Consumer Perceptions of Sustainable Farming Practices: A Best-Worst Scenario

TLDR: This paper used data collected from a national web-based survey of 1,002 households to examine consumer inferences and valuations of food products promoted as "sustainably produced" and found that consumers perceive farm size and local production as important elements of sustainable agriculture.

Abstract: Building on existing work evaluating food-attribute labels, we use data collected in 2010 from a national web-based survey of 1,002 households to examine consumer inferences and valuations of food products promoted as “sustainably produced.” A best-worst scale framework was implemented to identify how consumers deϐine “sustainably produced” and their preferences for each of the sustainable farming practices considered. The results suggest that consumers perceive farm size and local production as important elements of sustainable agriculture while economic attributes exhibit a signiϐicant amount of heterogeneity, indic ating segmentation in the sample and the potential for targeted marketing. Food produced using sustainable production practices is receiving increasing attention in both public and private arenas as a greater number of food products are being marketed and labeled using “sustainable” or “sustainably produced” certiϐication schemes for differentiation. As sustain ably produced food has gained market momentum, questions have arisen among researchers and marketers regarding what consumers perceive when faced with the label “sustainably produced.” Speciϐically, consumers want to know wh at claims of sustainability imply about the environmental, economic, and social factors associated with production and farmers want to know what consumers are willing to pay for this value-added attribute before either party invests heavily in sustainability certiϐication. The attributes that consumers desire in terms of food system sustainability recently have been studied in more detail. For example, a framework for evaluating consumer priorities with regard to sustainable food was built by Clonan et al. (2010) based on seven guiding principles of sustainability put forth by Sustain, 1 an alliance for better food and farming. The authors used a ϐive-point Likert Scale embedded in a structured questionnair e to explore attitudes toward sustainability components such as fair trade, organic and local production, and animal welfare. The study found that consumers responded

Figures

Table 4. Long-form Data Example

Table 1. Demographic Variables and Summary Statistics of Survey Participants

Table 5. Relative Importance of Sustainable Production Attributes: Logit Estimates

Table 2. Apple and Beef Survey Attributes

Table 6. LCA Parameter Estimates: Probability of Apple Production Attribute Chosen as Most Important by Cluster

Table 7. LCA Parameter Estimates: Probability of Beef Production Attribute Chosen as Most Important by Cluster

Full text

Agricultural and Resource Economics Review 42/2 (August 2013) 275–290
Copyright 2013 Northeastern Agricultural and Resource Economics Association
Consumer Perceptions of Sustainable
Farming Practices: A Best-Worst
Scenario
Hillary M. Sackett, Robert Shupp, and Glynn Tonsor
Building on existing work evaluating food-attribute labels, we use data collected in
2010 from a national web-based survey of 1,002 households to examine consumer
inferences and valuations of food products promoted as “sustainably produced.”
A best-worst scale framework was implemented to identify how consumers deine
“sustainably produced” and their preferences for each of the sustainable farming
practices considered. The results suggest that consumers perceive farm size and
local production as important elements of sustainable agriculture while economic
attributes exhibit a signiicant amount of heterogeneity, indicating segmentation in
the sample and the potential for targeted marketing.
Key Words: best-worst, consumer perceptions, sustainably produced food
Food produced using sustainable production practices is receiving increasing
attention in both public and private arenas as a greater number of food
products are being marketed and labeled using “sustainable” or “sustainably
produced” certiication schemes for differentiation. As sustainably produced
food has gained market momentum, questions have arisen among researchers
and marketers regarding what consumers perceive when faced with the label
“sustainably produced.” Speciically, consumers want to know what claims of
sustainability imply about the environmental, economic, and social factors
associated with production and farmers want to know what consumers are
willing to pay for this value-added attribute before either party invests heavily
in sustainability certiication.
The attributes that consumers desire in terms of food system sustainability
recently have been studied in more detail. For example, a framework for
evaluating consumer priorities with regard to sustainable food was built by
Clonan et al. (2010) based on seven guiding principles of sustainability put
forth by Sustain,
1
an alliance for better food and farming. The authors used
a ive-point Likert Scale embedded in a structured questionnaire to explore
attitudes toward sustainability components such as fair trade, organic and local
production, and animal welfare. The study found that consumers responded
1
See www.sustainweb.org/sustainablefood.
Hillary Sackett is an assistant professor in the Department of Economics and Business Management
at Westield State University. Robert Shupp is an assistant professor in the Department of
Agriculture, Food, and Resource Economics at Michigan State University, and Glynn Tonsor is
an associate professor in the Department of Agricultural Economics at Kansas State University.
Corresponding Author: Hillary Sackett
 Department of Economics and Business Management 
Westield State University
 577 Western Avenue  Westield, MA 01086  Phone 413.572.5253  Email
hsackett
@westield.ma.edu.
Financial support from the U.S. Department of Agriculture (USDA) is gratefully acknowledged.
The views expressed in this paper are the authors’ and do not necessarily represent the policies or
views of the sponsoring agencies.

276 August 2013 Agricultural and Resource Economics Review
positively to environmental responsibility metrics related to how their food was
produced. Similarly, using a discrete choice modeling method for evaluating
consumer attitudes regarding sustainability claims on food products, Saunders
et al. (2010) focused on results from a Likert Scale rating of sustainability
attributes in the context of carbon emissions and other contributions to global
climate change. However, the relatively limited literature on sustainable food
labels generally has not focused on identifying attributes of sustainability that
consumers believe are or should be important components of a “sustainably
produced” labeling scheme.
Batte (2010) reviewed several studies that identiied consumer-driven
changes in food marketing channels related to sustainable food claims. In
the review, Batte identiied three studies that supported the importance
of consumer demand in food-product-differentiation schemes. The irst, a
study by Onozaka and Thilmany McFadden (2010), found evidence from a
conjoint choice experiment of signiicant heterogeneity in valuing various
food-differentiation claims among shoppers in various marketing venues.
They noted that the consistent signiicance of self-perceived eficacy in the
psychographic model suggested that consumers who believe they have a role
to play in improving sustainability tend to value sustainable product claims
more. In another study, Onozaka and Thilmany McFadden (2011) explored the
increasing use of sustainable food labels by analyzing the interactive effects of
sustainable production claims and found that products that were locally grown
were the most highly valued. The authors suggest that consumers’ preference
for local food extends beyond basic quality characteristics. Increasingly, it is
related to sustaining the local economy by supporting the area’s farmers and
conserving local farm land. Batte (2010) concluded that further research was
needed to identify how consumer demand for sustainably produced food is
affected by the perceived importance of environmental, economic, and social
attributes used to differentiate products through certiication and labeling.
To address the void in the literature related to attributes of sustainability
that are important to consumers, we use a best-worst scaling framework
to examine consumer beliefs about the meaning of “sustainably produced”
in the context of food labels. Best-worst choice modeling is a relatively new
technique for analyzing consumer preferences or beliefs. It has been applied
in a variety of settings, including public health research by Flynn et al. (2007)
and international agribusiness marketing studies by Umberger, Stringer,
and Mueller (2010) and by Scarpa et al. (2011), to estimate desired tourism
beneits. In an agricultural and food system context, best-worst choice designs
and analyses have been previously demonstrated using empirical examples
involving the wine market by Casini and Corsi (2009), Cohen (2009), and
Mueller and Rungie (2009). More closely related to our work is a study by Lusk
and Parker (2009), which used best-worst scaling techniques to identify the
methods preferred by consumers for improving (reducing) the fat content
in ground beef. Most closely related to the current study is one by Lusk and
Briggeman (2009), which determined consumers’ relative attitudes toward
value-added food attributes such as safety, nutrition, taste, and price using the
best-worst framework. Here, we examine consumer attitudes toward value-
added sustainable production attributes in a similar fashion.
The primary purposes of this study are threefold. One goal is to introduce
an economic application of best-worst scaling for measuring the importance
of production attributes in consumer decision-making in the context of food

Consumer Perceptions of Sustainable Farming Practices 277Sackett, Shupp, and Tonsor
systems. Additionally, we seek to identify consumers’ perceptions of the
environmental, economic, and social indicators of sustainability currently used
by third-party agricultural certiiers as farm/ranch-level evaluation with a
goal of identifying the indicators that consumers see as important components
of food production. Finally, we aim to assess consumer heterogeneity in
perceptions of the importance of sustainable farming practices when choosing
“sustainably produced” food for insight into potential marketing strengths for
third-party certiiers.
Research Methodology
Best-Worst Scaling
Marketing surveys that measure attribute importance most often use a Likert
Scale ranking approach. However, the method has several known weaknesses.
First, scaled rating systems do not force respondents to make tradeoffs between
attributes. Additionally, Likert-Scale-ranked data defy natural interpretation
outside of the survey context. To address these issues, we implemented a
best-worst design to investigate preferences for and perceptions of alternative
sustainable farming practices. The survey instrument used to collect consumer
data was designed to simplify the choice task for respondents.
Best-worst analysis requires survey respondents to choose the most important
and least important attributes from a set of competing options simultaneously.
This method is commonly referred to as “maximum difference scaling” since
the attributes chosen should maximize the difference in utility realized by
a respondent on an underlying scale of preference. The measured level of
importance from the best-worst data analysis is applied to a standardized ratio
scale that determines the percentage difference in importance across attributes
with more certainty. The theoretical foundation for this analysis is provided by
Marley and Louviere (2005) in their development of probabilistic models for
analyzing best-worst choice tasks.
Best-worst scaling, as originally devised by Flynn and Louviere (1992),
is capable of addressing relative impacts on utility across attributes that
traditional discrete choice questions cannot. To observe tradeoff behavior in a
best-worst model, the speciication of attributes from a choice set of competing
alternatives is repeated over a number of variable choice subsets. In this way,
best-worst tasks provide more information than single choice designs while
forcing respondents to consider the extremes of their utility space. Ideally,
the stated preference outlined in each best-worst scenario approximates
observed consumer behavior in retail markets. We created each choice set
using a 210 main-effects orthogonal experimental design
2
that was balanced
with ten attributes, each exhibiting two levels. The speciic attributes chosen
for inclusion are described in the next subsection and further outlined under
Data Collection. The orthogonal experimental design yielded twelve alternative
choice sets, which were broken into two blocks and randomized across
participants. Each block contained two questions that involved ive alternatives,
three questions that involved six alternatives, and one question that included all
ten alternatives. The presence or absence of each sustainable farming attribute
was independent across choice subsets, allowing for identiication of relative
2
Linear Models and Analysis of Variance: Concepts, Models, and Applications, Volume II, David W.
Stockburger, Southwest Missouri State University, 1993.

278 August 2013 Agricultural and Resource Economics Review
preferences on a ratio scale. The additional utility or disutility from moving
between attribute levels was estimated using a logit model framework.
Sustainability Attributes
The U.S. Department of Agriculture’s (USDA’s) National Institute of Food and
Agriculture (NIFA) provides limited information on the purported sustainability
of particular farming and ranching practices. At a national level, sustainable
agriculture was irst addressed by Congress in the 1990 Farm Bill (Public
Law 101-624, Title XVI, Subtitle A, Section 1603), which deined sustainable
agriculture as used within the bill.
The term sustainable agriculture means an integrated system
of plant and animal production practices having a site-speciic
application that will, over the long term:
⦁ satisfy human food and iber needs
⦁ enhance environmental quality and the natural resource
base upon which the agricultural economy depends
⦁ make the most eficient use of nonrenewable resources and
on-farm resources and integrate, where appropriate, natural
biological cycles and controls
⦁ sustain the economic viability of farm operations
⦁ enhance the quality of life for farmers and society as
a whole.
However, USDA also warned that “guidelines about what speciic practices
meet long-term environmental, economic, and social goals and constitute
sustainable agriculture [are] still under debate” (National Agricultural
Library (NAL) 2009, p. 1). Therefore, when developing our consumer survey
to determine perceptions of practical farm-level components of a sustainable
production system, we followed USDA’s advice and included environmental,
economic, and social attributes of sustainability that address the last two points
of the preceding deinition.
The USDA website provides links (NAL 2009) to a handful of groups that
have developed standards and provide certiication services based on those
standards. The irst one listed by USDA is Food Alliance (www.foodalliance.org),
a nonproit organization that developed standards and operates a voluntary
certiication program. Food Alliance is the most comprehensive certiication
program for sustainable food production in North America. It employs
independent third-party inspectors to audit Food-Alliance-certiied businesses
to determine that they continue to meet the program’s standards and criteria.
Food Alliance outlines its crop- and livestock-speciic certiication criteria and
provides the following general requirements for certiied production.
⦁ Provide safe and fair working conditions
⦁ Ensure the health and humane treatment of animals
⦁ No use of hormones or subtherapeutic antibiotics
⦁ No genetically modiied crops or livestock

Consumer Perceptions of Sustainable Farming Practices 279Sackett, Shupp, and Tonsor
⦁ Reduce pesticide use and toxicity
⦁ Protect soil and water quality
⦁ Protect and enhance wildlife habitat
⦁ Continuously improve management practices
Many farming practices with sustainability characteristics could have been
included for the purposes of this study. We adhered as closely as possible to
Food Alliance’s whole farm/ranch evaluation criteria for crops and livestock
when choosing the sustainability attributes to include in the survey. Those
criteria are publicly available online and from the authors upon request.
Because there currently is no government-sponsored certiication of
sustainable food production, all claims of “sustainably produced” found on
food product labels in the United States are certiied by private third parties
such as Food Alliance or by the farm of origin. We are interested in identifying
the certiication-guided environmental, economic, and social attributes that
consumers perceive as important indicators of sustainability when faced with
these kinds of purchasing decisions.
The set of attributes used in this study offers insight into perceived
dimensions of sustainability and provides opportunities for expansion in future
work. Due to the nature of the variable choice sets, the analysis that follows is
conditional on the set of evaluated attributes (farming practices) and should
not be interpreted as informing sustainability attributes outside of this context.
Data Collection
To our knowledge, this study is the irst to use best-worst scaling to measure
consumer perceptions of the importance of production attributes in the
context of food system sustainability. We chose best-worst scaling because we
are especially interested in determining the relative importance of sustainable
agricultural certiication criteria currently used in food markets. To collect
consumer data for the analysis, we disseminated a national web-based survey
of 1,002 households in the summer and fall of 2010. Consumer respondents
were recruited by Decipher, a marketing-research and survey-programming
company. A summary of the population demographic statistics is provided in
Table 1.
We developed two versions of the survey based on the product to be
purchased—one for apples and one for beef. Table 2 lists the sustainable farming
practices included in accordance with Food Alliance’s crop-speciic (apple)
and livestock-speciic (beef) evaluation criteria as the best-worst attributes
from which survey respondents would choose. Many of these practices fall
under current organic-certiication guidelines endorsed by USDA, and all are
components of sustainable farm certiication by Food Alliance. The attributes
span the three-pronged sustainability framework suggested by Callens and
Tyteca (1999) and are in line with NIFA’s use of environmental, economic, and
social metrics for evaluation (NAL 2007).
We developed the choice sets for the analysis around ten attributes that
each were divided into two levels indicating the presence or absence of a
given farming practice. Consumers were shown a set of attributes and were
asked to indicate the one that was most important (best) and the one that was
least important (worst) in repeated choice opportunities in which the set of

Frequently asked questions

Q1. What contributions have the authors mentioned in the paper "Consumer perceptions of sustainable farming practices: a best-worst scenario" ?

Building on existing work evaluating food-attribute labels, the authors use data collected in 2010 from a national web-based survey of 1,002 households to examine consumer inferences and valuations of food products promoted as “ sustainably produced. ” A best-worst scale framework was implemented to identify how consumers de ine “ sustainably produced ” and their preferences for each of the sustainable farming practices considered. The results suggest that consumers perceive farm size and local production as important elements of sustainable agriculture while economic attributes exhibit a signi icant amount of heterogeneity, indicating segmentation in the sample and the potential for targeted marketing. 

Q2. What are the primary approaches used to analyze the best-worst ranked data?

Paired models are implemented to make inferences about the latent utility scale while marginal (count-based) models aggregate best and worst choices over all of the pairs that include a given attribute level to model choice frequencies. 

Q3. What is the ive-point scale used to assess consumer attitudes toward sustainability?

The authors used a ive-point Likert Scale embedded in a structured questionnaire to explore attitudes toward sustainability components such as fair trade, organic and local production, and animal welfare. 

Q4. What attribute was often chosen as least important in every scenario?

The attribute “Consumer food prices are affordable” was most often chosen as least important in every scenario and was therefore set as the base category. 

Q5. What percentage of the sample consists of consumers who believe that local production is a highly important?

For apples, the irst cluster, representing 14 percent of the sample, consists of consumers who believe that local production is a highly important aspect of sustainable production. 

Q6. What is the nature of the variable choice sets?

Due to the nature of the variable choice sets, the analysis that follows is conditional on the set of evaluated attributes (farming practices) and should not be interpreted as informing sustainability attributes outside of this context. 

Q7. How is the probability that attribute j is chosen as important?

Using the conditional logit parameter estimates (λj), the probability that attribute j is chosen as most important (best) and attribute k is chosen as least important (worst) is given by(3) Prob( j = best ∩ k = worst) = 

Q8. What are the important attributes of sustainable production systems?

This cluster of respondents most often chose prohibition of subtherapeutic antibiotics and bovine growth hormone along with pastured feed as the most important attributes of sustainable production systems. 

Q9. How did Saunders et al. (2010) evaluate consumer attitudes regarding sustainability?

using a discrete choice modeling method for evaluating consumer attitudes regarding sustainability claims on food products, Saunders et al. (2010) focused on results from a Likert Scale rating of sustainability attributes in the context of carbon emissions and other contributions to global climate change. 

Q10. What are the attributes of sustainability that are important to consumers?

Based on their initial results, size, scale, and geographic scope capture the attributes of sustainability that are most important to consumers, which is supported by the growing literature on the local food movement. 

Q11. What is the preference share sum for the given attribute?

The preference shares sum to one, and each represents the proportional share of importance for the given attribute relative to a value for consumer food prices, which is normalized to zero. 

Q12. How important is the attribute of farm size?

When standardized to the ratio scale, the sampled consumers reported that farm size is seven times more important in a sustainable agricultural system than animal health and safety.