Shaping the Rural Landscape: Institutions of Land Use Change in Non-Urbanized Areas in Poland

Abstract

Although extensive research has examined land use dynamics in urban settings, non-urbanized areas remain understudied, particularly in post-socialist contexts. This study investigates the political, economic, and institutional drivers of land use changes in the non-urbanized areas of Poland through the Political Market Framework. It focuses on how local government structures and interest group interactions shape land development decisions. The two-stage Heckman model was used to address sample selection bias. The results reveal three key findings: (1) municipal council composition significantly influences land use outcomes, with farmers and highly educated councilors generally opposing intensive development, while partisan councilors favor it; (2) interest groups, particularly production companies and the tourism sector, drive both the initiation and scale of land use changes; and (3) institutional factors, including agricultural land quality and farm size, moderate these relationships. This study advances our understanding of rural land use governance by demonstrating how political–economic forces interact with local institutional contexts to shape spatial development patterns. These findings have important implications for land use policy in transitional economies and contribute to broader discussions on rural governance.

1. Introduction

Land use changes in non-urbanized areas represent a critical environmental and socio-economic challenge, particularly in transitional economies where institutional frameworks are still evolving. While urbanization patterns and their drivers have been extensively studied [1,2,3,4,5,6,7,8,9,10], the complex dynamics of land use changes in non-urbanized regions remain understudied. These regions, characterized by lower population density and greater land availability, face unique development pressures that differ fundamentally from those of their urban counterparts [11]. These areas are often targets for agricultural expansion, resource extraction, and residential or commercial projects [12]. However, the socio-economic and financial limitations of local governments in these areas influence land use decisions [13,14,15].

Local governments in non-urbanized areas frequently depend on external funding or subsidies because of their limited tax bases, compelling them to promote development projects that promise to enhance local revenue generation or employment opportunities [16,17,18]. These circumstances, coupled with the influence of diverse stakeholders, including businesses and residents, create a governance dynamic that traditional land use change studies have largely overlooked. While extensive literature has concentrated on urban growth management and national-level policy implications, non-urbanized areas, particularly in post-socialist contexts such as Poland, confront distinct challenges and opportunities that warrant comprehensive investigation [19,20]. This research aims to investigate how local government structures and interest group interactions influence land use decisions in non-urbanized areas, using Poland as a case study.

This research addresses this knowledge gap by applying the Political Market Framework (PMF) to examine the political, economic, social, and institutional drivers of land use transformations in Poland’s non-urbanized areas, where local government entities must balance competing interests while making land use decisions under significant development pressures [3,5,7,8,21,22,23,24,25]. The findings of this research enhance our understanding of how political, economic, and institutional forces influence land use governance, with significant implications for regions worldwide facing comparable challenges. Through the PMF lens, this study not only highlights the importance of local governance structures in land use decision-making but also offers a detailed analysis of the role of interest groups and their interactions in shaping land use changes.

2. Theoretical Framework

2.1. Political Market Framework (PMF)

Land use changes represent multifaceted phenomena influenced by diverse political, economic, social, and institutional factors. This research employs the PMF to conduct a systematic analysis of these changes, conceptualizing policy outcomes as products of dynamic interactions between policy suppliers (decision-makers) and policy demanders (interest groups) [2,3,26]. In Poland, where local governance structures significantly influence land use decisions, the PMF provides an invaluable analytical lens for understanding the dynamic interplay between various actors, institutions, and incentives that generates observed patterns of land development.

The PMF’s theoretical foundation lies in institutional analysis, drawing substantially from Elinor Ostrom’s Institutional Analysis and Development (IAD) framework [27]. The framework conceptualizes policies as outcomes of negotiation processes between suppliers, which include decision-makers such as municipal councils, mayors, and demanders, which are interest groups competing for finite resources [26,27]. This negotiation process illuminates preferences for diverse policy outcomes and the propensity of decision-makers to prioritize certain interests (Figure 1). Through its focus on the formal and informal rules structuring decision-making processes, the PMF allows for a sophisticated examination of how local governance contexts in Poland shape land use outcomes.

Interest groups, comprising associations of individuals or organizations united by shared objectives, constitute crucial demand-side actors [28,29,30]. Unlike supply-side actors, who seek to control policy-making through electoral success, interest groups aim to influence powerholders indirectly by advocating for aligned policies. The literature presents different categorizations of interest groups, with Logan and Molotch [31] distinguishing between beneficiaries and defenders regarding urban growth. This distinction, based on anticipated resource values, elucidates the conflicting pressures confronting policy-makers. Beneficiaries typically pursue regulatory changes that enable more intensive resource utilization, primarily motivated by exchange value—financial gains from resource development. Conversely, defenders generally seek to maintain the status quo or limit development, often prioritizing user value—benefits derived from the current utilization or preservation.

Furthermore, supply-side actors encompass governmental entities with decision-making authority, including elected officials [26]. Their actions are often influenced by political incentives, such as re-election prospects, public opinion, and party alignment [32,33,34]. By considering career aspirations, ideological preferences, and operational costs, the framework transcends simplistic public interest or pure self-interest assumptions, explaining apparent policy contradictions and incremental changes in land use policy [26].

The interaction between the demand and supply sides occurs within the “action arena” in Ostrom’s IAD framework [27]. This arena facilitates the emergence and evolution of institutions that govern resource allocation. These institutions act as moderators, shaping policy outcomes by determining stakeholder access to decision-makers and interest aggregation methods [26,35]. The PMF’s emphasis on the mediating role of institutions, particularly through rules-in-use [36]—informal rules shaped by norms, customs, and community attributes—provides a comprehensive understanding of land use decision-making contexts. This institutional focus helps explain the varying outcomes across jurisdictions despite similar demand pressures or supply-side incentives.

2.2. Application of PMF in Other Studies

The PMF has demonstrated significant utility in examining land use policy and governance dynamics. Lubell et al. [7,8] explored the influence of institutional collective action and intergovernmental collaboration on land use outcomes, emphasizing the critical role of local governance structures and interest group dynamics. Their work revealed the importance of understanding these dynamics when analyzing land use policies.

Further studies extended this understanding by investigating how local governments manage growth through various policy instruments, reinforcing the relevance of the PMF in examining development pressures [21]. Hawkins [22] advanced this area of research by applying the framework to sustainable development policies, demonstrating how political institutions mediate between interest groups and policy-makers in land use decisions, balancing competing priorities such as economic growth and environmental sustainability.

Additionally, Deslatte et al. [3] investigated metropolitan land use decisions by examining political fragmentation and delays in policy implementation, which serve as strategic tools in growth management. Their findings highlighted the framework’s capacity to explain the complexities of urban governance and land use decisions. Similarly, Tong et al. [25] extended the framework’s application to infrastructure investment patterns in Chinese cities. These studies collectively confirm the effectiveness of the PMF in analyzing land use changes and governance dynamics, establishing its suitability for investigating the political and economic drivers of land use change in non-urbanized regions of Poland.

2.3. Literature Review and Hypotheses

Building on the theoretical foundation established through previous applications of the PMF, this study examines the factors influencing land use decisions in non-urbanized areas of Poland. The PMF suggests that both supply- and demand-side factors, along with their interactions, are fundamental for understanding these processes. Through a synthesis of the PMF’s elements, this study proposes the following hypotheses to investigate the drivers of land use change in non-urbanized areas of Poland.

2.3.1. Importance of Supply-Side Factors in Non-Urbanized Areas

Since 1990, Poland’s spatial planning system has transitioned from a centralized, state-controlled model to a decentralized framework, with municipalities playing a pivotal role in shaping land use policies. Legislative reforms shifted planning authority to municipalities, established structured procedures, and emphasized public participation, ensuring better integration across governance levels [37]. This transformation reflects Poland’s broader post-communist democratization, while highlighting the challenges of balancing local autonomy with national and regional planning priorities to achieve sustainable development. Numerous studies have emphasized the importance of mayors in land use policies [1,7,9,38], including research conducted in Poland [4]. This emphasis is logical, given that mayors typically function as executive authorities responsible for planning processes. However, the current literature exhibits a notable gap regarding the role of municipal councils, which serve as the primary legislative bodies responsible for approving zoning plans in Poland [37]. Consequently, this study concentrates on the composition of the municipal council and its role in shaping land use decisions. The council’s composition—comprising representatives from agricultural communities, such as farmers, individuals with higher education, and those affiliated with national political parties—significantly impacts the formulation and implementation of land use policies, particularly in contexts where diverse interests and motivations intersect to shape local governance.

• Composition of Local Councils: Farmers in Municipal Councils

Farmers face the threat of agricultural land conversion when alternative land uses offer greater economic returns [39,40,41]. However, in circumstances where agricultural activities remain economically viable, farmers frequently resist land use intensification to protect their livelihoods [42,43]. Farmers serving on municipal councils play a crucial role in shaping land use decisions because their vested interest in preserving agricultural land influences policy outcomes. Their representation typically strengthens advocacy for maintaining agricultural land usage and resistance against conversion pressures toward non-agricultural purposes [4,44]. Hence:

H1: 

A higher share of farmers in the municipal council is positively associated with increased resistance to land use changes that lead to non-agricultural development.

2.

Composition of Local Councils: Councilors with Higher Education

The educational background of councilors also plays a critical role in shaping land use policies. Councilors with a higher education demonstrate a greater propensity to advocate for policies that harmonize development with sustainability objectives, reflecting their comprehensive understanding of economic, social, and environmental implications [45,46]. A higher proportion of educated councilors can lead to more informed decision-making processes that prioritize sustainable land use, aiming to protect environmental and community values [2,3,7,10,21,24]. Hence:

H2: 

A higher share of councilors with a higher education in the municipal council is positively associated with increased resistance to land use changes.

3.

Composition of Local Councils: Partisan Councilors

The involvement of partisan councilors or those aligned with national parties substantially influences local land use policies. These representatives typically demonstrate a heightened responsiveness to the broader political environment and may shape their policy decisions with national political agendas in mind [9,10,46,47]. This influence is particularly pronounced in Poland, where national policies emphasize economic growth and development [19]. Moreover, Poland has received substantial European Union Cohesion Policy funding for infrastructure and modernization [48], which pressures local governments to align with pro-development goals. Consequently, a higher proportion of partisan councilors typically results in more unified and decisive policy-making, favoring land use changes consistent with larger political objectives. Hence:

H3: 

A higher share of partisan councilors in the municipal council is positively associated with increased land use changes toward more intensive development.

4.

Composition of Local Councils: Mayoral–Council Congruence

Political dynamics, particularly the level of mayoral authority, play a crucial role in shaping land use outcomes [6,47,49]. A stronger mayoral authority can either accelerate or hinder development projects contingent upon political agendas and the represented interests [8]. Strong council support for the mayor facilitates more cohesive policy-making [50,51]. This effect is particularly pronounced in non-urbanized areas with limited financial and political resources, where political cohesion disproportionately impacts policy outcomes. The level of support a mayor receives can greatly influence the extent of land development, especially when political actors seek additional revenue through land use changes [42,52,53]. Hence:

H4: 

Stronger congruence between the mayor and the council is positively associated with increased land use changes toward more intensive development.

2.3.2. Importance of Demand-Side Factors in Non-Urbanized Areas

5.

Beneficiary Interest Groups: Manufacturing and Production Companies

Local businesses, including production companies, have emerged as principal beneficiaries of land use policies in non-urbanized regions [24,31,54,55,56,57]. These areas offer extensive land availability at lower costs than urban centers, making them attractive for companies requiring extensive space for industrial facilities, such as factories, warehouses, or logistical centers [12,20,53]. Reduced land acquisition costs minimize initial capital expenditures and ongoing operating expenses, which is particularly advantageous for companies seeking to optimize their production budgets. Furthermore, local governments often provide incentives through tax concessions and flexible zoning regulations to attract production companies [49]. The economic needs of these areas often lead local governments to prioritize industrial growth over environmental conservation [58]. Such policies aim to generate jobs, increase tax revenues, and enhance the overall economic vitality in financially constrained areas. Additionally, a lower population density in non-urbanized areas typically results in reduced organized resistance to industrial expansion. This environment facilitates a more efficient negotiation of land use modifications and regulatory approvals. Hence:

H5a: 

The presence of manufacturing and production companies is positively associated with increased land use changes toward more intensive development.

A key characteristic of an interest group is its wealth and endowment in tangible resources, which reflect both the group’s income and size; these factors can serve as indirect measures of the group’s ability to influence the decision-making process [59]. Hence:

H5b: 

Higher contributions of property tax from businesses to municipal revenues are positively associated with increased land use changes toward more intensive development.

6.

Beneficiary Interest Groups: Tourism Sector

Tourism is another crucial factor that influences land use policies in non-urbanized areas. This sector stimulates local economies through revenue generation by creating local employment opportunities and supporting businesses that cater to tourists’ needs [60,61]. However, these economic benefits often come with an increased demand for land to develop tourism infrastructure such as recreational facilities, accommodation, and supporting amenities. Consequently, tourism-oriented policies often transform local landscapes by converting agricultural or natural land into commercial use [62,63,64]. This growth creates a “sustainability paradox” in which development potentially degrades the natural amenities and environmental assets that initially attract tourism [8]. The number of overnight stays can serve as a measure of the strength of the tourism sector in a municipality. This reflects the increasing demand for land development to accommodate tourists, including the construction of hotels, resorts, and other tourism-related infrastructure. Hence:

H6: 

The number of overnight stays provided is positively associated with increased land use changes toward more intensive development.

7.

Beneficiary Interest Groups: New Residents

Population growth through migration significantly shapes land use policies in non-urbanized regions [65]. As new residents move into these regions, they generate increased demand for housing [66], infrastructure [67], and services, driving intensive development [68]. Local governments typically respond by adapting policies to accommodate these demands through residential expansion, infrastructure enhancement, and enhanced public services [53,69]. While migration-driven growth presents significant governance challenges, especially for local governments with limited sustainable development capacity [70], migration-generated revenue through property taxation incentivizes residential development and infrastructure expansion to sustain financial health and meet the demands of their growing populations [71,72]. Hence:

H7a: 

A higher positive migration rate is positively associated with increased land use changes toward more intensive development.

H7b: 

A higher share of municipal revenue from property taxes paid by private individuals is positively associated with increased land use changes toward more intensive development.

8.

Defender Interest Groups: Residents with Higher Levels of Education

Defenders prioritize the user value of land, emphasizing the benefits derived from its current use, such as environmental quality and existing lifestyles. While any resident can be a part of a defender’s group, many residents lack the motivation to acquire the necessary knowledge or engage in decision-making processes because the provision of public goods and services is not typically conditional upon their participation. Defenders often react to ongoing regulations rather than proactively engaging in policy-making, which means that they may not always be well prepared or organized to influence decisions effectively [31].

However, residents with higher levels of education demonstrate increased civic engagement and environmental awareness [73] and advocate for sustainable development practices [2,3,7,10,21,24]. Hence:

H8: 

A higher share of residents with higher education levels is positively associated with increased resistance to land use changes toward more intensive development.

9.

Defender Interest Groups: Farmers

Farmers represent crucial defender interests in non-urbanized areas and often act as defenders of land use, prioritizing the current agricultural use of land over its potential development for other purposes [21,74]. Although defenders generally have limited organizational resources compared with beneficiaries [31], farmers maintain cohesive interests in agricultural land preservation [18,44,75]. As primary landowners and users, they hold valuable resources that they can utilize according to formal rules outlined in zoning plans. Their interests typically align with the preservation of the agricultural value of their land [42,43]. Hence:

H9a: 

A higher share of agricultural businesses within a municipality is positively associated with increased resistance to land use changes toward more intensive development.

Farmers’ influence on land use policies can also be measured indirectly through their financial contributions to municipal revenues through agricultural taxes. A significant share of revenue from these taxes indicates that local governments have a vested interest in maintaining agricultural land, as they serve as a crucial source of funding [76].

H9b: 

A higher share of municipal revenue from agricultural taxes is positively associated with increased resistance to land use changes toward more intensive development.

The impact of farmers on land use policies is not only shaped by their size and financial contributions but also by the economic viability of their farms. Farmers whose agricultural activities generate sufficient income are generally less inclined to support land conversion because they prioritize the user value of their land over their potential exchange value [42,77]. However, if their farms do not generate an adequate income or profits, it may increase farmers’ openness to alternative land uses or economic activities [39,40,41].

Small farms (under 1 hectare) in Poland often face limited economic viability, relying heavily on subsidies and government support [78,79]. Hence:

H9c: 

A high share of small farms positively mediates the influence of the supply side, leading to less resistance against land conversion and favoring policies that support more intensive land use changes.

Additionally, the condition of the agricultural land is another critical factor in determining productivity and sustainability. Farmers managing well-maintained agricultural land are likely to be strong advocates for policies that protect it from being converted to non-agricultural uses [43,78]. Hence:

H9d: 

The proportion of agricultural land in a good condition positively mediates the influence of the supply side, leading to reduced resistance against land conversion and favoring policies that support more intensive land use changes.

3. Data and Methods

3.1. Non-Urbanized Areas

The operationalization of non-urbanized areas relies on a one-dimensional spatial classification developed by Szmigiel-Rawska et al. [80]. Using static Corine Land Cover (CLC) data from the European Environment Agency, Poland’s 2477 municipalities are categorized as urbanized or non-urbanized based on the proportion of anthropogenic areas (CLC class 1: artificial surfaces) within their total area. Non-urbanized municipalities are defined as those where anthropogenic areas constituted up to 14% of the total area in 2012; there were 2087 such municipalities in 2012. These municipalities are predominantly characterized by agricultural land, forests, wetlands, or water bodies rather than human-transformed landscapes. Spatial classification is preferred over administrative definitions because it more accurately reflects actual land use patterns, avoids distortions caused by the heterogeneity of administrative boundaries, and provides a more precise identification of non-urbanized areas. This approach aligns with the focus of this study on land use dynamics and ensures compatibility with the land use change index (LUCI), which measures transitions to more intensive land uses.

3.2. Dependent Variable

The dependent variable captures actual land use data from 2012 to 2018 using the CLC dataset developed by the European Environment Agency. This comprehensive database classifies land cover into 44 categories, grouped under five main headings: artificial surfaces, agricultural areas, forests and semi-natural areas, wetlands, and water bodies. According to the Copernicus database, features are identified and classified on satellite images with a positional accuracy better than 100 m and maintain a minimum mapping unit of 5 hectares. With the smallest municipality in Poland covering 335 hectares and an average size of 12,624 hectares (median 11,190 hectares), these parameters enable the precise analysis of land use transformations within territorial units.

The current research uses the LUCI index as its dependent variable, following the methodology of Lubell et al. [8], Rasmussen and Weber [81], and Deslatte and colleagues [4]. This index measures the transitions from more extensive land uses (for example, forests and semi-natural areas) to more intensive applications (for example, artificial surfaces), calculated as the sum of land use changes toward more intensive categories, divided by the total area of each municipality. The relative measure (m²/km²) facilitates cross-municipal comparisons while controlling for size variations.

3.3. Independent Variables

The independent variables in this study are aligned with the PMF’s supply-side, demand-side, and contextual factors, and each set of variables is linked to a specific hypothesis (H1–H9). The primary data sources for the analyses in this study are the Main Statistical Office, the National Electoral Commission, and the Ministry of Finance (

Table 1. Independent variables, sources, and descriptive statistics.

Hypotheses	Variable	Description	Source	Min	Max	Mean	SD
Supply side (composition of local councils)
H1.	Farmers in municipal council	Average share of councilors declaring farmer as their occupation during the 2010–2014 term, %	Main Statistical Office	0	100.00	29.96	22.57
H2.	Councilors with higher education	Average share of councilors with higher education during the 2010–2014 term, %	Main Statistical Office	0	95.24	28.88	16.41
H3.	Partisan councilors	Share of councilors belonging to or supported by national parties elected in 2010, %	Main Statistical Office	0	100.00	25.00	26.07
H4.	Councilors supporting mayor	Mayor’s support in the council after the 2010 elections, based on councilors’ affiliation with the mayor’s electoral committee, %	National Electoral Commission	0	100.00	49.47	20.50
Demand side
Beneficiaries
H5.	Production companies	Average share of enterprises in the production, transportation, and storage sectors among total enterprises during the 2010–2014 term, % Average share of property tax from legal entities in municipal budget revenue during the 2010–2014 term, %	Main Statistical Office Ministry of Finances	1.28	45.76	10.89	4.22
				0.54	57.16	7.33	6.00
H6.	Tourism sector	Number of overnight stays provided per capita during the 2010–2014 term	Main Statistical Office	0.00	213.46	1.69	10.09
H7.	New residents	Average migration balance during the 2010–2014 term Average share of property tax from private persons in municipal budget revenue during the 2010–2014 term, %	Main Statistical Office Ministry of Finances	−329.75	968.50	1.76	68.05
				0.24	18.12	3.04	1.84
Defenders
H8.	Residents with higher level of education	Share of residents with higher education in the county in 2011, %	National Census	8.30	29.80	12.82	2.57
H9.	Farmers	Average share of enterprises in the agriculture sector among total enterprises during the 2010–2014 term, % Average share of agricultural tax in municipal budget revenue during the 2010–2014 term, % Share of farms with an area less than 1 hectare in the total number of farms, 2011, % Share of agricultural land in good agricultural condition in total agricultural land, 2011, %	Main Statistical Office Ministry of Finances Agricultural Census Agricultural Census	0.22	42.45	7.33	5.25
				0.00	15.61	2.99	2.39
				0.00	52.80	2.53	4.11
				0.00	100.00	96.48	5.59
Contextual variables
Wealth of residents		Average share of personal income tax in municipal budget revenue during the 2010–2014 term, %	Ministry of Finances	2.14	46.96	11.09	5.06
Distance to voivodeship		Distance to the voivodeship capital, km	Own calculations	6.50	219.00	75.52	27.26
Distance to county		Distance to the county capital, km	Own calculations	0.00	59.30	16.68	10.16
Large farms		Share of farms with an area greater than 15 hectares in the total number of farms, 2011, %	Agricultural Census	0.00	95.69	46.51	28.46
Expenditure of EU funds		Expenditures from European Union funds in PLN per capita, 2012–2018	Ministry of Finances	0.00	3633.42	177.36	146.61
Construction of national roads, highways		Dichotomous variable, 1 = highway and/or national road development between 2012 and 2018, 0 otherwise	General Direction for National Roads and Motorways	0.00	1.00
Exploitation fee		Average value of exploitation fee per capita from the years 2010–2012	Ministry of Finances	0.00	1382.00	9.25	46.69
Area of the municipality cover with forest		Proportion of municipal area covered by forests, 2012, %	Corine Land Cover	0.00	93.94	30.20	19.29
Area of the municipality cover with water		Proportion of municipal area covered by water, 2012, %	Corine Land Cover	0.00	80.68	1.74	4.46
Local spatial development plans		Proportion of municipal area covered by local spatial development plans, 2012, %	Main Statistical Office	0.00	100.00	30.04	40.66
Number of zoning decisions per 1 km2		Number of issued zoning decisions per 1 km2, 2011 2014	Main Statistical Office	0.00	14.27	0.97	1.43
Land use change index 2006–2012		Area of the municipality transformed toward more intensive use, m2/1 km2, 2006–2012	Own calculations	0.00	108,394.61	2042.23	6926.14

). All variables related to the supply side are operationalized through a description of the composition of the municipal council. To test H1, the presence of agricultural interest groups is measured by the proportion of councilors declaring farmers as their occupation during the 2010–2014 period. For H2, the proportion of councilors with a higher education is used. H3 is examined through the proportion of councilors affiliated with national political parties. Finally, H4 assesses mayoral support by measuring the proportion of councilors from the mayor’s electoral committee.

Interest group variables utilize dual measurement approaches where data permit: the presence and size of groups within the municipality and their financial contributions to municipal budgets. For H5, the presence of production companies is measured by the proportion of enterprises in the production, transportation, and storage sectors, along with the share of property taxes paid by legal entities. To assess H6, tourism activity is captured by the number of overnight stays provided per capita. H7 considers the influence of new residents, measured by the net migration balance and the share of property tax paid by individuals. For H8, the proportion of residents with a higher education is used. H9 is the most comprehensive hypothesis, operationalized by four variables: the share of agricultural enterprises, proportion of agricultural tax in municipal revenues, proportion of small-size farms, and agricultural land in good agricultural condition.

The analysis incorporates several contextual controls to account for the broader factors affecting land use changes. Resident wealth is measured through the share of Personal Income Tax (PIT) in the municipal budget, which reflects the economic capacity of local populations [82]. Proximity to voivodeship and county capitals is controlled for, as shorter distances to these urban centers tend to increase development pressure [77,83]. The proportion of large farms is also tested. Expenditure of European Union (EU) funds is another critical factor, as EU funding often supports infrastructure projects that drive land transformation [4,84,85], and in Poland, almost 40% of Cohesion Policy funding between 2007 and 2018 was allocated to non-urbanized areas [48]. Additional control variables include national road and highway construction (dichotomous), because significant land use changes are concentrated near major roadways that are under the jurisdiction of the central government. Similarly, the variable for coal mining—the exploitation fee—is included as a control to account for land use changes driven by mining activities. A significant portion of the observed changes involves the conversion of class 2.1 (arable land) to class 3.1 (mines, dumps, and construction sites), and local governments in Poland lack direct authority over the issuance of coal mining permits, as these decisions are made at the national level by the Ministry of Climate and Environment. There are also control variables related to land availability and scarcity [86,87], including the proportion of municipal area covered by forests and water bodies. Additionally, formal rules are captured through the proportion of municipal land covered by local spatial development plans and the number of zoning decisions issued per square kilometer, which describes whether changes occur more frequently within the framework of previously prepared plans or through more spontaneous, individual decisions. The final variable is the magnitude of the changes that occurred previously (2006–2012) in the municipality, to control for long-term processes that may take place in a municipality.

3.4. Two-Stage Heckman Selection Model Specification

This study employs a two-stage Heckman model to address potential sample selection bias, which can arise from varying degrees of land use transformation across municipalities. The Heckman model is particularly suitable in this context because of the presence of a large proportion of non-urbanized municipalities (1166 of 2087) where no land use changes were observed, alongside a subset of 921 municipalities where such changes were recorded. The selection bias in this study arises from the non-random nature of land use transformations, as only certain municipalities experience changes influenced by demand-side, supply-side, and contextual pressures. Ignoring this bias could lead to unreliable parameter estimates, either underestimating or overestimating the effects of key explanatory variables.

In the Heckman model, the error term from the selection equation is crucial for addressing potential selection bias. If this error term is uncorrelated with the independent variables, as assumed in traditional models, selection issues do not distort the results. However, if the error terms from the selection and outcome equations are correlated, selection bias occurs because the observations in the outcome equation are not randomly sampled from the population. This leads to inconsistent and biased parameter estimates [88]. The Heckman model corrects for this by incorporating the inverse Mills ratio (IMR) into the outcome equation. The IMR estimates the effect of the error term on the outcome and adjusts for selection bias introduced by the non-random sample. By including the IMR as an additional explanatory term, the Heckman model ensures unbiased and consistent parameter estimates, enabling a robust analysis despite the presence of selection bias [88].

To address the complexity of the influence of interest groups on land use change, two separate models were specified. The first incorporates tax-related variables into the selection equation and direct presence measures into the outcome equation. The second reverses this configuration, enabling a robust test of their respective roles in land use change dynamics. Both models incorporate variables affecting the likelihood and magnitude of land use changes, such as formal institutional rules (for example, local spatial plan coverage and zoning decisions), pressures for transformation (for example, EU fund allocations), and path dependency (for example, LUCI 2006–2012). Variables primarily influencing the probability of land use changes—such as proximity to voivodeship and county capitals, infrastructure development, and land availability—are included in the selection equation. Conversely, the outcome equations integrate variables that operationalize the interaction between supply-side factors (local government composition) and demand-side pressures (interest groups), reflecting core PMF dynamics (

Table 2. Model specifications.

Model Specification	I	II	III/IV
Selection equation	$y_s=a_i\beta +c_i\gamma +d_i\delta +u_i$	$y_s=b_i\beta +c_i\gamma +d_i\delta +u_i$	$y_s=a_i\beta +c_i\gamma +d_i\delta +u_i$
Outcome equation	$y_o=b_i\beta +c_i\gamma +e_i\delta +\lambda \theta +u_i$	$y_o=a_i\beta +c_i\gamma +e_i\delta +\lambda \theta +u_i$	$y_o=b_i\beta +c_i\gamma +\left(b_i\times c_i\right)\eta +e_i\delta +\lambda \theta +u_i$

). The significant and negative IMR values across all models confirm the presence of selection bias, further validating the use of this methodology. This study presents only significant interaction models.

• $a_i$: a vector of independent variables capturing the tax-related variables of interest groups’ presence;
• $b_i$: a vector of independent variables capturing the direct presence of interest groups;
• $c_i$: a vector of independent variables related to local government composition;
• $b_i\times c_i$: an interaction term representing the relationship between interest group characteristics and local government composition;
• $d_i$: a vector of contextual controls in the selection equation;
• $e_i$: a vector of contextual controls in the outcome equation;
• $\lambda$: derived from the selection equation, it adjusts for the potential selection bias that arises when the sample is non-random. Its inclusion ensures that the estimates of coefficients in the outcome equation are unbiased and consistent;
• $\beta , \gamma , \delta ,\theta ,\eta$: the vectors of coefficient groups of independent variables, representing the relationship between the variables and the dependent variable;
• $u_i$: the error term accounting for unobserved factors.

4. Results

The findings reveal that municipal council composition significantly impacts land use decisions in non-urbanized areas (

Table 3. Results.

	I	II	III	IV
Observations censored	1166	1166	1166	1166
Observations observed	921	921	921	921
(Intercept)	. −2.10 ** (0.70)	. −2.61 *** (0.74)	. −2.10 ** (0.70)	. −2.10 ** (0.70)
Residents with higher levels of education		17.46 *** (3.68)
New residents		0.002 ** (0.001)
Agricultural enterprises		. −0.14 (0.14)
Production enterprises		0.48 * (0.21)
Tourism		0.01. (0.00)
Property tax residents	0.68 *** (0.17)		0.68 *** (0.17)	0.68 *** (0.17)
Property tax legal entities	0.36 ** (0.11)		0.36 ** (0.11)	0.36 ** (0.11)
Agricultural tax	. −0.04 * (0.02)		. −0.04 * (0.02)	. −0.04 * (0.02)
Wealth of residents	0.72 ** (0.28)		0.72 ** (0.28)	0.72 ** (0.28)
Farmers in municipal council	. −0.33. (0.18)	. −0.62 *** (0.19)	. −0.33. (0.18)	. −0.33. (0.18)
Councilors with higher education	. −0.01 (0.23)	0.41. (0.23)	. −0.01 (0.23)	. −0.01 (0.23)
Councilors supporting mayor	. −0.07 (0.15)	. −0.10 (0.15)	. −0.07 (0.15)	. −0.07 (0.15)
Partisan councilors	0.18 (0.12)	0.17 (0.12)	0.18 (0.12)	0.18 (0.12)
Distance to voivodeship	0.00 (0.00)	0.002 ** (0.00)	0.00 (0.00)	0.00 (0.00)
Distance to county	. −0.01. (0.00)	. −0.01 *** (0.00)	. −0.01. (0.00)	. −0.01. (0.00)
Forest	. −0.05 (0.21)	0.45 * (0.18)	. −0.05 (0.21)	. −0.05 (0.21)
Water	. −0.09 (0.72)	0.52 (0.75)	. −0.09 (0.72)	. −0.09 (0.72)
Spatial development plans	0.08 (0.09)	0.09 (0.09)	0.08 (0.09)	0.08 (0.09)
National roads, highways	0.65 *** (0.19)	0.68 *** (0.19)	0.65 *** (0.19)	0.65 *** (0.19)
EU funds	0.00 (0.00)	. −0.00 (0.00)	0.00 (0.00)	0.00 (0.00)
LUCI 2006–2012	0.00 (0.00)	0.00 (0.00)	0.00 (0.00)	0.00 (0.00)
Small farms	. −2.23 * (0.90)	. −2.89 ** (0.91)	. −2.23 * (0.90)	. −2.23 * (0.90)
. Large farms	0.68 *** (0.15)	0.98 *** (0.14)	0.68 *** (0.15)	0.68 *** (0.15)
Agricultural land in good condition	0.43 (0.61)	0.56 (0.65)	0.43 (0.61)	0.43 (0.61)
Zoning decisions	0.09 ** (0.02)	0.10 ** (0.03)	0.09 ** (0.02)	0.09 ** (0.02)
Exploitation fee	0.01 *** (0.00)	0.01 *** (0.00)	0.01 *** (0.00)	0.01 *** (0.00)
(Intercept)	2693.74 (4058.23)	3806.83 (3828.04)	. −8601.03. (5017.49)	4316.81 (4134.07)
Residents with higher levels of education	3191.24 (17,248.56)		2488.83 (17,116.42	1468.28 (17,227.61)
New residents	0.10 (2.14)		0.85 (2.14)	0.09 (2.14)
Agricultural enterprises	. −283.33 (730.71)		. −323.59 (725.22)	. −337.43 (729.48)
Production enterprises	2298.87. (1246.55)		2266.35. (1237.24)	2173.26. (1245.19)
Tourism	27.06. (14.52)		27.08. (14.40)	25.62. (14.50)
Property tax residents		232.32 (944.08)
Property tax legal entities		15.71 (617.73)
Agricultural tax		. −1.38 (85.89)
Wealth of residents		1420.61 (1404.08)
Farmers in municipal council	. −902.10 (1076.67)	. −957.78 (1068.23)	. −1007.53 (1068.14)	. −1303.36 (1091.41)
Councilors with higher education	. −2842.24 * (1148.79)	. −3204.16 ** (1171.17)	. −2715.99 * (1139.54)	. −4034.18 ** (1284.47)
Councilors supporting mayor	608.60 (849.35)	616.35 (847.54)	741.43 (843.03)	673.25 (847.65)
Partisan councilors	1659.74 * (670.80)	1650.65 * (675.04)	34,460.74 *** (8806.78)	1644.01 * (669.02)
Spatial development plans	1139.97 * (504.63)	1033.80 * (502.69)	1175.39 * (500.55)	1143.64 * (503.28)
EU funds	2.34. (1.30)	3.15 * (1.36)	2.25. (129)	2.21. (1.30)
LUCI 2006–2012	0.20 *** (0.02)	0.20 *** (0.02)	0.19 *** (0.02)	0.20 *** (0.02)
Small farms	1783.90 (5265.82)	1572.37 (5276.47)	2195.84 (5223.48)	. −22,279.76. (808.68)
Large farms	623.86 (805.80)	256.94 (877.37)	331.98 (802.89)	438.78 (808.68)
Agricultural land in good condition	. −2052.53 (3630.57)	. −2316.73 (3656.98)	9669.98 * (4775.20)	. −2884.93 (3643.81)
Zoning decisions	103.41 (129.69)	85.71 (126.28)	101.86 (128.58)	106.54 (129.33)
Partisan councilors × agricultural land in good agricultural condition			. −33,999.96 *** (9102.79)
Councilors with higher education × small farms				54,869.66 * (26,707.59)
Councilors with higher education × large farms
Degrees of freedom	2046	2046	2045	2045
Adjusted R-squared	0.11	0.10	0.12	0.11
Invmillsratio	. −1274.12 * (639.21)	. −1233.35. (722.73)	. −1200.01. (658.19)	. −1238.78. (637.71)

Note: *** indicates p < 0.000, ** indicates p < 0.001, * indicates p < 0.05, . indicates p < 0.1. Bolded variables are statistically significant, regardless of the level of significance.

). Farmers in municipal councils consistently emerge as a significant factor in the selection equation across all models, albeit with marginal significance, while showing no significant impact in the outcome equation. This pattern indicates that agricultural representation primarily influences the occurrence rather than the magnitude of land use changes. The negative coefficient suggests that councils with a higher representation of farmers exhibit increased resistance to land use changes, reflecting farmers’ typical role as defenders of agricultural land. These findings align with H1, which posits that farmers prioritize preserving their land over development initiatives.

Councilors’ educational attainment demonstrates a varied significance across the stages of the analysis. Within the selection equation modeling change occurrence, the educational level achieves significance in one model. However, the outcome equation, examining the magnitude of change, reveals that councilors with a higher education consistently produce significant negative effects. This suggests that educated councilors moderate the scale of changes, likely prioritizing sustainability and cautious development, aligning with their broader understanding of long-term socio-economic and environmental implications. These observations support H2, which predicts that councilors with a higher education would correlate with resistance to intensive land use changes.

Next, partisan councilors—those affiliated with national political parties—exert a consistent positive significance in the outcome equation across all the models. Their presence correlates with more intensive land use changes, suggesting an alignment with pro-development agendas. This evidence supports H3, which proposes that partisan councilors, responsive to national growth policies, would drive larger-scale land use transformations in non-urbanized areas.

Interestingly, mayoral support within councils demonstrates no significant impact on either the selection or outcome equations across all the models. This finding suggests that political cohesion, as indicated by mayoral support within the council, does not directly influence land use decisions in this context, contrary to H4’s prediction that unified councils may facilitate development more easily.

Overall, the comprehensive analysis of council composition across both equations reveals that the variables demonstrate greater significance in the outcome equation, suggesting that while the presence of certain groups affects the occurrence of land use changes, their influence on the scale of changes is even more pronounced.

Interest groups play a crucial role in both the occurrence and magnitude of land use changes. Production companies exhibit positive significance in both the selection and outcome equations, confirming H5a. Their presence drives initiation and contributes to more extensive larger-scale transformations, indicating their role in promoting and shaping extensive land use changes.

The financial variable, such as property tax from legal entities, achieves significance exclusively in the selection equation, confirming H5b. This implies that financial pressure from businesses primarily influences change occurrence rather than the scale of those changes. This suggests that financial contributions from legal entities determine the likelihood of land conversion without directly impacting the magnitude of the transformation.

Similarly, the tourism sector demonstrates positive significance across both the occurrence and magnitude dimensions of land use changes, confirming H6. Municipalities with an active tourism industry are more likely to experience land use changes and extensive transformations. These findings support the view that tourism development plays a central role in driving significant land transformation, both in initiation and expansion.

New resident influence manifests primarily in the selection equation, showing a positive correlation with change occurrence and validating H7a. This indicates that an influx of new residents creates pressure without significantly affecting the transformation scale of land conversion. Similarly, the financial variable of property tax from private individuals appears only in the selection equation, confirming H7b. This further suggests that residential financial pressure triggers but does not determine the magnitude of change.

The presence of residents with a higher education shows a similar pattern, being significant only in the selection equation. Interestingly, the positive coefficient contradicts H8, which suggests that better-educated residents would resist land use change. This suggests that educational attainment correlates with the occurrence of change, without measurably impacting the transformation scale.

The final group of farmer variables reveals complex patterns. Agricultural enterprise share shows negative but insignificant effects across the models, failing to confirm H9a. However, the financial measure of agricultural tax demonstrates a negative significance in the selection equation, confirming H9b. This indicates that tax contributions from farmers effectively resist initial land use changes without influencing the subsequent transformation magnitude. The presence of larger farms significantly positively correlates with the occurrence of changes in the selection equation. Contrarily, a higher proportion of small farms is linked to resistance to land conversion. This finding generally contradicts expectations based on economic efficiency, where smaller farms are assumed to have a weaker stance in defending land from development.

Interestingly, small farms emerge significantly in the outcome equation through their interaction with the presence of councilors with a higher education, confirming H9c. While educated councilors are generally associated with smaller-scale land use changes, this relationship shifts to municipalities dominated by small farms. This interaction indicates that in municipalities where small farms prevail, educated councilors may be more willing to support land use changes, despite generally advocating for more sustainable development. This suggests that educated councilors act strategically, possibly recognizing that small, less viable farms may not have a sustainable future in agriculture and that land conversion might present an opportunity for broader economic development.

The condition of agricultural land outcomes through interaction effects confirms the mediating H9d hypothesis. Although the variable for land in good agricultural condition is independently insignificant, it becomes significant in the outcome equation when interacting with national party-affiliated councilors. These councilors are generally associated with promoting larger-scale changes. This interaction suggests a diminished pro-development influence among partisan councilors in areas with a higher share of well-maintained agricultural land. This indicates that even pro-development councilors may resist land use changes when the land in question is particularly valuable for agricultural purposes.

Several contextual variables demonstrate a significant influence on land use outcomes. The wealth of residents shows a positive significance in the selection equation, potentially reflecting that wealthier municipalities are more likely to experience land use changes. This may be related to the presence of better-educated individuals who typically earn more and seek higher-quality living conditions in non-urbanized areas, ultimately leading to more intensive land use changes. These variables have a correlation coefficient of 0.33, which is statistically significant at a p-value < 0.01.

The distance relationship between voivodeship and county capitals presents interesting dynamics. As hypothesized, a greater distance from county capitals is associated with a lower probability of land use changes. However, the opposite pattern is observed for voivodeship capitals, where land further from regional centers is more likely to experience changes. This might be due to a higher demand and competition for land closer to voivodeship capitals, driving up prices and making land conversion more difficult. In contrast, land near smaller centers is more economically available, leading to higher rates of development.

The allocation of EU funds affects the magnitude of land use changes rather than its occurrence. In all models, a positive significance indicates larger-scale changes in municipalities receiving greater EU funding support. This underscores the role of external financial support in driving more intensive development projects in non-urbanized areas.

Infrastructure development, particularly national roads and highways, consistently shows a positive significance in the selection equation, emphasizing that municipalities with better access to transportation networks are more likely to undergo development.

Natural resource extraction influences changes in non-urbanized municipalities. The presence of valuable natural resources creates strong economic incentives for local governments and industries to promote land conversions. However, its specific impact on the scale of these changes can vary depending on the resource type, extraction intensity, and the regulatory framework in place.

The availability of land through forests and water coverage shows an inconsistent predictive power. In one model, a higher share of forested areas is associated with an increased likelihood of changes, suggesting that forests may be particularly vulnerable to land use transformations. However, this result is inconsistent across all models, indicating that the impact of forest coverage on land use changes may depend on additional factors or vary across different contexts.

Spatial development plans have emerged as a crucial factor in shaping the scale of land use changes. The outcome equation depicts the positive significance associated with more intensive land use changes, suggesting that formal planning provides a structured environment that facilitates larger-scale development in non-urbanized areas. The importance of spatial development plans underscores the role of institutional arrangements in guiding land use outcomes, offering municipalities a clear path for implementing land transformations in a coordinated manner.

Conversely, zoning decisions achieve significance solely in the selection equation, indicating an influence over land use change occurrence without affecting the magnitude of these changes.

The final critical factor is the LUCI (2006–2012) variable, which demonstrates a consistent positive significance in the outcome equation. This suggests that past land use changes are a strong predictor of the scale of current transformations, indicating a cumulative process in land development. This indicates that previously transformed municipalities predict the current scale of change, reflecting a path-dependent trajectory in land use evolution.

5. Discussion

This study advances the literature on land use changes by applying the PMF to the Polish context, extending its applicability beyond urban and national settings. The findings underscore the critical role of political actors and interest groups in shaping land use decisions, yielding significant implications for policy formulation and governance in rural areas.

The findings confirm that municipal council composition is a decisive factor in land use changes, offering additional dimensions beyond previous research that predominantly focused on executive bodies, mainly in the United States [2,3,6,7,8,21,47]—or on a single-variable analysis of council characteristics in Europe [4,9,10,38]. The opposition of farmers in municipal councils to land use changes reinforces the role of agricultural interests in rural land preservation, aligning with studies documenting resistance to urbanization [42,43]. Similarly, councilors with a higher education demonstrate a tendency to curb the magnitude of land use changes, consistent with earlier research linking education with more sustainable development policies [2,3,7,10,21,24]. These educated councilors prioritize long-term planning in environmentally sensitive rural areas, although their positive interaction with small farms reveals strategic decision-making, potentially supporting the development of economically vulnerable farms.

In contrast, partisan councilors favor more intensive land use changes. While previous studies on partisanship have predominantly focused on the political orientation of councilors or mayors [9,10,46,47], this study specifically examines national party affiliations or the support they receive from these groups. In Poland, such affiliations significantly influence land use changes, regardless of political alignment. This suggests economic growth imperatives, particularly in closing development gaps with wealthier countries—as is the case with Poland in comparison to Western Europe—may overshadow traditional ideological divisions, especially when long-term economic gains are at stake. The interaction between partisan councilors and high-quality agricultural land reveals how political support for development may diminish when agricultural viability is threatened, highlighting the role of local economic conditions in shaping land use decisions.

The role of interest groups aligns with PMF predictions, emphasizing the influence of different actors in land use governance. Production enterprises have emerged as primary proponents of land use changes, reflecting the growing demand for non-agricultural land in non-urbanized areas for industrial or commercial development [24,54,55,56,57]. Similarly, municipalities with active tourism sectors are more likely to experience significant land transformations, aligning with studies on the sustainability paradox documenting natural resources’ vulnerability to developmental pressures [7,62,63,64].

Notably, Poland’s educated residents demonstrate effects contrary to international findings [2,3,7,10,21,24]. Rather than advocating conservation, a higher educational attainment correlates with increased development pressure, paralleling the impact of affluent residents. This suggests the prioritization of modernization and infrastructure enhancement over long-term environmental considerations, highlighting Poland’s unique developmental context, where education and wealth act as drivers of land use changes rather than constraints.

This research reveals crucial insights regarding future research on land use changes. The financial influence of interest groups operationalized through property taxes from residents and legal entities shows that municipalities with higher property tax revenues are more likely to permit land use changes (the selection step). However, financial pressure alone does not fully determine the magnitude of these changes (the outcome step). This suggests that, while financial motivations are critical in triggering development, they are less decisive in shaping their overall extent. Conversely, agricultural enterprises and taxes act as protective forces, resisting development to safeguard agricultural viability. Direct measures of interest groups more effectively capture the factors driving the magnitude of land use changes than indirect financial measures.

The institutional context, particularly the presence of spatial development plans, significantly influences land transformation. Municipalities with comprehensive spatial plans experience more significant land use changes, suggesting that these tools primarily stimulate development in the non-urbanized areas of Poland. However, these plans must balance development objectives with social impact considerations, including community fragmentation and environmental degradation—a concern previously advocated by scholars [89]. Moreover, zoning decisions can influence whether land use changes occur; however, they do not have a strong effect on how extensive these changes will be once they are set in motion.

Despite the above findings, the drivers of land use change in Poland differ significantly from other contexts, particularly Western Europe and North America. While much research in these regions highlights conservation priorities and environmental sustainability as dominant drivers, the findings in Poland suggest that economic growth imperatives, modernization, and development pressures take precedence. This reflects Poland’s transitional economic and political landscape, where balancing rural preservation with industrial and infrastructural development remains a significant challenge. Future studies should further explore these dynamics by directly comparing Poland’s drivers with those of other regions to better contextualize and refine the applicability of the PMF framework.

Policy recommendations should address these dynamics by establishing formal platforms for stakeholder dialogue to resolve land use conflicts, particularly among farmers, businesses, and residents. Protecting high-quality agricultural land through financial tools and stricter regulations is essential for balancing development with rural preservation. Nationally standardized sustainability metrics could provide municipalities with practical frameworks to monitor the long-term impacts of land use changes. Tourism strategies should prioritize sustainable development to mitigate environmental pressures. Additionally, mechanisms such as mandatory sustainability assessments for significant land use changes can help align municipal decisions with broader sustainability goals.

Implementing these recommendations requires tailored approaches to accommodate different political and economic contexts. In economically developed municipalities, tax incentives or grants for sustainable projects can encourage stakeholder engagement. In contrast, under-resourced municipalities may need direct government support, such as subsidies for agricultural land protection or funding for training programs on sustainable development planning. Partnerships between local governments, NGOs, and private stakeholders can also help address resource gaps, while promoting a shared responsibility for effective land use governance.

Despite its contributions, this study has certain limitations. The reliance on CLC data, with its 5-hectare minimum mapping unit, may overlook smaller-scale land use transformations, potentially underestimating specific dynamics. Furthermore, this study focuses exclusively on Poland, which may limit the generalizability of its findings to regions with significantly different governance and institutional frameworks. Future research should address these limitations by incorporating higher-resolution spatial data to capture finer-scale land use changes and by extending the analysis to include cross-national comparisons. Exploring the interplay of political and economic factors in other transitional economies, as well as contrasting the findings with those from more developed or underdeveloped regions, would provide valuable comparative insights and help refine the applicability of the Political Market Framework.

6. Conclusions

This study extends the PMF to land use governance in non-urbanized areas, emphasizing the roles of municipal council composition, interest groups, and institutional factors in shaping land use changes in Poland. Farmers and educated councilors resist intensive development, while partisan councilors and interest groups such as production enterprises and the tourism sector drive significant transformations. Spatial development plans and agricultural land quality further influence these dynamics, highlighting the tension between development pressure and rural preservation.

These findings provide actionable insights for policy development, including fostering stakeholder dialogue, protecting high-quality agricultural land, and implementing sustainability metrics. Despite its reliance on CLC data and its focus on Poland, this study offers a foundation for exploring similar dynamics in other transitional economies. Future research should incorporate higher-resolution data and cross-national comparisons to enhance the applicability of the PMF in diverse contexts.