Farmers’ willingness to invest in livestock disease control: the case of voluntary vaccination against bluetongue

Animal health authorities in the European Union nowadays consider voluntary approaches based on a neoliberal model of cost and responsibility sharing as a tool for controlling livestock diseases. Policy makers aim for policies that are soft and optional, and use insights from behavioural economics and social psychology. Voluntary approaches are flexible in terms of legislation and can be effective at lower costs, provided that farmers are willing to participate. In 2008, the Dutch animal health authorities used a voluntary vaccination approach to control an emerging bluetongue epidemic that started end of 2006. Nearly 60,000 holdings with ruminants were already affected by the end of 2007 and experts indicated that transmission could only be stopped through mass vaccination. Farmers were motivated to participate by informational and financial, incentive-based policy instruments. Economic theory predicts that farmers underinvest in private disease control measures in the presence of externalities. These studies, however, assume farmers only consider the private economic motives and that they only can be extrinsically motivated via (monetary) incentives. If the willingness to invest in livestock disease control is also driven by intrinsic and social motives, this could imply that not only financial compensation, but a mix of policy instruments is needed to make voluntary approaches work. The overarching research objective of this thesis was to assess the key determinants of farmers’ willingness to vaccinate against bluetongue and study the impact of different policy designs on the effectiveness of voluntary vaccination approaches to bluetongue disease control. A three-stage research approach was conducted. Two models of decision making, one from economics and one from social psychology, were first applied to the case study to obtain a solid understanding of important perceptions and motivations that farmers have to invest in livestock disease control. These motivations (sometimes incentives) and perceptions were then related to different attributes of a vaccination scheme to have a better understanding of how a higher uptake can be obtained. In the third stage, the effect of the interplay between farmers’ collective behaviour and disease epidemiology on disease rate and vaccination uptake was studied. Expected utility theory was used in combination with decision analysis and Monte Carlo simulation in chapter 2. The economic risk and monetary outcomes of the vaccination decision were considered, intrinsic or social motives ignored. The theoretical expectation from the analysis is that with high probabilities of herd exposure and disease effects at the start of the outbreak the farmer decides to vaccinate. Re-vaccination is uncertain during the course of the epidemic due to a lower probability of herd exposure and enduring protection against infection from previous vaccination. Factors that make re-vaccination more likely to happen are risk-averse behaviour and farm management aimed at the export of heifers. The decision moment – before or during an epidemic – and the characteristics of the disease – endemic, epidemic or emerging – are important factors in perceptions of disease risk. Chapters 3 to 5 used data from a survey that was based on the reasoned action approach. Data were analysed with a variety of statistical, mostly multivariate, techniques. The relative importance of the social-psychological constructs in predicting the intention to participate in a hypothetical reactive vaccination scheme against bluetongue was assessed in chapter 3. It was found that intended vaccination behaviour was mainly explained by farmers’ attitude, but also by social pressures from injunctive and descriptive norms. Perceived behavioural control was the least important predictor of intention. The most influential beliefs underlying the social-psychological constructs were assessed in chapter 4. Results suggested that instrumental beliefs (e.g. risk reduction) as well as experiential beliefs (e.g. animal welfare) were important drivers of the attitude towards vaccination against bluetongue. This indicates that in addition to monetary outcomes of the decision, at least a group of farmers also consider the non-monetary (or non-pecuniary) outcomes. The results further showed that the most influencing referents for the farmer are the veterinarian, his or her family members and colleague dairy farmers (peers). Two influencing control beliefs were associated with the provision of information and perceived trust and confidence in the vaccine safety, effectiveness and government approach to control the disease. The aim of chapter 5 was to explore factors that could explain heterogeneity in farmers’ attitudinal beliefs. In particular, perceived risk, measured by a relative risk attitude and risk perception, and the Big Five personality traits were associated with variability in these beliefs. Conscientiousness discriminated farmers into a group of ‘vaccination intenders’ and non-intenders although it remained somewhat unclear how it relates to the decision problem, as it can be a sense of duty, achievement striving or both. The perceived risk measures were related to the milk production intensity and also discriminated intenders from non-intenders. These differences in perceived risk indicated that farmers might not be commonly risk averse, however, it is important to account for the domain specificity of risk taking behaviour. A survey-based discrete choice experiment was used in chapter 6 to study more deeply farmers’ choices for different voluntary bluetongue vaccination scheme designs. A generalised random utility model of farmers’ behaviour allowed for heterogeneity in motives to invest in bluetongue disease control. Results showed that farmers have private economic motives (incentives) to participate in a vaccination scheme, such as to insure the production risk from disease infection and to maintain the export of heifers. Interaction effects found between social-psychological constructs and specific designs of policy instruments highlighted the importance of perceived trust and confidence in the vaccine safety and effectiveness and in the disease control strategy chosen by animal health authorities. Attitude interacted positively with government communication (information) provided via veterinarians. Descriptive norm interacted positively with a lower perceived probability of adverse effects. This suggests that farmers are more likely to vaccinate if they perceive that others in their social network perform vaccination without experiencing adverse effects. Injunctive norm interacted negatively with a higher level of government subsidy. This suggested a crowding-out mechanism through which subsidization adversely affect farmer’s motivation to comply with the vaccination policy. The interplay between farmers’ collective behaviour and bluetongue disease epidemiology was studied in chapter 7 with an agent-based model. The utility model specification from chapter 6 was used to describe the decision-making process of farmers. Other components that added to the dynamic nature of the model were a social network structure of the diffusion process of sharing information about vaccination status and a susceptible-latent-infectious-recovered model of disease spread. The effectiveness of different bluetongue vaccinations scheme designs was studied as measured by disease rate and vaccination uptake. Results of chapter 7 showed that vaccination schemes that focus more on motivating farmers via informational instruments were somewhat more effective than predicted from the comparative static analysis in chapter 6. Motivation via financial incentives resulted in a somewhat lower effectiveness than was predicted from that same model. This might be explained as an emergent effect that evolves under specific vaccination scheme designs from the interactions between farmers themselves and with the environment from which they observe the progress of the disease. These schemes focus more on serving the information needs of farmers and raising the perceived trust and confidence in the disease control approach rather than on incentivising with higher levels of subsidy. Three themes for livestock disease control emerged from the synthesis of the results in chapter 8, which were subsequently discussed in relation to the wider economic and (social) psychological literature. These themes coincide with shortcoming of the standard economic model of rational choice to describe and predict behaviour. The first theme was about understanding how farmers cope with risk in the context of livestock diseases. The second theme focused on the usefulness of financial compensation as a policy instrument. The third theme discussed the role of trust and social norms. After discussing the implications for policy making, main scientific contributions and suggestions for future research, the chapter concluded that: Dutch dairy farmers who operate large-scale and intensive farms or keep heifers for export are likely to have private economic motives to vaccinate against bluetongue (Chapter 2, 4, 5 and 6).Farmers’ willingness to vaccinate against bluetongue is mostly driven by attitude, followed by perceived social pressures from injunctive norms and descriptive norms. This implies farmers can be motivated intrinsically, extrinsically, or both (Chapter 3).Dutch dairy farmers have intrinsic motives to vaccinate against bluetongue. They do not want to be confronted with animal suffering but want to keep job satisfaction high from working with healthy animals (Chapter 4).Dutch dairy farmers have social motives to vaccinate against bluetongue. They consider what important referents, such as the veterinarian or family members, think they should do and take into account the perceived behaviour of peers (Chapter 3 and 4). Perceived risk, personality traits and past behaviour are important behavioural variables for explaining the heterogeneity in beliefs to vaccinate against bluetongue (Chapter 5). The efficacy of financial, incentive based instruments to motivate to vaccinate against bluetongue is heterogeneous and not necessarily positive for each farmer. They are not effective if farmers already expect a positive net benefit from vaccination or if they crowd-out the motivation to comply with the vaccination policy (Chapter 2, 4, 6, 7). The efficacy of informational policy instruments to motivate farmers to vaccinate against bluetongue is positively affected by farmers’ attitude towards vaccination and in case farmers perceive the communication channels used as credible and trustworthy (Chapter 3, 4, 6). The efficacy of social interaction mechanisms in policy making, such as the perceived social pressuretovaccinateagainstbluetongue,ispositivelyaffectedbyfarmers’trustandconfidence in the government approach to control the disease (Chapter 4, 6, 7).