Acta geographica Slovenica, 60-2, 2020, 7–20 FARM HOLDINGS AND THE OWNER’S RESIDENCE LOCATION IN THE ASPECT OF DIRECT PAYMENTS FROM THE EU: A CASE STUDY IN NINE REGIONS IN POLAND Katarzyna Kocur-Bera Agricultural areas for cultivation in Poland. K A T A R Z Y N A K O C U R - B E R A Katarzyna Kocur-Bera, Farm holdings and the owner’s residence location in the aspect of direct payments from the EU … DOI: https://doi.org/10.3986/AGS.6836 UDC: 332.334.4:63(438) 338.246.027(438) COBISS: 1.01 Katarzyna Kocur-Bera 1 Farm holdings and the owner’s residence location in the aspect of direct payments from the EU: A case study in nine regions in Poland ABSTRACT: Instruments promoting rural development have been implemented by many countries. Area- based payments for farmers allocated under the Common Agricultural Policy constitute one of such instruments in the European Union. The support system for rural areas, including the size of the declared reference parcels, is monitored as part of the cross-compliance mechanism. Parcels with unfavorable land- use patterns are more difficult to farm. According to estimates, more than 30% of agricultural farms in Poland fall into this category. This study proposes a universal algorithm for controlling the information submitted by farmers in payment applications. More than 76,000 applications were analyzed, and farms with the defective spatial structure of land were randomly selected. The results show that most errors occur in the case of land parcels situated the farthest from a farm holding (declared in the application), but the analysis revealed no strong correlation in this respect. KEY WORDS: rural area, land spatial structure, area-based payments, cross-compliance monitoring, Common Agricultural Policy, Poland Kmetijska gospodarstva in lokacije bivališč njihovih lastnikov z vidika neposrednih plačil iz EU: študija primera za del Poljske POVZETEK: Številne države imajo instrumente za podporo razvoju podeželja. Ena takih oblik podpore v državah Evropske unije so subvencije za kmetijske površine kmetijskih pridelovalcev, ki se izvajajo v okviru Skupne kmetijske politike. Podporni sistem za podeželje, vključno s površinami, ki jih kmetijski pridelo- valec prijavi za subvencije, je predmet nadzora. Kadar imajo parcele pomanjkljivo prostorsko strukturo, je kmetijska dejavnost na njih otežena. Takih posestev kmetijskih pridelovalcev je na Poljskem več kot 30 %. Članek predstavlja univerzalen algoritem za izvajanje kontrole podatkov v prijavah kmetijskih pridelovalcev. Analizirali smo več kot 76.000 prijav kmetijskih pridelovalcev, med katerimi smo naključno izbrali skupino kmetijskih pridelovalcev z razpršeno prostorsko strukturo. Rezultati kažejo, da se večina napak pojavlja pri zemljiščih, ki so najbolj oddaljeni od kmetijskega gospodarstva (navedeni v vlogi), vendar analiza ni pokazala močne korelacije v zvezi s tem. KLJUČNE BESEDE: podeželje, prostorska struktura zemljišč, subvencije za kmetijske površine, skupna kmetijska politika, spremljanje skladnosti, Poljska This article was submitted for publication on July 2 nd , 2018. Uredništvo je prejelo prispevek 2. julija 2018. 8 1 University of Warmia and Mazury in Olsztyn, Olsztyn, Poland katarzyna.kocur@uwm.edu.pl (https://orcid.org/0000-0001-7056-5443) 1 Introduction The main goals of the European Union’s Common Agricultural Policy (CAP) are to provide financial sup- port to farmers, increase the competitiveness of agricultural markets in the Member States, increase productivity by promoting technical progress in agriculture, stabilize the agricultural market and guar- antee fair incomes and a fair standard of living for the farming community. The last goal is addressed by the area-based payment scheme (Latruffe and Davidova 2007; Zadravec and Zalik 2009; Zygmunt et al. 2015; Janković et al. 2018). Polish farmers became entitled to financial support under the CAP after Poland joined the EU in 2004. Area-based payments are one of the key support instruments for the Polish agricul- tural sector. The Land Parcel Identification System (LPIS) was originally devised for registering agricultural reference parcels eligible for annual payments under the CAP (Grandgirard and Zielinski 2008; Kocur- Bera and Piórkowska 2018). The LPIS is a computerized system that identifies agricultural (reference) parcels based on aerial and spatial ortho-imagery (Milenov and Kay 2006; EC 2009). It is the key control mech- anism to verify the eligibility of reference parcels for direct payments (Sagris and Devos 2008; Zimmermann Acta geographica Slovenica, 60-2, 2020 9 40%60% . – . 61%80% . . – 81%100% . . – Legend dolnoslaskie lubuskie pomorskie zachodniopomorskie opolskie lodzkie slaskie malopolskie podkarpackie lubelskie mazowieckie kujawsko–pomorskie wielkopolskie podlaskie warminsko–mazurskie swietokrzyskie Figure 1: The percentage share of farm holdings in a region, where the farmstead is located farther than 10km from the closest cadastral parcel (Kozłowski 2015). Katarzyna Kocur-Bera, Farm holdings and the owner’s residence location in the aspect of direct payments from the EU … et al. 2016). EU Member States have defined reference parcels in different ways as cadastral parcels (CP), agricultural parcels (AP), farmer’s blocks (FB) (one or several agricultural parcels cultivated by a single farmer) or physical blocks (PB) (one or several agricultural parcels cultivated by one or several farmers) (Inan and Cete 2007; Sagris and Devos 2008; Inan et al. 2010; Levavasseur et al. 2016; Kocur-Bera 2019). In Poland, a cadastral parcel was accepted as a reference parcel (Kocur-Bera 2019). A reference parcel may contain one or many agricultural parcels declared for aid by one or several farmers, and it is provided with a unique identifier (Inan et al. 2010). An agricultural parcel (AP) determines the subject of the aid appli- cation, its geographic location and the extent (area) of the agricultural activity. It represents the land for which payments can be claimed, and it is also a subject of administrative crosschecks and control proce- dures established in the Integrated Administration and Control System (IACS). Due to the dynamic nature of agricultural activities, an agricultural parcel can be unstable over time and space (Inan et al. 2010; Kocur- Bera and Piórkowska 2018). The LPIS is the main component of the IACS. Around 5% of all beneficiaries of CAP programs are inspected for cross-compliance each year. Agricultural producers are required to analyze maps and indi- cate the size of their reference parcels in the payment application, excluding non-eligible areas and objects such as buildings, yards, shrubs, roads, forests, and lakes. Cross-compliance inspections are not easy to perform, particularly in farms with defective structures. Farms with faulty land spatial structures in Europe result mainly from historical social and economic processes (van Dijk 2003; Vitikainen 2004; Latruffe and Piet 2014; Janus et al. 2016; Leń 2018; Prus and Szylar 2018). Many researchers focus primarily on land fragmentation, defined as defective spatial structure (Bański 1999; Leń and Noga 2010; Sobolewska-Mikulska 2012; Leń 2018; Noga et al. 2018). Fragmentation is con- sidered to be a comprehensive set of farm holding components, such as size of parcel/ownership, dispersion of parcel, shape of parcel, accessibility of parcel, number of parcel, type of ownership, distance from farm- stead (van Dijk 2004; Demetriou, Stillwell and See 2012; Latruffe and Piet 2014; Demetriou 2014; Noga and Król 2016; Leń 2018). The distance between cadastral parcels and the farmstead is one of the attrib- utes of a defective spatial structure(Woch 2001; Harasimowicz, Janus and Ostragowska 2009; Janus 2018). According to Kozłowski (2015), in Poland, in the Zachodniopomorskie and Podlaskie voivodeships, over 10% of farm holdings have their farmsteads farther than 10 km from even a single cadastral parcel; in the Wielkopolskie, Mazowieckie and Lubelskie voivodeships – it is 8–10% of farms; in other seven voivodeships (Podkarpackie, Łódzkie, Dolnośląskie, Lubuskie, Kujawsko-pomorskie, Pomorskie, W armińsko-mazurskie)– it is 6–8%; and 4–6% in the remaining regions (Figure 1). Agricultural producers are unable to maintain parcels located too far away in good condition due to labor shortages, poor fertilization, high fuel- and time-related costs (Neupane 2000; van Dijk 2003; Niroula and Thapa 2005). Increased production costs, the effect of excessively long distances between parcels, hin- der agricultural development and reduce the competitiveness of farmers (Niroula and Thapa 2005). Even a small loss of agricultural payments poses a massive problem because area-based subsidies constitute the main source of income for Polish farmers. The purpose of the research is to examine the impact of the allocation of agricultural producer’s residence in relation to agricultural plots that are part of a farm. The procedure of selecting farms for cross-compliance inspections was described in the first part of the study. The distance of the parcel from the farmstead (owner’s residence) influences the quality of agri- cultural operations and is responsible for the discrepancies between the declared and the measured size of reference parcels. The resulting errors affect the value of financial support granted under the CAP . The correlations between errors in the declared size of reference parcels and other error-promoting factors were analyzed in the second part of the study. 2 Methods Farms applying for area-based payments have to submit an application to the local branch of the nation- al payment agency (the Agency for the Reconstruction and Modernization of Agriculture in Poland). As of 2018, applications have been submitted online (through the eWniosekPlus platform). Written applica- tions are accepted only under extraordinary circumstances, for example, when the applicant does not have Internet access. The applications are checked for completeness, coherence and data content, and the appli- cant is placed on a list of inspection candidates. Administrative controls can be carried out in a simplified 10 or a reference mode. In the simplified mode, the declared information is checked for compliance with the Land Administration System (LAS), whereas in the reference mode, the declared information is compared against the data declared by other farmers (to prevent double funding for the same parcel). The number of inspection candidates is determined at the beginning of each year. Various methods are used to select inspection candidates (Table 1), and most of them involve aerial imaging and on-site visits. Tolerance lev- els are set for dealing with differences between the declared and the measured size of the parcel. Farms, where tolerance levels are exceeded, are fined. The case study was carried out by selecting a group of farm holdings where at least one parcel is locat- ed at least 5 km from the farmstead. It was assumed that the farmstead is a farmer’s place of residence and the address declared in the EU payment applications. The farms were selected based on the procedure for monitoring cross-compliance in farms receiving area-based payments and the algorithm presented in Table 1. The allocation of cadastral parcels with respect to the farmstead is of particular interest to the control author- ity (Polish Agency for Restructuring and Modernization of Agriculture). The calculated differences in the parcel area declared in the payment application and the on-site measured area were added up and the gen- eral difference for each cadastral parcel was obtained. Subsequently, the errors were added up for each Acta geographica Slovenica, 60-2, 2020 11 Step 1 Start– The beneficiary submits an application (A). Step 2 The submitted documents are checked for completeness and coherence (B). and Simplified mode (B1) Data are checked against the applications submitted in the previous years and are compared with the LAS database. Reference mode (B2) Data are checked against the information declared by other applicants to prevent double funding for the same parcel. Step 3 Number of inspection candidates is determined (C). Step 4 Inspection candidates are selected (D). or Risk analysis method (D1) Risk is analyzed based on several risk factors, including errors in previous years, fines imposed in previous years, farm size and the number of declared reference parcels. Random method (D2) Inspection candidates are selected randomly . Manual method (D3) Applicants are inspected based on a justified suspicion of non-compliance or third-party reports. Step 4 A list of inspection candidates is developed (E). and Location-based method (E1) Farms are grouped based on their location. Area-based method (E2) Farms are grouped based on their size. Step 5 An inspection schedule is developed for the selected farms (F). Step 6 The inspection method is selected (G). or On-site visits (G1) Farms are inspected by on-site visits. Aerial imaging (G2) Farms are inspected by orthophoto map analysis and by local visits. Step 7 The inspector and the inspection date are selected (H). Step 8 Relevant documentation is developed (I). Step 9 Inspection (J). or Measurements are performed using a digital tachymeter (J1). Measurements are performed using a GPS tools (J2). Measurements are performed using a tape measure (J3). Combined measurements are performed with the use of tape measure and a digital tachymeter (J4). Measurements are performed by orthophoto map vectorization (J5). Step 10 The inspection report is developed (K). Step 11 Tolerance levels for the applied measuring device are set (L). Step 12 The severity of the cross-compliance breach is determined in the inspected farm (M). Step 13 The penalty rate is set (N). Step 14 The penalty is communicated to the beneficiary (O). Step 15 End– The penalty is imposed on the beneficiary (P). Table 1: The procedure for monitoring cross-compliance. Katarzyna Kocur-Bera, Farm holdings and the owner’s residence location in the aspect of direct payments from the EU … landowner studied. The assumptions made are consistent with the procedure for controlling the landown- ers. Sanctions and/or penalties are imposed when exceeding specific limits with respect to all cadastral parcels declared for payments from the EU. The distance of every analyzed cadastral parcel was measured from the geometric center of the agricultural parcel and the farmstead (Euclidean distance, in GIS) and it was expressed in kilometers. Initially, approximately 76,000 beneficiaries of CAP support schemes who reside in Poland were evaluated. Ultimately, 216 parcels owned by 18 beneficiaries were selected for analy- sis. The declared parcels were situated in nine Polish regions: Pomerania (Pomorze), West Pomerania (Zachodniopomorskie), Wielkopolska (Wielkopolskie), Lower Silesia (Dolnośląskie), Swietokrzyskie V oivodeship (Świętokrzyskie), Lodz V oivodeship (Łódkie), Subcarpathia (Podkarpackie), Lubusz V oivodeship (Lubuskie), and Warmia and Mazury (Warmińsko-Mazurskie) (Figure 2). The difference between the agricultural parcel area (a) declared in the application and the area measured on-site (m) was calculated using the equation: where: D ap – difference between the declared (A d ) and the measured (A m ) agricultural parcel area (error in declared area); A d – agricultural parcel area declared in the application; A m – agricultural parcel area measured on-site (with GPS tools, tachymeter). The correlation analysis revealed significant relationships between the observed differences for every cadastral parcel (Diff) and other land-use attributes, such as distribution (Dist), parcel area (Ar), number of parcels in a farm (Disp), number of regions where farm holdings are situated (RD) (Table 2). The observed correlations were interpreted on a six-point scale proposed by J. Guilford (Kocur-Bera 2016a). Table 2: Description of the land-use attributes. No. Attribute Symbol Description Measure 1 Difference Diff Difference between the cadastral parcel area declared in the application and the area measured on-site. % 2 Distance Dist Distance between the cadastral parcel and the landowner’s address declared in an EU application. km 3 Area (measured) Ar Area of a cadastral parcel. ha 4 Dispersion Disp A number of cadastral parcel belonging to the landowner. number 5 Regional Dispersion RD A number of regions where the cadastral parcels belonging to the landowner are situated. number 3 Results Farms were selected for analysis using the algorithm for selecting inspection candidates (Table 1). The analy- sis relied on the above algorithm to focus solely on farms with an allocation of the owner’s residence. Differences for each agricultural parcel were measured and the total sum of errors was calculated in rela- tion to each cadastral parcel. Differences in 216 plots in 18 farms were examined. In 58 cases the difference was different from zero. The results are presented in figure 3. In 83% of the parcels where inconsistencies were determined, the declared area was greater than the area used for agricultural production. In 39% of the analyzed farms, the average difference between the declared area and the measured area did not exceed 3%; in 44% of the analyzed farms, the difference was determined at 4–20%; and in 17% of the analyzed farms, the difference exceeded 20%. The average dif- ferences between the area declared in the application and the area measured on-site for evaluated farms are presented in Table 3. 12 Figure 2: The Polish regions included in the analysis (marked with blue color).p Figure 3: The differences between the declared area and the area measured on-site, for 216 cadastral parcels [%]. p p. 14 D ap = (A d –A m ) A d · 100 % Acta geographica Slovenica, 60-2, 2020 13 Finland Sweden Norway Russia Turkey Georgia Armenia Azerbaijan Belarus Lithuania Latvia Estonia Ukraine Moldova Romania Serbia Albania Greece Montenegro Hungary Slovakia Czech Republic Austria Croatia Slovenia Italy Monaco Andorra Portugal Spain Gibraltar (UK) France Jersey (UK) Ireland Iceland Faeroe Islands (Denmark) Denmark Germany Belgium Netherlands Luxembourg United Kingdom Switzerland Malta Bulgaria Katarzyna Kocur-Bera, Farm holdings and the owner’s residence location in the aspect of direct payments from the EU … 14 –150 –100 –50 0 50 100 1 5 9 13 17 21 25 29 33 37 41 45 49 53 57 61 65 69 73 77 81 85 89 93 97 101 105 109 113 117 121 125 129 133 137 141 145 149 153 157 161 165 169 173 177 181 185 189 193 197 201 205 209 213 Table 3: The average differences between the parcel area declared in the application and the area measured on-site in the evaluated farms. Number of The average differences Number of cadastral The average distance between Number of regions where the landowners in the area of all plots for one parcels on the farm the cadastral parcel and the cadastral parcel belonging landowner (in one farm)[%] landowner’s address declared to the landowner are situated in application [km] 1 –1.06 10 9.00 2 2 25.97 12 100.00 1 3 1.30 6 10.80 2 4 1.48 4 100.00 1 5 56.31 4 3.75 2 6 1.11 12 48.75 2 7 1.97 11 9.54 2 8 –8.20 9 33.33 2 9 22.00 6 100.00 1 10 9.66 15 100.00 1 11 16.25 18 77.22 3 12 6.27 16 89.37 3 13 7.44 48 85.62 4 14 –4.22 8 6.87 2 15 9.90 9 2.22 2 16 1.17 9 4.44 2 17 0.00 7 1.43 2 18 6.58 12 2.08 2 As the distance between the evaluated parcels and the farmstead was measured it was found that 39% of the analyzed agricultural parcels were situated at a distance of up to 20 km from the farmstead (Zone 1), 13% of the parcels were situated at a distance of 21–50 km (Zone 2), and 48% of the parcels were at a dis- tance of more than 50 km (Zone 3). The percentage differences in the area declared and measured on-site concerned 58 reference parcels, with an average difference of 30.75%. The smallest number of differences was observed in Zone 2 (21–50 km), it concerned 15 reference parcels and the average difference was 20.37%. The distribution of individual differences in zones is shown in Figure 4. Acta geographica Slovenica, 60-2, 2020 15 Figure 4: Diagram of the distribution of the percentage differences occurring in individual distance zones. Katarzyna Kocur-Bera, Farm holdings and the owner’s residence location in the aspect of direct payments from the EU … Table 4: Correlation coefficients (Diff – difference between the cadastral parcel area declared in the application and the area measured on-site differences; Dist – Distance between the cadastral parcel and the landowner’ s address declared in an EU application; Ar – Area of a cadastral parcel.; Disp – A number of cadastral parcel belonging to the landowner; RD – A number of regions where the cadastral parcels belonging to the landowner are situated. Variable Diff Dist Ar Disp RD Diff 1.0000 0.1514 0.0030 0.0923 0.0868 Dist 1.0000 0.0868 0.4622 0.2311 Ar 1.0000 –0.0195 0.0210 Disp 1.0000 0.8621 RD 1.0000 The calculated Pearson’s coefficients (Table 4) were analyzed to reveal that the difference (Diff) between the declared parcel area and the measured on-site was determined by the distance (Dist) between the ref- erence parcel and the landowner’ s address declared in an EU application. The correlation was weak (according to Guilford’s scale weak correlation is 0.1