Dairy farmers’ intention to use calf management technologies in four European countries: A QCA and PLS-SEM approach

Whilst livestock management technologies may help to improve productivity, economic performance, and animal welfare on farms, there has been low uptake of technologies across farming systems and countries. This study aimed to understand dairy farmers’ intention to use calf management technologies by combining partial least squares structural equation modelling (PLS-SEM) with qualitative comparative analysis (QCA). We evaluated the hypotheses that dairy farmers will intend to use calf technologies if they have sufficient competencies, sufficient materials, and positive meanings (e.g., attitudes or emotions) towards calf technologies, and they will not intend to use technologies if one of these elements is missing. An online survey was completed by 269 dairy farmers in Belgium, the Netherlands, Norway, and the UK. A PLS-SEM was developed, where the outcome was the number of calf management technologies that the respondent intended to use, and the latent constructs included meanings, materials, and competencies. QCA was then run separately for the datasets from each country. Intention to use technologies was the outcome, whereas positive meanings, sufficient materials, and sufficient competencies for technology use were conditions in the QCA. Evaluation of the PLS-SEM showed that reliability and validity of the latent constructs was appropriate for analysis. Assessment of the structural model indicated that having positive meanings regarding technologies significantly increased the number of calf technologies the farmer intended to use (β = 0.388, CI = 0.291 – 0.486). The QCA solutions show that the conditions for the intention to use, or not use, calf technologies differed between Belgium, the Netherlands, Norway, and the UK, but the presence (or absence) of positive meanings was consistently important. The solutions for Norway and Belgium aligned with our hypotheses, but the solutions for the Netherlands and UK did not. Some of the solutions exhibited features of causal complexity such as equifinality, conjunctural causation, and asymmetric causation, which would not be able to be easily identified using traditional regression analyses. This study highlights the causal complexity of technology use on farms as a social phenomenon. Furthermore, the study shows the usefulness of QCA for evaluating theoretical hypotheses regarding farmers’ behaviour. We suggest that researchers could use this method to investigate other practices on farms that may have causal complexity.