Table 1 summarizes the topics discussed in the literature and guidelines on the existing sustainability index in agriculture and food systems. Other than economic sustainability, some literature divided sustainable practices into three subcategories: environmental sustainability, external social sustainability, and internal social sustainability (Van Calker et al., 2005; Meul et al., 2008; Lebacq et al., 2013). According to studies of sustainability index in agriculture, internal social sustainability refers to employees’ work environment, education, and other forms of employee well-being, while external social sustainability refers to social values and concerns, such as community revitalization and job creation on the part of agriculture, food security, and the like (Lebacq et al., 2013). In Table 1, more than half of the indicators mentioned are related to energy, water, biodiversity, air quality, land (environmental), food safety, supplier social assessment (external social), and workplace conditions, labor health (internal social). Unlike environmental indicators, social indicators tend to be adopted differently, depending on the purpose and context. In this study, sustainability is quantified for each theme by adding topics considered issues in Japanese agriculture and food systems.

How can these various sustainability indicators be integrated? According to Table 1, more than half of the studies reviewed for this purpose report each indicator individually, without any integration description. However, some weight each indicator using input from experts and focus groups, integrating them into each theme (Van Calker et al., 2005; Meul et al., 2008; Ripoll-Bosch et al., 2012) or a single index (Gómez-Limón and Sanchez-Fernandez, 2010). On the other hand, a study analyzing the correlations among sustainability indicators in different countries has shown that there are synergies (significant positive correlations) among indicators of different themes (Zhang et al., 2021). Then, studies should include a perspective on the interrelationships among indicators across themes. One study that compares tools to assess the sustainability performance of farms highlights the importance of indicator interaction as a condition for understanding the complementarity of sustainability practices (Binder et al., 2010); however, none of the tools analyzed made any reference to indicator interactions (De Olde et al., 2016). Then, this study posits a hypothesis on the sustainability indices as follows:

To further enhance the validity of the indicator, firstly, the context in which it is used needs to be clarified (Latruffe et al., 2016). In comparative studies, the presence or absence of synergies among indicators varied widely across countries (Zhang et al., 2021). In other words, it is important to test the above hypotheses for a specific country or region. Secondly, this study uses practice-based indicators to quantify sustainability. Although performance-based indicators are generally preferred (FAO, 2013), the FAO’s guideline acknowledges that practice-based indicators may be the most practical for small-scale organizations, where measuring the performance of each indicator is not feasible. As this study focuses on small and medium-sized farms, practice-based indicators are used. These practice-based indicators have end-user validity in that it is easy for farmers to answer (Bockstaller and Girardin, 2003). Indicators with such contextual considerations and flexibility for users can also be used for farmers’ strategic decision-making (Coteur et al., 2016). In other words, the indexes in this study are also intended to be applied to farm management.

We use Meuwissen’s framework to evaluate the resilience of farming systems (Meuwissen et al., 2019). The conceptual framework is illustrated in Figure 2. First, all types of farms in Japan are included in the targeted farming system. Second, the COVID-19 pandemic and production cost escalation caused by Russia’s invasion of Ukraine are the challenges farmers face. Third, functions are determined by the most recent changes in the status of a farming business and the existence or non-existence of plans for some new activities over the next five years. Fourth, we clarify the type of resilience capabilities farmers exhibit, i.e., persistence or adaptation. Persistence corresponds to the recent business improvement, while adaptation corresponds to the presence or absence of new activities. Fifth, we focus on sustainability practices as resilience attributes.

Figure 3 shows the analytical framework used in this study. The processes used to develop each indicator and conduct the analysis are explained according to this framework.

Figure 4 depicts the percentage of respondents committed to sustainability practices by farm type. Regarding environmental sustainability, approximately 30% of rice and vegetable farms exhibit involvement in “reduction of chemical use” and “organic farming.” Notably, rice farming, which relies heavily on water as a natural resource, is characterized by a high degree of attention toward “water withdrawal and quality.” Over 50% of dairy and beef cattle farms pursue “cooperation of crop cultivation and livestock production.” This practice has garnered attention for its environmentally friendly aspects through resource recycling and its capacity to mitigate increasing production costs.

In the category of external social sustainability, “traceability” and “food safety” are the most prominent practices, implying widespread production process innovations. Most rice farms included in the survey have implemented “landscape management,” indicating that rice farming significantly impacts Japan’s rural landscape. For vegetable farms, “local employment” is a popular form of pursuing social sustainability. Meanwhile, many dairy and beef cattle farms engage in practices involving the “use of local resources.” However, there is limited interest in some issues where agriculture is expected to play a significant role, including “food culture diversity” and “food access.”

Concerning internal social sustainability, “workplace safety” is the top priority for respondents across all farming types. However, investment in human resources such as “adequate income” and “capacity development” is relatively low. Aside from vegetable farms, only about 10% of farms prioritize “workforce diversity.” These findings emphasize the need for sustainability assessments in Japanese agriculture to encompass not only the environment, local communities, and consumers but also the welfare of farm employees.

Based on a tetrachoric correlation analysis on sustainability indicators (see Supplementary Table S1), we employ a factor analysis to analyze the underlying farmers’ sustainability goals and use the factor scores as integrated sustainability indices. Figure 5 of the parallel analysis indicates that farmers’ sustainability goals can be characterized by six factors. Table 4 displays the factor loadings of these sustainability indicators within the six-factor model. Factor 1 includes high factor loadings for “local employment” and for all internal social sustainability elements, excluding “workforce diversity.” Hence, factor 1 is interpreted to represent the goal of job creation and investment in human capital, which we summarize as “caring for employees.” However, the factor loading for “Workplace Safety” is relatively low at 0.41, indicating that income and human resource development are more important in caring for employees.

Factor 2 incorporates food-related issues such as “food culture diversity,” “food education.” and “food access” into the environmental sustainability concerns represented by “reduction of chemical use” and “organic farming.” Thus, this factor addresses environmental and social issues linked to food and can be defined as “solving food challenges.” Factor 3 centers on conserving water and soil resources, encompassing aspects such as “water withdrawal and quality” and “soil quality and land degradation,” as well as the upkeep and enhancement of the region, including “landscape management” and “community development.” Accordingly, this factor is labeled as “caring for land and community,” signifying that farmers consider the conservation of local land and community as an integral part of their farming activities.

Factor 4 and factor 5 are composed exclusively of “cooperation of crop cultivation and livestock production” and “food safety,” respectively. Factor 4 also exhibits a relatively higher factor loading for “animal welfare,” which indicates it is a primary objective for livestock farmers. Factor 6 exhibits high factor loadings for climate change mitigation measures, including “renewable energy use” and “greenhouse gases reduction,” as well as for the “use of local resources.” Therefore, it can be defined as “use of natural and local resources,” as it aims to promote the efficient and integrated use of natural and local resources.

Four of these six factors denote the interactions between environmental sustainability and external or internal social sustainability indicators. This fact shows the importance of incorporating sustainability practices that align with farmers’ specific goals rather than solely analyzing the indicators or classifications obtained from standardized themes, supporting the hypothesis 1.

According to the results of the analysis of variance on the characteristics of the integrated sustainability indices by farming type and farming area type (Table 5), vegetable farms prioritize “caring for employees,” “solving food challenges,” and “food safety.” In contrast, rice farmers exhibit a significantly higher mean value for the “caring for land and community” factor than other farm types.

Examining the relationship between geographic type and sustainability goals shows that “caring for employees,” “solving food challenges,” and “food safety” are predominantly addressed in urban regions. This fact may be attributed to the higher demand for quality working conditions and more health-conscious consumers in urban areas, where conditions for agricultural production are not always favorable due to urban sprawl.

Table 2 displays the descriptive statistics concerning the resilience index. As for the persistence resilience index, fewer than 10% of respondents said business or financing conditions, production costs, or the business outlook improved.

Additionally, Table 2 also demonstrates the indicators of adaptive resilience. Less than half of farmers had plans for any activities in the next five years. The results of Table 2 indicate that a limited number of farmers possess resilience capacities, such as persistence and adaptation, in the current socio-economic environment.

Table 6 exhibits the factor analysis findings conducted for the persistence index. Figure 6 of parallel analysis for persistence indicates the presence of a one-factor model in this context. This fact substantiates that the four indicators above collectively form the persistence index. Similarly, Table 7 presents the outcomes of the factor analysis conducted for the adaptation index. Figure 6 of the parallel analysis for adaptation suggests a two-factor model as optimal. Factor 1 demonstrates substantial factor loadings for “Plan for new marketing channel,” “Plan for new business,” and “Plan for new product,” indicating that it serves as an index representing farmers’ inclination towards diversifying their farms. On the other hand, factor 2 solely displays a large factor loading for “Plan for new technology.” Hence, this factor signifies farmers’ inclination towards future technological transformation.

The results of the regression analysis of the relationship between resilience and sustainability, controlling for the fundamental variables (see Supplementary Table S2), are presented in Table 8. Regarding the integrated sustainability index, we note that “caring for employees,” “solving food challenges,” and “use of natural and local resources” are positively correlated with the persistence index, supporting the hypothesis 2. These sustainability practices are, therefore, key attributes of short-term resilience.

In addition to these three sustainability indices, the remaining three integrated sustainability indices, “caring for land and community,” “cooperation of crop cultivation and livestock production,” and “food safety,” are positively associated with the diversification index and technological change index, suggesting that a broader range of sustainability goals may potentially play a crucial role in the long-term adaptation aspect of resilience rather than in short-term persistence. This finding supports the hypothesis 3.

The positive effect of the persistence index on the diversification index suggests that demonstrating short-term resilience is an important step in attaining long-term resilience. The Aroian test estimated the indirect effects of each integrated sustainability index through the positive impact of the persistence index on the diversification index, and found that “caring for employees,” “solving food challenges,” and “use of natural and local resources” all had positive indirect effects (p < 0.01).

Table 8 also presents the resilience levels of four farm types. The coefficient for vegetable farms is the highest for the persistence index, while the coefficient for dairy farms is the lowest. These findings suggest that the effects of shocks such as the COVID-19 pandemic and rising production costs, which are the focus of this study, vary considerably across farm types. Regarding the adaptive resilience, rice and vegetable farms exhibit a greater propensity for planning various novel activities than livestock farms. Furthermore, rice farms predominantly prioritize the pursuit of technological transformation. Our results highlight a significant connection between integrated sustainability indices and farm resilience, even after accounting for farm types and other control variables.

The integrated sustainability index highlights the sustainability goals of Japanese farmers by accounting for the interactions between ranges of sustainability indicators. The findings show that four of the six factors we identify display linkages between environmental, external social, and internal social sustainability. This result confirms there is complementarity among sustainability practices (Binder et al., 2010) using practice-based indicators (FAO, 2013) of small and medium-sized farmers. As such, the results suggest that the indices used in this study are more adept at evaluating farmers’ sustainability endeavors than an index that integrates all indicators (Gómez-Limón and Sanchez-Fernandez, 2010) or that integrates indicators within a single theme (Van Calker et al., 2005; Meul et al., 2008; Ripoll-Bosch et al., 2012).

Using the indices constructed in this study reveals dissimilarities concerning certain sustainability practices by farm type, including consideration for employees and food-related challenges in vegetable farms, consideration for land and community in rice farms, employment of natural and local resources in dairy farms, and arable-livestock interdependence in beef cattle farms. Moreover, we demonstrate that the orientation of sustainability practices varies based on geography-based qualities. Notably, the findings that highlight differences in active consideration for employees and food-related challenges in urban areas, which face inherent farming difficulties, imply that sustainability practices are crucial in the continuity of farming in disadvantaged areas.

The post-COVID-19 economic turbulence precipitated a decline in economic performance for over half of all farmers during the study period. Furthermore, the proportion of farmers intending to alter their management practices within the forthcoming five years remained notably modest. Regression analysis elucidates substantial variations in economic performance across different farm types, notably highlighting the precarious situation within livestock farming sectors such as dairy and beef production. In essence, enhancing farm resilience emerges as a crucial imperative for the agricultural sector to surmount such economic adversities.

While previous study has shown a positive correlation between agricultural sustainability and resilience using country-level aggregate data (Volkov et al., 2022), this study has quantitatively demonstrated the impact of sustainability practices on farm resilience by indexing agricultural sustainability and farm resilience at the farm level. Moreover, by dividing agricultural resilience into persistence and adaptation, we have succeeded in approaching the important issue of identifying factors that lead to transformability resilience in agriculture (Meuwissen et al., 2021).

Firstly, the integrated sustainability indices developed in this study provide a better understanding of the relationship between farm resilience and sustainability. We find that “caring for employees,” “solving food challenges,” and “utilizing natural and local resources” are positively correlated with the farm resilience index for persistence. Consistent with the previous study (Meuwissen et al., 2018), one factor contributing to resilience could be the preservation and development of human resources through “caring for employees.” Secondly, “solving food challenges” is a resilience attribute for sustainable production. As shown in previous studies (Mastronardi et al., 2020, 2022; Yoshida and Yagi, 2021; Grigorescu et al., 2022; Little and Sylvester, 2022), diversifying to tackle various food-related challenges is also a crucial attribute of resilience. Finally, “utilizing natural and local resources” contributes to the resilience attribute of risk management by diversifying management resources. As shown in the previous study (Binder et al., 2010), this result suggests that implementing multiple practices aligned with each sustainability goal is critical to enhancing farm resilience.

On the other hand, we do not find a positive relationship between “caring for land and community” and the persistence index. This sustainability goal has traditionally been recognized as a critical function of Japanese rice farming, which helps preserve rural landscapes and communities. Nonetheless, if such activities are not linked to short-term farm resilience, this poses a significant challenge to the continuity of rural Japan. Similarly, “cooperation of crop cultivation and livestock production” and “food safety,” emphasized administratively in recent years, exhibit a weak relationship with farm resilience. In other words, while certain sustainability practices have a strong complementary relationship with farm resilience, others have a weak connection. The latter will only be able to maintain in the long term with measures to support continuous farm efforts.

Secondly, we observe a stronger positive correlation between sustainability practices and the long-term adaptive resilience indices. In other words, resilience attributes attained through sustainability practices are more closely associated with innovations that may significantly alter a farm’s management style in the long term rather than short-term financial performance improvements. As a form of diversification that transforms farm management and farm identity, sustainability practices promote stakeholder management and involvement, which are known to be highly effective in promoting transformability (Nera et al., 2020; Soriano et al., 2023). Furthermore, the adaptive resilience index has been shown to be influenced by the persistence resilience index. Consequently, the sustainability practices indirectly affect long-term resilience through short-term resilience improvement. Considering that many farm resilience studies indicate a lack of factors promoting transformability (Meuwissen et al., 2021), this study highlights the significant role that sustainability practices play in enhancing adaptive farm resilience.

The social implications of the above results are discussed. First, financial institutions can consider socially and environmentally sustainable practices in agriculture as a risk management tool and use them in their financing and investment decisions. Second, Japan’s Ministry of Agriculture, Forestry and Fisheries (MAFF) has developed guidelines on ESG regional finance. The results of this study provide evidence for the effectiveness of such guidance and also suggest a simplified method of identifying and interpreting sustainable practices. Third, for farmers who implement sustainable practices, the contribution of this study is that it clearly shows the process of achieving agricultural sustainability, in which farmers can attain economic sustainability by choosing multiple sustainable practices according to their business objectives.