motor vehicle the tax payer sums up the maximum weight of the lorry and trailer or semi-trailer. The parameters determined in this way are not always treated as a correct solution. It appears that the motor vehicle considered in the setting of this kind of maximum permissible weight cannot be identified only with the vehicle pulling the trailer or semi-trailer at this very moment. What matters is a potential motor vehicle that may be used to pull the trailer. Thus, a tax payer declares the maximum permissible weight of the trailer and semi-trailer together with the motor vehicle, which means: the semi-trailer maximum permissible weight and the smallest possible weight of the tractor able to pull the semi-trailer or the permissible maximum weight of trailer and the smallest possible maximum permissible weight of a potential ballast tractor (lorry) able to pull the trailer. It is important that each of these means of transport is subject to tax irrespective of being a part of a vehicle unit (motor vehicle with a trailer or semi-trailer) or being used independently by an entrepreneur. The aforementioned factors affect the level of highly diversified amount rates of tax. Their height depends on the so called external features, which may have an impact on the size of business activity pursued. Tax on means of transport has features of taxes 5  connected with the use of environment as the vehicles of higher weight and higher number of axles, being more environmentally undesirable, are levied with higher tax rates. Thus, the legislator takes care of the public interest diversifying tax rates7.
The height of rates of tax on means of transport applied in a commune is determined by the commune council; however they have to be included in between the minimum8 and maximum rates9. Importantly, a commune council must not set any rate at the level of 0 zlotys. In an economic sense it would mean the exemption of the vehicles from taxes, and this is settled by the legislator in Article 12 (tax exemptions). Moreover, what the doctrine indicates, a zero rate cannot be included in the idea of tax rate in the understanding of the analysed act of law because no tax amount may be calculated when applying a zero rate, and setting a rate at this level is identical to the lack of rate10. The literature also emphasises that the diversification of indexation methods of minimum and maximum rates cannot be rationally justified and in the extreme cases it could give rise to a paradoxical situation in which minimum rates could be higher than maximum rates11. In such a situation, in accordance with the content of Article 10 Section 3 the maximum rate is not taken into account and a commune council is obliged to accept the rate of tax on means of transport for all the types of vehicles at the level of a respective rate specified in Annexes 1-3 to the Act (i.e. the minimum rate). Let us emphasise that these regulations mean that commune councils lose their right to set tax rates. Only the minimum rate is valid then (higher than maximum rate), and a commune council can neither increase nor decrease it. In this way a constitutionally protected commune council competence to set tax rates within an appropriate scope has been limited12.
The rights on the part of a commune council with regard to tax rates do not concern only their determination but also the diversification of their height. When determining the rates for different kinds of taxed objects, the commune council takes into account in particular the impact of the means of transport on the natural environment, the year of production or the number of seats. The criteria indicated by the legislator should be treated as an example, and commune councils may adopt completely different criteria to diversify tax rates. But, let us enumerate evidence to prove that the procedure of tax rate diversification is partly limited. First, tax rates cannot be diversified on the subject basis. Diversification may only be effected with regard to the object criteria. Second, when diversifying taxes, the commune council cannot interfere in the statutory categories of the vehicles taxed. It is possible to introduce sub-divisions in addition to the statutory divisions. Third, the rates determined in the new classification cannot exceed maximum rates. Fourth, it is impossible to diversify rates with regard to the means of transport in the case of which minimum rates have been set.
The current preferences in tax on means of transport may be classified in the following groups: 1. Preferences of a systemic character, including: • means of transport in possession of diplomatic representations, • means of transport that constitute reserves for the purpose of military mobilisation and special-purpose vehicles • vintage vehicles within the meaning of road traffic regulations, • means of transport used in the so-called combined transport13. 2. Preferences introduced by commune councils in other forms than statutory tax exemptions. While considering the commune council exemption competences, one should focus on their scope. The legislator assumed in Article 12 Section 4 that a potential exemption enacted by a commune council may not refer to the means of transport covered by minimum rates. In an attempt to indicate the aim of such a regulation, it should be observed that since the legislator assumed such a solution (minimum rates), a commune council cannot introduce exemptions for these means of transport. An important problem refers also to the assessment of freedom given to a commune council in the exemption application. The statement that the only limitation in the implementation of exemptions by commune councils is the catalogue of exemptions of the aforementioned means of transport is fairly disputable. It is worth noting that the phrase used by the legislator: "A commune council has the right to introduce other object exemptions than those defined in Section 1", means the right to introduce only overall exemptions. If communes introduced on their territories a construction of a partial tax exemption or tax relief, it would be contrary to the content of Article 12 Section 4, as well as its 13 In the case o f this solution, the tax payer may receive a tax return depending on the num ber of runs with or w ithout the load. The reduction in the tax am ount is diversified and fluctuates from 25% of annual tax am ount (in the case o f 20 to 49 runs inclusive by a m eans of transport in railway transport in a fiscal year) to 100% of annual tax am ount (in the case of 100 runs and more). systemie interpretation14. In this situation another issue should be considered, i.e. giving more rights to commune councils to let them make resolutions on other preferences, e.g. tax reliefs. The following arguments may be mentioned in favour of the proposal of granting commune councils the right to enact reliefs in tax on means of transport15: • making use of non-fiscal tax functions to affect economic and social phenomena in the area (enacting tax reliefs communes may influence tax payers' decisions encouraging them to install equipment limiting fumes emission into the atmosphere), • giving communes complete tax authority within their own revenues, • adoption of uniform principles on the competence of commune councils in all local taxes16. The current preferences mean that the legislator has resigned from the solutions thanks to which priority used to be given to certain kinds of business pursuit and tax payers' conduct. Some symptoms of the application of a stimulative policy may be observed only within the procedure of tax return on account of combined transport. This solution is to be positively assessed as combined transport is used in Poland, in relation to the West European countries, unsatisfactorily (developing gradually though). The literature emphasises that the application of this solution should be particularly promoted due to the way means of transport are used to benefit the minimising of negative impact on the environment and road surface condition17.
To recapitulate this section of the article, it may be assumed that a desired feature of local taxes is an opportunity to affect the regional tax policy through their construction. Therefore, the idea of diversifying tax rates by commune councils is right. Thanks to it, tax payers may be encouraged to renew the possessed means of transport and to acquire vehicles in favour of the natural environment protection. In the context of enhancement of non-fiscal tax functions it would be well-founded to expand the rights of local entities and allow them to make resolutions not only with regard to non-statutory object exemptions but also tax reliefs18.

The Analysis and Evaluation o f Tax on M eans o f Transport as a Source o f Incom e o f City Counties
Tables 1 and 2 illustrate the 2007-2014 budget revenues of city counties on account of certain categories of their own sources of revenue19, with indication of these revenues that showed a yearly downward trend (darker areas in Table 2). Analysing the presented data, it should be observed that: -the volume of own budget revenues as well as an integral part, i.e. tax revenues generally show an upward trend: own revenues of nearly 28.5% from the level of 23 279.36 million zlotys to 29 909.31 million zlotys; local taxes of over 24% from the level of 7 403.55 million zlotys to 9 205.98 million zlotys; local taxes with active tax authority of 47.5% from the level of 5 448.81 million zlotys to 8 037.49 million zlotys. Only in the case of tax on means of transport there is a considerable decline in its impact on the budgets of city counties, from the level of 324.45 million zlotys to 307.35 million zlotys (decline of over 5%). It is commonly known that individual elements of local tax systems are characterised by different sensitivity to the course of the business cycle. Most local government revenues are highly sensitive to cyclical fluctuations in the economy: primarily the shares of local governments in income taxes, tax on civil law transactions or tax on means of transport20. Revenues from the aforementioned levies paid by entrepreneurs undoubtedly show close interdependence with their active business pursuit undertaken on the local markets. Property tax is less sensitive to business cycle fluctuations as its dues depend not on the value but on the property area21. Property tax is the most important source of own revenues of local governments from among all local taxes, providing local government budgets with high revenues also in the period of recession or economic slowdown (high growth dynamics). -tax on means of transport is not at present a stable source of city county revenue, and its significance in creating budgets of these entities has considerably decreased in comparison with 2007, when revenues as well as the share in their own revenues and in revenues from local taxes achieved a maximum level in the analysed period. Already 2008 saw a very clear decline in the local government revenues, and in the subsequent years this decline went even further. This phenomenon was, to a certain extent, caused by an economic slowdown, unfavourable factors typical of this sector (e.g. a decline in eastward freight in the international transport), as a result of which there was a rise in the number of corporate bankruptcies. In the years 2011-2014 revenues from tax on means of transport showed a definite upward trend. Thus, 2014 saw 307 million zloty worth of city county revenue from tax on means of transport, which amounted to only 1.03% of their own revenues, 3.34% of revenues from local taxes and only 3.82% of revenues from local taxes with active commune tax authority.  The research conducted in city counties confirms that tax on means of transport is primarily incurred by legal persons. To generalise, as much as 70% of the tax paid came from legal persons, and only 30% from natural persons (Table 3). The volume of revenue which city counties acquire from tax on means of trans port is determined by many diversified factors. Firstly, primarily by the ones beyond their control, for example the business cycle, the situation of the market of means of transport, level of prices, number of entrepreneurs, accessibility of credits and leases or the regulations included in the Act on local taxes and charges. Secondly, whether or not and to which extent they exercise their tax authority.
In order to illustrate the scale of application of competence of city authorities in the area of creation of revenue from tax on means of transport, a comparison was made of the effects of application of tax authority instruments with the total revenues from this tax totally, and also separately with regard to the tax paid by legal and natural persons (see Table 4). The rights exercised by cities in the analysed period had a diversified effect on revenues from TMT. In 2007 the fiscal effects of applica tion of tax authority instruments in relation to TMT in city counties amounted to 88.15 million zlotys, which accounted for 27.2% of revenues from this tax. In 2014 the amount was more than twice as high -nearly 182 million zlotys, which accounted for more than 59% of revenues from TMT. The competence to set tax rates was of the greatest significance in revenue reduction. Lower rates gave rise to a financial effect from nearly 83 million zlotys in 2007 to nearly 180 million zlotys in 2014. No wonder that a rising tendency recorded on absolute data occurred also in the relation of financial effects of upper rates reduction to tax revenues. The lost revenues rose from 25.5% in 2007 to over 58% in 2014 of revenues from TMT.
It is worth commenting on the scale of application of tax authority instruments by city counties with regard to individual groups of tax payers. In the years 2007-2014 legal persons had their share of 68% in them on the average and natural persons only 32%. Preferences given to legal persons rose more quickly -the rise in current prices of 117.95%, while preferences given to natural persons rose by 83.22%. Taking into account the relation of the financial effects of application of tax authority instruments to TMT revenues, we can see that the size of the financial loss was similar: on average nearly 44% in the case of legal persons and nearly 48% in the case of natural persons. The effects of application of tax authority instruments in TMT with regard to legal and natural persons using in their business pursuit certain means of transport were considerable, which means that town authorities willingly used their competences, primarily, as remarked, in the area of reduction in the upper TMT rates and their diversification. And the application of non-statutory reliefs and exemptions affected the level of revenues from tax on means of transport to a much lesser degree.
So far, our attention has been primarily paid to the fiscal effects of city county tax authority. However, one cannot ignore the fact that thanks to the reduction in maximum tax rates and also thanks to the diversification of their height it is possible to achieve certain goals, which are important from the perspective of the policy of socio-economic development of the local entity. They are, for example: more interest of the local entity on the part of transport sector companies, improvement in the natural environment or a better quality of the corporate transport base22. Hence, city counties, when diversifying TMT, used criteria mentioned in the Act, i.e. the impact of the vehicle on the environment or its year of production. The research conducted by J. Dziuba indicates that city counties often referred to the EURO norms, defining permissible fume emissions in the vehicles sold in the EU, they preferred vehicles possessing equipment reducing fume emissions, gas installation, electric drive or biofuel, as well as newer vehicles23.
Undeniably, lower local tax rates create more favourable conditions for companies to operate, but according to a lot of research conducted, they do not have a decisive significance in the selection of location for registration and business pursuit24. It appears that for companies it is not the height of tax rate that matters, but the stability, predictability and transparency of the commune tax policy25. It is also worth looking at Figure 1, which shows the significance of the tax policy pursued through tax rates in tax on means of transport in comparison with the decisions about all local taxes covered by city counties tax authority as defined by the Act. The presented chart indicates that in the analysed period the share of revenue losses on account of maximum rates reduction in TMT in the amount of loss of revenue on account of the application of this instrument of tax authority in all local taxes rose considerably (by more than 17 percentage points in relation to all tax payers, by more than 19 percentage points in relation to legal persons and by 14 percentage points in the case of natural persons).
The research indicates then that city counties as a whole group present an attitude consisting in expansive tax policy in the area of tax on means of transport. It may be expected that cities could see more effectiveness in the application of tax policy instruments, in particular in the case of TMT, where the tax base is more mobile than in the case of taxes on immovable property, generally more significant for communes. It cannot be forgotten that we deal, as proven, with a tax of small efficiency. A potential decline in revenue connected with the application of tax authority instruments by city counties is not generally too severe for their budgets. Hence, it cannot be excluded that TMT rates reduction may be treated by politicians as an action allowing them to demonstrate their openness to the support for local undertakings26.

Figure 1. The significance of tax policy of city counties in tax on means of transport in relation to all local taxes with active tax authority (in%) total in relation to legal persons in relation to natural persons
Source: authors' own materials based on the RB-27s Report data.

The Im pact o f A pplication o f Tax A u th o rity Instrum ents on th e Level o f B udget Revenues o f City Counties
Asking about possible effects of financial policy of city counties in tax on means of transport, the authors scrutinised at the beginning on the classical correlation of share growth of amounts due to the reduction in the upper tax rates in city counties revenues on account of transport tax (variable PU expressed in %) and a quotient of variable showing an inflation accounted relative growth of revenues on this account over the previous year (variable PD, expressed in %). A possible impact of PU on PD in a given year and in the next year was analysed. The correlation tables presented below show clearly that variable PU is generally negatively correlated with variable PD in a given year, which was to be expected as more tax preferences in the form of lower tax rates had to result in the decline in revenue growth. However, if one considers the impact of variable PU from a given year on variable PD in the next year, a positive correlation is seen, i.e. to cut it short: the tax policy works. The increase in tax preferences (enacting lower rates) gives rise to positive financial effects in transport tax collection in the following year. Let us assume though that it is only a preliminary observation.
A simple correlation analysis may be distorted even by single errors or outliers. In order to minimise the influence of such data on the analysis of the impact of PU on PD the authors processed the data. Firstly, dichotomisation of variables PU and PD was made to dichotomic variables u and d -if PUR < 0, then we assume uR = 0, and otherwise Ur = 1. The same -when variable PDr < 0, then we assume dR = 0, and otherwise dR = 1. We j oin bit variables from two consecutive years (R -1, R) obtaining variable Ur and variable Dr. These variables may assume values 0(00), 1 (01), 2(10) or 3 (11). We will show later that there is a significant correlation between Ur and Dr+1. The obtained contingency tables were used to complete a classical Pearsons independence test and also to calculate Spearmans rho correlation, useful here due to the type of variables.
In the next section of the article we show examples of models of the impact of PU in a given year R (PUr, R = 2008,..., 2013) on PD in the following year (PDr+1), primarily with regard to the tax policy addressed to natural persons, where it seems to have the greatest influence. A possible impact of tax policy "accompanies" the impact of dynamics of growth in transport tax revenues and the share of transport tax in total tax collection-the impact of both these variables is "negative" -the one that "rose" in transport tax collection "stronger" and had a higher transport share in all taxes and charges collected by the commune, had lower indicators of the dynamics of growth in the collection of this tax in the following year.
The analyses were carried out separately for the tax on means of transport paid by legal and natural persons. First, we present the correlations for TMT from legal persons. The statistical analysis shows that the completion of missing and potentially erroneous data gives unexpected results. Therefore, the authors decided to remove such data from the analysis in order to deal with the values for which a sensible processing is possible, for example we resigned from the indicators that would have to be divided by 0. If data PU, PD from all the years are treated as cross-sectional time data, there is a positive correlation of PU with PD from the same year (r = 0.1506, significance test value t = 2.563, p < 0.01, Students t-distribution with d f = 283 degrees of freedom) and insignificant (!) correlation with PD from the following year (r = -0.0839, significance test value t = -1.416, d f =283, p = 0.158). If these data are to be considered with regard to subsequent years separately (see Table 5), it is seen even more clearly that certain correlations, from among those specified in the correlation table, are not significant. For TMT from natural persons the calculated corrections indicate a clearer impact of tax policy on tax collection, though there are single outliers seen from the general "pattern" (see Table 6). If PU, PD data from all the years are considered as cross-sectional time data, there is a strong correlation of PU with PD from the same year (r = 0.3944, significance test value t = 7.155, Student's t-distribution with d f = 278 degrees of freedom) and a negative correlation with PD from the following year (r = -0.3646, significance test value t = -6.528, d f =278).
The correlations of variables PU with PD for natural persons are on average stronger and more stable over time than for legal persons (for the evaluation of effectiveness of tax policy pursued in year R the value of correlation of variable PUR with variable PDr+1 is significant here).
Additionally, the behaviour of variable PDr is shown in the following tables (Tables 7 and 8).  Apart from one outlier (for TMT from natural persons), a negative correlation is clearly seen. There is a negative dependence of PDr+1 on PDr -i.e. a positive tendency in growth in tax in a given year stimulates the occurrence of a negative tendency in the following year. On the contrary, if the growth in revenue from tax is on the decline in a given year, in the following year a stronger growth in revenues from tax is more likely to occur. It is as if the situation of fiscal effects of tax authority stabilised all by itself and in a sense averaged itself -an increasing growth is followed by a declining growth, and a declining one is followed by an increasing one.
As remarked before, in order to minimise the impact of outliers on the analysis of impact of tax policy on the size of tax collection, the data were subject to processing, described in detail at the beginning of the article. Thus, we will show now convergence tables worked out for variables UR and DR+1 for years R = 2009, 2010,..., 2013 (resulting from processing of original variables PU, PD from years R = 2008, 2 0 0 9 ,., 2013). Let us begin with such a table for TMT for legal persons and year R = 2009 -we will describe in a greater detail what was calculated and what the significance of the accompanying numbers is. So in Table 9 we examine the dependence of D2010 on U2009.  The value of Pearsons independence x2 = 9.47, which gives a probability of improper rejection of hypothesis of independence at the level p = 0.39 (distribution x2 with 9 degrees of freedom). The value of Spearmans rho correlation coefficient amounts to -0.1174, and the value of a respective significance test amounts to t = -0.938, p = 0.35 (Student's t-distribution with 63 degrees of freedom). The results of both these tests indicate an insignificant dependence between U2009 and D2010. The next tables, for the next years, will be completed with the values whose sense we described with the first convergence table (no. 9).    Source: authors' own materials based on the RB-27s Report data.
Next, there are convergence tables presented for tax on means of transport paid by natural persons, R = 2009,..., 2013.      We have already said that the correlations of variables PUr with PDR+1 for tax collected from legal persons do not indicate any significant "pattern" of dependence, if considered separately for every year. When considering these data as cross-sectional and time data, some weak dependences of this kind are seen. The situation is better in the case of tax collected from natural persons. The correlation dependence here is clearer and more stable -growing PUr affects a decline in variable PDr, and the same growing variable PUr affects a rise in variable PDr+1. This first dependence is a natural consequence of financial arithmetic -if communes increase the scale of application of their competence in a certain year, reducing the upper tax rates, the tax revenues in the same year will fall. The second dependence means that an increase in financial effects of commune tax authority will influence a positive tendency in tax collection in the following year. Thus, it may be stated that preferences given to tax payers in the form of lower tax rates in one year have a stimulative effect on the tax collection in the following year. TMT collection from natural persons brings about 3 times less revenue than from legal persons. There have to be other mechanisms here thanks to which the policy seems to bring positive effects in the case of natural persons, while in the case of legal persons it does not.
Recapitulating: for the collection of tax on means of transport from legal persons no regular "pattern" of dependence was found -there is no uniform, positive reaction of business entities to city counties tax policy. It is not much better in the case of collection of tax from natural persons. Pearsons independence test indicated a weak dependence of variable UR (a rise in the share of preferences in the form of lower rates in tax collection in year R) on variable Dr+1 (tendency of rise in tax col lection in year R+1). Spearmans rho correlation test did not confirm that. The size of the data set does not favour the application of Pearsons independence test and Spearmans rho correlation test. However, if we aggregate convergence tables from all the years (from 2009 to 2013), then for a 5 time bigger data set now, we will obtain new tables (Tables 19-21).  The presented tables indicate that the dependence is not very strong but significant. Pearsons independence test indicates a little stronger dependence for legal persons, while Spearmans rho correlation test indicates stronger dependence for natural persons. The latter, as shown in Tables 9-18, is less inclined to show a significant dependence, so the indication of Spearmans rho correlation for natural persons may be considered more significant. The indications of the classical Pearsons correlation test shown in Tables 5-6 may be used as an additional argument. They point to a more stable correlation dependence in the case of natural than legal persons.  The authors also analysed many econometric models to find possible dependences of variables for year R, including PUR, with variable PDR+1. For TMT collection from legal persons a "pattern" of dependence was sought, with independent variables UDr and PDr, and dependent variable PDr+1. In the models obtained for the following years a logical "pattern" of dependences could hardly be seen. If the 2009-2013 data are aggregated into one data series, then for independent variables PU and PD from a given year and variable PDnext from the next year as a dependent variable the following model was obtained: Here, also regression coefficients at variables PUR are positive, and regression coefficient at yariables PDr, are negative. The former means a positiye reaction to tax policy in relation to the tax paid by natural persons. The latter means that after a negative tendency in the growth in the tax from natural persons, in the following year there is a better chance for the occurrence of positiye tendency. Moreover, it should be observed that the model is coincident, i.e. the signs of regression coefficients are compliant with the signs of respective correlation coefficients. If the 2009-2013 data are combined in one cross-sectional time data series, then for independent variables PU and PD from a given year and variable PDnext from the next year as a dependent variable the following model is obtained: What does this model show in comparison with the former one? Firstly, a bigger % share of effects of reduction in upper rates in collection of tax on means of transport stimulates positive response to tax policy. Secondly, the bigger the % share of revenues is on account of tax on means of transport in the own revenues of city counties, the less effective their tax policy seems to be. Variables PU and PD work in the same way as in the model, in which only these variables were used as independent variables. It is a very interesting general question -in which communes (due to the value of a specific economic variable) tax policy works better, and in which ones it works worse. One of such definite answers is included in the above model -variables VU, UD are then a possibility of differentiating local entities with regard to the expected impact of tax policy on the volume of revenues in future periods. After removing variable VU from the set of independent variables we obtain the model: PDnext = 6.68% -7.5378*UD + 0.3261*PU -0.4523*PD If this model is compared with a respective model for tax on means of transport from legal persons presented before, it can be seen that: • A negative impact of variable UD is lower than for legal persons (increased share of tax on means of transport in total revenues of city counties in a given year more weakly affects the downward tendency in revenues from means of transport in the following year). • A positive impact of PU is also smaller than for legal persons (see a remark below). However, it should be considered here that variables PU and PD are generally negatively correlated every year; thus, PU growth causes the growth in PDnext (which results from the model), but PU growth is directly correlated with the PD decline, which also causes PDnext growth due to a negative correlation coefficient at PD. • A negative impact of variable PD is stronger than for legal persons (growth dynamics/decline in collection of tax on means of transport in a given year is subject to faster compensation in the following year; thus a relative growth in collection of tax on means of transport self-restricts itself more strongly).

Conclusion
Recapitulating the conducted research, we may state that tax on means of trans port is not a significant and stable source of income of city counties in Poland. In the course of analysed years the volume of revenues feeding their budgets on this account was only slightly reduced. The best year from the point of view of the effects of tax on means of transport in city counties budgets was definitely 2007, the first year of our analysis and at the same time the last pre-crisis year. Only in the last years of the research period was there a certain improvement, a growth in revenues received by city counties on account of TMT. The significance of this tax as a source of budget revenues of these entities decisively declined (minimum TMT share in own revenues and revenues on account of local taxes).
On the one hand, this phenomenon was caused by objective factors (macroeconomic), beyond any control of city counties. It should be emphasised that the regulations included in the Act on local taxes and charges were an important but permanent factor. In the analysed period the legislator unfortunately did not modify the laws regulating TMT. The most important changes, as a result of which the object scope of tax was narrowed and which seriously limited its role and at the same time its significance as an instrument of socio-economic policy of communes, were introduced much earlier.
O n the other hand however, it was caused by the application of certain instruments by city authorities (reduction in upper tax rates or introducing non-statutory reliefs and exemptions) in the framework of the local tax policy pursued. The research proves that city counties quite commonly used the rights given to them (a high and constantly growing share of effects of the use of tax authority in tax revenues, and also the growth in the significance of tax policy in TMT in comparison with tax policy in all local taxes). In particular, it dealt with the possibility of enacting lower tax rates than maximum rates defined in the Act, which as a result gave rise to a considerable decline in revenues on this account. Being aware however, of a small revenue potential of city counties from this tax (it has brought only 1% of city counties own revenues in recent years), we may record a marginal impact of the height of TMT rates on the revenue situation in city counties. The article also indicates that the diversification of rates, on the basis of the assumed criteria, might be in favour of rational activity stimulating the achievement of desired economic and social effects (improved natural environment thanks to lower fume emissions and reduced noise intensity or improvement in the transport base of companies operating in the commune). A substantive assessment of the effects of the outlined motivational function requires some separate in-depth research though.
Tax on means of transport is, as remarked, a local tax characterised by the greatest mobility of the tax base. Do the effects of upper rates reductions, particularly conspicuous in this tax, confirm tax competition between large cities? The article does not present strong evidence of tax competition in the area of tax on means of transport. It may only be speculated that in the case of tax whose tax base is more mobile than that of other local taxes, and also with regard to a tax of low efficiency (low risk for the revenue level), tax competition becomes more probable.
It was a very important issue for the authors to examine not only the use of instruments of the local tax policy, but also their practical effectiveness manifested in the budget revenue growth in the subsequent periods. The general conclusion from this part of analysis is as follows: the reduction in the upper rates of tax on means of transport gives rise to diversified effects in different periods. It more often leads to a decline in budget revenues. Thanks to the constructed correlation tables we managed to observe certain statistical dependences. A correlation has been confirmed between tax policy, primarily in the case of tax paid by natural persons, and a growing tendency in revenues from this tax. The authors made an attempt to create some models explaining the effectiveness of city counties tax policy in the area of tax on means of transport.