Appropriate-scale operation is important for integrating ordinary farm households into modern agriculture, yet the appropriate scale may differ substantially across producer types. Using micro-level data from the nationally representative 2021 China Household Finance Survey, this study develops an analytical framework integrating efficiency and resilience perspectives to compare ordinary farm households and family farms. Efficiency is the core empirical dimension and is assessed using cost–benefit analysis, the DEA-BCC model, and multivariate econometric methods. Resilience is treated as a theoretical perspective and a constraint on scale decisions; it is interpreted indirectly through the scale responses of the two groups to disaster shocks rather than measured using a standalone resilience index. The results identify distinct efficiency-based appropriate-scale ranges: 1–3 mu (approximately 0.07–0.20 hm²) for ordinary farm households and 100–200 mu (approximately 6.67–13.33 hm²) for family farms, with lower net profit per unit area at intermediate scales. Economic rationality and institutional constraints show marked heterogeneity across producer types: commercialization, land titling, subsidies, and credit access are more strongly associated with the scale expansion of family farms, whereas non-farm employment opportunities play a more decisive role for ordinary farm households; household labor and social networks contribute to scale expansion in both groups, and their between-group differences are not statistically significant. Disaster impact is associated with a significant reduction in the operational scale of ordinary farm households, whereas in the subgroup estimates the operational scale of family farms shows no significant association with disaster impact, suggesting that family farms are better able to maintain their operational scale. Digital technology also operates through differentiated pathways: it primarily alleviates information and factor constraints among ordinary farm households while strengthening efficiency, financial access, and organizational coordination among family farms. These findings support differentiated scale and digital-agriculture policies in economies characterized by heterogeneous farm structures.

Ecological efficiency is fundamental to the sustainable development of the livestock industry. Understanding how the digital economy enhances this efficiency is critical for promoting high-quality sectoral development. Using panel data from 30 Chinese provinces during 2011–2023, this study measures livestock ecological efficiency with a Super-SBM model and conducts empirical analyses using two-way fixed effects, mediation, spatial econometric, and panel threshold models. The results show that digital economic development significantly improves livestock ecological efficiency. This effect is realized through the expansion of livestock farming scale and the upgrading of the industry’s structure. Spatial panel analyses further indicate that digital economic development generates significant spatial spillover effects, which positively enhance ecological efficiency in neighboring regions. Further analyses reveal clear regional heterogeneity. Both the direct and spatial spillover effects decline progressively from eastern to western China. In addition, a significant single-threshold effect is identified. Based on these findings, this study proposes establishing cross-regional coordination mechanisms linking digital economy development with ecological compensation, implementing region-specific development policies, and strengthening ecological infrastructure within the livestock sector. These measures would help fully leverage the digital economy in improving livestock ecological efficiency.

Exploring the spatial correlation network (SCN) characteristics and influencing factors of the eco-efficiency of cultivated land use (ECLU) in Xinjiang’s counties is crucial for clarifying inter-regional spatial mechanisms and supporting cross-regional collaborative governance. Using a panel dataset covering 85 counties across Xinjiang, this study adopts the super-efficiency SBM model, revised gravity model, social network analysis and QAP model to quantify ECLU, and further investigate its spatial network features as well as driving mechanisms. The results showed that: (1) ECLU exhibited a fluctuating upward trend with significant regional differentiation, and northern Xinjiang performed notably better than southern Xinjiang. (2) SCN remained connected overall, but network density was low, average path length was long, and spatial transmission efficiency was relatively low. (3) Regional differences in economic levels, labor productivity, and industrial structure all had positive effects on the formation of the SCN throughout the study period. Regional differences in fiscal support for agriculture had positive effects only in 2014 and 2017, while differences in the soil and water coordination ratio had a negative effect in 2021. Future policies for sustainable cultivated land use should be differentiated and zone-specific, based on each county’s role within the correlation network, to promote coordinated improvement of ECLU across counties.