Do local governments use municipal companies for off-balance-sheet financing?

This paper aims to investigate whether municipalities in Poland use their municipal companies to increase debt capacity beyond the limitations imposed by the fiscal debt rules. The relationship between the debt of municipalities and the leverage of their subsidiaries is analysed from the perspective of corporate governance and agency theory. Municipalities are a particular example of stakeholders playing a dual role: principals (ultimate or controlling shareholders of municipal companies) and agents carrying out tasks imposed by the state (as defined in the Constitution) related to the provision of public (municipal) services to the residents under conditions of the public finance discipline. The agents of municipalities, in turn, are municipal companies whose owners or majority shareholders are municipalities, which can thus use the companies to achieve their objectives (public service delivery, off-balance-sheet financing). The implementation of the tasks of municipalities performed by municipal companies takes place based on management contracts, which have been in force in Poland since 2017, as well as in-house orders, commissioned to companies whose owner is a local government unit. To reflect the impact of this complex agency relationship, and off-balance-sheet financing on the debt of municipalities, explanatory variables from two sectors were taken into account: public and corporate finance.

Our study on the determinants of short- and long-term debt of municipalities in Poland was carried out using the Ordinary Least Squares (OLS) method. We use financial data of municipalities and municipal companies and structural and ownership data for 2018. Our research stands out from other studies because it combines public finance with private finance.

The structure of this article is adapted to realise its objectives. After the introduction, the second part of the paper reviews the literature on the principal–agency relationship between municipalities and their municipal companies, and debt from the perspective of public finance and corporate finance (capital structure). The third section presents the research design, the fourth part delivers results and the final section concludes.

The local government sector includes 2,477 municipalities (CSO, 2020). Municipalities, as local government units, belong to the public sector and are subject to the regulations of the Public Finance Act. The main task of municipalities is to provide public services for all residents to satisfy their needs. Fulfilling the tasks set for municipalities requires the decentralisation of public income and assets (Świrska, 2008, p. 167–180). The basis for financing local government units is their revenues and external sources of financing. Own revenues include taxes, which includes a share in personal income tax (PIT) and corporate income tax (CIT), local fees, income from municipal property, fees, interest on borrowings granted by municipalities, and other (Act of 11 September 2019 amending the Act on the income of local government units, Journal of Laws of 2019, item 1951).

Revenues from fees include stamp duty, trade fair fees, local, administrative, and other fees. Tax receipts include property taxes, share in PIT and CIT, agricultural, forestry, transportation, inheritance, donation, dog ownership, and civil law activities (Świrska, 2008, pp. 167–180). Apart from their income, the municipalities also obtain external sources of nonrefundable financing, including subsidies and grants, mainly financed from the state budget, the EU, and loans and borrowings.

Due to the multitude of public tasks that have to be done by municipalities, financial resources for individual purposes, including investments in municipal infrastructure, often seem to be insufficient. This is the reason why municipalities benefit from the possibility of debt. The phenomenon of indebtedness among local governments is common (zadluzenia.com), which translates into growing public debt. At the end of 2017, the total debt of Polish municipalities reached 57.205 billion PLN (Gazeta Krakowska, 7.09.2018). The largest Polish cities with the enormous budgets are nominally indebted: Warsaw (over 5 billion PLN), Wrocław and Łódź Ostrowice had the record debt per capita in 2017 (18.644 PLN), and as a result, Ostrowice was liquidated in 2019 (Serwis Samorządowy (Local Government Service) PAP), which is then followed by Rewal (14.924 PLN), Byczyna (12.060 PLN), Dziwnów (5.321 PLN), Wałbrzych (5.015 PLN), Sulmierzyce (4.904 PLN), Daszyna (4.651 PLN), Wrocław (4.410 PLN), Bielice (4.302 PLN) and Lublin (4.245 PLN). Seven municipalities have most signif icant debt problems: Ostrowice, Rewal, Białogard, Manowo, Bielice, Przybiernów and Brojce in West-Pomeranian province (Zadłużenie polskich gmin, 2017). In 2017, seventy-nine municipalities in Poland were not in debt at all, including Świdnik (https://www.portalsamorzadowy.pl/). However, the opportunities for local governments to incur liabilities are not arbitrary in the context of the choice of the purposes of going in debts but are constrained in terms of the volume of debt concerning the capacity to repay them. There are four legal limits to the amount of deficit and debt of territorial self-governments specified in the Public Finance Act (Act of 27 August 2009 on public finance).

These limitations specify the acceptable sources of liability coverage; the requirement to obtain an opinion of the Regional Chamber of Auditors on the possibility of financing the deficit, the limitation of the amount of the planned budgetary revenues of the units, and the limitation of the total amount of the local government debt, which by 2013 could not exceed 60% of the local government revenues established at the end of each quarter according to the values planned and at the end of the financial year according to the values executed (Satoła, 2012, pp. 77–91) and the level of debt service, which could not be higher than 15% of the local government revenues (Kopańska & Kopyt, 2018). The increase in municipalities’ debt led to the introduction in the Public Finance Act of 2009 of a ban on the current budget deficit and a new so-called individual debt ratio (IWZ) in the following form: IWZt=(Dbt-3+Smt-3-Wbt-3Dt-3+Dbt-2+Smt-2-Wbt-2Dt-2+Dbt-1+Smt-1-Wbt-1Dt-1)/3{IWZ_t}\left( {{{{Db_{t - 3}} + {Sm_{t - 3}} - {Wb_{t - 3}}} \over {{D_{t - 3}}}} + {{{Db_{t - 2}} + {Sm_{t - 2}} - {Wb_{t - 2}}} \over {{D_{t - 2}}}} + {{{Db_{t - 1}} + {Sm_{t - 1}} - {Wb_{t - 1}}} \over {{D_{t - 1}}}}} \right)/3 where Db – current revenues, Sm – income from the sale of assets, Wb – operating expenditures, D – total revenues in line with the municipality's budget, t – fiscal year.

The total debt repayment planned for the financial year, including interest, discount, and amounts resulting from the sureties and guarantees granted to total income in the financial year t may not exceed the individual debt ratio (IWZ). As the IWZ is more restrictive for a significant number of municipalities (Kopańska & Kopyt, 2018), an adjustment period was introduced, and the IWZ became effective from 2014.

In Poland, the municipal economy implemented by the local governments may take place within the framework of budget units, local government budgetary institutions and commercial law companies (i.e. limited liability companies). The most common form is municipal companies in which municipalities have various shares in equity, often even the majority ones (Żabski, 2014, pp. 409–423). Municipal services are infrastructural utilities necessary to ensure the basic needs of urban and rural residents. Such services include municipal transport, water and sewage systems, heating (combined heat and power plants), health care, and management of municipal administrative property, residential estate, and other facilities (Barwacz, 2009, pp. 169–179). Since these are public services, for the benefit of the local community, municipal enterprises are subject to control local authorities, i.e. municipalities. This form of ownership of municipalities leads to the existence of corporate governance, and in the case of almost 100% of the municipality's share in the equity of the municipal company, corporate supervision over the internal entity to which in-house contracts, regulated under the Public Procurement Act, apply.

The database used for the study was constructed from several data sources. Data on income and expenditure of Polish municipalities come from the Ministry of Finance. Data related to the local demographic structure come from the Local Data Bank of the Central Statistical Office, while financial and ownership data come from the ORBIS database. These are the data of municipalities and municipal companies located in Poland for 2018. The database consists of 2019 observations. We remove few observations due to the lack of data concerning total assets, profit before tax for 2018, and total assets for 2017 of municipal companies.

The dependent variables explained in the model are the municipalities’ short- and long-term debt. The explanatory variables, on the other hand, concern two sectors: public and private. Table 1 presents definitions of variables and data sources for municipalities, while Table 2 shows those for municipal companies. The leverage variable used in the first presented set of models is the long-term debt to total assets ratio of a municipal company. The remaining of the listed explanatory variables concerning municipal companies are the determinants of leverage well-recognised in the corporate finance literature. They were used in subsequent models. This allows for a more in-depth study of the role of characteristics of a municipal company determining its leverage in shaping a local debt of the municipality.

The study was carried out for all municipal companies’ industries in total and separately for six of the eight sectors in the database. The research carried out for each sector separately is motivated by the variety of characteristics (creditworthiness, debt capacity) between them and the different ways of managing companies’ finances. Two industries: administrative and support activities and agriculture, forestry, and fisheries, were omitted because the number of observations is too small. Table 3 contains a list of industries (according to top-level NACE sectors (sections)) with abbreviations and figures. Next, Table 4 presents municipal companies by number of shareholders, whereas Table 5 compiles descriptive statistics of the dependent variables – short-term and long-term debt of municipalities in Poland (logarithm). Table 6 presents descriptive statistics of independent variables used in the model, whose definitions (and the resulting units of measurement) are described in Tables 1 and 2.

The study was conducted using the Ordinary Least Square method. This method allows us to examine the correlation of debt with all independent variables. The creation of the following study was a several-stage process that aims to select the functional form of the model. The desired functional form is one that reflects as much as possible the theoretical basis for creating an econometric model. The final model is as follows: Short_debti,j=α+β1Revenue_diversificationi+β2Grant_revenuei+β3Unemployment_ratei+β4Income_pci+β5Population_densityi+β6Tax_burdeni+β7Tangibilityi,j+β8Sizei,j+β9ROAi,j+β10Tax_ratei,j+β11Growthi,j+ei,jLong_debti,j=α+β1Revenue_diversificationi+β2Grant_revenuei+β3Unemployment_ratei+β4Income_pci+β5Population_densityi+β6Tax_burdeni+β7Tangibilityi,j+β8Sizei,j+β9ROAi,j+β10Tax_ratei,j+β11Growthi,j+ei,j\matrix{ {Short\_{debt_{i,j}}} \hfill & { = {\rm{\alpha }} + {{\rm{\beta }}_1}{Re}venue\_{diversification_i} + {{\rm{\beta }}_2}Grant\_{revenue_i} + {{\rm{\beta }}_3}Unemployment\_{rate_i}} \hfill \cr {} \hfill & { + {{\rm{\beta }}_4}Income\_{pc_i} + {{\rm{\beta }}_5}Population\_{density_i} + {{\rm{\beta }}_6}Tax\_{burden_i} + {{\rm{\beta }}_7}{Tangibility_{i,j}}} \hfill \cr {} \hfill & { + {{\rm{\beta }}_8}{Size_{i,j}} + {{\rm{\beta }}_9}{ROA_{i,j}} + {{\rm{\beta }}_{10}}Tax\_{rate_{i,j}} + {{\rm{\beta }}_{11}}{Growth_{i,j}} + {{\rm{e}}_{i,j}}} \hfill \cr {Long\_{debt_{i,j}}} \hfill & { = {\rm{\alpha }} + {{\rm{\beta }}_1}{Re}venue\_{diversification_i} + {{\rm{\beta }}_2}Grant\_{revenue_i} + {{\rm{\beta }}_3}Unemployment\_{rate_i}} \hfill \cr {} \hfill & { + {{\rm{\beta }}_4}Income\_{pc_i} + {{\rm{\beta }}_5}Population\_{density_i} + {{\rm{\beta }}_6}Tax\_{burden_i} + {{\rm{\beta }}_7}{Tangibility_{i,j}}} \hfill \cr {} \hfill & { + {{\rm{\beta }}_8}{Size_{i,j}} + {{\rm{\beta }}_9}{ROA_{i,j}} + {{\rm{\beta }}_{10}}Tax\_{rate_{i,j}} + {{\rm{\beta }}_{11}}{Growth_{i,j}} + {{\rm{e}}_{i,j}}} \hfill \cr}

Table 7 presents the results of the regression of dependent variables: short-term and long-term debt using the characteristics of municipalities and debt (the lever variable) of municipal companies in the model (1) and (3) or determinants of the capital structure of companies (discussed in part 2.3, i.e., tangibility, firm size, return on assets (ROA), tax rate, growth) in the model (2) and (4). Due to the specificity of the sectors and the differentiation of debt of municipal companies among the sectors, the models were estimated separately for the sectors. The presented results enable verification of the H2 hypothesis. A positive correlation of financial leverage (lever) with dependent variables was expected. In the case of short-term debt, the hypothesis H2 was rejected. However, there are no grounds to reject the H2 hypothesis in the case of long-term debt. An increase in the leverage by one unit increases the long-term debt of the municipality per capita by 0.838 units, with the remaining variables in the model unchanged. Such conclusions lead to the extension of the study to include leverage determinants for both short- and long-term debt.

Table 8 presents regression results for short-term debt for particular sectors of municipal companies, while Table 9 for long-term debt. The differences between the significance of determinants and the direction of relations result from differences between sectors and their specificity. In the case of short-term debt, the income_pc variable relates positively to the dependent variable for four of the six branches: E, H, F, R. This means that more affluent municipalities with municipal companies operating in the water and sewage and waste management (E), passenger transport (H), construction (F) and sports facilities (R) sectors have higher short-term debt. Higher income per capita enables municipalities to service their debt. The results for these sectors are consistent with those for the total sample (2,019 observations). Short-term debt also positively correlates with the grant_revenue variable in the model for sectors L and D, which means that municipalities with higher revenues from subsidies, managing the real estate through municipal companies (L), and providing energy through combined heat and power plants operating as municipal companies (D), show higher short-term debt. Likewise, municipalities with a higher population density are more indebted in the short term, using municipal companies to provide water and sewage services and waste management (E), passenger transport (H), and property management (L). In contrast, the negative correlation of short-term debt with the tax_burden variable for sectors L and R indicates that municipalities with higher local fees revenue using municipal companies for property management (L) and sports facilities (R) and those with lower taxes paid by municipal companies providing property management services (L) or electricity, gas, steam, and air conditioning supply (D) (negative coefficient with tax_rate for sectors L and D) have less short-term debt. Larger (size) and with higher growth municipal companies providing passenger transport services (H) operate in municipalities with lower short-term debt, but municipal transport companies with higher tangible assets (to guarantee the debt) are found in municipalities with higher short-term debt. Larger municipal companies managing sports facilities or providing energy and heating to residents operate in municipalities with higher short-term debts. In the case of long-term debt, a positive correlation is with the revenue diversification and grant_revenue variables for sectors E and H, which means that a higher revenue diversification and grant receipts allow increasing the long-term debt of municipalities with municipal companies providing water, sewage, and waste management services (E) and transport (H). The higher grant receipts, in turn, reach the more indebted municipalities with municipal companies responsible for sports facilities (R). Municipalities with higher unemployment with municipal companies operating in the electricity, gas, steam, and air conditioning supply (D) sector are more indebted for the long term. In contrast, those with municipal transport companies (section H) have lower long-term debt. More affluent municipalities with shares in municipal companies managing real estate (L), providing construction services (F) or supplying energy and heat (D) have higher long-term debt (positive coefficient with the income_pc variable for the L, F, and D industries), probably due to higher capital intensity of the investment projects underway. Higher income enables them to service their debt. Municipalities with a higher population density and with water and sewage municipal companies, providing property management services, passenger transport, sports facilities management, and supplying electricity, gas, steam, and air conditioning are more indebted in the long term (positive coefficient at the population_density variable for E, L, H, R and D sectors). Capital expenditures (which cause the growth of assets) for the water and sewage and waste management industry (E), especially in the case of larger companies supplying electricity, gas, and steam (D) require higher long-term debt financing, while lower financing for construction companies (negative correlation with the growth variable for sector F).

Municipalities with higher income from local charges with their water and waste management companies (E), paying higher taxes (tax_rate for the E sector), i.e., more profitable have less long-term debts. From this, it can be concluded that in the water and sewerage and waste management industry, the burden of long-term financing is borne by municipal companies (via off-balance-sheet debt). Likewise, municipalities with more profitable (higher ROA) transport companies (H sector) benefit to a lesser extent from long-term debt financing thanks to shares in the revenues of municipal companies from these public services.

Statistical tests were carried out for all companies together on the final model form. For all p-value models, the statistic F is less than 0.05, i.e., the assumed level of significance, which means rejecting the null hypothesis about the total insignificance of variables in the model ().

The results of the regressions allow verifying the validity of the hypotheses. The first hypothesis (H1a and H1b) was confirmed for both long-term and short-term debt. Based on the obtained results, there are no grounds for rejecting H1a nor H1b. These hypotheses concern a positive relation of revenue diversification to debt levels. The value of the coefficient at the revenue_diversification variable indicates that as a result of an increase in overall revenue diversification by a unit, the short-term debt per capita increases by 1.356 units and by 2.059 units per capita in the case of long-term debt. This means that the revenue diversification of municipalities does not help them to reduce their debt.

On the contrary, too low fiscal autonomy in terms of determining income by municipalities may contribute to an increase in debt to meet the needs of residents. The H2 hypothesis concerned the positive correlation between the leverage of companies and the level of indebtedness of municipalities was verified based on the results presented in Table 6. The results allow concluding that the subsidised municipalities show higher long-term debts, more affluent municipalities (with higher income per capita), and higher population density are more indebted, both short- and long-term. On the other hand, a higher share of local fees in the municipalities’ income allows them to reduce their debts, both short- and long-term. Municipalities with larger municipal companies have higher long-term debt, and owners of more profitable companies are less in debt in the long run.

The aim of this paper has been realised via confirming that municipalities in Poland use their municipal companies to increase debt capacity beyond the limitations imposed by the fiscal debt rules. Based on the results of the study, the hypotheses were verified. A positive relation between debt levels and revenue diversification of municipalities and a positive relationship between debt levels and leverage of municipal companies were shown. The results provide evidence of the off-balance-sheet financing using debt taken by municipal companies owned by local governments. As income per capita increases and population density increases, both debt levels of municipalities increase. On the other hand, as the share of local fees in the municipality's income (and fiscal autonomy in terms of income shaping) increases, the average debt per capita decreases. Additionally, the subsidies received and the sizes of the municipal company are positively correlated with the long-term debt level. In contrast, the profitability of the municipal company is negatively associated with the municipality's debt. Larger companies generally show higher transparency due to the better quality of financial reporting, which allows them to obtain financing from the bank. On the other hand, municipalities with shares in more profitable companies partially obtain financing from dividends (shares in profits generated by the municipal companies), which allows them to reduce their long-term debt. The results of our study show the opposite relationship than in the case of cities in the USA, but consistent with the limited fiscal autonomy of Polish municipalities indicated in the literature. However, they should be approached with caution.