Introduction

The global concern for the sustainability of the agri-food sector, encompassing agriculture, livestock, and the agri-food industry, is growing due to its economic, social, and environmental implications (UN 2022). With the projected world population of 9.7 billion by 2050, addressing environmental degradation while increasing productivity is imperative (FAO 2017).

Initially, environmental concerns were overshadowed by the pursuit of increased agricultural productivity, leading to a 220.7% rise in cereal productivity between 1961 and 2021 through genetic improvements and production intensification (FAO 2017, 2019a, b; FAOSTAT 2022). However, this has resulted in significant negative impacts on ecosystems, including biodiversity loss, soil erosion, plastic contamination, and greenhouse gas emissions (Alcon and Zabala 2022; Borsato et al. 2018; Froldi et al. 2022; IPCC 2019; Omerkhil et al. 2020). The cumulative environmental impact, coupled with other anthropogenic activities, contributes to a substantial global environmental footprint (Keyber and Lenzen 2021; Rodríguez et al. 2022; Taylor et al. 2022).

While industrial food production drives socioeconomic development (Ayompe et al. 2021; Valera-Martínez et al. 2017), it concurrently generates social imbalances, land rights issues, and economic pressures on farmers, exacerbated by geopolitical instability (Abram et al. 2017; Castillo-Díaz et al. 2023a; Dislich et al. 2017; European Commission 2022; Moyo 2011).

Sustainable development as an indispensable goal

The commitment to sustainability, as outlined in the 2030 Agenda, emphasizes the interconnected dimensions of economic, social, and environmental sustainability (UN 2015, 2017). The 17 Sustainable Development Goals (SDGs) underpin the economic, social, and environmental aspects of food production, exemplified in targets such as SDG 2.4, aiming to "ensure the sustainability of food production systems" (UN 2015, 2017).

Amid global efforts to transition toward a sustainable economy, the European Union plays a leading role, proposing a circular economy model by 2050 to decouple economic growth from resource consumption (Cifuentes-Faura 2022; European Commission 2019, 2020). This paradigm shift, reflected in the European Green Pact, guides sectoral action plans and highlights the significance of national initiatives, like Spain's Plan for Adaptation to Climate Change 2021–2030 (MINECO 2021).

Need to quantify sustainability and methodological challenges

Quantifying the sustainability of food production through a composite indicator is crucial for strategic planning and managerial decision-making, particularly in agriculture, livestock farming, and the agri-food industry (Castillo-Díaz et al. 2023a; Suresh et al. 2022). This information, with relevance to food sovereignty and security, extends its utility to sectors like finance, identified as a key player in the green transition (EBA 2019). Therefore, it is necessary to develop a methodology to comprehensively assess the sustainability of farms and agri-food industries in order to obtain a sustainability score for agri-food activities.

Despite various attempts to quantify sustainability, challenges persist, including a lack of consensus, reliance on questionnaire-based methodologies, and multi-sectorial approaches (Bajardí-Azcárate et al. 2022; Suresh et al. 2022). This is due to the idiosyncrasies of each subsector (i.e., sub-activity within agriculture, livestock or agri-food industry) of the agri-food sector. The three basic rules of sustainability require the existence of a common ground, i.e., food production must involve the ecological, economic, and socio-cultural stability of the territories where it takes place (Bao et al. 2022; Rigby and Cáceres 2001; Suresh et al. 2022; Velten et al. 2015). Addressing these challenges requires a system of indicators offering a numerical value for comprehensive sustainability assessment. Recent developments in methodology have made it possible to quantify the social, economic, and environmental sustainability of agriculture, livestock, and the agri-food industry at a sectoral level (Castillo-Díaz et al. 2023a). However, specialized methodologies to assess sustainability specifically for agribusinesses and agri-food industries remain absent. The indicators used to evaluate sustainability in this sector and within these enterprises should differ due to their unique characteristics. While methods have been established to measure the sustainability of agricultural operations, livestock, and agri-food industries, they are not designed to collectively quantify the sustainability of agricultural businesses, livestock, and agri-food industries; rather, they assess each separately. Moreover, many of these methodologies were originally developed for other economic sectors and later adapted for agri-food enterprises using qualitative surveys (MAPA 2016). Thus, there is a pressing need to bridge this gap in the literature and develop a specific methodology to quantify the sustainability of these enterprises (Table 1).

Table 1 Models for determining the sustainability of the agri-food sector identified by the Ministry of Agriculture, Fisheries and Food of the Government of Spain (MAPA)
Full size table

The identified challenges underscore the necessity to develop a methodology enabling the quantification of economic, social, and environmental sustainability across farms and agri-food industries. This methodology should align with the value-chain image of the agri-food sector. Recognizing the specificities of subsectoral sustainability, common indicators across activities must be identified and evaluated (Malkina-Pykh 2000; MAPA 2016; Riley 2001; Suresh et al. 2022; Woodhouse et al. 2000). This model, beyond enhancing sustainability, holds significance for sectors like finance, requiring a protocol for quantifying the sustainability of the agri-food value chain. The aim of this research was to develop a methodology that could collectively quantify the economic, social, and environmental sustainability of agricultural businesses, livestock, and agri-food industries using a composite indicator. The research objectives, focusing on identifying indicators and proposing a methodological framework based on a composite indicator, guide the subsequent presentation of the methodology, results, and discussion, concluding with insights into limitations and future research directions.

Materials and methods

Systematic literature review

The initial phase encompassed a comprehensive qualitative systematic literature review of English-language scientific publications. The primary goal was to discern an initial set of indicators for subsequent evaluation by a group of experts, along with the methodology for constructing the composite indicator (refer to Section "Panel of experts").

This systematic review, crucial for obtaining information and guiding researchers in addressing a specific issue, namely the numerical quantification of sustainability within the agri-food sector, employed the "snowball" or "reference chain" protocol (Batlles-delaFuente et al. 2022; Kitchenham 2004; Tranfield et al. 2003). Scopus, a widely recognized repository for scientific publications, books, and conferences, served as the database due to its comprehensiveness (Elsevier 2022; FECYT 2022). The search strategy involved utilizing keywords such as agriculture, livestock, agri-food industry, economic sustainability, social sustainability, environmental sustainability, quantification, and their synonyms, leading to an initial pool of publications for the subsequent "snowball" systematic review. Additionally, a review of relevant reports from governments and international institutions contributed to the literature review (European Commission 2017, 2018c; FAO 2013, 2014). The systematic literature review spanned from January to September 2023, and the identified publications are listed in Appendix A.

Indicators selection

The literature review facilitated the identification of initial indicators for assessing economic, social, and environmental sustainability within the agri-food value chain during the subsequent research stage. The scientific literature has leveled severe critiques at the types of indicators that can be selected in research: reflective and causal/formative. Proponents of reflective indicators have advocated for banning the use of causal/formative indices. However, a review of the subject has established that the common criticisms made by advocates of reflective indicators lack a solid foundation, as they often involve logical fallacies and insufficient evidence. The conclusion of this research was clear: The construct formed by a battery of formative indicators should not be dismissed as it does not misrepresent reality. Therefore, it is determined that measurement should also incorporate causal formative indicators, as these should not be discarded due to their novelty or unsubstantiated positions (Bollen and Diamantopoulos 2017).

The selected indicators were required to exhibit a robust relationship among economic, social, and environmental components within the circular economy framework (Silvestri et al. 2022). Additionally, adherence to criteria outlined by relevant research and institutions was imperative, mandating indicators to be quantifiable, understandable, reliable, accessible, and aligned with sustainability goals (Castillo-Díaz et al. 2023b; OECD and European Commission, 2008; Opon & Henry 2019). Furthermore, a strict compliance check with the Pressure-State-Response (PSR) model principles, the globally endorsed system for sustainability index selection, was conducted (Hazbavi et al. 2020; Linster and Flecher 2001; Suresh et al. 2022; Wang and Wang 2021; Wolfslehner and Vacik 2008; Woodhouse et al. 2000). The PSR model, underpinned by causality, stipulates that human activities exert pressure on systems, influencing changes in the system's state. This model, recommended by entities like OECD and the United Nations Sustainable Development Commission, sets a maximum limit of 50 indicators per activity (Li et al. 2021; Linster and Flecher 2001). In this phase, indicators were assigned preliminary units in anticipation of potential modifications in the subsequent procedure (see Fig. 1 for the conceptual framework of the PSR model, adapted from Woodhouse et al. 2000).

Fig. 1
figure 1

Source: own elaboration adapted from Woodhouse et al. (2000)

Conceptual framework of the Pressure-State-Response (PSR) model.

Full size image

Questionnaire design

The third stage involved crafting a questionnaire based on recommendations from various institutions and prior research (González-Alzaga et al. 2022; MTASE and INSHT 2019; Roopa and Rani 2012). Key considerations included the questionnaire's effectiveness as a measurement unit, avoidance of excessive questions, and adherence to a logical sequence of inquiries.

Panel of experts

Objective and methodology

Following the initial research phase, a panel of agri-food sustainability experts was assembled to validate the methodological proposal and refine the initially selected battery of indicators. The Delphi methodology, renowned for its ability to achieve consensus through iterative surveys, was employed (Ghazy et al. 2022; Graham et al. 2003; Landeta 2006; Moutinho, Fernandes, and Rabechini 2024; Okoli and Pawlowski 2004). This method ensured respondent anonymity throughout, mitigating influences and promoting candid expression of opinions. This was one of the reasons for employing the Delphi methodology instead of other protocols, such as workshops or roundtable discussions (Dalkey and Helmer 1963; Yousuf 2007). Focus groups may not be appropriate if the experts are geographically dispersed or if the topic requires extensive individual contemplation rather than group discussion. The specialized literature has indicated the minimum criteria that Delphi analyses must meet to adhere to the required scientific rigor. These studies have determined the criteria that should be presented as quality indicators, which must be included in the final reports of a Delphi study (Landeta and Lertxundi 2024).

Participants

Sixteen experts participated in the initial round, with 10 contributing to the second, culminating in a consensus on the indicator set. The panel, comprising specialists in agri-food sustainability, met the minimum recommended threshold of 10 to 18 experts (Graham et al. 2003; Landeta 2006; Okoli and Pawlowski 2004). Our research falls within the minimum range recommended by the scientific literature (Landeta and Lertxundi 2024).

All the participants were university graduates, and half them had a doctorate. Half of the participants were from university centers, and 37.5% were from primary companies. The rest of the specialists came from administration and research centers. The experts consulted had a minimum of five years of experience in public or private organizations connected to agri-food sustainability (Table 2).

Table 2 Origin of the specialists involved in the round of consultations
Full size table

Communication and sampling strategy

Upon identifying panelists, collaboration was sought via email confirmation, aligning with established practices (Okoli and Pawlowski 2004). An online survey, employing Google Forms, was administered in two phases, featuring dichotomous and open-ended questions. Closed queries gauged the adequacy of selected indicators for characterizing sustainability, while open-ended questions allowed specialists to offer observations, including suggestions for new indicators or alterations (Landeta 2006). Throughout all survey phases, it was ensured that panelists could share any opinions they deemed appropriate, aiming to capture all possible information flows. The literature establishes that a quality criterion for Delphi studies is the appropriateness of the mechanisms for obtaining and managing expert knowledge (Landeta and Lertxundi 2024). Qualitative and quantitative insights were garnered during the September and October 2023 consultations.

Consensus criterion

The literature states that the panelists should be allowed to exchange anonymous opinions until a consensus is reached (Landeta and Lertxundi 2024). The study adhered to a consensus criterion requiring a total proportion of votes exceeding 70% for indicator inclusion. Indicators failing to secure a 70% favorable vote in the initial phase were excluded. The criteria were established based on the recommendations of previous literature (Bernabeu et al. 2021; Collado et al. 2022).

Results and discussion

This section outlines the specialized methodology for assessing the sustainability of agri-food systems by activity and subsector. The protocol, based on various stages, involves validating model indicators through a meticulous process.

Validation of model indicators

After identifying indicators from existing literature and selecting them through the PSR protocol, a panel of experts evaluated their suitability using the Delphi methodology.

First round of consultations

In the realm of economic indicators, the consensus on factors like public subsidies and liquidity ratio was elusive. Instead, the experts suggested incorporating indicators such as working capital and inflation, essential for gauging a company's economic standing (IMF 2023), in addition to other indicators of a social and environmental nature (Table 3). This nuanced economic characterization aligns with the trajectory of agri-food operators facing rising production costs in recent years (Castillo-Díaz et al. 2023b). Thus, the battery of indicators selected by the panelists makes it possible to characterize a company from an economic point of view (Table 4). One of the challenges observed during the indicator selection process, after the panelists had made their choices, was the non-selection of indicators with a selection rate close to 70%, despite some of them being highly recommended in the literature, such as the liquidity ratio.

Table 3 Indicators proposed by the panelists during the first round of consultations
Full size table
Table 4 Frequency of experts' (n = 16) selection on the adequacy of economic indices for measuring the economic sustainability of agricultural and livestock farms and agri-food industries
Full size table

Concerning social indicators, the panelists dismissed family/non-family distinctions and family employment due to complexities in calculations. Instead, they advocated for Annual Work Units (AWU) as a more relevant measure, rejecting certain social wage supplements and emphasizing the exclusion of certain ratios and indicators, such as occupational accidents, from the model. The panelists did not recommend the use of certain indicators, such as salary supplements beyond regulatory obligations, even though these could be considered indicators reflecting a company's social commitment to its employees (Tables 3 and 5).

Table 5 Frequency of expert (n = 16) selections on the adequacy of social indices to measure the social sustainability of farms and agri-food industries
Full size table

On the environmental front, the experts advised against using the number of environmental files as a parameter. They stressed the significance of renewable energy in determining sustainability, echoing global initiatives emphasizing its role in reducing greenhouse gas emissions (Gielen et al. 2019). For agriculture, indicators like nitrogen fertilizers consumption, water footprint, and alternatives to plastic gained prominence. The urgency of addressing plastic pollution was underscored, aligning with the circular economy principles (European Commission 2018a, 2018b; Mazur-Wierzbicka 2021; Paris et al. 2022) (Table 6). Additionally, they recommended the use of the average safety period of plant protection products as an indicator of the hazardousness of active substances applied in plant health actions (Table 3).

Table 6 Frequency of expert selections (n = 16) on the adequacy of environmental indices for measuring the environmental sustainability of farms
Full size table

In livestock production, the specialists highlighted energy consumption, wastewater generation ―produced mainly during the cleaning and disinfection of the facilities―, the total waste generated along with its relation to the amount of feed, and the reuse of manure. In livestock production, apart from energy consumption, a parameter of relevance for some subsectors, such as poultry meat and egg and white layer pigs, due to their high energy demand (Costantino et al. 2016), is waste management. Waste management is crucial due to the environmental hazard that can result from the poor management of certain materials, such as slurry, which can significantly reduce the environmental quality of ecosystems (Fangueiro et al. 2010; Sommer et al. 2013). It is noteworthy that zoosanitary products did not obtain the highest score when poor drug management can harm food safety (WHO 2015) (Table 7). As in agriculture, the panelists highlighted the need to characterize the degree of involvement of livestock operators in reducing the consumption of petrochemical plastic on their farms. Additionally, they indicated that compliance or non-compliance with the Integrated Health Plan for livestock farms should be included as an indicator (Table 3).

Table 7 Frequency of expert (n = 16) selections on the adequacy of environmental indices for measuring the environmental sustainability of livestock farms
Full size table

In the agri-food industry, the parameters that obtained the highest frequency of selection by the panelists were energy consumption, variables related to waste and its management, and water footprint (Table 8). Some countries have highlighted the need to reduce or eliminate food waste and the waste footprint of the activity (Cortes Generales 2023). These parameters can measure such influence. The agents consulted highlighted the need to include a variable to determine the type of packaging used in agri-food products, given the high amount of petrochemical plastic often used for packaging these products (Kirwan et al. 2011). It was surprising that the panelists recommended using the number of environmental files as an environmental variable in the agri-food industry, yet they dismissed its use in agricultural and livestock companies.

Table 8 Frequency of experts' (n = 16) selection on the adequacy of environmental indices to measure the environmental sustainability of agri-food industries
Full size table

Second round of consultations

In the second round of consultations, the ten panelists who responded indicated that the models of indicators selected were suitable for assessing the sustainability of farms and agri-food industries from a multisectoral point of view (Tables 9 and 10).

Table 9 Frequency of selection of the proposed model of indicators to assess the sustainability of farms and livestock farms and the agri-food industry
Full size table
Table 10 Indicators proposed to quantify the economic, social, and environmental sustainability of productive units in agriculture, livestock, and the agri-food industry
Full size table

Table 10 shows the final list of selected indicators and the negative or positive effect they have on sustainability, according to previous literature and the expert consultation carried out using the Delphi methodology. Table A1 shows details of the calculation of the ratios indicated in Table 10 (see the appendix A).

Design and Development of the questionnaire

The formulary is divided into four sections (see the appendix C). The first block focuses on describing and locating the production unit to be evaluated, along with its characteristics and location within the agri-food chain. The second block of questions characterizes the sustainability of each agricultural operation or agri-food industry based on the indicators identified in Table 10. This order is based on the design criteria recommended by previous research and organizations, including the most uncomfortable questions for the respondent (i.e., economic issues of the business) at an intermediate point of the questionnaire. The third block contains questions aimed at the sociodemographic characterization of the respondent. Finally, the fourth block of the questionnaire includes a dichotomous yes/no question to evaluate the respondent's opinion on whether the questionnaire is adequate to determine the sustainability of the agri-food sector. Additionally, an open-ended question with a limit of 200 words is included to allow the interviewed agent to propose improvement measures.

Composite indicator construction strategy

Standardization of sustainability indicators

The quantification of sustainability built from dimensional indexes entails a problem: The impossibility of developing a composite indicator that allows us to classify the sustainability of an activity from a system of indicators. This is due to the diversity of input units. Therefore, it is necessary to resort to mathematical models that allow us to normalize source information into dimensions without creating groups. Previous research that has attempted to quantify the economic, social, or environmental sustainability of economic activities and institutions such as the UN, the OECD, and the European Commission have recommended the following mathematical formulas (Bao et al. 2022; Hahn et al. 2009; Martínez 2009; Nasrnia and Ashktorab 2021; OECD and European Commission 2008; Omerkhil et al. 2020; Pandey and Jha 2012; Suresh et al. 2022), which are the recommended expressions for our model:

$${\text{Index}}_{{{\text{sd}}}} = {\raise0.7ex\hbox{${S_{d} - S_{\min } }$} \!\mathord{\left/ {\vphantom {{S_{d} - S_{\min } } {S_{\max } - S_{\min } }}}\right.\kern-0pt} \!\lower0.7ex\hbox{${S_{\max } - S_{\min } }$}}$$
(1)
$${\text{Index}}_{{{\text{sd}}}} = {\raise0.7ex\hbox{${S_{\max } - S_{d} }$} \!\mathord{\left/ {\vphantom {{S_{\max } - S_{d} } {S_{\max } - S_{\min } }}}\right.\kern-0pt} \!\lower0.7ex\hbox{${S_{\max } - S_{\min } }$}}$$
(2)

where,

Sd is the value of each indicator.

Smax and Smin are the maximum and minimum value of the set of indicators for the same indicator.

Based on the formulas indicated above, a value of Xn can be classified in the range set between the maximum (Xmax) and minimum (Xmin) value of the data set of an index in question. The maximum and minimum values are obtained from the data yielded by the production units evaluated (i.e., farms, livestock farms, and agri-food industries) for the same parameter. Therefore, the representativeness of the model will increase directly proportional to the number of surveys carried out, as this allows the limits of the proposed model to be extended (Castillo-Díaz et al. 2023).

The correct classification must take into account the positive (benefit) or negative (cost) impact that the indicator expresses on the sustainability of the production system being evaluated. The sign of the impact is shown in Table 10, and it has been established from previous literature. Thus, Eq. (1) will be used for indicators that contribute positively to the sustainability of the agri-food system, while Eq. (2) will be used for those that have a negative impact on it.

Constitution of economic, social, environmental, and global sustainability

Subcomponents of sustainability, including economic, social, environmental, and global dimensions, are constructed using a weighted sum of standardized indices. The balanced weighting across indicators within each subcomponent reflects existing literature recommendations (Castillo-Díaz et al. 2023; Ćulibrk et al. 2021; Nasrnia & Ashktorab 2021; Suresh et al. 2022). These works have recommended using similar weighting between the indicators framed within each subcomponent of sustainability (i.e., economic, social, and environmental) and the overall value of the same. Therefore, the recommended mathematical expression is as follows:

$${\text{IS}} = {\raise0.7ex\hbox{${\mathop \sum \nolimits_{i = 1}^{n} {\text{Index}}_{{{\text{sd}}}} }$} \!\mathord{\left/ {\vphantom {{\mathop \sum \nolimits_{i = 1}^{n} {\text{Index}}_{{{\text{sd}}}} } n}}\right.\kern-0pt} \!\lower0.7ex\hbox{$n$}}$$
(3)

where,

IS is the value of the economic (SE), social (SS) and environmental (SEE) and global (SG) subcomponent of the production unit evaluated.

N is the number of indicators.

As indicated in Section "Indicators selection", the influence of the variables may or may not characterize the construction of the latent variable. In our case, the variables measured construct the latent variable (i.e., economic, social, and environmental sustainability) and are causal/formative variables. Thus, there is no reason to exclude these variables (Bollen and Diamantopoulos 2017). Furthermore, the variables and the procedure were validated by the panelist group during the Delphi methodology.

Conclusions, limitations and future research lines

This study endeavors to construct a model for quantifying the sustainability of agricultural and livestock farms as well as agri-food industries. The initial phase involved an exhaustive literature review to discern pertinent economic, social, and environmental indicators. Subsequently, a panel of experts was convened, and the Delphi methodology was employed to refine, enhance, or validate the proposed indicator model.

The outcomes of this investigation posit that the sustainability of agricultural and livestock production units and the agri-food industry can be gauged effectively through a comprehensive set of indicators. This set comprises eight economic indices, five social indices, and 58 environmental indices. Feedback from the consulted experts prompted the exclusion of certain originally selected indicators that did not align with the model's objectives. Furthermore, emphasis was placed on the inclusion of additional indicators to augment the representativeness of the economy and bioeconomy, given their significance in the socioeconomic development of nations. Notably, the requisite information can be easily derived from farms and factories through the questionnaire devised in this study, subsequently standardized and represented by the methodology outlined herein.

The quantification and assessment of the triple dimension of sustainability prove pivotal in ascertaining the economic, social, and environmental viability of agri-food enterprises. This significance is underscored by the regulatory and strategic shifts observed in political developments, particularly within entities such as the European Union. The lack of a standardized methodology for numerically appraising the global sustainability of production units hampers their integration into evolving political landscapes. It not only constrains the formulation of action strategies but also impedes numerical comparisons between values calculated using disparate methodologies. The composite indicator proposed in this methodology facilitates the identification of specific subcomponents within agri-food sustainability that warrant closer attention. By scrutinizing individual indicators, as selected by the expert panel, a tailored action plan can be devised to address the distinct needs of each agri-food production unit.

Despite the valuable contributions of this methodological proposal to the existing body of knowledge, certain limitations merit acknowledgment. Primarily, the theoretical underpinnings, validated by expert consensus, necessitate practical validation through field tests. Such tests would elucidate aspects that may elude comprehension by practitioners in agriculture, livestock farming, and the agri-food industry. Secondly, the precision of the scale hinges on the critical data amassed from future surveys, as the maximum and minimum values in the dataset dictate the classification of a productive unit within the standardized range.

Future research lines should focus on experimental validation of the model developed in two phases. Initially, a validation of the methodology should be carried out involving approximately 15 agri-food companies, evenly distributed across the agriculture, livestock, and agri-food industry sectors. In the second phase, a larger-scale study should be conducted with the aim of achieving a 95% confidence level and a maximum error of 5%, to determine the economic, social, and environmental sustainability of agri-food companies in major food and beverage producing countries, such as Spain.