The Use of Agricultural Services in European Union Regions Differing in Selected Agricultural Characteristics

Abstract

Agricultural services, understood as the rental of machinery and equipment with appropriate labor, are one of the three types of production services in agriculture distinguished by European Union legislation. The aim of this paper is to identify clusters of regions in the European Union that differ in the level of use of agricultural services on farms and in selected characteristics related to production potential, labor input, and type of agricultural production. For this purpose, Ward’s method, from the group of hierarchical agglomerative cluster analysis methods, was used. Based on data on farms using agricultural services in 124 regions of the European Union, six clusters were formed. The study showed that agricultural services substitute for labor inputs in intensive agricultural production conditions, but in a situation with good technical equipment, farms may more often choose to employ hired workers. Such substitution does not occur in regions that are moderately and less well-equipped with machinery and equipment, because hired labor cannot completely replace the scarcity of machinery. The level of use of agricultural services is also related to the profile of the production carried out and the area of agricultural land, followed by the resources of land, capital, and labor. The level of economic development and historical background are also important.

1. Introduction

According to the theory of three sectors, with economic development, the role of agriculture systematically decreases, the role of industry increases, stabilizes and then decreases, and the role of services steadily increases [1]. This is accompanied by a change in the allocation of factors of production, which manifests in their successive fundamental absorption by the three distinguished spheres of activity [2,3,4,5,6,7,8,9]. A specific dimension of the impact of these regularities can also be observed within each of the sectors, which complement their potential through the use of production services in the manufacturing process in industry and agriculture, but also within the service sector. Indeed, business entities use various services that are no longer only in side activities, but commonly integrate services into the processes of core activities [1].

There are many types of services, and new ones are progressively emerging. One of the traditionally distinguished types are manufacturing services, which support the rationalization of manufacturing processes. Using the potential and knowledge of service providers makes it possible to reduce the costs associated with building and maintaining the producers’ own capacity and to achieve better production results and product quality. Services are also a vehicle for knowledge and progress in the organization of production and the technologies used. Currently, European agriculture is facing new challenges, related to the implementation of the European Green Deal (GD), the opening of the market for agricultural products from Ukraine and the EU–Mercosur Free Trade Area Agreement signed in 2019, which is still unratified [10]. In the face of changing formal and market operating conditions, farms will be forced to undertake costly adjustment measures. The scale of changes needed on farms will depend on the production profile, technologies used, fixed assets, and labor resources. The ability to adapt a farm will be determined mainly by its economic strength, production potential and production structure, and know-how. In this process, the use of services will be helpful or necessary, so it is important to study the coexistence of selected characteristics of farms and their use of services. Such knowledge provides a good starting point for further analysis, regarding opportunities, chances, and threats to the operation of agriculture in the future.

Council Regulation (EC) No. 138/2004 [11] distinguishes three groups of production services in agriculture. The first is agricultural services, i.e., the rental of machinery and equipment with the corresponding labor. This type of service can be roughly associated with the concept of mechanization services and is used primarily in crop production [1]. The second type is veterinary services, i.e., medicines, which are invoiced independently of the veterinarian’s fees and veterinary costs, which include medicines given directly by the veterinary service and included along with his fee. The third group is Financial Intermediation Services Indirectly Measured (FISIM), for which fees are charged indirectly, resulting from differences in interest rates on loans and deposits, rather than directly (e.g., as a commission). The paper focuses on agricultural services that relate directly (mainly) to crop production.

The aim of the paper is to identify clusters of regions in the European Union that differ in the level of use of agricultural services on farms and in selected characteristics related to production potential, labor input, and type of agricultural production. For this purpose, Ward’s method, from the group of hierarchical agglomerative cluster analysis methods, was used. The object of the study is the users of agricultural services in the European Union. The scope of the study included 124 regions comprising 27 European Union countries. Cyprus and Malta were omitted from the study due to the marginal importance of agriculture in their economies. Secondary data for 2021 from the Farm Accountancy Data Network (FADN) database [12] and the literature were used. The study adopted the FADN classification of regions appropriate to the data source used. Of the three groups of production services, defined in accordance with European Union legislation, agricultural services, i.e., the rental of machinery and equipment with corresponding labor, were studied. The paper presents a brief consideration of the nature of services and their definition. Then, the importance of production services in optimizing costs and inputs is pointed out, in view of the changes in the conditions of agricultural activity related to the Green Deal, the planned agreement with Mercosur and food imports from Ukraine. The empirical part of the paper presents clusters of European Union regions distinguished by selected indicators characterizing agriculture and the use of services, and discusses the reasons for the observed differences.

2. Background

There are many definitions of services, formulated at different stages of economic and social development. As early as 1965, E. Lipinski [13] wrote that there is no concept more full of ambiguities than services, and in 1973 T. Kotarbinski [14] argued that the formulation of a single, comprehensive definition, which would include all types of activities classified as services, is a foregone conclusion. The ambiguity of the concept of service was also pointed out by Jastrzębowski [15], Bywalec [16], and Janoś-Kresło [17], among others. Despite this, there have been many attempts to define services in an enumerative, negative or constructive way [1]. Enumerative definitions enumerate what activities count as services. Negative definitions specify what a service is not. For example, Lange [18] states that services are activities that do not directly produce objects. The authors of potential-oriented constructive definitions emphasize that a service provider maintains a certain service capacity through a combination of internal factors of production, and that new value, or service, is created only when the service provider comes into contact with the service recipient [8]. None of the formulated definitions have stood the test of time and changing economic conditions. New services emerged that did not fit into the categories provided by these definitions, and the very notion of service recipient and service provider also evolved. Taking into account the fact that under the conditions of globalization, the development of information technology, artificial intelligence (AI) technology, and the attribution of service provision to entities other than humans (e.g., the environment, economic sectors, Internet bots), today, it is not possible to establish a closed catalog of services. Therefore, a broad definition can be considered the most up-to-date one, according to which services are any activity as a result of which a specific utility is provided to another entity, not resulting from an employment relationship with the service recipient (when service providers are humans) or ownership/possession of the service provider by the service recipient (when service providers are artificial intelligence entities, enterprises, institutions, etc.) [1]. This definition is also applicable to agriculture, where a number of new AI-related services are emerging, provided automatically by AI without direct human involvement. Such services can be related to supply processes and distribution, but can also participate in technological processes, even in field work.

Another definition, according to which services are economic activities offered by one party to another, should also be considered valid and up-to-date: time-based services produce desired results for recipients, objects or other assets for which buyers are responsible. In the process of providing services, money, time, and effort are exchanged. Service buyers gain benefits from access to goods, labor, professional skills, facilities, networks, and systems. In doing so, however, there is no transfer of ownership of tangible goods. This definition is somewhat narrower in scope, as it excludes services provided outside of business activities, but it is sufficient in terms of the scope of the study conducted [19].

Among the many services present on the market, production services, which participate in manufacturing processes, are important. In the case of the agricultural sector, services respond to demands concerning the technological, economic, and environmental dimensions of agricultural activities. Their role and importance have changed over time, ranging from the need to close the technological gap, to increasing the rationality of farming, to modernizing farm operations and bringing them in line with European Union standards and the tenets of sustainable agriculture [20]. In the case of the agricultural sector, these are services that support agricultural production at all its stages: from the creation of conditions for production activities, supporting production from the financial and technological side, to the processing and distribution of produced food [1]. Production services, and among them agricultural services, enable the optimization of the use of land, labor, and capital, and thus support productivity and facilitate the adaptation of farms to changing operating conditions. The increase in the level of consumption of agricultural services in farm operations is also beneficial because of their income potential and optimization of the use of agricultural machinery. Farms with surplus equipment can provide agricultural services to those that do not have enough of it. In this way, the capabilities of the, often very efficient and modern, machinery are better utilized. The payback time for investments in these machines is shortened, and services become an additional source of income for machine owners. At the same time, service recipients avoid economically unjustified investments in technical equipment.

The better organization of agricultural services is one of the challenges of building a modern and computerized agriculture [1]. Services are also an alternative to investment at the individual farm level. As with previous technological advances, not all farmers will invest in new skills, and where technologies save labor, farms will expand [21]. Thus, as a cheaper and more convenient alternative to investment, in the face of changing agricultural conditions, services can play a significant role in optimizing production costs and facilitating the adaptation of the technical side of the production process to increasingly stringent technological and environmental requirements. In addition to economic arguments, the current political situation also supports this thesis. Investments in the development of one’s own machinery may not only be economically unprofitable, but they may also be incompatible with administratively imposed ways of conducting agricultural production and the anticipated direction of changes in this regard, for example, with the implementation of the assumptions of the Green Deal. The Green Deal assumes, among other things, the following: a 50% reduction in plant protection products, a 20% reduction in the use of mineral fertilizers, the designation of at least 10% of arable land for pro-environmental purposes (elements of the agricultural landscape), and the designation of 25% of agricultural land for organic farming. Achieving these goals without harming production volumes, if at all possible, requires a significant increase in labor and capital. The Green Deal also aims to expand protected land and sea areas in Europe, restoring degraded ecosystems by reducing the use and harmfulness of pesticides [22]. The priority of the GD is food security. The GD also aims to ensure (within the planet’s capacity) a sufficient supply of affordable and wholesome food, promote sustainable food production, encourage more sustainable food consumption, and support healthy eating [23]. However, according to the United States Department of Agriculture (USDA) [24], implementing GD solutions will reduce food production in the European Union by 12.0% and increase prices by an average of 17.0%. As a result, food exports will decrease by 20% and imports will increase by 2%. Similar conclusions are also provided by a report by a consortium of authors: the Institute of Rural and Agricultural Development of the Polish Academy of Sciences (IRWiR PAN)—the consortium leader, the Institute of Soil Science and Plant Cultivation—National Research Institute (IUNG-PIB), and the University of Life Sciences in Poznań (UPP) [25]. The report indicates the unfavorable effects of implementing the GD on agricultural production and farms in Poland. According to the authors, the full implementation of the GD will translate into a decrease in farmers’ income by at least 11% and cause food prices to rise. At the same time, the value of crop production will fall by 13% and crop acreage will decrease by 6%. In view of such predictions, and in view of the probable ratification of the EU–Mercosur Free Trade Area Agreement [10], signed in 2019 but “suspended” due to protests by European farmers, which opens the European agricultural market to cheaper food products from South American countries, the necessary investments may not be bearable for small- and medium-sized farms. This leaves the alternatives of giving up production activities and selling or leasing the land to more powerful entities, or using services if they are available and their cost is acceptable. Here, the key becomes the caveat of the GD goals cited above the limitation “within the planet’s capacity”, which in practice could mean agreeing to a drastic reduction in the profitability and scale of agricultural production in the EU in favor of food imports from abroad. The Mercosur agreement has not yet been ratified, but it is likely that similar effects, albeit on a much smaller scale, can be seen in the example of imports of cheap food from Ukraine [26]. In the situation of a drastically reduced profitability of production, agricultural policy, including the amount and structure of financial transfers to agriculture, will also be crucial. Proposals to date in this regard point to the pursuit of environmental goals at the expense of production, even at the cost of long-term or irreversible loss of production capacity [27,28,29]. Therefore, in the perspective of limited income from production and competition from cheaper products from outside the EU, it is crucial for the survival of farms to rationalize operating costs and maintain compliance with the European Union’s requirements. Production services are a helpful “tool” in this regard, which, rationally used, can facilitate the operation of farms in increasingly difficult economic conditions and in the face of increasingly stringent regulations.

3. Materials and Methods

This study constructs a typology of regions of European Union countries, using the Ward method, from the group of hierarchical agglomerative cluster analysis methods. Ward’s method allows objects to be combined into successive clusters based on the value of the similarity function. The more similar the objects are, the earlier they are combined with each other (minimization of the sum of squares of deviations of any two clusters that can be formed at each stage is performed) [30,31]. The Euclidean distance was used in forming clusters, defined as follows:

distance (x,y) = {∑_i (x_i − y_i)²}^½

(1)

A variance analysis approach was employed.

Clusters are arranged hierarchically so that lower-order clusters are included in higher-order clusters, according to the hierarchy of similarity that exists between objects [32,33]. Considering that the primary factor influencing clustering is the presence of mutually uncorrelated characteristics [34], the calculated indicators were evaluated through correlation coefficients, preceded by the standardization of the variables.

Calculations were carried out using Statistica 13.1 software. The basis for the implementation of Ward’s method in cluster analysis was the Lance–Williams algorithm [35,36], which is a general formula for updating the distance matrix between clusters in hierarchical clustering. With this algorithm, the distances between clusters can be effectively updated after each clustering step. The Lance–Williams algorithm provides a recursive formula for updating the distances between clusters during hierarchical agglomerative clustering. It is widely used in implementing clustering methods, including Ward’s method, as it efficiently handles the dynamic recalculation of distances when clusters are merged. Ward’s method aims to minimize the increase in the total within-cluster variance at each step of the clustering process. The Lance–Williams algorithm serves as the computational framework for this process by updating distances between the newly formed cluster and all remaining clusters [37].

The classification includes indicators reflecting the situation in agriculture within the framework of the use of services by farms. A number of attempts were made to create typologies on the basis of different sets of features. The set of features presented in the publication turned out to be the only one that simultaneously met the substantive and statistical criteria for selection. After eliminating variables that are highly correlated with each other, the typology was constructed using the following active indicators that characterize farms by their use of services:

• Utilized agricultural area (UAA) (ha);
• Own (unpaid) labor input (Annual Work Unit—AWU);
• The cost of purchasing services per hectare of UAA (euros);
• The cost of purchasing services per Annual Work Unit (AWU) (euros);
• The share of the cost of purchased services in indirect consumption (%);
• The share of crop production in the structure of agricultural production (%).

To determine the number of clusters, a dendrogram analysis was used, which presents the hierarchy of clustering units and the levels of clustering at which the units first merged [38,39,40,41]. The dendrogram is the result of applying a particular hierarchical clustering strategy and a measure of similarity or distance [21,38,42,43,44,45,46,47,48,49,50,51,52]. Determining the optimal number of clusters is among the most challenging decisions. Increasing the number of clusters enhances intra-cluster homogeneity, as reflected in the reduced variation of objects in the feature space. However, this also reduces the number of objects within each class and makes differences between cluster averages less significant. Conversely, reducing the number of clusters may, at a certain point, raise concerns about their homogeneity, potentially compromising the accurate identification of the underlying phenomena. Therefore, the determination of the number of classes and object memberships should aim for optimal intra-cluster homogeneity and inter-cluster heterogeneity [53]. Classification methods lack built-in procedures to determine the optimal number of classes, which is instead guided by the researcher’s knowledge and the outcomes of formal procedures [54].

Ward’s method aims to create relatively small clusters and is considered highly effective. Depending on the study’s assumptions, particularly the chosen taxonomic distance between objects based on the proposed set of characteristics, we can identify larger or smaller clusters, and consequently, a greater or lesser number of them. This approach allows us to consider not only statistical but also substantive criteria, resulting in a more accurate division into clusters [55].

The study used data collected and processed by the Farm Accountancy Data Network (FADN) system in 2021 for the European Union [12].

4. Research Results and Discussion

The services market in the European Union is diverse on both the supply and demand sides. A large number of players of different sizes operate on it and compete with each other. This applies to all the types of services mentioned [56]. On the other hand, farms make demands for different types of services, provided at different times and at a price they can accept. The analyzed regions differ in the values of the indicators included in the study. Western European countries are characterized by a more favorable land use structure and higher productivity than Central and Eastern European countries. This is due to their different histories. These divisions are also evident within individual countries, as their current borders are the result of a long evolution. They were shaped as a result of wars and international agreements. At the regional level, different farming patterns were adopted, often linked to socio-behavioral patterns, such as profit orientation or attachment to land as an intergenerational legacy. Among the contemporary events shaping the situation in European agriculture, it is worth firstly mentioning the divisions into political blocs after World War II, the political transition after 1989, the collapse of the Soviet Union, German reunification, and the enlargement of the European Union [57]. Structural changes in the agriculture of the former socialist bloc countries have shaped in them the structure of land use, in which large-scale farms operate alongside a large group of small- and medium-sized farms. Today, one of the most important factors shaping the structure of agriculture in the regions is the Common Agricultural Policy (CAP) [58].

History, different farming patterns, and varying levels of economic development mean that the level and extent of agricultural use varies from one region of the European Union to another, just as the agricultural characteristics of these regions differ. On the basis of the analysis, six clusters of regions were distinguished by selected agricultural characteristics and the level of use of agricultural services. The values of all considered (K = 6) characteristics for the studied (N = 124) EU regions were compiled into a K × N (6 × 124) dimensional data matrix. This became the basis for the construction of a typology of EU regions according to the use of services by farms. As a result of the agglomeration (Figure 1), it was found that it was optimal to divide the collection of EU regions into six clusters that differed from each other in the values of the characteristics adopted for the study (Figure 2).

The average cluster values of active indicators are given in

Table 1. Average cluster values of indicators characterizing agriculture and the use of agricultural services in 2021.

Indicator	Clusters						Total
	I	II	III	IV	V	VI
Active indicators (used in the classification process)
Utilized agricultural area (UAA) (ha)	25.7	56.1	28.1	453.9	98.3	90.5	53.2
Own (unpaid) labor (AWU)	0.9	1.1	1.3	0.9	1.3	1.3	1.2
Cost of purchasing services per hectare of UAA (euros)	43.9	68.8	36.2	80.7	301.9	198.2	90.8
Cost of purchasing services per AWU (euros)	834.5	2403.3	633.3	5723.6	12,034.2	8612.2	2063.3
Share of cost of purchased services in intermediate consumption (%)	4.9	5.8	3.8	5.9	17.4	9.6	6.1
Share of crop production in the structure of agricultural production (%)	77.9	34.5	67.3	62.1	81.4	42.9	66.0
Inactive indicators (other indicators that characterize the clusters)
Cost of purchasing services per farm (euros)	1289.0	3503.0	1310.5	36,625.0	22,772.0	16,647.0	3503
Service intensity of agricultural production (euros/1000 euros of production value)	30.6	132.1	32.1	76.6	128.5	163.0	66.1
Area of leased agricultural land (ha)	12.8	24.8	14.5	343.6	84.6	58.3	30.1
Hired labor input (AWU)	0.4	0.2	0.4	5.5	0.6	0.5	0.5
Machinery, equipment and means of transport (euros)	18,161.0	12,050.0	28,190.5	401,333.0	89,330.5	112,851.0	39,231
Land productivity (euro/ha)	2329.4	1904.3	2012.6	2059.3	2914.4	2796.8	2338
Labor productivity (euros/AWU)	40,901.4	57,885.4	41,034.4	144,396.8	145,143.0	132,116.3	66,090
Share of livestock production in the structure of agricultural production (%)	22.1	65.5	32.7	37.9	18.6	57.1	34.0
Gross value added (euro)	44,555.0	45,948.5	41,913.0	450,361.0	154,714.5	115,538.0	73,857.0

Source: own calculations based on [12].

, and the values of the measure of average differences, used to indicate the characteristics of each cluster, are shown in

Table 2. Values of the measure of differences in average of indicators characterizing agriculture and the use of agricultural services in 2021 *.

Indicator	Clusters
	I	II	III	IV	V	VI
Active indicators (used in the classification process)
Utilized agricultural area (UAA) (ha)	−0.8	0.1	−0.7	11.6	1.3	1.1
Own (unpaid) labor (AWU)	−1.6	−0.7	0.5	−1.5	0.7	0.8
Cost of purchasing services per hectare of UAA (euros)	−0.7	−0.3	−0.8	−0.1	3.0	1.5
Cost of purchasing services per AWU (euros)	−0.3	0.1	−0.4	1.0	2.7	1.8
Share of cost of purchased services in intermediate consumption (%)	−0.4	−0.1	−0.8	−0.1	4.1	1.3
Share of crop production in the structure of agricultural production (%)	0.8	−2.1	0.1	−0.3	1.1	−1.6
Inactive indicators (other indicators that characterize the clusters)
Cost of purchasing services per farm (euros)	−0.3	0.0	−0.3	3.9	2.2	1.5
Service intensity of agricultural production (euros/1000 euros of production value)	−0.7	1.2	−0.6	0.2	1.2	1.8
Area of leased agricultural land (ha)	−0.6	−0.2	−0.6	11.7	2.0	1.1
Hired labor input (AWU)	−0.1	−1.0	−0.3	16.9	0.7	0.2
Machinery, equipment, and means of transport (euros)	−0.5	−0.6	−0.2	7.8	1.1	1.6
Land productivity (euro/ha)	0.0	−0.5	−0.4	−0.3	0.7	0.6
Labor productivity (euros/AWU)	−0.6	−0.2	−0.6	1.9	1.9	1.6
Share of livestock production in the structure of agricultural production (%)	−0.8	2.1	−0.1	0.3	−1.1	1.6
Gross value added (euro)	−0.7	−0.6	−0.7	8.7	1.9	1.0

* The gray cells in the table show the indicators most characteristic in the clusters (assumed to be −1 and lower and 1 and higher). Source: own calculations based on [12].

.

Table 3. Characteristics of clusters of EU regions distinguished by selected indicators characterizing agriculture and the use of agricultural services in 2021.

Cluster	Characteristics	Number of Regions
I	With the lowest own labor inputs, small utilized agricultural area (UAA), low cost of purchasing services per hectare of UAA (euros), and high share of crop production in the agricultural production structure	30
II	With the lowest share of crop production in the agricultural production structure and low own (unpaid) labor (AWU)	10
III	With the lowest cost of purchased services per hectare of UAA and per AWU and the lowest share of the cost of purchased services in intermediate consumption	32
IV	With the largest average area of agricultural land, very low own labor input, highest hired labor input, and an average cost of purchased services per AWU	7
V	With a relatively high area of agricultural land, the highest cost of purchased services per hectare of UR, the highest cost of purchased services per AWU, the highest share of purchased services costs in intermediate consumption, and the highest share of crop production in the agricultural production structure	10
VI	With a high cost of purchased services per hectare of UAA, a very high cost of purchased services per AWU a high share of purchased services costs in intermediate consumption, and the lowest share of crop production in the agricultural production structure	35

Source: own calculations based on [12].

shows the characteristics of clusters of EU regions, distinguished by selected indicators characterizing agriculture and the use of agricultural services in 2021. As a result of the typology, six clusters were formed, the distribution of which is shown in Figure 3. A list of regions included in each cluster is provided in the Appendix A (

Table A1. A list of regions included in clusters.

Region Code	Name of the Region
Cluster I
HUN 0764	Észak-Magyarország
HUN 0768	Dunántúl
HUN 0767	Alföld
ESP 0560	Comunidad Valenciana
POR 0640	Alentejo e Algarve
ESP 0575	Andalucia
ROU 0840	Nord-Est
ELL 0470	Thessalia
ELL 0450	Makedonia-Thraki
ELL 0460	Ipiros-Peloponissos-Nissi Ioniou
ITA 0241	Trentino
ITA 0250	Liguria
ESP 0565	Murcia
ITA 0320	Sicilia
ITA 0311	Puglia
ITA 0303	Calabria
ESP 0525	La Rioja
ROU 0842	Sud-Muntenia
ROU 0841	Sud-Est
ITA 0312	Basilicata
ITA 0281	Marche
ROU 0844	Vest
ITA 0292	Abruzzo
ITA 0301	Molise
ITA 0291	Lazio
ITA 0302	Campania
ELL 0480	Sterea Ellas-Nissi Egaeou-Kriti
ESP 0530	Aragón
BGR 0836	Yugoiztochen
BGR 0833	Severoiztochen
Cluster II
POR 0650	Açores e Madeira
SVE 0730	Län i norra Sverige
ESP 0545	Castilla y León
ITA 0330	Sardegna
ESP 0555	Castilla-La Mancha
ESP 0520	Navarra
ESP 0500	Galicia
ESP 0505	Asturias
ESP 0510	Cantabria
IRE 0380	Ireland
Cluster III
ITA 0221	Valle d’Aosta
POL 0790	Wielkopolska and Slask
POL 0785	Pomorze and Mazury
ESP 0570	Extremadura
ESP 0550	Madrid
POL 0795	Mazowsze and Podlasie
ESP 0515	Pais Vasco
ROU 0847	Bucuresti-Ilfov
ROU 0843	Sud-Vest-Oltenia
HRV 0861	Jadranska Hrvatska (Adriatic Croatia)
BGR 0834	Yugozapaden
ITA 0244	Friuli-Venezia
ITA 0243	Veneto
ITA 0260	Emilia-Romagna
ESP 0535	Cataluña
POR 0630	Ribatejo e Oeste
DEU 0070	Rheinland-Pfalz
ITA 0270	Toscana
BGR 0832	Severen tsentralen
BGR 0831	Severozapaden
ROU 0846	Centru
ROU 0845	Nord-Vest
SVN 0820	Slovenia
POR 0615	Norte e Centro
LTU 0775	Lithuania
LVA 0770	Latvia
ESP 0540	Islas Baleares
ITA 0242	Alto-Adige
HRV 0862	Kontinentalna Hrvatska (Continental Croatia)
ITA 0222	Piemonte
POL 0800	Malopolska and Pogórze
BGR 0835	Yuzhen tsentralen
Cluster IV
DEU 0113	Mecklenburg-Vorpommern
SVK 0810	Slovakia
DEU 0112	Brandenburg
DEU 0116	Thüringen
DEU 0114	Sachsen
DEU 0115	Sachsen-Anhalt
CZE0745	Czech Republic
Cluster V
FRA 0203	Provence-Alpes-Côte dAzur
FRA 0201	Languedoc-Roussillon
FRA 0182	Aquitaine
FRA 0152	Alsace
FRA 0121	Île de France
FRA 0131	Champagne-Ardenne
FRA 0164	Poitou-Charentes
FRA 0133	Haute-Normandie
FRA 0134	Centre
FRA 0132	Picardie
ITA 0282	Umbria
FRA 0204	Corse
FRA 0183	Midi-Pyrénées
FRA 0136	Bourgogne
EST 0755	Estonia
SVE 0710	Slättbyggdslän
SUO 0670	Etelä-Suomi
SUO 0690	Pohjanmaa
DAN 0370	Denmark
FRA 0184	Limousin
FRA 0193	Auvergne
FRA 0151	Lorraine
FRA 0153	Franche-Comté
SVE 0720	Skogs- och mellanbygdslän
DEU 0060	Hessen
DEU 0100	Saarland
OST 0660	Austria
ITA 0230	Lombardia
DEU 0080	Baden-Württemberg
DEU 0090	Bayern
LUX 0350	Luxembourg
DEU 0015	Schleswig-Holstein/Hamburg
DEU 0030	Niedersachsen
SUO 0680	Sisä-Suomi
SUO 0700	Pohjois-Suomi
DEU 0050	Nordrhein-Westfalen
FRA 0163	Bretagne
FRA 0162	Pays de la Loire
FRA 0135	Basse-Normandie
FRA 0192	Rhônes-Alpes
FRA 0141	Nord-Pas-de-Calais
BEL 0343	Wallonie
NED 0360	The Netherlands
ESP 0580	Canarias
BEL 0341	Vlaanderen

Source: own study.

).

The first cluster consisted of 30 regions from the southern part of Europe, including 2 Bulgarian, 4 Greek, 5 Spanish, 11 Italian, 3 Hungarian, 1 Portuguese, and 4 Romanian regions (Figure 3). This cluster was mainly characterized by the lowest own labor input (AWU), at 0.9 AWU/1 ha of UR, and the smallest average of UAA, at 25.7 ha, and the largest share of crop production in the agricultural production structure. The cost of purchasing services per hectare of UAA was 43.9 euros/ha, slightly higher than in Cluster III and lower than in all other clusters. Compared to the average for all the regions studied (90.8 euros/ha), it was more than twice as low. The same was true for the cost of purchased services per AWU, whose value of 834.5 euros/AWU was slightly higher than in Cluster III and significantly lower than in all other clusters, and almost two and a half times lower than the EU average. Consequently, the value of the share of the cost of purchased services in intermediate consumption (4.9%) was also the smallest in this cluster.

The second cluster includes 10 regions from the southern and northern parts of Europe: 6 Spanish, 1 Italian, 1 Portuguese, 1 Swedish, and Ireland (Figure 3). This is the cluster with the lowest share of crop production in the agricultural production structure (34.5%). The average farm size here was 56.1 hectares, slightly higher than the average value in all EU regions studied (53.2 hectares).

The third cluster included 32 regions from mainly southern and central-eastern Europe (Figure 3): 4 Bulgarian, 5 Spanish, 2 Croatian, 7 Italian, 4 Polish, 2 Portuguese, 4 Romanian, and 1 German, as well as Lithuania, Latvia, and Slovenia (Figure 3). This cluster is mainly characterized by the lowest cost of purchased services per hectare of UAA (36.2 euros) and per AWU (633.3 euros), and the lowest share of the cost of purchased services in intermediate consumption (3.8%). The own labor input is 1.3 AWU, slightly higher than the EU average (1.2 AWU). The average farm area in this cluster is 28.1 hectares and is 1.9 times lower than the EU average. The share of crop production in the agricultural production structure (67.3%) is close to the EU average (66.0%).

In the fourth, least numerous (seven regions) cluster, five German regions, the Czech Republic and Slovakia can be mentioned (Figure 3). Here, were found agricultural enterprises of the largest average area of agricultural land (453.9 hectares), with very low own labor input (0.9 AWU), and an average value of service purchase cost per AWU (80.7 euros). Low employment results in a high cost of purchased services per hectare of UAA (5723.6 euros), but the share of these costs in intermediate consumption is 5.9%, which is 0.2 percentage points lower than the EU average. Similarly, the share of crop production in agricultural output is slightly lower in this cluster than the average for the entire EU, at 62.1%. This cluster is characterized by high labor productivity (144,396.8 euros/1 AWU), very good equipment in technical means of production (average 401,333.0 euros per farm). Perhaps somewhat surprisingly, despite this, land productivity is 2059.3 euros/ha, lower than the EU average by as much as 278.7 euros. In part, the explanation for this can be attributed to the location of the regions included in the cluster in the former German Democratic Republic and Czechoslovakia, which in the past were dominated by large-scale, state-owned farms. Despite the passage of time, this may have a negative impact on land productivity. Large-scale farms in eastern Germany are distinguished by high mechanization, but at the same time relatively low production intensity. Achieving high land productivity is also not helped by not very good soil quality [59,60]. The former Czechoslovakia is dominated by large farms, but they often lack flexibility and the ability to adapt quickly to changing market conditions, which lowers their productivity. A factor negatively affecting productivity in the area is the location of parts of the UAA in mountainous areas [61]. Some studies also indicate a negative relationship between larger farm size and production performance per hectare [62]. Large farms in these countries suffer from being trapped in the technology treadmill, which means improvements in the technologies used, increases in the scale of operations, and reductions in production costs, but at the same time, does not guarantee profitability cf. [63]. Cluster IV regions had the highest inputs of hired labor, at 5.5 AWU per hectare This dependence of large farms on hired labor, which is many times higher than in other cluster, may also reduce land productivity, counteracting the economies of scale achieved cf. [64].

The fifth cluster includes 10 French regions (Figure 3). This is the cluster with a relatively high average area of agricultural land (98.3 hectares), the highest cost of purchased services per hectare of UAA (301.9 euros/ha), the highest cost of purchased services per AWU (12,034.2 euros), the highest share of purchased services costs in intermediate consumption (17.4%), and the highest share of crop production in the agricultural production structure (81.4%). Regions included in this cluster also have the highest land productivity (2914.4 euros/ha) and the highest labor productivity (145,143.0 euros/AWU).

The sixth largest cluster of 35 regions included the regions of western and northern Europe: 2 Belgian, 7 German, 1 Spanish, 12 French, 2 Italian, 4 Finnish, and 2 Swedish, as well as Austria, Denmark, Estonia, Luxembourg, and the Netherlands (Figure 3). Included here are regions with a very high cost of purchased services per hectare of UAA (198.2 euros), a high cost of purchased services per AWU (8612.2 euros), a high share of purchased services costs in intermediate consumption (9.6%), and the lowest share of crop production in the agricultural production structure (42.9%).

The regions of the European Union are a heterogeneous collective in terms of the values of the characteristics studied. The level of use of agricultural services, as measured by the cost of purchasing services per hectare of UAA, 1 AWU, and the share of the cost of purchasing services in intermediate consumption, was highest in Cluster V, consisting of French regions with the highest share of crop production in the agricultural production structure and high average farm area. Cluster VI, with a similar average farm area and almost twice the share of crop production, was also characterized by a high but much lower level of use of agricultural services, which could be assumed to be due to a lower share of crop production. Farms in cluster IV with the largest average UAA, best equipped with machinery and technical equipment, purchased relatively fewer agricultural services, but compared to clusters I, II, and III, the level of use of services in cluster IV was high. In the regions included in it, in addition to their own technical means, the purchase of services was partially substituted by greater hired labor (5.5 AWU per hectare of UAA, while in the other clusters it was 0.2 to 0.6 AWU). These observations suggest that the level of use of agricultural services is primarily related to the profile of production carried out by farms and the area of farmland.

The cost of purchasing services per AWU, which varies by cluster (Table 1 and Table 2), shows that, especially under conditions of intensive agricultural production (clusters IV, V, VI), agricultural services replace own labor inputs. However, under conditions of good technical equipment, farms may choose to employ hired labor more often (cluster IV). The data in Table 1 and Table 2, however, do not allow us to conclude that regions moderately and less well equipped with machinery and equipment replace services with hired labor. This may be the case for certain branches of production that require manual labor, but for most field crops additional employment will not compensate for the scarcity of machinery and equipment. The alternative of accessing agricultural services may also be more cost-effective for these farms, compared to purchasing and operating the necessary machinery and equipment, which would often not be fully utilized, due to the fact that the scale of production is too small in relation to their productivity. Thus, in clusters IV, V and VI, well-equipped with machinery and equipment, the services complement the farms’ own potential, while in the less well-equipped clusters of I and II they compensate for deficiencies in this equipment. In cluster III, consisting mostly of regions of the former socialist bloc, the low level of use of agricultural services may be due, in addition to the characteristics taken into account in the typology, to the impact of an unquantifiable, decades-old tradition in farming methods and a lower level of wealth than in Western Europe [54]. In practical terms, this means using their own, often obsolete and depreciated, but still operable, machinery and equipment. Despite the relative cheapness of agricultural services, they are, in many cases, not competitive with owned machinery and equipment operated by farm owners and family members (especially if surplus labor is available in parallel). This, in turn, indicates the dependence of the level of service use on economic development, determining wealth and the distribution of employment between agriculture and other sectors of the economy. Among the post-socialist countries, only two, namely Poland and Romania, resisted collectivization during the period of dependence on the Soviet Union, preserving a dispersed structure of private ownership and individual land use. Small individual farms operated in a situation of machinery shortages, resulting from the system-conditioned promotion of state and cooperative land ownership. The solution for them turned out to be a specific form of production service provision, consisting of paid or unpaid neighborly assistance (in exchange for monetary remuneration or, more often, for an equivalent non-monetary benefit, such as crops, milk, and assistance in field work). To some extent, this form of service provision remains relevant even today, as manifested by the existence of a gray area of agricultural services and labor, unrecorded and invisible in public statistics. This is especially true for smaller farms. This may cause the actual level of service use in Poland and Romania to be slightly higher than indicated in the study, but the scale of this phenomenon is decreasing over time and remains practically impossible to determine precisely [1]. In addition, some services are not included in the survey presented here because small, non-commodity farms in these countries are not covered by FADN [12].

5. Conclusions

As a result of the study, six clusters of EU regions were formed, distinguished by selected characteristics of agriculture and use of agricultural services in 2021. The analysis conducted allows us to formulate the following observations and conclusions:

• Agricultural services replace personal labor inputs under conditions of intensive agricultural production. However, when farms are well-equipped technically, they may opt to hire additional labor. This substitution is weaker in moderately and poorly equipped regions, as hired labor cannot fully replace machine labor, which these farms often lack.
• In clusters well-equipped with machinery and equipment, services complement the farms’ own potential, and in less well-equipped clusters, services compensate for deficiencies in this equipment. The results obtained allow us to assume that the level of use of agricultural services is related primarily to the profile of production carried out by farms and the area of agricultural land, followed by the resources of land, capital, and labor.
• Since the analyzed regions differ not only in the characteristics considered in the typology, other factors, often unquantifiable or difficult to quantify, such as different farming traditions in Western European and post-socialist countries, should also be taken into account when formulating conclusions. Also important are prejudices resulting from their different histories, such as the degree of dispersion of land structure and land ownership, and the attitude of farm managers to the work they do (managerial or emotional).
• Due to farming traditions in regions located in former socialist countries with a dispersed land structure and land ownership, and with relatively high employment in agriculture, the use of services may not be very competitive with the use of one’s own, often obsolete and depreciated but still operable, machinery and equipment, despite the lower price of agricultural services than in other regions. This, in turn, indicates the dependence of the level of use of agricultural services on the level of economic development.
• The relatively low productivity of land in Cluster IV regions located in the former German Democratic Republic and Czechoslovakia, which in the past were dominated by large-scale, state-owned farms, is due to a combination of natural, technological, and economic factors. Among these are the inferior soil quality in the eastern German states compared to the western part of Germany and the dependence of the relatively large farms in Cluster IV on hired labor input. Striving to increase productivity by improving technical equipment is also costly, but does not always guarantee high land productivity.
• In Poland and Romania, on the other hand, the use of agricultural services may be underestimated, due to traditions of unregistered, paid or unpaid neighborly assistance provided in exchange for monetary remuneration or an equivalent non-monetary benefit, such as crops or milk, or the provision of undeclared labor on the farm of an informal service provider—usually a colleague or neighbor. However, the scale of this phenomenon is declining over time and its size in practice is impossible to determine precisely.