Introduction

New Zealand is under increasing pressure from terrestrial and aquatic pests, weeds, and diseases that threaten the country’s ecosystems and economy (Goldson et al. 2015). This pressure resulted in a visionary proposal to eliminate introduced rodents, mustelids and the common brushtail possum from New Zealand (Russell et al. 2015). This proposal eventuated in the establishment by the central government of Predator Free 2050 Ltd, a charity to direct Crown investment into an overarching goal: ridding forests of the devastating impacts of stoats, rats, and possums by 2050 (Department of Conservation 2021; Predator Free 2050 Ltd 2021).

The potential for people in urban areas to contribute to rat eradication on a large-scale is of national interest in New Zealand as a component in achieving the Predator Free 2050 vision (Russell et al. 2015; Peltzer et al. 2019). However, realising this potential means engaging and co-ordinating hundreds (if not thousands) of households (and businesses) in trapping and poisoning rats. A range of policy measures, including marketing, education, incentives, charges and regulations can be used to modify people’s behaviour and practice. For example, participation in an urban programme of rat trapping could be encouraged by offering incentives to households to install and monitor traps.

Choosing which policy measure to deploy depends on several factors, with the likelihood of householders responding favourably to the measure being, perhaps, the most critical. For example, incentives could be popular among householders but prohibitively expensive given the trapping densities that may be required. Regulations compelling the installation of traps could have the potential to change the behaviour of all households but may be unpopular among householders and monitoring and maintenance of traps may be problematic to enforce. Hence, knowing the likely response of householders to any proposed policy measure is crucial when choosing between policy measures (and knowing when there may be merit in combining them).

This paper contributes to the emerging literature on urban predator control by applying a dual-process model of adoption (Bagozzi 2006a, b) that can be used in urban as well as rural settings to understand and predict people’s behaviour and infer their responses to different policy measures for eradicating or controlling pests. We used the model to explain and predict the trapping of rats by urban households in New Plymouth, a large provincial city on the North Island of New Zealand. A programme to eliminate rats from the city is underway as part of the Predator-Free 2050 project (Taranaki Regional Council 2021).

Adaptable predators such as rats flourish in urban environments and there is an established literature on the ecology of common urban pests such as rats (Feng and Himsworth 2013; Himsworth et al. 2014; Youngsteadt et al. 2014; Parsons et al. 2017). There has been research on the potential public health (Wilson et al. 2018) and economic (Almeida et al. 2013) benefits of eradicating rats from urban environments, but the attitudes of urban residents towards participating in urban pest control or eradication appears to have received limited attention (German and Latkin 2016). Woolley et al. (2021) is an exception. To the best of our knowledge, this research is the first to investigate the attitudes, motivations and trapping behaviour of urban households using the dual-process model of adoption.

Theory

The dual-process model proposed by Bagozzi (2006a, b) builds on the extensive literature on intention to adopt (Ram and Sheth 1989; Bagozzi and Lee 1999; Oreg, 2003; Bagozzi and Warshaw 1990; Bagozzi 1992). Bagozzi and Lee (1999) observed that actual adoption involves both the intention to adopt and the translation of that intention into behaviour, which may not occur. The concept of ‘goal striving’ was developed to link intention with behaviour (Bagozzi and Dholakia 1999; Bagozzi and Lee 1999; Bagozzi 2007). Consequently, the dual-process model of consumer response to innovations has two components: goal setting, which describes the process of deciding to adopt; and goal striving, which describes the process of adopting. The goal setting process provides a foundation for identifying when motivation, and the factors that influence motivation, delay adoption. The goal striving process provides a foundation for identifying when it is the implementation of the decision to adopt, and the associated factors that influence implementation, that delay adoption.

This distinction is important when it comes to choosing policy measures to facilitate adoption. If motivation is absent, policy measures must be designed to create motivation; policy measures that simply remove apparent barriers will fail to have an effect. On the other hand, if motivation is present but there are barriers to the translation of intention into action, measures to remove barriers should be effective.

In the following, the theory, and its implication for hypotheses in the current context, are presented.

Goal setting

In the dual-process model (see Fig. 1) the first process is goal setting, and this process is triggered by awareness of an opportunity to achieve a goal (Bagozzi 2006a, b). This process determines the degree of interest the decision-maker has in achieving a goal, that is, ‘goal desire’ (Wright 2011). Insufficient interest halts any move to the conscious formation and use of attitudes and norms. Goal desire determines whether a goal will be adopted.

Bagozzi (2006a, b) proposes that five elements contribute to creating goal desire process. Two of these are ‘positive’ and ‘negative’ ‘anticipated emotions’. These emotions result from imagining success and failure, respectively, in goal attainment and their personal emotional consequences. These emotions could include happiness, excitement and pride or disappointment, anger and sadness. The likelihood of success or failure is not considered with anticipated emotions.

Another two elements that contribute to creating goal desire process are ‘anticipatory emotions’. These are emotional responses to the prospect of a future event. The emotions involved are hope and fear and depend in part on the perceived probability of an event occurring, that is, success or failure (Wright 2011). The final element in the process is ‘affect towards the means’ of striving for the goal. This is the personal emotional appeal of the methods, processes, actions and so on believed to be required to pursue the goal (Bagozzi 2006a, b). In sum, the desirability of a goal is determined by the perceived attractiveness of likely short-term and long-term consequences of attaining the goal, and judgements about consequences arise from personal evaluations, evaluations by significant others, and the emotional experience associated with moving towards the goal (Oettingen and Gollwitzer 2001).

Fig. 1
figure 1

Simplified representation of the dual-process model. Notes: Plain text indicate key variables used in the study. Source: Adapted from Bagozzi (2006a, b)

Full size image

Fundamentally, goal desire arises from the alignment of the goal with the functional, experiential and self-expressive needs of the decision-maker (Oettingen and Gollwitzer 2001; Ryan et al. 1996). These needs underlie anticipated emotions and, in conjunction with consideration of the magnitude of consequences and risks, underlie anticipatory emotions.

In other words, the personal importance of a goal will be judged on its implications for satisfying functional, experiential and self-expressive needs (Oliver 1997; Assael 1989; Broderick 2007).

In marketing terms, the more personally important a goal is, the greater will be personal ‘involvement’ with the goal (Assael 1998; Stankevich 2017) and the greater the time and effort that will be devoted to searching for information in relation to the goal, evaluating that information, and making decisions about pursuing the goal (Celsi and Olson 1988; Poiesz and Bont 1995). Consequently, involvement with a goal can be interpreted as a measure of goal desire. If involvement with a goal is weak, goal desire is weak, and the goal setting process halts. The matter of how to achieve the goal is not considered. The rubicon is not crossed (Oettingen and Gollwitzer 2001). The failure to act arises from a lack of involvement with, of interest in, the goal.

Involvement has been employed to understand people’s responses to biosecurity issues (Bewsell et al. 2012), farmers’ approaches to nutrient budgeting (Bewsell and Brown 2011) and adoption of agricultural technologies and practices (Kaine and Wright 2022), campers’ firewood collection behavior (Daigle et al. 2018), the adoption of measures to prevent the spread of COVID-19 (Kaine et al. 2022; Kaine and Wright 2023) and policy design (Howlett 2018).

The strength of goal desire, together with attitude towards the goal and subjective norms about the goal, determines goal intention, the commitment to act to achieve the goal. In the context we are considering here, this means that the intention to reduce the population of rats (goal intention) will be influenced by involvement with, attitudes towards, and subjective norms about, reducing rat populations. This yields:

Hypothesis 1. The intention to reduce the population of rats (goal intention) will be influenced by involvement with, attitudes towards, and subjective norms about, reducing rat populations.

The commitment to a goal, goal intention, must then be translated into a set of specific behaviours to be implemented, which is ‘behavioural desire’. The factors that moderate the translation of goal intention into behavioural desire are those identified in mainstream models of consumer behaviour, such as attitudes, outcome expectancies, and subjective norms (Fishbein and Ajzen 1975; Ajzen 2001, 2002). The strength of behavioural desire, like the strength of goal desire, will depend on the alignment of the behaviour with the functional, experiential and self-expressive needs of the decision-maker. Consequently, involvement with the behaviour can be interpreted as a measure of the strength of behaviour desire. If involvement with the behaviour is weak, the goal setting process halts; the matter of how to implement the behaviour is not considered. The failure to act arises from a lack of interest in the behaviour. For example, a strong preference for using baits to control rats rather than traps, or perceptions that traps are dangerous or ineffective, will lead to the abandonment of trapping as a behavioural option.

Behavioural desire is then translated into specific ‘behavioural intentions’ which may be moderated by perceptions of behavioural control such as ‘self-efficacy’ (Bandura 1997). Consequently, the strength of behavioural desire, together with attitude towards the behaviour and subjective norms about the behaviour, determine behavioural intention, which is the commitment to enact the behaviour. In the context we are considering here this means that the intention to trap rats (behavioural intention) will be influenced by involvement with, attitudes towards, and subjective norms about, trapping rats.

Goal striving

As actual and intended behaviour are not always highly correlated (Bagozzi and Lee 1999), the factors that influence the correlation between intended and actual behaviour are considered explicitly in the goal striving component of the dual-process model. Consideration of these factors is particularly important both in forecasting rates of adoption of the behaviour and in highlighting what opportunities, if any, there may be to influence the translation of intention into action.

The first stage in the goal striving component is the choice of how the behavioural intention will be fulfilled. Alternative means by which this may be done (e.g., which type of trap to use) are evaluated in terms, particularly, of self-efficacy, ‘outcome expectancy’ and affect, which is like or dislike of a means. These elements of appraisal need to be integrated to make a choice (Wright 2011). For example, implementing the intention to trap rats will be evaluated in terms of perceptions of: the personal possession of the skills and effort required to trap (self-efficacy); the usefulness of trapping in reducing the number of rats (outcome expectancy); and the emotions that are expected to be experienced when trapping (affect towards means).

It is in this stage that barriers to adoption may appear. For example, a perceived absence of trapping skills, or time constraints, may, by engendering feelings of low self-efficacy, result in a decision to abandon trapping. These are barriers to implementing the behavioural intention.

The second stage is ‘action planning’. Action planning ‘involves decisions as to when, where, how and how long, to act’ (Wright 2011, p18). In this stage, situational cues for the timing of specific actions are considered, such as when and where to place traps (Stronge and Kaine 2020). The third stage in goal striving is the implementation of the plan, which is the commencement of action in pursuit of the goal.

The fourth stage of goal striving consists of the control processes exercised over the planned actions such as tracking progress, identifying opportunities and hindrances, and revising plans accordingly. Appraisals of progress will lead to affective responses which may influence the continuation of striving. For example, positive affect will evoke an intention to stay the course. A negative affect may evoke greater effort; alternatively, it may result in changes in goals, a redefinition of success, or a judgement of failure and abandonment of goal striving (Bagozzi 2006b). The final stage is the outcome: adoption, trial or failure to adopt, which will generate emotions. The decision to persist with or abandon trapping will depend, then, on revisions, in the light of experience, to evaluations of self-efficacy, the outcomes experienced and affect towards means.

Consequently, the translation of goal intention to an implemented behaviour (via behavioural desire) will be influenced by the strength of behavioural desire together with the attitude towards, and social norms about, the behaviour. In addition, the translation of goal intention into actual behaviour will also be influenced by preferences regarding alternative behaviours, and factors that influence goal striving such as outcome expectancy and affect. Note that the anticipated and experienced outcomes and emotions of trapping will supersede (by reinforcing or contradicting) attitudes and social norms towards trapping rats formed at the behavioural intention stage. Consequently, attitudes and norms formed in the behavioural intention stage become redundant once the behaviour is implemented.

In the context we are considering here, this means that the trapping of rats (behaviour) will be influenced by involvement with trapping rats (strength of behavioural desire), together with time constraints and preferences regarding baits and commercial pest services (all of which influence the translation of goal intention into behavioural intention). Also exerting an influence will be the anticipated and experienced outcomes, and the anticipated and experienced emotions, arising from trapping rats (goal striving). This leads to:

Hypothesis 2. The trapping of rats (behaviour) will be influenced by involvement with trapping rats (strength of behavioural desire), time constraints, preferences regarding baits and commercial pest services, and anticipated and experienced outcomes and experienced emotions (goal striving).

In principle, behaviour can also be explained by commencing with goal intention and moving through behavioural desire and behavioural intention to finish with goal striving. This means that the trapping of rats (behaviour) should also be predicted by goal intention and goal striving, a function of perceptions of self-efficacy and the anticipated or experienced outcomes and emotions of trapping rats. This leads to:

Hypothesis 3. The trapping of rats (behaviour) will be influenced by goal intention, time constraints, preferences regarding baits and commercial pest services, and anticipated and experienced outcomes and experienced emotions (goal striving).

Case study

Background

New Plymouth is located on the west coast of New Zealand’s North Island and has a population of approximately 86,000 (NZ Stats 2021). Predator Free 2050 funded research to provide insights into the popularity, or otherwise, of a program to promote urban trapping in New Plymouth (Taranaki Regional Council 2021). As part of this project, a survey of New Plymouth residents was undertaken (Kaine and Stronge 2020). The hypotheses put forward in the preceding section were tested using some of the data collected in this survey. Note that the theoretical basis for the survey was the I3 Response Framework (Kaine et al. 2010). This limited the comprehensiveness of our analysis as data on some key variables, such as goal intention, were not collected. The questionnaire used in the survey is described below.

Methods

The questionnaire contained two sets of scales to elicit householders’ involvement and attitudes in relation to reducing rat numbers and trapping rats (see Supplement A). The first set of scales measured their involvement with the idea of reducing rat numbers and their involvement with the idea of trapping rats. Involvement was measured using a condensed version of the Laurent and Kapferer (1985) involvement scale, with respondents rating statements for each of the five components of involvement: functional, experiential, self-expressive, consequence and risk involvement (see Table 1 for an example).

Similar statements were formulated for involvement with trapping rats. Respondents rated their agreement with each statements using a five-point Likert scale ranging from strongly disagree (1) to strongly agree (5). Involvement scores were interpreted as low (1-–2), mild (2-–3), moderate (3-–4) and high (4-–5).

The second set of scales measured attitudes, and attitude strength, towards trapping rats. Attitudes were measured using a simple, evaluative Likert scale. The strength of respondents’ attitudes to rat trapping was expected to vary depending on the strength of their involvement with trapping. Attitude strength was measured as the mean absolute deviation from the mid-point of the scale.

The questionnaire included three questions relating to the intentions of householders to reduce rat numbers. These were:

  • I feel some responsibility for reducing the number of rats in my area.

  • I am prepared to take some kind of action to reduce rat numbers in my area.

  • I am prepared to make sacrifices to reduce the number of rats in my area.

  • It’s important to work together to reduce the number of rats.

In addition, there were four questions that sought to establish the subjective norms of householders with respect to reducing rat numbers. These were:

  • Most people I know feel some responsibility for reducing the number of rats.

  • I think nearly everyone is prepared to take some kind of action to reduce rat numbers.

  • Most people are prepared to make sacrifices to reduce rat numbers.

  • Most people know we have to work together to reduce the number of rats in our area.

The questionnaire included a series of questions formulated to discover respondents’ beliefs about the advantages and disadvantages of reducing rat numbers, and of trapping to achieve this. Respondents rated their agreement with the belief statements using a five-point Likert scale ranging from strongly disagree (1) to strongly agree (5).

Table 1 Involvement with reducing rat numbers
Full size table

Information was sought on whether respondents trapped rats, and their experiences if they had. Respondents who did not trap rats were asked about their reasons for not doing so (see Table 2). Respondents’ answers were averaged to create variables representing respondents’ anticipated outcome expectancies (for those that did not trap) or their experienced outcomes (for those that did trap) and their affect towards means.

The ordering of the statements in the involvement, attitude, and belief scales was randomised to avoid bias in responses. A series of questions was included concerning respondents’ age, gender, education, income, residence, and location. The questionnaire was approved for distribution by Manaaki Whenua – Landcare Research’s social ethics process (application 1920/10).

The questionnaire was administered online and by telephone by a New Zealand market research company. Telephone respondents were randomly selected from a database of urban addresses in New Plymouth. Internet respondents were randomly selected from a database of panellists with urban addresses in New Plymouth. Unlike telephone respondents, internet respondents receive compensation from the panel owner for competing surveys. They also have greater flexibility with respect to when they participate. Consequently, we expected respondents who completed the questionnaire via the telephone to exhibit higher involvement with reducing rats and, possibly, trapping than respondents who completed an online questionnaire, as the former would be more likely to be motivated by an intrinsic interest to participate. The survey was open for approximately 10 weeks beginning in December 2019, and we received 436 responses. A complete report on the sample and our results can be found in Kaine and Stronge (2020).

Hypothesis 1

regarding goal intention was tested using linear regression analysis while hypotheses 2 and 3 regarding trapping behaviour were tested using binomial logistic regression.

Results

Approximately 61 per cent of respondents were women. The age distribution of the sample was marginally older than census estimates for New Plymouth and the sample had a higher average level of education than census estimates for New Plymouth. The overwhelming majority of respondents lived in a house (71 per cent) with most of the remaining respondents (17 per cent) living in apartments, townhouses, or units. A small proportion of respondents (12 per cent) lived on farmlets or lifestyle blocks bordering the city.

Statistical tests, i.e. Cronbach’s alpha (Carmines and Zeller 1979), indicated that the involvement scales were reliable; that is, they were internally consistent in the sense that scores on related statements were highly correlated with each other. Cronbach’s alpha was 0.79 for involvement with reducing rat numbers and 0.78 for involvement with trapping rats.

Table 2 Real and imagined experience with trapping
Full size table

Involvement with reducing rat numbers and with using traps to reduce rat numbers were not related to the age, gender or property type of respondents. There was a statistically significant, but inconsequential, association between level of education and involvement with trapping. There was no association between level of education and involvement with reducing rat numbers. Respondents who completed the questionnaire via the telephone exhibited marginally higher involvement with reducing rats, but not with trapping, than did respondents who were registered members of a market survey panel. Attitudes towards rats and trapping were unrelated to demographic characteristics.

Beliefs, attitudes and involvement

We confirmed that attitudes are strongly, and plausibly, associated with beliefs by estimating two regressions using beliefs about the effects of rats, and beliefs about the advantages and disadvantages of traps, to predict attitudes towards reducing numbers of rats and using traps to do this, respectively. The results were satisfactory with beliefs explaining approximately 33% of the variance in respondents’ attitudes.

Involvement with, and attitudes towards, reducing numbers of rats were primarily influenced by concerns about the effects of rats on the environment and human health, together with concerns about the ethics of killing any animal (see Table 3). As expected, we found that respondents’ involvement with the goal of reducing rat numbers was strongly influenced by the strength of their goal desire as measured by involvement (see Table 4). Attitudes towards using traps to reduce rat numbers were influenced by concerns about the relative effectiveness of traps and poison baits, concerns that traps are a risk to people’s health and a danger to native birds, and that traps are a cruel and inhumane way to kill animals (see Table 4).

As expected, we found that the strength of respondents’ attitudes towards the goal of reducing rat numbers was influenced by the strength of their goal desire as measured by involvement (see Table 5). The strength of their attitudes towards using traps to reduce rat numbers was influenced by the strength of their attitude towards the goal and the strength of their behavioural desire (see Table 5).

Table 3 Beliefs, attitudes and involvement regarding rats
Full size table
Table 4 Beliefs, attitudes and involvement regarding trapping
Full size table
Table 5 Attitude strength and involvement
Full size table

Predicting goal intention

Following Bagozzi (2006a, b), the strength of goal desire, together with attitude towards the goal and subjective norms about the goal, determine goal intention, the commitment to act to achieve the goal. Consequently, we hypothesised that the intention to reduce the population of rats (goal intention) would be influenced by involvement with, attitudes towards, and subjective norms about, reducing rat populations (Hypothesis 1). The regression results were consistent with this hypothesis (see Table 6).

Predicting behaviour

We expected (Hypothesis 2) that the actual trapping of rats would be influenced by involvement with trapping rats, perceived obstacles to trapping and the anticipated or experienced outcomes and emotions from trapping rats. Since the anticipated and experienced outcomes and emotions of trapping will supersede attitudes towards trapping rats formed at the behavioural intention stage, attitude towards trapping was not included in the regression. The regression results were consistent with this hypothesis (see Table 7) apart from the estimated coefficient for involvement with trapping rats being less than, instead of greater than, one.

The estimated likelihood ratios indicate that trapping behaviour is highly sensitive to the anticipated outcome expectancies (for those who don’t trap) and experienced outcomes (for those who do trap). Trapping behaviour is also sensitive to involvement with trapping (reflecting the strength of behavioural desire).

We also hypothesised (Hypothesis 3) that the actual trapping of rats would be influenced by the strength of the intention to reduce the number of rats (goal intention), self-efficacy and the anticipated or experienced outcomes of trapping rats (goal striving). The regression results were consistent with this hypothesis (see Table 7) with respect to preparedness to take responsibility and intention to act. However, the estimated parameters for willingness to make sacrifices and the importance of working with others were not statistically significant in their respective regressions. These results suggest that (a) a preparedness to make sacrifices to reduce rat numbers does not extend to trapping rats and (b) that householders will trap rats irrespective of the importance they attach to the need to work together to reduce rat numbers.

Again, the estimated likelihood ratios indicate that trapping behaviour is highly sensitive to the anticipated outcome expectancies (for those who don’t trap) and experienced outcomes (for those who do trap). Trapping behaviour is also sensitive to the strength of the intention to reduce the number of rats (strength of goal intention).

Table 6 The influence of involvement, attitudes and subjective norms on goal intentions
Full size table
Table 7 The influence of behavioural and goal intentions on trapping behaviour
Full size table

The estimated likelihood ratios confirm that trapping behaviour is highly sensitive to the anticipated outcomes (for those who don’t trap) and experienced outcomes (for those who do trap). Trapping behaviour is also sensitive to feeling some responsibility for reducing rat numbers and being willing to act to reduce rat numbers (strength of goal desire), as well as to the emotional rewards from trapping (anticipated or experienced) and perceptions of the time and effort entailed in trapping. Overall, the results indicate that motivational strength with respect to goal desire plays an important role in determining whether householders trap rats or not.

The estimated probability of trapping rats for each household in the sample (using the coefficient estimates in the second column of Table 7) is graphed in Figs. 2, 3 and 4 together with goal desire, behavioural desire and anticipated and experienced outcomes, respectively. Households were ordered by the estimated probability of trapping rats.

Comparison of the figures clearly reveals a weaker association between the estimated probability of trapping and both goal desire (Fig. 2) and behavioural desire (Fig. 3) compared to anticipated and experienced outcomes (Fig. 4). This indicates first that, although most householders in our sample had moderate-to-high goal desire, this did not translate directly into trapping behaviour. Second, even though most householders in our sample had moderate behavioural desire, this did not translate strongly into trapping behaviour. Third, anticipated and experienced outcomes, a key component of goal striving, strongly influenced householders’ probability of trapping rats.

Discussion

Following Bagozzi (2006a, b), we hypothesised that the strength of goal desire, together with attitude towards the goal and subjective norms about the goal, determine goal intention, the commitment to act to achieve the goal. This hypothesis was supported. We also hypothesised that the translation of goal intention through to an implemented behaviour (via behavioural desire) would be influenced by factors that influence goal striving such as outcome expectancy and affect. We found this hypothesis was supported in the context of householders’ desire to reduce rat numbers and their adoption of trapping to achieve this goal.

Motivation and barriers to adoption

This latter finding is important as the model of behaviour adoption proposed by Bagozzi (2006a, b) has profound implications for those who seek to influence the behaviour of the public to achieve policy objectives. The model provides a sound, theory-based method for distinguishing between the failure of members of the public to adopt a behaviour desired by policy makers because of an absence of motivation, and the failure of members of the public to adopt it because of the presence of a barrier to adoption. The former represents differences in the goals and preferences between the public and the policy maker.

A barrier is an obstacle that prevents progress; a barrier implies the presence of motivation. This means that the idea that there are ‘barriers’ to people trapping rats must be treated cautiously. The term ‘barrier’ implies there is an intention to trap rats but an obstacle hampering the actioning of that intention. In terms of social psychology, a factor can be said to be a barrier when it prevents goal desire, manifest as a behavioural intention, from becoming actual behaviour (Bagozzi 2006a, b). Logically, the barrier must be an influential component of goal striving (Bagozzi 2006a, b).

In contrast, a lack of intention should not be interpreted as a ‘barrier’ (Woolley et al. 2021). This situation is qualitatively different from one in which the behavioural intentions of the individual do align with the policy outcome, but some obstacle is preventing the individual from acting on those intentions. The latter implies that behaviour will align with the policy outcome once the obstacle is removed. The former is a more profound problem for the policy maker; it requires either:

  • reconfiguring the goals or behavioural intentions of the individual to better align with the policy outcome or measure,

  • revising the policy outcome or measure to better align with the needs of the individual, or.

  • accepting limits on achieving the policy outcome.

To describe differences in desired goals as barriers to adopting the behaviour desired by a policy maker is problematic and unhelpful. Consider, for example, the magnitude of the challenges facing a promotional campaign to promote rat trapping by changing householders’ beliefs about, and attitude towards, trapping rats; that is, a campaign to reconfigure householders’ behavioural intentions to better align with the policy measure.

Changing motivation

In principle, knowing the reasons why householders want to reduce rat numbers provides a foundation for influencing their willingness to trap rats and participate in a rat trapping programme. For example, householders’ attitude towards reducing rat numbers in New Plymouth was primarily motivated by concerns for the potential harms rats pose for biodiversity and the environment, and for the health of themselves and their families. Consequently, a promotional campaign to promote trapping might focus on promoting the potential of urban trapping to reduce these harms. Promotional campaigns are regularly suggested in the literature (Gramza et al. 2016; Klapwijk et al. 2016; Niemiec et al. 2016; Novoa et al. 2017). Given that most respondents had moderate-to-high involvement with reducing rat numbers, most householders may well notice, and pay attention to, such a campaign. However, the impact of such a campaign will be somewhat limited unless (1) it substantially increases goal desire in relation to reducing rat numbers and (2) any increase in goal desire translates into an increase in behavioural desire and behavioural intention with respect to trapping rats.

Fig. 2
figure 2

Goal desire (involvement with reducing rat numbers) and probability of trapping

Full size image
Fig. 3
figure 3

Behavioural desire (involvement with trapping) and probability of trapping

Full size image
Fig. 4
figure 4

Anticipated and experienced outcomes and probability of trapping

Full size image

Unfortunately, respondents’ attitudes towards using traps were strongly influenced by perceptions that trapping is a danger to native birds and that traps are a cruel and inhumane way to kill animals. This meant that, at least for some respondents who did not trap rats, that strong goal desire did not result in strong behavioural desire. In fact, the opposite is the case. Hence, an increase in goal desire need not automatically translate into an increase in behavioural desire and may evoke counter-productive, approach-avoidance goal conflict behaviour with respect to trapping rats among those who don’t trap rats by reminding them of their aversion to the use of traps (Elliot 1999; Townsend and Busemeyer 1989; Dunn et al. 2018; Kelly et al. 2018). That is, promoting greater goal desire simultaneously draws attention, through goal striving processes, to trapping such that the attractiveness of eliminating rats is balanced by the unattractiveness of the method of doing so, leading to sustained indecision. This raises doubts about the effectiveness of promotional efforts to reconfigure householders’ goals and strengthen their motivation to reduce rat numbers (goal desire) to better align with the policy measure of trapping rats. Approach-avoidance conflict may also explain why greater involvement with rat trapping reduced the likelihood of trapping, after allowing for the effects of anticipated and experienced outcomes and affect towards means (see Table 7).

Our results indicated that most respondents had moderate involvement with using traps to reduce rat numbers. In other words, most respondents had a moderate level of behavioural desire with respect to trapping rats. The implication is that most householders were motivated, but not strongly motivated, to trap rats. As a consequence, the translation of their behavioural desire into a behavioural intention will be sensitive to factors such as anticipated outcome expectancies and affect towards means. This suggests that efforts to encourage more favourable outcome expectancies regarding rat trapping by, for example, offering convincing evidence that trapping is a practical, useful and helpful way of making a difference to the environment, are likely to meet with success in the long term. This is an example of modifying beliefs to increase the conversion of goal desire into behavioural desire.

Relatedly, as most respondents had moderate involvement with using traps to reduce rat numbers, a promotional campaign seeking to promote trapping by counteracting perceptions that trapping is a danger to native birds and that traps are a cruel and inhumane way to kill animals is likely to engage most householders (Priluck and Till 2004). Such a campaign might emphasise the safety of traps, and the speed and efficacy with which they function. This is another example of changing beliefs to improve the conversion of goal desire into behavioural desire.

Barriers to adoption of rat trapping

The principal barrier to the adoption of rat trapping that we identified was the perceived time and effort required to trap rats. Given that most respondents had moderate involvement with using traps to reduce rat numbers, they will also be sensitive to any perceived barriers to trapping and, further, the removal of any perceived barriers to trapping is unlikely to prompt a substantial proportion of respondents to commence rat trapping themselves, in the absence of a favourable re-evaluation of outcome expectancies. Relatedly, the same reasoning leads to the conclusion that offering small incentives (e.g., free traps) to encourage trapping is unlikely to be effective in increasing the adoption of rat trapping.

As most respondents were moderately-to-strongly involved with the idea of reducing rat numbers, and given an apparent absence of strong opposition to the use of traps to reduce rat numbers, we conclude that most respondents in our sample were likely to support, at least in principle, a trapping programme by permitting the installation of traps on their properties which could be serviced by programme volunteers. This is consistent with the views of householders obtained in interviews (Stronge and Kaine 2020) and experience from apparently similar urban predator control in Wellington (Predator Free Wellington 2019). Given that most respondents exhibited only moderate involvement with trapping, participation in such a programme should be made as simple and easy as possible. This amounts to revising the policy measure to better fit the needs of the community by reducing the time and effort householders themselves need to devote to using traps to reduce rat numbers.

The time and resources required to implement such a measure, which would have impacts immediately, may compare favourably to the time and resources required to change householders’ perceptions about the expected outcomes of householders trapping rats, and which would be unlikely to have any immediate impact.

Our results illustrate the importance of considering both attitude and motivational strength when designing policy measures that rely on public participation. Typically, favourable attitudes towards an outcome or action are taken to mean (often implicitly) a high likelihood that members of the public will pay attention to promotional efforts about the outcome or action, and (more explicitly) a high likelihood that the outcome will be supported, or the action taken. Such assumptions are embedded in traditional behavioural models based on attitudes as exemplified by Rogers (1975), Ajzen and Fishbein (1977), Bandura (1997), Petty and Cacioppo (1979), Janz and Backer (1984), Schwarzer (1992), Witte (1992), and others. Unless a favourable attitude is accompanied by at least moderate goal and behavioural desire, the propensity to act is likely to be limited.

In the context we have investigated here, discriminating between involvement (i.e., goal desire and behavioural desire) and attitude provides insights into:

  • Distinguishing between householders that will participate in a trapping programme by themselves trapping and householders that would participate by allowing trapping to occur on their properties. This has implications for the design of the programme and the allocation of programme resources.

  • Assessing the ease, or otherwise, of co-ordinating trapping across many households. If the involvement of householders with trapping is primarily mild-to-moderate the time and effort they will actually devote to trapping will be limited, which has implications for when they will trap and how frequently they will set, check, and maintain traps. Reliably achieving consistency and co-ordination in trapping in these circumstances is problematic and suggests the programme may have a higher chance of success by relying on professionals and mobilising a workforce of trained volunteers.

Our results are based on the analysis of data that was collected by Kaine and Stronge (2020) who were testing a framework proposed by Kaine et al. (2010), not the dual-process model proposed by Bagozzi (2006a, b). Consequently, although the dual-process model can be viewed as providing the theoretical foundation for Kaine et al. (2010), our application of it has been limited because data on some variables in the dual-process model such as anticipated and anticipatory emotions, affect towards means, social norms regarding trapping and behavioural intentions regarding trapping, were not collected by Kaine and Stronge (2020).

Conclusion

The model proposed by Bagozzi (2006a, b) provides a systematic basis for government agencies to develop a mix of strategies that target relevant differences in the determinants of the propensity of people to change their behaviour in response to a policy measure. By application in a predator-control case study, we have shown the model has merit in predicting the behavioural intentions of people and that it can be employed to help identify nuanced strategies to promote the adoption of desired behaviour. Hence, the model may provide a basis for targeting policy resources, thereby reducing the risk of over-investing in policy measures that are likely to have little impact or under-investing in alternative measures that are likely to have greater impact.