Living in sprawling areas: a cost–benefit analysis in Poland

Abstract

Suburbanization still remains the key urbanization process in the world. Urban sprawl, as a specific manifestation of suburban morphology, belongs to the urban phenomena that is very often studied but still remains ambivalent. The ambivalence of the assessment results from the problems of weighing costs and benefits generated by sprawl. At the same time, there is a gap in international research on the costs and benefits of urban sprawl for households. Therefore, the aim of this article is to analyze and evaluate the household budgets in Poland in terms of potential costs and benefits of living in an urban sprawl area. The research hypothesis that households causing urban sprawl achieve net budgetary benefits was also adopted. In order to achieve the aim and verify the hypothesis, the data on the budgets of Polish households purchased from the Polish Central Statistical Office was used, and a questionnaire survey was carried out to determine the conclusions. The research hypothesis was not confirmed. The analysis and evaluation of many income and expenditure items of household budgets indicate that net costs in a significant group of budget items is high, while net benefits are relatively low. Households relocating to the sprawl zone, striving to meet their own needs, do not fully take into account all operating expenditures and ultimately incur net costs.

1 Introduction

Urban sprawl belongs to the urban phenomena that is very often studied but still remain ambivalent (Varela-Candamio 2017). Although urban sprawl can lead to serious consequences in reducing the quality of life of residents (Xu and Gao 2021; Shao et al. 2021), the phenomenon is still progressing in Europe (Hesse and Siedentop 2018). The economic consequences of urban sprawl are also the subject of contemporary research, dating back to the mid-twentieth century (Mumford 1961; Brueckner 2000; Chin 2002; Mendonca et al. 2020). The economic costs are primarily increases in public expenditures to build, expand, and maintain infrastructure and public services, longer commuting distances, energy consumption, and negative market impacts on the urban core (McHarg 1969; RERC 1974; Jackson 1985; Downs 1994; Bank of America 1995; Fulton et al. 2002; Brueckner and Largey 2008; Morelli and Salvati 2010; Daneshopur and Shakibamanesh 2011; OECD 2012). The negative impact of sprawl on household budgets has also been empirically proven: primarily in terms of commuting (Young et al. 2016; DiBartolomeo and Turnbull 2021; Larson and Zhao 2020). Yet households migrate to suburbs, making the location decision with a rather subequivalent cost–benefit calculus in mind. Indeed, lower suburban land or house prices and the free of chargé opportunity to use the environmental amenities for recreation become factors that attract these households to the suburbs (O'Toole 2009; Mendonca et al. 2020). These decisions are also supported by income growth and infrastructure investments that reduce travel times in the suburban zone (Klumer et al. 2014). The costs of living in an urban sprawl zone today are amortized by the growth of mobile occupations, i.e., those working professionally in more than one city, the development and economic importance of the Internet, and other means of telecommunication (Anas and Rhee 2007; Anas 2012).

The current state of research on the financial consequences of sprawl area households distinguishes studies on income (Hummel 2020) and on expenditure, which can be grouped into three streams: (1) transportation; (2) property maintenance; (3) consumption and lifestyle (DiBartolomeo and Turnbull 2021; Larson and Zhao 2020; Zhao and Wan 2021; Proque et al. 2020; Manganelli i in. 2020; Gielen et al. 2021; Jadach-Sepioło and Zathey 2021; Jończy et al. 2021; Śleszynski et al. 2020a; Dittmar 2019; Hilal et al. 2018; Acolin and Reina 2022). In economics, the assessment of suburban household income and expenditure is a complex process due to the coexistence of selected microeconomic mechanisms like: financial constraints and opportunities; preferences and utility maximization of the place of residence (Boeckermann et al. 2019; Lotfi et al. 2019; Acolin and Reina 2022; Booi and Boterman 2020). These mechanisms, along with migration and the ineffectiveness of spatial policies, are recognized as some of the most important determinants of urban sprawl (Daneshopur and Shakibamanesh 2011; Bhatta 2010; Śleszyński et al. 2020; Szafrańska et al. 2019). Thus, household economic calculation can distinguish categories of objective costs, but also subjective costs regarding preferences and utility. Additionally, consequences of sprawl are hard to be clearly evaluated, because the benefits collide with the problem of costs (Chin 2002; Wassmer 2002). Complexities in valuing costs and benefits in terms of the location of households in urban sprawl are solved by assessing net marginal costs (Gordon and Richardson 1996; Brueckner 2000). This refers to the additional costs of households created by a location decision in a suburban area. This viewpoint reveals the need to assess the unit costs and benefits of urban sprawl borne by households in order to demonstrate the efficiency of sprawl.

As Booi and Boterman (2020) note, despite narratives about the values of urban living, suburbanization still remains a key urbanization process. Moreover, the European Commission (2006) estimates that eastern and southern countries are the ones most at risk from an explosive process of urban sprawl. Research on the progress of urban sprawl and its costs in Poland is being currently conducted (Jadach-Sepioło and Zathey 2021; Jończy et al. 2021; Śleszyński et al. 2020; Lityński 2021). However, these studies are dominated by macroeconomic approaches. What is lacking are research approaches that balance in detail the costs and benefits of households located in an urban sprawl zone. There is a similar gap in the current state of knowledge and research in the international literature. Hence, the aim of this paper is to analyze and evaluate household budgets in Poland in terms of potential costs and benefits of living in an urban sprawl area. The research hypothesis that households causing urban sprawl achieve net budgetary benefits was also adopted. In order to achieve the aim and verify the hypothesis, the following structure of the article was adopted: Sect. 2 The Theoretical Framework is a review of the literature on the consequences of location in a sprawl area for household income and expenditure. Attention is paid here to the mentioned factors, namely three streams of research (transportation; property maintenance; consumption and lifestyle). Section 3 Materials and Methods is a presentation of the method of measuring the net budgetary effect for households causing urban sprawl in Poland. The research uses the data on Polish household budgets purchased from the Central Statistical Office, and in order to draw conclusions, a survey was conducted. Section 3. Results presents the results of the analysis and evaluation of Polish household budgets causing urban sprawl. Section 4 Discussion attempts to compare the results of the study with the literature discussed in Sect. 2. Section 5 Conclusions provides a response to the research hypothesis, general conclusions and recommendations for spatial policy based on the results of the study.

2 The theoretical framevork

2.1 The institutional context of urban sprawl in Poland

Suburbanization processes in Poland, especially in the form of urban sprawl, were launched after 1989 along with the political changes in the country, i.e. the transition from a centrally planned economy to a market economy. The significant increase in urban sprawl dates back to the second half of the 1990s. (Zborowski et al. 2012; Jakóbczyk-Gryszkiewicz 2011; Gorzelak and Smętkowski 2005; Lisowski 2005). During this period, a high level of demand for housing was reported, which was determined by several factors (Rembarz 2009; Brade et al. 2009; Zawadzki 2000; Jutajewska 2010): a decrease in housing investments in cities, a persistent shortage of apartments, an increase in the number of households, the need for housing self-fulfillment of the emerging middle class. The new middle class, including its preferences, are considered important factors in urban sprawl (O'Sullivan 2012; Booi and Boterman 2020). Housing preferences in the context of urban sprawl include (De Vos et al. 2016; Acolin and Reina 2022; Lotfi et al. 2019; Boeckermann et al. 2019; Acioly and Davidson 1996; Proque et al. 2020; Bhatta 2010): a preference for rural location, a desire to increase household living space, preference of single-family housing. These preferences are creating new demand for residential properties located in suburban areas and are valid both in Poland and other countries of the world (Barnes et al. 2001; Lityński and Hołuj 2017; Szafrańska et al. 2019; Dolan 2019; Śleszynski et al. 2020b).

With the adoption of the market economy by Poland, agglomerations with new, fragmented and autonomous local self-government authorities showed significant weaknesses in response to social housing needs. This resulted in dynamic changes in the spatial structure in connection with the individual and non-synchronized development of areas (Martyniuk-Pęczek et al. 2018). Urban sprawl was first observed in large urban areas, and then in medium-sized areas (Gałka and Warych-Juras 2011).

The systemic transformation in Poland, unfortunately, started the state of a deepening crisis in the sphere of Polish spatial management, the effect of which is the spontaneous development of the suburban space, and defective Polish law is indicated among the more important reasons. The sources of the crisis include (Parysek 2016; Kowalewski and Nowak 2018; Rozbiewski et al. 2019; Śleszyński et al. 2020): (a) lack of tradition and experience in building a democratic country and strong political and economic position of developers who are opposed to public control of urbanization processes; (b) neoliberal ideology of successive governmental authorities that made spatial management the subject of the market game: Polish liberalism led to the domination of property rights over the public interest and the primacy of the right to build over property rights; (c) attitude of society expressed in helplessness and distrust, which arose from a lack of knowledge: new urbanization challenges, which are of concern to civilized countries, are not of interest to Polish authorities, society and—to a low degree of—science.

The spatial growth of Polish urban areas through urbanization of the areas surrounding the core cities was not planned from the beginning of the transformation period. In the first place, the areas surrounding the city became an attractive speculative and investment offer due to the legal solutions introduced in 1989 of the Geodetic and Cartographic Law (Ustawa z dnia 17 maja 1989 r. Prawo geodezyjne), which allowed for the division of agricultural land into smaller plots—up to 0.3 ha—without the requirement to obtain an administrative decision approving the division. This law allowed farmers to divide agricultural land and sell it on the free market. Agricultural plots were generally divided without urban planning, i.e. freely, being the subject of a quick transaction on the market of land for single-family housing. Martyniuk-Pęczek et al. (2018) prove that the intensification of urban sprawl processes was also a derivative of the new tax policy (Ustawa z dnia 26 lipca 1991 o podatku…). At that time, construction and housing allowances were introduced to improve the poor condition of housing construction. Due to the lack of a developed mortgage lending system, these reliefs were used by households with financial resources—at least from the social middle class. The reliefs were in force in the years 1992–2004 and were related to, inter alia, with the purchase of land for housing development, with the simultaneous lack of the requirement to start a construction investment immediately.

At the same time, in the first half of the 1990s, the obligatory spatial planning was abolished, which meant that new construction investments could be built in areas for which no spatial plan was in force. The elimination of the obligatory nature then resulted from the ideological negation of the solutions used in the times of socialism, regardless of their legitimacy and effectiveness. At the same time, it was not noticed that the obligatory spatial planning was applied even in Western European countries. These restrictions were initially expressed in 1994 in the Act of on Spatial Development (Ustawa z 7 lipca 1994 r. o zagospodarowania przestrzennym). However, this act limited the possibility of issuing building permits in areas without a plan, and in the situation of low efficiency of communes in developing plans, it extended the validity of the existing ones (Dutkowski 2012). This situation was changed in 2003 by the Act on Spatial Planning and Development (Ustawa z 27 marca 2003 r. o planowaniu i zagospodarowaniu przestrzennym), which finally annulled the existing "old" spatial development plans and introduced administrative decisions (the so-called WZ) weakening public control over the space.

The research of Kowalewski and Nowak (2018) shows that half of the investments in Poland were carried out in areas without local plans. Replacing local plans by the WZ triggered corruption-generating phenomena in local government administration, which were related, for example, to (Zdasień 2003; NIK 2004): too broad scope of competences concentrated in the hands of a single, unsupervised official issuing the WZ; discretion in issuing WZ; combining the employment of public officials with private service activities provided in the same administrative area. This state of affairs continues even today. As a result, urban sprawl is intensifying Poland, as indicated by the OECD Report (OECD 2018a). In the period 1990–2014, population density declined in Polish agglomerations by 21%, and this decline occurred primarily after 2000 (OECD 2018b).

2.2 Household income potential

The primary household budget constraint is income (Mattingly and Morrissey 2014, Hummel 2020). The relationship between suburbanization and income attainment is captured in two cross-sections. The first, refers to a trend of research on the consequences of suburbanization of a monocentric city. In this trend, the administrative borders of the city do not play an important role. Both the labour market and the markets for goods and services are not limited by them. The basic role is played by the household remaining in the area of influence of the city. The larger it is economically, the more strongly it influences the labour market in terms of the quantity and quality of jobs and wages earned by households (O'Sullivan 2012). At the same time, in the communities surrounding the city, the spill-over effect can be observed, associated with wage levels that remain slighty lower, but close to the core city (Hołuj 2018). The second trend, refers to business suburbanization. Population suburbanization, which results in urban sprawl, is accompanied by the relocation of economic activity. Consequently, economic activity intensifies in suburban areas (Li et al. 2019). Suburban community characteristics, including demographic structure, education, and income, are attractive to manufacturers and service providers (Mayer et al. 2017). This is because suburban labor supply represents a sufficient level with wages lower than in the urban center; at the same time, proximity to the city enables professional training of workers (Sokołowicz and Zasina 2016; Hilal et al. 2018; Wu et al. 2020). This means that part of the households living in the sprawl zone receive lower income.

In the context of assessing the financial possibilities of suburban households, their ability to incur and repay debts is important. The ability to incur liabilities is determined by the income mentioned above, as well as by the assets already accumulated by households. In connection with the above the concept of an intergenerational assets transfer should be mentioned. This concept is related to the transfer of personal assets (e.g. real estate, cash, etc.) to subsequent generations (Bywalec 2017). In the last decades of the twentieth century, these transfers noticeably intensified in Poland due to the accumulation of fixed assets accompanying dynamic economic growth, which was not disturbed by adverse events such as wars (Śleszyński i in. 2021). Additionally, the impact of intergenerational assets transfer on suburbanization is compounded by the attitude of farmers, i.e. the original landowners, who, in order to secure the next generation, transfer land to them for house construction. Thus, households, through the accumulated and transferred wealth, increase their ability to incur and repay debts—especially mortgages. The availability of external means of financing property purchases, including mortgages and their low interest rates, is considered an important factor in urban sprawl (Hamidi et al. 2016; De Decker and Pascal 2011). Access to external sources of finance allows households to purchase property before they have accumulated the financial resources to buy it. The mortgage market for economic and urban development is of considerable importance. Properly developed credit mechanisms stimulate the development of cities, allowing the population to secure their living conditions earlier. Pathological credit mechanisms, on the other hand, may result in a spatial crisis in the form of left-over and uninhabited neighborhoods or a financial crisis of international scale (Rosenblatt and Sacco 2018).

2.3 Household expenditures

The stream of research on the transportation consequences of living in a sprawl area is dominated by the view of higher burdens on household budgets than households living in core cities (Dittmar 2019; Zao and Wan 2021). The problem of excess transportation costs in the context of households' decision to suburbanize plays a dominant role, as it is underpinned by weighing the private benefits of living in a suburban area (e.g., lower property costs) against the private costs of that decision. The higher the private benefits, the further away from the city center the household locates. At the same time, a morphological feature of urban sprawl is the spontaneous location of development. Such localization, making it difficult for public authorities to carry out road investments along the shortest line, results in a longer distance from the suburbs to the center (Śleszyński et al. 2020; Jadach-Sepioło and Zathey 2021). This distance primarily results in a longer commute. Also, it is common for suburban households to use a car for travel due to the long distances to destinations (Manganelli et al. 2020). Consequently, suburban areas experience higher fuel consumption, more cars per household, and excessive congestion (Travisi et al. 2010; Young et al. 2016; Proque et al. 2020). Hence, in low-density areas, higher spending has been observed in terms of (Ford et al. 2015; El-Geneidy et al. 2016; Mouter and Chorus 2016; Ojeda-Cabral and Chorus 2016; Litman 2022): fuel, insurance, vehicle maintenance, and the periodic purchase of additional vehicles. The phenomenon of longer commutes is also accompanied by fatigue affecting lower labor productivity (Litman 2022). In turn, low labor productivity can translate into lower household income. In contrast, research by Anas (2012) indicates that sprawl may not always result in negative impacts on household transportation costs. In his study, average vehicle use can be reduced due to adjustments in workplace location and the expansion of public transportation.

The stream of research on the real estate implications of sprawl area living distinguishes lower burdens on household budgets due to: property purchase and maintenance, lack of financially accessible homes in urban centers, and lower taxes. In general, property purchase costs are lower in rural areas than in the urban core, which encourages the choice of a suburban location (Ewing and Cervero 2010a). Most often, residents of urban areas seek an urban location, but lower property costs encourage them to choose a rural location (Mendonca et al. 2020). Parallel to the relatively low cost of suburban real estate is the problem of lack of financially accessible real estate for households in the central city (Kane et al. 2014). Financial accessibility, in the context of spontaneous suburbanization processes, is understood as the amount of real estate purchase and maintenance expenditures achievable for a middle-income household. The lack of financially accessible properties in the city, which in terms of area and quality meet the expectations of average households, encourages settlement in rural areas under the economic influence of the central city (Mattingly and Morrissey 2014). Real estate costs also include taxes and fees paid. In Poland, the amount of public tributes related to real estate is not high and does not play a major role in decision-making. It is different in countries where these tributes depend on the value of the property and their annual amount is very high, e.g. in the United States. Nevertheless, it should be noted that costs per capita related to social and technical infrastructure and public services are higher in suburbs than in the city (Gielen et al. 2021; Hortas-Rico 2014). Similarly, the cost of ongoing maintenance of this infrastructure (Ermini and Santolini 2017). Therefore, it is considered that the taxation of entities located outside the city should be higher to secure the infrastructure costs. However, in Poland it turns out that tax rates do not depend on the specifics of the spatial structure of the municipality, and in many cases taxes turn out to be lower in rural municipalities than in core cities (Lityński 2019). At the same time, the problem of the local tax system is the absence or low degree of development tributes imposed on investors developing space (Brueckner and Kim 2003). The issue here is the failure of the paid public tributes to cover the costs of infrastructure development to be brought to new construction investments. Infrastructure costs are shifted to the local government, which in Poland is obliged to secure the needs of local communities. In this light, for investors, the costs of constructing a real estate in the suburbs turn out to be limited only to projects realized on the owned land, without taking into account the necessity of its connection with the surrounding area. Therefore, the failure to include infrastructure costs in the investment costs of entities carrying out construction projects means that they do not affect the budget burden of households, and thus encourage the suburbanization of the population. The transfer of infrastructure costs to local authorities is not a phenomenon that is characteristic only of Polish conditions. Quigley and Rosenthal (2005) also notice this problem in the United States. In Canada, the problem was solved by introducing development charges, i.e. a fee charged to developers at the time of obtaining a building permit in order to cover the costs of infrastructure necessary to provide municipal services of a new investment, such as road, communication, water and sewage infrastructure, community centres, fire and police facilities (Baumeister 2012).

The stream of research on consumption and lifestyle of sprawl area households distinguishes the impact of location of the burden on household budgets from daily functioning (Ewing and Cervero 2010b; Li et al. 2019; Dodds and Dubrovinsky 2015). In general, the existential costs associated with daily household functioning appear to be lower in rural areas than in the urban center, which encourages the choice of a suburban location (Ewing and Cervero 2010a). However, this is not always the case, as in Poland, but also in other countries, the suburban zone is largely inhabited by consumers representing the middle or upper class (Bhatta 2010; Lityński and Hołuj 2017). For consumers with at least middle-income, shopping in malls has become an important part of their modern lives, and the diverse needs of this class are met in service-diversified stores within a single mall. Consumers, especially middle-aged couples with children, fully reflect the features and benefits of the services that are offered at the mall. These include the presence of entertainment venues alongside grocery, clothing, and beauty shops. The mall makes it possible to make one major purchase less often than several smaller ones, which is an advantage when living in a suburban area and depending on the car (Li et al. 2019). Suburbanization of middle-income populations also triggers a mechanism for increased household spending in suburban shopping malls. Household spending on shopping in suburban malls is significantly higher than spending in downtown malls. Living in a suburban area near a shopping mall, results (Li et al. 2019): first, in the abandonment of time-consuming and expensive commutes to shopping malls in the city; second, in more frequent visits to the suburban shopping mall as a local center for social life, entertainment, and leisure activities. Since the end of the second half of the twentieth century, research on the relationship between consumption and leisure time has been available. The shopping mall has become not only a place for shopping, but also a center of integration of shopping with leisure and tourism. The success of suburban shopping malls has for many years been determined by the entertainment component (Dodds and Dubrovinsky 2015). Research is also available indicating the varied purposes of visiting a suburban shopping mall (Chai et al. 2007): half of consumers for dining, a quarter for entertainment. In the cited studies, the share of consumers coming to a suburban mall for shopping purposes was much lower than in a downtown mall. The mechanism of increased spending may be some variation of Veblen's paradox or the bandwagon effect, whose motives for spending are also socio-culturally based and involve both luxury and common goods. At the same time, the propensity to consume is determined not only by short-term needs but also by long-term needs. In economics, the propensity to consume in the long-term horizon is explained by the life cycle hypothesis (Modigliani 1986). The basic assumption of the hypothesis is the rationality of household financial behavior, including planning expenditures for the entire period of one's life (Browning and Crossley 2001; Kawiński 2015; Bywalec 2017). Life cycle phases (pre-production, production, post-production) imply income levels, preferences, and the size and structure of household expenditures. Among the assumptions of the concept is that households seek to maximize the marginal utility of consumption given the income they earn. Kawiński (2015) notes that the size and structure of consumption depends to a limited extent on household rationality. On the other hand, Bywalec emphasizes (2017) that the dynamic civilizational changes of the past decades resulting in an increase in the wealth of societies, weaken the role of income in consumption decisions. This is because a shift of consumption determinants to cultural and socio-psychological factors is observed, which are characterized by significant variability and individualization.

3 Materials and methods

The aim of the study is to estimate the net financial costs and benefits to households living in urban sprawl areas. The study area is the communes where urban sprawl occurs in Poland (Lityński and Hołuj 2020). The study area is therefore the communes around the largest Polish cities that are the capitals of regions, which are indicated in Fig. 1.

Fig. 1

Source Own study based on Lityński and Hołuj (2020)

Research area: communes within urban sprawl, Poland.

The research hypothesis that households causing urban sprawl achieve net budgetary benefits was adopted. Proposed method and data sources adopted to verify the research hypothesis depreciate the problem of the functional structure of agglomerations, i.e. monocentric vs. polycentric. This is due to the fact of studying the structure of urban sprawl zone household budgets, not the flows between the center and the sprawl zone. For example, in the case of transportation expenditures, the direction and length of trips are not analyzed, but the transportation expenditures themselves without the importance of the destination.

Identifying the net effect involves comparing the finances of the analyzed group with those of the control group. The analyzed group consists of households that moved out of the city and now live in an urban sprawl zone. The control group, on the other hand, consists of households that are similar in type but whose new location decisions resulted in them remaining in the core city. For this identification of the net effect, d Cohena was used (Livingston et al. 2009):

$$d\,\mathrm{ Cohena}=\frac{{\mathrm{Av}}_{\mathrm{ag}}-{\mathrm{Av}}_{\mathrm{cg}}}{\mathrm{s}}$$

where Av_ag – average for the analyzed group, Av_cg – average for the control group, s – population standard deviation

The index of d Cohena evaluates the effect in the following ranges: 0 is no effect; 0–0.2 small; 0.2–0.5 moderate; 0.5–0.8 large and statistically significant; > 0.8 very large. When d Cohena occurs with a positive sign, then the effect is evaluated positively as a net benefit from household functioning in the urban sprawl zone; when it occurs with a negative sign it is evaluated negatively as a net cost.

There are two problems associated with the proposed research method regarding the identification of households for the analyzed and control group in Polish public statistics. The first problem arises from the fact that Polish statistics lack precise financial data for households causing urban sprawl. In the absence of specific data, approximation solutions are used in economics. Approximation consists of creating a solution that approximates the problem, which cannot be represented exactly in analytical form. Approaches of replacing unavailable data with other measures are known in regional studies, e.g.: Zaucha et al. (2015) use Personal Income Tax and Agricultural Tax Revenues to identify local GDP. Data on household finances is available for a fee from the Central Statistical Office and is a database from the Household Budget Survey (CSO HBS). This database collects data from an annual survey in which selected households record their income and expenditure in special booklets. This is therefore precise data. It gives the possibility to select respondents with respect to many categories e.g. year of house construction. However, a limitation of the collection is the inability to indicate the specific location of the household. The indication of location is limited to six categories of localities in relation to population, i.e.: above 500 thousand; 200–499 thousand; 100–199 thousand; 20–99 thousand; below 20 thousand; rural. Thus, the research proposes to first conduct a dedicated survey of households causing urban sprawl for the selected urban area, and then search CSO HBS databases for a group of households with a similar profile to those surveyed. In other words, appropriately filtered CSO HBS data can provide similar results to surveys among households causing urban sprawl. The second problem is to determine the control group, i.e. in relation to which the previously identified households should be compared. The mentioned mechanism of household preferences and utility maximization in terms of place of residence, and financial constraints and opportunities seems to be important here. For the selection of the control group, it is necessary to identify such households that chose to locate in a new house/apartment in the city instead of living in an outer zone. That is, the control group can be selected households from the core city.

The research method adopted takes into account the research problems, which involves three steps as shown in Fig. 2.

Fig. 2

Source: Own study

Research procedure.

The first step is to obtain a research sample that can be used as a model for large-scale studies. The first stage is therefore a survey among households that cause urban sprawl in the area surrounding the city of Kraków, Poland. Such a sample will be the basis of searching for a group of households with a similar profile in the next step. This involves searching for similar households in the public statistics of the CSO HBS. The technique used to identify the sample is a questionnaire survey, among households experiencing urban sprawl in the Krakow area. The survey technique used was PAPI. The drawing of such households for the survey was based on two address databases of the Central Center for Geodetic and Cartographic Documentation, Poland for the years 2015 and 2017. The addresses of households that live in new buildings in suburbs were drawn, i.e. those that appeared in the 2017 database, which were not present in 2015. Thus, 120 addresses were drawn from a database of 518 addresses. The survey was conducted by individuals with experience at the CSO in collecting financial data. The substantive scope of the survey was household income and expenditure. The structure of questions in the survey questionnaire corresponded to the structure of questions in the annual survey of household budgets conducted by the CSO (CSO HBS). The structure of the surveyed household income and expenditure is presented in Table 1.

Table 1 Surveyed household income and expenditures

Quantitative questions about finances were expanded to include open-ended questions that enrich the ability to formulate conclusions about how households function in urban sprawl. Basic metrics of the surveyed households are provided in Table 2.

Table 2 Metrics of surveyed households

Having the research sample, we may proceed to the second step of the research, which is approximation. The aim of the approximation is to identify in the CSO HBS databases such a set of households in the Kraków area, which would be similar to the research sample. Realization of such an approximation aim is connected with searching for such an arrangement of filters in CSO HSB that gives the same results as in the case of surveys. Checking the similarity of the survey results with the set of CSO HSB was done using the technique: t-test and gamma coefficient (Siegel and Castellan 1988). Using these techniques, two hypotheses were statistically tested:

H₀

Av_qu = Av_sz

H₁

Av_qu ≠ Av_sz

The null hypothesis is that the survey average is the same as the CSO HSB average. The alternative hypothesis assumes that the survey average is different from the CSO HSB average. The income and expenditures for which H₀ is satisfied are sought. Since the surveys were conducted in 2018 and the available CSO HSB databases were for 2016, the database was updated using wage growth and inflation data. The database filtering pattern obtained as a result of the approximation, is applicable not only to the Kraków area, but also to other urban areas studied. The database filtering pattern was used in the third step.

The aim of stage three is to estimate the net costs and benefits of operating households causing urban sprawl. This stage uses only the CSO HBS database. This is because it was assumed on the basis of the results of step two that not only in the Kraków area, but also in the whole country, appropriately filtered data from the databases gives the same results as the surveys. Therefore, step three consists of comparing selected income and expenditure of households causing urban sprawl (the analyzed group) within the control group. The control group consists of similarly filtered households in the databases, but located in the core city instead of the outer zone. The third step is to conduct research not only in the Krakow area, but also in other urban areas in Poland accepted for research. These areas are indicated in Fig. 1. The comparison of the two sets was made based on the d Cohena index mentioned earlier.

However, the proposed research method, like much of the original concept, may have some limitations. The biggest appears to be the generalization of the study group's adoption through the use of derived filtering of CSO data. Indeed, household budget statistics lack filtering items that would indicate that a particular household migrated from a particular city. Nonetheless, the proposed research procedure allows one to get as close as possible to inferring the financial consequences of urban sprawl for households. Limitations mean that the results of the research can advance the discussion of the financial consequences of urban sprawl by adding new elements of knowledge in terms of the balance of costs and benefits for households. However, the limitations of the proposed method make it impossible to prove these financial consequences with certainty.

4 Results

4.1 Test sample

The results of test sample obtained from the survey is shown in Fig. 3. The figure includes both the analyzed income and expenditure categories of the surveyed households.

Fig. 3

Source Own study

Test sample: average monthly income and expenditure of households in urban sprawl zone based on surveys (€, 2018).

In terms of household income, the highest is from employment at €1,311. Income from self-employment is slightly lower at €1,225. The third category in terms of amount is rental income from owned property at €320. These are generally income from apartment rental in the core city. These apartments are those properties that were occupied by households before they migrated to the urban sprawl zone. Household budgets are also fed by social benefits at €320.

In terms of monthly expenditures of the interviewed households, the high expenses that can be attributed to living and properties are of importance. These are mortgages €320 and, in the case of households without their own property, property rentals €252. Loans €252 and food €182 also make up a significant group of expenses, as do vehicle usage costs: fuel €72, repairs €137 and insurance €171. A detailed list of income and expenditure levels and other survey statistics can also be found in Table 4. Whereas, Table 3 presents the place of money spending of the surveyed households that live in the sprawl area of Kraków.

Table 3 Place of money spending by the surveyed households

One of the costs of urban sprawl cited in the literature is the negative market impact on the core city (McHarg 1969; RERC 1974; Jackson 1985; Downs 1994; Bank of America 1995; Fulton et al. 2002; Brueckner i Largey 2008; Morelli i Salvati 2010; Daneshopur i Shakibamanesh 2011; OECD 2012). This cost results from an outflow of consumers to the suburban zone and a reduction in the concentration of demand based on residents' expenditures, the effects of which are observed on the supply side, e.g. through a reduction in the distribution of services. On the other hand, due to suburban migration, consumer demand shifts territorially to the suburban zone. However, there is no assessment of the scale of the shift in consumer demand in Polish studies. At the same time, it should be noted that respondents to the open question expressed the opinion that it is cheaper in the suburban zone, i.e. the prices of consumer goods and services are lower than in the city. Hence, in the surveys conducted, respondents were asked to estimate the share of expenditure incurred in the suburban municipality of residence and in the core city.

The largest part of the budget of the surveyed households is expenditure on food. Suburban households spend more money on food in the city, as 48%, while in grocery stores located in urban sprawl they leave 44% of their expenses. The rationale for the relatively high share of food expenditures in the city can go in two cross-section directions. The first, grocery purchases are made on commuting to/from work. The second, the assortment structure of suburban stores does not match the preferences of migrants. Differences in food prices between the core city and the suburban zone do not amortize the signaled two disadvantages for consumers.

A high proportion of the budget of sprawl households is also spent on vehicle repairs. This is a category that, along with fuel expenditures, is part of the urban sprawl losses recognized in the literature. These households, often described as car-dependent, are forced to own a car because of the need to commute, and often they own more than one vehicle. These conditions determine higher expenditures. It would also seem, the availability in the sprawl area of vehicle repair facilities and repair prices are lower than in the core city, which encourages the use of sprawl facilities. However, it turns out that nearly 60% of vehicle repair expenses are incurred in the city, while only 34% are incurred in the commune of residence. Again, two lines of justification are perceived. The first, vehicles are serviced at authorized auto repair shops that have a location in the core city. This servicing formula generally applies to new cars. However, in the case of slightly older cars whose warranty period has ended and there is no obligation to use expensive authorized service centers, consumer habits may play an important role. This refers to the habit, often long-standing, to the facilities that migrants used before living in the city. This is the second direction of justification for the high share of expenditures made in the city. In the case of fuel expenditures, where the majority of respondents (54%) indicated the city as the place of expenditure, the primary justification may be refueling vehicles on the way to or from work. It also happens that urban fuel stations are more reliable in terms of fuel quality than sprawl stations owned by less recognizable fuel companies.

Other budget categories such as clothing, electronics, pharmaceuticals, sports and restaurants also have a higher share of spending in the city than in the suburban zone. The only category for which respondents indicated that the sprawl zone predominates is education. Consequently, sprawl households leave 56% of their consumer spending in the city, while only 25% in the sprawl zone. It seems that the mentioned assortment structure of sprawl consumption supply does not match the preferences of migrants. The spending habits acquired during the period of living in the core city and making purchases on the way to or from work also play a role. Although prices are lower in the sprawl zone, the determinants of spending in the city make a big difference in the location of shopping.

4.2 Approximation

Searching the CSO HSB databases for a set of households similar to the respondents' households involved testing the H₀ and H₁ hypotheses. A total of 70 approaches to hypothesis verification were conducted using t-test and gamma coefficient. Finally, the results of the approximation show that in order to talk about the similarity of households, the CSO HBS database should be filtered according to the following criteria: location in the considered region; location outside the core city; period of house construction after 2007, obtaining income from non-agricultural activities, location in villages and towns with a population of up to 99 thousand. The results of the approximation are presented in Table 4.

Table 4 Approximation statistics: average, t-test and gamma coefficient

There are some categories in the income and expenditure set that should be dropped in the subsequent research procedure. These are financial categories for which the counts are too small in the HSB CSO databases. Such a group includes property rental income and expenditures for: property rental, hot water, association, propoperty insurance, health insurance, and vehicle insurance.

In the set of examined expenditures there are categories for which the value of the test verification homogeneity of variance is greater than 0.05 (V_sd²), which does not give grounds to reject H₀. At the same time, for these variables the probability t-test is less than 0.05 (p). This means that for these variables the alternative hypothesis H₁ is positively verified, i.e.: the average between the questionnaire and the CSO database are significantly different. From the point of view of research procedure, these expenditures are also omitted in further analysis and are: alcohol, cigarettes & drugs; medical services; TV & internet.

The remaining group of income and expenditure are the categories for which (V_sd²) > 0,05, (p) > 0.05 and gamma coefficient is high, which means that hypothesis H₀ is positively verified. This group of income and expenditure is considered similar, which from the research point of view allows for its inclusion in the third step of the research. The income group includes: employment, non-agricultural self-employment, social benefits. The group of expenses includes: food; clothing; cold water; gas; electricity; garbage; sewage; heating; maintenance; property tax; mortgages; loans; pharmaceuticals; fuel; vehicle repairs; transport tickets; phone; electronic devices; sports; restaurants; tourism & cinemas; education.

4.3 Cost–benefit analisys

The results of the approximation, i.e. the database filtering pattern and the set of income and expenditures fulfilling H₀, were used as a starting point for the cost–benefit analysis. This analysis applies not only to households in the Kraków area, but also to those located in other examined urban areas. The presented cost–benefit analysis uses the averages for income and expenditure for the analyzed group and the control group, which are presented in Fig. 4. Based on the results for the analyzed group and the control group, d Cohena analysis was performed, which is presented in Table 5.

Fig. 4

Source Own study

Average monthly household income and expenditure of the analyzed group and the control group (€, 2018).

Table 5 d Cohena rezults

When analyzing income, each category is lower than the control group. The most important income categories of each household, i.e. employment or non-agricultural self-employment, represent d Cohena at − 0.39 and − 0.34 respectively. These values are in the range (− 0.0; − 0.5) indicating net moderate costs. Income from social benefits, is also lower in the analyzed group, with d Cohena − 0.13. Thus, households in the analyzed group also have net costs from income stemming from social benefits, with relatively low net costs.

For expenditures, both net costs and net benefits were identified. In terms of net benefits these are: clothing, gas, mortgages, loans, pharmaceuticals, sports, tourism & cinemas, educations. Net benefits for clothing and sports are insignificant as d Cohena is 0.03 and 0.05 respectively. Also relatively low net benefits were identified for pharmaceuticals (0.12) and tourism & education (0.11). However, there is a group of expenditures from which sprawl area households obtain slightly higher net benefits. These are expenditures on: gas (0.27), mortgages (0.41), loans (0.29), and education (0.42). The listed financial categories are in the range (0.2; 0.5) which means moderate net benefits of living in sprawling area.

However, a larger group of effects than net benefits are net costs to households living in sprawling areas. Fourteen expenditure categories were identified that are detrimental to household budgets. Bearing in mind the ranges against which the d Cohena is applied, the examined net costs can be divided into four groups: small, moderate, large, very high. The group of small net costs includes expenditures on: vehicle repairs (− 0.01), electronic devices (− 0.4), restaurants (− 0.14), transportation tickets (− 0.18). The net moderate cost group included expenses for: maintenance (− 0.25), phone (− 0.26), heating (− 0.36), garbage (− 0.38), food (− 0.43). The net cost group at large, in the group included expenses for: fuel (− 0.52), cold water (− 0.59), sewage (− 0.72). The net cost group at very high, group included expenses for: property tax (− 1.41), electricity ( − 1.52).

5 Discussion

Urban sprawl is associated with financial mechanisms in household budgets. These mechanisms are expressed in net benefits and net costs and concern both income and expenditure of households.

On the income side, the financial mechanism is expressed in the net costs incurred, which represent a moderate level. The net costs, which we can call losses, are mainly caused by lower wages and business income. These results confirm that households located in a sprawl area may have lower income due to local employment (Mayer et al. 2017; Wu et al. 2020). Conversely, households that commute to the urban center may also earn lower incomes due to proffesional stagnation resulting from reduced labor productivity (Litman 2022). As a consequence of commuting and fatigue, and further lower labor productivity, suburban households have lower professional outcomes than core city residents. This results in slower promotion or lower wages over several years. In contrast, the presented research results do not support the observation that income does not depend on where one lives in a metropolitan area (Mattingly and Morrissey 2014; O'Sullivan 2012). Thus, the observation that both the labor market and the markets for goods and services are not administratively limited, and the primary role is played by remaining in the area of influence of the city, which is the metropolitan core, is not confirmed.

On the expenditure side, the financial mechanism involves both net costs and benefits. The findings will therefore be presented in the context of research streams discussed in the literature related to: transportation; property maintenance; consumption and lifestyle.

Transportation consists of budget items such as fuel, vehicle repairs, and transportation tickets. In each of these financial categories, sprawling area households incur net costs, where fuel is among the large net costs. These results are therefore consistent with previous knowledge indicating that sprawling area households incur significant transportation costs (Ford et al. 2015; El-Geneidy et al. 2016; Mouter and Chorus 2016; Ojeda-Cabral and Chorus 2016; Young et al. 2016). In the sprawl area in Poland, the road network is not functional. The roads are winding, with low capacity (Śleszyński et al. 2020). Therefore, travel time is increased. These conditions determine higher expenditures on fuel and vehicle repairs, and the long distance of residence to different destinations also entails higher expenditures on public transport tickets. Thus, the results of this study are consistent with previous research indicating that the disorganized location of suburban developments, while causing longer distances, also results in more frequent use of the car for travel, and consequently greater fuel consumption and higher expenditures (Litman 2022). The opinion of households surveyed regarding transportation costs also indicates that nearly 60% of expenditures are incurred in the city, while only 34% are incurred in the community of residence. This implies a shift away from the lower priced services of a suburban no-toll station or car service in favor of more expensive entities located in the city. The use of these services in the core city is explained primarily by consumer habits or the higher reliability of cities entities. This habituation is associated with years of use of auto services used prior to residence change from the core city. The reasoning is similar for fueling. The majority of households indicated the city as the place to purchase fuel. City fuel stations are more reliable in terms of fuel quality than suburban stations owned by less recognizable fuel companies. As a consequence of frequent vehicle use, longer distances and repairs, and refueling in the city, transportation expenses are very high resulting in a net cost to sprawl households.

Property maintenance consists of expense categories such as cold water, gas, electricity, garbage, sewage, heating, maintenance, propoerty tax, and mortgages. The analysis of property maintenance expenses distinguishes both net costs and benefits. In terms of net benefits, gas and mortgages were identified and are moderate. In contrast, the remaining categories are net costs. Moderate net costs are: maintenance, heating and garbage. Large net costs are: cold water and sewage. Very high net costs are: property tax and electricity. When balancing benefits against costs, net costs dominate. These results are contrary to the state of the art in suburban real estate maintenance costs. Generally in the literature, property purchase and maintenance costs are lower in rural areas than in the urban center, which encourages the choice of suburban location (Ewing and Cervero 2010a; Kane et al. 2014; Mattingly and Morrissey 2014). Thus, a conclusion can be made about the apparent lower cost of maintaining an urban sprawl property without considering all cost categories. One example is property taxes. The amount of property tax rates in the suburbs should be higher to secure infrastructure costs (Ermini and Santolini 2017; Hortas-Rico 2014). But in Poland, tax rates do not depend on the specific spatial structure of the municipality, and in many cases rates turn out to be lower in rural communes than in core cities. This is because in Polish tax rates are set annually by the communes council. Authorities of suburban communes, wishing to attract residents and further increase their income, lower property tax rates. In Poland, this is a manifestation of race to the bottom (Gerlak et al. 2020). Thus, in the metropolitan area there is a large variation of these rates. Moreover, property taxes are calculated on the area of the property, not on its value as in many countries. Lower tax rates encourage households to relocate to sprawl area. However, despite the lower tax rates, the considerable area of a sprawl house, but especially of a plot of land, leads to a much higher property tax burden. In the presented research, the net cost generated on this account were identified as very high.

Keeping consumption and lifestyle in mind, in the presented research it refers to: food, loans, clothing, electronic devices, sports, restaurants, tourism & cinemas. Most of these expenditures do not generate a noticeable losses or benefits. However, food and laons expenditures are already associated with noticeable net costs. Although the urban sprawl zone appears to be an area of lower grocery prices according to the surveyed households, it appears that these households use urban stores. The preference for using urban grocery stores is related to the fact that they are made on the occasion of commuting to and from work. The second factor shaping the preferences in question, is the unsatisfactory assortment structure of suburban stores. The differences in food prices between the core city and the sprawl zone do not amortize these two disadvantages for consumers. Since households in the urban sprawl zone are located far away from the stores in which they shop, they stock up in purchases, often in shopping malls. Thus, similar to previous studies, household spending in shopping malls is significantly higher for suburban households (Li et al. 2019). At the same time, a significant proportion of consumer spending is done on credit or paid by credit card. This means that loans of sprawl households generate net costs. As a result, the food and loan expenses of these households turn out to be high, generating significant budget losses.

High consumption expenditures with lower incomes of households located in the urban sprawl zone can be related to the life cycle hypothesis (Modigliani 1986; Browning and Crossley 2001; Kawiński 2015; Bywalec 2017). The current state of research on the hypothesis provides justification about the diminishing role of income level in consumption decisions. This is based on observation in a shift of consumption determinants to cultural and socio-psychological factors, which are characterized by significant variability and individualization, is observed. The size and structure of consumption to a limited extent depends on the rationality of households (Kawiński 2015). Hence, sprawling households, despite lower incomes formulate a consumption structure at the level of wealthier, urban households. Migration to the sprawl area does not lead to a reconfiguration of the structure of consumption, and further, of expenditures for its satisfaction.

6 Conclusions

Financial mechanisms occurring in the budgets of households located in the urban sprawl zone distinguish both net costs and benefits. The most important sources of income, ie employment or non-agricultural self-employment, for sprawl zone households are lower than for cities households. Cohen's d index is − 0.39 and − 0.34, respectively, which was assessed as a net loss moderate. For expenses, both net costs and net benefits have been identified. The group of expenses for which net benefits were demonstrated is not numerous, and their level was defined as the moderate benefits (Cohen’s d 0.2; 0.5). These benefits relate to expenditure on: education, gas, mortgages and loans. In turn, the group of net costs includes more numerous financial items. There are 14 expenditure categories that are unfavorable to household budgets. Among them, there are categories that were assessed as large and very high net costs (Cohen’s d ≥ 0.8), e.g. fuel, cold water, sewage, property tax, electricity. Losses prevail on the expenditure side of the household budgets of the sprawl zone. Taking into account losses in income, a negative budget balance for living in sprawl becomes noticeable. Hence, the research hypothesis that households causing urban sprawl achieve net budgetary benefits should be falsified.

The results of the research add value to the current state of knowledge especially by indicating the estimated amounts of losses and benefits for household budgets from living in urban sprawl area. The research also verifies many other international studies in Polish conditions. The conclusions can be presented as follows:

On the income side, the net costs from living in an urban sprawl zone are proven. These results confirm that sprawling households may have lower incomes due to local employment (Mayer et al. 2017; Wu et al. 2020) or reduced productivity due to long time-consuming commutes (Litman 2022). In contrast, the research results presented here do not support the observation that income does not depend on where one lives in a metropolitan area (Mattingly and Morrissey 2014; O'Sullivan 2012).

On the expenditure side, the net cost and benefits mechanism was demonstrated, which concerns: transportation; property maintenance; consumption and lifestyle. In terms of transportation, high net costs have been proven, which is consistent with previous knowledge (Ford et al. 2015; El-Geneidy et al. 2016; Mouter and Chorus 2016; Ojeda-Cabral and Chorus 2016; Young et al. 2016). In the sprawling area of Poland, the road network is not functional. It has been shown that not only longer distances affect the amount of losses, but also drivers' habits of servicing and refueling their vehicles in a more expensive city. In terms of property maintenance, both net benefits and costs are shown. When balancing benefits with costs, net costs dominate. These results are at odds with the state of the art in suburban property maintenance costs, where maintenance costs are lower (Ewing and Cervero 2010b; Kane et al. 2014; Mattingly and Morrissey 2014). In terms of consumption and lifestyle, on the other hand, it has been shown that spending on food and laons is already associated with noticeable net costs. This confirms other studies that shopping expenditures, especially in shopping malls, are significantly higher for suburban households (Li et al. 2019). Also observed, the diminishing role of income level in consumption decisions due to cultural and sociopsychological factors. Suburban households, despite their lower income formulate a consumption structure at the level of wealthier, urban households.

The conclusion can be made that households located in a sprawl area are individuals who incur financial losses to maximize housing preferences. The research presented here partially supports the perception of suburbia as an area where the ongoing costs of daily living are lower (Ewing and Cervero 2010a). Originally assumed by residents making the decision to relocate to an urban sprawl area, the vision of lower daily living costs did not in fact fully materialize. Rather, research supports Parysek's (2004) thesis that individual households relocating to a sprawl zone in pursuit of their own needs and ambitions do not fully consider all costs of functioning. It is about the lack of awareness of households about the differences in living costs in several dimensions: Firstly, households do not include all components that make up their income, especially in terms of expenses, in the economic calculation. Secondly, within the considered financial components, they do not take into account their full scope (e.g. they equate transport expenses with fuel costs only; they do not take into account the more frequent costs: repairs, tire and oil changes, changes to another car due to faster depreciation, etc.). Thirdly, households demonstrate a low accuracy of forecasting even those costs that they were aware of (e.g. they assumed that they would spend their expenses in a cheaper suburban zone, but in fact they would spend them in a more expensive city due to: better assortment, habits, along the way for work, etc.). The differences in living costs in the above-mentioned cross-sections are especially noticeable in the results of the survey research. However, due to the territorial scope of these studies, i.e. one agglomeration, the issue of awareness of differences in living costs seems to be an interesting direction of future research, but with a wider territorial scale, allowing for generalization.

The results of the research can be applied in formulating spatial policy—especially in Poland. The dominant results indicate the negative impact of urban sprawl on household budgets. It should be noted that in Poland urban sprawl results mainly from the location decisions of households and to a small extent from the decisions of business entities. Therefore, if location decisions in the sprawl area have negative effects on both households and local government budgets, the general assessment of urban sprawl should be negative. The identified losses and benefits, as well as the financial mechanisms, should nuance the public policy addressed to the problems generated by urban sprawl. Spatial policy should focus on directing urban sprawl. Such a policy should stimulate the emergence of compact suburban settlements. Such structures, in turn, will have a positive impact on the burden not only on households but also on municipalities by reducing length of streets and its costs of maintenance. The proposed direction of urban sprawl expressed in the stimulation of compact suburban settlements also results from the fact that urban sprawl belongs to the category of natural spatial processes. The naturalness results from the preferences of households, which cannot be influenced by public authorities and, in principle, cannot be stopped by settlement processes. Thus directing urban sprawl, which in the first place in public policy is expressed in the formulation of objectives. The formulation of goals should take into account several postulates. Firstly, it should prevent from locating residential developments at a distance from the existing ones, stimulating development in the vicinity of other buildings. Secondly, it should be stimulated to increase the spatial continuity of development in communes, by supporting the location in the free spaces between existing buildings. This increases the efficiency of utilization of the existing municipal infrastructure and further expenditures on its maintenance and construction. Thirdly, communes should be provided with instruments to coordinate individual investments that would enable new developments to be grouped into compact settlements. Fourthly, the three postulates should be tightened together with the distance from the core city. While instruments stimulating localisation in a dense, concentrated, continuous and grouped into compact settlements could prevail in the communes directly adjacent to the core city, restrictions limiting new development may be imposed in other communes.