Assessment of Concept between Rural Development Challenges and Local Food Systems: A Combination between Multi-Criteria Decision Analysis and Econometric Modelling Approach

Abstract

This study investigates the influence of social, economic, and environmental impacts on the promotion and marketing systems of local food products from mountain areas in Slovenia. These impacts were assessed using an econometric modelling approach. Two econometric models were developed (one for production and one for marketing). The case studies of local food products were selected from Slovenian mountain regions, most of which were from less-favored areas (LFAs). A majority of the selected food commodities were of high quality, with or without protected designations. Data collection was carried out via interviews. Due to production limitation on mountain areas in Slovenia and other constraints in LFAs, the socio-economic and environmental impacts on success of production and marketing systems need to be clarified. These relations present a potential impact on the wider socio-economic development in the region. The empirical results, obtained using an econometric modelling approach, clearly show the importance of encouraging the socio-economic and environmental impacts in ensuring the marketing and production potentials of local food products. The result express good relationships, and cooperation between the actors in the food supply chains contributing to a successful marketing system and production system of local food products (small, average, large) is dependent on the available local labour in mountain rural areas.

1. Introduction

In recent years, the production and processing of local food products has had a significant impact on the development of rural areas. Successful local food marketing processes provide an opportunity for small agri-food businesses in short food supply chains [1]. In studies [2,3], the connection between the characteristics of regional food and rural development was recognized, and the authors developed a technique for evaluating the success of the promotion of the protected designation of origin and the protected geographical indication in a pilot study in Italy. The local food products were closely connected with short food supply chains and presented a key element in understanding new patterns of rural development [4,5]. Searching the development opportunities of rural areas can be found through the assessment processes and with different methodological approaches. One methodological approach (multi-criteria decision analysis (MCDA) technology) for assessing the successes of production and marketing system of local food product was explained in a study as well as for other investigation purposes [6,7,8,9,10,11]. Moreover, econometric methodology was used by the authors of [12] for impact analysis of a young farmers’ support program on the development process of the Slovenian dairy sector. In this survey, the authors analyzed only political aspects, and for this reason, a lack of studies was identified related to methodological approaches that can be widely expressed for analyzing broader aspects of the impact of production and marketing systems on social and economic indicators and vice versa. This presents a challenge in terms of combining the relevant methodology and upgrading the topic. One of the measurement econometric models based on a methodological approach was presented in study [13] for assessment of the sustainability of equestrian tourism in Slovenia.

Local food products in Slovenia need good support from promotion systems of local food products, as they are closely connected to qualities related to geographical origin and they represent one of the qualities that defines value-based food chains (also known food chains with added values). Sufficient production and good promotion of local food products are two main conditions for the existence of value-based food chains (VBFC) and further development of rural LFA areas in Slovenia. Due to a lack of relevant research studies supported by quantitative models for assessing the economic impacts of VBFC on primary actors, some authors [14] decided to develop the econometric model to evaluate the economic viability of value-based food chains related to dairy farmers. This analysis was focused on disclosing the economic returns yielded by the econometric modelling approach—a translog profit function. Another study [15] investigated how involvement in value-based food chains can improve the economic situation of dairy farms in mountain regions. The results gave empirical outcomes of positive/negative consequences of involvement the farmers in VBFC. Study [16] explicitly outlined the characteristics of successful local food production and marketing processes and identified the shortcomings in each process. Although the economic impacts on primary actors are well researched, studies connected to further unexplored possibilities of agro-food chains are still missing. Because of this, there is importance to empirically approaching the linkage of production and marketing in the case of Slovenian local food products. The underlying reasons for successful and unsuccessful local food production and marketing systems have not yet been elucidated. Many authors have explored the effects of successful production and marketing on developing rural areas, but the goals of the production process, the actors involved, and the marketing approaches adopted have not been investigated.

The aim of this paper is to show how different activities (monetary and material investments) can influence the successful producing and selling of local food. The added value of this research is in the combination of two different methods (multi-criteria decision analysis and the econometric modeling approach) for measuring the impacts on the final assessment of local food products. Multi-criteria analysis is very useful in practice, but it includes only qualitative values. Thus, it is impossible to measure and identify empirically the size of the impact of the parameters on the variables studied. From this perspective, the logical next step was recognized in examining how the different socio-economic and environmental impacts can improve the marketing and production systems of local food products. The research question of this paper is quite simple but very interesting for local communities, local decision makers, and policy decision makers to identify the importance of connections between the impacts of different socio-economic and environmental factors on the production and marketing processes of local food. The research hypothesis for this study is as follows: The investigated connections of social, economic, and environmental improvements and successful development process of local food products on local areas can raise the concept of rural development challenges for improving the local economy and the quality of rural people.

2. Materials and Methods

2.1. Data Collection

Slovenian mountains and hilly regions encompass more than 70% of the country’s total area. The research focus is primarily on Alpine and pre-Alpine areas in Slovenia, due to the existence of VBFC in mountain areas. The term “local mountain food products” was used to avoid any dual identity of the products. In mountain areas, mountain products are considered local products and sold as such, but in non-mountain areas, they are mostly considered to be special products, reflecting the image of the mountain area. By using the term “local mountain food products”, we attempted to emphasize the local origin of the products, as well as local sales/local marketing channels.

Data were collected from 20 face-to-face interviews carried out by well-qualified examiners, and the interviewers were leaders/managers of local action groups and managers (coordinators) responsible for promotion projects of local food products. Before structuring the questionnaire, the literature review was performed by experts involved in the study section of the model development. To obtain answers that were as unbiased as possible, the questionnaire included mostly closed questions (expresses as descriptive estimates), administered in the north and northwest parts of Slovenia. Collected information were used as data for ten independent variables (x) (

Table 1. Description of dependent and independent data of econometric models.

Variable	Name of Variable	Description of Variable	Type of the Variable	Variable in Model	The Purpose of the Variable Expression
Ya	Marketing system	Variable expresses the results of the success of marketing system of local food products from multi-criteria decision analyzing. Taken from study [6]	Dependent variable	Belongs to model 1	Not relevant
Yb	Production system	Variable expresses the results of the success of production system of local food products from multi-criteria decision analyzing. Taken from study [6]	Dependent variable	Belongs to model 2	Not relevant
X1	Tourism	Variable expresses the invigoration of the tourism on local areas and tourist farms with supporting and developing applied projects of local community	Independent variable	Belongs to model 1	Socio-economic impact
X2	Employment in rural areas	Variable expresses the invigoration of the activities for decreasing the unemployment rates in rural areas	Independent variable	Belongs to model 2	Economic impact
X3	Sustainable development of rural areas	Variable expresses the invigoration of the development activities on local areas with supporting and developing applied projects of local community	Independent variable	Belongs to model 1	Socio-economic and environmental impacts
X4	Promotion of rural areas	Variable expresses the invigoration of the promotion activities on local areas with supporting and developing applied projects of local community	Independent variable	Belongs to model 1	Economic impact
X5	Cross-sectoral cooperation	Variable expresses the invigoration of the cross-sectoral cooperation activities between actors in local food supply chains	Independent variable	Belongs to model 1	Social impact
X6	Traditional knowledge	Variable expresses the invigoration of the transmission of traditional knowledge between old and young generation	Independent variable	Belongs to model 1	Social impact
X7	Higher profits	Variable expresses the invigoration of the development of local food chains with the aim to improve profits for local farmers	Independent variable	Belongs to model 2	Economic impact
X8	Access to new markets	Variable expresses the activities of local community for development of new market positions	Independent variable	Belongs to model 1	Economic impact
X9	Investments in farm modernization	Variable expresses the invigoration of the investments in farm modernization	Independent variable	Belongs to model 2	Economic impact
X10	Overall environmental protection	Variable expresses the invigoration of the environmental protection activities of local areas	Independent variable	Belongs to model 2	Environmental impact

). The independent variables taken into consideration were tourism, employment in rural areas, sustainable development of rural areas, promotion of rural areas, cross-sectoral cooperation, traditional knowledge, higher profits, access to new markets, investments in farm modernization, and environmental protection. Some of the independent variables were used in “marketing model” (also known as model 1) and some of them were used in “production model” (also known as model 2). The choice depended on experts’ opinions, which were based on preliminary literature review. More complex decisions, also based on relevant literature review, were made in the case of the “marketing model”. For example, the investigation questions between the “marketing model” and development of local tourism were based on literature [17,18]. The investigation questions between the “marketing model” and cross-sectoral cooperation were based on literature review [19]. Furthermore, the investigation questions between successful marketing approach and creating new marketing channels were based on the literature review [20]. Creating the “production model” was easier, because it was created by well-known indicators (such as employment rate, profits, and investments) which were directly connected to the production system. For example, when compiling the qualitative data of independent variables, different qualitative parameters, such as number of local projects, investment funds, local activities, interested of local community, etc., were collected from interviewers. Answers relating to independent variables in the questionnaire were ranked from 1 to 5, with 1 being “not important” and 5 being “very important.” The data conversion process was done by the researchers working on rural development issues also considering the experiences and knowledge from the field work. The questions pertained to the possible impacts (based on the managers’ opinions) of 10 different independent variables further expressing socio-economic and environmental impacts that affect production and marketing systems of local food products.

An important step for the aim of this paper and its novelty is defining the purpose of independent variable expression connected to socio-economic and/or environmental impacts. In study [21] authors described one of the socio-economic and environmental impact assessment processes. They stated that this method may differ from other methods in the essence of scientific and technical prospects, but the main aspect of socio-economic impact assessment is that all identified impacts are expressed in economic terms. This statement was taken into account in this survey, and the authors use their own methodological approach and data collection technique. In Table 1, the data in last column present the purpose of variable expression of socio-economic and/or environmental impacts.

The length of the interview is approximately 3 h per each actor from all sections of the local food product industry (production system, processing system, marketing system, distribution, local action groups (LAG), local policy decision makers, consumers, tourism sector, non-government organizations, etc.).

Data collection, processing, and analysis were carried out between 2018 and 2020 and are based on 10 Slovenian traditional local food products from Alpine and Pre-Alpine areas. Mountain areas in Slovenia are known as the Julian Alps and the Kamnik-Savinja Alps located in the north and northwest part of Slovenia. The northeast part of Slovenia and north part of middle Slovenia are known as the Pre-Alpine area. The selection of the products was based on the mountain areas in Slovenia, choosing some traditional food products with protected designations (e.g., Upper Savinja stomach sausage, Bovški cheese, and Tolminc cheese) and some without. The way in which the food product is marketed also dictated the selection of each food product, rather than the product’s ingredients or quality characteristics [6]. The local food products selected are promoted by seven local action groups in the mountain areas in Slovenia and are recognized at a local level. We collected 130 quantitative data for dependent and independent variables.

After getting results for the independent variables, the next process was converting the already known qualitative data of dependent variables (y) from the DEX model into quantitative data. Final assessments of local food products (taken from [6]—Figure 1 and Figure 2) present the dependent variables. Assessments were converted from descriptive estimations into quantitative (also numerical) data, which were used in econometric modeling in this survey. For example, the assessments of “marketing system” in the DEX model can get three different ratings (“successful, partial successful and not successful”); they were converted into quantitative numerical values used for econometric modelling purposes (values from 1 to 3) (

Table 2. Qualitative DEX results of the marketing system of local food products and their conversation to quantitative values for econometric models [6].

No. of Food Product from [6]	Name of Food Product	Qualitative Values of the Marketing System from DEX Model	Converted Quantitative Values for “Marketing Model”	Qualitative Values of the Production System from DEX Model	Converted Quantitative Values for “Production Model”
I	Upper Savinja stomach sausage (dried meat)	Successful	3	Large	3
II	Dried fruit	Partially successful	2	Average	2
III	Solčava sirnek (dairy product from fresh milk)	Partially successful	2	Small	1
IV	Rye bread	Partially successful	2	Average	2
V	Cider	Not successful	1	Small	1
VI	Bovški cheese (sheep cheese)	Not successful	1	Average	2
VII	Tolminc (cow cheese)	Successful	3	Large	3
VIII	Pohorje pot	Successful	3	Small	1
IX	Tarragon cake	Partially successful	2	Small	1
X	Jetrnica (sausage)	Not successful	1	Average	2

).

2.2. Methodological Approach

The methodological approach in this study was based on econometric modeling using part of the input data from study [6].

The description of the methodological approach used in this study was given in three steps. The first step involved conversion processes of input data already explained in sub-section “data collection”, which were included in the SPSS statistical package. The second step of the methodological approach was to specify the econometric models. Two different models have been defined. First (1) describes the influences of socio-economic and environmental impacts on the marketing systems of local food products. Second (2) describes the influences of socio-economic and environmental impacts on production systems of local food products. The first model was named “MARKETING MODEL” and the second model was named “PRODUCTION MODEL”. Equations below present the structure of the models. Descriptions of the dependent (y) and independent (x) variables were presented in Table 1.

Y_a = c + αx₃ + βx₄ + γx₅ + δx₆ + εx₉ + u_t

(1)

Y_b = c + αx₁ + βx₂ + γx₇ + δx₈ + ρx₁₀ + u_t

(2)

where: Y and X—description from Table 1, c—constant, and ${\mathrm{u}}_{\mathrm{t}}$—random residual.

Four different mathematical types of the models have been tested (linear—“LIN”, logarithmic—“LOG”, linear-logarithmic—“LIN-LOG”, and logarithmic-linear—“LOG-LIN”). It means that eight different types of models were tested in total. By calculating the elasticities from the most suitable type(s), the different socio-economic and environmental impacts were identified that can improve marketing and production systems of local food products.

The third step of the methodological approach was testing the models. Econometric results have been evaluated given economic, statistical, and econometric criteria (the same criteria were used and described in previous papers [14,15]). Finally, the elasticity values for most suitable socio-economic and environmental impacts were calculated and the empirical results discussed. At the end of the estimation process some model(s) is/are chosen as the best predictor(s). In the case of MARKETING MODEL, this is the LIN-LOG variant, whereas in the case of PRODUCTION MODEL, it is the LOG variant.

We choose the elasticity calculation approach to reach our study aim. Elasticity is a measure of a variable’s sensitivity to a change in another variable; most commonly this sensitivity is the change in quantity demanded relative to changes in other factors, such as price. It is commonly used in basic economics and the business world. However, as we commonly see that multi-criteria decision analysis can be used as an assessment tool [6,10,11], we decided to use elasticity calculation approach to investigate the socio-economic issue in this survey. This is possible because of the quantitative data that was converted from qualitative data. The same methodology approach was used by authors in study [12] in another rural development context. Elasticity in LOG models can be read from coefficients, whereas in LIN-LOG models it should be calculated with equation—b*(1/y′), where “b” is coefficient of variable and “y′” is mean value of dependent variable.

3. Results and Discussion

After model evaluation processes and estimation tests were performed, the autocorrelation was identified in six of the eight models. Durbin–Watson (DW) tests were lower than optimum values (1.8–2.2). Only two models were recognized as suitable predictors from which the elasticities values for socio-economic and environmental impact can be computed. In

Table 3. Results of the models which were taken into consideration.

Name of the Model	Type of the Model	Durbin–Watson Test (DW)	Significance Value of the Model (Sig.)	Model Formulation	Significance Value of the Most Suitable Socio-Economic and Environmental Impacts (Sig.)	Elasticity Values
Marketing model	LIN-LOG	2.040	0.011	Ya = −0.013 − 0.480x3 + 0.591x4 + 1.410x5 + 0.869x6 − 0.540x9 + ut	X5 = 0.002	0.705
Production model	LOG	1.848	0.228	Yb = 1.387 + 1.177x1 − 2.072x2 + 0.278x7 + 0.475x8 − 0.759x10 + ut	X2 = 0.038	−2.072

, the characteristics of the best evaluated models and elasticity values for most suitable socio-economic and environmental impacts are presented:

According to the calculated results in Table 3, the interpretation of the results is a very important aspect. For MARKETING MODEL we choose LIN-LOG type of the model as the best evaluated model. There is no positive or negative autocorrelation in the model and the DW test is 2.040. The DW test is normally used to detect the autocorrelation in the residuals from a statistical model or regression analysis. Because some qualitative (and lately quantitative) data from interviewees express the numerical data from profits, prices, etc., the DW test should be taken into consideration.

The significance value of the MARKETING MODEL is below (* p < 0.05) as well as the Sig. value of independent value X₅. The result express good relationships and cooperation between the actors in the food supply chains contributing to a successful marketing system. Hypothetically, results of the elasticity show that increasing the process of the cross-sectoral cooperation activities between actors in local food supply chains by 1% can improve the final assessment of local food products marketing system by 0.7%. In practice it means that additional activities for cooperation growth between actors through the food chain (i.e., new organizational scheme of food chain, signing new long-term contracts between actors, improving the up-stream and down-stream communication between actors, equivalent treatment of all actors in food chain, and rights and duties are evenly distributed among all actors) can increase the success of marketing systems of local food products.

Activities for successful marketing depends on all the actors involved having an organized approach to marketing activities. This statement also confirms our experience from fieldwork in the case of a sausage called Jetrnica. In this instance, there were more than 500 actors involved in the production process, and there were plenty of wholesale marketing opportunities. However, the actors were not interested in cooperating to improve the marketing system. In this case, on the one hand, the consequence was expressed as a lack of cooperation, resulting in weak food product promotion, and, on the other hand, all quantities were sold in a face-to-face manner. In contrast, two or more actors cooperated in promoting and marketing Tolminc cheese and Upper Savinja stomach sausage. However, the number of actors involved was much lower than in the Jetrnica case. In the case of Tolminc cheese, there were three different actors in the marketing system: farmers, local public institutions, and individuals involved in promoting alternative ways of marketing. Farmers and local public institutions were responsible for promoting Upper Savinja stomach sausage. We can conclude that even if production capacities of Tolminc cheese and Upper Savinja stomach sausage is higher compared to Jetrnica case, the cooperation of a few actors in marketing system can handle and sell all of the food products.

According to PRODUCTION MODEL, there are no optimal options in the interpretation of the results. Only one of the models could be potentially taken into consideration—the LOG model. The significance value is 0.228, whereas the significance values of X₂ are lower (* p < 0.05). The promotion of activities aimed at decreasing unemployment rates in rural areas for 1% can subsequently lead to a decline in a rate of success of production system of local food products for −2.072%. It is our understanding that the results interpretation can get in the way where the size of the production system of local food products (small, average, large) is dependent on the available local labour in mountain rural areas. This is also supported by our experiences from the field work and some literature review. Some authors [22,23] estimated the importance of local food to be much broader than mere consumer perception; local food also has a significant environmental and socioeconomic role and is closely linked to wider rural development issues, such as sustainable local businesses, retaining money in the local economy, the presence of successors on farms, and maintaining and preserving agricultural land in mountain areas.

Outcomes from the survey identify some discussion areas, such as farms located in (unfavourable) mountain areas are characterized by their ability to connect between each other, that is, by their multiple functions going far beyond the sole food production function. Achieving sustainability in agri-food chains requires coordinated interactions between chain actors. It is especially important in countries such as Slovenia where the average size of the farm is stated in 6.9 ha. Thus, limited the production capacities and some added values should be introduced to stabilize the economic situations of farms. This can be done with producing high quality food products, such as presented in this paper, but their long-term success is tied to different socio-economic and environmental impacts that can affect the marketing and production processes. Although the farmers can handle the selling of local food products with the individual performance on the market, it is dangerous for further sustainable growth, even if healthy growth of local food chains can be achieved when they value place before volume.

4. Conclusions

The model of the authors of [6] was upgraded by econometric modeling to identify the different socio-economic and environmental impacts that may affect the production and marketing processes of local food from mountain areas. Evaluated case study is relevant for Slovenian rural areas where many similar cases have the properties of Alpine and Pre-Alpine farming characteristics. From this aspect it could be relevant in other neighboring countries such as Austria and Italy. The relevant rural priorities could also be extended in mountainous Gorski Kotar in Croatia. The upgraded methodology showed that an econometric modeling approach is important when assessing food supply chain processes, as it provides clear empirical results rather than qualitative results. Decision-makers can use the results to define priorities (via a hierarchical structure) for production and marketing, regardless of the origin of the local food (food products from mountains, non-mountain regions, or less-favored areas). The main finding here is the identification of the potential influences of socio-economic and environmental impacts on the production and marketing of local food products. Previous models identified only the main bottlenecks along the food supply chain. The take-home message of this paper is identification of the socio-economic and environmental influences on production and marketing systems that are important for the success in the production and marketing systems of local food.

Local policy decision-makers and others can use the obtained results to encourage local food producers and to prioritize investment in improving socio-economic and environmental factors that affect local food production and marketing. We envisage that empirical evidence obtained in this study can have some “trigger effects” on the policy decision makers who must eventually realize how important it is to support development of local food chains, especially in countries such as Slovenia, where there is a considerable lack of motivation to keep small farming in mountain and hilly regions. Targeting the stated research hypothesis in the introduction section we can confirm it, but we were expecting more methodological relevant connections and relations between observed impacts and local food systems. This preliminary research could have greater validation in relation to the possible model of sustainable development of these areas if the basket of local food products were greater and the assessment of the socio-economic and environmental impacts of the territory under consideration could be better assessed in the whole of the local economy.

Concerning the number of environmental impacts, these should be numerous and presents the limit of this and challenge for further study.