Final report for GNC21-337
Project Information
The emergence of niche crop markets and local food and drink value chains have offered small farmers an opportunity to diversify crop selection and increase profitability. However, burgeoning markets are subject to boom-and-bust cycles that can leave farmers without an end-buyer. The craft alcohol movement has become one of these budding markets, with the number of local breweries, cideries, distilleries, and wineries growing steadily over the past decade. With this, the North Central region has seen a boost in specialty crop production to supply locally-sourced inputs to these craft producers.
The 2018 Farm Bill also legalized industrialized hemp production, offering yet another specialty crop for local farmers to consider. Industrialized hemp has several end-uses including fiber and feed, but the use of particular relevance here is the commodity’s cannabidiol (CBD) content. While countless CBD products have entered commerce—from CBD-infused muscle rubs to lotions to edible food and beverage—little is known about the long-run potential for cannabis-infused products, leaving farmers wondering about the economic longevity of producing for the industry. Further, many states have cannabis with tetrahydrocannabinol (THC), which could serve as an economic complement or substitute for craft beverages. Loopholes in the 2018 Farm Bill also allowed states without recreational cannabis to produce and distribute certain types of THC-infused products. It is unknown to what extent this policy reform will impact craft beverage markets and stakeholders across alcohol and cannabis supply chains (from the farm to the end consumer).
This project analyzes consumer preference for alcohol and cannabis products to uncover the preferences for the two goods and identify the demand for cannabis-infused, non-alcoholic beverages. In achieving these objectives, we evaluate how these recent reforms will likely affect the alcohol supply chain and offer market insight to farmers and craft beverage producers on how to adapt to these evolving markets.
We use data from an online survey distributed to a representative panel of U.S. households to understand consumer preferences, attitudes, and consumption habits of alcohol and cannabis products. Examining these two markets within a single experiment allows us to identify how alcohol markets could change given cannabis reform. Moreover, we complement the primary experiment with secondary and tertiary experiments, enabling a deeper evaluation of these markets.
Some of the primary research findings are as follows:
- More than 50% of U.S. adults are willing to try hemp-derived beverages.
- The average willingness to pay for CBD- and THC-infused beverages varies by product type (e.g., sparkling water, coffee, tea) and cannabinoid (i.e., CBD and THC).
- Beer drinkers expect hemp-derived beverages to have similar safety and nutrition scores as traditional beers, but at higher prices.
- Some consumers view alcohol and cannabis products as complements, while others view them as independent markets.
- The hemp-derived beverage market remains a fraction of the size of the U.S. alcohol market, but sales projections are increasing as the market develops.
Output includes two peer-reviewed publications as well as one paper that is currently under review. Information has been disseminated through conference presentations, social media campaigns, and discussions with industry representatives. These efforts will continue as our final paper is accepted for publication. Ultimately, the hemp-derived beverage market is growing, and this project provides an important first step into a scientifically-driven understanding of its market potential.
We identify three primary learning outcomes associated with this grant proposal. First is a more thorough understanding of the alcohol and cannabis laws regulating production, distribution, and commerce, emphasizing the North Central region. This includes analyzing the loopholes in the 2018 Farm Bill that have allowed select THC-infused beverage sales in states without recreational cannabis. This understanding is obtained through a review of the academic literature and state/federal policies. Knowledge is disseminated to our target audience through presentations at conferences and workshops, academic publications, and media engagement.
Second, by identifying consumer preferences for alcohol and cannabis products, we analyze anticipated purchasing patterns within these industries. Generating and sharing this knowledge with our target audience will leave members of the alcohol supply chain better equipped to manage these evolving markets.
The final learning outcome is understanding the demand for hemp-derived beverages, as this demand affects farmers and beverage producers alike. Uncovering consumer demand for cannabis-infused beverages will equip farmers and beverage producers with the knowledge to make production decisions.
The proposal also has three intended action outcomes. Primarily, the expected output will better inform decision-making on niche crop selection, mitigating uncertainty about the longevity of these markets. Next, the output will influence local craft beverage producers' production decisions as they continue to innovate. Finally, the proposal seeks to heighten industry collaboration amongst stakeholders across the alcohol and cannabis supply chains, increasing market coordination, participation, and lobbying efforts.
Cooperators
- (Researcher)
Research
We construct a survey instrument to achieve our main objectives. The initial proposal outlined three surveys, each focusing on cannabis and a single alcohol type. Instead, we use one survey that combines cannabis and all alcohol types because of insights from conversations with industry members and advancements in the choice modeling literature. In brief, our discussions have conveyed that having a holistic view of the beverage market is essential (i.e., do not consider the relationship between one type of alcohol and cannabis in isolation). Many adults consume more than one type of alcoholic product, and we wanted to create a hypothetical setting that mimicked the real-world purchasing environment. If we only focused on one alcohol type at a time, we may bias the estimates since we do not observe the whole market. Recent advancements in the economics literature provide a convenient tool to address this change in the study's framework.
The survey begins with a series of demographic questions to ensure we obtain a sample representative of the U.S. population with respect to gender, age, income, education, and region. Following these questions, respondents participate in a hypothetical economic experiment. This serves as the primary portion of the study. In this experiment, known as a basket-based choice experiment, consumers are presented with multiple types of alcohol and cannabis products that vary in price. Consumers are asked to select the alcohol and cannabis products they would choose if they faced this decision in real life. In analyzing the responses to this experiment, we address the primary objective of this study, which is to understand whether alcohol and cannabis are economic complements or substitutes.
A secondary objective of this study is to analyze the demand for cannabis-infused beverages. The experiment described above partially addresses this goal by including cannabis-infused sparkling water in the experimental framework. For a more comprehensive analysis, however, respondents answer a series of follow-up questions regarding their willingness to pay (WTP) for various cannabis-infused beverages. Specifically, respondents are asked to state their WTP for 16 products, including CBD- and THC-infused options, using double-bounded contingent valuation methods. We ask about the WTP for a 12 oz. can of non-alcoholic beer, wine, ready-to-drink, and cider for both CBD and THC. We also ask about their WTP for 12 oz. cans of sparkling water, sweetened iced tea, and iced coffee to understand the demand for more general cannabis beverage product offerings (as opposed to adult beverage-inspired cannabis drinks). As stated preference methods can have hypothetical bias, leading some consumers to overstate their WTP, we include cheap talk scripts following our profession's best practices. We also use methodological techniques that allow us to deflate WTP estimates using insights from the marketing literature.
After the WTP questions for various cannabis-infused beverages, we ask respondents additional questions about their personal and household characteristics to perform market segmentation analysis. The survey instrument received IRB approval from Michigan State University, and we hired Qualtrics to collect data. Ultimately, we were able to collect data from 2,685 U.S. adults 21+.
Data collection occurred in May 2023 (following IRB approval in March 2023), and primary data analysis occurred from June 2023 to February 2024. However, data analysis remains ongoing. Nonetheless, the current project updates include three academic paper drafts, two that have been accepted for publication and one under review. We summarize the key insights from these papers and list their stage in the peer-review process below.
- "Beer drinker perceptions of cannabis-infused beverages," published in the British Food Journal. Given the rise in non-alcoholic beers, this paper uses survey data from 1,094 U.S. beer consumers to understand attitudes toward and willingness to try CBD- and THC-infused, non-alcoholic beverages. (Note: the sample of beer drinkers listed here is a subset of the full sample.) There are two main components of the study. First, we use regression analysis to uncover which consumers are most likely to try cannabis-infused beverages. Within the survey, respondents report whether they would be willing to try CBD- and THC-infused beverages. Their responses are then used in multinomial logistic regression analysis to understand the characteristics and behaviors most associated with a willingness to try these beverages. The results suggest that more than 50% of beer drinkers in the sample would be willing to try cannabis-infused beverages, and age is the leading predictor. Additionally, craft beer drinkers are more likely to be willing to try THC-infused beverages. Secondly, in developing a hypothetical craft brewing company in the survey, we assess perceptions of alcoholic beers relative to cannabis-infused beverages. The results suggest beer drinkers expect similar safety and nutrition ratings for the traditional and cannabis-infused alternatives. They also expect price premiums on the cannabis-infused options relative to conventional beer. (Full citation: Staples, A.J. (2024). Beer drinker perceptions of cannabis-infused beverages. British Food Journal, 127(2), 451-475. https://doi.org/10.1108/BFJ-02-2024-0209)
- "Canning cannabis: Consumer preferences for CBD- and THC-infused beverages," published in the Journal of Wine Economics: This article uses survey data and double-bounded contingent valuation methods from the total sample of 2,685 U.S. consumers to understand their attitudes toward and willingness to try CBD- and THC-infused, non-alcoholic beverages. This study offers three key findings and builds off the findings presented in the first paper. First, in targeting a nationally representative sample of U.S. adults with respect to different demographic characteristics (rather than just the beer drinkers previously), the study suggests that more than 50% of respondents would be willing to try cannabis-infused beverages, with age being the leading predictor of openness to consumption. Secondly, in constructing 16 hypothetical cannabis-infused beverages and estimating willingness to pay for each product, we show that, on average, consumers are willing to pay $0.60 more for THC-infused beverages than CBD-infused beverages, with juices, iced coffee, and sweetened iced tea being the most popular product options. Finally, it is important to understand the demographics and consumer characteristics driving the acceptance and willingness to pay for new, emerging products. The study determines which consumers are willing to try these products and how much each consumer is willing to pay. In doing so, it evaluates the characteristics associated with higher potential CBD- and THC-infused beverage consumption. The results reveal that age, alcohol use, and recent cannabis consumption best predict a willingness to try and a higher willingness to pay. (Full citation: Staples, A.J. (2024). Canning cannabis: Consumer preferences for CBD- and THC-infused beverages. Journal of Wine Economics, 19(4), 313 - 334. https://doi.org/10.1017/jwe.2024.21).
- "Sin goods in a basket-based choice experiment: Who bundles alcohol and cannabis?" under review at the Journal of Experimental and Behavioral Economics: This paper presents the results of the basket-based choice experiment using data from 2,010 U.S. adults. The sample size differs from the second paper because we had to exclude respondents who took the survey on their phones due to display differences. This paper achieves the main objective of this proposal in analyzing alcohol and cannabis preferences. It also identifies the consumers most likely to simultaneously consume these substances. Through basic summary statistics, multivariate logistic modeling, and exploratory regression analysis, we challenge the dichotomous generalization of classifying alcohol and cannabis as complements or substitutes. Instead, we show that most consumers view these substances as independent markets, while a non-trivial population views them as complements. The exploratory regression results reveal that males and younger consumers are most likely to bundle these substances. We also show that personality correlates with this bundling behavior, where higher agency, openness, neuroticism, extraversion, and lower conscientiousness predict simultaneous purchasing behavior. These results offer valuable insights that can be used in health policies but also by stakeholders.
Other researchers, extension educators, and news media are aware of the work being conducted, which has also allowed us to effectively disseminate results to our target stakeholders across these supply chains.
Educational & Outreach Activities
Participation Summary:
The study was first presented at the 2022 Agricultural and Applied Economics Association (AAEA) Annual Meeting in Anaheim, CA, where we discussed the project's framework and the next steps in data collection. It was also discussed in an MSU Extension hop working group meeting, which improved our experimental design and increased awareness of the project.
The data were then collected in May 2023, and we have since prepared three articles for peer review. The results have also been presented at two international conferences in the summer of 2024, including Beeronomics (Milan, Italy) and the American Association of Wine Economists Annual Meeting (Lausanne, Switzerland). This has allowed us to disseminate key results to a global network of industry and academic professionals, receive important feedback on our project, and generate avenues for future research.
As we wrap up the project and await the publication of our final paper, we will continue our social and news media engagement to share findings with our relevant target audience.
Project Outcomes
From an economic perspective, the emergence of craft beverage and cannabis markets has allowed farmers to diversify crop selection, which could increase farm profitability. By uncovering consumer demand, willingness to pay, etc., the results assess the long-term viability of these markets to help industry stakeholders think about risk and expected future profits. As for the social benefits, many studies have shown that craft beverage manufacturers and cannabis markets have boosted state and local economies through job creation, tax revenue, etc. In sharing our results, we provide insights into the demand for different alcohol and cannabis products. Highlighting these consumer results will have implications that affect local beverage manufacturers and cannabis producers, which affects on-farm decision-making. This project provides an important first step in our understanding of the hemp-derived beverage market, and there remain several avenues for future research.
During the first year of this project, I increased my understanding of (i) the regulatory landscape of alcohol and cannabis supply chains; (ii) the manufacturing process of cannabis-infused beverages; and (iii) the delicacies of niche supply chains as they relate to farm profitability and sustainability. In addition, I have increased my survey and experimental design skills, including learning novel mathematical models to model consumer choice in our alcohol and cannabis setting.
During the second year of the project, I (i) improved my survey design capabilities, (ii) learned new analytical tools, and (iii) improved my software programming proficiencies. These advancements have translated to multiple papers seeking publication that highlight the complexity of this new emerging market.
During the third year of the project, I continued to build the skills developed in the first two years while also crafting my writing skills to publish the research findings.
Outside of my own personal experiences, knowledge on the demand for and perceptions of cannabis-infused beverages has been shared with a variety of groups. While data were not collected at conferences and workshops due to the timing and location of these events, researchers, industry groups, and other organizations have gained insights from this project. For example, on the academic side, the Journal of Wine Economics paper was published in late-November and has nearly 300 views. The British Food Journal paper was published in December 2024 and is about to eclipse 100 downloads.
Outside of the academic papers, my LinkedIn posts about these two papers also reached a wide and diverse audience audience. The post summarizing my Journal of Wine Economics paper received 2,600+ impressions and reached nearly 1,500 unique members. While most of these individuals work in higher education or research, 15% work in food and beverage services or manufacturing. The post for the British Food Journal publication received 1,500+ impressions and reached 880 unique members. Again, roughly 15% of these members work in food and beverage services or manufacturing.
I have had multiple researchers and industry representative email me about the papers following outreach campaigns. This includes an an hour-long meeting with industry representatives at Sightlines who work in this space to discuss the project output. They found the exploration of different beverage categories and the calculation of WTP to be a fascinating angle. The research has also resulted in another professor emailing me to begin an early-stage collaboration building on this work. The research has also been cited in an industry newsletter describing the case for low-dose hemp-derived beverages based on existing consumer demand.