1. Introduction
Understanding the evaluative processes in consumers’ decisions to sample a novel food has received a great deal of attention in recent times [
1,
2,
3]. Increased use of novel technology in food production, increased globalization and exposure to new markets for consumers, and a need to develop foods with greater environmental sustainability provides a potentially increased market for novel food products. Understanding consumer perceptions and attitudes toward novel foods have become an integral component of the development process for these products to succeed in the market [
4,
5]. There are several standardized psychometric instruments that evaluate discrete factors of novel foods, general food choice, and personal attributes which might influence novel food acceptance but there is not currently a standardized instrument that allows evaluation of intention to consume foods that applies broadly across all types of novel foods. A common framework employed in the evaluation of discrete novel foods is the Theory of Planned Behaviour [
6,
7]. Consistent with other studies [
8,
9] the current research has utilized a two-stage approach to item development and data collection under the TPB framework to evaluate consumers’ intention to consume hemp food.
Hemp food produced from the seeds of the plant
cannabis sativa is a recent introduction to the Australian market, having passed legislative requirements for approval as a novel food in November 2017 [
10]. Hemp food was previously illegal in Australia due to a perception of issues related to controlling product levels of tetrahydrocannabinol (THC), the psychoactive property of cannabis. As a consequence, the approval of hemp food now includes a requirement that all production, importing, and manufacture of hemp food products undergo a stringent compliance process aimed at keeping THC levels well below prescribed percentages that would potentially induce a psychoactive effect in the consumer [
11]. Research of the effects of THC in hemp food at designated percentages (0.001% legal limit and twice this concentration) has determined that “consumption of low-content THC oil does not result in positive biological assessments” [
12] (p. 101). Hemp food is marketed to consumers as having several health benefits and promoted to producers as an environmentally sustainable and high-yielding crop. Despite the purported benefits, however, hemp food does not seem to have infiltrated the Australian market to a degree that might be expected [
13]. Hemp food presents as a unique type of novel food, with many associations to other uses of cannabis from building products and textiles to medicinal products and illicit uses [
14]. Evaluating the impact these associations might have on the acceptance and intention to consume hemp food has the potential to not only assist the Australian hemp food industry to better understand consumer perceptions of their products but may also provide a protocol for evaluation of novel foods to be developed in the future. As innovation in food production increases to meet the demands of sustainability and increasing populations, early evaluation of potential novel foods is essential.
1.1. Theoretical Background
The theory of planned behavior posits that intention to perform a behavior is correlated to actual behavior in an extension to the theory of reasoned action [
6] where behavioral and normative beliefs are reported to contribute to attitudes toward a behavior, and subjective beliefs form through perceptions of societal norms regarding that behavior Cognitive evaluations of these attitudes and beliefs contribute to consumers’ intention to perform the behavior themselves. Expanding on this, TPB includes volitional control as an additional consideration, reporting that people must have a perception of control over the behavior if the intention is to be realized as actual behavior [
7]. Several hundred studies have applied TPB within a variety of disciplines with accuracy [
7] and the utility of the theory in the evaluation of novel food acceptance has been supported by experimental validation of the intention–behavior relationship [
1,
8,
15]. A review of 42 studies employing TPB as a measure of intent to consume a discrete food while evaluating the intention–behavior relationship reported a strong association overall (
r = 0.45) between intention to consume and actual consumer behavior.
1.2. Methodology
There is currently no standardized psychometric instrument available that is appropriate for capturing the unique associations attributed to a discrete novel food within a TPB framework. Consistent with previous novel food research this study has developed a fit-for-purpose scale for use in the study. Examination of currently available literature did not discover adequate research on consumer perceptions and attitudes toward hemp food to enable the development of the TPB items without consumer consultation. As a result, a mixed-method (qual-QUANT) approach as described by Teddlie and Tashakkori [
16] was adopted for the project which was conducted under a pragmatic framework. While pragmatism is more of a philosophical framework than a methodology, it allows for the methodology appropriate to each phase to be applied and a re-evaluation of the findings from each to form meta-inferences beyond those yielded from the findings within each phase. The first phase consisted of semi-structured interviews which informed the development of TPB items, and the second phase employed the TPB items in a questionnaire that was disseminated online nationally to evaluate consumer acceptance of hemp food in the Australian market.
4. Discussion
The use of TPB as a theoretical framework for the evaluation of acceptance of a novel food is not a new concept, however, evaluating the behavioral beliefs that form attitudes toward the food as discrete positive and negative entities is a new approach. Negative attitudes toward factors related to a novel food have been previously reported as significant to their acceptance [
2,
8,
20]. For example, [
2] included negative attitude as a discrete construct in their study of acceptance of insects as a novel food. The current study accepted empirical evidence of the importance of negative factors in the evaluation of novel foods and developed discrete positive and negative attitudinal constructs from qualitative surveys of consumer behavioral beliefs. This approach proved beneficial to understanding the acceptance of hemp food as the negative attitude items while loading onto a single factor and correlating positively with each other, had differential effects on the construct associated with the psychological benefits of Cannabidiol (CBD) oil. Originally developed as a component of positive attitudes, anomalies detected early in the analysis of the data led to the construct defined as relating to CBD being identified as a potential mediator between negative attitudes and intention to consume under the TPB framework. While the negative association of the CBD constructs with three of the items was theoretically and statistically supported, the positive association with two of the items (NA1: Eating hemp food might result in a positive roadside drug test; NA5: People don’t know enough about the effects of eating hemp food) was counterintuitive and is discussed below.
In the qualitative phase of the study, a perception of an association between hemp food and CBD oil provided an impetus for the development of the items relating to the psychological benefits of CBD oil. Reduced anxiety and increased relaxation were perceived by survey participants as a positive aspect of hemp food through this association. While these benefits are not promoted as applicable to hemp food, it is the perception of the benefits through association with CBD oil that contributes to a consumer’s intention to consume the food. What has been revealed through the application of SEM, however, is a differential influence of the perception of CBD on intention to consume hemp food, dependent upon which negative aspect of attitude is being addressed. On further examination, there appear to be three possible interpretations of the findings. The first is that despite the association of hemp food with CBD oil being perceived as a positive aspect of the product, there remains a belief that; a: There remains an association with both THC and CBD and a subsequent belief that potential for testing positive to THC after consumption of hemp food exists, and b: That not enough is known about the effects of eating hemp food. This association would be potentially problematic for the hemp food industry as it may suggest that despite assurances from industry and government sources that THC in hemp food is below detectable levels [
12], consumer acceptance of this assertion appears to be low. The overall mediating effect of the association of hemp food with CBD oil, however, has a positive effect on the intention to consume hemp foods and can therefore be viewed as benefitting hemp food acceptance. While this may indicate that an association with CDB oil has thus far influenced the acceptance of hemp food, the long-term outcome from such an association would not be beneficial. It has been demonstrated that knowledge of hemp food may be lacking within the marketplace but as consumer education increases and the distinction between CBD oil and hemp food becomes more apparent, any positive influence the misperception has had to that point may be reversed.
The second interpretation is that a perception of anxiolytic and relaxation benefits in hemp food is attributed to a direct association with THC and purported side effects of illicit and “medicinal” use of marijuana. However, while this would explain the relationship between each of the negative attitudes and the CBD construct, it would only account for the increase in intention to consume hemp food if consumers were hopeful of and actively seeking THC in hemp food. While this may potentially apply to a particular subset of the population it is less likely to be discerned in a sample that has been deemed representative of the Australian population.
The third interpretation may simply be that for at least some consumers, there is confusion surrounding the differences between CBD oil, THC, and hemp oil and food which results in the anomalies in the way that the CBD related survey items were interpreted by the consumer, the perception of it as a positive or negative aspect dependent upon whether the two entities are viewed as having different or similar properties. The more likely explanation might be somewhat a combination of the first and third scenarios and suggests the challenge for the hemp food industry to separate hemp food from any association with either CBD oil or THC, whether this is through increased consumer education or the development of strains of Cannabis sativa which are guaranteed completely void of THC. Knowledge of the processes of varietal development is beyond the scope of this research and the authors acknowledge that the latter may not even be possible.
A positive attitude did not predict the intention to consume hemp food. That is, despite a belief in the physical and psychological benefits of hemp food, a positive attitude toward hemp food did not play a significant role in intention to consume it. The factor which provided the greatest contribution to intention to consume hemp food was consumers’ normative beliefs. Subjective norms focused on consumers’ perceptions of what family, friends, peers and doctors would think of them consuming hemp food and whether they felt others like them would consume it as well. A high level of concern for the opinion of others and the mediating effect of the CBD construct on negative attitudes suggests that for this sample of the Australian population there remains a widespread stigma associated with hemp consumption. This stigma may be difficult to eradicate as it is the artifact of the previous illegal status of hemp food combined with decades of propaganda against cannabis use in any form [
14,
24]. Consumers may find it difficult to disassociate hemp food from the more illicit uses of cannabis and seek the approval of others for confirmation of the acceptability of consuming hemp products. The normative beliefs and the mediated influence of negative beliefs through association with CBD oil are the only contributing factors to intention to consume hemp food in this study.
A final analysis of the findings from across the two phases of this research consisted of a meta-inference of qualitative with the quantitative findings under a pragmatic framework. Here it was revealed that the application of the theory of planned behavior and SEM in the quantitative phase two of this study was consistent with the outcomes of the initial interviews. Survey participants in the qualitative phase of the current study generally held positive attitudes toward hemp food and indicated they were likely to consume them at some time in the future, however, they also made contradictory statements indicating their reluctance to consume them at a time when they had to drive a car or attend their workplace, for example, yet denied they believed they would test positive. The issue of an association of hemp food with CBD oil and THC is complex and may not be easily resolved.
Limitations and Future Research
This research evaluates the acceptance of hemp food in the Australian population and evaluates behavioral and normative beliefs within a population where consumption of cannabis products has been illegal until recent times. The attitudes and beliefs surrounding cannabis may not extrapolate to a population where cannabis is a legal entity, or where hemp foods have been available for longer periods of time. However, the protocol of assessing negative attitudes as a separate construct under TPB may have utility for future research of alternate novel foods where highly salient negative associations are made to the food from external sources.
The findings from the current study suggest some ambiguity may exist within the population between CBD oil and THC, the two more widely known properties of cannabis, and its association with hemp food, specifically hemp seed oil. Despite being highly conspicuous of hemp food, consumers may not be fully versed in their properties, effects, health benefits, etc., or aware of the differences between Cannabis sativa used in hemp food production, and Cannabis indica which is better known as marijuana. This may have contributed to the responses regarding the anxiolytic and relaxation effects of hemp food. Future research could focus on consumer understanding of hemp food’s specific properties and improve consumer education to address the identified lack of understanding. This may be of benefit to both the hemp food industry, and the medicinal cannabis quarter.
The findings also point to a potential for the development of a strain of cannabis fit for human consumption which is free from THC. While it is beyond the scope of this research and the competency of the researchers to suggest how or if this might be achieved, it would appear that innovation in food production is currently achieving goals never before imagined. Genetic modification, 3D printed foods, and foods produced using ultrasound or infusion heat treatment, for example, are highly advanced production methods for novel foods that may only be the tip of the innovation in the food production iceberg.