THREE ECONOMIC ESSAYS ON THE U.S. MEAT AND POULTRY INDUSTRIES
By
Kelsey A. Hopkins
A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Agricultural, Food, and Resource Economics – Doctor of Philosophy
2024
ABSTRACT
This dissertation focuses broadly on the U.S. domestic meat and poultry industries, more
specifically on issues related to developing and adopting policies and programs for two niche
markets – halal and farm animal welfare friendly. Halal meat and poultry products meet the
religious dietary restrictions for Muslim consumers, specifically those related to the slaughter
process for animals. The first two essays focus on the halal market, while the third focuses on farm
animal welfare policy adoption.
The first chapter is concerned with meat and poultry processors’ and retailers’ decision-
making patterns related to supplying halal meat and poultry in the U.S. domestic market. There
has been strong ongoing demand growth for halal meat and poultry products in the U.S., but a
relative dearth of processors and retailers entering the market to supply these goods. This essay
seeks to understand if there are differences in preferences and business decision-making behavior
between agents in the halal market and agents outside to suggest methods in which to increase
market participation. To do this, I utilize a mixed methods design consisting of qualitative
interviews with halal processors and retailers and survey data from halal and non-halal processors
and retailers. My methodology consists of analyzing Likert scale data using descriptive statistics,
principal components analysis, and k-means clustering to reveal patterns and group respondents
for comparison. My results show that businesses that may expand into the halal market have been
established longer and more likely to be retailers or further processors.
The second chapter focuses on market participants’ preferences for designing a U.S.
national halal meat and poultry certification program. The development of such a program serves
as a potential solution to food fraud stemming from an overabundance of confusing and commonly
contradictory certifications already in the market, similar to the issues that lead to the creation of
the USDA Organic standard. I again use a mixed methods approach of halal consumer, retailer,
and processor qualitative interviews paired with national stacked surveys containing best-worst
scaling questions to investigate preferences for the design of a U.S. national halal meat and poultry
certification program. Results show that the market overall prefers that program designers consider
most carefully Who/What is Certified, Halal Standards, and Costs. Additional results show
preferences for which organizations should be involved in setting and/or enforcing this program,
namely government (enforcement only), non-government, religious, and certifier organizations.
Finally, the data indicated that multiple transparency and traceability measures should be included
to ensure a robust and trustworthy program. In all, this chapter aids in bolstering halal meat
consumer confidence in product authenticity and improves the equity of the U.S. food system.
The final chapter explores modeling of farm animal welfare regulation adoption across the
U.S. In the U.S., 19 state-level bills and ballot initiatives concerning farm animal welfare (FAW)
have been adopted across 12 states. In this chapter, I and my co-authors seek to model the evolution
of the state-level FAW regulatory landscape as a function of legislature characteristics and
constituent demographics. More specifically, we utilize a two-stage model known as a multinomial
endogenous switching regression to assess whether and when a given state considers FAW
measures, and if so, the likelihood the measures are passed. Using this model, we estimate the
likelihood of FAW adoption for all 50 states. Additionally, we find that the cost to the egg and
pork industries to upgrade to cage- and crate-free production methods in the states most likely to
pass a FAW regulation in the future is small relative to the size of the industry. Our findings will
assist producers and industry stakeholders in gauging the future of the regulatory landscape and
provide guidance on whether to upgrade existing enclosures to comply with mandates on the
horizon or to continue operating with “conventional” enclosures.
To my mom for her unwavering support and belief in me – not to mention her stellar
proofreading skills.
To my husband, Alex, for his love and encouragement in everything I do.
To my twin daughters, Violet and Vivienne, who serve as both heralds of my next chapter and
strong motivators in the completion of this degree.
iv
ACKNOWLEDGEMENTS
Completing this dissertation would not have been possible without support and
encouragement from many channels. First and foremost, I thank my spectacular superwoman of
an advisor, Dr. Melissa G.S. McKendree, for her faith in me and the many hours we spent together
working towards the finish line of this degree. I thank my committee – Professors Kimberly Chung,
David Ortega, and Felicia Wu – for their expertise and experience that greatly improved this
dissertation.
I am also thankful for the friendships I have made while at Michigan State, and how these
wonderful humans shaped my time there. I am especially grateful for my friendships with Sarah
Klammer, Angelos Lagoudakis, and Caitlin and Matt Herrington – for all the coffees drank,
dinners eaten, problems solved, laughs had, and memories made.
Finally, I am thankful for my family. I give thanks for my mom, Sharon McCoy, for
everything she has done for me and the sacrifices she has made, and my husband, Alexander
Hopkins, for his love and belief in me.
v
Islamic Dietary Laws
PREFACE
In order for a meat product to be halal – “permissible” for consumption – there are many
qualifications that must be met; if they are not, the product is haram, or “forbidden.” According to
Bonne and Verbeke (2007, 2008), most practicing Muslims require halal certifications that ensure
that Islamic dietary laws are followed at all stages of the supply chain. Islamic dietary restrictions
are nuanced, and as with many religious texts, open to interpretation. Therefore, I detail only the
most widely accepted views and dietary laws here.
Muslim teachings prohibit consumption of certain species or types of meat, namely pork
and dead meat. Dead meat refers to an animal whose spinal cord is severed in the process of
slaughter, rendering the heart unable to pump and thus the animal does not die of exsanguination.
Additionally, animals must be raised on a natural, vegetarian diet that excludes filth or any animal
proteins. Animals must be treated humanely; they must be well-nourished, not stressed before
slaughter, the knife should not be sharpened in front of the animal, and no animal should witness
the slaughter of another animal. Animals must be alive – both heart and brain still fully functioning
– at the time of slaughter and must die of blood loss. As such, there is much debate about the use
of electrical stunning prior to slaying an animal; many Muslims are opposed to the practice because
of the risk of premature death, but it is common and considered humane in conventional U.S.
slaughter.
Preferably, animals should be slaughtered by hand; machine slaughter is not fully accepted.
The knife used should be sharp enough to kill the animal with one cut. The slaughter person must
be a sane adult Muslim who invokes the name of Allah prior to each individual slaughter.
Recordings of blessings that are played on loop are not recognized as halal. Finally, cross-
vi
contamination with haram products renders halal products haram. As such, all steps of the
slaughter process and the following supply chain members must be certified halal.
vii
TABLE OF CONTENTS
CHAPTER 1. PROCESSOR AND RETAILER MOTIVATIONS FOR HALAL MEAT AND
POULTRY MARKET PARTICIPATION: PROFIT, PIETY, AND PURPOSE .......................... 1
1. Introduction and Motivation ................................................................................................... 1
2. Research Methods ................................................................................................................... 4
3. Quantitative Data Collection Process ................................................................................... 10
4. Quantitative Analysis Methods ............................................................................................. 14
5. Summary Statistics................................................................................................................ 21
6. Results & Discussion ............................................................................................................ 31
BIBLIOGRAPHY ..................................................................................................................... 47
CHAPTER 2. MARKET PARTICIPANTS’ PREFERENCES FOR A NATIONAL HALAL
MEAT CERTIFICATION PROGRAM ....................................................................................... 52
1. Introduction and Motivation ................................................................................................. 52
2. Current Landscape of U.S. Domestic Halal Certifications ................................................... 54
3. Halal Certification Around the World .................................................................................. 57
3. Mixed Methods and Survey Design ...................................................................................... 59
4. Methodology & Statistical Framework for Analysis ............................................................ 74
5. Results and Discussion ......................................................................................................... 78
6. Implications for Implementation ......................................................................................... 102
7. Conclusions, Limitations, and Future Work ....................................................................... 105
BIBLIOGRAPHY ................................................................................................................... 109
CHAPTER 3. RESOLVING THE REALITY GAP IN FARM REGULATION VOTING
MODELS .................................................................................................................................... 114
1. Introduction ......................................................................................................................... 114
2. Background ......................................................................................................................... 117
3. Methodology ....................................................................................................................... 121
4. Results ................................................................................................................................. 132
5. Implications ......................................................................................................................... 142
6. Conclusion .......................................................................................................................... 146
BIBLIOGRAPHY ................................................................................................................... 148
APPENDIX A.1 INTERVIEW PROCESS AND ANALYSIS OVERVIEW ............................ 153
APPENDIX A.2 PROCESSOR, RETAILER, AND CONSUMER SURVEY QUESTIONS .. 155
APPENDIX A.3 BEST-WORST SCALING RESULTS AND ADDITIONAL TABLES ....... 221
APPENDIX A.4 FARM ANIMAL WELFARE REGULATIONS ADDITIONAL TABLES .. 236
APPENDIX A.5 FARM ANIMAL WELFARE DATA COLLECTION METHODS .............. 242
APPENDIX A.6 CALCULATION OF FARM ANIMAL WELFARE INDUSTRY COSTS .. 245
viii
CHAPTER 1. PROCESSOR AND RETAILER MOTIVATIONS
FOR HALAL MEAT AND POULTRY MARKET
PARTICIPATION: PROFIT, PIETY, AND PURPOSE
1. Introduction and Motivation
Demand for halal meat and poultry (hereafter referred to as halal meat for brevity) products
in the U.S. is growing quickly. The population of U.S. Muslim consumers is increasing rapidly
and is projected to double by 2050, from about 3-5 million to 6-10 million people (Pew Research
2018). This growth in demand is in part due to immigration patterns that grow the consumer base,
but also the vertical mobility by second and third generation Muslims who have begun to consume
more meat products (Bereaud-Blackler 2004, Bonne & Verbeke 2007). Additionally, non-Muslim
consumers have demonstrated demand for halal meat products (Campbell et al. 2011).
This rapid growth in demand for halal meat products represents an opportunity for U.S.
meat processors and retailers to enter the market and expand halal meat availability. Despite this
strong demand and the corresponding opportunities, domestic supply of halal meat products is
relatively low. Further, halal meat and poultry products are not readily available outside of major
metropolitan areas or areas with relatively concentrated Muslim populations. Considering this
largely untapped market opportunity, this chapter addresses several research questions. First, how
did current halal meat retailers and processors decide to offer halal meat products in their stores?
Are there incentives and barriers that exist when entering the halal meat market, and how do these
retailers and processors perceive them? Further, what motivations do current halal meat retailers
and processors have for supplying [certified] halal meat? Taking this research one step further,
how do the answers to these questions change for non-halal meat processors and retailers, and what
comparisons can be made?
1
As with any market, participants face incentives and barriers to entry; thus, understanding
how participants view these incentives and barriers is essential to devising policies to increase
halal market participation. Compounding these aspects, the U.S. domestic halal meat market has
multiple religious community considerations in addition to the usual pecuniary motivations that
influence processor and retailer decision making in specialty market participation.
The U.S. domestic halal meat and poultry market is understudied. To my knowledge, there
is no agricultural economic academic research published to date that focuses on the U.S. halal
market. However, there is a recent study discussing religiosity and minority community behavioral
patterns and their relation to Muslims’ desires to consume halal meat (Mumuni et al 2018). The
majority of agricultural economic research into the supply side of halal meat markets has been
conducted in Europe (e.g., Ahmed 2008, Fuseini et al. 2017, Fuseini et al. 2021, Lever et al. 2010,
Lever & Miele 2012, Masudin et al. 2020, Tieman et al. 2012), Southeast Asia (e.g., Ab Rashid &
Bojei 2019, Salindal 2019, Shahijan et al. 2014, Tieman et al. 2012,), Australia (e.g., Zulfakar et
al. 2013, Zulfakar et al. 2018, Zulfakar et al. 2019), and Brazil (e.g., de Araújo 2019). In general,
these papers find that there are hurdles to halal meat market participation – including obtaining
certifications, implementing halal slaughter and processing methods, and ensuring supply chain
integrity. However, they also note that the benefits to supplying halal meat products are numerous,
including entry into niche markets, the ability to charge a price premium, stronger transparency in
the meat supply chain, and access to export markets. The U.S. meat market faces many of the same
obstacles and has the potential to reap many of these benefits.
Research into the U.S. meat sector in general has uncovered several barriers to and
incentives for market entry. Barriers include strong regulatory standards and food safety
regulations that can be daunting to potential entrants (Worosz et al. 2008); for halal and other
2
specialty or niche meat sectors, these standards include additional regulations and requirements to
achieve certification, likely compounding these issues. Additionally, there are issues related to
labor shortages and the seasonality of the livestock slaughter and processing industry (Lewis &
Peters 2011, Choe 2023), which can lead to logistical challenges for processors. However, research
in New England found that livestock producers commonly demanded more slaughter and
processing capacity than regional slaughterhouses were able to supply (Lewis & Peters 2011),
which suggests there is room for more processors in the market. Indeed, during the height of the
COVID-19 pandemic, processing capacity was a concern, largely due to continued labor shortage
concerns related to virus transmission and policies and regulations impacting supply chain
resiliency (Hobbs 2021, Ijaz et al. 2021, Larue 2021, Taylor et al. 2020, Weersink et al. 2021).
Thus, there are multiple incentives to market entry, including major market opportunities and
government grant programs – such as the 2022 Meat and Poultry Inspection Readiness Grant
(MPRIG) – designed to offset the cost of inspected meat processing for small and very small meat
processors (USDA-AMS 2022).
This essay contributes to the literature in several ways. First, this study serves to define the
baseline for the current state of the industry for future work, given the dearth of research on the
U.S. domestic halal meat supply chain. Second, this research contributes to a better understanding
of the motivations and attitudes of current halal meat processors and retailers towards market
participation, as well as the barriers to and incentives for entry they face(d). Additionally, it is the
first-ever evaluation of non-halal processors’ and retailers’ pecuniary and non-pecuniary
motivations for their decision to not enter the halal meat market, including their perceptions of
barriers to and incentives for halal meat market entry. This assessment provides necessary
identification of common concerns, misconceptions, and other factors dissuading processors and
3
retailers from entering the U.S. domestic halal meat market that can potentially be addressed
through increased outreach and education. Further, by utilizing a mixed methods approach with
active stakeholder engagement to ensure supply-side market participants’ concerns and opinions
are heard and incorporated into study design, this chapter expands the current body of literature on
addressing systemic inequities in the U.S. food system. In all, this work contributes to efforts to
promote equitable access to food for Muslim consumers through investigating methods to increase
the supply of authentic halal meat products nationwide.
2. Research Methods
The unique religious and niche nature of the halal market, coupled with the lack of research
to understand the behavior of U.S. halal meat suppliers requires a methodological approach
allowing for exploration and learning feedback. As such, I use a mixed methods approach
composed of both qualitative and quantitative components. The qualitative component consisted
of in-depth semi-structured interviews with halal meat and poultry market participants – halal meat
and poultry processors and retailers, as well as Muslim halal meat and poultry consumers. The
quantitative component consisted of nationwide online surveys to collect further data; the structure
and content of these surveys was informed by the data obtained in the qualitative interviews.
The synergy between methods is exhibited in two ways. In the first stage of this project, I
formally interviewed multiple halal meat processors and retailers; interviews were qualitative and
open-ended. These interviews were necessary to learn about the current state of the industry and
understand the nuances of market participation, halal production methods, and product
certification. I used these conversations to inform the development of the second stage of this
project – further narrowing my research questions and the design of the quantitative data collection
tools. From the qualitative interviews, I determined the most appropriate research questions to
4
explore in relation to halal meat processing and retailing revolve around barriers and incentives to
market participation, as well as perceptions of halal meat. This focus led me to decide that my
quantitative data collection should utilize Likert scale type survey questions to measure
perceptions of barriers and incentives to entry, as well as survey questions to benchmark current
levels of knowledge about halal meat. The data from the quantitative portion of this chapter is
analyzed and interesting, unexpected, or particularly important results are highlighted and
discussed in the context of information collected from the qualitative interviews. In this manner,
the research comes full circle and incorporates findings from both methodologies throughout the
project.
The two components of this chapter’s methods are described in more detail in the next
subsections.
2.1 Phase 1: Qualitative Exploration of U.S. Domestic Halal Meat and Poultry Industry
The qualitative portion of this study was designed using suggested methods from Patton
(2014), Rubin and Rubin (2011), and Maxwell (2012). Interview questions were grouped by
topic/area of interest and were open-ended to allow interviewees room for robust answers (Rubin
and Rubin 2011). These qualitative interviews were crucial components of this study; they
provided the opportunity to obtain information about the U.S. domestic halal meat and poultry
industry previously unknown to me, as this market is understudied. Information collected via
qualitative interviews included retailers’ and processors’ motivations for supplying halal meat and
their preferences between different certifications to use in their operations. Collecting information
directly from market participants in this way allowed me to design a more robust and relevant
research program for the quantitative portion of this project. The qualitative interviews also
provided vital context and explanatory power for the quantitative survey findings. The interview
5
guide for these qualitative interviews, information for how interviews were conducted, and the
outline of the interview analysis process are found in Appendix A.1.
Interviewees were recruited from lists of retailers and processors registered as certified to
supply halal meat by a reputable halal meat certification organization active in the industry for two
decades. These lists are available on the certification organization’s website. The interview
candidates were narrowed to those in Midwestern states for ease of access and improved likelihood
of name recognition for Michigan State University. A series of eight in-depth, semi-structured
interviews with Midwestern halal meat retailers and 10 in-depth, semi-structured interviews with
Midwestern halal meat processors were conducted December 2021 through April 2022.
2.2 Phase 2: Online Survey Methods for Further Exploration of Themes
The second stage of this chapter involves conducting two national online surveys, one with
a sample of meat retailers and another with meat processors. The goal of these surveys is to elicit
nationally representative opinions related to the information collected and to dig deeper into the
common themes discovered via the qualitative interviews. The sample for both retailers and
processors included those who currently offer halal products and those that do not. Including both
current halal meat market participants and those outside the market is important to be able to
compare differences in motivations and perceived barriers and incentives. This comparison will
allow us to understand what could increase market participation by meat processors and retailers.
Potentially, the results will inform policies to increase market participation, thereby increasing
supply of halal meat products and fulfilling demand. Each survey includes Likert scale questions
to elicit participants’ attitudes towards different incentives and barriers to supplying halal meat
and poultry products and using halal certifications. The surveys also included questions about
6
participant and business demographics, with emphasis on religious and cultural demographic
information.
In the qualitative interviews, the majority of halal meat processors and retailers described
strong support for increased certification utilization across the halal meat supply chain. Muslim
processors and retailers discussed the importance of supplying halal meat to support their cultural
and religious community members, while non-Muslim owners cited access to niche markets as a
motivation for supplying halal meat products. Along these lines, most of the processors
interviewed expressed interest in adopting new traceability strategies – including certifications,
blockchain, and revised (electronic) paperwork for inventory – in their operations to strengthen
the integrity of their operation and the halal meat market in general. Additionally, processors and
retailers reported a variety of barriers to and incentives for halal meat market participation that
influenced their decision making. Barriers or challenges related to halal meat market participation
included racial, ethnic, cultural, or religious biases and discrimination, limited access to financial
assistance such as Islam-compliant business loans, input shortages, and fraudulent competition.
Incentives included a desire to supply a necessary niche product to their community, the conviction
that financially supporting themselves/their families through halal methods was morally correct,
the belief that they are supplying a higher quality product, and the opportunity to access a niche
market.
Thus, the areas of interest for Likert scale questions were selected based on common
themes from the Phase 1 qualitative interviews. These qualitative interview findings suggested that
the Likert scale questions in the quantitative survey should be designed to include questions to
evaluate participants' attitudes towards potential barriers and incentives to halal meat market
participation and their knowledge of halal meat religious requirements. An example of these
7
questions for retailers and processors are included in Figure 1, Figure 2, Figure 3, and Figure 4.
Participants selected how likely each option was or would be an incentive (motivation) or
disincentive (barrier) for adding a halal program to their operations.
Finally, members of the Islamic community were asked to look over the survey for clarity
and to ensure there were no misrepresentations prior to distribution. Both the processor and retailer
surveys in entirety can be found in Appendix A.2.
Figure 1: Example Likert Scale Questions for Halal Retailing Incentives
8
Figure 2: Example Likert Scale Questions for Halal Processing Incentives
Figure 3: Example Likert Scale Questions for Barriers to Halal Retailing
9
Figure 4: Example Likert Scale Questions for Barriers to Halal Processing
3. Quantitative Data Collection Process
3.1 Data Collection Process & Summary Statistics
Surveying supply-side agents is notoriously difficult relative to surveying consumers.
Typically, supply-side studies receive a very low number of responses, due to a variety of reasons
– a smaller population to sample from, a lack of relevant pre-established survey panels, and the
opportunity cost of a business owner’s time being just a few. Nevertheless, I conducted multiple
efforts over many months to contact and obtain survey responses from both the processor and
retailer populations for this study; details on these efforts are given below.
10
3.1.1 Processor Sample and Data Collection
I recruited processor participants from three sources: 1) USDA Food Safety and Inspection
Service (FSIS) list of registered meat processors, 2) registered processors on Halal Monitoring
Services’ (HMS) website, and 3) the American Association of Meat Processors (AAMP). Poultry,
lamb, beef, and goat processors were included in the sample – I excluded processors that only
processed pork since the species is not acceptable for consumption in Islam. While the processors
listed on HMS’ website are known to be halal, it is also likely that some processors on the USDA
FSIS database and the AAMP membership list also process halal meat or poultry products, though
the exact number is unknown.
The USDA FSIS database lists all 6,788 USDA-inspected processing establishments.
Removing pork-only, siluriform-only, and egg-only facilities left 5,859 establishments. Of these
establishments, 2,736 are classified as “very small” by the USDA with less than 10 employees or
less than $2.5 million in annual sales, 2,656 are classified as “small” by the USDA with 10-499
employees and 440 are “large” establishments with 500 or more employees. I conducted stratified
random sampling of the three groups of establishments in the USDA FSIS data file, using Excel
to generate random number lists to select establishments from the populations. As very small,
small, and large processors make up 46.7%, 45.3%, and 7.5% of the total population, respectively,
these percentages were used to determine how many establishments to sample from each group.
There were 1,049 processors contacted from the USDA FSIS database including 20 large
processors, 451 small processors, and 578 very small processors. Establishments were called by a
team of undergraduate research assistants beginning in early November 2022 to determine who at
the establishment should respond to the survey and obtain email addresses. Email addresses for
11
establishments without a phone number listed or those that did not answer or return calls were
retrieved from business websites when available.
Individual Qualtrics survey URLs were sent via email to the USDA FSIS sample. The first
round of emails was sent using MS Word mail merge on December 9, 2022, with follow-up
reminder emails on December 13 and 16, 2022. The next reminder email was sent using Constant
Contact on January 11, 2023. From the first round of emails, approximately 50 bounced back as
undeliverable, and 12 businesses responded saying they did not qualify or would not be taking the
survey. Thus, a total of 987 processors from the USDA FSIS list received the survey.
There were an additional 58 registered processors on the HMS website which were all
included in the sample. Emails were obtained from the certifier, and individualized Qualtrics
survey URLs were sent using MS Word mail merge in January 2023, with follow-up reminder
emails in January and February 2023. Additionally, these processors were given the option to take
the survey in either Arabic or Urdu if they preferred.
The AAMP membership list was contacted via an association representative, who
distributed an anonymous Qualtrics survey URL to the membership email listserv in March 2023,
with reminder emails in March and April 2023. It is likely that many AAMP members received
the survey who also were included in my USDA FSIS recruitment efforts. However, the response
rate for both samples was very low, so I do not anticipate there were any duplicate survey
responses.
Despite contacting businesses in three different samples, I received only 195 total
responses, with only 95 remaining after data cleaning. I received responses mainly from very
small, small, and medium processing plants, both because these make up over 90% of meat
12
processors in the nation and, anecdotally, because larger processors typically do not respond to
surveys. Summary statistics and a discussion of valid responses is presented later in this chapter.
3.1.2 Retailer Sample and Data Collection
Non-halal and halal meat retailers were recruited between February 2023 and October
2023. Non-halal retailers were recruited via multiple state-level grocers associations and from a
membership list from the National Grocers Association (NGA). First, I attempted to recruit
retailers via the state-level grocers associations and the NGA email listservs with the assistance of
association representatives. However, only 18 responses were received via these efforts, so another
recruitment approach was needed. In May of 2023, a team of undergraduate research assistants
called retailers from the 2019 winter NGA membership list (National Grocers Association 2019)
– the most recent available online – between May 2023 and September 2023 to collect point of
contact email addresses. After removing closed businesses, a total of 946 retail stores were called,
and 236 email addresses were obtained. As in the processor case, it is possible that some of these
retailers actually did have a halal program at the time of the survey, though the exact number of
these stores is unknown.
Known halal retailers were recruited from halal certifiers’ online lists of registered
businesses and through a nationwide web scraping of Yellow Pages using the following key terms
and phrases: “halal meat grocery store,” “halal meat,” “Indo-Pak grocery,” “African grocery,” and
“international grocery store.” The results of the web scraping were compiled, then duplicates,
unrelated businesses, and closed businesses were removed from the list. A team of undergraduate
research assistants called the remaining 919 stores between July 2023 and October 2023 to collect
email contact information; 96 email addresses were obtained.
13
Emails with survey links were sent three times to each category of retailers between August
25 and October 10, 2023. Incentive payments of $25 were offered for complete and quality
responses, though not all respondents claimed their incentive. In total, 50 responses were collected
from the retailer samples, and after data cleaning, 39 viable survey responses remained.
4. Quantitative Analysis Methods
The Likert scale data collected from the surveys was analyzed using three methods: Count
data and descriptive statistics, principal component analysis (PCA), and k-means clustering. The
data in this study consists of two small samples from related populations – meat and poultry
processors and retailers. While there are certainly similarities between businesses within each of
these two populations (e.g., retailer-to-retailer or processor-to-processor), I am interested in
similarities and differences between businesses across these two populations (e.g., retailer-to-
processor). Specifically, I am interested in ascertaining classifying businesses by their patterns of
behavior so that generalizations based on these classifications can be made and used to prescribe
potential methods to increase halal meat market participation. Thus, I first present an overall
summary of the data collected by establishment type (processor or retailer) and halal status using
descriptive analysis. Then, I undertake a more complex statistical analysis by pooling the data and
conducting PCA and k-mean clustering on the combined data to generate variables (PCs) and sort
businesses based on these common characteristics (k-means).
4.1 Principal Components Analysis Statistical Framework
There are multiple methods, including econometric regressions, available to analyze Likert
scale data; however, the data in this study requires a method that can handle both multicollinearity
between large numbers of variables and a relatively small sample size. The analysis method that
best fits this description is PCA, which is commonly used to condense high-dimensional Likert
14
scale responses into a lower-dimensional form to identify overlapping variability in the sample.
PCA and k-means clustering are commonly used together for statistical analysis of multi-
dimensional data sets. K-means clustering has been used in conjunction with PCA to identify
market segments for agricultural products including seafood (Hanson et al. 1994), apples (Bejaei
et al. 2020), and beer (Malone & Lusk 2018, and has been popular in the marketing and data-
mining literature (Arabie & Hubert 1996). It is used to group similar participants without requiring
large sample sizes nor some of the more rigorous assumptions required by discrete choice
experiments and conjoint analysis (Arabie & Hubert 1996).
PCA was first developed in the early 1900s, when Pearson (1901) and Hotelling (1933)
endeavored to mathematically define patterns in data to describe large numbers of variables using
a smaller subset of independent variables. These independent variables – the principal components
– are chosen to maximize their explanatory power for the total variance of the original variables,
and the components that are derived in this way were termed ‘principal components.’ In the
successive century since PCA was developed, thousands of studies and papers across all fields of
science have used this method to condense complex data to discover important patterns (e.g., Hsu
et al. 2009, Sinha et al. 1969, Calder et al. 2001).
PCA is an orthogonal linear transformation; it translates higher-dimensional data into a
lower-dimensional coordinate system so that the first coordinate – that is, the first principal
component (PC1) – represents some linear projection of the greatest variance of the data.
Additional principal components (e.g., PC2 and PC3) would lie on the second, third, and so forth
coordinates with their corresponding second, third, and so forth greatest variances in the data. The
underlying concepts and procedures are illustrated mathematically below.
15
Suppose that I have a random vector:
𝑋 =
⎞
⎟
𝑋!
𝑋"
⋮
𝑋#⎠
⎛
⎜
⎝
with population variance-covariance matrix
𝑣𝑎𝑟(𝑋) = Σ = 1
" ⋯ 𝜎!#
𝜎!
⋮
⋮
⋱
"
𝜎#! ⋯ 𝜎#
5
Equation 1
Equation 2
Consider the linear combinations, each of which can be thought of as a linear regression
predicting 𝑌$ from 𝑋!, 𝑋", … , 𝑋#:
𝑌! = 𝑒!!𝑋! + 𝑒!"𝑋" + ⋯ + 𝑒!#𝑋#
𝑌" = 𝑒"!𝑋! + 𝑒""𝑋" + ⋯ + 𝑒"#𝑋#
⋮
𝑌# = 𝑒#!𝑋! + 𝑒#"𝑋" + ⋯ + 𝑒##𝑋#
Since the 𝑌$ are functions of random data, I have:
#
#
𝑣𝑎𝑟(𝑌$) = ;
%&!
; 𝑒$%𝑒$’𝜎%’
’&!
= 𝑒$
( ; 𝑒$
#
#
𝑐𝑜𝑣>𝑌$, 𝑌)? = ;
%&!
; 𝑒$%𝑒)’𝜎%’
’&!
= 𝑒$
( ; 𝑒)
Equation 3
Equation 4
Now, the first principal component (PC1) accounts for as much variation in the data as
possible. It is expressed as the linear combination of x-variables that has maximum variance among
16
all the linear combinations. I must maximize the variance subject to the constraint that the sum of
the squared coefficients 𝒆𝟏 equals one. Expressed mathematically this is:
#
#
𝑣𝑎𝑟(𝑌!) = ;
%&!
; 𝑒!%𝑒!’𝜎%’
’&!
= 𝑒!
( ; 𝑒!
such that:
#
"
(𝑒! = ; 𝑒!%
𝑒!
%&!
= 1
Equation 5
Equation 6
Additional principal components are computed in much the same way, with the added
constraint that the components are uncorrelated. That is, for the ith PC, I find the vector 𝒆𝒊 of
coefficients to solve:
such that
and
Equation 7
Equation 8
#
#
𝑣𝑎𝑟(𝑌$) = ;
%&!
; 𝑒$%𝑒$’𝜎%’
’&!
= 𝑒$
( ; 𝑒$
#
"
(𝑒! = ; 𝑒!%
𝑒!
%&!
= 1
#
#
𝑐𝑜𝑣(𝑌!, 𝑌$) = ;
%&!
; 𝑒!%𝑒)’𝜎%’
’&!
= 𝑒!
( ; 𝑒$ = 0
#
#
𝑐𝑜𝑣(𝑌", 𝑌$) = ;
%&!
; 𝑒"%𝑒)’𝜎%’
’&!
= 𝑒"
( ; 𝑒$ = 0
17
⋮
#
#
𝑐𝑜𝑣(𝑌$,!, 𝑌$) = ;
%&!
; 𝑒$,!%𝑒)’𝜎%’
’&!
= 𝑒$,!
( ; 𝑒$ = 0
Equation 9
To solve for the coefficients 𝑒$), I must use the eigenvalues and eigenvectors of the
variance-covariance matrix Σ. Letting 𝜆! ≥ 𝜆" … ≥ 𝜆# be the eigenvalues of Σ and 𝒆𝟏, 𝒆𝟐, … , 𝒆𝒑
be the corresponding eigenvectors, I have that the elements for the eigenvectors are the PC
coefficients. Additionally, the variance of the ith PC is equivalent to the ith eigenvalue:
𝑣𝑎𝑟(𝑌$) = 𝑣𝑎𝑟>𝑒$!𝑋! + 𝑒$"𝑋" + ⋯ + 𝑒$#𝑋#? = 𝜆$
Now, to obtain the principal components of a sample of our data, I must compute the
corresponding sample eigenvalues 𝜆E
$ and eigenvectors 𝒆F 𝒊 of the sample variance-covariance
matrix S. I can then define each estimated PC as a linear combination using the eigenvectors as
Equation 10
coefficients:
𝑌G! = 𝑒̂!!𝑋! + 𝑒̂!"𝑋" + ⋯ + 𝑒̂!#𝑋#
𝑌G" = 𝑒̂"!𝑋! + 𝑒̂""𝑋" + ⋯ + 𝑒̂"#𝑋#
⋮
𝑌G# = 𝑒̂#!𝑋! + 𝑒̂#"𝑋" + ⋯ + 𝑒̂##𝑋#
In PCA, I seek to retain minimal PCs so that the proportion of the variance is described
by the first g PCs is large and ideally close to one. That is:
Equation 11
18
! + 𝜆E
𝜆E
! + 𝜆E
𝜆E
" + ⋯ + 𝜆E
/
" + ⋯ + 𝜆E
#
≅ 1
This allows us to obtain the simplest possible relationship between the original data and
the new, condensed variables (the PCs). If the first few PCs explain a small amount of variation, I
need more of them to explain a desired percentage of total variance resulting in a large g. To
determine the number of PCs for this study, I retained any PC that accounts for at least 10% of the
Equation 12
common variance (Malone and Lusk 2017).1
4.2 K-means Clustering Motivation and Statistical Framework
K-means clustering is advantageous for analyzing the data in this study because it is
straightforward to implement and interpret – preferable attributes for performing an exploratory
study. Additionally, k-means clustering is ideal as it segments the market into groups that exhibit
common views.
Given the exploratory nature of this study, I conduct simple nonhierarchical cluster analysis
where individuals are grouped using the least squares method to minimize the Euclidean distance
within a specified k number of groups or clusters. This algorithm partitions the data space in a way
so that data points within the same cluster are as similar as possible (intra-class similarity) with
respect to their PC scores, while data points from different clusters are as dissimilar as possible
(inter-class similarity). In k-means, each cluster is characterized by its centroid, which is the
arithmetic mean of the data points assigned to the cluster, but it might not be a member of the
dataset. K-means randomly selects data points as possible centroids of the clusters and then
1 I used the principal components analysis commands in Stata to conduct this analysis, using the nine common
Likert scale questions between the processor are retailer surveys as the independent variables for which to develop
the PCs from. I did not use an orthogonal rotation method, as I am not using PCA to compare across different scales
or groups of questions in the data. Specifically, the code used was: pca PricePremium - LackKnowledge
19
iteratively recalculates new centroids to converge to a final clustering of the data points. K-means
assigns every data point in the dataset to the nearest centroid, meaning that a data point is in a
given cluster if it is closer to that cluster’s centroid than any other centroid. The algorithm repeats
the selection of centroids and sorting of data points until the sum of distances between data points
and their given centroid is minimized, the maximum number of iterations is reached, or there are
no changes in centroid values.2
The Likert scale data resulted in a 9-dimensional space. I indexed this data using the three
PCs, and then partition the observations into three groups (clusters) based on their PC values. As
such, the specific objective of this cluster analysis is to minimize the within-group variation of the
PC values. Specifically, the k-means method in our context minimizes the squared distance from
each observation (𝑥$) to the center of the observation’s associated cluster (𝑋K$0):
2
min(𝑑𝑖𝑠𝑡𝑎𝑛𝑐𝑒1) = 𝑚𝑖𝑛U;(𝑥$ − 𝑋K$0)"
$&!
Equation 13
Once I select the k PCs that best describe most of the variance, I use these PCs as
independent variables on which to cluster or “sort” individuals into groups based on their
individual predicted PC values. After condensing the data in this manner, k-means clustering is
applied to group the observations into clusters that exhibit similar response patterns. That is, rather
than grouping variables as is done in PCA, cluster analysis allows us to group people – or in our
case, businesses. In this study, cluster analysis will combine all participants’ responses and then
2 I conducted k-means cluster analysis in Stata using the cluster kmeans command on the three PCs selected from
the PCA to sort the data into three groups with a set starting seed value and a maximum of 1,000,000 iterations. In
particular, the code used is: cluster kmeans pc1 pc2 pc3, k(3) name(threegroups) start(random(12281996))
iterate(1000000)
20
sort them into new groups; I anticipate this segmentation will group some non-halal processors
and retailers with their halal counterparts, revealing what common characteristics of non-halal
businesses are most suitable to future expansion into the halal market.
To determine the unique characteristics of the different market segments created via k-
mean clustering, F-tests on the means between groups for each demographic category are
conducted. Values with 10% or stronger statistical significance are then evaluated with pairwise t-
tests to discover more specific differences between the groups. Ultimately, this will allow for
policy prescriptions tailored to their unique positions in the market that are not easily seen when
analyzing participants within their original (halal versus non-halal) categories.
5. Summary Statistics
The summary statistics of respondents are given by business type in the following
subsections.
5.1 Processor Summary Statistics
The processor survey received 195 responses. After dropping ineligible and responses that
were less than 50% complete, 95 responses remained. Establishments represented in this data are
in 39 states.3
5.1.1 Processor Respondent Characteristics
Summary statistics of processor respondent characteristics are given in Table 1.
Respondents were mainly male and white. The majority had at least a 4-year college degree. When
asked their political affiliation, most said they were Republicans or Independents, although 31%
3 The states represented in the data are Alabama, Arkansas, California, Colorado, Connecticut, Florida, Georgia,
Illinois, Indiana, Kansas, Kentucky, Louisiana, Maine, Massachusetts, Michigan, Minnesota, Mississippi, Missouri,
Montana, Nebraska, Nevada, New Jersey, New York, North Carolina, North Dakota, Ohio, Oklahoma, Oregon,
Pennsylvania, Rhode Island, South Dakota, Tennessee, Texas, Utah, Virginia, Washington, West Virginia,
Wisconsin, and Wyoming.
21
opted not to disclose their political affiliation. Twenty-one percent indicated they were second-
generation immigrants (i.e., their parent(s) were born outside of the U.S., but they were born in the
U.S.). Immigration information is relevant to this study, as more recent immigrants or younger
generations of immigrants are more likely to identify with and participate in the cultural and
religious heritage from their home or ancestral country (Lopez 2017). The most recent waves in
immigration of Muslims to the U.S. occurred in the 1980s and in the 2000s onward, as the U.S.
relaxed immigration laws and civil unrest and climate catastrophes in the Arab world and
Southeast Asia contributed to a refugee crisis (Carmichael 2020); as such, it is possible that more
younger generations of immigrants are active in the halal meat and poultry supply chain. Further,
21% of the sample indicated they are religious. As halal is a religious product, this information is
potentially important for understanding processors’ motivations - or lack thereof - for providing
halal products.
22
Table 1: Processor Respondent Summary Statistics
Category
Gender (n = 94)
Male
Female
Prefer not to disclose
Education Level (n = 94)
High School
Some College
2-Year Degree (Associates)
4-Year Degree (Bachelor's)
Master's Degree
Professional Degree
Race (n = 94)
White
Black
Native American or Alaskan Native
Native Hawaiian or Pacific Islander
Asian
Other
Prefer not to disclose
Hispanic (N = 93)
No
Yes
Prefer not to disclose
Political Party (N = 94)
Democrat
Republican
Independent
Other
Prefer not to disclose
2nd Generation Immigrant (N = 94)
No
Yes
Prefer not to disclose
Currently Religious (N = 94)
Yes
No
Prefer not to disclose
Previously Religious (N = 20)
Yes
No
Prefer not to disclose
23
Percent
76%
19%
5%
6%
13%
10%
44%
22%
5%
76%
1%
0%
0%
3%
6%
14%
89%
3%
8%
5%
35%
26%
3%
31%
66%
21%
13%
21%
56%
22%
45%
40%
15%
5.1.2 Processor Establishment Characteristics
Processing establishment summary statistics are given in Table 2. Establishments
represented in the sample consisted of facilities that slaughtered, processed, both slaughtered and
processed, or conducted other meat and poultry product handling (“other”). Seven percent of
respondents only slaughtered meat or poultry, 44% of respondents only processed meat or poultry,
45% both slaughtered and processed meat or poultry products, and 3% of respondents conducted
other meat and poultry product handling. Most respondents represented very small or small
processors; the mean and median business size of our sample was 62 and 20 employees,
respectively. This is not surprising, as over 95% of meat and poultry processing establishments in
the U.S. are small or very small per USDA standards. On average, these establishments had been
open for 31 years. Unsurprisingly, given the nature of meat and poultry processing, 48% of
respondents indicated their establishment is in a rural area.
Many slaughter and processing establishments handle multiple species. Of the 95
establishments in our sample, 83% indicated they handled beef, 34% veal, 55% lamb, 46% pork,
35% turkey, 38% chicken, 55% goat, and 35% handled exotics (game, specialty poultry, etc.).
Respondents were asked if they currently or had ever slaughtered, processed, or handled
halal meat or poultry products. Thirty-four percent responded they were currently operating a halal
program, 4% had previously had a halal program but did not currently, and 62% indicated they
had never operated a halal program at their establishment.
24
Table 2: Processing Establishment Summary Statistics, N = 95
Category
Establishment Type
Slaughter without processing
Processing without slaughter
Slaughter and processing
Other
Location
Rural
Suburban
Urban
Prefer not to disclose
Type of Animal
Beef
Veal
Lamb
Pork
Turkey
Chicken
Goat
Exotics
Halal Status
Current Halal
Past Halal
Never Halal
Number of Employees Mean Median
Year Established
62
20
Mean Median
1992
2001
Percent
7%
44%
45%
3%
48%
23%
23%
6%
83%
34%
55%
46%
35%
38%
55%
35%
34%
4%
62%
Min
2
Min
1902
Max
850
Max
2022
Looking more specifically at the 32 establishments that reported currently operating a halal
program, over three-quarters indicated they have a zabiha (hand-slaughtered) halal program (Table
3). Further, over one-half of these establishments reported operating a halal program for more than
7 years. Most of these establishments are certified by a third-party halal certification group, which
is not surprising given our sampling strategy. For operations that conduct halal slaughter or
processing, most reported that halal slaughter or processing was over 50% of their operation.
25
Table 3: Current Halal Establishment Summary Statistics, N = 32
Category
Halal slaughter method
Zabiha
Machine
Unsure/Don’t know
Years operating a halal program
< 1 year
1-3 years
4-6 years
7-10 years
11-20 years
> 20 years
Certified by 3rd party
Yes
No
Unsure
Percent of slaughter that is halal
< 10%
11-25%
26-50%
50-75%
> 75%
Percent of processing that is halal
< 10%
11-25%
26-50%
50-75%
> 75%
Percent
77%
6%
16%
6%
29%
13%
16%
13%
23%
78%
16%
6%
9%
4%
4%
74%
9%
24%
12%
4%
4%
56%
5.2 Retailer Summary Statistics
The retailer survey received 50 responses. After dropping ineligible responses and
responses that were less than 50% complete, 41 responses remained. The survey was distributed
nationwide; stores represented in this data are located in 21 states.4
4 The states represented in the data are Alaska, California, Hawaii, Illinois, Indiana, Maine, Maryland,
Massachusetts, Michigan, Montana, Nebraska, New Jersey, North Carolina, North Dakota, Ohio, Rhode Island,
South Dakota, Texas, Utah, Virginia, and Wisconsin.
26
5.2.1 Retailer Respondent Characteristics
Retailer respondent demographic summary statistics are given in Table 4. Like in the
processor survey, respondents were mainly male and white, and the majority had at least a
bachelor’s degree. When asked their political affiliation, most indicted they were Republicans or
Independents, although 24% opted not to disclose their political affiliation. Fifteen percent
indicated they were second generation immigrants (i.e., their parent(s) were born outside of the
U.S., but they were born in the U.S.). Further, 36% of the sample indicated they are currently
religious.
27
Table 4: Retailer Respondent Demographic Summary Statistics
Category
Gender (n = 33)
Male
Female
Education Level (n = 32)
High School
Some College
2-Year Degree (Associates)
4-Year Degree (Bachelor’s)
Master’s Degree
Professional Degree
Race (n = 36)
White
Black
Native American or Alaskan Native
Native Hawaiian or Pacific Islander
Asian
Other
Prefer not to disclose
Hispanic (n = 33)
No
Yes
Political Party (n = 33)
Democrat
Republican
Independent
Other
Prefer not to disclose
1st or 2nd Generation Immigrant (n = 33)
No
Yes
Prefer not to disclose
Currently Religious (n = 33)
Yes
No
Prefer not to disclose
Previously Religious (n = 12)
Yes
No
Percent
79%
21%
3%
13%
28%
25%
22%
9%
66%
0%
5%
2%
10%
2%
2%
97%
3%
6%
27%
36%
6%
24%
82%
15%
3%
36%
48%
15%
58%
42%
28
5.2.2 Retail Store Characteristics
Summary statistics of the retail stores in our sample are given in Table 5. Respondents
consisted of grocery stores (78%), butcher shops or delis (14%), and “other” (8%) retail stores –
likely ethnic grocery store-restaurant hybrids, which are common in areas with high ethnic
populations. Most retailers represented in the sample were very small or small independent
retailers, with an average of 45 employees. This is not surprising, as we sampled mainly from the
National Grocers Association membership list. Further, approximately 33% of retail grocery stores
in the U.S. are small or very small independent retailers, per the National Grocers Association. On
average, these stores have been open for 44 years. Over half of respondents indicated their store is
in a rural area.
Respondents were asked if they currently or had ever sold halal meat or poultry products:
22% responded they were currently selling halal meat or poultry products, 10% had previously
sold halal meat or poultry products, 51% had never sold halal meat or poultry products at their
store, and 17% were unsure if they had ever sold halal. For the purposes of the survey and this
analysis, those who were unsure if they had ever sold halal were classified as “never sold halal”
for further questions about their operations.
29
Table 5: Retail Store Summary Statistics (n = 41)
Participant Type
Grocery Store
Butcher Shop/Deli
Other
Rural
Suburban
Urban
Currently selling halal
Previously sold halal
Never sold halal
Unsure if ever sold halal
Number of Employees
Frequency
33
5
3
20
11
2
9
4
21
7
Mean
45
Mean
1979
Year Established
Percentage
80%
12%
7%
61%
33%
6%
22%
10%
51%
17%
Min
600
Max
2021
Min
1
Min
1867
Retail food stores sell multiple species of meat or poultry; this information is presented in
Table 6. Of the 41 stores in our sample, 100% indicated they sold beef, 41% veal, 56% lamb, 85%
pork, 95% turkey, 100% chicken, 17% goat, and 34% exotics (game, specialty poultry, etc.). For
the 13 stores that currently sell or previously sold halal meat or poultry products, 62% sold halal
beef, 38% halal veal, 46% halal lamb, 38% halal turkey, 92% halal chicken, 54% halal goat, and
8% halal exotics.
Table 6: Percent of Retail Stores that Carry Different Species of Meat and Poultry
Type of Animal
Beef
Veal
Lamb
Pork
Turkey
Chicken
Goat
Exotics
Current and Past Halal Stores Only, n = 13
62%
38%
46%
0%
38%
92%
54%
8%
All Stores, n = 41
100%
41%
56%
85%
95%
100%
17%
34%
The summary statistics for the nine current halal retailers are given in Table 7. Most of
these stores sell zabiha (hand-slaughtered) halal products. Further, all stores in the sample had sold
30
halal meat and poultry products for less than a decade at the time of the survey. Most of these
stores are certified by a third-party halal certification group. However, halal meat and poultry sales
make up less than 50% of each store’s overall meat and poultry sales.
Table 7: Current Halal Retailer Summary Statistics (n=9)
Category
Zabiha
Machine
Unsure/don’t know
Years Selling Halal Meat & Poultry
< 1 year
1-3 years
4-6 years
6-10 years
11-20 years
> 20 years
Store is Halal Certified
Yes
No
Unsure
Percent of Meat & Poultry Sales that
are Halal
< 10%
11-25%
26-50%
50-75%
> 75%
Frequency Percent
7
1
2
2
1
3
3
0
0
5
2
2
1
2
6
0
0
70%
10%
20%
22%
11%
33%
33%
0%
0%
56%
22%
22%
11%
22%
67%
0%
0%
6. Results & Discussion
The results of the analyses presented in this paper shed light on potential avenues to
increase the availability of halal meat and poultry products. In the following subsections, results
are given by analysis method – Likert scales, PCA, and k-means clustering - and related discussion
is included.
31
6.1 Processors’ and Retailers’ Perceptions of Motivations and Barriers
Likert scales were used to assess processors’ and retailers’ perceptions of motivations
(incentives) and barriers (disincentives) in the decision to establish a halal program at their
operation. I am interested in the differences between businesses who currently or have never
offered halal meat or poultry products, as these differences may indicate avenues for supporting
increased market entry into the halal processing and retailing industry. Note, I do not report
specific statistics for past halal processors and retailers out of this analysis, as this group includes
only eight businesses.
6.1.1 Likert Scales: Motivations and Incentives
Retailers and processors were asked to indicate which motivations or incentives they
considered for currently or potentially selling or processing halal meat or poultry products. An
example of these questions for retailers and processors are included above in Figure 1 and Figure
2, respectively. In answering these questions, participants selected how likely each option was or
would be an incentive or motivation for adding a halal program to their operations.
The results of these groups of Likert questions are in Table 8, reported as percentages.
Results are presented by type of establishment – retailers and processors, halal and non-halal – and
by level of incentive in decision-making.
32
Table 8: Results of Likert Scale Questions for Halal Incentives
Incentives
Considered for
Halal
Higher Price
Access New
Markets
Supply Minority
Communities
Financial Aid to
Establish
Assistance from
Organizations
Compete with
Similar
Businesses
Likert
Scale
Response
Likely
Neutral
Unlikely
Likely
Neutral
Unlikely
Likely
Neutral
Unlikely
Likely
Neutral
Unlikely
Likely
Neutral
Unlikely
Likely
Neutral
Unlikely
Halal
Retailers
n = 9
0%
33%
67%
67%
11%
22%
0%
89%
11%
N/A
N/A
N/A
33%
22%
44%
56%
11%
33%
Non-Halal
Retailers
n = 26
31%
38%
31%
42%
31%
27%
19%
54%
27%
N/A
N/A
N/A
23%
31%
46%
8%
50%
42%
All Retailers
n = 39
20%
37%
39%
49%
22%
24%
15%
59%
22%
N/A
N/A
N/A
27%
27%
41%
17%
41%
37%
Halal
Processors
n = 18
26%
37%
37%
78%
17%
6%
56%
28%
17%
11%
11%
78%
17%
22%
61%
N/A
N/A
N/A
Non-Halal
Processors
n = 57
19%
32%
49%
28%
25%
47%
16%
37%
47%
30%
25%
46%
30%
28%
42%
N/A
N/A
N/A
All
Processors
n = 79
21%
34%
45%
40%
24%
35%
25%
36%
38%
25%
23%
51%
26%
28%
45%
N/A
N/A
N/A
33
Nine retailers and 18 processors reported that they are currently supplying halal meat or
poultry products and answered the Likert scale questions in the survey. When making the decision
to offer halal meat or poultry products at their business, the ability to sell halal products at a higher
price than conventional products was not an incentive for current halal retailers and processors.
However, the ability to access new markets was a strong incentive for both retailers and processors.
Surprisingly, the ability to supply minority communities was not an incentive for halal retailers,
with most current halal retailers indicating it was a neutral factor in their decision. This may be
due to current halal retailers operating in areas of the country in which there are large Muslim
populations, in which these retailers may not see Muslims as a minority in their area. Indeed,
according to the Public Religion Research Institute (2023), Muslim-Americans are mainly
concentrated in major metropolitan areas. However, the ability to supply products for minority
communities was generally a strong incentive for current halal processors. Obtaining assistance
from outside organizations to set up a halal program at their stores was in general a slight incentive
for current halal retailers, and obtaining financial aid and assistance setting up a halal program
were not incentives that were valued by current halal processors. Finally, the ability to compete
with similar businesses was generally a strong incentive for current halal retailers and a slight
incentive for halal processors.
For the 26 retailers and 57 processors that have never had a halal program, I asked which
of these six factors would incentivize them to add a halal program. The majority indicated that all
six factors would not be an incentive or would be a neutral factor in their decision. Access to new
markets (42%) was the factor most likely to incentivize retailers to add a halal program to their
store – determined by the overall percentage of the time retailers selected the factor as likely was.
This is unsurprising, as it is likely that retailers that have never sold halal products are not members
34
of the Muslim community; most retailer respondents in our survey who said they are religious
indicated they were not Muslim. Therefore, they are unlikely to have considered adding halal
products to their stores and may not have known what halal was prior to receiving this survey.
Indeed, several survey respondents indicated in the comments box they did not know what halal
was prior to taking the survey. The factors most likely to incentivize processors to add a halal
program to their establishment were financial aid to establish a halal program (30%), assistance
from organizations to set up a halal program (30%), and the ability to access new markets (28%),
though these are not practically significant figures.
6.1.2 Likert Scales: Barriers and Disincentives
Participants were also asked to indicate which potential barriers or disincentives they faced
or considered for currently or potentially offering halal meat or poultry products at their operations.
An example of this question for retailers is given above in Figure 3 and an example of this question
for processors is given in Figure 4. When answering these questions, participants could select how
likely each option was or would be a barrier in their decision to supply or not supply halal meat or
poultry products.
The descriptive statistics of these responses are given in Table 9, reported as percentages.
Results are presented by type of establishment – retailers and processors, halal and non-halal – and
by level of disincentive in decision-making.
35
Table 9: Results of Likert Scale Questions for Halal Disincentives
Barriers Considered
for Halal
Likert Scale
Response
Costs to Participate
Discrimination from
Regulators
Backlash from Non-
Muslims
Lack of Muslim Labor
Limited Local Market
Lack of Halal
Knowledge
Unlikely
Neutral
Likely
Unlikely
Neutral
Likely
Unlikely
Neutral
Likely
Unlikely
Neutral
Likely
Unlikely
Neutral
Likely
Unlikely
Neutral
Likely
Current
Halal
Retailers
n = 9
44%
33%
22%
67%
33%
0%
100%
0%
0%
N/A
N/A
N/A
44%
11%
44%
56%
44%
0%
Current
Halal
Processors
n = 19
56%
31%
13%
72%
22%
6%
72%
22%
6%
11%
50%
39%
68%
5%
26%
72%
6%
22%
Non-Halal
Processors
n = 57
All
Processors
n = 77
19%
33%
47%
53%
44%
4%
56%
39%
5%
25%
35%
40%
23%
30%
47%
21%
39%
40%
26%
31%
43%
56%
41%
4%
59%
35%
5%
20%
39%
41%
34%
25%
41%
32%
33%
35%
Non-Halal
Retailers
n = 26
All
Retailers
n = 39
26%
26%
38%
49%
36%
5%
62%
23%
5%
N/A
N/A
N/A
15%
8%
67%
28%
36%
26%
21%
25%
46%
46%
39%
7%
54%
32%
7%
N/A
N/A
N/A
7%
7%
79%
21%
36%
36%
36
For the nine halal retailers and 19 halal processors in our sample, costs to participate in a
halal certification program to sell products at their businesses were not a major barrier for retailers
nor processors. Evidence from interviews conducted in the first phase of this project indicated that
multiple halal processors have experienced discriminatory or hostile behavior from USDA meat
and poultry plant inspectors. However, the results of the national online survey show that many
current halal retailers, processors, and regulators did not consider discrimination from regulators
or inspectors to be a barrier to halal production.
Halal retailers and processors were not concerned about backlash from their non-Muslim
customers. However, while some halal processors slaughter exclusively halal, some of the meat
and poultry processing plants interviewed for this project that have a halal program said that they
run halal on their processing lines relatively infrequently – from once or twice a month to once
every few months – while the rest of their processing is for non-halal customers. Lack of Muslim
labor was a concern for some halal processors. In interviews, some processors said they must
contract with their certifier to bring in a Muslim slaughter person for the days they run halal on
their line and cited reasons related to their local area (e.g., rural non-Muslim communities). For
current halal retailers, a limited local market was equally indicated to be a barrier and not a barrier.
Contrastingly, a limited local market was not a barrier for processors; this is not surprising, as meat
processors typically distribute their products outside of their local market, regardless of whether
they are halal or not. Finally, a lack of halal knowledge was not a challenge faced by current halal
retailers when establishing the program at their store – again, likely because many of these retailers
are themselves Muslim. However, a lack of halal knowledge was a challenge faced by 22% of
halal processors when establishing the program at their establishment – this may be because of the
37
numerous and rigorous standards that must be upheld for halal slaughter. It may also be because
many of these processors run halal infrequently and are likely not themselves Muslim.
The 26 retailers and 57 processors who have never offered halal meat or poultry products
at their businesses in general differ from halal retailers and processors in what challenges they
perceive they would face if they were to offer halal meat or poultry products at their establishments
in the future. These retailers and processors perceive that the costs to participate in a halal certified
program would be prohibitive. Discrimination from regulators and backlash from non-Muslim
consumers are not seen as major challenges for retailers nor processors. Retailers’ and processors’
lack of concern about discrimination and backlash is likely because these businesses are, in
general, not themselves ethnic or religious minorities, and therefore are unlikely to be worried
about this behavior. Non-halal retailers are strongly concerned about the lack of a local market for
halal meat and poultry products, as 79% of them indicated this would be a barrier for their store to
offer halal, while processors are not as concerned about the lack of a local market. This makes
sense, as meat and poultry products are commonly shipped across state lines from processors to
retail stores, while retailers rely on customers near their stores. Finally, non-halal retailers and
processors indicated that a lack of knowledge about halal meat and poultry would be a disincentive
or a neutral factor in their decision to offer halal meat or poultry products in the future.
6.2 Principal Components Analysis and K-Means Clustering
PCA was conducted on the combined processors’ and retailers’ data sets, again excluding
past halal retailers. This allows us to analyze patterns across all halal and non-halal businesses
together and potentially reveal commonalities that point to outreach methods or policy design to
facilitate halal market expansion. Additionally, PCA facilitates analysis of how participants
38
responded to all questions, instead of looking at each question’s response individually. This allows
me to better pinpoint participants’ overall attitudes and opinions towards halal business decisions.
In the data, an increasing Likert scale value for a motivation (incentive) indicates an
increase in the attractiveness of that motivation; that is, higher values correspond to a business
feeling more incentivized. Contrastingly, an increasing Likert scale value for a barrier
(disincentive) indicates an increase in how prohibitive it is; that is, higher values correspond to a
business feeling more disincentivized. Only the Likert scale questions that were included on both
the retailer and the processor surveys were used for the PCA.
The results of PCA for this analysis are given in Table 10 and Table 11. A PC was retained
if it accounted for at least 10% of the variation in the data. The PCA resulted in three latent factors
representing a total of 71.84% of the variance in the original dataset; PC1 represented 37.29% of
the variation, PC2 represented 22.35%, and PC3 represented 12.20%. Thus, there are now three
new variables (PC1, PC2, and PC3) that contain information and patterns between the nine
variables in the original dataset (the motivations and barriers). These three new variables now
facilitate more straightforward analysis of the responses, as they represent the three major trends
or themes in the data in which we can classify behavior. I call these new variables “Concerns”
(PC1), “Motivators” (PC2), and “Niche Minority Considerations” (PC3) and explain their
elements below.
PC1, “Concerns,” has positive correlations of approximately equal magnitude with all the
features of a business’ decision of whether to offer halal products; thus, businesses with high values
of PC1 are equally motivated by the incentives and dissuaded by the barriers. PC2, “Motivators,”
has a positive correlation with all motivations and a negative correlation with all barriers; thus,
businesses with high values for PC2 consider the motivations to be attractive and the barriers to
39
not be prohibitive in their decision-making process. PC3, “Niche Minority Considerations,” has a
positive correlation with the first three motivations: Price Premium, Access to New Markets, and
Providing a Religious Minority Product, as well as two of the barriers: Discrimination from
Regulating Bodies and Backlash from Non-Muslim Customers. It follows that businesses with
high values of PC3 have mixed opinions on the motivations and barriers, specifically those related
to supplying religious minority products when deciding to offer halal at their business.
Table 10: Eigenvalues and Proportion of Variance for Principal Components Analysis of
Retailer and Processor Data, n = 109
Principal Components Metrics
Eigenvalues
Proportion of Variance
PC1
3.36
37.29%
PC2
2.01
22.35%
PC3
1.10
12.20%
PC1
Price Premium
Motivation or Barrier
Table 11: Loadings for Principal Components Analysis of Retailer and Processor Data, n =
109
Motivation
or Barrier
Motivation
Motivation Access to New Markets
Motivation
Motivation Technical Assistance Setting Up
Barrier
Barrier
Barrier
Barrier
Barrier
Halal Certification Program Costs
Discrimination from Regulating Bodies
Backlash from Non-Muslim Customers
Limited Local Market Opportunities
Lack of Halal Knowledge
0.26
0.47
0.47
0.29
-0.17
-0.36
-0.40
-0.22
-0.18
0.04
0.13
0.32
-0.12
-0.40
0.49
0.50
-0.40
-0.21
0.38
0.28
0.21
0.36
0.37
0.31
0.28
0.34
0.41
Providing a Religious Minority Product
PC2
PC3
K-means clustering was then conducted using the three PCs from the PCA as the
underlying factors on which to sort the respondents. Three groups (k = 3) were chosen due to the
small number of responses, as conducting k-means with more groups in each data set resulted in
at least one group with very small membership relative to the other groups.
Table 12 gives the group sizes and their mean PC values. Group 1 had 35 members, with
the most negative average values for PC1 “Concerns” (-1.82), PC2 “Motivators” (-1.03), and PC3
40
“Niche Minority Considerations” (-0.23) of the three groups, though the magnitude of the PC3
averages is small. The negative values of all three PCs combined suggest that the motivations are
strongly negatively weighted for Group 1, while the weight placed on the barriers is close to zero
in magnitude. Group 1 is therefore least likely to find the motivations for offering halal meat and
poultry products to be attractive and are neutral towards the barriers to offering halal. When
comparing their opinions to the operational and demographic data (Table 13), the findings are
logical; these are the smallest businesses on average and contain a large percentage of rural
businesses, as well as a large percentage of non-halal businesses and a small percentage of halal
businesses. Taken together, it is possible that these smaller, more rural businesses do not have the
local market to sell halal products to or may not have the capacity to add another program to their
establishment. As such, I would not expect these businesses to be strong candidates for adding a
halal program to their operation in the future.
Group 2 had 28 members, with the most positive average value (1.56) for PC2
“Motivators” of the three groups. The average values for Group 2’s PC1 “Concerns” (-0.42) and
PC3 “Niche Minority Considerations” (0.07) are the mid-range value relative to Groups 2 and 3,
though PC3 is relatively small in magnitude (Table 12). Thus, Group 2 was most strongly
positively influenced by the motivations and negatively influenced by the barriers in PC2.
However, Group 2’s PC1 magnitude indicates a slight counteracting negative influence to PC2.
Therefore, Group 2 contains businesses that are most likely to find motivations for offering halal
to be attractive and the barriers to be prohibitive. The conflicting negative and positive influence
of PC2 and PC1 indicate the barriers are overall more prohibitive in Group 2’s decision-making
than the motivations are attractive. Group 2 contains the highest percentage of retailers, the highest
percentage of halal businesses, the lowest percentage of non-halal businesses, are the largest on
41
average, the most recently established, and are more likely to be in urban or suburban areas than
the other two groups (Table 13). Since Group 2 has the highest percentage of halal businesses, it
is likely that businesses in this group that do not currently offer halal products are good candidates
to do so in the future, as they must have other qualities in common that align with halal market
participation. Indeed, these businesses are more likely to be in areas with larger Muslim consumer
populations and may have the capacity to add an additional program to their operation.
Additionally, retailing of halal products requires much less capital investment and knowledge of
the halal slaughter process than what is required for slaughterhouses and processors, and therefore
may allow these businesses to adopt halal more easily. However, as these businesses are slightly
younger on average, they may not have the capital to support expanding into a new market, which
may explain their larger concerns with the barriers to market entry. Additionally, it is worth noting
that Group 2 is the smallest group, with only 13 non-halal businesses, meaning the proportion of
businesses that are good candidates to adopt a halal program in the future is small relative to the
overall sample of this study.
Group 3 had 46 members, with the most positive values for PC1 “Concerns” (1.64), and
mid-range values for PC2 “Motivators” (-0.17) and PC3 “Niche Minority Considerations” (0.13)
that are relatively small in magnitude. Thus, Group 3 was most strongly influenced by PC1, with
slight counteracting effects from PC2 and PC3. Group 3 therefore contains businesses that are
likely to find the motivations for halal to be attractive and the barriers to be prohibitive, with the
influence of the motivations and barriers fairly equal and opposite each other relative to Group 2.
Group 3’s businesses are on average the oldest of the three groups, have the highest percentage of
processors, and a relatively high percentage of retailers. Retailing and processing halal products is
logistically simpler than slaughtering, which would make adoption of a halal program more
42
straightforward for these businesses than groups with more slaughterhouses. Additionally, older
businesses are likely well-established and could have more access to capital resources to expand
their operations. Thus, these businesses may be good candidates for expanding into the halal meat
and poultry business in the future if given enough information on the motivations and barriers
specific to their operations (e.g., a feasibility study).
Table 13 gives operation and respondent demographic percentages across the three groups
of businesses. Note that not all percentages sum to 100% within a group’s category (e.g., race), as
some respondents declined to answer all survey questions. From the F-tests and resulting t-tests, I
find that there exist significant differences for non-halal, current halal, only processors, number of
employees, at most high school diploma holders, and Asians. There are statistically significant
differences at the 1% level between Groups 1 and 2 for the percentages of non-halal and current
halal businesses, and a difference at the 5% level for the number of employees. There also exists
a statistically significant difference between Groups 1 and 3 at the 5% level for high school
diploma holders. Finally, there exists a statistically significant difference between Groups 2 and 3
at the 1% for halal and non-halal businesses, and at the 5% level for number of employees, Asian
respondents, and high school diploma holders. Altogether, I see the most distinct operational
demographic patterns between Groups 1 and 3 versus Group 2.
Table 12: K-Means Groups and Means of PC1, PC2, and PC3, Retailers and Processors, n
= 109
Group
1
2
3
Number of Members Mean of PC1 Mean of PC2 Mean of PC3
35
28
46
-1.82
-0.42
1.64
-1.03
1.56
-0.17
-0.23
0.07
0.13
43
Table 13: Demographic Composition of Groups, n = 109
Demographics
Group 1
(n = 35)
Group 2
(n = 28)
Group 3
(n = 46)
85.71%a***
14.29% a***
34.29%
5.71%
28.57%
25.71%
1991
28a**
51.43%
20.00%
8.57%
86.96% c***
13.04% c***
21.74%
2.17%
36.96%
34.78%
1986
32 c**
52.17%
23.91%
8.70%
46.43%a***, c***
53.57% a***, c***
32.14%
7.14%
21.43%
35.71%
1993
63a**, c**
39.29%
32.14%
17.86%
Operation Demographics
Non-Halal Businesses
Current Halal Businesses
Slaughter and Process
Only Slaughter
Only Process
Retailer
Year Established
Number of Employees
Rural
Suburban
Urban
Respondent Demographics
White
78.26%
Black
0.00%
0.00% c**
Asian
Native American or Alaskan Native
2.17%
Hispanic
4.35%
Democrat
4.35%
Republican
32.61%
Independent
28.26%
First or Second-Generation Immigrant
4.35%
Female
23.91%
0.00%b**, c**
High School
Bachelor’s
32.61%
Graduate Degree
21.74%
Religious
50.00%
Notes: F-tests conducted on all three groups and statistically significant findings evaluated with pairwise t-
tests. Statistical significance between groups from the pairwise t-tests is indicated as follows: a indicates
significance between Groups 1 and 2, b indicates significance between Groups 1 and 3, and c indicates
significance between Groups 2 and 3. Additionally, *** indicates p < 0.01, ** indicates p < 0.05, and *
indicates p < 0.1.
75.00%
3.57%
10.71% c**
0.00%
3.57%
10.71%
28.57%
32.14%
10.71%
10.71%
10.71% c**
35.71%
14.29%
42.86%
62.86%
0.00%
5.71%
2.86%
0.00%
5.71%
25.71%
22.86%
5.71%
20.00%
11.43%b**
25.71%
17.14%
54.29%
7. Conclusions, Limitations, and Future Work
This study utilized a mixed methods approach consisting of qualitative interviews and
nationwide online surveys with U.S. meat and poultry processors and retailers. The purpose of this
research was to investigate the decision-making process for offering halal meat and poultry
44
products; specifically, the incentives and barriers that exist when entering the halal meat market
and how retailers and processors perceive them. The goal of this research was to determine
characteristics of businesses that are mostly likely to adopt a halal program in the future and how
to facilitate their market participation. Both halal and non-halal businesses were sorted into three
groups via principal components analysis (PCA) and non-hierarchical k-means clustering using
the data collected on the motivations and barriers related to offering [certified] halal meat and
poultry products.
Altogether, my analysis sheds light on potential avenues for supply-side expansion of the
U.S. domestic halal meat and poultry industry. By studying patterns in businesses’ perceptions of
the motivations and barriers and operational demographics, I can determine common
characteristics that lend themselves well to the adoption of a halal program. Indeed, the analysis
shows that urban, suburban, retail, younger, and larger businesses are more likely to currently have
a halal program (Group 2), while businesses that may expand into the halal market in the future
are older and more likely to be retailers or processors instead of slaughterhouses. As retailing and
processing of halal meat and poultry products is generally less capital intensive and requires less
knowledge and skill to implement, these businesses are likely the strongest candidates to expand
into halal in the future. Furthermore, if this expansion in the downstream supply chain can be
achieved, it may make adding a halal program upstream to existing slaughtering facilities more
attractive and provide demand-driven pressure for halal slaughterhouse market entry.
There are a few notable limitations of this study. First, collecting data from supply-side
agents in any market is notoriously difficult, which leads to lower response rates and lower
numbers of quality observations in our dataset. Thus, more complex statistical analyses were not
possible. Second, there are relatively few halal meat and poultry businesses in the U.S. compared
45
to traditional meat and poultry businesses, so halal businesses are not equally represented in our
survey sampling. Finally, even though efforts were made to survey participants in their native
language, it is likely that some halal businesses did not take the survey due to English language
barriers.
There is still need for more research focused on the U.S. domestic halal meat supply chain,
and future studies can expand on this work in a few ways. As my analysis was exploratory in
nature, additional research can and should be done into the noteworthy motivations and barriers
for the different types of businesses in this study – halal retailers, halal processors, non-halal
retailers, and non-halal processors. A deeper understanding of these businesses’ differences in
perceptions of adopting a halal program would help design more effective policies and incentive
structures for a more robust halal meat supply chain. To achieve this, future work would benefit
from larger sample sizes to facilitate more advanced statistical techniques for data analysis. There
are two methods that I believe could be effective for increasing sample size when working with
these populations. First, utilizing a team of researchers to conduct in-person or virtual (e.g., Zoom
or phone) surveys in real-time may increase response rates and quality, as processors and retailers
are typically less likely to complete surveys. Secondly, some of the U.S. Muslim community are
nonnative English speakers, especially older individuals and recent immigrants. As such,
researchers may benefit from close partnerships with native Arabic and Urdu speakers when
collecting data from halal businesses to increase participation rates. Altogether, these suggestions
will help future work make meaningful contributions to our understanding and support of this
unique market, as well as add to the literature on meat and poultry businesses’ decision making
and design of policy to support supply chain development.
46
BIBLIOGRAPHY
Ab Rashid, N., & Bojei, J. (2019). The relationship between halal traceability system adoption and
environmental factors on halal food supply chain integrity in Malaysia. Journal of Islamic
Marketing, 11(1), 117–142. https://doi.org/10.1108/JIMA-01-2018-0016
Ahmed, A. (2008). Marketing of halal meat in the United Kingdom: Supermarkets versus local shops.
British Food Journal, 110(7), 655–670. https://doi.org/10.1108/00070700810887149
Arabie, P., & Hubert, L. (1996). Advances in Cluster Analysis Relevant to Marketing Research. In W.
Gaul & D. Pfeifer (Eds.), From Data to Knowledge (pp. 3–19). Springer Berlin Heidelberg.
https://doi.org/10.1007/978-3-642-79999-0_1
Bejaei, M., Cliff, M. A., & Singh, A. (2020). Multiple Correspondence and Hierarchical Cluster
Analyses for the Profiling of Fresh Apple Customers Using Data from Two Marketplaces.
Foods, 9(7), 873. https://doi.org/10.3390/foods9070873
Bergeaud-Blackler, F. (2004). Social definitions of halal quality: the case of Maghrebi Muslims in
France.
Bonne, K., & Verbeke, W. (2007). Religious values informing halal meat production and the control
and delivery of halal credence quality. Agriculture and Human Values, 25(1), 35–47.
https://doi.org/10.1007/s10460-007-9076-y.
Bonne, K., & Verbeke, W. (2008). Muslim consumer trust in halal meat status and control in
Belgium. Meat Science, 79(1), 113–123. https://doi.org/10.1016/j.meatsci.2007.08.007.
Bonne, K., Vermeir, I., & Verbeke, W. (2008). Impact of Religion on Halal Meat Consumption
Decision Making in Belgium. Journal of International Food & Agribusiness Marketing, 21(1),
5–26. https://doi.org/10.1080/08974430802480628.
Calder, A. J., Burton, A. M., Miller, P., Young, A. W., & Akamatsu, S. (2001). A principal
component analysis of facial expressions. Vision Research, 41(9), 1179–1208.
https://doi.org/10.1016/S0042-6989(01)00002-5
Campbell, H., Murcott, A., & MacKenzie, A. (2011). Kosher in New York City, halal in Aquitaine:
Challenging the relationship between neoliberalism and food auditing. Agriculture and Human
Values, 28(1), 67–79. https://doi.org/10.1007/s10460-010-9260-3
Carmichael, L. (2020). Muslims in America. Zwemer Center for Muslim Studies.
https://www.zwemercenter.com/muslims-in-america/
Choe, J. (2023). Outlook for Livestock and Poultry in 2022. USDA’s 99th Agricultural Outlook
Forum. United States Department of Agriculture.
https://www.usda.gov/sites/default/files/documents/2023AOF-livestock-poultry-outlook.pdf
47
Fuseini, A., Hadley, P., & Knowles, T. (2021). Halal food marketing: An evaluation of UK halal
standards. Journal of Islamic Marketing, 12(5), 977–991. https://doi.org/10.1108/JIMA-02-2020-
0037
Fuseini, A., Wotton, S. B., Knowles, T. G., & Hadley, P. J. (2017). Halal Meat Fraud and Safety
Issues in the UK: A Review in the Context of the European Union. Food Ethics, 1(2), 127–142.
https://doi.org/10.1007/s41055-017-0009-1
de Araújo, S.H. (2019). Assembling halal meat and poultry production in Brazil: Agents, practices,
power, and sites. Geoforum, 100, 220–228. https://doi.org/10.1016/j.geoforum.2019.01.014
Durán, K., & Pipes, D. (2002). Muslim Immigrants in the United States. Center for Immigration
Studies. CIS.org. https://cis.org/Report/Muslim-Immigrants-United-States
Hanson, G. D., Rauniyar, G. P., & Herrmann, R. O. (1994). Using consumer profiles to increase the
U.S. market for seafood: Implications for aquaculture. Aquaculture, 127(4), 303–316.
https://doi.org/10.1016/0044-8486(94)90233-X
Hobbs, J. E. (2021). The Covid-19 pandemic and meat supply chains. Meat Science, 181, 108459.
https://doi.org/10.1016/j.meatsci.2021.108459
Hotelling, H. (1933). Analysis of a complex of statistical variables into principal components. Journal
of Educational Psychology, 24(6), 417–441. https://doi.org/10.1037/h0071325
Hsu, F.-C., Kritchevsky, S. B., Liu, Y., Kanaya, A., Newman, A. B., Perry, S. E., Visser, M., Pahor,
M., Harris, T. B., Nicklas, B. J., & for the Health ABC Study. (2009). Association Between
Inflammatory Components and Physical Function in the Health, Aging, and Body Composition
Study: A Principal Component Analysis Approach. The Journals of Gerontology Series A:
Biological Sciences and Medical Sciences, 64A (5), 581–589.
https://doi.org/10.1093/gerona/glp005
Ijaz, M., Yar, M. K., Badar, I. H., Ali, S., Islam, Md. S., Jaspal, M. H., Hayat, Z., Sardar, A., Ullah,
S., & Guevara-Ruiz, D. (2021). Meat Production and Supply Chain Under COVID-19 Scenario:
Current Trends and Future Prospects. Frontiers in Veterinary Science, 8, 660736.
https://doi.org/10.3389/fvets.2021.660736
Larue, B. (2021). COVID‐19 and labor issues: An assessment. Canadian Journal of Agricultural
Economics/Revue Canadienne d’agroeconomie, 69(2), 269–279.
https://doi.org/10.1111/cjag.12288
Lever, J., & Miele, M. (2012). The growth of halal meat markets in Europe: An exploration of the
supply side theory of religion. Journal of Rural Studies, 28(4), 528–537.
https://doi.org/10.1016/j.jrurstud.2012.06.004
Lever, J., Puig De La Bellacasa, M., Miele, M., & Higgin, M. (2010). From the Slaughterhouse to the
Consumer: Transparency and information in the distribution of Halal and Kosher meat. Dialrel.
48
Malone, T., & Lusk, J. L. (2018). If you brew it, who will come? Market segments in the U.S. beer
market. Agribusiness, 34(2), 204–221. https://doi.org/10.1002/agr.21511
Lopez, M. H. (2017, December 20). Hispanic identity fades across generations as immigrant
connections fall away. Pew Research Center’s Hispanic Trends Project.
https://www.pewresearch.org/hispanic/2017/12/20/hispanic-identity-fades-across-generations-as-
immigrant-connections-fall-away/
Masudin, I., Jie, F., & Widayat, W. (2020). Impact of halal supplier service quality and staff readiness
to adopt halal technology on halal logistics performance: A study of Indonesian halal meat
supply chain. International Journal of Agile Systems and Management, 13(3), 315-338.
Maxwell, J. A. (2012). Qualitative research design: An interactive approach. Sage Publications.
Miles, M. B., Huberman, A. M., & Saldaña, J. (2020). Qualitative data analysis: A methods
sourcebook (Fourth edition). SAGE.
Mumuni, A. G., Veeck, A., Luqmani, M., Quraeshi, Z. A., & Kamarulzaman, Y. (2018). Religious
identity, community, and religious minorities’ search efforts for religiously sanctioned food: The
case of halal food in non-Muslim majority markets. International Journal of Consumer Studies,
42(6), 586–598. https://doi.org/10.1111/ijcs.12423
National Grocers Association. (2019). NGA Retail Membership List by State – Winter 2019. National
Grocers Association. https://www.nationalgrocers.org/wp-content/uploads/2018/12/NGA-Retail-
Membership-List-Winter-2019-by-State.pdf
Patton, M. Q. (2014). Qualitative research & evaluation methods: Integrating theory and practice.
Sage Publications.
Pearson, K. (1901). LIII. On lines and planes of closest fit to systems of points in space. The London,
Edinburgh, and Dublin Philosophical Magazine and Journal of Science, 2(11), 559–572.
https://doi.org/10.1080/14786440109462720
Public Religion Research Institute. (2023, November 3). The 2020 PRRI Census of American religion.
https://www.prri.org/research/2020-census-of-american-religion/#:~:text=Muslim%20
PRRI.
Americans%20are%20primarily%20concentrated,County%2C%20New%20York%20(4%25)
Rubin, H. J., & Rubin, I. S. (2011). Qualitative interviewing: The art of hearing data. Sage Publications.
Saldaña, J. (2013). The coding manual for qualitative researchers (2. ed). SAGE Publ.
Salindal, N. A. (2019). Halal certification compliance and its effects on companies’ innovative and
market performance. Journal of Islamic Marketing, 10(2), 589–605.
https://doi.org/10.1108/JIMA-04-2018-0080
49
Shahijan, M.K., Rezaei, S., Nigel Preece, C., & Khairuzzaman Wan Ismail, W. (2014). Examining
retailers’ behaviour in managing critical points in Halal meat handling: A PLS analysis. Journal
of Islamic Marketing, 5(3), 446–472. https://doi.org/10.1108/JIMA-10-2013-0070
Sinha, R. N., Wallace, H. A. H., & Chebib, F. S. (1969). Principal-Component Analysis of
Interrelations Among Fungi, Mites, and Insects in Grain Bulk Ecosystems. Ecology, 50(4), 536–
547. https://doi.org/10.2307/1936244
Stata. (2013, February 11). Statistical Software for data science | Stata. Stata.com. Retrieved August
8, 2022, from https://www.stata.com/manuals13/mvfactor.pdf
Taylor, C. A., Boulos, C., & Almond, D. (2020). Livestock plants and COVID-19 transmission.
Proceedings of the National Academy of Sciences, 117(50), 31706–31715.
https://doi.org/10.1073/pnas.2010115117
Tieman, M., van der Vorst, J. G. A. J., & Che Ghazali, M. (2012). Principles in halal supply chain
management. Journal of Islamic Marketing, 3(3), 217–243.
https://doi.org/10.1108/17590831211259727
United States Department of Agriculture (USDA) Agricultural Marketing Service (AMS).
(2022). Meat and poultry inspection readiness grant. Meat and Poultry Inspection Readiness
Grant | Agricultural Marketing Service. Retrieved August 5, 2022, from
https://www.ams.usda.gov/services/grants/mpirg
Verbeke, W., Rutsaert, P., Bonne, K., & Vermeir, I. (2013). Credence quality coordination and
consumers’ willingness-to-pay for certified halal labelled meat. Meat Science, 95(4), 790–797.
Weersink, A., Von Massow, M., Bannon, N., Ifft, J., Maples, J., McEwan, K., McKendree, M. G. S.,
Nicholson, C., Novakovic, A., Rangarajan, A., Richards, T., Rickard, B., Rude, J., Schipanski,
M., Schnitkey, G., Schulz, L., Schuurman, D., Schwartzkopf-Genswein, K., Stephenson, M., …
Wood, K. (2021). COVID-19 and the agri-food system in the United States and Canada.
Agricultural Systems, 188, 103039. https://doi.org/10.1016/j.agsy.2020.103039
Winship, C., & Mare, R. D. (1984). Regression Models with Ordinal Variables. American
Sociological Review, 49(4), 512. https://doi.org/10.2307/2095465
Zulfakar, M. H., Jie, F., & Chan, C. (2013, December). Critical success factors for a successful
implementation of halal red meat supply chain in Australia: meat processor’s perspective. In 11th
ANZAM Operations, Supply Chain and Services Management Symposium, Melbourne (pp. 20-
21).
Zulfakar, M. H., Chan, C., & Jie, F. (2018). Institutional forces on Australian halal meat supply chain
(AHMSC) operations. Journal of Islamic Marketing, 9(1), 80–98. https://doi.org/10.1108/JIMA-
01-2016-0005
50
Zulfakar, M. H., Chan, C., Jie, F., & Sundram, V. P. K. (2019). Halal accreditation and certification in
a non-Muslim country setting: Insights from Australia halal meat supply chain. International
Journal of Supply Chain Management, 8(1), 10-17.
51
CHAPTER 2. MARKET PARTICIPANTS’ PREFERENCES FOR
A NATIONAL HALAL MEAT CERTIFICATION PROGRAM
1. Introduction and Motivation
There are over 200 different halal food certification organizations around the globe
(Zabihah 2021, Verify Halal 2022), and in the U.S. alone, there are over 20 verified halal
certification organizations. Given that the global Muslim population does not have a universally
accepted and verified halal meat standard and the resulting astoundingly large number of
certifications and certifying organizations operating in the market, it is unsurprising that halal
foods are the fourth most likely food in the U.S. to be fraudulent (FSNS 2020). In most cases, halal
meat and poultry certification organizations have their own internal halal standards that they use
to evaluate and then grant certification; these depend on how strictly they interpret the religious
laws from The Qur’an and the teachings of the Prophet Muhammad (peace be upon him) written
in The Hadith that govern halal meat. The plethora of distinct and sometimes contradictory
certifications can make selecting a product confusing for consumers, accurately sharing
information with customers difficult for retailers, and determining requirements for halal meat
processing challenging for processors.
There is an overabundance and a lack of clarity and standardization in certifications in the
U.S. halal meat and poultry market. The current halal meat and poultry market situation is similar
to the U.S. organic market prior to the development of the USDA organic certification, when there
were many different and competing organic certifications available for producers, and consumers
necessarily incurred search and information collection costs to determine which labels and
standards were on products (Lohr 1998). As in the case of the pre-USDA certification organic
market, this likely makes it challenging for participants to engage fully in the halal meat and
poultry market and reduces the domestic market’s competitiveness relative to imported products
52
(Lohr 1998). Further, an overabundance of certification and regulations serves as a barrier to
market entry for specialty meat products (Worosz et. al 2008) – especially in the case of smaller
or new operations, as previously seen in the organic farming industry (Guthman 1998). These
barriers result in a loss of market efficiency, and in the long term, a harm to consumers who are
thereby underserved. Thus, this research aims to undertake the first step in the process of
developing a U.S. NHMC program to help streamline market participation.
These shared concerns suggest there is room for a verification strategy in the form of a
unified U.S. national halal meat and poultry certification (NHMC) program. Correspondingly, the
purpose of this study is to understand what the U.S. domestic halal market participants want out
of a NHMC program, and how these desires compare to what is feasible to implement. Therefore,
to determine what such a program would look like and how it would operate, this chapter addresses
three major research questions. First, what characteristics of a hypothetical NHMC are important
to halal meat processors, halal meat retailers, and Muslim halal meat consumers? Second, are there
differences in preferences for NHMC program characteristics between groups, and if so, what is
the nature of these differences? And finally, how could differences in preferences between groups
impact the design and implementation of a NHMC program? This research will answer these
questions by investigating the preferences for a NHMC program at three levels of the market
(processors, retailers, and Muslim halal meat consumers). The differences in preferences across
groups will be compared to each other and will serve as a baseline understanding of the market’s
desires, which future research can expand upon.
In all, this research contributes to the literature on labeling and certification for improving
trust in products with credence attributes. This chapter provides an overview of the variety of
different halal meat certifications, certification programs, and certifying bodies around the world,
53
with a direct comparison to the current U.S. halal meat certification landscape, which to the
author’s knowledge has not been explicitly explored before. Altogether, the findings of this chapter
will aid in bolstering halal meat consumer confidence in the authenticity of their products, as well
as improve the equity of the U.S. food system.
2. Current Landscape of U.S. Domestic Halal Certifications
As early as the 1970s and 1980s, U.S. Muslim consumers – concerned about the
authenticity of the halal meat and poultry products they were buying – began to demand more
verification for their products in the form of third-party labels and certifications. From this
movement, certifying organizations were established; these organizations typically charge a fee –
per unit, an annual lump sum, or some formulaic combination – to inspect slaughter, processing,
and retail facilities and provide certification that their products and/or establishment conforms to
a given set of halal standards.
As the demand for halal meat and poultry products has grown and evolved over the past 50
years, so too has the number of certifiers, the variety of certifications, and the ways in which
certifiers operate. Today, there are several major halal meat and poultry certifiers and another
dozen or so smaller local or regional certification agencies operating in the U.S.5 These certifiers
are broken into two organizational categories: non-profit foundations that rely mainly on volunteer
work and donations from the public to operate, and for-profit businesses which charge fees for
their services and retain employees to conduct the certification process.6 The non-profit
foundations include Halal Monitoring Services (HMS) of Shariah Board of America and Halal
Food Standards Alliance of America (HFSAA), among others. The for-profit businesses include
5 Not all of these certifiers work exclusively with meat and poultry products.
6 These classifications are based upon what the author learned in qualitative interviews with halal meat and poultry
certifiers during this project and are not necessarily mutually exclusive nor immutable.
54
Islamic Services of America (ISA), Islamic Society of the Washington Area (ISWA), Halal
Transactions of Omaha (HTO), Islamic Food and Nutrition Council of America (IFANCA), and
American Halal Foundation (AHF), among others.
Many of the for-profit certifiers have certification via the United States Department of
Agriculture’s Food Safety and Inspection Service (USDA-FSIS) to operate in the import and
export markets, while the non-profit foundations generally do not.7 For-profit certifiers offer both
zabiha and machine-slaughtered halal certifications, while the non-profit certifiers focus mainly
on the zabiha market.8 Correspondingly, for-profit certifiers typically contract with larger domestic
and multinational corporations and establishments looking for international business opportunities,
while non-profits operate mainly in the domestic small and local business arena. Thus, for-profit
certifiers currently make up most of the certification market by both quantity of products certified
and revenues received, therefore holding the majority of power in the halal certification market.
However, the non-profit certifiers work with a larger number of domestic businesses, though the
quantity of products and overall revenues is lower.
As with any market, power dynamics and market structure play a pivotal role in market
growth, policy development, and the implementation of new methods of production.9 Indeed, the
U.S. domestic halal meat and poultry certification market is no exception. According to multiple
certifiers and Muslim consumer advocacy groups, efforts have been underway for over a decade
to create a set of industry-wide U.S. national halal standards, though progress in reaching
agreements on what specifications and requirements to include has been slow. For-profit and non-
7 These certifications are a service provided by the Food Safety and Inspection Service (FSIS) of the United States
Department of Agriculture (USDA) and requires businesses to pay a fee to obtain export certification and/or import
products from foreign countries.
8 The term “zabiha” refers to hand-slaughtered animal products. More information is given in the preface to this
dissertation.
9 Historically, markets with an imbalance of power (e.g., those with monopolies or oligopolies) see declines in
growth, equity, innovation, and efficiency losses (Washington Center for Equitable Growth 2018).
55
profit certifiers commonly disagree on which standards should be included in a U.S. national halal
certification, as well as how such a certification should be implemented and verified. What’s more,
Muslim halal consumers have varying opinions due to the extremely diverse nature of the
population in terms of ethnicity, religious sect, immigration and citizenship status, and other
cultural factors. The dissenting views on halal certification from both the supply and demand side
of the market contribute to the ongoing struggle to develop a single cohesive certification program.
Further complicating the matter are the views of the general public and federal and state
government representatives. Islam is a minority religion in the U.S., and multiple domestic and
international events over the past several decades have reinforced prejudices that negatively impact
U.S. Muslims; indeed, a Pew Research poll revealed that a significant percentage of Americans
believe Muslims face “a lot of” discrimination and are viewed more negatively than other religions
(Mohammed 2021). As such, it is unlikely that many government representatives will move to
make changes to state or federal laws to improve halal verification and transparency, nor support
the development of a national-level halal meat and poultry certification. Instead, general efforts to
increase authenticity in the food system, strengthen traceability, and improve labeling
requirements are more realistic avenues for addressing halal certification issues.
2.1 Legal Enforcement of Certifications in the U.S.
In the U.S., there are a lack of labeling regulations in place to ensure that consumers know
which standards are involved in the production of food products. Outside of a handful of states
that have instituted forms of consumer-right-to-know legislation (California, Illinois, Maryland,
Michigan, Minnesota, New Jersey, New York, Texas, and Virginia), there is no federal law or
regulation that requires a business that claims to be certified to supply a given specialty product to
prove that they are certified by displaying the certificate or registering with the state department
56
of agriculture. Furthermore, these state laws that require proof of certification to be readily
available often are not enforced. Altogether, there is ample room for fraudulent behavior by
uncertified actors throughout the U.S.; this is likely a large factor contributing to halal food being
the fourth most fraudulent industry in the U.S.
When looking to certify an attribute, one of the main questions is who should be the
certifying body. Many studies have focused on preferences for certification entities and have found
that government certification is in general more trusted than third-party certification in a wide
variety of contexts; see for example McKendree et al. (2013), Johnston et al. (2001), Ortega et al.
(2011, 2012), and Sønderskov & Daugbjerg (2011). However, for the halal market in the U.S., the
First Amendment prohibits the government from setting any religious product standards, as this
would be an infringement of religious freedoms. California, Illinois, Maryland, Michigan,
Minnesota, New Jersey, New York, Texas, and Virginia have all circumvented these constitutional
rights concerns by enacting laws in the spirit of consumer-right-to-know considerations. These
include allowing the state to enforce the proper use of halal labels, including outlawing mislabeling
and false representation, requiring businesses to display their halal meat Department of Agriculture
registration, and requiring separate facilities/machinery for halal products to ensure no cross
contamination can occur (Illinois Halal Food Act 2002). A similar law exists in many states for
kosher meat. A federal law and enforcement akin to these states’ laws would provide another layer
of authentication to prevent halal food fraud.
3. Halal Certification Around the World
Of the plethora of certification programs operating around the world, there are several that
oversee halal production standards on a national or even multi-national scope – such as the
Department of Islamic Development Malaysia (JAKIM) program or the Australian national halal
57
meat program (Rhaman et al. 2018, Australian Halal 2022). Though the U.S. does not have the
same institutional framework as these countries, lessons can be learned from national-level
programs already in place.
In countries like Australia, Singapore, the United Arab Emirates (UAE), and Malaysia, the
national halal meat certification (NHMC) standard is set by, and administered through, the federal
government. This allows for a generalization of the halal meat standards within these countries;
there is only one halal meat standard or a set of predetermined levels of halal meat standards. This
makes adoption by supply-side agents in the halal meat market straightforward, as they can either
adhere to the universal standard or not – there is no room for interpretation or for confusion as to
which certification a supplier needs to enter the market. This is also advantageous for consumers,
as the halal meat standards will be identical across the country, so there is no room for
miscommunication or confusion with competing labels or standards. On the flip side, this does
remove diversity from the halal meat market, and so consumers who desire different standards for
halal meat products may still need to look elsewhere for their products. While a program in which
local, state, or federal government sets halal standards may not be possible in the U.S. due to
concerns about First Amendment violations, the structure and development of a national program
is still useful to understand. Furthermore, a U.S. national program could potentially have
government involvement in the enforcing of standards set by non-governmental organizations.
In contrast, in other countries, there is no national halal program, leaving room for multiple
third-party agents, such as Islamic community groups, imams (faith leaders in Islam), and
certifying corporations to design their own halal meat standards and certification implementation
processes. This system is akin to what is currently in place in the U.S., Canada, and most other
countries across the globe. This system can be very confusing to consumers and supply-side agents
58
alike, especially when there is little to no transparency in the standards required to obtain a given
certification. Consumers are faced with an overload of information and sorting out which
certifications align with their beliefs is mentally taxing. Supply-side agents have other issues –
selecting one certification to use in their operation may preclude them from some market
opportunities or result in backlash from consumer groups who do not agree with the standards they
follow. These concerns are especially relevant in the U.S., as, to the author’s knowledge, the U.S.
has the largest number of halal meat certifiers and certifications currently available in the market
of any country in the world. However, this assortment of certifications can also be a boon – for
consumers with very niche requirements, they may be able to find a certification that aligns with
their beliefs, whereas a universal standard may not align. Altogether, the future of U.S. domestic
halal certification is complex, and understanding and addressing the market’s needs will require
significant effort on behalf of researchers and policy makers alike.
3. Mixed Methods and Survey Design
As discussed, the U.S. domestic halal meat and poultry market has distinctive niche
religious requirements that must be upheld. The nature of the halal market, coupled with the lack
of research to understand the certification landscape of the U.S. halal meat market poses an
exciting opportunity to focus this study on multiple groups throughout the domestic market. This
ensures that the main actors in the halal meat market that would be impacted by a certification –
consumers, retailers, and processors – are included in the study. This understudied market also
requires a methodological approach designed for exploration and learning feedback. Therefore, I
employ a mixed methods research design using both qualitative interviews and quantitative survey
methods in this project. The synergy between methods is exhibited in two ways. In the first stage
of this project, I conducted qualitative interviews with halal meat processors, retailers, and Muslim
59
consumers. These qualitative interviews were necessary to develop an understanding of and elicit
opinions on halal meat market certification practices in the U.S. domestic market. I used the data
from these conversations to narrow my research questions and design quantitative data collection
tools. From these interviews, I determined the most appropriate questions to explore in relation to
halal meat and poultry certification revolve around an overabundance of certifications, unclear
standards, and a lack of transparency in certification use and enforcement. These qualitative
interview findings in turn inform the quantitative components of the study. In particular, the
patterns uncovered in the qualitative research indicated that the quantitative data collection should
utilize best-worst scaling (BWS) to measure processors’, retailers’, and consumers’ preferences
for possible attributes of such a program.
3.1 Phase 1: Qualitative Interview Methods
The qualitative portion of this project was designed with methods from Patton (2014),
Rubin and Rubin (2011), and Maxwell (2012). The purpose of these qualitative interviews was to
obtain information about the U.S. domestic halal meat and poultry industry’s certification
standards and process previously unknown to me, as this market is understudied. Information
collected via qualitative interviews included Muslim halal consumers’, halal retailers’, and halal
processors’ opinions on the current certification landscape and their preferences for future
certification development. Collecting information directly from market participants in this way
allowed me to design a more robust and relevant research program for the quantitative portion of
this project. The qualitative interviews also provided vital context and explanatory power for the
quantitative survey findings. Interview questions were grouped by topic and were open-ended to
allow for robust answers (Patton 2014, Rubin and Rubin 2011, Maxwell 2012). The interview
60
guide for these interviews, information for how interviews were conducted, and the outline of the
interview analysis process are found in Appendix A.1.
Supply-side interviewees were recruited from lists of retailers and processors registered as
certified to supply halal meat by a reputable halal meat certifier. The lists of supply-side interview
candidates were narrowed to those in Midwestern states for ease of access and improved likelihood
of name recognition for Michigan State University. A series of eight in-depth, semi-structured
interviews with Midwestern halal meat retailers and 10 in-depth, semi-structured interviews with
Midwestern halal meat processors were conducted December 2021 through April 2022. Supply-
side interviews were conducted via Zoom or over the phone. Consumer interviewees were
recruited in person from five different certified halal meat retailers in Illinois, Wisconsin, and
Michigan; the list of interested participants was then randomly sampled from. Selected consumers
were interviewed over Zoom or via phone and were compensated for their participation with a $50
gift card. In total, 12 consumer interviews were conducted in April 2022 through June 2022.
3.2 Phase 2: Quantitative Survey Methods
The second stage of this chapter involves conducting three stacked national online surveys,
one with a sample of meat processors, one with a sample of meat retailers, and another with
Muslim halal meat consumers.10 These parallel stacked surveys are designed to assess preferences
related to the development of a hypothetical U.S. NHMC program. The surveys had three major
purposes: first, to further explore the patterns uncovered in the qualitative interviews; second, to
be able to gauge these groups’ preferences and opinions at a national level; and third, to directly
compare preferences and response patterns across the three market groups – a key advantage of
stacked surveys. This method is particularly attractive for application in this study, as consumers,
10 Surveys were available in English, Arabic, and Urdu for all study participants.
61
retailers, and processors are the main actors in the halal meat market that would be impacted by a
certification and should therefore be included. Additionally, the shared desire for a NHMC
program across the three groups interviewed suggests that comparing preferences directly in this
manner will be useful in determining consensus or differences in opinions.
To compare preferences for attributes of a U.S. NHMC program directly across the three
groups using a consistent and unitless ranking system, I use best-worst scaling (BWS), also known
as maximum difference (max-diff) scaling or most-least scaling. BWS is a choice analysis method
that asks participants to repeatedly – over a series of different choice sets – select the most
preferred (best) and least preferred (worst) options out of a given set of items and allows the
researcher to thereby understand preferences between items. BWS was introduced by Finn and
Louviere (1992) with theoretical properties of probabilistic, best-worst choice models being more
recently explained by Marley and Louviere (2005). BWS is advantageous relative to Likert scale
questions, as Likert scales make it challenging to distinguish the actual importance of attributes.
For example, multiple Likert scale questions can all have the same mean level of importance.
Likert scale questions are additionally problematic due to scale subjectivity – what is considered
a “4” on one individual’s scale may be a “5” on another (Lusk and Briggeman 2009, Lusk and
Parker 2009, Wolf and Tonsor 2013). Additionally, as BWS is a tradeoff method, I achieve shares
of importance that can be directly interpreted from a ratio scale. The sum of shares among all
attributes analyzed must equal one. If attributes j and k have importance shares of 0.2 and 0.1,
respectively, attribute j is two times as important as k. BWS also provides five to ten times more
differentiation than most scaling methods, such as the aforementioned Likert scales (Horne 2012).
Directly interpretable shares and more differentiation provides further insight into the exact
importance of each attribute.
62
In each survey, participants were asked a series of seven BWS questions with different
combinations of seven attributes of a potential future U.S. NHMC program. In these questions,
participants were asked to select which of three attributes they thought was the most important and
which was the least important to consider when designing a U.S. NHMC program. An example of
one of the questions that participants saw is shown in Figure 5.
Figure 5: Example Best-Worst Question for Potential National Halal Certification
3.3 Attribute Selection for the Best-Worst Scaling Survey Instrument
The BWS for this survey includes seven attributes for consideration. The attributes for the
BWS questions were selected based on the characteristics of current certification programs around
the world (both within and outside of halal meat), discussion in the literature, and information
collected from interviews. Justification for these choices is broken down by attribute below. These
attributes differ slightly in their precise language across the three surveyed groups to better fit the
audience; attributes included here are cost, administrative body, traceability, information collected
or available, certification type, audit characteristics, and halal type. The description of these
attribute categories and the relevant questions for each surveyed group are presented in Table 14.
63
Table 14: Best-Worst Scaling Attributes and Question Language by Group
Attribute wording by group
Referred to in
paper as
Consumers
Retailers
Processors
Costs
Additional costs I would need to pay
for certified products
Costs associated with certification program involvement
(certification fees, infrastructure, labor)
Enforcement
What organization enforces the program standards
Benefits
Confidence gained by purchasing a
certified product
Benefits associated with certification program involvement
(access to new markets, price premiums, etc.)
Information What information is available to me
What information is passed on to my customers
e
t
u
b
i
r
t
t
A
Who/What
Certified
What or who should be required to be certified (products, retailers, slaughterhouses, etc.)
Inspections
How retailers and processors are
audited/inspected
How my suppliers’ & my operation
are inspected
How my operation is
inspected
Halal
Standards
What halal standards are required (hand- versus machine-slaughtered, stunning or no stunning, etc.)
64
3.3.1 Costs
Cost is a major component of a certification’s success and is therefore vital to include in
the BWS questionnaire. Supply-side agents are unlikely to adopt a new certification, technology,
or production method if they perceive the costs of adoption outweigh the benefits (Pearson &
Henryks 2008, El-Osta & Morehart 2000, Ugochukwu & Phillips 2018). Likewise, consumers are
unlikely to pay the price premium for a certified product if the premium elevated the cost of the
product outside of their budget (Pearson & Henryks 2008). The qualitative data indicated that
supply-side agents were concerned about bringing new or additional certification programs into
their operations, as these are likely to raise their cost of production, which would be mostly passed
onto their consumers. Similarly, consumers interviewed were also hesitant about potential
increased prices on their halal meat products, which are already more expensive than non-halal
products.
3.3.2 Enforcement
It is also essential to consider preferences that market participants may have for a NHMC
program’s administering body, as opinions on government versus non-government oversight are
varied. McKendree et al. (2013), Johnston et al. (2001), Ortega et al. (2011, 2012), and Sønderskov
and Daugbjerg (2011) have all found that certifications administered by a government agency are
in general more trusted than those run by a third-party organization. However, in the case of halal
meat, which is a religious standard, a government-run certification program would likely violate
the First Amendment, and so an alternate organizer is needed. Further, halal meat processors and
retailers interviewed communicated trepidation with additional government involvement in their
operations. In contrast, most consumers interviewed were in favor of increased government
involvement in regulating halal certified meat products.
65
3.3.3 Benefits
Given the rampant food fraud found in the halal market (FSNS 2020) and market agents’
corresponding concerns about transparency, the benefits to consumers and supply chain agents
alike of a NHMC program are necessary to consider. On the supply side, access to additional
markets is possible, and consumer trust and loyalty can be improved using certifications. Wary
consumers are likely to be willing to pay more for certified products to have enhanced peace of
mind. Processors and retailers interviewed described interest in using a national certification to
access new markets and share product attributes easily; consumers likewise would like the ability
to verify product information more easily.
3.3.4 Information
Market agents have preferences for what types of information is collected throughout the
supply chain. There is a fine line for some individuals between collecting relevant and necessary
information for ensuring a certification is met and being overly intrusive in an operation. Some
processors interviewed were concerned that collecting additional data on their day-to-day
operations would slow down their processing. Others expressed displeasure with the idea of yet
another agency “sticking their noses” into business.
Further, it is important to consider what, if any, of the information collected should be
available for downstream agents to access. In interviews with processors and retailers, some
discussed experiences with adding a halal meat program to their organization, only to be met with
backlash from non-Muslim consumers. As such, some processors and retailers, especially larger
non-Muslim owned operations may be hesitant to have their certification information publicly
available. However, Muslim-owned processing plants and retailers, as well as the Muslim
consumers interviewed, want this information to be readily available to the public. This increased
66
availability of information can aid in halal meat consumers’ search efforts and help supply chain
agents compete.
3.3.5 Who or What is Certified
Certification type refers to what the certification program applies to – in this case, whether
individual halal meat products should be certified (i.e., a brand of chicken breasts), whether an
entire operation should be certified (i.e., a slaughterhouse or a retailer), the entire supply chain
(i.e., processors, wholesalers/distributors, and retailers), or some combination of these options.
Current halal meat certification agencies in the U.S. certify at all these levels; some certify
individual products or supply chain agents (e.g., Islamic Services of America 2022), and some
certify the entire supply chain (e.g., Halal Monitoring Services 2022). In interviews, members of
all three groups – consumers, retailers, and processors – had preferences between the types of
certifications they preferred, with some insisting on full supply chain certification and others
accepting piecemeal product certificates.
3.3.6 Inspections
Audit characteristics refers to the type of audit (i.e., scheduled or surprise) and how often
a certification must be renewed or confirmed (i.e., yearly, monthly, daily). Some of the certifiers
in the U.S. have a scheduled annual audit for renewal for their certifications (e.g., Islamic Services
of America), while others have both a scheduled annual audit and surprise audits (e.g., Halal
Monitoring Services). Further, in the case of meat, it is important to consider the value to the
consumer in confirming that the certification has been upheld for each day a processor conducts
halal slaughter. This distinction in audit type and frequency may be important to consumers who
do not trust current certifications currently in the U.S.
67
3.3.7 Halal Standards
Last, but likely most important to Muslim halal meat consumers, is preferences for halal
type, which are likely to have a strong impact on preferences for a NHMC program’s attributes.
There is a very long list of attributes that consumers could prefer to be certified as part of a halal
standard. These include whether the animal was hand- or machine-harvested, pre-stunned before
harvest, facing Mecca at the time of slaughter, harvested by a Muslim or a Person of the Book, and
many other considerations. These attributes vary in their ease of implementation in industrialized
food systems; hand slaughter results in a slower line speed than machine slaughter but facing
Mecca may be relatively simple to accommodate. Regardless of the ease of adopting these
methods, these attributes are vitally important to Muslims and non-negotiable for many. For
instance, in interviews, consumers said they have never and will never eat machine-slaughtered
halal meat products; that is, they require a halal meat certification that uses hand-slaughter
methods. On the other hand, processors interviewed are concerned about the competitiveness of
their operations when their line speed is significantly slower under hand-slaughter methods.
3.4 Additional Survey Questions & Analysis Methods
In addition to the BWS questionnaire, the survey included follow-up questions for each
attribute category to clarify what preferences the three groups have. For example, for the attribute
“Enforcement and Regulation,” the follow-up question shown to participants is given in Figure 6.
The full list of follow-up questions for the seven attributes are given in Figure 7. These follow-up
questions elicit more specific preferences that the surveyed groups have and allow for a more
detailed comparison of choices across the market groups.
68
Figure 6: Example Best-Worst Scaling Follow-Up Question
69
Figure 7: List of Follow-Up Best-Worst Scaling Questions
Costs
• C: How much would they be willing to pay in addition for this
certification on a product (%)
• P&R: List of costs they may consider important in their decision-
making process
Enforcement
• A list of possible organizations that could run the program (e.g., NGO,
religious group, current certifiers, etc.)
Benefits
• P&R: List of benefits they could consider important
Information
• C: Types of info that should be included on packages/in stores
• P&R: Types of info that should be passed on to customers
Who/What Certified
• What should be certified/carry a certification stamp (products, retailers,
suppliers, etc.)
Inspection
• List of inspection types/techniques (scheduled, random, etc.) for
certification renewal
Halal Standards
• List of standards that could be included in a certification
There were also demographics questions included in the surveys. These consisted of the
standard socioeconomic questions for individuals typically seen in survey data collection, as well
as questions for consumers about their ethnic and cultural background, immigration status or
70
generation of citizenship,
religious history
(i.e., born and
raised Muslim versus
converting/reverting to the faith), subsect of Islam that they practice, and other cultural or religious
characteristics that could influence perceptions and preferences related to the U.S. domestic halal
meat market. For processors and retailers, the survey also included operation demographics
questions. The survey instruments were reviewed with members of the Islamic community for
clarity and to ensure there were no misrepresentations.
3.5 Participant Recruitment
The recruitment processes for each of the three groups surveyed – processors, retailers, and
consumers – are given below.
3.5.1 Processor Recruitment
I recruited processor participants from three sources: 1) USDA Food Safety and Inspection
Service (FSIS) list of registered meat processors, 2) registered processors on Halal Monitoring
Services’ (HMS) website, and 3) the American Association of Meat Processors (AAMP). Poultry,
lamb, beef, and goat processors were included in the sample. While the processors listed on HMS’
website are known to be halal, it is also likely that some processors on the USDA FSIS database
and the AAMP membership list also process halal meat or poultry products, though the exact
number is unknown.
The USDA FSIS database lists 5,859 USDA-inspected poultry, lamb, beef, and goat
processing establishments. Of these establishments, the USDA classifies 2,736 as “very small”,
2,656 as “small”, and 440 as “large” establishments.11 I conducted stratified random sampling of
the three groups of establishments in the USDA FSIS data file, using Excel to generate random
11 The USDA classifies processors as “very small” if they have less than 10 employees or less than $2.5 million in
annual sales, “small” if they have 10-499 employees, “large” if they have 500 or more employees.
71
number lists to select establishments from the populations.12 Establishments were called by a team
of undergraduate research assistants beginning in early November 2022 to determine who at the
establishment should respond to the survey and obtain email addresses. Email addresses for
establishments without a phone number listed or those that did not answer or return calls were
retrieved from business websites when available.
Individual Qualtrics survey links were sent via email to the USDA FSIS sample. The first
round of emails was sent using MS Word mail merge on December 9, 2022, with follow-up
reminder emails on December 13 and 16, 2022. The next reminder email was sent using Constant
Contact on January 11, 2023. From the first round of emails, approximately 50 bounced back as
undeliverable, and 12 businesses responded saying they did not qualify or would not be taking the
survey. Thus, a total of 987 processors from the USDA FSIS list received the survey.
There were an additional 58 registered processors on the Halal Monitoring Services’
(HMS) website which were all included in the sample. Emails were obtained from the certifier,
and individualized Qualtrics survey links were sent using MS Word mail merge in January 2023,
with follow-up reminder emails in January and February 2023. Additionally, these processors were
given the option to take the survey in either Arabic or Urdu if they preferred.
The AAMP membership list was contacted via an association representative, who
distributed an anonymous Qualtrics survey link to the membership email listserv in March 2023,
with reminder emails in March and April 2023. It is likely that many AAMP members received
the survey who also were included in my USDA FSIS recruitment efforts. However, the response
12 As very small, small, and large processors make up 46.7%, 45.3%, and 7.5% of the total population, respectively,
these percentages were used to determine how many establishments to sample from each group. There were 1,049
processors contacted from the USDA FSIS database including 20 large processors, 451 small processors, and 578
very small processors.
72
rate for both samples was very low, so I do not anticipate there were any duplicate survey
responses.
Despite contacting businesses in three different samples, I received only 195 total
responses, with only 95 complete responses remaining after data cleaning. I received responses
mainly from very small, small, and medium processing plants, both because these make up over
90% of meat processors in the nation and, anecdotally, because larger processors typically do not
respond to surveys.
3.5.2 Retailer Recruitment
General and halal meat retailers were recruited between February 2023 and October 2023.
Non-halal retailers were recruited via multiple state-level grocers associations and from a
membership list from the National Grocers Association (NGA). First, I attempted to recruit
retailers via the state-level grocers associations and the NGA email listservs with the assistance of
association representatives. However, only 18 responses were received via these efforts, so another
recruitment approach was needed. In May of 2023, a team of undergraduate research assistants
called retailers from the 2019 winter NGA membership list (National Grocers Association 2019)
– the most recent available online – between May 2023 and September 2023 to collect point of
contact email addresses. After removing closed businesses, a total of 946 retail stores were called,
and 236 email addresses were obtained. As in the processor case, it is possible that some of these
retailers actually did have a halal program at the time of the survey, though the exact number of
these stores is unknown.
Known halal retailers were recruited from halal certifiers’ online lists of registered
businesses and through a nationwide web scraping of Yellow Pages using the following key terms
and phrases: “halal meat grocery store,” “halal meat,” “Indo-Pak grocery,” “African grocery,” and
73
“international grocery store.” The results of the web scraping were compiled, and a team of
undergraduate research assistants called the 919 stores between July 2023 and October 2023 to
collect email contact information; 96 email addresses were obtained.
Emails with survey links were sent three times to each category of retailers between August
25 and October 10, 2023. Incentive payments of $25 were offered for complete and quality
responses, though not all respondents claimed their incentive. In total, 50 responses were collected
from the retailer samples, and after data cleaning, 39 viable survey responses remained.
3.5.3 Consumer Recruitment
For Muslim halal meat and poultry consumer recruitment, I partnered with Qualtrics ™ to
collect responses to ensure a nationally representative population to sample from. Multiple rounds
of recruitment were necessary due to the relatively small number of these consumers nationwide.
To qualify for the survey, consumers needed to be practicing Muslims, over the age of 18, the
primary grocery shopper for their household, and have bought a halal meat or poultry product in
the last 12 months. In total, 507 complete and clean responses to the online survey were collected
between May 2023 and December 2023.
4. Methodology & Statistical Framework for Analysis
Multiple methods for data analysis are used in this chapter. First, descriptive statistics are
used to detail the characteristics of survey participants and general patterns in their responses.
Then, BWS scores are used to determine which attribute to use as the base case for further analysis
with random parameter logits (RPLs) models. In particular, common practice is to set the base case
for a logit model as the largest or smallest BWS score. From the RPLs, BWS shares are calculated
and compared across groups using Poe tests to determine what statistically significant differences
in preferences exist between market groups.
74
4.1 Count Data and Descriptive Best-Worst Scaling Statistics
A descriptive analysis of the BWS data was conducted using BWS scores. BWS scores are
computed as the number of times an attribute was selected as most preferred, minus the number of
times that attribute was selected as least preferred, divided by the number of times the attribute
appears in the design. In this study, each attribute appeared three times in the design. These BWS
scores provide another descriptive method for analyzing the data, as well as inform the selection
of the base-case in logit models. In the RPL models discussed in the next section, I selected the
base case as the smallest BWS score (Inspections) for two out of the three survey groups.
4.2 Best-Worst Scaling Theoretical Foundation & Analytical Methodology
BWS is rooted in random utility theory (RUT), which assumes that agents seek to
maximize their expected utility subject to the choices they are presented (McFadden 1974). In
RUT, it is assumed that the relative preference for object A over object B is a function of the
relative frequency with which A is chosen as better than B for an individual (Louviere et al. 2013).
Individuals make choices randomly, with some error involved, to maximize their utility.
The best-worst scaling method presents each individual multiple answer options (in our analysis,
attributes of a NHMC program) and asks them to select one as “best” (or most important) and one
as “worst” (or least important). In practice, the BWS method consists of a series of several
questions, each comprised of different combinations of attributes per question. According to RUT,
the utility for respondent n in selecting alternative i in choice set t is:
𝑈!"# = 𝑉!"# + 𝜀!"#
where 𝑉!"# is the deterministic portion of utility dependent upon the attributes of the alternative
and 𝜀!"# is the stochastic component of utility, which is independently and identically distributed
over all alternatives and choice scenarios.
Equation 14
75
Generally, when respondents are presented with a choice set, they make choices based on
maximizing the utility they can receive from each alternative in the choice set. For example, in
making a choice between alternative j and alternative k, respondent n will pick alternative j over
alternative k when:
𝑈!$# > 𝑈!%# 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑗 ≠ 𝑘
Equation 15
Given that each choice set has J attributes, the pair of attributes chosen represents a choice
from all J(J-1) possible pairs, which maximizes the difference in importance. Following Lusk and
Briggeman (2009) and McKendree et al. (2018), let the true or latent unobservable level of
importance for individual n be represented by 𝐼3) = 𝜆) + 𝜖3) where 𝜆) represents j’s location on
the scale of importance and 𝜖3) is the random error term.
The probability that pair (j, k) is chosen out of a choice set with J attributes, where j
represents the most important attribute and k represents the least important attribute, is the
probability that the difference between j and k is larger than all the J(J – 1) – 1 other possible
differences in the choice set. When the error terms 𝜖3) are independent and identically distributed
type I extreme value, the multinomial logit (MNL) form of this probability is:
𝑃𝑟𝑜𝑏!(𝑗 𝑐ℎ𝑜𝑠𝑒𝑛 𝑎𝑠 𝑚𝑜𝑠𝑡, 𝑘 𝑎𝑠 𝑙𝑒𝑎𝑠𝑡) =
𝑒 &!’&"
(
)*+
𝑒 &!’&" − 𝐽
(
∑ ∑
,*+
From this probability statement, by maximizing the log-likelihood function, parameters 𝜆$ can be
estimated. When doing this, the dependent variable is 1 for the chosen most-least attribute pair and
Equation 16
0 for the remaining J(J – 1) – 1 pairs.
An MNL assumes respondents have homogenous views of the attributes analyzed.
However, past studies have found heterogenous preferences among agricultural supply chain
76
agents (e.g., McKendree et al. 2018, Schulz and Tonsor 2010, Ortega et. al 2019) and consumers
in relation to agricultural products (e.g., Ortega et. al 2011, Bazzani et. al 2017, Ubilava & Foster
2009, McKendree et. al 2013). Not only do individuals behave differently, it is likely they also
have different motivations behind these decisions. Accordingly, to account for response
heterogeneity, both an uncorrelated and correlated random parameters logit (RPL) were estimated
(Boxall and Adamowicz 2002) and compared to the MNL for each of the three groups surveyed.
The MNL and RPL models were conducted using NLogit 6 on the BWS data for all three
groups in the study. In each group, the correlated and uncorrelated RPLs resulted in statistically
significant standard deviations for all attributes and statistically significant coefficients for at least
some of the attributes; in the case of the correlated RPLs, the covariances were also strongly
statistically significant, confirming heterogeneity of opinions within each group. Thus, I chose to
use the results of the correlated RPLs for further analysis and comparisons across the groups.
RPL coefficient estimates cannot be directly interpreted. However, a “share of importance”
estimate based on a ratio scale can be calculated for each of the seven attributes of a hypothetical
U.S. national halal program included in the BWS portion of the surveys:
Share of importance for attribute j =
∑
4!"#
%
$&’
#
4!$
Equation 17
These shares provide a more intuitive approach to analyzing the data. The shares and p-
values were calculated in MATLAB with the coefficient matrices from NLogit 6. The sum of
shares among all seven attributes analyzed must equal one (or 100%). If attribute j has an
importance share of 0.3 (30%) and attribute k has an importance share of 0.1 (10%), then j is three
times as important as k. Krinsky-Robb confidence intervals (Krinsky & Robb 1986) were
calculated as a conservative way to compare statistical differences across importance shares both
77
within and across survey groups. Additionally, to test for differences in preference shares across
the samples, I use the full combinatorial method from Poe et. al (2005). This allows me to draw
conclusions for which attributes are most important to different groups, and how these preferences
are displayed across the market. The Krinsky-Robb and Poe tests were conducted in MATLAB
using the coefficient and variance matrices obtained from modeling the correlated RPLs in NLogit
6.
5. Results and Discussion
The results of the analysis methods used in this chapter and corresponding discussion are
presented by method type in the sections below.
5.1 Descriptive Statistics
Participant demographics by survey group are presented in Table 15 and Table 16. For
retailers and processors, the demographic statistics are representative of the survey respondent;
relevant operation demographics are included in Table 17.
78
Table 15: General Participant Demographics by Survey Group
Category
Processors
(%)
Retailers
(%)
Consumers
(%)
Gender (n = 94, n = 33, n = 507)
Male
Female
Prefer not to disclose
Education Level (n = 94, n = 32, n = 507)
Less than High School
High School
Some College
2-Year Degree (Associates)
4-Year Degree (Bachelor's)
Master's Degree
Professional Degree
Prefer not to Disclose
Race (n = 94, n = 36)
White
Black
Native American or Alaskan Native
Native Hawaiian or Pacific Islander
Asian
Other
Prefer not to disclose
Political Party (n = 94, n = 33, n = 507)
Democrat
Republican
Independent
Other
Prefer not to disclose
1st Generation Immigrant (n = 94, n = 33, n = 507)
No
Yes
Prefer not to disclose
2nd Generation Immigrant (n = 94, n = 33, n = 507)
No
Yes
Prefer not to disclose
Currently Religious (n = 94, n = 33)
Yes
No
Prefer not to disclose
Previously Religious (n = 20, n = 12)
Yes
No
Prefer not to disclose
76%
19%
5%
0%
6%
13%
10%
44%
22%
5%
0%
76%
1%
0%
0%
3%
6%
14%
5%
35%
26%
3%
31%
82%
10%
9%
66%
21%
13%
21%
56%
22%
45%
40%
15%
79
49%
50%
<1%
2%
20%
18%
11%
27%
17%
5%
<1%
40%
21%
28%
1%
9%
70%
27%
4%
52%
41%
8%
79%
21%
0%
0%
3%
13%
28%
25%
22%
9%
0%
66%
0%
5%
2%
10%
2%
2%
6%
27%
36%
6%
24%
91%
6%
3%
82%
15%
3%
36%
48%
15%
58%
42%
0%
Increased specificity for consumer demographics is important to include, as preferences
and opinions of Muslim consumers will be influenced more strongly and in different ways by their
race, ethnicity, sect of Islam, and status as a convert or revert to Islam (Table 16).
Percentage
45%
27%
23%
2%
8%
16%
<1%
3%
1%
Table 16: U.S. Muslim Halal Consumer Demographics: Race, Ethnicity, and Religion (n =
507)
Category
Race and Ethnicity
White
Black or African American
Middle Eastern or North African
American Indian or Alaskan Native
Asian
South or Southeast Asian
Native Hawaiian or Other Pacific Islander
Other
Prefer not to Disclose
Sect of Islam
Sunni
Shia or Shiite
Ibadi
Non-denominational
Other
Prefer not to disclose
Convert or Revert to Islam
Years Since Converting or Reverting to Islam (n = 165)
0-5 years
6-10 years
11-15 years
16-20 years
21-25 years
26-30 years
31-35 years
Over 35 years
Prefer not to Disclose
Previous Religion (n = 165)
Christianity
Hinduism
Buddhism
Judaism
Sikhism
Other
None
Prefer not to Disclose
23%
17%
17%
10%
7%
7%
5%
8%
6%
50%
9%
2%
5%
2%
5%
19%
7%
63%
11%
6%
13%
2%
5%
33%
80
Unsurprisingly, the demographic patterns of the supply-side agents differ from consumers.
While the majority of supply-side agents were male, white, and had at least a 4-year degree,
consumers were evenly split amongst genders, were a much wider variety of races, and were
generally less educated overall. Supply-side agents were more commonly Republicans or
Independents, while consumers were more likely to be Democrats. Consumers were more likely
to be first- or second-generation immigrants and 100% were currently religious (non-religious
individuals were screened out of the survey), compared to 21% and 36% of processors and
retailers, respectively. These general differences in survey groups may help to explain differences
in preferences and opinions across groups for a potential U.S. NHMC program.
Muslim consumers’ demographics provide even more detailed information that can explain
variability in preferences. While the largest group of consumers is white, there are also large
percentages of other races and ethnicities represented in the survey. Likewise, the majority of
consumers represented are Sunni Muslims (the most common across the world and the most likely
to follow stricter halal dietary standards), though other denominations are well represented.
Further, one-third of the sample were Islamic converts (also commonly referred to as reverts), with
the majority converting within the last 15 years. Recent converts may display different preferences
than life-long Muslims. Finally, one-half of converts were previously Christians, who may display
different preferences or opinions than converts from other religions or those who had not
previously followed a religion.
81
Retailers (n = 41)
Processors (n = 95)
48%
23%
23%
6%
7%
44%
45%
3%
N/A
N/A
N/A
Table 17: Processing and Retailing Establishment Summary Statistics
Category
Establishment Type
Slaughter without processing
Processing without slaughter
Slaughter and processing
Other
Grocery Store
Butcher Shop/Deli
Other
Location
Rural
Suburban
Urban
Prefer not to disclose
Type of Animals
Processed/Sold
Beef
Veal
Lamb
Pork
Turkey
Chicken
Goat
Exotics
Halal Status
Current Halal
Past Halal
Never Halal
Unsure
No. of Employees
83%
34%
55%
46%
35%
38%
55%
35%
34%
4%
62%
N/A
Mean
62
Mean
1992
Min
2
Min
1902
Max
850
Max
2022
Mean
45
Mean
1979
Year Established
N/A
N/A
N/A
N/A
80%
12%
7%
61%
33%
6%
N/A
100%
41%
56%
85%
95%
100%
17%
34%
22%
10%
51%
17%
Min
1
Min
1867
Min
600
Max
2021
The processors and retailer operation demographics in Table 17 show that supply-side
responses came from a variety of different business types, ages, and sizes. For processors, many
establishments slaughtered multiple species of animals; likewise, retailers sold products from
82
multiple species. Finally, halal businesses were represented in each sample. The variety of
businesses and heterogeneity in their characteristics helps to ensure the data collected is fairly
representative of each group’s national population.
5.2 Attribute Preferences
The count data and corresponding descriptive discussion of the best-worst question
responses is given in Appendix A.3. Figure 8 presents the descriptive results based on the BWS
scores. The scores reveal the most and least important attributes to consider for each of the three
groups. The BWS scores provide information on the ideal base case for the RPL model; in this
instance, I selected the least important attribute for two of the three groups (Inspections) as the
base case.
83
Figure 8: Best-Worst Scores for Processors, Retailers, and Consumers
Processor Best-Worst Scores
Costs
Information
Inspections
Enforcement
Who/What
Certified
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
Halal Standards
Benefits
Retailer Best-Worst Scores
Costs
Enforcement
Information
Inspections
Halal Standards
Who/What
Certified
-0.5
-0.4
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
Benefits
Consumer Best-Worst Scores
Costs
Information
Who/What Certified
Enforcement
Inspections
Halal Standards
Benefits
-0.3
-0.2
-0.1
0
0.1
0.2
0.3
0.4
0.5
The coefficients for the correlated RPLs are given in Table 18, while the additional
variance and correlation matrices for the correlated RPLs and the results for the uncorrelated RPLs
and MNLs are given in Appendix A.3. In all models, the Inspection attribute of a hypothetical U.S.
NHMC program – has been dropped to avoid multicollinearity. Therefore, the coefficients reflect
the importance of each of the six attributes relative to Inspection, which was normalized to zero
for identification purposes as it was the least important attribute to processors and retailers, and
the second least important attribute to consumers (based on the BWS scores).
84
The correlated RPL results are generally consistent with the BWS scores. For processors,
all attributes of the best-worst scales are statistically significant at least at the 5% level, excluding
Benefits, which was not significant. For retailers, all attributes are statistically significant at least
at the 10% level, excluding Enforcement and Halal Standards; for consumers, all attributes are
statistically significant at the 10% level, excluding Benefits. The results indicate that Who/What
is Certified is the most preferred attribute for retailers and processors, closely followed by Halal
Standards and Costs for processors, as well as Costs for retailers. For consumers, Halal Standards
and Who/What is Certified are the top two preferred attributes. On the other hand, Information
was the least preferred attribute (least important to consider) for processors and retailers, while
consumers indicated that the Costs were least important.
The importance shares from the correlated RPLs are given in Table 19. Importance shares
results for the uncorrelated RPLs and MNLs are given in Appendix A.3. Results show that 23%
of processors view the attributes Who/What is Certified and Halal Standards as the most important
attribute for program designers to consider when developing a hypothetical U.S. NHMC program.
Who/What is Certified was the most important attribute for program designers to consider for
retailers (27%) and the second most important for program designers to consider for consumers
(18%). Halal Standards was the fourth most important attribute for program designers to consider
for retailers (13%) and the most important attribute for program designers to consider for
consumers (22%). I see the most consensus in rankings of attribute importance between the two
supply chain members (processors and retailers), while the preferences of consumers are more
divergent. These differences between the supply and demand sides of the market are unsurprising
and illustrate the general challenges of food policy making and, more specifically, the diversity of
85
preferences for what is most important to affected market participants when designing a NHMC
program.
Finally, I conducted pair-wise Poe tests on the preference shares associated with each of
the seven attributes considered in the BWS for a hypothetical NHMC program. Table 20 presents
the p-values from the Poe tests, with values under 0.05 indicating the two groups’ shares are
statistically different at the 5% level. Poe test results for the uncorrelated RPLs and MNLs are
given in Appendix A.3. Overall, the results reveal patterns of preferences between the three groups
in the study. Unsurprisingly, the Poe test shows that consumers’ preferences across the attributes
of a hypothetical national U.S. halal certification program are significantly different from those of
the processors and retailers. Consumers place stronger, more positive emphasis on Inspection
relative to retailers and Halal Standards. On the other hand, consumers place stronger and more
negative emphasis on the Costs and Who/What is Certified than both processors and retailers, and
Benefits and Inspection when compared to processors. As for processors and retailers, their
responses are overall very similar; there are only two statistically significant differences on the
attributes Enforcement and Benefits. The BWS shares indicated that processors valued
Enforcement more than retailers, while retailers valued Benefits more than processors.
86
Table 18: Correlated Random Parameters Logit Results of Best-Worst Scaling for Hypothetical U.S. National Halal Meat and
Poultry Certification Program: Processors (n = 82), Retailers (n = 33), and Consumers (n = 507)
Best-Worst Scaling Attributes
Cost of Certification
Enforcement and Regulation
Information Collected/Available
Who/What Must be Certified
Which Halal Standards are Included
Benefits of Certification
Processors
Coefficient
(st. error)
0.92***
(0.16)
0.29**
(0.14)
-0.81***
(0.16)
0.99***
(0.15)
0.97***
(0.15)
-0.047
(0.15)
Retailers
Coefficient
(st. error)
0.91***
(0.25)
-0.18
(0.23)
-0.79***
(0.24)
1.05***
(0.23)
0.35
(0.24)
0.40*
(0.24)
Consumers
Coefficient
(st. error)
-0.35***
(0.05)
-0.09*
(0.05)
-0.20***
(0.05)
0.29***
(0.05)
0.52***
(0.05)
0.07
(0.05)
McFadden Pseudo R2
N
Log likelihood
AIC
AIC/N
Note: *** indicates p < 0.01, ** indicates p < 0.05, and * indicates p < 0.1. Base attribute normalized to zero is Inspection Type/Frequency.
0.23
574
-860.64
1775.30
3.09
0.24
231
-342.56
739.10
3.20
0.12
3549
-6053.56
12161.1
3.43
87
Table 19: Best-Worst Scaling Shares of Preferences for Hypothetical U.S. National Halal Meat and Poultry Certification
Program: Consumers, Retailers, & Processors: Correlated Random Parameters Logit
National Halal Meat & Poultry
Certification Program Attribute
Cost of Certification
Enforcement and Regulation
Information Collected/Available
Who/What Must be Certified
Inspection Frequency and Type
Which Halal Standards are Included
Benefits of Certification
Processors (n = 82)
Share
0.22***
0.12***
0.04***
0.23***
0.09***
0.23***
0.08***
95% CI
[0.18, 0.26]
[0.10, 0.14]
[0.03, 0.05]
[0.20, 0.27]
[0.07, 0.10]
[0.19, 0.27]
[0.07, 0.10]
Retailers (n = 33)
Share
0.24***
0.08***
0.04***
0.27***
0.09***
0.13***
0.14***
95% CI
[0.17, 0.30]
[0.06, 0.11]
[0.03, 0.06]
[0.21, 0.34]
[0.07, 0.13]
[0.09, 0.19]
[0.10, 0.19]
Consumers (n = 507)
95% CI
Share
[0.09, 0.10]
0.09***
[0.11, 0.13]
0.12***
[0.10, 0.12]
0.11***
[0.17, 0.19]
0.18***
[0.12, 0.14]
0.13***
[0.21, 0.24]
0.22***
[0.13, 0.15]
0.14***
Note: *** indicates p < 0.01, ** indicates p < 0.05, and * indicates p < 0.1.
Table 20: P-values from Poe tests for Hypothetical U.S. National Halal Meat and Poultry Certification Program: Pair-wise
Comparisons Between Processors (n = 82), Retailers (n = 33), and Consumers (507)
National Halal Meat & Poultry
Certification Program Attribute
Cost of Certification
Enforcement and Regulation
Information Collected/Available
Who/What Must be Certified
Inspection Frequency and Type
Which Halal Standards are Included
Benefits of Certification
Processors vs. Retailers
Processors vs. Consumers Retailers vs. Consumers
0.68
0.03
0.34
0.16
0.33
1.00
0.00
0.00
0.68
0.01
0.00
0.00
0.58
0.00
0.00
0.99
0.01
0.00
0.02
0.00
0.44
Note: Values that are statistically significant at the 5% level or better are bolded.
88
5.3 Follow-Up Questions: Certification Specifics
Follow-up questions for each of the BWS attributes allow me to dig deeper into the nature
of participants’ preferences for a hypothetical U.S. NHMC program. Descriptive statistics for these
follow-up questions are given by survey group in Table 21-Table 30. Note that the number of
respondents for processors and retailers is slightly higher for these questions than the BWS
questions, as not all processors and retailers completed all BWS questions, thus excluding these
participants from the MNL and RPL panel analyses.
The BWS scores, RPL coefficients, and BWS shares for the three groups provided
consistent preference patterns for different attributes of a hypothetical U.S. NHMC program. To
summarize, processors’ results showed Who/What is Certified, Halal Standards, and Costs as the
most important attributes for program designers to consider when developing a hypothetical U.S.
NHMC program. Retailers’ results showed Who/What is Certified and Costs as the two most
important attributes for program designers to consider when developing a hypothetical U.S.
NHMC program. Muslim halal consumers’ results identified Halal Standards and Who/What is
Certified as the two most important attributes for program designers to consider when developing
a hypothetical U.S. NHMC program. I now discuss these results in more detail by group – and by
halal versus non-halal supply chain members where applicable – using additional information
gathered from the follow-up questions and qualitative interview quotes for further context.
Table 21 provides more detail on the Halal Standards preferred by group. Between non-
halal processors and halal processors only, the results show an increased preference for each of the
given possible standards by halal processors. This increase in strength of preferences makes sense,
as halal processors are both more knowledgeable of the halal process and have more at stake when
defining standards than those currently outside of the market. These responses also support the
89
interview findings that halal processors generally favor standards that are more closely associated
with hand slaughter (e.g., zabiha slaughter method, slaughterers of Muslim faith, individual spoken
blessings, and no stunning).
For halal consumers, the results show more variability in preferences, with consumers
indicating stronger preferences for additional quality attributes not typically included in current
halal certifications used in the U.S. (e.g., animals face Mecca and non-GMO) (Table 21). These
variations in consumers’ preferences and the generally smaller percentages of consumers
indicating that any one standard was preferred are unsurprising given the diversity in Islamic sect,
racial, and ethnic backgrounds of the consumer sample. Furthermore, consumers interviewed in
the earlier phase of this project who are less strict in their halal dietary requirements indicated little
to no preference between zabiha and machine slaughter or the faith of the slaughterer, for example.
Table 22 provides more information on opinions on the attribute Who/What is Certified.
For all retailers and processors, the majority want a hypothetical U.S. national halal certification
program to require certification of individual products (>63%) and supply chain members (>50%).
For those that want supply chain members to have certification, the supply chain agents indicated
slaughter (>91%) and processing (>83%) establishments were the most preferred to be certified.
A higher percentage of halal processors and retailers indicated that supply chain agents should be
certified, which matches the opinions of nearly all halal processors and retailers interviewed. The
interviewees valued a halal supply chain with strict and well-defined certifications for each
member to ensure their good name and product quality were maintained. Additionally, halal
retailers interviewed indicated they rely heavily on their reputation within their local communities
and word-of-mouth of happy customers; thus, strong certification utilization is important.
90
I asked processors and retailers to indicate which of the given costs of a U.S. NHMC
program would be most important to their businesses (Table 23). The cost categories indicated as
important were consistent between non-halal processors and halal processors, with the most
notable difference being the costs of potential establishment modifications; 67% of non-halal
processors and 34% of halal processors indicated this was an important factor. The higher
percentage of non-halal processors who considered the costs of establishment modifications to be
important is logical, as these processors do not currently have a halal program and therefore their
establishments may not currently have the infrastructure to add a halal program. There were
notable differences in which cost categories were most important for non-halal retailers and halal
retailers. Forty-one percent of non-halal retailers and 67% of halal retailers indicated certification
fees were an important factor, while 50% of non-halal retailers and 33% of halal retailers indicated
the costs of potential establishment modifications were important.
Table 21: Market Participants’ Preferred Halal Standards for a U.S. National Halal Meat
and Poultry Certification Program
Preferred Halal
Standards
Zabiha (hand-slaughter)
Machine slaughter
Slaughterers of Muslim
faith
Slaughterers of Jewish or
Christian faith
Individual spoken
blessings
Animals not stunned
Animals face Mecca
Non-GMO
Other
Non-Halal
Processors
(n = 66)
76%
9%
Halal
Processors
(n = 32)
88%
22%
Non-Halal
Retailers
(n = 32)
22%
16%
Halal
Retailer
s (n = 9)
78%
0%
Halal
Consumers
(n = 507)
53%
19%
6%
9%
16%
22%
3%
22%
13%
67%
33%
33%
44%
33%
11%
0%
58%
14%
36%
35%
33%
32%
0%
12%
8%
8%
3%
3%
3%
8%
75%
16%
47%
38%
22%
6%
6%
91
Table 22: Market Participants’ Opinions on Who or What Should be Certified Under a
U.S. National Halal Meat and Poultry Certification Program
Non-
Halal
Retailers
(n = 32)
63%
50%
Non-Halal
Processors
(n = 66)
Halal
Processors
(n = 32)
Halal
Retailers
(n = 9)
Halal
Consumers
(n = 507)
Who or What Certified
75%
72%
79%
77%
78%
67%
80%
58%
Individual Products
Supply Chain Members
Which Supply Chain
Members? (n = 61, 38,
22, 16, 392)
Slaughter
Establishments
Processing
Establishments
Transportation/
Distributors
Retailers/Wholesalers
Restaurants and Food
Service
Other
92%
84%
24%
32%
34%
5%
96%
91%
57%
65%
61%
0%
94%
83%
88%
100%
31%
44%
38%
0%
100%
100%
100%
0%
76%
66%
52%
66%
57%
1%
Table 23: Costs of U.S. National Halal Meat and Poultry Certification Considered Most
Important
Program Cost Category
Certification Fees
Establishment Modifications
Increased Labor Hours Needed
Cost of Traceability Equipment
Other
Non-Halal
Processors
(n = 66)
35%
67%
42%
38%
2%
Halal
Processors
(n = 32)
47%
34%
28%
31%
0%
Non-Halal
Retailers
(n = 32)
41%
50%
28%
44%
0%
Halal
Retailers
(n = 9)
67%
33%
33%
44%
0%
5.4 Follow-Up Questions: Program Implementation and Transparency
The development of a U.S. NHMC program will be complex and subject to a wide array
of market participants’ preferences, as seen in previous U.S. national certification program
development processes (e.g., USDA Organic). Nonetheless, the results of my surveys and analysis
provide initial context and guidance for the potential future development of such a program. There
are no glaring differences in opinions and preferences that would preclude the development of a
92
national program across the three study groups. However, other factors must be considered when
designing a national-level certification program; specifically, who would set the standards for and
enforce such a program, how certifications would be administered and audited, and the amount
and nature of program transparency.
5.4.1 Program Standards and Enforcement
Halal is a religious attribute of a food product and is defined by interpreting religious texts;
thus, the First Amendment to the U.S. Constitution prohibits federal, state, and local governments
from setting or specifying halal standards. Therefore, the choice of which halal standards to include
or exclude from a national program would fall to non-government organizations (NGOs) –
potentially including consumer advocacy groups, religious organizations, certifier and producer
organizations, and supply chain members. Table 24 shows market participants’ opinions of which
of these groups should have a say in setting these standards. Reaching a consensus on halal
standards amongst so many different voices is likely to be difficult; therefore, identifying which
types of organizations should lead this effort is critical. In interviews, multiple halal processors,
halal retailers, and Muslim halal consumers all indicated that religious organizations were their
most preferred option for setting halal standards, followed to a lesser extent by certifiers and non-
government groups. Indeed, based on the results of market participants’ opinions presented here,
the ideal candidates to work together to set halal standards should be religious (44-89%), certifier
(31-48%) and non-government (11-28%) organizations.
Enforcement of a predetermined religious standard is not prohibited by the First
Amendment; therefore, options for program enforcement include federal and state government
organizations such as the USDA or state-level departments of agriculture. With this expansion of
options, the organizations most preferred to be in charge of program management and enforcement
93
are religious (22-67%), U.S. government (11-53%), and certifier organizations (19-35%). These
findings match interview findings; multiple halal processors, retailers, and consumers all expressed
their desire for a U.S. government-backed certification program, with input from religious and
certifier organizations.
94
Table 24: Market Participants’ Opinions on Who Should Set and Enforce Standards for a U.S. National Halal Meat and
Poultry Certification Program, Aggregated
Organizations
Non-Halal
Processors
(n = 66)
Halal
Processors
(n = 32)
Non-Halal
Retailers
(n = 32)
Halal
Retailers
(n = 9)
Halal
Consumers
(n = 507)
Set
U.S. Government Organizations N/A
State Government Organizations N/A
23%
Non-Government Organizations
49%
Religious Organizations
32%
Certifier Organizations
20%
Producer Organizations
28%
Slaughterers & Processors
5%
Wholesalers & Distributors
3%
Retailers & Restaurants
6%
Other
Note: Due to the religious freedom protections of the First Amendment, governments are prohibited from setting or defining religious practices; in this case, halal
standards.
Enforce
44%
18%
14%
30%
33%
11%
11%
3%
0%
3%
Enforce
51%
35%
21%
49%
35%
23%
31%
23%
18%
0%
Enforce
53%
22%
19%
44%
19%
0%
19%
0%
0%
3%
Enforce
11%
22%
44%
67%
33%
0%
22%
0%
0%
0%
Enforce
28%
28%
16%
22%
34%
6%
13%
3%
3%
0%
Set
N/A
N/A
28%
63%
31%
9%
28%
3%
3%
6%
Set
N/A
N/A
25%
66%
48%
30%
29%
42%
22%
0%
Set
N/A
N/A
28%
44%
34%
19%
6%
25%
6%
0%
Set
N/A
N/A
11%
89%
33%
22%
0%
33%
0%
0%
95
5.4.2 Certification Administration and Auditing
Proper enforcement of a U.S. NHMC program will require defined standards for
administering certifications and auditing certified agents and products. However, administration
and auditing of certifications is more than just setting standards – a defined method and timeline
for inspections is ideal to ensure initial and ongoing compliance. Table 25 shows survey
participants’ preferences for the types of inspections or audits that should be conducted to ensure
supply chain members’ compliance and retain certification. The results show that the majority of
the market prefers random or a mixture of scheduled and random inspections or audits. This is
unsurprising, as multiple interview participants across all three groups expressed their concern that
exclusively using scheduled audits could allow for certifications to be granted to dishonest supply
chain members. Furthermore, these preferences are in line with the nature in which many
certification programs are managed in a variety of contexts, and therefore should not pose a
challenge to implement.
Table 25: Market Participants’ Opinions for how Certified Suppliers Under a U.S.
National Halal Meat and Poultry Certification Program Should be Audited
Inspection Type
Non-Halal
Processors
(n = 38)
47%
5%
Halal
Processors
(n = 23)
22%
22%
Non-Halal
Retailers
(n = 16)
44%
6%
Halal
Retailers
(n = 6)
0%
50%
Halal
Consumers
(n = 392)
19%
29%
Scheduled
Random
Mixed Scheduled
& Random
Other
Note: The number of participants is lower for this question, as only survey participants who indicated that they
preferred supply chain members be required to hold certification saw this question in the survey.
50%
52%
42%
57%
50%
0%
0%
5%
0%
0%
5.4.3 Program Transparency and Traceability
Finally, it is vital to consider the transparency and traceability of a certification granted
under a hypothetical U.S. NHMC program. Consumers are increasingly interested in having access
to information about where and how their food is produced, as well as having the ability to self-
96
authenticate labels and other quality indicators in real-time. Likewise, supply chain members are
in favor of employing transparency and traceability efforts to ensure their products are viewed as
authentic and trustworthy to their customers. I asked survey participants about who should have
access to four main types of information that relate to a hypothetical U.S. NHMC program; the
results are shown in Table 26 - Table 29.
First, survey participants were asked which groups should have access to traceability
information for individual products (Table 26) and who should have access to a list of all certified
establishments (Table 27). The majority of survey participants indicated that the general public,
slaughter and processing establishments, wholesalers and distributors, and retailers and restaurants
should have access to both of these types of information. Halal supply chain members were more
in favor of these groups having access to this information than non-halal supply chain members.
Second, survey participants were asked who should have access to a list of all enforcement
agencies that would be responsible for issuing and auditing certifications (Table 28) and who
should have access to the list of halal standards included in the certification (Table 29). Again,
most survey participants indicated that the general public, slaughter and processing establishments,
wholesalers and distributors, and retailers and restaurants should have access to this information;
additionally, the majority of halal processors again indicated that government organizations should
be able to access this information. Again, halal supply chain members were more in favor of these
groups having access to this information than non-halal supply chain members. Finally, I asked
survey participants how the general public should have access to any of the traceability or
transparency information asked about in these four questions; these results are shown in Table 30.
. All three groups were strongly in favor of online access and halal processors, non-halal
and halal retailers, and consumers also favored the ability to use a QR code or cell phone app.
97
These responses show that generally, there is strong interest in ensuring a U.S. NHMC program
has robust transparency and traceability attributes. Overall, the results of these survey questions
align with opinions expressed by interviewees; the majority of interviewees were in favor of a U.S.
NHMC with robust transparency and traceability attributes.
98
Table 26: Market Participants’ Opinions on Which Groups Should Have Access to Traceability Information for Individual
Products Under a U.S. National Halal Meat and Poultry Certification Program
Groups That Should Have Access to
Traceability Information
General Public
Slaughter & Processing Establishments
Wholesalers & Distributors
Retailers & Restaurants
Government Organizations
None
Non-Halal
Processors
(n = 66)
59%
59%
59%
52%
41%
8%
Halal
Processors
(n = 32)
84%
75%
81%
84%
66%
0%
Non-Halal
Retailers
(n = 32)
53%
53%
59%
53%
31%
3%
Halal
Retailers
(n = 9)
56%
78%
78%
56%
33%
0%
Halal
Consumers
(n = 507)
62%
59%
55%
55%
32%
2%
Table 27: Market Participants’ Opinions on Which Groups Should Have Access to a List of all Certified Establishments Under
a U.S. National Halal Meat and Poultry Certification Program
Groups That Should Have Access to
List of all Certified Establishments
General Public
Slaughter & Processing Establishments
Wholesalers & Distributors
Retailers & Restaurants
Government Organizations
None
Non-Halal
Processors
(n = 66)
64%
59%
64%
58%
47%
9%
Halal
Processors
(n = 32)
88%
81%
88%
88%
66%
0%
Non-Halal
Retailers
(n = 32)
59%
53%
56%
56%
38%
3%
Halal
Retailers
(n = 9)
67%
67%
67%
56%
33%
0%
Halal
Consumers
(n = 507)
60%
59%
55%
55%
33%
2%
99
Table 28: Market Participants’ Opinions on Which Groups Should Have Access to a List of all Enforcement Agencies Under a
U.S. National Halal Meat and Poultry Certification Program
Groups That Should Have Access to
List of all Enforcement Agencies
General Public
Slaughter & Processing Establishments
Wholesalers & Distributors
Retailers & Restaurants
Government Organizations
None
Non-Halal
Processors
(n = 66)
55%
56%
53%
45%
49%
11%
Halal
Processors
(n = 32)
81%
75%
75%
75%
66%
0%
Non-Halal
Retailers
(n = 32)
53%
53%
59%
53%
34%
3%
Halal
Retailers
(n = 9)
78%
56%
67%
56%
33%
0%
Halal
Consumers
(n = 507)
60%
53%
53%
50%
35%
2%
Table 29: Market Participants’ Opinions on Which Groups Should Have Access to a List of Halal Standards Included in a U.S.
National Halal Meat and Poultry Certification Program
Groups That Should Have Access to
List of Halal Standards Used
General Public
Slaughter & Processing Establishments
Wholesalers & Distributors
Retailers & Restaurants
Government Organizations
None
Non-Halal
Processors
(n = 66)
64%
62%
59%
52%
48%
8%
Halal
Processors
(n = 32)
81%
81%
78%
72%
69%
0%
Non-Halal
Retailers
(n = 32)
53%
53%
53%
50%
44%
3%
Halal
Retailers
(n = 9)
78%
56%
67%
56%
33%
0%
Halal
Consumers
(n = 507)
57%
58%
57%
55%
33%
2%
100
Table 30: Market Participants’ Opinions on how the General Public Should Have Access to Information Related to a U.S.
National Halal Meat and Poultry Certification Program
Information Access Method
Halal
Processors
(n = 29)
100%
Online website
52%
Using a QR code or cell phone app
31%
Freedom of Information Act (FOIA)
0%
Other
Note: The number of participants is lower for this question, as only survey participants who indicated that the general public should have access to information
related to a U.S. NHMC program saw this question in the survey.
Halal
Consumers
(n = 411)
74%
53%
55%
<1%
Non-Halal
Processors
(n = 46)
100%
39%
35%
0%
Non-Halal
Retailers
(n = 20)
95%
65%
35%
0%
Halal
Retailers
(n = 7)
100%
86%
0%
0%
101
6. Implications for Implementation
As previously discussed, the modern-day U.S. halal meat and poultry certification
landscape is complex. There are many competing players with differing standards, approaches to
certification, and market segments. This complicates efforts to develop a uniform standard for the
U.S. market. However, halal meat and poultry market participants’ interest in a U.S. government-
backed approach to increased market regulation is coming to fruition, though not specifically for
halal products. The USDA under the Biden administration has made efforts to expand
transparency, diversity, and accessibility in local meat and poultry processing, via the Executive
Order on Promoting Competition in the American Economy, the 2021 American Rescue Plan, and
the USDA’s Meat and Poultry Supply Chain initiatives, among other methods (The United States
Government 2022, United States Department of Agriculture 2023).
Additional concerns of halal market participants can be addressed by the federal
government. Supply-side survey respondents indicated that the potential costs– such as
certification fees and establishment modifications – of participating in a NHMC program are of
concern; however, it is possible that federal grant programs could alleviate these financial burdens.
The Biden administration has provided the USDA with increased funding for grant programs
including the Meat and Poultry Inspection Readiness Grant Program (MPIRG), the Meat and
Poultry Processing Capacity Technical Assistance Program (MPPTA), the Local Meat Capacity
Grant (LocalMCap), and the Indigenous Animals Grant (IAG), which together aim to support the
growth of the meat and poultry supply chain, increase access to inspection and certification
programs, expand domestic processing capacity, and improve the ability of independent facilities
to serve more customers in more markets (United States Department of Agriculture 2023). These
grants are available to very small and small meat processors to “Increas[e] access to
102
slaughter/processing facilities for smaller farms and ranches, new and beginning farmers and
ranchers, socially disadvantaged producers, veteran producers, and/or underserved communities”
and promote efforts for “developing new and expanding existing markets” (United States
Department of Agriculture Agricultural Marketing Service 2024). These efforts are important
steps for the overall U.S. meat and poultry supply chain that also benefit the halal market.
Furthermore, additional rulemaking endeavors and investigations are underway by the
USDA’s Packers and Stockyards Division to expand and strengthen the USDA’s ability and
authority to foster and regulate a more diverse, equitable, and transparent meat and poultry supply
chain (The United States Government 2022). Results of these investigations may provide new legal
precedent for enforcing and regulating transparency and labeling in the meat and poultry supply
chain, which will address some of halal market participants’ concerns about the appropriate
utilization of halal certifications and information accessibility. Likewise, the new rules will enable
the USDA to prosecute unfair, deceptive, and anticompetitive behavior in the meat and poultry
industry, which will strengthen the authenticity of labeled products. Altogether, the financial and
legal efforts at the federal level point to increased interest in revitalizing the U.S. domestic meat
and poultry industry. This revitalization will include diversification of the products available to
consumers – such as certified halal products – and improvements in enforcement of traceability,
labeling, and other authenticity verification strategies that would positively impact the certification
landscape of the halal meat and poultry market. While these efforts are not explicitly directed at
the U.S. domestic halal meat and poultry certification landscape, they have and will continue to
improve upon certification-related issues such as transparency, information access, certification
costs and benefits, enforcement, and regular inspections – all of which are areas of concern for the
halal market.
103
Taken together, public opinion and current presidential leadership do not provide a clear
answer as to whether a NHMC program will be developed in the near future, nor how it would
likely be implemented. However, a more transparent, well-regulated U.S. domestic halal meat and
poultry market is possible without direct federal intervention in the halal market, as discussed
above. Further, consumer right-to-know legislation provides legal protections for enforcing
appropriate certification utilization without violating the religious protections in the First
Amendment. Thus, we have the necessary legal framework to facilitate the proper use of a U.S.
domestic NHMC program, were one to be developed. The only remaining piece of the certification
implementation process is the certification program itself. The results of this study help identify
market’s preferences for the attributes of a NHMC program. However, a significant hurdle in
developing a NHMC will be reaching a consensus on standards amongst the many different groups
in the market. If this can be achieved, U.S. domestic halal meat certifiers develop the NHMC
program, thereby simplifying and strengthening U.S. domestic halal meat and poultry certification.
Finally, the NHMC program can be implemented effectively by coordinating with federal and state
governments as needed for proper labeling enforcement and management.
However, the development of a national certification program in the U.S. is complex, as
seen in the years-long refinement and adoption of the USDA organic certification. Nonetheless,
the logistical and institutional experience of developing the USDA organic certification can inform
and aid in the process of designing a U.S. national halal meat certification program. Most notably,
perhaps, is the way in which program development can incorporate the preferences of stakeholders
and consumers. The variation in opinions and preferences for a NHMC program’s standards and
structure from market agents and how they may impact how a NHMC program would operate are
important to consider for the program to be successful. Additionally, when developing such a
104
program, the U.S. can take cues from current halal meat and poultry certification programs in place
throughout the world, in terms of organizational structure, transparency standards, implementation
process, and many other attributes.
7. Conclusions, Limitations, and Future Work
There is a lack of clarity and standardization in certifications in the U.S. halal meat and
poultry market that makes it challenging for participant – including processors, retailers, and
consumers – to engage fully and confidently in the market. Indeed, in interviews conducted for
this research, Muslim halal meat consumers and current halal meat retailers and processors
expressed concern over the lack of transparent and standardized certification requirements and
shared desires for the development of such a system. These issues leave much to be desired for
Muslim Americans. Thus, the purpose of this study was to understand what the U.S. domestic halal
market participants want out of a NHMC program, and how these desires compare to what is
feasible to implement. I achieved this objective by investigating the preferences for a NHMC
program for processors, retailers, and Muslim halal meat consumers. I used a mixed methods
design employing qualitative interviews and stacked surveys with BWS. In all, the findings of this
chapter aid in bolstering halal meat consumer confidence in the authenticity of their products, as
well as improve the equity of the U.S. food system.
My analysis reveals the preferences of market participants for the design and
implementation of a U.S. NHMC program and the potential challenges that may be faced in
developing such a program. By studying patterns in processors’, retailers’, and consumers’
preferences for different attributes of a U.S. NHMC program, I determine common characteristics
that should be carefully considered in program design to meet the market’s needs. Results show
that the market overall prefers that program designers consider most carefully Who/What is
105
Certified, Halal Standards, and Costs when developing a U.S. NHMC program. The results show
that the implementation of a NHMC designed by non-governmental agencies and backed by
federal consumer-right-to-know legislation may be ideal. In this manner, the First Amendment is
not violated, but Muslim Americans are granted further religious security. Finally, the data
indicated that multiple transparency and traceability measures should be included to ensure a
robust and trustworthy program.
The hurdles to designing a U.S. NHMC program described in this research are notable, but
not insurmountable. The design of a U.S. NHMC program will involve multiple groups’ opinions
and the need to consider a variety of religious and non-religious preferences for program attributes.
The most difficult part of the process will be reaching a consensus on standards amongst many
different groups without government involvement, and then coordinating with federal and state
governments as needed for proper enforcement and management without violating the First
Amendment. Despite this challenge, the results of the analyses described in this work provide
detailed information on which attributes are important to consider while developing a U.S. NHMC
program, and also shed light on how to best suit the needs and wants of the market. With this
information, program designers will be well equipped to develop a U.S. NHMC program.
Furthermore, steps to strengthen the U.S. meat and poultry industry’s transparency, equity, and
authenticity in general have already begun to be implemented by the Biden administration and the
USDA, and consumer right-to-know legislation is in place to enforce proper certification
utilization in the food system overall. Thus, the U.S. domestic halal meat and poultry market is in
106
a prime position to implement a national certification; the only missing piece is the certification
program itself.
Despite these novel findings, there are some limitations of this study. First, collecting data
from supply-side agents in any market is notoriously difficult, which led to lower response rates
and lower numbers of quality observations in my BWS datasets. As such, the sample sizes for
processors and retailers were much smaller than the sample of consumers and may not be
representative of the industry in this analysis. However, the findings from the interviews are
consistent with those from the surveys, suggesting that the results and conclusions are reasonably
sound and representative of the market despite the small sample sizes.
Moving forward there is need for additional research focused on the potential structure and
design of a U.S. NHMC program. As my analysis is exploratory in nature, additional research
should be conducted to better describe the attribute preferences expressed in the BWS. A deeper
understanding of these preferences would aid in the design of a more effective program. To achieve
this, future work would benefit from larger sample sizes for processors and retailers to ensure more
representative data. As in the case of Essay 1, there are two methods that I believe could be
effective for increasing sample size when working with these populations. First, utilizing a team
of researchers to conduct in-person or virtual (e.g., Zoom or phone) surveys in real-time may
increase response rates and quality, as processors and retailers are typically less likely to complete
surveys. Secondly, some of the U.S. halal processor and retailer communities are nonnative
English speakers, especially older individuals and recent immigrants. As such, researchers may
benefit from close partnerships with native Arabic and Urdu speakers when collecting data from
halal businesses to increase participation rates.
107
Additionally, future investigation into the design and implementation of a U.S. NHMC
program would benefit from more information on the process for making complex food and
agricultural policies, regulations and laws, especially in the case of religious standards. Potential
avenues to acquire this knowledge would be interviews with representatives of the USDA,
Congresspersons, individuals who helped to design the USDA Organic standards, and government
or certifier representatives from other countries around the world that have national level halal
meat and poultry certification programs.
Overall, my findings and additional suggestions for additional research will help future
work make meaningful contributions to our understanding and support of regulation and
certification within this unique market, as well as add to the literature on the design of policies to
support both supply-side agents and consumers.
108
BIBLIOGRAPHY
An Update on Food Fraud. (2020, January 31). Food Safety Net Services (FSNS). Retrieved
November 15, 2021, from https://fsns.com/2020/01/31/an-update-on-food-fraud/#.
Australian Halal. (2022). Australian Halal. https://australiahalal.com/en/index.html
Bandoim, Lana. 2019. “Can Blockchain and Chip Technology Improve Beef Sourcing
Transparency?” Forbes. Accessed Aug 8, 2022.
https://www.forbes.com/sites/lanabandoim/2019/04/30/can-blockchain-and-chip- technology-
improve-beef-sourcing-transparency/#441ecf473284
Batz, F. ‐J., Peters, K. J., & Janssen, W. (1999). The influence of technology characteristics on the
rate and speed of adoption. Agricultural Economics, 21(2), 121–130.
https://doi.org/10.1111/j.1574-0862.1999.tb00588.x
Bazzani, C., Caputo, V., Nayga, R. M., & Canavari, M. (2017). TESTING COMMITMENT COST
THEORY IN CHOICE EXPERIMENTS: COMMITMENT COST IN CHOICE
EXPERIMENTS. Economic Inquiry, 55(1), 383–396. https://doi.org/10.1111/ecin.12377
Boxall, P.C., Adamowicz, W.L. Understanding Heterogeneous Preferences in Random Utility
Models: A Latent Class Approach. Environmental and Resource Economics 23, 421–446 (2002).
https://doi.org/10.1023/A:1021351721619
Caswell, J. A. (1998). How Labeling of Safety and Process Attributes Affects Markets for Food.
Agricultural and Resource Economics Review, 27(2), 151–158.
https://doi.org/10.1017/S106828050000647X.
Collart, A. J., & Canales, E. (2022). How might broad adoption of blockchain‐based traceability
impact the U.S. fresh produce supply chain? Applied Economic Perspectives and Policy, 44(1),
219–236. https://doi.org/10.1002/aepp.13134
Economically Motivated Adulteration (Food Fraud). (2021, November 4). U.S. Food & Drug
Administration (FDA). Retrieved November 15, 2021 from
https://www.fda.gov/food/compliance-enforcement-food/economically-motivated-adulteration-
food-fraud.
El-Osta, H. S., & Morehart, M. J. (2000). Technology Adoption and Its Impact on Production
Performance of Dairy Operations. Review of Agricultural Economics, 22(2), 477–498.
https://doi.org/10.1111/1058-7195.00034
Finn, A., & Louviere, J. J. (1992). Determining the Appropriate Response to Evidence of Public
Concern: The Case of Food Safety. Journal of Public Policy & Marketing, 11(2), 12–25.
https://doi.org/10.1177/074391569201100202
109
Guthman, J. (1998). Regulating Meaning, Appropriating Nature: The Codification of California
Organic Agriculture. Antipode, 30(2), 135–154. https://doi.org/10.1111/1467-8330.00071
Halal Food Act, 410 Ill. Compiled Statute 637 (2002).
https://www.ilga.gov/legislation/ilcs/ilcs3.asp?ActID=1581&ChapterID=35.
Halal Food Market: Global Industry Trends, Share, Size, Growth, Opportunity and Forecast 2021-
2026. (2021, April). ResearchandMarkets.com. Retrieved November 15, 2021 from
https://www.researchandmarkets.com/reports/5311860/halal-food-market-global-industry-trends-
share?utm_source=BW&utm_medium=PressRelease&utm_code=5w4lr3&utm_campaign=1527
350+-+%241.9+Trillion+Halal+Food+Market+-
+Global+Industry+Trends%2c+Share%2c+Size%2c+Growth%2c+Opportunity+and+Forecast+t
o+2026&utm_exec=chdo54prd.
Halal Food Market in U.S. 2020-2024. (2020, June). ResearchandMarkets.com. Retrieved November
15, 2021 from https://www.researchandmarkets.com/reports/5125564/halal-food-market-in-us-
2020-
2024?utm_source=BW&utm_medium=PressRelease&utm_code=nvjb5t&utm_campaign=14495
38+-+Halal+Food+Market+in+U.S.+2020-
2024%3a+Market+Size%2c+Forecasts%2c+Trends%2c+Drivers%2c+and+Challenges&utm_ex
ec=chdo54prd.
Halal Monitoring Services. (2022). About HMS. Halal Monitoring Services – About Us. Retrieved
August 8, 2022, from https://hmsusa.org/about.html
Hassan, F., & Hanif, A. (2017). Halal issues in processed food: Misuse of the Halal logo. Journal of
Emerging Economies and Islamic Research, 5(3), 1. https://doi.org/10.24191/jeeir.v5i3.8826
Horne, J. (2012). How anchored tradeoffs reveal customer preferences. Market Strategies
International.
Islamic Services of America. (2022). Halal Meat & Halal Food Certification: ISA. Islamic Services
of America. Retrieved August 8, 2022, from https://www.isahalal.com/about-isa/what-I-do.
Johnston, R. J., Roheim, C. A., Donath, H., Asche, F., Johnston, R. J., Roheim, C. A., Donath, H., &
Asche, F. (2001). MEASURING CONSUMER PREFERENCES FOR ECOLABELED
SEAFOOD: AN INTERNATIONAL COMPARISON. https://doi.org/10.22004/AG.ECON.31157
Lin, W., Ortega, D. L., Ufer, D., Caputo, V., & Awokuse, T. (2020). Blockchain‐based traceability
and demand for U.S. beef in China. Applied Economic Perspectives and Policy, aepp.13135.
https://doi.org/10.1002/aepp.13135
Lohr, L. (1998). Implications of Organic Certification for Market Structure and Trade. American
Journal of Agricultural Economics, 80(5), 1125–1129. https://doi.org/10.2307/1244216
110
Louviere, J., Lings, I., Islam, T., Gudergan, S., & Flynn, T. (2013). An introduction to the application
of (case 1) best–worst scaling in marketing research. International Journal of Research in
Marketing, 30(3), 292–303. https://doi.org/10.1016/j.ijresmar.2012.10.002
Lusk, J. L., & Briggeman, B. C. (2009). Food Values. American Journal of Agricultural Economics,
91(1), 184–196. https://doi.org/10.1111/j.1467-8276.2008.01175.x
Marley, A. A. J., & Louviere, J. J. (2005). Some probabilistic models of best, worst, and best–worst
choices. Journal of Mathematical Psychology, 49(6), 464–480.
https://doi.org/10.1016/j.jmp.2005.05.003
McFadden, D. (1974). Conditional logit analysis of qualitative choice behavior. Zarembka (Ed.),
Frontiers of Econometrics, Academic, New York.
McKendree, M.G.S., Widmar, N. O., Ortega, D. L., & Foster, K. A. (2013). “Consumer
preferences for verified pork-rearing practices in the production of ham products.” Journal of
Agricultural and Resource Economics, 38(3), 397-417.
McKendree, M. G. S., Tonsor, G. T., & Wolf, C. A. (2018). Animal Welfare Perceptions of the U.S.
Public and Cow-Calf Producers. Journal of Agricultural and Applied Economics, 50(4), 544–
578. https://doi.org/10.1017/aae.2018.14
Mohamed, Besheer. (2021, September 1). Muslims are a growing presence in U.S., but still face
negative views from the public. Pew Research Center. https://www.pewresearch.org/short-
reads/2021/09/01/muslims-are-a-growing-presence-in-u-s-but-still-face-negative-views-from-
the-public/
Mohamed, Besheer. (2018, January 3). New estimates show U.S. Muslim population continues to
grow. Pew Research. https://www.pewresearch.org/fact-tank/2018/01/03/new-estimates-show-u-
s-muslim-population-continues-to-grow/.
Ortega, David L., and Christopher A. Wolf. “Demand for Farm Animal Welfare and Producer
Implications: Results from a Field Experiment in Michigan.” Food Policy, vol. 74, Jan. 2018, pp.
74–81. DOI.org (Crossref), doi: 10.1016/j.foodpol.2017.11.006
Ortega, D. L., Wang, H. H., Wu, L., & Olynk, N. J. (2011). Modeling heterogeneity in consumer
preferences for select food safety attributes in China. Food Policy, 36(2), 318–324.
https://doi.org/10.1016/j.foodpol.2010.11.030
Ortega, D. L., Ward, P. S., & Caputo, V. (2019). Evaluating producer preferences and information
processing strategies for drought risk management tools in Bangladesh. World Development
Perspectives, 15, 100132. https://doi.org/10.1016/j.wdp.2019.100132
Pearson, D., & Henryks, J. (2008). Marketing Organic Products: Exploring Some of the Pervasive
Issues. Journal of Food Products Marketing, 14(4), 95–108.
https://doi.org/10.1080/10454440801986421
111
Poe, G. L., Giraud, K. L., & Loomis, J. B. (2005). Computational Methods for Measuring the
Difference of Empirical Distributions. American Journal of Agricultural Economics, 87(2), 353–
365. http://www.jstor.org/stable/3697850
Rahman, A.A, Md. Ismail, C.T. & Abdullah, N.A. (2018). Regulating Halal Food Consumption:
Malaysian Scenario. International Journal of Law, Government and Communication, 3 (13),
313-321.
Sønderskov, K. M., & Daugbjerg, C. (2011). The state and consumer confidence in eco-labeling:
Organic labeling in Denmark, Sweden, The United Kingdom, and The United States. Agriculture
and Human Values, 28(4), 507–517. https://doi.org/10.1007/s10460-010-9295-5.
Stata. (2013, February 11). Statistical Software for data science | Stata. Stata.com. Retrieved August
8, 2022, from https://www.stata.com/manuals13/mvfactor.pdf.
Statista Research Department. (2016, Oct. 3). Major growth opportunities of halal food market
worldwide as of 2016. Statista. https://www.statista.com/statistics/784921/halal-food-global-
market-opportunities/.
The United States Government. (2022, January 3). Fact Sheet: The Biden-Harris Action Plan for a
Fairer, More Competitive, and More Resilient Meat and Poultry Supply Chain. The White
House. https://www.whitehouse.gov/briefing-room/statements-releases/2022/01/03/fact-sheet-
the-biden-harris-action-plan-for-a-fairer-more-competitive-and-more-resilient-meat-and-poultry-
supply-chain/
Tian, Feng. (2016). An agri-food supply chain traceability system for China based on RFID &
blockchain technology. 2016 13th International Conference on Service Systems and Service
Management (ICSSSM), 1–6. https://doi.org/10.1109/ICSSSM.2016.7538424
Ubilava, D., & Foster, K. (2009). Quality certification vs. product traceability: Consumer preferences
for informational attributes of pork in Georgia. Food Policy, 34(3), 305–310.
https://doi.org/10.1016/j.foodpol.2009.02.002
United States Department of Agriculture. (2023). USDA Announces Funding Availability to Expand
Meat and Poultry Processing Options for Underserved Producers and tribal Communities.
United States Department of Agriculture. https://www.usda.gov/media/press-
releases/2023/04/19/usda-announces-funding-availability-expand-meat-and-
poultry#:~:text=These%20new%20grant%20programs%2C%20the,midsized%20meat%20and%
20poultry%20processors
United States Department of Agriculture, Agricultural Marketing Service. (2024). Meat and Poultry
Inspection Readiness Grant. Meat and Poultry Inspection Readiness Grant | Agricultural
Marketing Service. https://www.ams.usda.gov/services/grants/mpirg
112
Ugochukwu, A. I., & Phillips, P. W. B. (2018). Technology Adoption by Agricultural Producers: A
Review of the Literature. In N. Kalaitzandonakes, E. G. Carayannis, E. Grigoroudis, & S.
Rozakis (Eds.), From Agriscience to Agribusiness (pp. 361–377). Springer International
Publishing. https://doi.org/10.1007/978-3-319-67958-7_17
Verify Halal. (2022). Verify halal. Verify Halal. Retrieved August 6, 2022, from
https://www.verifyhalal.com/
Washington Center for Equitable Growth. (2018, May 23). Market Power in the U.S. Economy Today.
Washington Center for Equitable Growth. https://equitablegrowth.org/market-power-in-the-u-s-
economy-today/
Winship, C., & Mare, R. D. (1984). Regression Models with Ordinal Variables. American
Sociological Review, 49(4), 512. https://doi.org/10.2307/2095465
Wolf, C. A., & Tonsor, G. T. (2013). Dairy Farmer Policy Preferences.
https://doi.org/10.22004/AG.ECON.158286
Zabihah. (2021). Halal certifying authorities. https://www.zabihah.com/aut.
113
CHAPTER 3. RESOLVING THE REALITY GAP IN FARM
REGULATION VOTING MODELS
A version of this Chapter was previously published in Food Policy and is reproduced with the
permission of the coauthors. DOI: https://doi.org/10.1016/j.foodpol.2022.102357
Hopkins, K.A., McKendree, M.G.S., & Schaefer, K. Aleks. (2022). Resolving the reality gap in
farm regulation voting models. Food Policy, 112, 102357.
© 2022 Elsevier Ltd. All rights reserved.
1. Introduction
Throughout the world, many national, provincial, state, and local jurisdictions engage in
legislative efforts to regulate food production systems beyond federal or overarching requirements.
This behavior has been seen recently with members of the European Union (EU), including
Germany, Italy, Spain, France, and the United Kingdom (UK)13, who have passed additional
agricultural production regulations beyond the EU's standards (Vogeler, 2019a, 2019b). Similarly,
over the past two decades in the United States (US), some states have made concerted efforts to
wrest from Congress the regulatory control of food production systems within their borders (Neill
et al., 2020). These state regulatory efforts tend to focus on socially controversial agricultural
practices, such as genetically modified varieties or use of production enclosures or farming
practices deemed not to promote farm animal welfare (FAW).
The resulting laws can have substantial negative economic effects for agricultural
producers, including unfunded mandates and vote-buy gaps (Sumner et al., 2008). Unfunded
mandates in agriculture arise when a law passes that requires changes to agricultural production
13 These regulations were passed prior to the UK's withdrawal from the EU in 2020.
114
practices but provides no monetary assistance to the producer to implement these changes (Paul et
al., 2019). A vote-buy gap occurs when citizens vote or express support for a law to regulate
products, but then demonstrate little demand for these specialized products (Norwood et al., 2019).
Further, these regulations effect producers and consumers both in-state (Sumner et al., 2008;
Malone & Lusk, 2016; Mullally & Lusk, 2018; Ortega & Wolf, 2018) and out-of-state (Carter et
al., 2021; Carter & Schaefer, 2019; Sumner, 2017).
The rise of piecemeal state legislation surrounding labeling of genetically modified foods
led Congress to implement the National Bioengineered Food Disclosure Standard (NBFDS),
which preempts state efforts to regulate the labeling of genetically modified foods (Bovay &
Alston, 2018). However, states have almost-unfettered power to develop new mandates or
restrictions in the area of FAW. As shown in Figure 9, this authority has culminated in 19 state-
level bills and ballot initiatives concerning FAW across 13 states through 2020 (Animal
Agricultural Alliance, 2021). In light of the substantial negative economic effects of FAW
regulations for agricultural producers and other stakeholders, a natural question emerges – can one
empirically assess how and why these measures occur in US states' regulatory landscapes?
Figure 9: Timeline of All Enacted FAW Regulations 2000-2021
Note: Bills are above the horizontal axis, have square markers, and are blue in color; ballot initiatives are below the
axis, have round markers, and are orange in color. The bills in Kentucky and Ohio were administrative regulations or
revised statutes. Regulations occurring after 2020 were not included in this article’s analysis due to limited data
availability.
115
We seek to model the evolution of the state-level FAW regulatory landscape as a function
of legislature characteristics and constituent demographics. More specifically, we utilize a two-
stage model to assess (i) whether and when a given state considers FAW measures, and (ii) if so,
the likelihood the measures are passed. Using this model, we estimate the likelihood of FAW
adoption outcomes for all 50 states. Using these predictions, we then estimate the cost to the egg
and pork industries to upgrade to cage- and crate-free production methods in the states most likely
to pass a FAW regulation in the future. We believe this exercise will assist producers and industry
stakeholders in gauging the future of the regulatory landscape and provide guidance on whether to
upgrade existing enclosures to comply with mandates on the horizon or to continue operating with
“conventional” enclosures.
Of course, we are not the first to attempt to understand the uptake of state-level farm
regulation. Videras (2006) first analyzed whether religious demographic variables could be used
to predict voting outcomes in the context of the 2002 Florida Animal Cruelty Amendment. Results
showed that Catholicism and Evangelism had strong, conflicting effects on support for the FAW
ballot initiative. Smithson et al. (2014) expanded on (Videras, 2006) by analyzing the demographic
drivers of voting under the 2008 California Proposition 2 ballot initiative. The authors further
created predictions for all 50 states to determine which states and animal agriculture industries
have a high probability of future FAW regulations. Bovay and Alston (2016) develop a similar
approach to model the probability of genetically modified organism (GMO) labeling restrictions
across all 50 states based on California Proposition 37. Similarly, Bovay and Sumner (2019) used
voting results from both California Proposition 2 and the Prevention of Farm Animal Cruelty Act
in Massachusetts in 2016 to draw connections between political party affiliation and support for
FAW initiatives.
116
However, one persistent puzzle in this line of research is that studies tend to over-predict
state-level FAW regulation. Smithson et al. (2014), for example, predicted between 46% and 100%
of all 50 US states would pass a FAW law through a ballot initiative, if such initiatives were
considered. Similarly, Bovay and Sumner (2019) predicted that nearly all 50 states would have
passed FAW regulations in the 2008 and 2016 election years. If FAW measures are so universally
popular, why has this widespread adoption not already occurred?
We posit that this gap between the literature and reality is a function of inherent selection
in whether and when states consider this type of regulation. In contrast to previous papers,14 our
model utilizes a two-stage, three-part process to account for potential selection bias between bills
and ballot initiatives by state legislators. A large body of research in political science has
established how the use of ballots versus bills functions in state legislature behavior (Matsusaka,
1992; Boehmke & Patty, 2007). Consider, for example, the different costs to legislators and citizen
groups of voting on a proposed law. It is much cheaper for an interest group to lobby a handful of
legislators than the population of a state, and so this is usually the first step (Matsusaka, 1992). For
the legislators, costs are more complex: they must weigh their personal views, party views, and
the views of their constituents before deciding to vote. Our specification allows us to consider the
effect that legislative decisions have on the success of a proposed bill or ballot initiative, especially
when using these models to forecast voting outcomes in other states.
2. Background
In the US, individual states can enact their own FAW laws and regulations, so long as these
laws do not contradict laws passed by the federal government. This type of legislative process is
similar to that exercised by members of the EU, in which some countries have chosen to enact
14 Previous papers include an earlier version of this work, Hopkins et al. (2020).
117
their own laws to regulate agricultural production that go beyond what is required by the EU. In
the US, FAW legislation is passed at the state level through two major avenues: through a bill
voted on in the legislature or through a ballot initiative voted on directly by citizens.
Since 2007, 11 FAW laws have been enacted through the legislative process. All but one
US state has a bicameral legislative body, which is a two-body legislature made up of the State
Senate and State House of Representatives.15 The process to pass a legislative bill in the US
involves several stages. Typically, one or more representatives drafts a bill to present to the
legislative body that they reside in.16 The bill will then be considered by smaller, more focused
committees within the respective body, and if it passes in committee(s), it can be voted on by the
entire body.17 Once a bill passes by majority vote in either the House or Senate, it is sent to the
other legislative body and goes through the entire process again. If it passes a vote in the second
legislative body, it is then referred to the Governor of the state, who can either sign it into law or
veto it.18
Ballot initiatives are an option to create laws in 24 of the 50 US states.19 A ballot initiative
is typically proposed by citizens of the state, an interest group, or some other non-governmental
organization (NGO). Ballot initiatives in the US fall into two categories: direct and indirect ballot
initiatives. Direct ballots bypass the legislature at every step of the ballot initiative process; that is,
they do no not require approval or action by legislatures to be placed on a ballot, so long as a state's
15 Nebraska is the only state with a unicameral legislature, meaning it only has one legislative body.
16 Representatives may draft legislation to support their own beliefs, the beliefs of their constituents, or in response
to lobbying from interest groups.
17 It is common that bills do not make it through committee or do not receive a vote on the floor of the Senate or
House once through a committee.
18 If there is two-thirds support of the vetoed law in both legislative bodies, vetoed laws can be passed into law
without the Governor's signature, though this is rare.
19 Each state has its own requirements and processes that must be followed before the initiative can be considered,
and these requirements vary greatly across states; therefore, our empirical analysis cannot differentiate between
direct versus indirect ballot initiatives.
118
signature and legal filing requirements have been met. On the other hand, indirect ballot initiatives
must be approved by the state legislature before they can appear on a ballot. In either case, typically
a subset of the state legislature, such as a Budget Committee, is involved in the ballot initiative
process to the extent that they conduct a financial and legal analysis of the proposed law to present
to the petitioners. Some states also require the legislature to hold formal hearings or an open forum
about the initiative proposed. At any point in the direct or indirect ballot initiative process, the
legislature can decide to pass a bill to enact the regulations proposed in the ballot initiative. Once
an initiative is on a ballot, it is voted on by citizens and will become law if it passes with a majority
of the vote.20,21 Six FAW laws have been passed through ballot initiatives since 2002.
A list of the FAW bills and ballot initiatives analyzed in this article is given in Table 31.
Of the laws enacted, 10 of them involve confinement standards for egg-laying hens or the sale of
eggs from hens raised in battery cages, 11 involve confinement standards for gestating sows, and
11 involve confinement of veal calves. There were an additional two bills regulating the
confinement of veal calves and gestating sows that were vetoed by the states' governors (MI 2019
and NJ 2013).
20 Some states have specific requirements for majority vote, such as a 60% super-majority or a majority of all ballots
cast, even if a person declined to vote on the initiative.
21 Once passed by majority vote, the outcome may need to be confirmed by the legislature before being passed into
law and may still require the Governor's signature. However, this additional legislative confirmation process is
normally more of a formality; once the citizens have expressed majority support for a ballot initiative, the law will
be enacted.
119
Table 31: Farm Animal Welfare Legislative Bills and Ballot Initiatives Analyzed
Type
Ballot
Title of Legislation
Amendment 10 – HSUS Ballot Initiative:
Gestating Sows
Proposition 204
Or. Rev. Stat. §600.150
SB 201
Proposition 2
Mich. Comp. Laws. Ann. §287.746
Ballot
Bill
Bill
BallotD
Bill
BallotD Amendment 2 – Livestock Care Standards
Year State
FL
2002
Industry Affected
Pork
Pork, Veal
Pork
Pork, Veal
2006 AZ
2007 OR
2008 CO
2008 CA Eggs, Pork, Veal
2009 MI
Eggs, Pork, Veal
2009 OH Veal
Bill
Bill
Bill
Bill
Bill*
Ballot
Bill*
Bill
BallotD
Bill
Bill
Bill
Amendment
2010 CA Eggs
AB 1437
SB 805 – Relating to egg-laying hens
2011 OR Eggs
Wash. Rev. Code §69.25.065 and §69.25.107 2011 WA Eggs
SB 2191
SB 1921
Prevention of Farm Animal Cruelty Act
SB 660
HB 7456 – Unlawful Confinement of a
Covered Animal
Proposition 12
SB 174
S.B. 1019
HB 2049 – Concerning commercial egg layer
operations
Pork, Veal
Pork
RI
2012
2013 NJ
2016 MA Eggs, Pork, Veal
2018 MI
RI
2018
Eggs, Pork
Eggs, Pork, Veal
2018 CA Eggs, Pork, Veal
2019 MI
Eggs
2019 OR Eggs
2019 WA Eggs
* Represents a bill that passed but was vetoed by the state’s governor. All other bills or ballots have been enacted. D
Indicates a direct ballot initiative. All other ballot initiatives are indirect.
Relationships between demographic characteristics and support for FAW are well
documented in agricultural economics and political science literature and are therefore important
to consider when studying FAW regulation adoption. Smithson et al. (2014) found that an increase
in median household income and an increase in poverty rate both correlated with a decrease in
support for FAW regulations. Educational achievement has been shown to correlate with lower
support for regulations to increase FAW, as more educated individuals are more likely to view
animal and human similarities and differences more scientifically and this may change their views
on FAW (Jerolmack, 2003).
Previous research suggests that religion plays a large role in an individual's view of the
natural world and thus impacts views on animals and animal welfare (Videras, 2006). For instance,
120
Catholics tend to be more supportive of animal welfare issues than Protestants and Evangelicals
(Smithson et al., 2014; Oldmixon, 2017). Overall, non-religious, non-Christians, and Catholics are
most in favor of FAW over Christians (Cornish et al., 2016; Jerolmack, 2003; Flynn, 2001).
Jerolmack (2003) found that Jewish and other religions were more likely to support animal rights.
In Islam, concern for animal welfare and animal rights are key moral and religious values
(Gharebaghi et al., 2007). Several studies have also shown that non-white Americans tend to view
regulation to increase FAW more positively (Jerolmack, 2003; Franklin et al., 2001; Nibert, 1994;
Peek et al., 1996; Uyeki & Holland, 2000). Over the past two decades, FAW has become an
increasingly politicized issue in the US (Feindt et al., 2020; Lai et al., 2021; Vogeler, 2020).
Membership in the Democratic party has been linked to higher concern for animal welfare in a
wide variety of past studies and contexts, including McKendree et al. (2014), Deemer and Lobao
(2011), Czech and Borkhataria (2001), Miele et al. (1993), and Heleski et al. (2006). Furthermore,
from past studies using voter data, we know that liberals are more supportive of animal welfare
measures, in general, than conservatives (Smithson et al., 2014).
3. Methodology
We utilize a two-stage, three-part multinomial endogenous switching regression (MESR)
to model the implementation of FAW regulations and account for decision selection bias.22,23 In
22 Selection bias is a common challenge in studies using nonrandomized data to model decisions and outcomes.
Methodologically, most studies in this area have used propensity score matching (PSM); however, the PSM
approach does not correct selection bias from unobserved factors (Abdulai, 2016; Jaleta et al., 2016). Unlike PSM,
MESR models employ a selection correction method by calculating an Inverse Mills Ratio (IMR) using the theory of
truncated normal distribution to correct selection bias (Bourguignon et al., 2007). The IMR is the ratio of the
probability density function to the complementary cumulative distribution function of a distribution. This technique
is commonly used in development economics to account for unobserved heterogeneity and selection bias in farmers'
cropping decisions (Kassie et al., 2015; Di Falco, 2014).
23 One method for adjusting for this bias may be Heckman's (Heckman, 1976) two-stage model; however, we must
consider the different levels of data aggregation, inclusion of population weights, use of both panel and cross-
sectional data sets, and likelihood of unobserved heterogeneity in our analysis. Thus, the Heckman approach would
give inconsistent estimates if selection bias originating from observed and unobserved heterogeneity is not
addressed.
121
the MESR framework, the two stages are modeled simultaneously. A simple schematic of our
MESR model is depicted in Figure 10. We hypothesize the likely source of selection bias occurs
at the state legislative level – legislatures may endogenously self-select different FAW legislative
actions, and decisions are likely to be influenced by unobserved factors that may be correlated
with outcome variables. Additionally, accounting for institutional characteristics is crucial to
understanding the relationship between policy changes and politics, particularly in FAW where
passing regulations using democratic instruments such as ballot initiatives is common (Vogeler,
2020). The MESR allows us to take into account the characteristics of individual states'
governments and their impact on the likelihood of a FAW bill or ballot initiative occurring and
passing. These relationships are modeled in Stage 1 (Legislature Action Decisions) of our model.
We then account for decision selection bias in our second stage FAW regulation voting outcome
models – Stage 2.1 (Ballot Initiative Voting Outcomes) and Stage 2.2 (Legislative Bill Voting
Outcomes) by including the first-stage Inverse Mills Ratio (IMR) as an explanatory variable and
bootstrapping to compute coefficient and standard error estimates.
3.1. Econometric Model
We discuss each of the components in turn.
122
Figure 10: Farm Animal Welfare Voting Two-Stage Selection Model Schematic
3.1.1 Stage 1: Legislative Action Decisions
Here, we model state legislatures' choice of alternative legislative actions for addressing
FAW concerns – no action, bill proposed, or ballot initiative allowed – using a multinomial logit
selection (MNLS) model accounting for unobserved heterogeneity (Equation 18). Stage 1 is
aggregated at the state level and consists of 20 years of annual panel data for 49 of the 50 states (n
= 980).24 We cluster the standard errors by state. The dependent variable can take three values:
y = 0 if no action was taken, 𝑦 = 1 if a ballot initiative was placed on a ballot, or 𝑦 = 2 if a bill
was proposed to the state legislature.
We estimate the following first-stage model:
𝑌$ = {0,1,2} = (𝑋$
( × β) + (𝐷$
( × δ) + (𝐶$
( × κ) + ε$
where 𝑌$ is the predicted action outcome of a state legislature, 𝑋$ is the matrix of state legislature
political variables – dummy variables to indicate whether the state house, senate and governor are
all of the same political party (denoted TRIFECTA_D and TRIFECTA_R), and continuous
Equation 18
24 Nebraska is excluded from this data set due to its unique unicameral state government system and because the
state's legislators are not required to affiliate with a political party.
123
variables indicating the percent of house and senate seats occupied by Democrats (denoted
HOUSE%D and SENATE%D) – with β the corresponding coefficients. 𝐷$ is the matrix of
legislative characteristic variables – counts of previous animal welfare legislation (denoted
COUNT_PASSED_PREV), a dummy variable indicating whether the state had previously passed
FAW regulation (denoted PREV_LAW),25 and a dummy variable indicating whether the state
allows ballot initiatives (denoted ALLOW_BALLOT) – with δ the corresponding coefficients. 𝐶$ is
the matrix of state agricultural industry density variables (denoted HENS_PER_1000 and
HOGS_PER_1000) which are the number of egg-laying hens and gestating sows per 1,000 people
in each state, respectively – with κ the corresponding coefficients. ε$ is the error term.
The IMRs for each of the second stage regressions are then calculated (Equation 19 and
Equation 20) from their respective estimated outcome probabilities in the MNLS model:
𝐼𝑀𝑅$ =
𝑓(𝑌$)
𝐹K(𝑌$)
𝐹K(𝑌$) = 𝑃𝑟(𝑌$ > 𝑦$) = k 𝑓(𝑠)𝑑𝑠
6
7(
Equation 19
Equation 20
where 𝑖 = 0,1,2 for the no action, ballot, and bill outcomes, respectively, 𝑓(𝑌$) is the standard
normal probability density function (PDF), 𝐹K(𝑌$) is the standard normal cumulative distribution
function (CDF), and 𝑠 is the integration argument.
25 COUNT_PASSED_PREV and PREV_LAW account for the "snowball effect" of previous laws. A "snowball
effect" is a situation in which one event influences the likelihood that a similar event occurs (Matsusaka, 2005).
These two variables also proxy for the diffusion of media coverage related to FAW within and between states,
increased public awareness of FAW, and other time effects.
124
3.1.2. Stage 2: Ballot and Bill Voting Outcomes
In the second stage, we evaluate the voting outcomes for a ballot initiative (Stage 2.1) or a
legislative bill (Stage 2.2) using ordinary least squares (OLS) with IMRs as additional covariates
to account for selection bias from time-varying unobserved heterogeneity.
Stage 2.1 Ballot Initiative Voting Outcomes: In Stage 2.1, we model county-level voting
outcomes for six ballot initiatives in Arizona (AZ; 2006), California (CA; 2008 & 2018), Florida
(FL; 2002), Massachusetts (MA; 2016), and Ohio (OH; 2009) (n = 299). To model these voting
outcomes, we estimate a log-linear model:
𝑙𝑛 m
𝑉0
1 − 𝑉0
o = (𝐼𝑀𝑅0
( × ϕ) + (𝐵0
( × ζ) + 𝑢0
Equation 21
where the dependent variable is the log-odds of the predicted “yes” portion of the ballot initiative
vote 𝑉0. 𝐼𝑀𝑅0 is the IMR value from the Stage 1 ballot outcome with ϕ the corresponding
coefficient.26 𝐵0
is
the matrix of county demographics
including people per farm
(PEOPLE_PER_FARM),27 percent of voters in the county who voted for the Democratic
presidential candidate in the year closest to when the initiative was on the ballot (%DEMOCRAT),
median household income in thousands of dollars (HOUSEHOLD_INCOME_1000), the percent
of people in poverty (POVERTY_RATE), the percent of persons of 25+ years of age with a
bachelor's degree (EDUCATION), the percent of white (%WHITE), Black (%BLACK), and
Hispanic
citizens
(%HISPANIC),
and
the
percent
of Mainline
Protestants
(%MAINLINE_PROTESTANT), Evangelical Protestants (%EVANGELICAL_PROTESTANT),
26 A table showing the distribution of the Stage 1 ballot IMRs is given in Appendix A.4.
27 This variable (PEOPLE_PER_FARM) is a proxy for citizens' familiarity with agriculture.
125
and Catholics (%CATHOLIC) – with ζ the corresponding coefficients.28 There is little accurate
data available for Jewish and Muslim populations at the county level, and so we unfortunately do
not include these religions in our model.29 𝑢0 is the error term. Standard errors in the model are
clustered by state, as we are aggregating data from five different states.
Stage 2.2 Legislative Bill Voting Outcomes: In Stage 2.2, we model 13 individual bill voting
outcomes in California (CA; 2010), Colorado (CO; 2008), Michigan (MI; 2009, 2018 & 2019),
New Jersey (NJ; 2013), Oregon (OR; 2007, 2011 & 2019), Rhode Island (RI; 2012 & 2018), and
Washington (WA; 2011 & 2019) (n = 1,583). To estimate bill-voting outcomes, we use a linear
probability model (LPM) with continuous state legislative-district level demographic data to
predict the vote of individual legislators.30 The dependent variable can take two values: 𝑍 = 0 if
the legislator voted “no”, was absent, or declined to vote on an FAW bill, or 𝑍 = 1 if the legislator
voted “yes”. We estimate the following model:
𝑍) = {0,1} = >𝐼𝑀𝑅)
( × τ? + >𝐿)
( × γ? + 𝑢)
where 𝑍) is the actual vote of a given legislator. 𝐼𝑀𝑅) is the vector of IMR values from the Stage
1 bill outcome with τ the corresponding coefficient. 𝐿) is the matrix of state legislative district
Equation 22
28 These variables were selected to be consistent with previous literature on the relationships between demographic
characteristics and FAW ballot initiative outcomes (Videras, 2006; Smithson et al., 2014; Bovay & Sumner, 2019),
as well as documented correlations between demographics and support for FAW in general (McKendree et al., 2014;
Jerolmack, 2003; Deemer & Lobao, 2011; Czech and Borkhataria, 2001; Miele et al., 1993; Heleski et al., 2006;
Oldmixon, 2017).
29 We collected county-level data from ARDA the Association of Religious Data Archives
(http://www.thearda.com/QL2010/) for the percentage of Jewish and Muslim citizens for the bill and ballot states
involved in our Stage 2 models. This data was very sparse, as these populations are small relative to the overall
population of the US. Roughly 3/4 of the counties collected did not report a percentage estimate for Muslim citizens
and about 2/3 did not report a percentage for Jewish citizens. We attempted to interpolate the data using fractional
probit and OLS models; however, the results were poor. For example, the model predicted negative Jewish and
Muslim population estimates for some counties, which is nonsensical.
30 In order to use the MESR model with IMRs, the second-stage models must be linear, that is, a probit or logit
function is not compatible. Further, an LPM is as appropriate as a logit or probit for this data so long as our standard
errors are robust (Bellemare, 2015).
126
demographic variables. These variables are equivalent to those in Stage 2.1, except at the
legislative district, rather than county, level.31 Additionally, we include a dummy variable to
indicate the political party of the legislator (denoted DEMOCRAT). Corresponding coefficients are
represented by γ. 𝑢) is the error term.
Stage 2 Bootstrapping and Standard-Error Clustering: To help with bias and inconsistency
and improve inference in the IMR-adjusted second stage models, we bootstrap coefficients and
standard errors using a stationary cluster block bootstrap method. This allows us to estimate
coefficients simultaneously across the two stages of our model (Politis & White, 2004; Politis &
Romano, 1994; Hall et al., 1995). Our block bootstrapping procedure randomly selects 45 states
to use in our model and tests out-of-sample prediction accuracy using the five states randomly
excluded in each repetition.32
3.2 Data by Stage
In this section, we describe our data selection and collection process, aggregation levels,
and summary statistics by stage. More explicit information on how data were accessed and data
sources are given in Appendix A.5.
The variable names, descriptions, and summary statistics for Stage 1 (Legislative Action
Decisions) are included in Table 32. The majority of our data were collected from the respective
31 As above, these include people per farm (PEOPLE_PER_FARM), median household income
(HOUSEHOLD_INCOME_1000), the percent of people in poverty (POVERTY_RATE), the percent of persons 25+
years of age with a bachelor's degree (EDUCATION), the percent of white (%WHITE), Black (%BLACK), and
Hispanic citizens (%HISPANIC), and the percent of Mainline Protestants (%MAINLINE_PROTESTANT),
Evangelical Protestants (%EVANGELICAL_PROTESTANT), and Catholics (%CATHOLIC). We selected the
variables based on documented correlations between demographics and support for FAW (McKendree et al., 2014;
Jerolmack, 2003; Deemer & Lobao, 2011; Czech & Borkhataria, 2001; Miele et al., 1993; Heleski et al., 2006;
Oldmixon, 2017) and for comparison to the ballot model.
32 We chose to bootstrap using blocks rather than only clustering by state, as only clustering standard errors by state
in the Stage 1 MNLS model would likely lead to correlation between observations from the same state. We
conducted 1,000 repetitions with 35, 40, and 45 states per block in Stage 1. Models were robust to block size and
thus we chose the 45 block model.
127
state government websites and Ballotpedia.org. Agricultural industry data was collected from
USDA QuickStats and the Livestock Marketing Information Center (LMIC).
Variable names, descriptions, and summary statistics for Stage 2.1 (Ballot Initiative Voting
Outcomes) are given in Table 33. Counties with less than 2,000 people were dropped, then each
county within a state was given a weight corresponding to the fraction of its population relative to
the overall state population, minus any dropped counties.33
Variable names, descriptions, and summary statistics for Stage 2.2 (Legislative Bill Voting
Outcomes) are given in Table 34. There is no need to weight this data, as legislative districts are
drawn to be proportional based on the most recent census data.
The majority of the demographic data for Stage 2.1 and 2.2 were downloaded from the
United States Census website (census.gov). Data were used from the census year closest to the
year the ballot or bill was considered. Stage 2.1 data on vote outcomes were collected from each
state's records, available online. Stage 2.2 data on legislators' individual votes was collected from
each state legislature's records, available online. Farm data was collected from USDA QuickStats.
All religion data were retrieved from the Association of Religious Data Archives.34
33 Weighting is important when evaluating demographic makeup, as weighting the data by county population
ensures more realistic and accurate predictions. In the U.S., state-level elections are determined by a majority vote;
therefore, relative populations should be considered when collecting the data. Our weighting procedure is similar to
that used by Smithson et al. (2014). This weighting method is intuitive: if County A has two times the population of
County B, then the demographic percentages and voting outcomes from County A will be given two times the
importance of those from County B in the state-wide calculations.
34 As religious data is not reported for all states by legislative district, religious affiliation data for legislative districts
was obtained by aggregating and averaging the data from counties in the district.
128
Table 32: Variable Names and Descriptions for Stage 1 – Legislative Action Decisions
Variable Name
Description
𝑌$, 𝑖 = 0,1,2
HENS_PER_1000
HOGS_PER_1000
TRIFECTA_D
TRIFECTA_R
HOUSE%D
SENATE%D
COUNT_PASSED_PREV
PREV_LAW
ALLOW_BALLOT
Categorical variable = 0 if the state legislature took no action,
= 1 if the state legislature allowed a ballot initiative, and = 2 if
the state legislature proposed a legislative bill
Continuous variable equal to the number of egg-laying hens
per 1,000 people in a state in a given year
Continuous variable equal to the number of gestating sows per
1,000 people in a state in a given year
Dummy variable = 1 if all three bodies of state legislative
branch are controlled by Democrats, = 0 otherwise
Dummy variable = 1 if all three bodies of state legislative
branch are controlled by Republicans, = 0 otherwise
Continuous variable equal to the percent of the state’s House of
Representatives that belong to the Democratic Party
Continuous variable equal to the percent of the state’s Senate
that belong to the Democratic Party
Discrete variable equal to the total number of farm animal
welfare regulations in place throughout the US
Dummy variable = 1 if the state has a previous farm animal
welfare law in place, = 0 otherwise
Dummy variable = 1 if the state allows ballot initiatives, = 0
otherwise
Mean
Standard
Deviation
Frequency
(%)
𝑖 = 0, 95.61%
𝑖 = 1, 0.92%
𝑖 = 2, 3.47%
1430.57
2474.54
61.78
123.08
0.21
0.33
0.41
0.47
48.76
16.89
47.04
19.11
8.60
0.12
0.48
7.55
0.32
0.50
129
Table 33: Variable Names and Descriptions for Stage 2.1 – Ballot Voting Outcomes
Variable Name
𝑙𝑛 m
𝑉0
1 − 𝑉0
o
PEOPLE_PER_FARM
HOUSEHOLD_INCOME_1000
POVERTY_RATE
EDUCATION
%WHITE
%BLACK
%HISPANIC
%DEMOCRAT
%CATHOLIC
%EVANGELICAL_PROTESTANT
%MAINLINE_PROTESTANT
Description
Log of the odds of success of the ballot initiative
Continuous variable equal to the number of people per farm in the
county
Continuous variable equal to the median household income in the
county in thousands of dollars
Continuous variable equal to the percent of people in the county who
are below the poverty line
Continuous variable equal to the percent of people in the county 25+
years old with a 4-year college degree
Continuous variable equal to the percent of people in the county who
are white
Continuous variable equal to the percent of people in the county who
are Black
Continuous variable equal to the percent of people in the county who
are Hispanic
Continuous variable equal to the percent of voters in the county who
voted for the Democratic presidential candidate in the presidential
election year closest to when the initiative was on the ballot
Continuous variable equal to the percent of people in the county who
are Catholic
Continuous variable equal to the percent of people in the county who
are Evangelical Protestants
Continuous variable equal to the percent of people in the county who
are Mainline Protestants
Mean
0.31
Standard
Deviation
0.65
1392.42
9326.10
49.56
16.26
5.19
5.23
20.69
9.98
71.99
21.01
5.92
7.44
15.42
17.32
47.42
12.06
17.67
14.04
12.87
8.89
6.52
4.99
130
Table 34: Variable Names and Descriptions for Stage 2.2 – Legislative Bill Voting Outcomes
Variable Name
𝑍)
PEOPLE_PER_FARM
HOUSEHOLD_INCOME_1000
POVERTY_RATE
EDUCATION
%WHITE
%BLACK
%HISPANIC
DEMOCRAT
%CATHOLIC
%EVANGELICAL_PROTESTANT
%MAINLINE_PROTESTANT
Description
Discrete variable = 1 if the legislator voted yes on a farm animal welfare
bill, = 0 otherwise
Continuous variable equal to the number of people per farm in the
legislative district
Continuous variable equal to the median household income in the
legislative district in thousands of dollars
Continuous variable equal to the percent of constituents in the legislative
district who are below the poverty line
Continuous variable equal to the percent of constituents in the legislative
district 25+ years old with a 4-year college degree
Continuous variable equal to the percent of constituents in the legislative
district who are white
Continuous variable equal to the percent of constituents in the legislative
district who are Black
Continuous variable equal to the percent of constituents in the legislative
district who are Hispanic
Dummy variable = 1 if legislator belongs to the Democratic Party
Continuous variable equal to the percent of constituents in the legislative
district who are Catholic
Continuous variable equal to the percent of constituents in the legislative
district who are Evangelical Protestants
Continuous variable equal to the percent of constituents in the legislative
district who are Mainline Protestants
Mean
Standard
Deviation
0.80
0.40
395.71
2358.62
60.57
18.43
13.76
6.92
30.98
14.01
80.81
16.94
9.04
13.90
13.32
14.42
0.59
0.49
21.20
13.47
10.05
4.77
5.75
22.28
131
4. Results
We present the results of each stage of our model in turn.
4.1. Stage 1: Legislature Action Decisions
Results of the multinomial logit model for Stage 1 are in Table 35. with results for the
ballot outcome on the left and bill outcome on the right. The base case is a legislature not taking
any action on FAW regulations; that is, neither a bill nor ballot initiative is proposed. For ease of
interpretation, Stage 1 model coefficients are given in relative risk ratio (RRR) format.35 A value
greater than one means the risk of the outcome falling in the comparison group rather than the base
group increases as the variable increases. A RRR less than one means the risk of the outcome
falling in the comparison group over the base group decreases as the variable increases. For
example, the coefficient of 1.695 on PREV_LAW for the bill outcome means that a bill is 69.5%
more likely to occur than no action, all else equal. Similarly, the coefficient of 0.281 on
TRIFECTA_R in the ballot column means that a ballot is 71.9% [1.00 - 0.281 = 0.719] less likely
to occur than no action, all else equal.
A state's ties to the pork and egg industries resulted in interesting impacts on the likelihood
that a bill or ballot would be proposed over no action taken. The ratio of gestating sows and egg-
laying hens to 1,000 people within the state, HOGS_PER_1000 and HENS_PER_1000, have
notable predicted influences. While the number of egg-laying hens per 1,000 people does not have
a correlated effect on the likelihood a ballot or bill is proposed, the number of gestating sows per
1,000 people does. As the number of gestating sows per 1,000 people increases, we see a
corresponding 5.2% decrease in the likelihood that a ballot is proposed and a 0.8% decrease in the
likelihood that a bill is proposed. This implies that states with higher ratios of gestating sows to
35 A table with non-transformed coefficients is in Appendix A.4.
132
people are less likely to propose FAW regulations. This is sensible. States in which this ratio is
higher likely have larger populations of people familiar with the pork industry, and this familiarity
will influence the peoples' choice to regulate FAW in pork production. Further, state legislators in
states with prominent agricultural industries likely have closer connections with local Farm
Bureaus and other agricultural lobbyists. Introducing legislation that is counter to these groups'
interests can have high political costs from the legislators' perspective.
The history of FAW regulations is also important. As the number of previous FAW laws
in the nation overall rises (COUNT_PASSED_PREV), we see a correlation with an increase in the
likelihood that either a ballot or a bill is proposed in any state, though this correlation is only
significant in the bill case. These results support (Matsusaka, 2005) who asserts that there can be
a bandwagon effect where legislation in one state leads to proposals of similar legislation in other
states. When there is already a FAW law in place in a state (PREV_LAW), a bill is more likely to
be proposed in a given state during a given year than no action taken. This association makes sense,
as citizens of a state that already had a FAW regulation in place have likely been exposed to more
media coverage of FAW issues. On the other hand, in states with a previous law, it is 51.5\% less
likely that a ballot is proposed than no action taken. This makes sense as citizens are unlikely to
feel the need to petition for a ballot initiative for another FAW law if FAW concerns have already
been addressed. The option within a state to take action through a ballot initiative
(ALLOW_BALLOT) correlates with an over 18-fold (19.391) increase in the likelihood that a ballot
will be proposed in a given year than no action taken.
Interestingly, the influence of political party and legislative composition was more
pronounced in the bill model. Our results show that the presence of a Democratic trifecta
(TRIFECTA_D) is associated with a significant increase in the likelihood that a bill is proposed
133
and a decrease in the likelihood that a ballot is proposed. On the other hand, a Republican trifecta
(TRIFECTA_R) correlates with a decrease in the chances of either a FAW bill or ballot, though
the correlation is only significant in the case of a bill. As the percent of Democrats in the House
(HOUSE%D) increases, we see an increase in the likelihood of proposing a bill or a ballot, though
these effects are not statistically significant. These findings are in line with previous studies, which
found that Democrats tend to support FAW more than Republicans (McKendree et al., 2014;
Deemer and Lobao, 2011; Czech and Borkhataria, 2001; Miele et al., 1993; Heleski et al., 2006).
On the other hand, as the percent of Democrats in the Senate increases (SENATE%D), we see a
small associated decrease in the likelihood of a proposed bill or ballot. Finally, the constant terms
in both the bill and ballot outcomes are very close to zero, meaning the probability of no action is
almost 100% more likely than a bill or ballot. Our out-of-sample model predictions (the five states
randomly excluded from the clustering in each of our 1,000 bootstrapping repetitions) are over
92% accurate.36
In Table 36 we present the predicted mean probability of the three outcomes in a given
state for a given year from 2000 to 2019. We note that the mean predicted probability for no action
in a state in each year is about 96%, going as low as about 49% and as high as nearly 100%. These
predictions suggest that the FAW bills and ballots that are currently in place were unlikely to occur.
4.2. Stage 2.1: Ballot Initiative Voting Outcomes
The results of the OLS regression for the ballot initiative model in Stage 2.1 are given in
Table 37 with the ballot model on the left in column (1). The dependent variable in the ballot
36 See Appendix A.4 for descriptive statistics on the accuracy of out-of-sample predictions.
134
regression is the log-odds of success of the initiative. The results of the model are contingent on a
ballot being allowed and put to a vote.37
In our ballot model, all variables are statistically significant. We note first that
PEOPLE_PER_FARM, %DEMOCRAT, %WHITE, %BLACK, and %HISPANIC are all positive,
indicating that an increase in these variables is associated with an increase in the likelihood of a
ballot initiative's success. These effects are in line with those in Smithson et al. (2014). On the
other hand, an increase in the percent of adults with at least a 4-year degree (EDUCATION) and
the percent of citizens living below the poverty line (POVERTY_RATE) both correlate to a
statistically significant decrease on the success of a ballot initiative. The sign on POVERTY_RATE
matches the models presented in Smithson et al. (2014). The negative association of increased
education on support for FAW regulations is supported by Jerolmack (2003), who suggest that
educated individuals are more likely to view animal and human similarities and differences more
scientifically. Additionally, the positive sign on the coefficient for HOUSEHOLD_INCOME_1000
indicates a predicted increase in support for FAW regulations as median household income
increases – this makes sense, as households with higher income levels have more disposable
income to spend on specialty products.
An increase in the percent of Catholic citizens (%CATHOLIC) suggests a slight decrease
in the likelihood that a FAW ballot initiative succeeds, which is in line with previous findings
(Smithson
et
al.,
2014; Videras,
2006; Oldmixon,
2017). The
signs
of
%EVANGELICAL_PROTESTANT and %MAINLINE_PROTESTANT are negative and positive,
respectively. Our results for Evangelical Protestants align with results in previous papers
(Smithson et al., 2014; Videras, 2006; Oldmixon, 2017). While our results for Mainline Protestants
37 K-density plots showing the distribution of the R-squared values, IMRs, and residuals for the ballot model are
given in Appendix A.4. Additionally, descriptive statistics for the IMRs are given by year in Appendix A.4.
135
conflict with (Videras, 2006) and Smithson et al. (2014), we believe the inclusion of more data in
our model, particularly from states with higher percentages of Mainline Protestants, allows for
more robust estimates of the effects on support for FAW regulations of numbers of Mainline
Protestants. The IMR values from Stage 1 (BALLOT_IMR) are statistically significant and positive.
Thus, there exists a selection process that would bias results if not taken into account.38 Finally,
the constant term has a positive and significant value, suggesting that a ballot is likely to pass once
considered.
Table 35: Stage 1 Legislature Action Decisions Output
VARIABLES
HENS_PER_1000
HOGS_PER_1000
COUNT_PASSED_PREV
PREV_LAW
ALLOW_BALLOT
TRIFECTA_D
TRIFECTA_R
HOUSE%D
SENATE%D
CONSTANT
(1)
Ballot Outcome
1.000
(<0.001)
0.948**
(0.024)
1.050
(0.045)
0.485
(0.467)
19.391***
(21.053)
0.537
(0.312)
0.281
(0.328)
1.019
(0.042)
0.995
(0.038)
0.001***
(0.002)
(2)
Bill Outcome
1.000
(<0.001)
0.992**
(0.003)
1.097***
(0.038)
1.695
(0.793)
1.443
(0.617)
2.442*
(1.233)
0.179*
(0.159)
1.045
(0.034)
0.987
(0.026)
0.002***
(0.002)
Observations
Note: Robust standard error form in parentheses, *** p<0.01, ** p<0.05, * p<0.1. Standard errors are given in relative
risk ratio format.
980
38 An unadjusted ballot model, similar in structure to those estimated by Smithson et al. (2014) and Videras (2006) is
included in Appendix A.4.
136
Table 36: Mean Predicted Probability of Stage 1 Legislative Action Decisions Outcomes
Max. Prob.
0.99940
0.51945
0.09639
Predicted Outcome Mean Prob.
No Action
Bill Proposed
Ballot Proposed
Note: 36 FAW bills were proposed between 2000 and 2019, but only 13 made it to a vote in a state legislature; the
“Bill Proposed” outcome does reflect this full consideration rate.
Min. Prob.
0.47826
0.00021
0.00014
Std. Dev.
0.06766
0.06342
0.01403
0.95714
0.03367
0.00918
Table 37: Stage 2.1 Ballot Initiative and Stage 2.2 Legislative Bill Voting Outcomes
Variables
PEOPLE_PER_FARM
%DEMOCRAT
DEMOCRAT
%WHITE
%BLACK
%HISPANIC
EDUCATION
POVERTY_RATE
HOUSEHOLD_INCOME_1000
%CATHOLIC
%EVANGELICAL_PROTESTANT
%MAINLINE_PROTESTANT
BALLOT_IMR
BILL_IMR
CONSTANT
(1)
Ballot Model
<0.001***
(<0.001)
0.014***
(<0.001)
0.008***
(<0.001)
0.001***
(<0.001)
0.008***
(<0.001)
-0.009***
(<0.001)
-0.023***
(<0.001)
0.002***
(<0.001)
-0.003***
(<0.001)
-0.034***
(<0.001)
0.030***
(<0.001)
0.110***
(<0.001)
0.650***
(0.001)
(2)
Bill Model
<0.001***
(<0.001)
0.208***
(0.001)
<0.001***
(<0.001)
0.002***
(<0.001)
0.002***
(<0.001)
<0.001***
(<0.001)
-0.006***
(<0.001)
-0.001***
(<0.001)
<0.001***
(<0.001)
0.003***
(<0.001)
0.013***
(<0.001)
0.009***
(<0.001)
0.678***
(0.004)
Observations
Note: Robust standard errors in parentheses. Coefficients and errors are obtained from cluster bootstrapping. ***
p<0.01, ** p<0.05, * p<0.1. All variables in (1) are weighted by county population as a proportion of the total state
population.
1,583
299
137
4.3. Stage 2.2: Legislative Bill Voting Outcomes
The results of the LPM regression for the legislative bill model in Stage 2.2 are given in
Table 37 on the right in column (2). The dependent variable here is the vote of the legislator, with
a value of 1 equating to a “yes” vote and a value of 0 otherwise. The results of this model are
contingent on a bill being proposed and put to a vote.
An increase in the number of people per farm (PEOPLE_PER_FARM) correlates with a
significant increase in the likelihood of a legislator voting "yes" on a FAW bill, though the
magnitude of this coefficient is small. This implies that legislators from districts with relatively
less representation and familiarity with agriculture are more likely to vote in support of FAW
regulations. Legislators who identify as Democrats are significantly more likely to vote “yes” on
an FAW regulation than to vote “no”, as evidenced by the positive coefficient on DEMOCRAT.
Again, this aligns with previous literature that Democrats are more likely to support FAW than
Republicans (McKendree et al., 2014; Deemer & Lobao, 2011; Czech & Borkhataria, 2001; Miele
et al., 1993; Heleski et al., 2006). Increases in the percent of white (%WHITE), Black (%BLACK),
Hispanic (%HISPANIC), adults with at least a 4-year college degree (EDUCATION), Catholic
(%CATHOLIC), Evangelical Protestant (%EVANGELICAL_PROTESTANT), and Mainline
Protestant (%MAINLINE_PROTESTANT) constituents all correlate with an increased likelihood
that a legislator votes “yes” on a FAW bill. However, all these effects are small. The positive signs
on EDUCATION, %CATHOLIC, and %EVANGELICAL_PROTESTANT are opposite to the signs
in the ballot model. Increases in the percent of constituents living below the poverty line
(POVERTY_RATE) and the median household income (HOUSEHOLD_INCOME_1000) in the
legislative district are associated with a decrease in the likelihood that a legislator votes “yes” on
a FAW bill, but again, these effects are small. The negative correlation between increases in
138
HOUSEHOLD_INCOME_1000 and the likelihood that a legislator supports a FAW bill differs
from the correlation found in the ballot model. The IMR values from Stage 1 are statistically
significant, confirming the need to include the selection process from Stage 1. Additionally, the
constant coefficient, which is statistically significant and positive, suggests that once a bill is put
to a vote, it is more likely to pass than fail.
4.4. Ballot (Stage 2.1) and Bill (Stage 2.2) Model Comparisons
There are general differences in the importance and influence of demographics between
the bill and ballot initiative models. The only demographic variable that has a meaningful
magnitude associated with success of a FAW regulation in the bill model is the legislator's political
party (DEMOCRAT). In the ballot initiative model, the demographic variables are all relatively
similar in their overall magnitude and correlation to voting outcomes. When comparing these two
models, we conclude that legislators tend to vote along party lines when it comes to FAW
regulations. For ballot initiatives, constituents' preferences and demographic characteristics more
directly impact the outcome of FAW regulation, and as such, we see that the effect of demographic
variables are more influential on voting outcomes. Indeed, Tolbert and Smith (2006) find that
policies that result from a popular vote are more likely to be representative of voter preferences
than policies that result from legislative votes.
4.5. Ballot Model Predictions
To understand the implications of using a multi-stage model and accounting for the
decision of whether or not a ballot goes before the people, we generate predictions for all 50 states
for our novel IMR adjusted ballot model using data from 2019 (
139
Figure 11). Accurate predictions are shaded in dark grey in
140
Figure 11. We use the ballot model to predict outcomes, but count either a bill,
administrative action, or a ballot as a pass in “reality”. Since only 24 of the 50 states have a ballot
process, we use the outcomes from our model as a prediction of the opinions of a state's population.
Our predictions for the non-ballot states should be considered as indicators for public opinion
towards FAW, which influence the likelihood of a legislature proposing a bill but do not indicate
that a legislature will necessarily pass a law. These results are also presented in Table 51 in
Appendix A.4. In reality, 24% of all 50 states and 41.7% of the 24 ballot initiative states have
passed FAW regulations. With our ballot model we predict that 26% of all 50 states and 30.4% of
the 24 ballot states would pass FAW regulations. The predicted pass rates given by our novel IMR
adjusted model are close to the actual percentage of states with FAW regulations in place.
Furthermore, our model is 74% and 75% accurate in predicting the presence of a FAW law in all
50 states and the 24 ballot states, respectively. These results show that the inclusion of legislature-
level behavior in our multi-stage model is important to consider and yields predictions that are
more consistent with actual FAW regulation outcomes.
141
Figure 11: Map of Ballot Model Predictions
WA*
✓L
OR*
✓L
ID*
MT*
WY*
NV*
UT*
CA*
✓B, L
ND*
SD*
NE*
MN
IA
CO*
✓L
KS
MO*
AZ*
✓B
NM
OK*
TX
AR*
LA
AK*
HI
NH
VT
NY
PA
VA
NC
ME*
✓A
MA*
✓B
✓L
RI
CT
NJ
V
✓L
DE
MD
WI
MI*
✓L
IL*
IN
OH*
✓B
WV
✓A
KY
TN
AL
GA
MS*
SC
FL*
✓B
Powered by Bing
© GeoNames, Microsoft, TomTom
Note: States shaded in grey represent where the model prediction matches reality. States with a (P) have passed a
FAW regulation as of 2019. The subscripts “A”, “B”, and “L” correspond to regulations passed through an
administrative regulation, a ballot initiative, and a legislative bill, respectively. The superscript “V” denotes that the
regulation was vetoed by the governor of the state and therefore is not in effect. States with a (*) allow ballot initiatives.
5. Implications
Predicting the outcome of future FAW regulations is important for several reasons.
Producers, animal agriculture supply chain stakeholders, and consumers are all impacted by the
outcomes of FAW regulations. In some cases, such as in California (CA) and Michigan (MI), FAW
regulations also prohibit the import of products not produced in a manner that adheres to their state
regulation(s). As such, consumers within a state with a FAW regulation can be precluded from
purchasing these products, which negatively effects some consumers' welfare. These stakeholder
effects could be seen in other instances of national, provincial, state, and local government
regulations (such as in the EU) that go beyond what is required by the overarching government or
organization.
142
The state-level regulatory process of the US is likely to remain the same in the future, with
all states offering legislative bills and only 24 states allowing ballot initiatives as a means for
passing new laws. As such, our predictions can assist producers and industry stakeholders in
gauging the future of the regulatory landscape and provide guidance on whether to upgrade
existing production methods to comply with anticipated mandates. Our model predicts where new
FAW regulations are most likely to be passed; namely in the seven states that were predicted by
our model to have a FAW law in place in 2019 but did not. These states are Alaska (AK),
Connecticut (CT), Delaware (DE), Maryland (MD), Nevada (NV), New Hampshire (NH), and
Vermont (VT). In Table 38 and Table 39, we present the predicted percentage of the population
within these states that would be in favor of FAW regulations, the number of egg-laying hens and
gestating sows in the state, the relative size of the state's industry to the national total, and the
estimated costs to the industry in each of these states to update to cage-free egg and crate-free pork
production methods. We assume that 18.32% of egg-laying hens and 18.67% of gestating sows
are already in cage-free and crate-free housing systems.39
Table 38: Annual Cost to Update Cage-Free Egg Production in Seven States Predicted to
Pass FAW Regulation
State
Number of
Egg-Laying
Hens
Percent of Egg-
Laying Hens in
the Nation
Alaska*
Connecticut
Delaware
Maryland
Nevada
New Hampshire
Vermont
Total
Note: States with a (*) allow ballot initiatives. We assume a $6.95 increase in production cost per hen per year. Collar
values have been inflated to 2022 dollars.
8,360
3,249,703
3,249,703
2,971,918
15,964
246,099
173,241
9,914,988
0.002
0.882
0.882
0.807
0.004
0.067
0.047
2.69
Annual Cost to
Update to
Cage-Free
System
$47,477.80
$18,455,593.01
$18,455,593.01
$16,878,005.68
$90,662.15
$1,397,636.25
$983,863.82
$56,308,831.72
Percent of
Population in
Favor of FAW
Regulation
57.224
57.659
61.573
59.082
55.794
62.032
63.689
N/A
39 For a detailed explanation of how we calculated these numbers and the estimated costs to the other states without
a FAW regulation in place, please see Appendix A.6.
143
Table 39: Annual Cost to Update to Crate-Free Pork Production in Seven States Predicted to Pass FAW Regulation
State
Alaska*
Connecticut
Delaware
Maryland
Nevada
New Hampshire
Vermont
Total
Note: States with a (*) allow ballot initiatives. We assume an increase in production costs of $51.53 (lower bound) and $87.77 (upper bound) per sow per year.
Dollar values have been inflated to 2022 dollars.
Annual Cost to
Update to Crate-Free
System Lower Bound
$16,754.39
$25,131.59
$167,543.93
$154,978.13
$16,754.39
$25,131.59
$33,508.79
$439,802.81
Annual Cost to
Update to Crate-Free
System Upper Bound
$28,534.83
$42,802.24
$285,348.25
$263,947.13
$28,534.83
$42,802.24
$57,069.65
$749,039.16
Percent of
Population in Favor
of FAW Regulation
57.224
57.659
61.573
59.082
55.794
62.032
63.689
N/A
Percent of
Gestating Sows in
the Nation
0.003
0.005
0.031
0.029
0.003
0.005
0.006
0.080
Number of
Gestating
Sows
400
600
4,000
3,700
400
600
800
10,500
144
As seen in Table 38, these seven states make up almost 3% of national egg production.
Matthews and Sumner (2015) estimate an annual increase of $6.95 per hen in 2022 dollars to
update to a cage-free production method. The total estimated annual cost to update all seven states'
egg industries to cage-free production methods is over $56.3 million annually. Likewise, in Table
39, these seven states make up less than 1% of national pork production. Ortega and Wolf (2018)
estimate an annual increase of between $51.53 and $87.77 per sow in 2022 dollars to update to
crate-free production methods. The estimated cost to update all seven states' pork industries to
crate-free production is between $439.8 thousand and $749.1 thousand annually. These estimated
values do not take into account any effects on interstate commerce. In the case that future FAW
regulations impose restrictions on what products can be sold within a state, the predicted costs to
consumers would be added to these estimates.
Outside of the US, the correlations we find between demographic variables and predicted
public support for FAW regulations are likely transmutable. These associations can help inform
policy makers and industry stakeholders of potential future FAW regulations in countries or
organizations with similar legislative and regulatory processes to the US. Furthermore, we have
provided a novel two-stage, three-part MESR analysis method to incorporate multiple stages of
regulatory processes into predictions that can be applied to other legislative processes throughout
the world. Additionally, an MESR model like the one presented here can be used to model the
regulatory process for other types of agricultural policies and laws, such as laws related to
regulating agricultural pollution.
Further, as affluence increases in developing countries worldwide, it is likely that there will
be an increase in these consumers' activity in food system regulations in the future. Our results
shed light on likely regulatory outcomes in these countries. Countries in which ballot initiatives or
145
a similar process are allowed can anticipate citizens' voting behavior through analyzing
demographic characteristics. On the other hand, countries in which laws are only passed through
a legislature of elected representatives can anticipate that political party-line voting will occur.
6. Conclusion
Nineteen state-level bills and ballot initiatives concerning farm animal welfare (FAW) have
been adopted across 12 states. In this research, we seek to model the evolution of the state-level
FAW regulatory landscape as a function of legislature characteristics and constituent
demographics. More specifically, we utilize a two-stage model to assess (i) whether and when a
given state considers FAW measures, and (ii) if so, the likelihood the measures are passed. We
find that state legislature characteristics influence the likelihood of taking FAW regulatory action
differently between ballot initiatives and legislative bills. Moreover, political party has a stronger
influence on the outcome of votes on legislative bills, while demographics have a stronger effect
on the outcome of votes on ballot initiatives. Finally, we find that new FAW regulations are most
likely to be passed in Alaska, Connecticut, Delaware, Maryland, Nevada, New Hampshire, and
Vermont. We estimate the costs to the egg and pork industries to update to cage- and crate-free
production systems in these states to be small.
Of course, our analysis is not without limitations. For example, our analysis necessarily
reduces nuanced regulation into a binary outcome. However, not all FAW regulations are alike.
Underlying “yes” or “no” outcomes we have modeled are distinct rules that may affect markets
differently. We are unable to account for these differences in our specifications. Further, there are
likely multiple factors leading to over-prediction in FAW regulation voting models, including the
presence of social desirability bias (SDB) in public voting (Lai et al., 2021). Moreover, and perhaps
most importantly, our analysis relies on an assumption that the future – both in terms of what
146
policies are considered and how those considerations play out – is like the past. The US is
constantly experiencing changes in discourse, policy environments, and business strategies at the
firm level. These changes may impact the probabilities that FAW legislation is proposed and how
it is voted on, so out of sample predictions in the future using our model may not be possible. To
the extent that the COVID-19 pandemic, ongoing climate crisis, and other current issues take
center stage in the policy arena, these considerations could “crowd out” agricultural policies, such
as the farm animal welfare regulations considered in this analysis. This could reduce the ongoing
external validity of our results. Despite these concerns, our model does provide a good indication
of FAW support within each state and can serve as a tool for policy makers, industry members,
and other associated groups to understand the FAW regulatory landscape and provide guidance on
whether to upgrade existing enclosures to comply with mandates on the horizon or to continue
operating with “conventional” enclosures.
147
BIBLIOGRAPHY
Abdulai, Abdul Nafeo. 2016. “Impact of conservation agriculture technology on household
welfare in Zambia.” Agricultural economics, 47(6): 729-741.
Animal
Agricultural
Alliance.
2021.
“Issues:
Legislation.”
https://animalagalliance.org/issues/legislation/?fbclid=IwAR08JAohHlmE3zOpGobAOy
VBE_9OlB4jEtFAwDXau5L2iUPvjSWh64FLhEs
Bellemare, Marc F. 2015. “A rant on estimation with binary dependent variables (technical)Marc.”
Boehmke, Frederick J, and John W Patty. 2007. “The selection of policies for ballot initiatives:
What voters can learn from legislative inaction.” Economics & Politics, 19(1): 97-121.
Bourguignon, François, Martin Fournier, and Marc Gurgand. 2007. “Selection bias corrections
based on the multinomial logit model: Monte Carlo comparisons.” Journal of Economic
Surveys, 21(1): 174-205.
Bovay, John, and Daniel A Sumner. 2019. “Animal welfare, ideology, and political labels:
evidence from California’s proposition 2 and Massachusetts’s question 3.” Journal of
Agricultural and Resource Economics, 44(1835-2019-1542): 246-266.
Bovay, John, and Julian M Alston. 2016. “GM labeling regulations by plebiscite: analysis of voting
on Proposition 37 in California.” Journal of Agricultural and Resource Economics, 161-
188.
Bovay, John, and Julian M Alston. 2018. “GMO food labels in the United States: Economic
implications of the new law.” Food Policy, 78: 14-25.
Carter, Colin A, and K Aleks Schaefer. 2019. “Impacts of Mandatory GE Food Labeling: A Quasi-
Natural Experiment.” American Journal of Agricultural Economics, 101(1): 58-73.
Carter, Colin A, K Aleks Schaefer, and Daniel Scheitrum. 2021. “Piecemeal Farm Regulation and
the US Commerce Clause.” American Journal of Agricultural Economics, 103(3): 1141-
1163.
Cornish, Amelia, David Raubenheimer, and Paul McGreevy. 2016. “What we know about the
public level of concern for farm animal welfare in food production in developed countries.”
Animals, 6(11): 74.
Czech, Brian, and Rena Borkhataria. 2001. “The relationship of political party affiliation to
wildlife conservation attitudes.” Politics and the Life Sciences, 3-12.
Deemer, Danielle R, and Linda M Lobao. 2011. “Public concern with farm-animal welfare:
Religion, politics, and human disadvantage in the food sector.” Rural Sociology, 76(2):
167-196.
148
Di Falco, Salvatore. 2014. “Adaptation to climate change in Sub-Saharan agriculture: assessing
the evidence and rethinking the drivers.” European Review of Agricultural Economics,
41(3): 405-430.
Feindt, Peter H, Sandra Schwindenhammer, and Jale Tosun. 2020. “Politicization, depoliticization
and policy change: A comparative theoretical perspective an agri-food policy.” Journal of
Comparative Policy Analysis: Research and Practice, 1-17.
Flynn, Clifton. 2001. “Acknowledging the “Zoological connection”: A sociological analysis of
animal cruelty.” Society & Animals, 9(1): 71-87.
Franklin, Adrian, Bruce Tranter, and Robert White. 2001. “Explaining support for animal rights:
to humans, nonhuman animals, and
two recent approaches
A comparison of
postmodernity.” Society & Animals, 9(2): 127-144.
Gharebaghi, Reza, MR Vaez Mahdavi, Hasan Ghasemi, Amir Dibaei, and Fatemeh Heidary. 2007.
“Animal rights in Islam.” AATEX, 14: 63.
Hall, Peter, Joel L Horowitz, and Bing-Yi Jing. 1995. “On blocking rules for the bootstrap with
dependent data.” Biometrika, 82(3): 561-574.
Heckman, James J. 1976. “The common structure of statistical models of truncation, sample
selection and limited dependent variables and a simple estimator for such models.” In
Annals of economic and social measurement, volume 5, number 4. 475-492. NBER.
Heleski, Camie R, Angela G Mertig, and Adroaldo J Zanella. 2006. “Stakeholder attitudes toward
farm animal welfare.” Anthrozoös, 19(4): 290-307.
Hopkins, Kelsey A, Melissa GS McKendree, K. Aleks Schaefer, and Emma D Rice. 2020.
“Understanding the US Publics’ Voting on Animal Welfare and Genetically Modified
Organism Labeling Ballot Initiatives.” Agricultural and Applied Economics Association
Annual Meeting, [Paper Presentation]. Virtual.
Jaleta, Daniel, Boniface Mbilinyi, Henry Mahoo, and Mulugeta Lemenih. 2016. “Eucalyptus
expansion as relieving and provocative tree in Eithiopia.” Journal of Agriculture and
Ecology Research International, 1-12.
Jerolmack, Colin. 2003. “Tracing the profile of animal rights supporters: A preliminary
investigation.” Society & Animals, 11(3): 245-263.
Kassie, Menale, Hailemariam Teklewold, Paswel Marenya, Moti Jaleta, and Olaf Erenstein. 2015.
“Production risks and food security under alternative technology choices in Malawi:
Application of a multinomial endogenous switching regression.” Journal of Agricultural
Economics, 66(3): 640-659.
149
Lai, Yufeng, Albert Boaitey, and Kota Minegishi. 2021. “Behind the veil: Social desirability bias
and animal welfare ballot initiatives.” Food Policy, 102184.
Langemeier, Michael. 2019. “Long-Term Trends In Pigs Per Litter.”
Malone, Trey, and Jayson L Lusk. 2016. “Putting the Chicken Before the Egg Price: An “Ex Post”
Analysis of California’s Battery Cage Ban.” Journal of Agricultural and Resource
Economics, 518-532.
Matsusaka, John G. 1992. “Economics of direct legislation.” The Quarterly Journal of Economics,
107(2): 541-571.
Matsusaka, John G. 2005. “Direct democracy works.” Journal of Economic Perspectives, 19(2):
185-206.
Matthews, William A, and Daniel A Sumner. 2015. “Effects of housing system on the costs of
commercial egg production.” Poultry Science, 94(3): 552-557.
McKendree, Melissa GS, Candace C Croney, and NJ Olynk Widmar. 2014. “Effects of
demographic factors and information sources on United States consumer perceptions of
animal welfare.” Journal of Animal Science, 92(7): 3161-3173.
Miele, Joseph, Leanne Tingley, Robert Kimball, and John Broida. 1993. “Personality differences
between pro- and antivivisectionists.” Society & Animals, 1(2): 129-144.
Mullally, Conner, and Jayson L Lusk. 2018. “The impact of farm animal housing restrictions on
egg prices, consumer welfare, and production in California.” American Journal of
Agricultural Economics, 100(3): 649-669.
Neill, Clinton L, Rodney B Holcomb, and Jayson L Lusk. 2020. “Estimating potential beggar-thy-
neighbor effects of state labeling programs.” Agribusiness, 36(1): 3-19.
Nibert, David A. 1994. “Animal rights and human social issues.” Society & Animals, 2(2): 115-
124.
Norwood, Franklin Bailey, Glynn Tonsor, and Jayson L Lusk. 2019. “I will give you my vote but
not my money: Preferences for public versus private action in addressing social issues.”
Applied Economic Perspectives and Policy, 41(1): 96-132.
Oldmixon, Elizabeth A. 2017. “Religious representation and animal welfare in the US senate.”
Journal for the Scientific Study of Religion, 56(1): 162-178.
Ortega, David L, and Christopher A Wolf. 2018. “Demand for farm animal welfare and producer
implications: Results from a field experiment in Michigan.” Food Policy, 74: 74-81.
150
Paul, Andrew S, Jayson L Lusk, F Bailey Norwood, and Glynn T Tonsor. 2019. “An experiment
on the vote-buy gap with application to cage-free eggs.” Journal of Behavioral and
Experimental Economics, 79: 102-109.
Peek, Charles W, Nancy J Bell, and Charlotte C Dunham. 1996. “Gender, gender ideology, and
animal rights advocacy.” Gender & Society, 10(4): 464-478.
Politis, Dimitris N, and Halbert White. 2004. “Automatic block-length selection for the dependent
bootstrap.” Econometric Reviews, 23(1): 53-70.
Politis, Dimitris N, and Joseph P Romano. 1994. “The stationary bootstrap.” Journal of the
American Statistical Association, 89(428): 1303-1313.
Smithson, Katie, Max Corbin, Jayson L Lusk, and F Bailey Norwood. 2014. “Predicting state-
wide votes on ballot initiatives to ban battery cages and gestation crates.” Journal of
Agricultural and Applied Economics, 46(1379-2016-113917): 107-124.
Sumner, Daniel A. 2017. “Economics of US state and local regulation of farm practices, with
emphasis on restrictions of interstate trade.” Annual Review of Resource Economics, 9: 13-
31.
Sumner, Daniel A, J Thomas Rosen-Molina, William A Matthews, Joy A Mench, and Kurt R
Richter. 2008. Economic effects of proposed restrictions on egg-laying hen housing in
California. University of California Agricultural Issues Center Davis, CA.
Tolbert, Caroline J, and Daniel A Smith. 2006. “Representation and direct democracy in the United
States.” Representation, 42(1): 25-44.
United Egg Producers. 2022. “Facts and Stats.”
Uyeki, Eugene S, and Lani J Holland. 2000. “Diffusion of pro-environment attitudes?” American
Behavioral Scientist, 43(4): 646-662.
Videras, Julio. 2006. “Religion and animal welfare: Evidence from voting data.” The Journal of
Socio-Economics, 35(4): 652-659.
Vogeler, Colette S. 2019a. “Market-Based governance in farm animal welfare: A comparative
analysis of public and private policies in Germany and France.” Animals, 9(5): 267.
Vogeler, Colette S. 2019b. “Why do farm animal welfare regulations vary between EU member
states? A comparative analysis of societal and party political determinants in France,
Germany, Italy, Spain, and the UK.” JCMS: Journal of Common Market Studies, 57(2):
317-335.
151
Vogeler, Colette S. 2020. “Politicizing farm animal welfare: A comparative study of policy change
in the United States of America.” Journal of Comparative Policy Analysis: Research and
Practice, 1-18.
World Animal Protection. 2021. “Quit Stalling: Are companies making good on promises to end
sow confinement?”
152
APPENDIX A.1 INTERVIEW PROCESS AND ANALYSIS
OVERVIEW
The interviews were audio or video recorded using Zoom or a recorder app, then
transcribed within 24 hours of the interview – using an automated transcription service – and then
I further cleaned and edited the transcripts for grammar and transcription errors. The analysis of
the interview data took part in eight main stages, an expanded version of what is suggested as the
typical structure of qualitative analysis by Miles, Huberman, and Saldaña (2020). Retailers and
processors were analyzed separately, though the steps were identical. These stages are illustrated
in Figure 1.
I identified recurring themes, concepts, and motivations in the interviews and defined
explicit major themes and concepts. Using these concepts and themes, I developed a concept map
and the first round of subject codes following techniques and suggestions from Saldaña (2013).
Then I reviewed the transcripts again, and refined the concept map and codes, to allow for the
addition of new codes/themes or to condense codes/themes together. I applied each cycle of codes
to the interviews until I was satisfied that the codes accurately captured the information needed for
the next phase of data collection.
153
Read through
of transcripts
Identification
of major
themes and
concepts
Concept map
and code
revising
Development
of concept
map and
codes
Application
of codes to
transcripts
Summary
statement
writing
Answering
research
questions
Compilation
of findings
and drawing
conclusions
Figure 1: Stages of Qualitative Data Analysis
For each interview transcript, a summary statement was composed for each of the codes.
Then, these individual summaries were condensed into a single summary statement for each
subject code to convey general interview findings. This technique is used in qualitative research
to display data in an organized manner and see trends or gaps in information (Miles, Huberman,
and Saldaña 2014). These summaries were used to draw conclusions and provide answers to the
research questions and inform the design of the online survey. I intend to review all transcripts at
least one additional time after the quantitative data is collected and analyzed; I think it is prudent
to also review the transcripts through the lens of quantitative findings and potential
interconnections and effects between the groups of interviewees rather than assuming each group
operates in a vacuum.
154
APPENDIX A.2 PROCESSOR, RETAILER, AND CONSUMER
SURVEY QUESTIONS
Meat Processor Survey
Q1 You are being asked to participate in a research study of meat and poultry slaughter and
processing establishment preferences. Processing includes packing, freezing, canning, salting,
smoking, and eviscerating meat products. You should feel free to ask the researchers any
questions you may have. Your participation in this study will take about fifteen to twenty
minutes. You will be asked to respond to a series of questions about how you make decisions at
your establishment. There are 46 questions asking about your preferences for business practices
in addition to questions asking about your establishment’s operations. I also ask some basic
demographic questions. This project will assist researchers to benchmark awareness of food-
related issues and study events that could affect demand. You can choose to not complete the
survey without penalty.
Study Title: U.S. Meat Industry Overview: Consumer Preferences, Retailer Motivations, and
Processor Practices Researcher Title and Contact Information: Melissa G.S. McKendree, PhD,
mckend14@msu.edu and Kelsey Hopkins, PhD Candidate, hopki190@msu.edu, 847-513-1708
Department and Institution: Dept. of Agricultural, Food, and Resource Economics, Michigan
State University Sponsor: USDA-Agriculture and Food Research Initiative
The researchers will not have access to your name or your establishment's name. At no point
will a data file be constructed in which your personal information is linked with your responses.
The data will only be released in summaries in which no individual’s answers can be identified.
You have the right to say no to participating in this research. You can stop at any time after you
have started. There will be no consequences if you stop and you will not be criticized. You will
not lose any benefits that you normally receive.
If you have concerns or questions about this study, such as scientific issues, how to do any part
of it, or to report an injury, please contact the researcher Melissa G.S. McKendree, 202 Morrill
Hall of Agriculture, East Lansing, MI, 48824, mckend14@msu.edu. If you have questions about
your role and rights as a research participant, would like to obtain information or offer input, or
would like to register a complaint about this study, you may contact, anonymously if you wish,
the Michigan State University’s Human Research Protection Program at 517-355-2180, Fax 517-
432-4503, or e-mail irb@msu.edu or regular mail at 4000 Collins Rd, Suite 136, Lansing, MI
48910.
Continuing with the survey means that you voluntarily agree to participate in this research study.
Q2 What type of establishment do you have?
155
Note: By definition, "processing" includes packing, freezing, canning, salting, smoking, and
eviscerating/cutting meat products after slaughter.
o Slaughter establishment (without further processing)
o Processing establishment (without slaughter)
o Slaughter and processing establishment
o Distributor/storage facility
o Other (e.g., custom butcher shop, on-farm custom exempt slaughter)
o None of the above
Q3 Does your establishment slaughter or process ONLY pork, seafood, fish, and/or egg
products?
o Yes
o No
o Not applicable
Q4 To the best of your knowledge, which of these are requirements for a meat or poultry product
to be halal? Select all that apply.
▢ A blessing was spoken over the animal at the time of slaughter
▢ It is free of alcohol or alcohol derivatives
▢ The animal is slaughtered by a Muslim or Person of the Book (Jewish or Christian)
▢ The animal or meat product is imported from a Muslim majority country
▢ It is free of pork/porcine products or derivatives
▢ ⊗I am not sure/don’t know
Q5 Has your establishment ever supplied halal meat or poultry products?
o Yes, currently supplying halal meat/poultry products
o Yes, supplied halal meat/poultry products in the past but not currently
o No, never supplied halal meat/poultry products
Q6 Is your establishment certified for any other niche/specialty or value-added products (e.g.,
organic, kosher, antibiotic free, no added hormones, grass-fed, humanely raised, branded, etc.)?
o No
o Yes; please specify: __________________________________________________
Q7 In what year did your establishment begin operations? _______________________
Q8 In what U.S. state is your establishment?
▼ Alabama ... I do not reside in the United States
Q9 Approximately how many people are employed at your establishment?_______________
Q10 Approximately what percentage of your products are exported for international sale?
o 0%
o 1-25%
o 26-50%
o 51-75%
o 75% or more
o Not applicable/unsure
Q11 What type(s) of animal(s) do you slaughter/process at your establishment?
156
Slaughter
Process
Beef
Veal
Lamb or sheep
Pork
Turkey
Chicken
Goat
Other (e.g., deer,
bison, or exotic fowl)
Q15 Please select the top three (3) motivations for your decision to supply halal meat or poultry
products:
Neither Slaughter nor
Process
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢ My establishment has a halal program so I can sell our products at a premium price
▢ My establishment has a halal program so I can sell to more customers
▢ My establishment has a halal program so I can compete with other similar businesses
with halal programs
▢ My establishment has a halal program so I can provide halal products for Muslim
communities
▢ My establishment has a halal program so I can provide halal products for people with
diverse cultural backgrounds
Q12 What type of halal meat or poultry does your establishment provide? (Select all that apply)
▢ Zabiha (hand-slaughtered) halal
▢ Machine-slaughtered halal
Display This Question:
If Q12 = Zabiha (hand-slaughtered) halal
Q13 For your method of zabiha (hand-slaughtered) halal, is the cut to the animal's throat vertical
or horizontal?
o Vertical
o Horizontal
o I don't know/ I am not sure
Q14 How many years have you had a halal program at your establishment?
o Less than 1 year
o 1-3 years
o 4-5 years
o 6-10 years
o 11-20 years
o More than 20 years
o Q94 Please indicate to what extent each of the following INCENTIVIZED you to start
a halal program to your establishment:
157
Neutral
Not an
incentive
o
o
o
An
incentive
o
o
o
o
o
o
Higher price for my products
Access to new markets to sell my products
Supplying niche religious products to minority
communities
The ability to apply for grants or financial aid to help set
up the program
Assistance from organizations (certifiers, universities,
etc.) to coordinate setting up the halal program
Competing with similar businesses that had a halal
program
Q93 Please indicate to what extent each of the following served as a BARRIER when you started
a halal program to your establishment:
o
o
o
o
o
o
o
o
o
Costs associated with certification program involvement
(certification fees, infrastructure, traceability technology
materials, labor, etc.)
Religious or racial discrimination from regulating bodies
Backlash or displeasure from my non-Muslim customers
Lack of Muslim laborers available near me
Limited local market opportunities to sell my product
Lack of knowledge for how to implement a halal program at
my operation
Display This Question:
Not a
barrier
o
o
o
o
o
o
Neutral A
barrier
o
o
o
o
o
o
o
o
o
o
o
o
If Q2 = Slaughter establishment (without further processing)
Or Q2 = Slaughter and processing establishment
Or Q2 = Other (e.g., custom butcher shop, on-farm custom exempt slaughter)
Q16 Approximately what percent of your establishment’s total slaughter is halal?
o Less than 10%
o 11-25%
o 26 - 50%
o 51-75%
o More than 75%
o Not applicable
Display This Question:
If Q2 = Processing establishment (without slaughter)
Or Q2 = Slaughter and processing establishment
Or Q2 = Other (e.g., custom butcher shop, on-farm custom exempt slaughter)
Q17 Approximately what percent of your establishment’s total processing is halal?
o Less than 10%
o 11-25%
o 26-50%
o 51-75%
o More than 75%
158
o Not applicable
Q18 Approximately what percentage of your halal products are exported for international sale?
o 0%
o 1-25%
o 26-50%
o 51-75%
o 75% or more
o Not applicable/unsure
Display This Question:
If Q18 = 0%
Q19 Why does your establishment not export halal meat or poultry products for international
sale? Select all that apply.
▢ Our halal certification program does not support exports
▢ I do not know how to coordinate exporting our halal products
▢ I do not have excess halal product available for export
▢ I do not want to export our halal products
▢ Other; please specify: __________________________________________________
Q20 Is your operation certified by a third party to provide halal meat or poultry products?
o Yes
o No
o I am not sure
Display This Question:
If Q20 = Yes
Q21 Which organization(s) provides your halal certification? Select all that apply.
Islamic Services of America (ISA)
Halal Food Standards Alliance of America (HFSAA)
Halal Monitoring Services (HMS)
Halal Transactions of Omaha (HTO)
Other; please specify:
Display This Question:
Current
certifier
▢
▢
▢
▢
▢
Past
certifier
▢
▢
▢
▢
▢
Never been
my certifier
▢
▢
▢
▢
▢
If Q2 = Processing establishment (without slaughter)
Or Q2 = Slaughter and processing establishment
Or Q2 = Other (e.g., custom butcher shop, on-farm custom exempt slaughter)
Q22 Approximately what percent of your establishment’s total processing was halal?
o Less than 10%
o 11-25%
o 26-50%
o 51-75%
o More than 75%
o Not applicable
Display This Question:
If Q2 = Slaughter establishment (without further processing)
Or Q2 = Slaughter and processing establishment
159
Or Q2 = Other (e.g., custom butcher shop, on-farm custom exempt slaughter)
Q23 Approximately what percent of your establishment’s total slaughter was halal?
o Less than 10%
o 11-25%
o 26-50%
o 51-75%
o More than 75%
o Not applicable
Q24 Approximately what percentage of your halal meat or poultry products were exported for
international sale?
o 0%
o 1-25%
o 26-50%
o 51-75%
o 75% or more
o Not applicable/unsure
Q25 What type of halal meat or poultry did your establishment provide? (Select all that apply)
▢ Zabiha (hand-slaughtered) halal
▢ Machine-slaughtered halal
Display This Question:
If Q25 = Zabiha (hand-slaughtered) halal
Q26 For your method of zabiha (hand-slaughtered) halal, was the cut to the animal's throat
vertical or horizontal?
o Vertical
o Horizontal
o I don't know/ I am not sure
Q27 Please select the top three (3) motivations for your decision to supply halal meat or poultry
products in the past:
▢ My establishment had a halal program so I could sell our products at a premium price
▢ My establishment had a halal program so I could sell to more customers
▢ My establishment had a halal program so I could compete with other similar businesses
with halal programs
▢ My establishment had a halal program so I could provide halal products for Muslim
communities
▢ My establishment had a halal program so I could provide halal products for people with
diverse cultural backgrounds
Q28 How many years did you have a halal program at your establishment?
o Less than 1 year
o 1-3 years
o 4-5 years
o 6-10 years
o 11-20 years
o More than 20 years
Q29 Was your operation certified by a third party to provide halal meat or poultry products?
o Yes
o No
160
o I am not sure
Display This Question:
If Q29 = Yes
Q30 Why did you end your halal certification? (Check all that apply)
▢ Costs became prohibitive
▢ Wanted to offer different products
▢ Standards were too strict/hard to meet
▢ Poor working relationship with certifier
▢ Insufficient demand for halal products
▢ Other; please specify: __________________________________________________
Display This Question:
If Q29 = Yes
Q31 Which organization(s) was your halal certification from? Select all that apply.
▢ Islamic Services of America (ISA)
▢ Halal Food Standards Alliance of America (HFSAA)
▢ Halal Monitoring Services (HMS)
▢ Halal Transactions of Omaha (HTO)
▢ Other; please specify: __________________________________________________
Q32 Have you ever considered adding a halal program to your establishment again?
o Yes
o No
Q33 Please indicate how likely each of the following are to INCENTIVIZE you to add
a halal program back to your establishment:
Unlikely Neutral
o
o
o
Higher price for my products
Access to new markets to sell my products
Supplying niche religious products to minority
communities
The ability to apply for grants or financial aid to help set up
the program
Assistance from organizations (certifiers, universities, etc.)
to coordinate setting up the halal program
Competing with similar businesses that had a halal program
Q34 Please indicate how likely each of the following serve as a BARRIER in your decision to
add a halal program back to your establishment:
o
o
o
o
o
o
o
o
o
o
o
o
Likely
o
o
o
161
Unlikely Neutral
o
Costs associated with certification program involvement
(certification fees, infrastructure, traceability technology
materials, labor, etc.)
Religious or racial discrimination from regulating bodies
Backlash or displeasure from my non-Muslim customers
Lack of Muslim laborers available near me
Limited local market opportunities to sell my product
Lack of knowledge for how to implement a halal program
at my operation
Q35 Have you ever considered adding a halal program to your establishment?
o
o
o
o
o
o
o
o
o
o
o
Likely
o
o
o
o
o
o
o Yes
o No
o I do/did not know what halal is
Q36 Are you interested in learning more about halal meat production opportunities for your
establishment?
o Yes
o No
Q37 Please indicate how likely each of the following are to INCENTIVIZE you to add
a halal program to your establishment:
Unlikely Neutral
Likely
o
o
o
o
Higher price for my products
Access to new markets to sell my products
Supplying niche religious products to minority communities
The ability to apply for grants or financial aid to help set up
the program
Assistance from organizations (certifiers, universities, etc.)
to coordinate setting up the halal program
Competing with similar businesses that had a halal program
Q38 Please indicate how likely each of the following serve as a BARRIER in your decision to
add a halal program to your establishment:
o
o
o
o
o
o
o
o
o
o
o
o
o
o
Unlikely Neutral
Likely
o
o
o
Costs associated with certification program involvement
(certification fees, infrastructure, traceability technology
materials, labor, etc.)
Religious or racial discrimination from regulating bodies
Backlash or displeasure from my non-Muslim customers
Lack of Muslim laborers available near me
Limited local market opportunities to sell my product
Lack of knowledge for how to implement a halal program at
my operation
Q39 There is not a NHMC program in the U.S. I am interested in your opinions to help design a
future national U.S. meat and poultry halal certification program. If you do not currently have
a halal program, imagine that you are considering adding a halal program to your establishment.
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
In the following section of the survey, you will be presented seven scenarios. Please consider the
162
three factors presented, and indicate which one factor is the least important and which one is the
most important to you when designing a national U.S. halal meat and poultry certification
program. Please select one factor as least important AND one factor as most important in each
question.
The questions look similar but contain different comparisons of factors. Please treat each
question individually.
To help, I have given an example below with ice cream, where flavor is the most important
factor and price is the least important factor in your decision to buy ice cream.
EXAMPLE
Of the following three factors, which one is the least important and which one is the most
important in your decision to buy ice cream?
Q40 Of the following three factors, which one is the least important and which one is the most
important to consider when designing a national U.S. halal meat and poultry certification
program?
Least Important
(Check only one)
Most Important
(Check only one)
o
o
o
⊗Which group(s) will enforce the program (e.g.,
government, religious organization, private non-
religious organization)
⊗What will be required to be certified (e.g.,
products, retailers, slaughter and processing
establishments)
⊗What halal standards will be required (e.g.,
hand versus machine slaughter, stunned or not
stunned)
o
o
o
Q41 Of the following three factors, which one is the least important and which one is the most
important to consider when designing a national U.S. halal meat and poultry certification
program?
Least Important
(Check only one)
o
o
o
⊗Which group(s) will enforce the program (e.g.,
government, religious organization, private non-
religious organization)
⊗Inspection process (e.g., frequency, random or
scheduled)
⊗Benefits associated with certification program
involvement (e.g., access to new markets, price
premiums)
Most Important
(Check only one)
o
o
o
163
Q42 Of the following three factors, which one is the least important and which one is the most
important to consider when designing a national U.S. halal meat and poultry certification
program?
Least Important
(Check only one)
Most Important
(Check only one)
o
o
o
⊗Costs associated with certification program
involvement (e.g., certification fees, infrastructure,
traceability technology materials, labor)
⊗Which group(s) will enforce the program (e.g.,
government, religious organization, private non-
religious organization)
⊗What information will be passed on to my
customers and how they access it (e.g., only
available through the Freedom of Information Act
(FOIA) versus accessible online)
o
o
o
Q43 Of the following three factors, which one is the least important and which one is the most
important to consider when designing a national U.S. halal meat and poultry certification
program?
Least Important
(Check only one)
Most Important
(Check only one)
o
o
o
⊗What information will be passed on to my
customers and how they access it (e.g., only
available through the Freedom of Information Act
(FOIA) versus accessible online)
⊗Inspection process (e.g., frequency, random or
scheduled)
⊗What halal standards will be required (e.g.,
hand versus machine slaughter, stunned or not
stunned)
o
o
o
Q44 Of the following three factors, which one is the least important and which one is the most
important to consider when designing a national U.S. halal meat and poultry certification
program?
Least Important
(Check only one)
o
o
o
⊗Benefits associated with certification program
involvement (e.g., access to new markets, price
premiums)
⊗What halal standards will be required (e.g.,
hand versus machine slaughter, stunned or not
stunned)
⊗Costs associated with certification program
involvement (e.g., certification fees, infrastructure,
traceability technology materials, labor)
Most Important
(Check only one)
o
o
o
164
Q45 Of the following three factors, which one is the least important and which one is the most
important to consider when designing a national U.S. halal meat and poultry certification
program?
Least Important
(Check only one)
Most Important
(Check only one)
o
o
o
⊗Costs associated with certification program
(e.g., facility modifications, traceability equipment,
certification fees, labor)
⊗What will be required to be certified (e.g.,
products, retailers, slaughter and processing
establishments)
⊗Inspection process (e.g., frequency, random or
scheduled)
o
o
o
Q46 Of the following three factors, which one is the least important and which one is the most
important to consider when designing a national U.S. halal meat and poultry certification
program?
Least Important
(Check only one)
Most Important
(Check only one)
⊗What information will be passed on to my
customers and how they access it (e.g., only
available through the Freedom of Information Act
(FOIA) versus accessible online)
⊗Benefits associated with certification program
involvement (e.g., access to new markets, price
premiums)
⊗What will be required to be certified (e.g.,
products, retailers, slaughter and processing
establishments)
165
Q47 If you were to use a national halal certification at your establishment, would you want it to
be internationally accredited or recognized for export?
o Yes
o No
o Maybe
o I would not want to use a national halal certification at my establishment
Q48 Should individual products or supply chain members be required to have a national halal
certification to ensure authentic halal meat and poultry products?
o Individual products
o Supply chain members (e.g., processor, wholesaler, retailer)
o Both individual products and supply chain members
Display This Question:
If Q48 = Supply chain members (e.g., processor, wholesaler, retailer)
Or Q48 = Both individual products and supply chain members
Q49 Which members of the supply chain should be required to have a national halal
certification? Select all that apply.
▢ Slaughter establishments
▢ Processing establishments
▢ Distributors or transportation services
▢ Retailers and wholesalers
▢ Restaurants and food service
▢ Other; please specify: __________________________________________________
Q50 Under a national halal certification, how should a certified business be inspected?
o Pre-scheduled inspections
o Random/surprise inspections
o A mixture of pre-scheduled and random inspections
o Other; please specify: __________________________________________________
Q51 Which benefits of a national halal certification would be most important to your business?
Select up to three.
▢ Access to new domestic (U.S.) markets
▢ Access to new export (international) markets
▢ Ability to charge a higher price for my products
▢ Increased consumer trust
▢ Ease of identifying a certified product
▢ Ease of communicating product attributes
▢ Other; please specify: __________________________________________________
▢ ⊗None of the above
Q52 Which costs of a national halal certification would be most important to your business?
Select up to three.
▢ Reoccurring certification fees
▢ Cost of establishment modifications
▢ Increased labor hours needed
▢ Cost of traceability and/or verification equipment
▢ Other; please specify: __________________________________________________
166
▢ ⊗None of the above
Q53 In your opinion, what standards should be included in a national halal certification for
meat and poultry? Select all that apply.
▢ Zabiha (hand-slaughter)
▢ Machine-slaughter
▢ Slaughterers of Muslim faith
▢ Slaughterers of Christian or Jewish faith
▢ Individual spoken blessings
▢ Animal(s) not stunned
▢ Animal(s) face Mecca at time of slaughter
▢ GMO-free
▢ Other; please specify: __________________________________________________
▢ ⊗I don't know
Q54 Please indicate which parties should have access to the names of establishment(s) at which
a halal meat or poultry product was slaughtered and/or processed using a national halal
certification. Select all that apply.
▢ General public
▢ Processors/slaughterers
▢ Wholesaler/distributor
▢ Retailers/restaurants
▢ Government organizations
▢ ⊗None of the above
Q55 Please indicate which parties should have access to a national list of halal certified
meat/poultry establishments certified with a national halal certification. Select all that apply.
▢ General public
▢ Processors/slaughterers
▢ Wholesalers/distributors
▢ Retailers/restaurants
▢ Government organizations
▢ ⊗None of the above
Q56 Please indicate which parties should have access to the name(s) of enforcement agencies
that certify halal establishments with a national halal certification. Select all that apply.
▢ General public
▢ Processors/slaughterers
▢ Wholesalers/distributors
▢ Retailers/restaurants
▢ Government organizations
▢ ⊗None of the above
Q57 Please indicate which parties should have access to information regarding the halal
standards used in slaughter and/or processing under a national halal certification. Select all that
apply.
▢ General public
▢ Processors/slaughterers
167
▢ Wholesalers/distributors
▢ Retailers/restaurants
▢ Government organizations
▢ ⊗None of the above
Display This Question:
If Q54 = General public
Or Q55 = General public
Or Q56 = General public
Or Q57 = General public
Q58 How should the general public be able to access information related to a national halal
certification program for meat/poultry? Select all that apply.
▢ Online (e.g., company website, online database)
▢ Using a QR code/cell phone app
▢ Freedom of Information Act (FOIA) request
▢ Other; please specify: __________________________________________________
Q59 Who should set standards for a new national halal certification for meat and poultry?
Select all that apply.
Note: The U.S. and state governments are not legally allowed to define standards related to
religious products.
▢ Non-government organizations
▢ Religious organizations
▢ Certifier-led organizations
▢ Producer-led organizations
▢ Wholesalers/distributors
▢ Slaughterers/processors
▢ Retailers/restaurants
▢ Other; please specify: __________________________________________________
Q60 Who should enforce a new national halal certification for meat and poultry? Select all that
apply.
▢ U.S. government organization (e.g., the USDA)
▢ State government organization (e.g., state department of agriculture)
▢ Non-government organizations
▢ Religious organizations
▢ Certifier-led organizations
▢ Producer-led organizations
▢ Slaughterers/processors
▢ Wholesalers/distributors
▢ Retailers/restaurants
▢ Other; please specify: __________________________________________________
Q61 Please select the degree to which you agree or disagree with the following statements:
168
Agree Neutral Disagree
o
o
o
o
o
o
o
o
o
Halal meat or poultry tastes better than non-halal meat or
poultry
All halal meat or poultry slaughter or processing
establishments must be halal certified
Halal establishments that are certified have a stronger
reputation than halal establishments that are not certified
Halal meat/poultry is more sanitary than non-halal meat/
poultry
The halal slaughter process is more humane for the animal
Establishments that are not halal certified cannot be trusted to
supply authentic halal products
Halal meat and poultry is higher quality than non-halal meat
and poultry
Halal meat and poultry is healthier than non-halal meat and
poultry
If an establishment has a good reputation for supplying halal
meat and poultry products, it does not need to be certified as
halal
Q62 What is your role or job at your establishment? __________________________________
Q63 What is your current age?
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o 18-24 years old
o 25-34 years old
o 35-44 years old
o 45-54 years old
o 55-64 years old
o 65-74 years old
o 75 years or older
o Prefer not to disclose
Q64 Were you born in the U.S.?
o Yes
o No, I was born in this country: __________________________________________
o Prefer not to disclose
Q65 Were your parents and grandparents born in the U.S.?
o Yes
o No; they were born in this/these countries:____________________________________
o Prefer not to disclose
Display This Question:
If Q64 = No, I was born in this country:
Or Were you born in the U.S.? Text Response Is Not Empty
Q66 How long have you lived in the U.S.?
o 0-5 years
o 6-10 years
o 11-15 years
o 16-20 years
o 21-25 years
169
o Over 25 years
o Prefer not to disclose
Q67 What is your gender?
o Male
o Female
o Prefer to self-describe: __________________________________________________
o Prefer not to disclose
Q68 What is the highest level of education you have completed?
o Less than High School
o High school graduate or GED
o Some college
o 2-year degree (Associates)
o 4-year degree (BA, BS)
o Professional degree (JD, MD, PhD, etc.)
o Prefer not to disclose
Q69 Are you of Hispanic, Latino, or Spanish origin?
o No
o Yes; please specify: __________________________________________________
o Prefer not to disclose
Q70 What is your race? Select all that apply.
▢ White
▢ Black or African American
▢ Native American or Alaska Native
▢ Native Hawaiian or Pacific Islander
▢ Asian
▢ Other; please specify: __________________________________________________
▢ ⊗Prefer not to disclose
Q71 Which political party do you most identify with?
o Democrat
o Republican
o I am an independent
o Other; please specify: __________________________________________________
o Prefer to not disclose
Q72 Which best describes the area in which you live?
o Rural
o Suburban
o Urban
o Prefer not to disclose
Q73 Do you consider yourself to be religious?
o No
o Yes, I am: __________________________________________________
o Prefer not to disclose
Display This Question:
If Q73 = No
Q74 Have you followed a religion in the past even if you do not do so now?
o No
170
o Yes, I was: __________________________________________________
o Prefer not to disclose
Q75 Do you have any final thoughts or comments you wish to share about using/not using a
halal meat program at your establishment?
________________________________________________________________
________________________________________________________________
171
Meat Retailer Survey
Q1 You are being asked to participate in a research study of meat and poultry retailers. You
should feel free to ask the researchers any questions you may have. Your participation in this
study will take about fifteen to twenty minutes. You will be asked to respond to a series of
questions about how you make decisions at your retail establishment. There are questions asking
about your preferences for business practices in addition to questions asking about your store’s
operations. I also ask some basic demographic questions. This project will assist researchers to
benchmark awareness of food-related issues and study events that could affect demand. You can
choose to not complete the survey without penalty.
Study Title: U.S. Meat Industry Overview: Consumer Preferences, Retailer Motivations, and
Processor Practices
Researcher Title and Contact Information: Melissa G.S. McKendree, PhD, mckend14@msu.edu
and Kelsey Hopkins, PhD Candidate, hopki190@msu.edu, 847-513-1708
Department and Institution: Dept. of Agricultural, Food, and Resource Economics, Michigan
State University
Sponsor: USDA-Agriculture and Food Research Initiative
At no point will a data file be constructed in which your personal information is linked with your
responses. The data will only be released in summaries in which no individual’s answers can be
identified. You have the right to say no to participating in this research. You can stop at any
time after you have started. There will be no consequences if you stop, and you will not be
criticized. You will not lose any benefits that you normally receive.
If you have concerns or questions about this study, such as scientific issues, how to do any part
of it, or to report an injury, please contact the researcher Melissa G.S. McKendree, 202 Morrill
Hall of Agriculture, East Lansing, MI, 48824, mckend14@msu.edu. If you have questions about
your role and rights as a research participant, would like to obtain information or offer input, or
would like to register a complaint about this study, you may contact, anonymously if you wish,
the Michigan State University’s Human Research Protection Program at 517-355-2180, Fax 517-
432-4503, or e-mail irb@msu.edu or regular mail at 4000 Collins Rd, Suite 136, Lansing, MI
48910.
Continuing with the survey means that you voluntarily agree to participate in this research
study.
Please certify that you are over 18 years of age and agree to voluntarily participate in this survey.
o I am over 18 and agree to participate
o I am not over 18 or do not agree to participate
172
Q144 Please ensure you are carefully reading through the survey questions and making
thoughtful selections in order to qualify for the $25 survey incentive. Any nonsense answers,
quality issues, or speeding will be disqualified without incentive.
o Agree
o Disagree
Skip To: End of Survey If Q144 = Disagree
Q2 Do you represent (work for or own) a retail store that sells meat or poultry products?
o No
o Yes
Q3 What is your role or job at your store?____________________________________________
Q4 Which of the following best describes your store?
o Supermarket (e.g., Walmart, Target, Meijer)
o Club membership store (e.g., Costco, Sam's Club)
o Grocery store (e.g., independent small local or regional store)
o Convenience store
o Butcher shop/deli
o Other, please specify: __________________________________________________
o None of the above
Q5 To the best of your knowledge, which of these are requirements for a meat or poultry product
to be halal? Select all that apply.
▢ A blessing was spoken over the animal at the time of slaughter
▢ It is free of alcohol or alcohol derivatives
▢ The animal is slaughtered by a Muslim or Person of the Book (Jewish or Christian)
▢ The animal or meat product is imported from a Muslim majority country
▢ It is free of pork/porcine products or derivatives
▢ ⊗I am not sure/don’t know
Q6 Has your retail store ever sold halal meat or poultry products?
o Yes, currently selling halal meat/poultry products
o Yes, sold halal meat/poultry products in the past but not currently
o No, never sold halal meat/poultry products
o I am unsure/don't know
Q7 Does your retail store sell any other certified niche/specialty or value-added meat or poultry
products besides halal (e.g., organic, kosher, antibiotic free, no added hormones, grass-fed,
humanely raised, branded)?
o No, no other specialty meat or poultry products
o Yes, I sell other specialty meat or poultry products. Please specify: __________________
Q8 The following questions ask for basic information about your retail store. If you are
responsible for multiple retail stores, please answer the questions based on your primary or
flagship retail location.
Q9 In what year did your store begin operations? Please enter a full year, such as 1990. _______
Q10 In what U.S. state is your store?
▼ Alabama ... My store is not in the United States
173
Q11 In what type of area is your store located?
o Rural or countryside
o Suburban or small-mid size city
o Urban or large city
o Prefer not to disclose
Q12 Approximately how many people are employed at your store full time or full time
equivalent? Note: For example, if an employee works 20 hours per week, they are considered 0.5
full time equivalent.
_________________________________________________
Q13 The following questions ask you about the meat and poultry products you sell in your
store.
Q14 Where or how do you sell meat or poultry products at your store? Select all that apply.
o Products are sold prepackaged on shelves (shelf stable products)
o Products are sold prepackaged in a refrigerated or frozen display case (case ready)
o Products are sold prepackaged in a refrigerated or frozen display case (not case ready/cut
in store) Products are sold at a service deli counter
o Products are sold at a service butcher counter
o Other location or method. Please specify: ______________________________________
174
I do not
sell
products
from this
species
at my
store
▢
▢
▢
▢
▢
▢
▢
▢
Q15 What species and types of meat or poultry products do you sell at your store?
Frozen
fully
cooked
Deli
meats/
poultry
Canned,
smoked, or
cured
meat/poultry
products
Ready to
eat
products
(e.g.,
snack
sticks,
jerky)
Frozen
whole
muscle cuts
or ground
(e.g.,
chicken
breasts,
roasts,
ground
turkey)
Fresh
whole
muscle
cuts or
ground
(e.g.,
chicken
breasts,
roasts,
ground
turkey)
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
Beef
Veal
Lamb or
sheep
Pork
Turkey
Chicken
Goat
Other
(e.g.,
deer,
bison,
camel,
or exotic
fowl)
Q16 How many years have halal meat/poultry products been sold at your store?
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
o Less than 1 year
o 1-3 years
o 4-5 years
o 6-10 years
o 11-20 years
o More than 20 years
Q17 Approximately what percent of your store’s total meat/poultry sales are halal?
o Less than 10%
o 11-25%
o 26 - 50%
o 51-75%
o More than 75%
Q18 What standard of halal meat or poultry does your store sell? Select all that apply.
▢ Zabiha (hand-slaughtered) halal
▢ Machine-slaughtered halal
▢ ⊗I am not sure/don't know
Display This Question:
175
If Q18 = Zabiha (hand-slaughtered) halal
Q19 For your zabiha (hand-slaughtered) halal products, is the cut to the animal's throat vertical
or horizontal?
o Vertical (up and down)
o Horizontal (ear to ear)
o I don't know/ I am not sure
Q20 What halal meat or poultry products do you sell at your store? Select all that apply.
▢ Fresh whole muscle cuts or ground (e.g., chicken breasts, roasts, ground turkey)
▢ Frozen whole muscle cuts or ground (e.g., chicken breasts, roasts, ground turkey)
▢ Frozen fully cooked products
▢ Ready to eat shelf stable products (e.g., snack sticks, jerky)
▢ Deli meats/poultry
▢ Canned, smoked, or cured meats/poultry
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q20 = Fresh whole muscle cuts or ground (e.g., chicken breasts,
roasts, ground turkey)
Q21 What types of halal fresh whole muscle cuts do you sell at your store? Select all that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q20 = Frozen whole muscle cuts or ground (e.g., chicken breasts,
roasts, ground turkey)
Q22 What types of halal frozen whole muscle cuts do you sell at your store? Select all that
apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q20 = Frozen fully cooked products
Q23 What types of halal frozen fully cooked products do you sell at your store? Select all that
apply.
▢ Beef
▢ Veal
176
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q20 = Ready to eat shelf stable products (e.g., snack sticks, jerky)
Q24 What types of halal ready to eat (shelf stable) products do you sell at your store? Select all
that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q20 = Deli meats/poultry
Q25 What types of halal deli products do you sell at your store? Select all that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q20 = Canned, smoked, or cured meats/poultry
Q26 What types of halal canned, smoked, or cured products do you sell at your store? Select all
that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If What halal meat or poultry products do you sell at your store? Select all
that apply. Text Response Is Not Empty
Q27 What types of halal ${Q20/ChoiceTextEntryValue/7} products do you sell at your
store? Select all that apply.
▢ Beef
177
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Q28 Would you like to sell other halal meat or poultry products at your store that you do not
currently offer?
o Yes. Please specify other halal meat or poultry products you would like to sell: ________
o No
Q29 Please select the top three (3) motivations for your decision to sell halal meat/poultry
products:
▢ My store offers halal meat/poultry products so I can sell these products at a higher retail
margin
▢ My store offers halal meat/poultry products so I can attract more customers
▢ My store offers halal meat/poultry products so I can compete with other similar
businesses that offer halal meat/poultry products
▢ My store offers halal meat/poultry products so I can provide halal meat/poultry products
for Muslim communities
▢ My store offers halal meat/poultry products so I can provide halal meat/poultry products
for people with diverse cultural backgrounds
Q30 Are any of the halal products your establishment provides certified by a third party? (e.g.,
individual products have a stamp or label that says "halal")
o Yes
o No
o I am not sure
Q31 Is your store certified by a third party to provide halal meat or poultry products?
o Yes
o No
o I am not sure
Display This Question:
If Q31 = Yes
Q32 Which organization(s) provides your store's halal certification? Select all that apply.
Islamic Services of America (ISA)
Halal Food Standards Alliance of America (HFSAA)
Halal Monitoring Services (HMS)
Halal Transactions of Omaha (HTO)
The Islamic Society of the Washington Area (ISWA)
Other. Please specify:
Display This Question:
If Q30 = Yes
Current
certifier
▢
▢
▢
▢
▢
▢
Past
certifier
▢
▢
▢
▢
▢
▢
Never been
my certifier
▢
▢
▢
▢
▢
▢
Q33 Which organization(s) provides your products' halal certification? Select all that apply.
178
Current
certifier
Past
certifier
Never
been my
certifier
Unsure if
they have
ever been
my certifier
▢
▢
▢
▢
▢
▢
Islamic Services of America (ISA)
Halal Food Standards Alliance of America
(HFSAA)
Halal Monitoring Services (HMS)
Halal Transactions of Omaha (HTO)
The Islamic Society of the Washington Area
(ISWA)
Other. Please specify:
▢
Q34 Please indicate to what extent each of the following MOTIVATED your store to start a
halal program:
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
▢
Neutral
Not a
motivation
o
o
o
A
motivation
o
o
o
Higher retail margin for halal products
Access to new customers to sell halal products
Supplying niche religious products to minority
communities
The ability to compete with other similar businesses
that offer halal meat/poultry products
Assistance from organizations (certifiers, producer
groups, etc.) to coordinate finding a supplier
Q35 Please indicate to what extent each of the following was a CHALLENGE when you started
a halal program to your store:
o
o
o
o
o
o
o
o
o
Costs associated with certification program involvement
(certification fees, infrastructure, traceability technology
materials, labor, etc.)
Religious or racial discrimination from regulating bodies
Backlash or displeasure from my non-Muslim customers
Limited local customer base to sell halal products
Lack of knowledge for how to sell halal products at my
store
Display This Question:
Not a
challenge
o
o
o
o
o
Neutral
o
o
o
o
o
A
challenge
o
o
o
o
o
If Q35 = Costs associated with certification program involvement (certification fees,
infrastructure, traceability technology materials, labor, etc.) [ Neutral ]
Or Q35 = Costs associated with certification program involvement (certification fees,
infrastructure, traceability technology materials, labor, etc.) [ A challenge ]
Q36 In the previous question, you indicated that costs were a challenge or barrier when
beginning your halal program. Which of the following costs were the most challenging? Select
all that apply.
▢ Reoccurring certification fees
179
▢ One-time infrastructure costs (e.g., additional refrigerators or shelves)
▢ Increased labor
▢ Traceability technology costs
▢ Other. Please specify: __________________________________________________
Q37 How many years did you sell halal meat/poultry products at your store?
o Less than 1 year
o 1-3 years
o 4-5 years
o 6-10 years
o 11-20 years
o More than 20 years
Q38 Approximately what percent of your store’s total meat/poultry sales were halal?
o Less than 10%
o 11-25%
o 26-50%
o 51-75%
o More than 75%
Q39 What standard of halal meat or poultry products did your store sell? Select all that apply.
▢ Zabiha (hand-slaughtered) halal
▢ Machine-slaughtered halal
▢ ⊗I don't know/am not sure
Display This Question:
If Q39 = Zabiha (hand-slaughtered) halal
Q40 For your zabiha (hand-slaughtered) halal products, was the cut to the animal's throat
vertical or horizontal?
o Vertical (up and down)
o Horizontal (ear to ear)
o I don't know/ I am not sure
Q41 What halal meat or poultry products did you sell at your store? Select all that apply.
▢ Fresh whole muscle cuts or ground (e.g., chicken breasts, roasts, ground turkey)
▢ Frozen whole muscle cuts or ground (e.g., chicken breasts, roasts, ground turkey)
▢ Frozen fully cooked products
▢ Ready to eat shelf stable products (e.g., snack sticks, jerky)
▢ Deli meats/poultry
▢ Canned, smoked, or cured meats/poultry
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q41 = Fresh whole muscle cuts or ground (e.g., chicken breasts,
roasts, ground turkey)
Q42 What types of halal fresh whole muscle cuts did you sell at your store? Select all that
apply.
▢ Beef
▢ Veal
▢ Lamb
180
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q41 = Frozen whole muscle cuts or ground (e.g., chicken breasts,
roasts, ground turkey)
Q43 What types of halal frozen whole muscle cuts did you sell at your store? Select all that
apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q41 = Frozen fully cooked products
Q44 What types of halal frozen fully cooked products did you sell at your store? Select all that
apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q41 = Ready to eat shelf stable products (e.g., snack sticks, jerky)
Q45 What types of halal ready to eat (shelf stable) products did you sell at your store? Select all
that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q41 = Deli meats/poultry
Q46 What types of halal deli products did you sell at your store? Select all that apply.
▢ Beef
▢ Veal
181
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q41 = Canned, smoked, or cured meats/poultry
Q47 What species of halal canned, smoked, or cured products did you sell at your store? Select
all that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If What halal meat or poultry products did you sell at your store? Select all
that apply. Text Response Is Not Empty
Q48 What species of halal ${Q41/ChoiceTextEntryValue/7} products did you sell at your
store? Select all that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Q49 Please select the top three (3) motivations for your decision to sell halal meat/poultry
products:
▢ My store offered halal meat/poultry products so I could sell those products at a higher
retail margin
▢ My store offered halal meat/poultry products so I could attract more customers
▢ My store offered halal meat/poultry products so I could compete with other similar
businesses that offered halal meat/poultry products
▢ My store offered halal meat/poultry products so I could provide halal meat/poultry
products for Muslim communities
▢ My store offered halal meat/poultry products so I could provide halal meat/poultry
products for people with diverse cultural backgrounds
Q50 Why did you stop offering halal? (Check all that apply)
▢ Costs became prohibitive
▢ Wanted to offer different products
▢ Standards were too strict/hard to meet
182
▢ Poor working relationship with certifier
▢ Insufficient demand for halal meat/poultry products
▢ Insufficient supply of halal meat/poultry products
▢ Other. Please specify: __________________________________________________
Q51 Were any of your halal meat/poultry products certified by a third party?
o Yes
o No
o I am not sure
Q52 Was your store certified by a third party to provide halal meat or poultry products?
o Yes
o No
o I am not sure
Display This Question:
If Q51 = Yes
Q53 Which organization(s) was your store's halal certification from? Select all that apply.
▢ Islamic Services of America (ISA)
▢ Halal Food Standards Alliance of America (HFSAA)
▢ Halal Monitoring Services (HMS)
▢ Halal Transactions of Omaha (HTO)
▢ The Islamic Society of the Washington Area (ISWA)
▢ Other. Please specify: __________________________________________________
▢ ⊗Unsure/don't know
Display This Question:
If Q51 = Yes
Q54 Which organization(s) were your products' halal certification from? Select all that apply.
▢ Islamic Services of America (ISA)
▢ Halal Food Standards Alliance of America (HFSAA)
▢ Halal Monitoring Services (HMS)
▢ Halal Transactions of Omaha (HTO)
▢ The Islamic Society of the Washington Area (ISWA)
▢ Other. Please specify: __________________________________________________
▢ ⊗Unsure/don't know
Q55 Have you ever considered selling halal meat/poultry products at your store again?
o Yes
o No
Q56 Please indicate how likely each of the following are to MOTIVATE you to
sell halal meat/poultry products at your store again:
183
Unlikely Neutral
Higher retail margin for halal products
Access to new customers to sell halal products
Supplying niche religious products to minority communities
The ability to compete with other similar businesses that
offer halal meat/poultry products
Assistance from organizations (certifiers, producer groups,
etc.) to coordinate finding a supplier
Q57 Please indicate how likely each of the following would be a CHALLENGE if you were to
sell halal meat/poultry products at your store again:
o
o
o
o
o
o
o
o
o
o
o
Likely
o
o
o
o
Costs associated with certification program involvement
(certification fees, infrastructure, traceability technology
materials, labor, etc.)
Religious or racial discrimination from regulating bodies
Backlash or displeasure from my non-Muslim customers
Limited local customer base to sell halal products
Lack of knowledge for how to sell halal products at my
store
Display This Question:
Unlikely Neutral
o
o
o
o
o
o
o
o
o
o
Likely
o
o
o
o
o
If Q57 = Costs associated with certification program involvement (certification fees,
infrastructure, traceability technology materials, labor, etc.) [ Neutral ]
Or Q57 = Costs associated with certification program involvement (certification fees,
infrastructure, traceability technology materials, labor, etc.) [ Likely ]
Q58 In the previous question, you indicated that costs would be a challenge or barrier when
starting a halal program at your store again. Which of the following costs would be the most
challenging? Select all that apply.
▢ Reoccurring certification fees
▢ One-time infrastructure costs (e.g., additional refrigerators or shelves)
▢ Increased labor
▢ Traceability technology costs
▢ Other. Please specify: __________________________________________________
Q59 Have you ever considered selling halal meat/poultry products at your store?
o Yes
o No
o I do/did not know what halal is
Q60 Are you interested in learning more about halal meat/poultry retail opportunities for your
store?
o Yes
o No
Display This Question:
If Q59 = Yes
Or Q60 = Yes
Q61 What halal meat or poultry products would you like to sell at your store? Select all that
apply.
184
▢ Fresh whole muscle cuts or ground (e.g., chicken breasts, roasts, ground turkey)
▢ Frozen whole muscle cuts or ground (e.g., chicken breasts, roasts, ground turkey)
▢ Frozen fully cooked products
▢ Ready to eat shelf stable products (e.g., snack sticks, jerky)
▢ Deli meats/poultry
▢ Canned, smoked, or cured meats/poultry
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q61 = Fresh whole muscle cuts or ground (e.g., chicken breasts,
roasts, ground turkey)
Q62 What types of halal fresh whole muscle cuts would you like to sell at your store? Select all
that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q61 = Frozen whole muscle cuts or ground (e.g., chicken breasts,
roasts, ground turkey)
Q63 What types of halal frozen whole muscle cuts would you like to sell at your store? Select
all that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q61 = Frozen fully cooked products
Q64 What types of halal frozen fully cooked products would you like to sell at your
store? Select all that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
185
If Q61 = Ready to eat shelf stable products (e.g., snack sticks, jerky)
Q65 What types of halal ready to eat (shelf stable) products would you like to sell at your
store? Select all that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q61 = Deli meats/poultry
Q66 What types of halal deli products would you like to sell at your store? Select all that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q61 = Canned, smoked, or cured meats/poultry
Q67 What species of halal canned, smoked, or cured products would you like to sell at your
store? Select all that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If What halal meat or poultry products would you like to sell at your
store? Select all that apply. Text Response Is Not Empty
Q68 What species of halal ${Q61/ChoiceTextEntryValue/7} products would you like to sell at
your store? Select all that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
186
Q69 Please indicate how likely each of the following are to MOTIVATE you to sell halal meat
or poultry products at your store:
Unlikely Neutral
Likely
o
o
o
o
Higher retail margin for halal products
Access to new customers to sell halal products
Supplying niche religious products to minority communities
The ability to compete with other similar businesses that
offer halal meat/poultry products
Assistance from organizations (certifiers, producer groups,
etc.) to coordinate finding a supplier
Q70 Please indicate how likely each of the following serve as a CHALLENGE in your decision
to sell halal meat or poultry products at your store:
o
o
o
o
o
o
o
o
o
o
o
Costs associated with certification program involvement
(certification fees, infrastructure, traceability technology
materials, labor, etc.)
Religious or racial discrimination from regulating bodies
Backlash or displeasure from my non-Muslim customers
Limited local customer bases to sell halal products
Lack of knowledge for how to sell halal products at my
store
Display This Question:
Unlikely Neutral
o
o
o
o
o
o
o
o
o
o
Likely
o
o
o
o
o
If Q70 = Costs associated with certification program involvement (certification fees,
infrastructure, traceability technology materials, labor, etc.) [ Neutral ]
Or Q70 = Costs associated with certification program involvement (certification fees,
infrastructure, traceability technology materials, labor, etc.) [ Likely ]
Q71 In the previous question, you indicated that costs would be a challenge or barrier when
starting a halal program at your store. Which of the following costs would be the most
challenging? Select all that apply.
▢ Reoccurring certification fees
▢ One-time infrastructure costs (e.g., additional refrigerators or shelves)
▢ Increased labor
▢ Traceability technology costs
▢ Other. Please specify: __________________________________________________
Q146 The following question is to verify that you are paying attention. Please select the animal
that has hooves from the options below.
o Cat
o Horse
o Dog
o Hamster
Skip To: End of Survey If Q146 != Horse
Q72 There is not a NHMC program in the U.S. I am interested in your opinions to help design a
future national U.S. meat and poultry halal certification program. If you do not currently
sell halal meat or poultry products, imagine that you are considering selling halal meat or
187
poultry products at your store.
In the following section of the survey, you will be presented seven scenarios. Please consider the
three factors presented, and indicate which one factor is the least important and which one is the
most important to you when designing a national U.S. halal meat and poultry certification
program. Please select one factor as least important AND one factor as most important in each
question.
The questions look similar but contain different comparisons of factors. Please treat each
question individually.
To help, I have given an example below with ice cream, where flavor is the most important
factor and price is the least important factor in your decision to buy ice cream.
EXAMPLE
Of the following three factors, which one is the least important and which one is the most
important in your decision to buy ice cream?
Please click the arrow to continue.
Q73 Of the following three factors, which one is the least important and which one is the most
important to consider when designing a national U.S. halal meat and poultry certification
program?
Least Important
(Check only one)
Most Important
(Check only one)
o
o
o
⊗Which group(s) will enforce the program (e.g.,
government, religious organization, private non-
religious organization)
⊗What will be required to be certified (e.g.,
products, retailers, slaughter and processing
establishments)
⊗What halal standards will be required (e.g.,
hand versus machine slaughter, stunned or not
stunned)
o
o
o
188
Q74 Of the following three factors, which one is the least important and which one is the most
important to consider when designing a national U.S. halal meat and poultry certification
program?
Least Important
(Check only one)
Most Important
(Check only one)
o
o
o
⊗Which group(s) will enforce the program (e.g.,
government, religious organization, private non-
religious organization)
⊗Inspection process (e.g., frequency, random or
scheduled)
⊗Benefits associated with certification program
involvement (e.g., access to new markets, higher
retail margin, increased consumer trust)
o
o
o
Q75 Of the following three factors, which one is the least important and which one is the most
important to consider when designing a national U.S. halal meat and poultry certification
program?
Least Important
(Check only one)
Most Important
(Check only one)
o
o
o
⊗Costs associated with certification program
involvement (e.g., certification fees, infrastructure,
traceability technology materials, labor)
⊗Which group(s) will enforce the program (e.g.,
government, religious organization, private non-
religious organization)
⊗What information will be passed on to my
customers and how they access it (e.g., only
available through the Freedom of Information Act
(FOIA) versus accessible online)
o
o
o
Q76 Of the following three factors, which one is the least important and which one is the most
important to consider when designing a national U.S. halal meat and poultry certification
program?
Least Important
(Check only one)
Most Important
(Check only one)
o
o
o
⊗What information will be passed on to my
customers and how they access it (e.g., only
available through the Freedom of Information Act
(FOIA) versus accessible online)
⊗Inspection process (e.g., frequency, random or
scheduled)
⊗What halal standards will be required (e.g.,
hand versus machine slaughter, stunned or not
stunned)
o
o
o
189
Q77 Of the following three factors, which one is the least important and which one is the most
important to consider when designing a national U.S. halal meat and poultry certification
program?
Least Important
(Check only one)
Most Important
(Check only one)
o
o
o
⊗Benefits associated with certification program
involvement (e.g., access to new markets, higher
retail margin, increased consumer trust)
⊗What halal standards will be required (e.g.,
hand versus machine slaughter, stunned or not
stunned)
⊗Costs associated with certification program
involvement (e.g., certification fees, infrastructure,
traceability technology materials, labor)
o
o
o
Q78 Of the following three factors, which one is the least important and which one is the most
important to consider when designing a national U.S. halal meat and poultry certification
program?
Least Important
(Check only one)
Most Important
(Check only one)
o
o
o
⊗Costs associated with certification program
(e.g., facility modifications, traceability equipment,
certification fees, labor)
⊗What will be required to be certified (e.g.,
products, retailers, slaughter and processing
establishments)
⊗Inspection process (e.g., frequency, random or
scheduled)
o
o
o
Q79 Of the following three factors, which one is the least important and which one is the most
important to consider when designing a national U.S. halal meat and poultry certification
program?
Least Important
(Check only one)
Most Important
(Check only one)
⊗What information will be passed on to my
customers and how they access it (e.g., only
available through the Freedom of Information Act
(FOIA) versus accessible online)
⊗Benefits associated with certification program
involvement (e.g., access to new markets, higher
retail margin, increased consumer trust)
⊗What will be required to be certified (e.g.,
products, retailers, slaughter and processing
establishments)
190
Q80 If there were a national halal certification, should individual products or supply chain
members be required to have a certification to ensure authentic halal meat and poultry products?
o Individual products
o Supply chain members (e.g., processor, wholesaler, retailer)
o Both individual products and supply chain members
Display This Question:
If Q80 = Supply chain members (e.g., processor, wholesaler, retailer)
Or Q80 = Both individual products and supply chain members
Q81 Which members of the supply chain should be required to have a national halal
certification for their establishment? Select all that apply.
▢ Slaughter establishments
▢ Processing establishments
▢ Distributors or transportation services
▢ Retailers and wholesalers
▢ Restaurants and food service
▢ Other. Please specify: __________________________________________________
Display This Question:
If Q80 = Supply chain members (e.g., processor, wholesaler, retailer)
Or Q80 = Both individual products and supply chain members
Q82 Under a national halal certification, how should a certified business be inspected?
o Pre-scheduled inspections
o Random/surprise inspections
o A mixture of pre-scheduled and random inspections
o Other. Please specify: __________________________________________________
Q83 Under a national halal certification, how should a certified meat or poultry product be
verified or traced? Select all that apply.
▢ Paper trail/certificates
▢ Online universal internet or cloud-based system (e.g., blockchain, RFID)
▢ Online store-specific system (e.g., store records)
▢ Laboratory tests to ensure no pork DNA
▢ Government audits
▢ Third party certifier audits (non-government)
▢ Other. Please specify: __________________________________________________
Q84 Which benefits of a national halal certification would be most important to your business?
Select all that apply.
▢ Access to new markets
▢ Ability to receive a higher retail margin for my products
▢ Increased consumer trust
▢ Ease of identifying a certified product
▢ Ease of communicating product attributes
▢ Other. Please specify: __________________________________________________
Q85 Which costs of a national halal certification would be most important to your business?
Select all that apply.
▢ Reoccurring certification fees
191
▢ Cost of establishment modifications (e.g., more shelves/coolers)
▢ Increased labor hours needed
▢ Cost of traceability and/or verification equipment
▢ Other. Please specify: __________________________________________________
Q86 In your opinion, what standards should be included in a national halal certification for
meat and poultry? Select all that apply.
▢ Zabiha (hand-slaughter)
▢ Machine-slaughter
▢ Slaughterers of Muslim faith
▢ Slaughterers of Christian or Jewish faith
▢ Individual spoken blessings
▢ Animal(s) not stunned
▢ Animal(s) face Mecca at time of slaughter
▢ Non-GMO
▢ Other. Please specify: __________________________________________________
Q87 For a national halal certification, please indicate which parties should have access to the
names of slaughter and/or processing establishment(s) at which an individual halal meat or
poultry product was produced. Select all that apply.
▢ General public
▢ Processors/slaughterers
▢ Distributors/wholesalers
▢ Retailers/restaurants
▢ Government organizations
▢ ⊗None of the above
Q88 Please indicate which parties should have access to a national list of halal certified
meat/poultry establishments certified with a national halal certification. Select all that apply.
▢ General public
▢ Processors/slaughterers
▢ Distributors/wholesalers
▢ Retailers/restaurants
▢ Government organizations
▢ ⊗None of the above
Q89 Please indicate which parties should have access to the name(s) of enforcement agencies
that certify halal establishments with a national halal certification. Select all that apply.
▢ General public
▢ Processors/slaughterers
▢ Distributors/wholesalers
▢ Retailers/restaurants
▢ Government organizations
▢ ⊗None of the above
Q90 Please indicate which parties should have access to information regarding the halal
standards used in slaughter and/or processing under a national halal certification. Select all that
apply.
192
▢ General public
▢ Processors/slaughterers
▢ Distributors/wholesalers
▢ Retailers/restaurants
▢ Government organizations
▢ ⊗None of the above
Display This Question:
If Q87 = General public
Or Q88 = General public
Or Q89 = General public
Or Q90 = General public
Q91 How should the general public be able to access information related to a national halal
certification program for meat/poultry (e.g., where it is was produced, what halal standards were
used)? Select all that apply.
▢ Online (e.g., company website, online database)
▢ Using a QR code/cell phone app
▢ Freedom of Information Act (FOIA) request
▢ Other. Please specify: __________________________________________________
Q92 Who should set standards for a new national halal certification for meat and poultry?
Select all that apply.
Note: U.S. and state governments are not legally allowed to define standards for religious
products.
▢ Non-government organizations
▢ Religious organizations
▢ Certifier-led organizations
▢ Producer-led organizations
▢ Slaughterers/processors
▢ Distributors/wholesalers
▢ Retailers/restaurants
▢ Other. Please specify: __________________________________________________
Q93 Who should enforce a new national halal certification for meat and poultry? Select all that
apply.
▢ U.S. government organization (e.g., the USDA)
▢ State government organization (e.g., state department of agriculture)
▢ Non-government organizations
▢ Religious organizations
▢ Certifier-led organizations
▢ Producer-led organizations
▢ Slaughterers/processors
▢ Distributors/wholesalers
▢ Retailers/restaurants
▢ Other. Please specify: __________________________________________________
Q94 Please select the degree to which you agree or disagree with the following statements:
193
Agree Neutral Disagree
o
o
o
o
o
o
o
o
o
Halal meat or poultry tastes better than non-halal meat or
poultry
All halal meat or poultry retail stores must be halal certified
Halal retail stores that are certified have a stronger reputation
than halal retail stores that are not certified
Halal meat/poultry is more sanitary than non-halal meat/
poultry
The halal slaughter process is more humane for the animal
Retail stores that are not halal certified cannot be trusted to
supply authentic halal products
Halal meat and poultry is higher quality than non-halal meat
and poultry
Halal meat and poultry is healthier than non-halal meat and
poultry
If a retail store has a good reputation for supplying halal meat
and poultry products, it does not need to be certified as halal
Q145 The following question is to verify that you are a real person. Which of the following is
equal to 10 plus 21?
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o
o 11
o 17
o 31
Skip To: End of Survey If Q145 != 31
Q95 What is your current age?
o 18-24 years old
o 25-34 years old
o 35-44 years old
o 45-54 years old
o 55-64 years old
o 65-74 years old
o 75 years or older
o Prefer not to disclose
Q96 Were you born in the U.S.?
o Yes
o No, I was born in this country: ______________________________________________
o Prefer not to disclose
Q97 Were your parents and grandparents born in the U.S.?
o Yes
o No, they were born in this/these countries: ____________________________________
o Prefer not to disclose
Display This Question:
If Q96 = No, I was born in this country:
Or Were you born in the U.S.? Text Response Is Not Empty
Q98 How long have you lived in the U.S.?
o 0-5 years
o 6-10 years
194
o 11-15 years
o 16-20 years
o 21-25 years
o Over 25 years
o Prefer not to disclose
Q99 What is your gender?
o Male
o Female
o Prefer to self-describe: __________________________________________________
o Prefer not to disclose
Q100 What is the highest level of education you have completed?
o Less than High School
o High school graduate or GED
o Some college
o 2-year degree (Associates)
o 4-year degree (BA, BS)
o Professional degree (JD, MD, PhD, etc.)
o Prefer not to disclose
Q101 Are you of Hispanic, Latino, or Spanish origin?
o No
o Yes; please specify: __________________________________________________
o Prefer not to disclose
Q102 What is your race? Select all that apply.
▢ White
▢ Black or African American
▢ Native American or Alaska Native
▢ Native Hawaiian or Pacific Islander
▢ Asian
▢ Other. Please specify: __________________________________________________
▢ ⊗Prefer not to disclose
Q103 Which U.S. political party do you most identify with?
o Democrat
o Republican
o I am an independent
o Other. Please specify: __________________________________________________
o Prefer to not disclose
Q104 Which best describes the area in which you live?
o Rural
o Suburban or mid-size city
o Urban or large city
o Prefer not to disclose
Q105 Do you consider yourself to be religious?
o No
o Yes, I am: __________________________________________________
o Prefer not to disclose
Display This Question:
195
If Q105 = No
Q106 Have you followed a religion in the past even if you do not do so now?
o No
o Yes, I was: __________________________________________________
o Prefer not to disclose
Q107 Do you have any final thoughts or comments you wish to share about selling/not selling
halal meat or poultry at your store?
________________________________________________________________
________________________________________________________________
Q143 Thank you for completing this survey! If you would like to retrieve your $25 reward for
participating, please enter your email address in the box below, and I will reach out to you with
your gift card. If you do not wish to receive your gift card, you can click the "Next" button at the
bottom of the screen to skip this question. Please note that in order to qualify for this gift card,
your survey response will need to meet data quality standards and will be reviewed by the
research team. Gift cards will be sent out on a weekly or bi-weekly basis, so it may take up to
14 days to receive your gift card. Thank you for your participation in this research project!
Email address for gift card:__________________________________________________
196
Halal Meat Consumer Survey
Q1 You are being asked to participate in a research study of U.S. Muslim consumer preferences
for halal meat and poultry retail purchases. Your participation in this study will take about
twenty (20) minutes. You will be asked to respond to a series of questions about how you
purchase halal meat, your preferences for retail locations, and halal meat certifications. We also
ask some basic demographic questions. This survey will assist researchers to anticipate the
demand for various halal meat products and improve awareness of halal-related issues or events
that could affect demand. Researchers are required to provide a consent form to inform you
about the research study, to convey that participation is voluntary, to explain risks and benefits of
participation, and to empower you to make an informed decision. You should feel free to ask the
researchers any questions you may have.
Study Title: U.S. Meat Industry Overview: Consumer Preferences, Retailer Motivations, and
Processor Practices
Researcher Title and Contact Information: Melissa G.S. McKendree, PhD,
mckend14@msu.edu and Kelsey Hopkins, PhD candidate, hopki190@msu.edu, 847-513-1708
Department and Institution: Department of Agricultural, Food, and Resource Economics,
Michigan State University
Sponsor: USDA-Agriculture and Food Research Initiative
The risks associated with this study are minimal. The risks are not greater than those ordinarily
encountered in daily life. Moreover, you may stop the survey at any time. The data will only be
released in summaries in which no individual’s answers can be identified. You have the right to
say no to participate in the research. You can stop at any time after it has already started. There
will be no consequences if you stop, and you will not be criticized. You will not lose any
benefits that you normally receive.
If you have concerns or questions about this study, such as scientific issues, how to do any part
of it, or to report an injury, please contact the researcher Melissa G.S. McKendree, 202 Morrill
Hall of Agriculture, East Lansing, MI, 48824, mckend14@msu.edu. If you have questions or
concerns about your role and rights as a research participant, would like to obtain information or
offer input, or would like to register a complaint about this study, you may contact, anonymously
if you wish, the Michigan State University’s Human Research Protection Program at 517-355-
2180, Fax 517-432-4503, or e-mail irb@msu.edu or regular mail at 4000 Collins Rd, Suite 136,
Lansing, MI 48910.
Continuing with the survey means that you voluntarily agree to participate in this research
study. Please click the arrow at the bottom of the screen to continue.
Q2 Please ensure you are carefully reading through the statements and making thoughtful
selections in order to qualify for the incentive. Any nonsense answers, keyboard slamming,
quality issues, or speeding will be disqualified without incentive.
o Agree and continue
o Disagree and exit
Skip To: End of Block If Please ensure you are carefully reading through the statements and
making thoughtful selections i... = Disagree and exit
197
Q3 Are you 18 years or older and live in the U.S.?
o Yes
o No
Skip To: End of Block If Are you 18 years or older and live in the U.S.? = No
Q4 Have you purchased a halal meat or poultry product in the last twelve (12) months?
o Yes
o No
Skip To: End of Block If Have you purchased a halal meat or poultry product in the last twelve
(12) months? = No
Q5 Do you identify as Muslim?
o Yes
o No
Skip To: End of Block If Do you identify as Muslim? = No
Q6 Are you one of the primary grocery shoppers for your household?
o Yes
o No
Skip To: End of Block If Are you one of the primary grocery shoppers for your household? = No
Q7 Do you follow a vegetarian, vegan, or pescatarian diet? (That is, you do not eat meat.)
o Yes
o No
Q8 How often do you eat halal meat products?
o Always. I never eat non-halal meat products.
o Very often. It is rare that I eat meat that is non-halal.
o Often. I frequently eat halal meat products.
o Somewhat often. I eat halal meat products, but I eat non-halal meat products just as
frequently.
o Almost never. I typically eat non-halal meat products.
o Never.
Q9 If halal meat products are not available, will you purchase kosher meat products to eat
instead?
o Always. I always purchase kosher meat if halal meat is not available.
o Very often. It is very common that I purchase kosher meat if halal meat is not available.
o Often. I frequently purchase kosher meat if halal meat is not available.
o Somewhat often. I purchase kosher meat about half of the time if halal meat is not available.
o Almost never. It is rare that I purchase kosher meat if halal meat is not available.
o Never. I do not purchase kosher meat if halal meat is not available.
198
Q10 If halal meat products are not available, will you purchase vegetarian or pescatarian
(fish/seafood) options instead?
o Always. I always purchase vegetarian or pescatarian options if halal meat is not available.
o Very often. It is very common that I purchase vegetarian or pescatarian options if halal meat
is not available.
o Often. I frequently purchase vegetarian or pescatarian options if halal meat is not available.
o Somewhat often. I purchase vegetarian or pescatarian options about half of the time if halal
meat is not available.
o Almost never. It is rare that I purchase vegetarian or pescatarian options if halal meat is not
available.
o Never. I do not purchase vegetarian or pescatarian options if halal meat is not available.
Q11 Please indicate whether you agree or disagree with the following statements:
▼ Strongly agree ... Not applicable/ Unsure
▼ Strongly agree ... Not applicable/ Unsure
▼ Strongly agree ... Not applicable/ Unsure
⊗I prefer that my halal meat retailer is
reliable
⊗I prefer that my halal meat retailer is
considerate and nice
⊗I prefer that my halal meat retailer is well-
stocked
⊗I prefer that my halal meat retailer sells
quality halal meat products
⊗I prefer that my halal meat retailer is
transparent and honest
Q12 What is the name of the store you purchase the majority of your everyday halal meat and
poultry products from?
Note: "Everyday" halal meat and poultry products are those you purchase for daily
consumption outside of religious holidays such as Eid.
o Name of store: __________________________________________________
Q13 This section of the survey asks you about your buying relationship with
${e://Field/Store%20Name}.
Q14 Please indicate whether you agree or disagree with the following statements:
▼ Strongly agree ... Not applicable/ Unsure
▼ Strongly agree ... Not applicable/ Unsure
${e://Field/Store%20Name} demonstrates empathy and kindness
toward me and treats everyone fairly
${e://Field/Store%20Name} openly shares information, motives,
and choices in straightforward and plain language
${e://Field/Store%20Name} consistently and dependably delivers
on their promises
${e://Field/Store%20Name} communicates product characteristics
in plain and easy to understand language
${e://Field/Store%20Name} resolves issues in an adequate and
timely manner
▼ Strongly agree ...
Not applicable/ Unsure
▼ Strongly agree ...
Not applicable/ Unsure
▼ Strongly agree ...
Not applicable/ Unsure
▼ Strongly agree ...
Not applicable/ Unsure
▼ Strongly agree ...
Not applicable/ Unsure
199
Q15 Please indicate whether you agree or disagree with the following statements:
${e://Field/Store%20Name} quickly resolves issues with safety,
security and satisfaction in mind
${e://Field/Store%20Name} values and respects everyone,
regardless of background, identity or beliefs
${e://Field/Store%20Name} values the good of society and the
environment, not just profit
${e://Field/Store%20Name} is upfront about how they make and
spend money from our interactions
${e://Field/Store%20Name} is clear and upfront about fees and
costs of products, services and experiences
Q16 Please indicate whether you agree or disagree with the following statements:
▼ Strongly agree ...
Not applicable / Unsure
▼ Strongly agree ...
Not applicable / Unsure
▼ Strongly agree ...
Not applicable / Unsure
▼ Strongly agree ...
Not applicable / Unsure
▼ Strongly agree ...
Not applicable / Unsure
${e://Field/Store%20Name}'s products are good quality, accessible
and safe to use
${e://Field/Store%20Name}'s prices are good value for the money ▼ Strongly agree ...
▼ Strongly agree ...
Not applicable/ Unsure
${e://Field/Store%20Name}'s employees and leadership are
competent and understand how to respond to my needs
${e://Field/Store%20Name} can be counted on to improve the
quality of their products and services
${e://Field/Store%20Name} consistently delivers products, services,
and experiences with quality
Q17 Please indicate whether you agree or disagree with the following statements:
Not applicable/ Unsure
▼ Strongly agree ...
Not applicable/ Unsure
▼ Strongly agree ...
Not applicable/ Unsure
▼ Strongly agree ...
Not applicable/ Unsure
⊗${e://Field/Store%20Name} sells quality halal meat products
⊗${e://Field/Store%20Name} takes care of their employees
▼ Strongly agree ...
Not applicable/ Unsure
▼ Strongly agree ...
Not applicable/ Unsure
▼ Strongly agree ...
Not applicable/ Unsure
▼ Strongly agree ...
Not applicable/ Unsure
▼ Strongly agree ...
Not applicable/ Unsure
⊗${e://Field/Store%20Name}'s marketing and communications are
accurate and honest
⊗${e://Field/Store%20Name} creates long term solutions and
improvements that work well for me
⊗${e://Field/Store%20Name} facilitates digital interactions that
run smoothly and work when needed (e.g., placing online or phone
orders)
Q18 Please rate your overall relationship (in terms of trustworthiness, friendliness, reliability,
and transparency) with ${e://Field/Store%20Name}. A score of "100" indicates your relationship
with ${e://Field/Store%20Name} is ideal or perfect, a score of "0" indicates your relationship is
nonexistent.
0 10 20 30 40 50 60 70 80 90 100
200
Relationship
Q19 In this section of the survey, we will be asking you questions about
${e://Field/Store%20Name} and the halal meat or poultry products you purchase there.
Q20 What type of store is ${e://Field/Store%20Name}?
o Small ethnic grocery store (for example, an Asian or Indo-Pakistani grocery store)
o Small local or regional non-ethnic grocery store
o Box/chain store (for example, Walmart, Whole Foods, Meijer)
o Membership store (for example, Costco, Restaurant Depot)
o Butcher shop (that is, they sell only meat or poultry products)
o Online retailer (for example, One Stop Halal, Crescent Foods)
o Other. Please specify: __________________________________________________
Q21 How often do you purchase halal meat or poultry products from
${e://Field/Store%20Name}?
o Daily
o 4-6 times a week
o 2-3 times a week
o Once a week
o Every other week
o Once a month
o Every other month
o 3-5 times per year
o 1-2 times per year
Q22 When you purchase halal meat or poultry, do you purchase only for your household, or do
you purchase on behalf of a group or multiple households?
o Only my household
o Multiple households or a group
Q23 When you buy halal meat or poultry products from ${e://Field/Store%20Name}, what kinds
of products do you purchase? (Select all that apply)
▢ Fresh whole cuts or ground products (for example, chicken breasts or ground beef purchased
at a butcher counter)
▢ Pre-packaged fresh whole cuts or ground products (for example, steak or ground turkey
available in coolers/on refrigerated shelves)
▢ Frozen packaged whole cuts (for example, frozen chicken breasts)
▢ Refrigerated processed products (for example, lunch/deli meat, hot dogs)
▢ Frozen processed products (for example, frozen dinners, frozen hamburger patties)
▢ Ready to eat processed products (for example, jerky, snack sticks)
▢ Other. Please specify: __________________________________________________
Display This Question:
If When you buy halal meat or poultry products from ${e://Field/Store%20Name}, what
kinds of product... = Fresh whole cuts or ground products (for example, chicken breasts or
ground beef purchased at a butcher counter)
Q24 When shopping at ${e://Field/Store%20Name}, which types of halal meat or poultry do
you purchase as fresh whole cuts (for example, chicken breasts or ground beef purchased from a
201
butcher counter)?
Select all that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If When you buy halal meat or poultry products from ${e://Field/Store%20Name}, what
kinds of product... = Pre-packaged fresh whole cuts or ground products (for example, steak or
ground turkey available in coolers/on refrigerated shelves)
Q25 When shopping at ${e://Field/Store%20Name}, which types of halal meat or poultry do
you purchase as pre-packaged refrigerated whole cuts (for example, steaks or ground turkey
available in coolers/on refrigerated shelves)? Select all that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If When you buy halal meat or poultry products from ${e://Field/Store%20Name}, what
kinds of product... = Frozen packaged whole cuts (for example, frozen chicken breasts)
Q26 When shopping at ${e://Field/Store%20Name}, which types of halal meat or poultry do
you purchase as frozen packaged whole cuts (for example, chicken breasts available in
freezers)? Select all that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If When you buy halal meat or poultry products from ${e://Field/Store%20Name}, what
kinds of product... = Refrigerated processed products (for example, lunch/deli meat, hot dogs)
Q27 When shopping at ${e://Field/Store%20Name}, which types of halal meat or poultry do
you purchase as refrigerated processed products (for example, lunch/deli meat, hotdogs)?
Select all that apply.
▢ Beef
▢ Veal
202
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If When you buy halal meat or poultry products from ${e://Field/Store%20Name}, what
kinds of product... = Frozen processed products (for example, frozen dinners, frozen hamburger
patties)
Q28 When shopping at ${e://Field/Store%20Name}, which types of halal meat or poultry do
you purchase as frozen processed products (for example, frozen dinners, frozen hamburger
patties)? Select all that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If When you buy halal meat or poultry products from ${e://Field/Store%20Name}, what
kinds of product... = Ready to eat processed products (for example, jerky, snack sticks)
Q29 When shopping at ${e://Field/Store%20Name}, which types of halal meat or poultry do
you purchase as ready to eat processed products (for example, jerky, snack sticks)?
Select all that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
▢ Chicken
▢ Other. Please specify: __________________________________________________
Display This Question:
If If When you buy halal meat or poultry products from ${e://Field/Store%20Name}, what
kinds of product... Text Response Is Not Empty
Q30 When shopping at ${e://Field/Store%20Name}, which types of halal meat or poultry do
you purchase as ${Q23/ChoiceTextEntryValue/6}?
Select all that apply.
▢ Beef
▢ Veal
▢ Lamb
▢ Goat
▢ Turkey
203
▢ Chicken
▢ Other. Please specify: __________________________________________________
Q31 Is ${e://Field/Store%20Name} certified by a halal certification organization (for example,
the store has a sign/certificate displayed from a certifier)?
o Yes. Please specify certifier if known: ____________________________________
o No
o I don't know
Q32 Does ${e://Field/Store%20Name} sell individually halal certified meat products? (for
example, a package has a label or stamp on it that says "halal")
o Yes. Please specify certifier if known: _____________________________________
o No
o I don't know
Q33 Please select all factors that contributed to your decision to purchase from
${e://Field/Store%20Name}. (Select all that apply)
▢ Recommendations from friends or family
▢ Good customer service
▢ Online reviews
▢ Recommendations from mosque, Islamic community center, or other religious leader/group
▢ Reputation or history as a trustworthy retailer
▢ Store carries a halal certification for all meat or poultry products
▢ Certified halal meat products available
▢ Owner / staff is Muslim
▢ Retailer serves my country's community (for example, Indo-Pakistani, Somali, Turkish)
▢ Low/no sales tax
▢ Near other businesses I shop at
▢ Low/fair prices
▢ Near where I live
▢ Near where I work
▢ Cleanliness / sanitary environment
▢ Quality of products available
▢ Other. Please specify: __________________________________________________
Q34 What standard of halal meat or poultry products do you purchase?
o Zabiha (hand-slaughtered) halal
o Machine-slaughtered halal
o Whichever is available
o I do not have a preference
o I do not know the difference between Zabiha and machine-slaughtered halal
Q35 What standards do you prefer when you purchase your halal meat and poultry products?
Select all that apply.
▢ Zabiha (hand-slaughter)
▢ Machine-slaughter
▢ Slaughterers of Muslim faith
▢ Slaughterers of Christian or Jewish faith
204
▢ Individual spoken blessings
▢ Animal(s) not stunned
▢ Animal(s) face Mecca at time of slaughter
▢ Does not contain genetically modified organisms (non-GMO)
▢ Other. Please specify: __________________________________________________
Q36 Where or from whom do you get information about what standards are required for meat
or poultry to be halal? Select all that apply.
▢ Friends / family
▢ Religious leader / group (for example, from your imam, your mosque, or Islamic community
center)
▢ Social media page(s) (for example, a Facebook group)
▢ Store websites / advertisements
▢ Halal certifiers
▢ Other. Please specify: __________________________________________________
Q37 Where or from whom do you get information about where to find authentic halal meat or
poultry? Select all that apply.
▢ Friends / family
▢ Religious leader / group (for example, from your imam, your mosque, or Islamic community
center)
▢ Social media page(s) (for example, a Facebook group)
▢ Store websites / advertisements
▢ Halal certifiers
▢ Consumer apps / websites (for example, Scan Halal)
▢ Other. Please specify: __________________________________________________
Q56 The following question is to verify that you are a real person. Which of the following is
equal to 10 + 21?
o 9
o 31
o 53
Skip To: End of Block If The following question is to verify that you are a real person. Which of
the following is equal t... != 31
Q38 Please read these instructions and descriptions carefully, as they are important for the
next questions.
In this section of the survey, you will be asked eight (8) questions where you will choose
between two (2) different stores to purchase halal meat and poultry products, or have the option
to not shop at either store.
When comparing the stores, you will see different characteristics: The use of a comprehensive
store-wide halal certification A halal certification only on individual products Your one-way
travel time to the store
the stores are the same (for example, prices and variety of products available) - that is,
your choice should depend on the information shown.
In the following questions, you will be shown pictures of the stores you are choosing between to
Your relationship with the retailer. All other characteristics of
205
help you to visualize your choice.
____________________________________________________________________________
Halal Certifications
For the purpose of this survey, the halal standards for these certifications are those that you
prefer (for example, hand or machine cut, stunned or not stunned).
On the front of these stores, you will see different halal certification logos, which are also
shown here with a description:
The store above with the green 8-point star is certified halal, meaning that all the meat and
poultry products sold inside are certified as halal. The store above with the blue 18-point star
carries halal certified meat and poultry products, but not all meat and poultry products sold
inside are necessarily certified as halal. Halal certified meat and poultry products would be
labeled individually with the halal certification. The store above with the green 8-point star
and blue 18-point star is both halal certified and carries halal certified meat and poultry
products, meaning that all the meat and poultry products sold inside are certified as halal and all
of these products are additionally individually labeled as halal certified.
____________________________________________________________________________
Travel Time
Travel time is the one-way time shown is the minutes it takes you to get to the retailer from
your home. Travel times will be either 15, 30, 45, or 60 minutes one-way.
____________________________________________________________________________
Relationship
You will see four (4) different possible levels of your relationship with the retailer; these are
given below with definitions.
No relationship with this retailer: You have never purchased from this store before and do not
know anything about their trustworthiness, friendliness, reliability, or transparency.
Relationship I have with my current retailer: This is the current relationship you have with
the retailer you purchase from most frequently, in terms of trustworthiness, friendliness,
reliability, and transparency.
Best retailer relationship I have experienced: This is the best relationship you have had with a
retailer in terms of trustworthiness, friendliness, reliability, and transparency.
Ideal or perfect retailer relationship: This is the most trustworthy, friendly, reliable, and
transparent retailer you can imagine purchasing halal meat and poultry products from.
Q39 IMPORTANT: Previous similar surveys have found that people often state they are willing
to shop at different stores when they are not actually willing to do so. Accordingly, it is
important that you make each of your upcoming selections like you would if you were actually
facing these exact choices; that is, noting that choosing to shop at one store means that you
would not shop at the other location for your halal meat or poultry products. The accuracy of
your responses is very important, as the information collected here will be used to help design
future policies and regulations for halal meat and poultry products.
206
Please read carefully and be aware that every question has different information even
though they may look very similar.
Q40 Imagine you are grocery shopping and want to purchase halal meat or poultry products. Of
the stores below, which would you prefer to purchase from, taking into account halal
certifications, one-way travel time to the store, and your relationship with the retailer?
o
No certifications present 60 minutes total time spent traveling to the store Best retailer
relationship I have experienced
Store and product certifications present 45 minutes total time spent traveling to the store
Ideal or perfect retailer relationship
o
o I would not shop for halal meat products at either of these two stores
Q41 Imagine you are grocery shopping and want to purchase halal meat or poultry products. Of
the stores below, which would you prefer to purchase from, taking into account halal
certifications, one-way travel time to the store, and your relationship with the retailer?
o
No certifications present 15 minutes total time spent traveling to the store Ideal or perfect
retailer relationship
Product and store certifications present 60 minutes total time spent traveling to the store
No relationship with this retailer
o
o I would not shop for halal meat products at either of these two stores
Q42 Imagine you are grocery shopping and want to purchase halal meat or poultry products. Of
the stores below, which would you prefer to purchase from, taking into account halal
certifications, one-way travel time to the store, and your relationship with the retailer?
o
Product certification present 30 minutes total time spent traveling to the store No
relationship with this retailer
Store certification present 15 minutes total time spent traveling to the store Relationship I
have with my current retailer
o
o I would not shop for halal meat products at either of these two stores
Q43 Imagine you are grocery shopping and want to purchase halal meat or poultry products. Of
the stores below, which would you prefer to purchase from, taking into account halal
certifications, one-way travel time to the store, and your relationship with the retailer?
o
Product certification present 45 minutes total time spent traveling to the store Relationship I
have with my current retailer
Store certification present 30 minutes total time spent traveling to the store Best retailer
relationship I have experienced
o
o I would not shop for halal meat products at either of these two stores
Q44 Imagine you are grocery shopping and want to purchase halal meat or poultry products. Of
the stores below, which would you prefer to purchase from, taking into account certifications,
one-way travel time to the store, and your relationship with the retailer?
o
Store certification present 45 minutes total time spent traveling to the store No relationship
with this retailer
Product certification present 30 minutes total time spent traveling to the store Relationship
I have with my current retailer
o
o I would not shop for halal meat products at either of these two stores
Q45 Imagine you are grocery shopping and want to purchase halal meat or poultry products. Of
the stores below, which would you prefer to purchase from, taking into account halal
certifications, one-way travel time to the store, and your relationship with the retailer?
207
o
o
Store certification present 30 minutes total time spent traveling to the store Relationship I
have with my current retailer
Product certification present 15 minutes total time spent traveling to the store Best retailer
relationship I have experienced
o I would not shop for halal meat products at either of these two stores
Q46 Imagine you are grocery shopping and want to purchase halal meat or poultry products. Of
the stores below, which would you prefer to purchase from, taking into account halal
certifications, one-way travel time to the store, and your relationship with the retailer?
o
Store and product certifications present 15 minutes total time spent traveling to the store
Best retailer relationship I have experienced
No certifications present 60 minutes total time spent traveling to the store Ideal or perfect
retailer relationship
o
o I would not shop for halal meat products at either of these two stores
Q47 Imagine you are grocery shopping and want to purchase halal meat or poultry products. Of
the stores below, which would you prefer to purchase from, taking into account halal
certifications, one-way travel time to the store, and your relationship with the retailer?
o
Store and product certifications present 60 minutes total time spent traveling to the store
Ideal or perfect retailer relationship
No certifications present 45 minutes total time spent traveling to the store No relationship
with this retailer
o
o I would not shop for halal meat products at either of these two stores
Q48 Imagine you are grocery shopping and want to purchase halal meat or poultry products. Of
the stores below, which would you prefer to purchase from, taking into account halal
certifications, one-way travel time to the store, and your relationship with the retailer?
o
No certifications present 15 minutes total time spent traveling to the store No relationship
with this retailer
Store and product certifications present 60 minutes total time spent traveling to the store
Relationship I have with my current retailer
o
o I would not shop for halal meat products at either of these two stores
Q49 Imagine you are grocery shopping and want to purchase halal meat or poultry products. Of
the stores below, which would you prefer to purchase from, taking into account halal
certifications, one-way travel time to the store, and your relationship with the retailer?
o
No certifications present 60 minutes total time spent traveling to the store Relationship I
have with my current retailer
Store and product certifications present 45 minutes total time spent traveling to the store
Best retailer relationship I have experienced
o
o I would not shop for halal meat products at either of these two stores
Q50 Imagine you are grocery shopping and want to purchase halal meat or poultry products. Of
the stores below, which would you prefer to purchase from, taking into account halal
certifications, one-way travel time to the store, and your relationship with the retailer?
o
Product certification present 45 minutes total time spent traveling to the store Best retailer
relationship I have experienced
Store certification present 30 minutes total time spent traveling to the store Ideal or perfect
retailer relationship
o
o I would not shop for halal meat products at either of these two stores
208
Q51 Imagine you are grocery shopping and want to purchase halal meat or poultry products. Of
the stores below, which would you prefer to purchase from, taking into account halal
certifications, one-way travel time to the store, and your relationship with the retailer?
o
Product certification present 30 minutes total time spent traveling to the store Ideal or
perfect retailer relationship
Store certification present 15 minutes total time spent traveling to the store No relationship
with this retailer
o
o I would not shop for halal meat products at either of these two stores
Q52 Imagine you are grocery shopping and want to purchase halal meat or poultry products. Of
the stores below, which would you prefer to purchase from, taking into account halal
certifications, one-way travel time to the store, and your relationship with the retailer?
o
Store certification present 30 minutes total time spent traveling to the store Best retailer
relationship I have experienced
Product certification present 15 minutes total time spent traveling to the store Ideal or
perfect retailer relationship
o
o I would not shop for halal meat products at either of these two stores
Q53 Imagine you are grocery shopping and want to purchase halal meat or poultry products. Of
the stores below, which would you prefer to purchase from, taking into account halal
certifications, one-way travel time to the store, and your relationship with the retailer?
o
Store certification present 45 minutes total time spent traveling to the store Ideal or perfect
retailer relationship
Product certification present 30 minutes total time spent traveling to the store No
relationship with this retailer
o
o I would not shop for halal meat products at either of these two stores
Q54 Imagine you are grocery shopping and want to purchase halal meat or poultry products. Of
the stores below, which would you prefer to purchase from, taking into account halal
certifications, one-way travel time to the store, and your relationship with the retailer?
o
Store and product certifications present 60 minutes total time spent traveling to the store
No relationship with this retailer
No certification present 45 minutes total time spent traveling to the store Relationship I
have with my current retailer
o
o I would not shop for halal meat products at either of these two stores
Q55 Imagine you are grocery shopping and want to purchase halal meat or poultry products. Of
the stores below, which would you prefer to purchase from, taking into account halal
certifications, one-way travel time to the store, and your relationship with the retailer?
o
Store and product certifications present 15 minutes total time spent traveling to the store
Relationship I have with my current retailer
No certification present 60 minutes total time spent traveling to the store Best retailer
relationship I have experienced
o
o I would not shop for halal meat products at either of these two stores
Q57 Please select the degree to which you agree or disagree with the following statements:
209
Halal meat and poultry is healthier to eat than non-halal meat and
poultry
I will never eat meat or poultry that is not halal certified
▼ Strongly agree ...
Not applicable / Unsure
▼ Strongly agree ...
Not applicable / Unsure
▼ Strongly agree ...
Not applicable / Unsure
▼ Strongly agree ...
Not applicable / Unsure
Halal meat and poultry is cleaner/more hygienic than non-halal
meat or poultry
I have access to good information about halal certified meat and
poultry in the U.S.
Halal meat and poultry tastes better than non-halal meat and poultry ▼ Strongly agree ...
Q58 Please select the degree to which you agree or disagree with the following statements:
Not applicable / Unsure
I will not buy a meat or poultry product if my peers or family have
doubts about whether it is truly halal
I always check labels to see if all ingredients are halal before
purchasing
Halal meat and poultry slaughtering is more humane than non-halal
meat and poultry slaughtering
It is easy for me to tell if a meat or poultry product is halal
I am willing to travel extra miles to get authentic halal meat or
poultry products
I check to see if a restaurant serves halal food before I eat there
▼ Strongly agree ...
Not applicable / Unsure
▼ Strongly agree ...
Not applicable / Unsure
▼ Strongly agree ...
Not applicable / Unsure
▼ Strongly agree ...
Not applicable / Unsure
▼ Strongly agree ...
Not applicable / Unsure
▼ Strongly agree ...
Not applicable / Unsure
Q59 Please select the degree to which you agree or disagree with the following statements:
I am willing to pay more for meat or poultry that has been certified
as halal
If halal meat or poultry is not available, I will chose a seafood or
vegetarian option instead
I always check to see if a meat or poultry product is certified halal
before eating it
I have enough knowledge about U.S. halal meat and poultry to tell
the difference between halal certified and non-halal meat and
poultry
Q60 Now we will move onto the next portion of the survey, which will ask you questions about
your opinions about a national halal certification program.
Q61 There is not a NHMC program in the U.S. We are interested in your opinions to help design
a future national U.S. meat and poultry halal certification program.
▼ Strongly agree ...
Not applicable / Unsure
▼ Strongly agree ...
Not applicable / Unsure
▼ Strongly agree ...
Not applicable / Unsure
▼ Strongly agree ...
Not applicable / Unsure
In the following section of the survey, you will be presented seven (7) scenarios. Please consider
the three (3) factors presented, and indicate which one (1) factor is the least important and which
one (1) is the most important to you when designing a national U.S. halal meat and poultry
certification program. Please select one factor as least important AND one factor as most
important in each question.
210
The questions look similar but contain different comparisons of factors. Please treat each
question individually.
To help, we have given an example below with ice cream, where flavor is the most important
factor and price is the least important factor in your decision to buy ice cream.
EXAMPLE
Of the following three (3) factors, which one (1) is the least important and which one (1) is the
most important in your decision to buy ice cream?
Q62 Of the following three (3) factors, which one (1) is the least important and which one (1) is
the most important to consider when designing a national U.S. halal meat and poultry
certification program?
Least Important
(Check only one)
o
Most Important
(Check only one)
o
⊗Which group(s) will enforce the program (for
example, government, religious organization, private
non-religious organization)
⊗What will be required to be certified (for
example, products, retailers, slaughter and processing
establishments)
⊗What halal standards will be required (for
example, hand versus machine slaughter, stunned or
not stunned)
Q63 Of the following three (3) factors, which one (1) is the least important and which one (1) is
the most important to consider when designing a national U.S. halal meat and poultry
certification program?
Least Important
(Check only one)
o
Most Important
(Check only one)
o
⊗Which group(s) will enforce the program (for
example, government, religious organization, private
non-religious organization)
⊗Inspection process (for example, frequency,
random or scheduled)
⊗Benefits associated with certified products (for
example, transparency, reliability, quality)
o
o
o
o
o
o
o
o
211
Q64 Of the following three (3) factors, which one (1) is the least important and which one (1) is
the most important to consider when designing a national U.S. halal meat and poultry
certification program?
Least Important
(Check only one)
o
Most Important
(Check only one)
o
Q65 Of the following three (3) factors, which one (1) is the least important and which one (1) is
the most important to consider when designing a national U.S. halal meat and poultry
certification program?
Least Important
(Check only one)
o
Most Important
(Check only one)
o
⊗Costs associated with certified products (for
example, higher prices)
⊗Which group(s) will enforce the program (for
example, government, religious organization, private
non-religious organization)
⊗What product information I have access to and
how I can access it (for example, only available
through the Freedom of Information Act (FOIA)
versus accessible online)
⊗What product information I have access to and
how I can access it (for example, only available
through the Freedom of Information Act (FOIA)
versus accessible online)
⊗Inspection process (for example, frequency,
random or scheduled)
⊗What halal standards will be required (for
example, hand versus machine slaughter, stunned or
not stunned)
Q66 Of the following three (3) factors, which one (1) is the least important and which one (1) is
the most important to consider when designing a national U.S. halal meat and poultry
certification program?
Least Important
(Check only one)
o
Most Important
(Check only one)
o
⊗Benefits associated with certified products (for
example, transparency, reliability, quality)
⊗What halal standards will be required (for
example, hand versus machine slaughter, stunned or
not stunned)
⊗Costs associated with certified products (for
example, higher prices)
o
o
o
o
o
o
o
o
o
o
o
o
212
Q67 Of the following three (3) factors, which one (1) is the least important and which one (1) is
the most important to consider when designing a national U.S. halal meat and poultry
certification program?
Least Important
(Check only one)
o
Most Important
(Check only one)
o
o
o
⊗Costs associated with certified products (for
example, higher prices)
⊗What will be required to be certified (for
example, products, retailers, slaughter and
processing establishments)
⊗Inspection process (for example, frequency,
random or scheduled)
o
o
Q68 Of the following three (3) factors, which one (1) is the least important and which one (1) is
the most important to consider when designing a national U.S. halal meat and poultry
certification program?
Least Important
(Check only one)
Most Important
(Check only one)
⊗What product information I have access to and
how I can access it (for example, only available
through the Freedom of Information Act (FOIA)
versus accessible online)
⊗Benefits associated with certified products (for
example, transparency, reliability, quality)
⊗What will be required to be certified (for
example, products, retailers, slaughter and processing
establishments)
213
Q69 Should individual products or supply chain members be required to have a national halal
certification to ensure authentic halal meat and poultry products?
o Individual products
o Supply chain members (for example, processor, wholesaler, retailer)
o Both individual products and supply chain members
Display This Question:
If Should individual products or supply chain members be required to have a national halal
certifica... = Supply chain members (for example, processor, wholesaler, retailer)
Or Should individual products or supply chain members be required to have a national halal
certifica... = Both individual products and supply chain members
Q70 Which members of the supply chain should be required to have a national halal
certification? Select all that apply.
▢ Slaughter establishments
▢ Processing establishments
▢ Distributors or transportation services
▢ Retailers and wholesalers
▢ Restaurants and food service
▢ Other. Please specify: __________________________________________________
Q71 Under a national halal certification, how should a certified business be inspected?
o Pre-scheduled inspections
o Random/surprise inspections
o A mixture of pre-scheduled and random inspections
o Other. Please specify: __________________________________________________
Q72 If there were a national halal certification for meat and poultry, what should be the
minimum requirements for the certification? Select all that apply.
▢ Zabiha (hand-slaughter)
▢ Machine-slaughter
▢ Slaughterers of Muslim faith
▢ Slaughterers of Christian or Jewish faith
▢ Individual spoken blessings
▢ Animal(s) not stunned
▢ Animal(s) face Mecca at time of slaughter
▢ Not genetically modified (Non-GMO)
▢ Other. Please specify: __________________________________________________
Q73 Please indicate which parties should have access to the names of establishment(s) at which
a halal meat or poultry product was slaughtered and/or processed using a national halal
certification. Select all that apply.
▢ General public
▢ Processors/slaughterers
▢ Distributors/wholesalers
▢ Retailers/restaurants
▢ Government organizations
▢ ⊗None of the above
214
Q74 Please indicate which parties should have access to a national list of halal certified
meat/poultry establishments certified with a national halal certification. Select all that apply.
▢ General public
▢ Processors/slaughterers
▢ Distributors/wholesalers
▢ Retailers/restaurants
▢ Government organizations
▢ ⊗None of the above
Q75 Please indicate which parties should have access to the name(s) of enforcement agencies
that certify halal establishments with a national halal certification. Select all that apply.
▢ General public
▢ Processors/slaughterers
▢ Distributors/wholesalers
▢ Retailers/restaurants
▢ Government organizations
▢ ⊗None of the above
Q76 Please indicate which parties should have access to information regarding the halal
standards used in slaughter and/or processing under a national halal certification. Select all that
apply.
▢ General public
▢ Processors/slaughterers
▢ Distributors/wholesalers
▢ Retailers/restaurants
▢ Government organizations
▢ ⊗None of the above
Display This Question:
If Please indicate which parties should have access to the names of establishment(s) at
which a hala... = General public
Or Please indicate which parties should have access to a national list of halal certified
meat/poult... = General public
Or Please indicate which parties should have access to the name(s) of enforcement agencies
that cert... = General public
Or Please indicate which parties should have access to information regarding the halal
standards use... = General public
Q77 How should the general public be able to access information related to a national halal
certification program for meat/poultry? Select all that apply.
▢ Online (for example, company website, online database)
▢ Using a QR code/cell phone app
▢ Freedom of Information Act (FOIA) request
▢ Other. Please specify: __________________________________________________
Q78 Who should set standards for a new national halal certification for meat and poultry? Select
all that apply.
Note: U.S. and state governments are not legally allowed to define standards for religious
products.
215
▢ Non-government organizations
▢ Religious organizations
▢ Certifier-led organizations
▢ Producer-led organizations
▢ Slaughterers/processors
▢ Distributors/wholesalers
▢ Retailers/restaurants
▢ Other. Please specify: __________________________________________________
Q79 Who should enforce a new national halal certification for meat and poultry? Select all that
apply.
▢ U.S. government organization (for example, the USDA)
▢ State government organization (for example, state department of agriculture)
▢ Non-government organizations
▢ Religious organizations
▢ Certifier-led organizations
▢ Producer-led organizations
▢ Slaughterers/processors
▢ Distributors/wholesalers
▢ Retailers/restaurants
▢ Other. Please specify: __________________________________________________
Q80 Please select the degree to which you agree or disagree with the following statements:
Agree Neutral
o
o
Disagree
o
o
o
o
o
o
o
o
o
All halal meat or poultry slaughter or processing
establishments must be halal certified
Halal establishments that are certified have a stronger
reputation than halal establishments that are not certified
Establishments that are not halal certified cannot be trusted to
supply authentic halal products
If an establishment has a good reputation for supplying halal
meat and poultry products, it does not need to be certified as
halal
Q81 Now we are going to ask you questions about yourself. These responses will only be used
by the research team and will not be shared with any identifying information attached.
Q82 Which branch of Islam do you practice?
o Sunni
o Shia or Shiite
o Ibadi
o Non-denominational
o Other. Please specify: __________________________________________________
o Prefer not to disclose
Q83 Have you converted or reverted to Islam?
o Yes
o No
o Prefer not to disclose
o
216
Display This Question:
If Have you converted or reverted to Islam? = Yes
Q84 Which religion did you follow prior to converting or reverting to Islam?
o Christianity
o Hinduism
o Buddhism
o Judaism
o Sikhism
o Other. Please specify: __________________________________________________
o None/ not applicable
o Prefer not to disclose
Display This Question:
If Which religion did you follow prior to converting or reverting to Islam? = Christianity
Q85 Which branch of Christianity did you follow prior to converting or reverting to Islam?
o Catholicism
o Protestant (for example, Baptist, Lutheran, Methodist, Presbyterian, Pentecostal,
Episcopalian)
o Orthodox (for example, Greek, Eastern)
o Prefer not to disclose
Display This Question:
If Have you converted or reverted to Islam? = Yes
Q86 When did you convert to Islam?
o 0-5 years ago
o 6-10 years ago
o 11-15 years ago
o 16-20 years ago
o 21-25 years ago
o 26-30 years ago
o 31-35 years ago
o Over 35 years ago
o Prefer not to disclose
Q87 I consider myself:
o A vegetarian
o A vegan
o None of the above
o Prefer not to disclose
Q88 What is your current age?
o 18 - 24 years old
o 25 - 34 years old
o 35 - 44 years old
o 45 - 54 years old
o 55 - 64 years old
o 65 - 74 years old
o 75 years or older
o Prefer not to disclose
Q89 What is your gender?
217
o Male
o Female
o Non-binary / third gender
o Prefer to self-describe: __________________________________________________
o Prefer not to disclose
Q90 What is your current marital status?
o Single, Never Married
o Married
o Separated
o Divorced
o Widowed
o Prefer not to disclose
Q91 How many people (including yourself) live in your household?
o 1
o 2
o 3
o 4
o 5 or more
o Prefer not to disclose
Q92 Are there children under the age of 12 living in your household?
o Yes
o No
o Prefer not to disclose
Q93 Have you ever received food stamps?
o Yes
o No
o Prefer not to disclose
Q94 Are you currently on food stamps?
o Yes
o No
o Prefer not to disclose
Q95 In what U.S. state do you live?
▼ Alabama ... Wyoming
Q96 What is your ZIP code? _____________________________________________________
Q97 Please select your U.S. citizenship status.
o U.S. Citizen
o Lawful Permanent Resident
o Temporary Resident (e.g., visitor, student)
o Prefer not to disclose
Q98 Were you born in the U.S.?
o Yes
o No, I was born in this country: ____________________________________________
o Prefer not to disclose
Q99 Were your parents born in the U.S.?
o Yes
o No, they were born in this country: __________________________________________
218
o Prefer not to disclose
Display This Question:
If Were you born in the U.S.? = No, I was born in this country:
Or Or Were you born in the U.S.? Text Response Is Not Empty
Q100 How long have you lived in the U.S.?
o 0-5 years
o 6-10 years
o 11-15 years
o 16-20 years
o 21-25 years
o 26-30 years
o 31-35 years
o More than 35 years
o Prefer not to disclose
Q101 What is the highest level of education you have completed?
o Less than High School
o High School/GED
o Some College
o 2-Year College Degree (Associates)
o 4-Year College Degree (BA, BS)
o Master's Degree
o Professional Degree (Ph.D., J.D., M.D., etc.)
o Prefer not to disclose
Q102 What is your approximate annual household income before taxes?
o Less than $20,000
o $20,000 - $39,999
o $40,000 - $59,999
o $60,000 - $79,999
o $80,000 - $99,999
o $100,000 - $119,999
o $120,000 - $139,999
o $140,000 - $159,999
o $160,000 or greater
o Prefer not to disclose
Q103 Which category best describes you? Select all that apply.
▢ White (for example, German, Irish, English, Italian, Polish, French, etc.)
▢ Hispanic, Latino or Spanish origin (for example, Mexican or Mexican American, Puerto
Rican, Cuban, Salvadoran, Dominican, Colombian, etc.)
▢ Black or African American (for example, African American, Jamaican, Haitian, Nigerian,
Ethiopian, Somalian, etc.)
▢ Middle Eastern or North African (for example, Lebanese, Iranian, Egyptian, Syrian,
Moroccan, Algerian, etc.)
▢ American Indian or Alaskan Native (for example, Navajo nation, Blackfeet tribe, Mayan,
Aztec, Native Village or Barrow Inupiat Traditional Government, Nome Eskimo
Community, etc.)
▢ Asian (for example, Chinese, Korean, Japanese, etc.)
219
▢ South or Southeast Asian (for example, Indian, Pakistani, Filipino, Vietnamese, Malaysian,
etc.)
▢ Native Hawaiian or Other Pacific Islander (for example, Native Hawaiian, Samoan,
Chamorro, Tongan, Fijian, etc.)
▢ Other. Please specify: __________________________________________________
▢ Prefer not to disclose
Q104 Which U.S. political party do you most identify with?
o Democratic
o Republican
o I am an independent
o Other. Please specify: __________________________________________________
o Prefer not to disclose
Q105 Which best describes the area in which you live?
o Rural
o Suburban
o Urban
o Prefer not to disclose
Q106 Do you have any final comments to share about your responses or this study?
________________________________________________________________
________________________________________________________________
220
APPENDIX A.3 BEST-WORST SCALING RESULTS AND
ADDITIONAL TABLES
The count data of the best-worst question responses is given in Figure 12, Figure 13, and
Figure 14. Figure 12 aggregates these responses from halal and non-halal processors, Figure 13
gives the responses from halal processors only, and Figure 14 gives the responses from non-halal
processors only. We exclude past halal processors from this analysis. Looking at the data in Figure
12, we may conclude that the meat and poultry processing industry overall is not strongly
concerned about what requirements for information and transparency across the supply chain may
be imposed by a national halal certification. Likewise, the benefits of certification and inspection
type and frequency appear to have low importance to the industry overall. Costs of certification,
which standards are included in the certification, and what establishments are required to carry a
certification are overall more important to the industry.
However, when disaggregated into halal and non-halal processors, the count data provides
a more nuanced insight into the market’s preferences. For halal processors, which standards are
included in a certification, what establishments are required to be certified, and who is in charge
of enforcing the certification are highly important, while the remaining four factors are relatively
unimportant to halal processors. For non-halal processors, the most important characteristics to
consider when designing a national halal certification program are costs, what is included in the
certification, and which establishments are required to be certified. The type of information and
transparency passed along the supply chain appears very unimportant, while the remaining three
factors are of relatively mild importance. When comparing halal and non-halal processors, I
conclude that halal processors are overall more interested in the rigor of the standards and
enforcement integrity of a potential national halal certification, while non-halal processors are
most concerned with the costs and how they may need to adjust their business to meet standards.
Figure 12: Best-Worst Count Data by Attribute of Potential National Halal Certification
Program, Current Halal and Non-Halal Processors Combined, n = 98
All Processors Best-Worst Count Data
Which standards are included
45%
21%
34%
What is required to be certified
Costs of certification
41%
38%
16%
22%
Who is in charge of enforcement
28%
32%
Benefits of certification
22%
Inspection frequency and type
22%
34%
33%
Type of information passed down the supply chain
10%
49%
43%
39%
40%
43%
46%
41%
% of time chosen as most important
% of time chosen as least important
% of time not chosen
0%
10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
221
Figure 13: Best-Worst Count Data by Attribute of Potential National Halal Certification
Program, Current Halal Processors, n = 32
Halal Processors Best-Worst Counts
Which standards are included
66%
8%
27%
What is required to be certified
48%
18%
Costs of certification
16%
50%
34%
34%
Who is in charge of enforcement
41%
27%
32%
Benefits of certification
17%
Inspection frequency and type
16%
38%
40%
Type of information passed down the supply chain
14%
37%
46%
44%
49%
% of time chosen as most important
% of time chosen as least important
% of time not chosen
0%
10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Figure 14: Best-Worst Count Data by Attribute of Potential National Halal Certification
Program, Non-Halal Processors, n = 66
Non-Halal Processors Best-Worst Counts
Which standards are included
What is required to be certified
39%
41%
28%
34%
16%
43%
Costs of certification
49%
12%
39%
Who is in charge of enforcement
24%
34%
Benefits of certification
28%
34%
Inspection frequency and type
25%
32%
Type of information passed down the supply chain
7%
56%
43%
38%
43%
37%
% of time chosen as most important
% of time chosen as least important
% of time not chosen
0%
10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
222
Table 40: Multinomial Logit Results of Best-Worst Scaling for Hypothetical U.S. National
Halal Meat and Poultry Certification Program: Processors (n = 82), Retailers (n = 33), and
Consumers (n = 507)
Best-Worst Scaling Attributes
Cost of Certification
Enforcement and Regulation
Information Collected/Available
Who/What Must be Certified
Which Halal Standards are Included
Benefits of Certification
Processors
Coefficient
(st. error)
0.46***
(0.11)
0.17
(0.11)
-0.41***
(0.11)
0.60***
(0.11)
0.55***
(0.11)
0.01
(0.11)
Retailers
Coefficient
(st. error)
0.39**
(0.17)
-0.14
(0.17)
-0.46***
(0.17)
0.62***
(0.17)
0.22
(0.17)
0.14
(0.17)
Consumers
Coefficient
(st. error)
-0.25***
(0.04)
-0.07
(0.04)
-0.16***
(0.04)
0.22***
(0.04)
0.37***
(0.04)
0.06
(0.04)
N
Log likelihood
AIC
AIC/N
Note: *** indicates p < 0.01, ** indicates p < 0.05, and * indicates p < 0.1. Base attribute normalized to zero is
Inspection Type/Frequency.
3549
-6202.82
12417.6
3.499
231
-388.49
789.0
3.415
574
-960.86
1933.7
3.369
223
Table 41: Best-Worst Scaling Shares of Preferences for Hypothetical U.S. National Halal Meat and Poultry Certification
Program: Consumers, Retailers, & Processors: Multinomial Logit
Processors (n = 82)
Share
0.18***
0.13***
0.07***
0.20***
0.11***
0.19***
0.11***
National Halal Meat & Poultry
Certification Program Attribute
Cost of Certification
Enforcement and Regulation
Information Collected/Available
Who/What Must be Certified
Inspection Frequency and Type
Which Halal Standards are Included
Benefits of Certification
Consumers (n = 507)
95% CI
Share
[0.10, 0.11]
0.11***
[0.12, 0.13]
0.13***
[0.11, 0.12]
0.12***
[0.16, 0.18]
0.17***
[0.13, 0.14]
0.14***
[0.19, 0.21]
0.20***
[0.14, 0.15]
0.14***
95% CI
[0.15, 0.20]
[0.11, 0.15]
[0.06, 0.09]
[0.18, 0.23]
[0.10, 0.13]
[0.17, 0.22]
[0.10, 0.13]
95% CI
[0.14, 0.22]
[0.08, 0.13]
[0.06, 0.10]
[0.18, 0.27]
[0.10, 0.15]
[0.12, 0.19]
[0.11, 0.17]
Share
0.18***
0.11***
0.08***
0.22***
0.12***
0.15***
0.14***
Retailers (n = 33)
Note: *** indicates p < 0.01, ** indicates p < 0.05, and * indicates p < 0.1.
Table 42: P-values from Poe tests for Hypothetical U.S. National Halal Meat and Poultry Certification Program: Pair-wise
Comparisons Between Processors (n = 82), Retailers (n = 33), and Consumers (507), Multinomial Logit
National Halal Meat & Poultry
Certification Program Attribute
Cost of Certification
Enforcement and Regulation
Information Collected/Available
Who/What Must be Certified
Inspection Frequency and Type
Which Halal Standards are Included
Benefits of Certification
Processors vs. Retailers
Processors vs. Consumers Retailers vs. Consumers
0.54
0.05
0.40
0.19
0.26
0.97
0.05
0.00
0.35
1.00
0.01
0.00
0.32
0.00
0.00
0.94
1.00
0.00
0.15
0.00
0.36
Note: Values that are statistically significant at the 5% level or better are bolded.
224
Table 43: Uncorrelated Random Parameters Logit Results of Best-Worst Scaling for
Hypothetical U.S. National Halal Meat and Poultry Certification Program: Processors (n =
82), Retailers (n = 33), and Consumers (n = 507)
Best-Worst Scaling Attributes
Cost of Certification
Enforcement and Regulation
Information Collected/Available
Who/What Must be Certified
Which Halal Standards are Included
Benefits of Certification
Processors
Coefficient
(st. error)
0.80***
(0.15)
0.25*
(0.14)
-0.74***
(0.14)
0.92***
(0.14)
0.95***
(0.14)
-0.04
(0.14)
Retailers
Coefficient
(st. error)
0.67***
(0.23)
-0.41*
(0.23)
-0.78***
(0.23)
0.92***
(0.22)
0.43*
(0.23)
0.16
(0.22)
Consumers
Coefficient
(st. error)
-0.34***
(0.05)
-0.10**
(0.05)
-0.20***
(0.05)
0.27***
(0.05)
0.49***
(0.05)
0.07
(0.04)
McFadden Pseudo R2
N
Log likelihood
AIC
AIC/N
Note: *** indicates p < 0.01, ** indicates p < 0.05, and * indicates p < 0.1. Base attribute normalized to zero is
Inspection Type/Frequency.
0.12
3549
-6095.59
12215.2
3.442
0.21
231
-353.08
730.2
3.161
0.21
574
-879.30
1782.6
3.106
225
Table 44: Best-Worst Scaling Shares of Preferences for Hypothetical U.S. National Halal Meat and Poultry Certification
Program: Consumers, Retailers, & Processors: Uncorrelated Random Parameters Logit
National Halal Meat & Poultry
Certification Program Attribute
Cost of Certification
Enforcement and Regulation
Information Collected/Available
Who/What Must be Certified
Inspection Frequency and Type
Which Halal Standards are Included
Benefits of Certification
Processors (n = 82)
Share
0.20
0.12
0.04
0.23
0.09
0.23
0.09
95% CI
[0.16, 0.24]
[0.09, 0.14]
[0.03, 0.05]
[0.19, 0.27]
[0.07, 0.11]
[0.19, 0.27]
[0.07, 0.11]
Retailers (n = 33)
Share
0.21
0.07
0.05
0.27
0.11
0.17
0.13
95% CI
[0.16, 0.27]
[0.05, 0.10]
[0.03, 0.07]
[0.20, 0.34]
[0.08, 0.14]
[0.12, 0.22]
[0.09, 0.17]
Consumers (n = 507)
95% CI
Share
[0.09, 0.10]
0.10
[0.11, 0.13]
0.12
[0.10, 0.12]
0.11
[0.16, 0.18]
0.17
[0.13, 0.14]
0.13
[0.21, 0.23]
0.22
[0.13, 0.15]
0.14
Note: *** indicates p < 0.01, ** indicates p < 0.05, and * indicates p < 0.1.
Table 45: P-values from Poe tests for Hypothetical U.S. National Halal Meat and Poultry Certification Program: Pair-wise
Comparisons Between Processors (n = 82), Retailers (n = 33), and Consumers (507), Uncorrelated Random Parameters Logit
National Halal Meat & Poultry
Certification Program Attribute
Cost of Certification
Enforcement and Regulation
Information Collected/Available
Who/What Must be Certified
Inspection Frequency and Type
Which Halal Standards are Included
Benefits of Certification
Processors vs. Retailers
Processors vs. Consumers Retailers vs. Consumers
0.57
0.01
0.30
0.16
0.16
0.97
0.03
0.00
0.68
1.00
0.00
0.00
0.72
0.00
0.00
1.00
1.00
0.00
0.06
0.03
0.19
Note: Values that are statistically significant at the 5% level or better are bolded.
226
Figure 15: All Output of Correlated Random Parameters Logit Results of Best-Worst
Scaling for Hypothetical U.S. National Halal Meat and Poultry Certification Program,
Consumers (n = 507)
Iterative procedure has converged
Normal exit: 38 iterations. Status=0, F= .6053561D+04
-----------------------------------------------------------------------------
Random Parameters Multinom. Logit Model
Dependent variable CHOICE
Log likelihood function -6053.56104
Restricted log likelihood -6906.03512
Chi squared [ 27](P= .000) 1704.94815
Significance level .00000
McFadden Pseudo R-squared .1234390
Estimation based on N = 3549, K = 27
Inf.Cr.AIC = 12161.1 AIC/N = 3.427
---------------------------------------
Log likelihood R-sqrd R2Adj
No coefficients -6906.0351 .1234 .1221
Constants only can be computed directly
Use NLOGIT ;...;RHS=ONE$
At start values -6202.8157 .0241 .0226
Note: R-sqrd = 1 - logL/Logl(constants)
Root Likelihood:Geom. Mean of P^ .1816
Warning: Model does not contain a full
set of ASCs. R-sqrd is problematic. Use
model setup with ;RHS=one to get LogL0.
---------------------------------------
Response data are given as ind. choices
Replications for simulated probs. =1000
Used Halton sequences in simulations.
RPL model with panel has 507 groups
Fixed number of obsrvs./group= 7
Number of obs.= 3549, skipped 0 obs
227
Figure 15 (cont’d)
--------+------------------------------------------------------------------
--
| Standard Prob. 95% Confidence
CHOICE| Coefficient Error z |z|>Z* Interval
--------+------------------------------------------------------------------
--
|Random parameters in utility functions..........................
A1| -.34640*** .04825 -7.18 .0000 -.44098 -.25183
A2| -.09280* .04750 -1.95 .0507 -.18589 .00030
A3| -.20008*** .04678 -4.28 .0000 -.29177 -.10839
A4| .28766*** .04682 6.14 .0000 .19591 .37942
A6| .51522*** .04916 10.48 .0000 .41887 .61157
A7| .06852 .04628 1.48 .1387 -.02219 .15924
|Diagonal values in Cholesky matrix, L...........................
NsA1| .88157*** .05371 16.41 .0000 .77630 .98683
NsA2| .88719*** .04802 18.47 .0000 .79307 .98131
NsA3| .59115*** .04163 14.20 .0000 .50955 .67274
NsA4| .60445*** .04064 14.87 .0000 .52480 .68410
NsA6| .58399*** .04228 13.81 .0000 .50112 .66685
NsA7| .36926*** .03704 9.97 .0000 .29665 .44186
|Below diagonal values in L matrix. V = L*Lt.....................
A2:A1| -.21233*** .04916 -4.32 .0000 -.30868 -.11598
A3:A1| -.33569*** .04866 -6.90 .0000 -.43106 -.24031
A3:A2| -.29119*** .04222 -6.90 .0000 -.37394 -.20844
A4:A1| -.08463* .04841 -1.75 .0804 -.17952 .01026
A4:A2| -.28091*** .04249 -6.61 .0000 -.36418 -.19764
A4:A3| .10671*** .03916 2.73 .0064 .02996 .18346
A6:A1| -.13966*** .04991 -2.80 .0051 -.23749 -.04183
A6:A2| -.32461*** .04452 -7.29 .0000 -.41186 -.23735
A6:A3| .21022*** .04121 5.10 .0000 .12946 .29098
A6:A4| .65385*** .04287 15.25 .0000 .56983 .73788
A7:A1| -.38867*** .04841 -8.03 .0000 -.48356 -.29378
A7:A2| -.11276*** .04168 -2.71 .0068 -.19445 -.03106
A7:A3| .31559*** .03918 8.06 .0000 .23880 .39238
A7:A4| .26119*** .03784 6.90 .0000 .18703 .33535
A7:A6| .08272** .03635 2.28 .0229 .01146 .15397
228
Figure 15 (cont’d)
|Standard deviations of parameter distributions..................
sdA1| .88157*** .05371 16.41 .0000 .77630 .98683
sdA2| .91224*** .04805 18.99 .0000 .81807 1.00641
sdA3| .73955*** .04251 17.40 .0000 .65623 .82287
sdA4| .68031*** .04087 16.65 .0000 .60020 .76041
sdA6| .96831*** .04050 23.91 .0000 .88894 1.04769
sdA7| .68905*** .04121 16.72 .0000 .60827 .76982
|Covariances of Random Parameters................................
A2:A1| -.18718*** .03935 -4.76 .0000 -.26430 -.11007
A3:A1| -.29593*** .03810 -7.77 .0000 -.37061 -.22125
A3:A2| -.18706*** .04227 -4.43 .0000 -.26991 -.10421
A4:A1| -.07461* .04097 -1.82 .0686 -.15491 .00569
A4:A2| -.23125*** .03825 -6.05 .0000 -.30623 -.15627
A4:A3| .17329*** .03888 4.46 .0000 .09708 .24950
A6:A1| -.12312*** .04109 -3.00 .0027 -.20367 -.04258
A6:A2| -.25833*** .04053 -6.37 .0000 -.33777 -.17889
A6:A3| .26567*** .04215 6.30 .0000 .18307 .34828
A6:A4| .52066*** .05060 10.29 .0000 .42148 .61983
A7:A1| -.34264*** .03733 -9.18 .0000 -.41580 -.26948
A7:A2| -.01751 .04338 -.40 .6865 -.10253 .06751
A7:A3| .34986*** .04672 7.49 .0000 .25830 .44143
A7:A4| .25612*** .04050 6.32 .0000 .17674 .33551
A7:A6| .37631*** .04928 7.64 .0000 .27972 .47290
--------+------------------------------------------------------------------
--
***, **, * ==> Significance at 1%, 5%, 10% level.
Model was estimated on Jan 01, 2024 at 09:24:34 PM
---------------------------------------------------------------------------
--
Correlation Matrix for Random Parameters
--------+-----------------------------------------------------
Cor.Mat.| A1 A2 A3 A4 A6 A7
--------+-----------------------------------------------------
A1| 1.00000 -.23276 -.45391 -.12440 -.14423 -.56407
A2| -.23276 1.00000 -.27727 -.37262 -.29245 -.02786
A3| -.45391 -.27727 1.00000 .34443 .37099 .68657
A4| -.12440 -.37262 .34443 1.00000 .79037 .54638
A6| -.14423 -.29245 .37099 .79037 1.00000 .56400
A7| -.56407 -.02786 .68657 .54638 .56400 1.00000
229
Figure 16: All Output of Correlated Random Parameters Logit Results of Best-Worst
Scaling for Hypothetical U.S. National Halal Meat and Poultry Certification Program,
Retailers (n = 33)
Iterative procedure has converged
Normal exit: 21 iterations. Status=0, F= .3425588D+03
---------------------------------------------------------------------------
--
Random Parameters Multinom. Logit Model
Dependent variable CHOICE
Log likelihood function -342.55881
Restricted log likelihood -449.50524
Chi squared [ 27](P= .000) 213.89288
Significance level .00000
McFadden Pseudo R-squared .2379203
Estimation based on N = 231, K = 27
Inf.Cr.AIC = 739.1 AIC/N = 3.200
---------------------------------------
Log likelihood R-sqrd R2Adj
No coefficients -449.5052 .2379 .2197
Constants only can be computed directly
Use NLOGIT ;...;RHS=ONE$
At start values -388.4877 .1182 .0971
Note: R-sqrd = 1 - logL/Logl(constants)
Root Likelihood:Geom. Mean of P^ .2270
Warning: Model does not contain a full
set of ASCs. R-sqrd is problematic. Use
model setup with ;RHS=one to get LogL0.
---------------------------------------
Response data are given as ind. choices
Replications for simulated probs. =1000
Used Halton sequences in simulations.
RPL model with panel has 33 groups
Fixed number of obsrvs./group= 7
Number of obs.= 231, skipped 0 obs
230
Figure 16 (cont’d)
--------+------------------------------------------------------------------
--
| Standard Prob. 95% Confidence
CHOICE| Coefficient Error z |z|>Z* Interval
--------+------------------------------------------------------------------
--
|Random parameters in utility functions..........................
A1| .91392*** .24643 3.71 .0002 .43093 1.39691
A2| -.17735 .23345 -.76 .4474 -.63491 .28021
A3| -.79397*** .24061 -3.30 .0010 -1.26556 -.32237
A4| 1.05049*** .23288 4.51 .0000 .59405 1.50692
A6| .35458 .23941 1.48 .1386 -.11467 .82382
A7| .39813* .23971 1.66 .0967 -.07168 .86795
|Diagonal values in Cholesky matrix, L...........................
NsA1| 1.85616*** .29287 6.34 .0000 1.28215 2.43017
NsA2| 1.73116*** .26483 6.54 .0000 1.21210 2.25022
NsA3| 1.08359*** .22861 4.74 .0000 .63552 1.53166
NsA4| .23369 .19074 1.23 .2205 -.14015 .60752
NsA6| .01543 .19983 .08 .9385 -.37623 .40709
NsA7| 1.18455*** .22404 5.29 .0000 .74543 1.62367
|Below diagonal values in L matrix. V = L*Lt.....................
A2:A1| .56945** .23733 2.40 .0164 .10430 1.03460
A3:A1| .31574 .23456 1.35 .1783 -.14399 .77547
A3:A2| .35253 .22054 1.60 .1099 -.07973 .78479
A4:A1| .71177*** .23914 2.98 .0029 .24307 1.18047
A4:A2| .28354 .20243 1.40 .1613 -.11321 .68029
A4:A3| -.09126 .18869 -.48 .6286 -.46109 .27858
A6:A1| -.30952 .25500 -1.21 .2248 -.80930 .19026
A6:A2| .95030*** .23396 4.06 .0000 .49174 1.40886
A6:A3| .63784*** .23022 2.77 .0056 .18662 1.08905
A6:A4| -1.66497*** .28356 -5.87 .0000 -2.22074 -1.10919
A7:A1| 1.32502*** .26996 4.91 .0000 .79592 1.85413
A7:A2| .33871 .21834 1.55 .1208 -.08923 .76665
A7:A3| -.66067*** .20993 -3.15 .0016 -1.07213 -.24922
A7:A4| .92612*** .23406 3.96 .0001 .46738 1.38486
A7:A6| -1.00791*** .21907 -4.60 .0000 -1.43727 -.57855
231
Figure 16 (cont’d)
|Standard deviations of parameter distributions..................
sdA1| 1.85616*** .29287 6.34 .0000 1.28215 2.43017
sdA2| 1.82241*** .26765 6.81 .0000 1.29784 2.34699
sdA3| 1.18243*** .22299 5.30 .0000 .74539 1.61948
sdA4| .80619*** .23139 3.48 .0005 .35268 1.25971
sdA6| 2.04403*** .23082 8.86 .0000 1.59164 2.49643
sdA7| 2.36297*** .20718 11.41 .0000 1.95690 2.76905
|Covariances of Random Parameters................................
A2:A1| 1.05699* .54743 1.93 .0535 -.01595 2.12993
A3:A1| .58607 .47567 1.23 .2179 -.34623 1.51837
A3:A2| .79009* .46112 1.71 .0866 -.11369 1.69387
A4:A1| 1.32116** .55949 2.36 .0182 .22458 2.41773
A4:A2| .89617* .47633 1.88 .0599 -.03742 1.82976
A4:A3| .22581 .30547 .74 .4598 -.37291 .82452
A6:A1| -.57452 .46464 -1.24 .2163 -1.48518 .33615
A6:A2| 1.46887*** .55642 2.64 .0083 .37831 2.55944
A6:A3| .92844** .40721 2.28 .0226 .13033 1.72656
A6:A4| -.39814 .46307 -.86 .3899 -1.30575 .50946
A7:A1| 2.45946*** .75575 3.25 .0011 .97822 3.94069
A7:A2| 1.34090** .60810 2.21 .0274 .14904 2.53275
A7:A3| -.17813 .44834 -.40 .6911 -1.05685 .70060
A7:A4| 1.31586** .54121 2.43 .0150 .25511 2.37661
A7:A6| -2.06715*** .78743 -2.63 .0087 -3.61048 -.52382
--------+------------------------------------------------------------------
--
***, **, * ==> Significance at 1%, 5%, 10% level.
Model was estimated on Jan 01, 2024 at 07:44:43 PM
---------------------------------------------------------------------------
--
Correlation Matrix for Random Parameters
--------+-----------------------------------------------------
Cor.Mat.| A1 A2 A3 A4 A6 A7
--------+-----------------------------------------------------
A1| 1.00000 .31247 .26703 .88288 -.15143 .56074
A2| .31247 1.00000 .36665 .60996 .39432 .31138
A3| .26703 .36665 1.00000 .23688 .38414 -.06375
A4| .88288 .60996 .23688 1.00000 -.24161 .69074
A6| -.15143 .39432 .38414 -.24161 1.00000 -.42798
A7| .56074 .31138 -.06375 .69074 -.42798 1.00000
232
Figure 17: All Output of Correlated Random Parameters Logit Results of Best-Worst
Scaling for Hypothetical U.S. National Halal Meat and Poultry Certification Program,
Processors (n = 96)
Iterative procedure has converged
Normal exit: 32 iterations. Status=0, F= .8606388D+03
---------------------------------------------------------------------------
--
Random Parameters Multinom. Logit Model
Dependent variable CHOICE
Log likelihood function -860.63878
Restricted log likelihood -1116.95243
Chi squared [ 27](P= .000) 512.62728
Significance level .00000
McFadden Pseudo R-squared .2294759
Estimation based on N = 574, K = 27
Inf.Cr.AIC = 1775.3 AIC/N = 3.093
---------------------------------------
Log likelihood R-sqrd R2Adj
No coefficients -1116.9524 .2295 .2222
Constants only can be computed directly
Use NLOGIT ;...;RHS=ONE$
At start values -960.8607 .1043 .0958
Note: R-sqrd = 1 - logL/Logl(constants)
Root Likelihood:Geom. Mean of P^ .2233
Warning: Model does not contain a full
set of ASCs. R-sqrd is problematic. Use
model setup with ;RHS=one to get LogL0.
---------------------------------------
Response data are given as ind. choices
Replications for simulated probs. =1000
Used Halton sequences in simulations.
RPL model with panel has 82 groups
Fixed number of obsrvs./group= 7
Number of obs.= 574, skipped 0 obs
233
Figure 17 (cont’d)
--------+------------------------------------------------------------------
--
| Standard Prob. 95% Confidence
CHOICE| Coefficient Error z |z|>Z* Interval
--------+------------------------------------------------------------------
--
|Random parameters in utility functions..........................
A1| .91671*** .15806 5.80 .0000 .60691 1.22651
A2| .28945** .14281 2.03 .0427 .00955 .56935
A3| -.81308*** .15676 -5.19 .0000 -1.12032 -.50584
A4| .98937*** .14577 6.79 .0000 .70366 1.27509
A6| .97008*** .15223 6.37 .0000 .67172 1.26845
A7| -.04691 .14636 -.32 .7486 -.33378 .23995
|Diagonal values in Cholesky matrix, L...........................
NsA1| 2.09648*** .21355 9.82 .0000 1.67792 2.51503
NsA2| 1.31529*** .15257 8.62 .0000 1.01627 1.61432
NsA3| 1.46573*** .15789 9.28 .0000 1.15628 1.77518
NsA4| .82066*** .12589 6.52 .0000 .57392 1.06740
NsA6| 1.31810*** .15276 8.63 .0000 1.01871 1.61750
NsA7| 1.39940*** .15315 9.14 .0000 1.09923 1.69958
|Below diagonal values in L matrix. V = L*Lt.....................
A2:A1| .53761*** .15864 3.39 .0007 .22669 .84854
A3:A1| -.18392 .15577 -1.18 .2377 -.48922 .12138
A3:A2| .34598** .13530 2.56 .0106 .08081 .61115
A4:A1| -.28950* .15286 -1.89 .0582 -.58911 .01011
A4:A2| .04337 .12925 .34 .7372 -.20996 .29670
A4:A3| .31768*** .12000 2.65 .0081 .08249 .55287
A6:A1| .45964*** .16344 2.81 .0049 .13931 .77997
A6:A2| .32291** .13953 2.31 .0206 .04945 .59638
A6:A3| .37361*** .12727 2.94 .0033 .12415 .62306
A6:A4| -.70761*** .13096 -5.40 .0000 -.96430 -.45093
A7:A1| -.74733*** .16135 -4.63 .0000 -1.06356 -.43110
A7:A2| .59629*** .13425 4.44 .0000 .33316 .85941
A7:A3| .83063*** .13309 6.24 .0000 .56977 1.09149
A7:A4| -.16944 .12007 -1.41 .1582 -.40477 .06589
A7:A6| .24544** .12172 2.02 .0438 .00688 .48401
234
Figure 17 (cont’d)
|Standard deviations of parameter distributions..................
sdA1| 2.09648*** .21355 9.82 .0000 1.67792 2.51503
sdA2| 1.42092*** .16430 8.65 .0000 1.09890 1.74295
sdA3| 1.51720*** .15719 9.65 .0000 1.20912 1.82528
sdA4| .92742*** .12721 7.29 .0000 .67810 1.17674
sdA6| 1.64111*** .14855 11.05 .0000 1.34996 1.93225
sdA7| 1.91084*** .13770 13.88 .0000 1.64094 2.18073
|Covariances of Random Parameters................................
A2:A1| 1.12709*** .38691 2.91 .0036 .36875 1.88543
A3:A1| -.38558 .32129 -1.20 .2301 -1.01529 .24414
A3:A2| .35619* .21533 1.65 .0981 -.06585 .77823
A4:A1| -.60693* .31371 -1.93 .0530 -1.22178 .00793
A4:A2| -.09859 .19115 -.52 .6060 -.47324 .27605
A4:A3| .53389** .21289 2.51 .0121 .11663 .95114
A6:A1| .96363** .38080 2.53 .0114 .21729 1.70998
A6:A2| .67184** .26875 2.50 .0124 .14509 1.19858
A6:A3| .57479** .23677 2.43 .0152 .11073 1.03885
A6:A4| -.58109*** .17591 -3.30 .0010 -.92586 -.23631
A7:A1| -1.56676*** .31778 -4.93 .0000 -2.18960 -.94393
A7:A2| .38252 .24389 1.57 .1168 -.09549 .86053
A7:A3| 1.56123*** .33940 4.60 .0000 .89603 2.22643
A7:A4| .36703 .22372 1.64 .1009 -.07145 .80552
A7:A6| .60279** .27302 2.21 .0273 .06768 1.13789
--------+------------------------------------------------------------------
--
***, **, * ==> Significance at 1%, 5%, 10% level.
Model was estimated on Jan 01, 2024 at 08:01:49 PM
---------------------------------------------------------------------------
--
Correlation Matrix for Random Parameters
--------+-----------------------------------------------------
Cor.Mat.| A1 A2 A3 A4 A6 A7
--------+-----------------------------------------------------
A1| 1.00000 .37835 -.12122 -.31216 .28008 -.39110
A2| .37835 1.00000 .16522 -.07482 .28811 .14088
A3| -.12122 .16522 1.00000 .37943 .23085 .53852
A4| -.31216 -.07482 .37943 1.00000 -.38179 .20711
A6| .28008 .28811 .23085 -.38179 1.00000 .19222
A7| -.39110 .14088 .53852 .20711 .19222 1.00000
235
APPENDIX A.4 FARM ANIMAL WELFARE REGULATIONS
ADDITIONAL TABLES
Table 46: Stage 1 Legislative Action Decision Outcomes Coefficients
VARIABLES
HENS_PER_1000
HOGS_PER_1000
COUNT_PASSED_PREV
PREV_LAW
ALLOW_BALLOT
TRIFECTA_D
TRIFECTA_R
HOUSE%D
SENATE%D
CONSTANT
(1)
Ballot Outcome MNL
Coefficients
<0.001
(<0.001)
-0.054**
(0.025)
0.049
(0.043)
-0.724
(0.963)
2.965***
(1.086)
-0.622
(0.580)
-1.268
(1.165)
0.019
(0.041)
-0.005
(0.039)
-6.740***
(1.645)
980
Observations
Note: Robust standard errors in parentheses. *** p<0.01, ** p<0.05, * p<0.1.
Table 47: Out of Sample Stage 1 Accurate Prediction Percentage
(2)
Bill Outcome MNL
Coefficients
<-0.001
(<0.001)
-0.008**
(0.003)
0.093***
(0.034)
0.528
(0.468)
0.367
(0.428)
0.893*
(0.505)
-1.719*
(0.888)
0.044
(0.032)
-0.013
(0.026)
-6.106***
(1.077)
Predicted Outcome Accuracy
Mean
0.931
Note: We used clusters of 45 states in the Stage 1 bootstrap; the remaining 5 states are used for out of sample
predictions for each repetition. An accurate prediction is when the real-world outcome matches the overall prediction
of the model. As there are three possible outcomes (no action, ballot proposed, bill proposed), the predicted probability
for any given outcome must be larger than the remaining two outcomes for it to be considered the model’s overall
prediction.
Standard Deviation
0.218
236
Figure 18: K-Density Plots: Ballot Model Residuals (top), Ballot Model R-Squared
(middle), and Ballot Model Inverse Mills Ratios (bottom)
K-Density Ballot Model Residuals
y
t
i
s
n
e
D
5
.
1
1
5
.
0
0
4
0
3
y
t
i
s
n
e
D
0
2
0
1
0
8
0
.
6
0
.
y
t
i
s
n
e
D
4
0
.
2
0
.
0
-4
-2
0
Residuals
kernel = epanechnikov, bandwidth = 0.0262
2
4
K-Density Ballot Model R-squared
.3
.4
.5
R-Squared
.6
.7
kernel = epanechnikov, bandwidth = 0.0007
K-Density Ballot Model Inverse Mills Ratio
-200
-150
-100
Ballot Inverse Mills Ratio
-50
0
kernel = epanechnikov, bandwidth = 0.2838
237
Table 48: Ballot IMR by Year
Year
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
Mean
-12.726
-12.693
-12.679
-12.646
-12.598
-12.570
-12.297
-12.261
-12.006
-11.876
-12.081
-12.503
-12.163
-12.237
-12.424
-12.163
-12.053
-12.042
-11.931
-11.751
Standard Deviation
2.826
2.835
2.695
2.720
2.788
2.788
2.766
2.570
2.566
2.620
2.630
2.756
2.767
2.789
2.719
2.546
2.804
2.886
2.843
2.525
Table 49: Coefficients with of Percentage of “Yes” Votes as Independent Variable
Variables
PEOPLE_PER_FARM
%DEMOCRAT
%WHITE
%BLACK
%HISPANIC
EDUCATION
POVERTY_RATE
HOUSEHOLD_INCOME_1000
%CATHOLIC
%EVANGELICAL PROTESTANT
%MAINLINE PROTESTANT
BALLOT_IMR
CONSTANT
Observations
Note: Coefficients in (1) are obtained from cluster bootstrapping. *** p<0.01, ** p<0.05, * p<0.1. To obtain the
coefficients in terms of the percentage of “yes” votes from our adjusted model, we must exponentiate to remove the
natural log and rearrange terms. The rearranged equation is:
(2)
Transformed Coefficients
0.500***
0.504***
0.502***
0.500***
0.502***
-0.498***
-0.494***
0.501***
-0.499***
-0.491***
0.507***
0.527***
0.657***
(1)
OLS Coefficients
0.000***
0.014***
0.008***
0.001***
0.008***
-0.009***
-0.023***
0.002***
-0.003***
-0.034***
0.030***
0.110***
0.650***
299
𝑉) =
𝑒𝑥𝑝[(𝐼𝑀𝑅 × ϕ) + (𝐵* × ζ))]
1 + 𝑒𝑥𝑝[(𝐼𝑀𝑅 × ϕ) + (𝐵* × ζ))]
Equation 23
238
Table 50: Ballot Model Without MESR and IMRs
VARIABLES
PEOPLE_PER_FARM
%DEMOCRAT
%WHITE
%BLACK
%HISPANIC
EDUCATION
POVERTY_RATE
HOUSEHOLD_INCOME_1000
%CATHOLIC
%EVANGELICAL PROTESTANT
%MAINLINE PROTESTANT
CONSTANT
Unadjusted (no IMR)
Ballot Model Coefficients
<0.001
(<0.001)
0.014**
(0.006)
0.009**
(0.004)
-0.006
(0.009)
0.008
(0.005)
-0.006
(0.007)
-0.008
(0.014)
0.005
(0.003)
0.004
(0.004)
-0.037**
(0.016)
0.029***
(0.010)
-0.881
(0.622)
Observations
R-squared
Note: Robust standard errors are in parentheses, *** p<0.01, ** p<0.05, * p<0.1. All variables are weighted by county
population as a proportion of the total state population.
299
0.311
239
Table 51: Comparing Real World Outcomes to Ballot Model Predictions
State
Alabama
Alaska*
Arizona*
Arkansas*
California*
Colorado*
Connecticut
Delaware
Florida*
Georgia
Hawaii
Idaho*
Illinois*
Indiana
Iowa
Kansas
KentuckyR
Louisiana
Maine*R
Maryland
Massachusetts*
Michigan*
Minnesota
Mississippi*
Missouri*
Montana*
Nebraska*
Nevada*
New Hampshire
New JerseyV
New Mexico
New York
North Carolina
North Dakota*
Ohio*
Oklahoma*
Oregon*
Reality
No regulation in place
No regulation in place
Regulation in place
No regulation in place
Regulation in place
Regulation in place
No regulation in place
No regulation in place
Regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
Regulation in place
No regulation in place
Regulation in place
No regulation in place
Regulation in place
Regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
Regulation in place
No regulation in place
Regulation in place
Ballot Model
Fail
Pass
Fail
Fail
Pass
Fail
Pass
Pass
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Pass
Pass
Pass
Pass
Fail
Fail
Fail
Fail
Fail
Pass
Pass
Fail
Fail
Fail
Fail
Fail
Fail
Fail
Pass
240
Table 46 (cont’d)
State
Ballot Model
Fail
Pennsylvania
Fail
Rhode Island
Fail
South Carolina
Fail
South Dakota*
Fail
Tennessee
Fail
Texas
Fail
Utah*
Pass
Vermont
Fail
Virginia
Pass
Washington*
Fail
West Virginia
Fail
Wisconsin
Fail
Wyoming*
26%
All 50 States
30.4%
Ballot States
74%
Accuracy All 50 States
Accuracy Ballot States
75%
Note: States with a (*) are states that allow ballot initiatives. (V) indicates that a FAW law passes by the legislature
was vetoed by the governor and is therefore not in effect in reality. Shaded cells correspond to predictions that match
what exist in reality. The superscript (R) indicates a state that has a FAW regulation in place that was not passed
through a bill or ballot.
Reality
No regulation in place
Regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
No regulation in place
Regulation in place
No regulation in place
No regulation in place
No regulation in place
24%
41.7%
N/A
N/A
241
APPENDIX A.5 FARM ANIMAL WELFARE DATA
COLLECTION METHODS
Stage 1:
State legislature political party information was collected from state legislature websites.
In some instances, these websites were incomplete, and these gaps were filled in by the lead author
using a range of sources, including Ballotpedia.org, state and local newspapers, and political party
websites. Industry size information for egg-laying hens was downloaded from USDA QuickStats
and industry size information for gestating sows was downloaded from the Livestock Marketing
Information Center (LMIC).
Stage 2:
for
the
file
The
2006
Records
Votes for State Initiative
Arizona 2006, Proposition 204: Votes on Proposition 204 came from the Arizona’s Secretary of
State
Election
General
(https://apps.azsos.gov/election/2006/General/ElectionInformation.htm). The
“Official
Election Results (PDF)” was downloaded on February 8, 2019. Data regarding the voting outcome
of Proposition 204, by county, is found on page 15 of the document.
California 2008, Proposition 2: Votes on Proposition 2 came from California’s Statement of Vote
from the 2008 General Election ( https://www.sos.ca.gov/elections/prior-elections/statewide-
file
election-results/presidential-general-election-november-4-2008/statement-vote/).
“Complete Statement of Vote (PDF)” was downloaded on January 11, 2019. Data regarding the
voting outcome of Proposition 2, by county, is found on page 57 of the document.
California 2018, Proposition 12: Votes on Proposition 12 came from California’s Statement of
Vote from the November 6, 2018 General Election (https://www.sos.ca.gov/elections/prior-
elections/statewide-election-results/general-election-november-6-2018/statement-vote/). The file
“Complete Statement of Vote (PDF)” was downloaded on February 10, 2019. Data regarding the
voting outcome of Proposition 12, by county, is found on page 98 of the document.
Florida 2002, Amendment 10: Votes on Amendment 10 came from Florida’s Department of State
2002
results
(https://results.elections.myflorida.com/downloadresults.asp?ElectionDate=11/5/2002&DATAM
ODE=).
Massachusetts 2016, Question 3: Votes on Question 3 came from Massachusetts’ Secretary of
the
results
General
(http://electionstats.state.ma.us/ballot_questions/view/2741/).
Ohio 2009, Issue 2: Votes on Issue 2 came from the Ohio’s Statement of Vote from the November
3, 2009 General Election (https://www.sos.state.oh.us/elections/election-results-and-data/2009-
election-results/state-issue-2-november-3-2009/). The file “State Issue 2 Official Results:
November 3, 2009” was downloaded on February 10, 2019.
Commonwealth
Elections
Election
General
2016
Votes for State Legislation
California 2010: The
https://leginfo.legislature.ca.gov/faces/billVotesClient.xhtml?bill_id=200920100AB1437%20
Colorado 2008: The
http://www.leg.state.co.us/clics/cslFrontPages.nsf/PrevSessionInfo?OpenForm
legislators’ votes on
legislators’ votes on
record of
record of
this bill
this bill
came
came
from
from
242
came
came
came
this bill
this bill
this bill
record of
record of
record of
record of
2007: The
legislators’ votes on
legislators’ votes on
legislators’ votes on
legislators’ votes on
from
Michigan 2009: The
http://www.legislature.mi.gov/(S(45zoivtdskmlve2f2f03jh3a))/mileg.aspx?page=getObject&obje
ctName=2009-HB-5127
Michigan 2018: The
from
http://www.legislature.mi.gov/(S(rvjeyosf413g1xjdeb3o32sh))/mileg.aspx?page=GetObject&obj
ectname=2017-SB-0660
Michigan 2019: The
from
http://www.legislature.mi.gov/(S(trsczgqyfjcjfrm3ho5b0gni))/mileg.aspx?page=getObject&objec
tName=2019-SB-0174
New Jersey 2013: The
https://www.njleg.state.nj.us/bills/BillsByNumber.asp%20
this
of
record
Oregon
https://olis.oregonlegislature.gov/liz/2007R1/Measures/Overview/SB600
Oregon
this
of
https://olis.oregonlegislature.gov/liz/2011R1/Measures/Overview/SB805
Oregon
this
of
https://olis.oregonlegislature.gov/liz/2019R1/Measures/Overview/SB1019
Rhode Island 2012: The
from
http://webserver.rilin.state.ri.us/search/search.idq?CiRestriction=SB+2191&CiMaxRecordsPerPa
ge=25&CiScope=%2FJournals12%2F&CiSort=DocTitle%5Ba%5D&HTMLQueryForm=%2Fse
arch%2Fsearch%2Easp&Abstractt=1
Rhode Island 2018: The
http://webserver.rilin.state.ri.us/journals18/senatejournals18/senatejournals18.html
Washington 2011: The
https://app.leg.wa.gov/billsummary?BillNumber=5487&Year=2011&Initiative=false
Washington 2019: The
https://app.leg.wa.gov/billsummary?BillNumber=2049&Year=2019&Initiative=false
legislators’ votes on
legislators’ votes on
legislators’ votes on
legislators’ votes on
this bill came
this bill came
this bill came
this bill came
this bill came
2011: The
2018: The
record of
record of
legislators’
legislators’
legislators’
record of
record of
record
record
votes
votes
votes
came
came
came
from
from
from
from
from
from
from
bill
bill
bill
on
on
on
all
data
Demographic Information
File Download website
For
(https://www.census.gov/support/USACdataDownloads.html), all reference codes are found in the
file, Mastdata.xls, under “Reference Information Files”. The data available closest to the vote in
question was used for both counties and legislative districts.
the Counties Data
retrieved
from
Vote for Democratic Presidential Candidate
Arizona 2006, Proposition 204: Votes for the Democratic ticket, Kerry-Edwards, in the 2004
Presidential election came from Arizona’s Secretary of State Record for the 2004 General Election
(https://apps.azsos.gov/election/2004/Info/ElectionInformation.htm).
California 2008, Proposition 2: Votes for the Democratic ticket, Obama-Biden, in the 2008
Presidential Election came from the California’s Statement of Vote Records for the 2008 General
(https://www.sos.ca.gov/elections/prior-elections/statewide-%20election-
Election
results/presidential-general-election-november-4-2008/statement-vote/).
California 2018, Proposition 12: Votes for the Democratic ticket, Clinton-Kaine, in the 2016
Presidential election came from the California’s Statement of Vote Records for the 2016 General
Election (https://elections.cdn.sos.ca.gov/sov/2016-general/sov/2016-%20complete-sov.pdf).
243
6,
of
on
the
2012
Votes
General
November
Florida 2002, Amendment 3: Votes for the Democratic ticket, Clinton-Gore, in the 1996
Presidential election came from Florida’s Department of State records for the November 5, 1996
Abstract
Election
(https://results.elections.myflorida.com/Index.aspElectionDate=11/5/1996&DATAMODE=).
*Note: The 1996 Presidential election was used opposed to the 2000 Presidential election due to
the controversial results in Florida in 2000.
Massachusetts 2016, Question 3: Votes for the Democratic ticket, Clinton-Kaine, in the 2016
Presidential election came from the Massachusetts’ Secretary of the Commonwealth results for the
2016 General Election
(https://elections.cdn.sos.ca.gov/sov/2016-%20general/sov/2016-
complete-sov.pdf).
Ohio 2009, Issue 2: Votes for the Democratic ticket, Obama-Biden, in the 2008 Presidential
election came from the Ohio’s Secretary of State Results for the November 4, 2008 General
Election
(https://www.sos.state.oh.us/elections/election-resuls-and-%20data/2008-election-
results/).
Median Household Income
Data regarding median household income is found on the “Small Area Income and Poverty
Estimates
the US Census Bureau website
on
(https://www.census.gov/programs-surveys/saipe.html).
Program”
(SAIPE)
page
Percent of People all Ages in Poverty
Data regarding poverty is found on the “Small Area Income and Poverty Estimates (SAIPE)
the US Census Bureau website (https://www.census.gov/programs-
Program” page on
surveys/saipe.html).
Persons 25+ years of age with a Bachelor’s degree or higher
Education data
(https://www.census.gov/support/USACdataDownloads.html).
found on
is
the USA Counties Data File Download website
Race – White, Black, Hispanic
Race
on
is
(https://www.census.gov/support/USACdataDownloads.html).
found
data
the USA Counties Data File Download website
Religious Data – Catholic, Mainline Protestant, Evangelical Protestant
Religious
of Religious Data Archives
the Association
(http://www.thearda.com/QL2010/). Collect data from the “Percent” column for the 2010 religious
census.
came
from
data
People per Farm
Information on farm numbers per county was downloaded from USDA QuickStats.
244
APPENDIX A.6 CALCULATION OF FARM ANIMAL WELFARE
INDUSTRY COSTS
Eggs
The calculation of the estimated cost to the egg industry of updating to cage-free egg
production in states that do not currently have a FAW regulation in place utilized data from the
United Egg Producers (United Egg Producers, 2022), USDA Quick Stats, and Matthews and
Sumner (2015). Data from 2017 (the most recent available) on the inventory of egg-laying hens in
each state was downloaded from USDA Quick Stats. In five states – Arizona, Connecticut,
Delaware, Kansas, and Maine – the number of egg-laying hens was not provided. For these five
states, we took the number of egg-laying hens in the nation overall, subtracted the total known
from the 45 states that reported inventory numbers, and then divided the unaccounted for inventory
evenly between these five states. According to the United Egg Producers, at the end of 2020, 28%
of all egg-laying hens were in cage-free systems. We removed the percentage accounted for by the
seven states that have a cage-free regulation – California, Colorado, Massachusetts, Michigan,
Oregon, Rhode Island, and Washington. We assumed that 100% of the egg-laying hens in these
states were in cage-free systems. We then calculated the number of egg-laying hens in cage-free
systems in the remaining 43 states without cage-free regulations and computed the new percentage
of egg-laying hens in cage-free systems. This updated estimate is 18.32% of egg-laying hens,
meaning 81.68% of all egg-laying hens are in conventional housing in the 43 states without a cage-
free regulation. This percentage of conventionally housed hens was multiplied by the total
inventory in a given state to calculate the number of egg-laying hens in conventional housing.
To estimate the cost of upgrading to cage-free production for these 81.68% of hens, we
used estimated changes in cost of producing a dozen eggs under conventional versus cage-free
systems from Matthews and Sumner (2015), scaled to the average number of eggs produced per
hen in 2020 and inflated to 2022 dollars. According to the United Egg Producers (United Egg
Producers, 2022), on average a hen laid 296 eggs in 2020. Converting this number to dozens of
eggs, we have the estimated cost of upgrading from conventional to cage-free production per egg-
laying hen, which is $6.95 per hen per year in 2022 dollars. We present the predicted percentage
of the population within the 31 states not predicted by our model to be likely to pass a FAW
regulation, the number of egg-laying hens in those states, the relative size of the states’ industry to
the national total, and the estimated costs to the industry in each of these states to update to cage-
free egg production methods Table 52.
Pork
The calculation of the estimated cost to the pork industry of updating to crate-free pork
production in states that do not currently have a FAW regulation in place utilized data from World
Animal Protection (World Animal Protection, 2021), the Livestock Marketing Information Center
(LMIC), Purdue University’s Center for Commercial Agriculture (Langemeier, 2019), and Ortega
and Wolf (2018). Data from 2019 (corresponding to the year for the data used in our predictions)
on the inventory of gestating sows in each state was downloaded from LMIC.
According to World Animal Protection’s “Quit Stalling” report on crate-free pork
production (World Animal Protection, 2021), at the end of 2020, about 25% of all gestating sows
were in crate-free systems. We removed the percentage accounted for by the eleven states that
have a cage-free regulation – Arizona, California, Colorado, Florida, Maine, Massachusetts,
Michigan, Ohio, Oregon, Rhode Island, and Washington. We assumed that 100% of the gestating
245
sows in these states were in crate-free systems. We then calculated the number of gestating sows
in crate-free systems in the remaining 39 states without crate-free regulations and computed the
new percentage of gestating sows in crate-free systems. This updated estimate is 18.67% of
gestating sows, meaning 81.34% of all gestating sows are in conventional housing in the 39 states
without a crate-free regulation. This percentage of conventionally housed gestating sows was
multiplied by the total inventory in a given state to calculate the number of gestating sows in
conventional housing.
To estimate the cost of upgrading to crate-free production for these 81.34% of sows, we
used estimated changes in cost per weaned pig under conventional versus two types of crate-free
systems from Ortega and Wolf (2018), scaled to the average number of pigs produced per sow in
2019 and inflated to 2022 dollars. According to the Purdue University Center for Commercial
Agriculture (Langemeier, 2019), in 2019, on average a sow had two litters of piglets per year, with
an average litter size of 11, for an estimated total of 22 piglets per sow per year. Converting to cost
per sow, we have the estimated cost of upgrading from conventional to crate-free production per
sow, which is between $51.53 and $87.77 per sow per year in 2022 dollars. We present the
predicted percentage of the population within the 31 states not predicted by our model to be likely
to pass a FAW regulation, the number of gestating sows in those states, the relative size of the
states’ industry to the national total, and the estimated costs to the industry in each of these states
to update to crate-free pork production methods in Table 53.
246
Table 52: Cost to Update to Cage-Free Egg Production in 31 States not Predicted to Pass a
FAW Regulation
State
Alabama
Arkansas*
Georgia
Hawaii
Idaho*
Illinois*
Indiana
Iowa
Kansas
Louisiana
Minnesota
Mississippi*
Missouri*
Montana*
Nebraska*
New Jersey
New Mexico
New York
North Carolina
North Dakota*
Oklahoma*
Pennsylvania
South Carolina
South Dakota*
Tennessee
Texas
Utah*
Virginia
West Virginia
Wisconsin
Wyoming*
Percent of
Population in
Favor of FAW
Regulation
21.451
34.367
35.904
38.745
42.278
46.797
23.841
0.171
5.630
34.514
11.274
20.817
11.958
39.778
N/A
55.172
38.939
54.224
10.165
12.272
0.345
61.928
25.678
0.011
20.968
34.773
52.484
51.360
39.490
48.259
5.369
Number of
Egg-Laying
Hens
Percent of Egg-
Laying Hens in
the Nation
Cost to Update to
Cage-Free
System
7,867,738
12,285,533
17,966,521
192,185
472,192
5,470,158
26,354,377
56,554,774
3,249,703
1,970,896
10,849,607
5,828,262
11,306,386
931,006
7,353,761
1,631,775
102,020
6,058,141
14,160,452
81,364
3,354,460
26,317,523
4,002,121
2,708,331
1,986,321
21,006,254
4,480,850
2,447,718
1,215,655
7,639,627
29,550
2.137
3.336
4.879
0.052
0.128
1.485
7.157
15.358
0.882
0.535
2.946
1.583
3.070
0.253
1.997
0.443
0.028
1.645
3.845
0.022
0.911
7.147
1.087
0.735
0.539
5.704
1.217
0.665
0.330
2.075
0.008
$44,682,163.73
$69,771,540.05
$102,034,794.87
$1,091,449.87
$2,681,655.17
$31,065,916.96
$149,670,793.32
$321,183,759.74
$18,455,593.01
$11,193,038.93
$61,616,682.76
$33,099,647.82
$64,210,804.99
$5,287,334.49
$41,763,204.75
$9,267,115.62
$579,388.17
$34,405,168.05
$80,419,510.00
$462,079.39
$19,050,523.92
$149,461,493.46
$22,728,696.07
$15,381,052.24
$11,280,640.02
$119,297,932.97
$25,447,475.93
$13,900,988.62
$6,903,902.46
$43,386,684.26
$167,819.26
$1,509,948,850.89
265,875,261
Total
Note: States with a (*) are states that allow ballot initiatives. The predicted percent of the population in favor of a
FAW regulation in Nebraska could not be estimated since Nebraska was left out of the Stage 1 model due to its
unicameral state legislature.
72.20
N/A
247
Table 53: Cost to Update to Crate-Free Pork Production in 31 States not Predicted to Pass
a FAW Regulation
Percent of
Population
in Favor of
FAW
Regulation
Number
of
Gestating
Sows
21.451
34.367
35.904
38.745
42.278
46.797
23.841
0.171
5.630
34.514
11.274
20.817
11.958
39.778
N/A
55.172
38.939
54.224
10.165
12.272
0.345
61.928
25.678
0.011
20.968
34.773
52.484
51.360
39.490
48.259
5.369
15,312
57,816
21,114
2,225
8,238
464,442
280,559
917,567
174,810
2,018
572,545
47,797
334,240
20,933
391,551
685
318
10,923
896,231
35,147
425,387
103,064
9,195
167,015
15,466
83,017
16,842
8,460
1,362
43,716
16,842
Percent
of
Gestating
Sows in
the
Nation
0.271
1.025
0.374
0.039
0.146
8.232
4.973
16.263
3.098
0.036
10.148
0.847
5.924
0.371
6.940
0.012
0.006
0.194
15.885
0.623
7.540
1.827
0.163
2.960
0.274
1.471
0.299
0.150
0.024
0.775
0.299
Cost to Update
to Crate-Free
System Lower
Bound
Cost to Update
to Crate-Free
System Upper
Bound
$1,141,393.00
$670,175.71
$6,634,346.82
$3,895,396.30
$2,568,134.25
$1,507,895.34
$142,674.13
$83,771.96
$706,236.92
$414,671.22
$79,540,824.82
$46,702,869.62
$34,598,475.37
$20,314,701.14
$88,798,281.25 $151,234,572.76
$25,253,320.17
$14,827,637.53
$28,534.83
$16,754.39
$85,247,789.83
$50,053,748.16
$6,348,998.57
$3,727,852.37
$72,193,107.37
$42,388,613.50
$4,565,572.01
$2,680,702.83
$54,858,201.16
$32,210,319.95
$57,069.65
$33,508.79
$28,534.83
$16,754.39
$370,952.73
$217,807.10
$78,955,075.55 $134,470,363.04
$5,207,605.57
$3,057,676.67
$62,063,244.48
$36,440,804.10
$15,480,142.59
$9,089,258.03
$1,319,735.66
$774,890.66
$36,310,564.88
$21,319,964.70
$3,887,869.91
$2,282,786.00
$19,261,006.91
$11,309,215.07
$11,984,626.52
$7,036,844.93
$477,958.32
$280,636.08
$49,935.94
$29,320.19
$7,062,369.20
$4,146,712.19
$5,171,937.04
$3,036,733.67
State
Alabama
Arkansas*
Georgia
Hawaii
Idaho*
Illinois*
Indiana
Iowa
Kansas
Louisiana
Minnesota
Mississippi*
Missouri*
Montana*
Nebraska*
New Jersey
New Mexico
New York
North Carolina
North Dakota*
Oklahoma*
Pennsylvania
South Carolina
South Dakota*
Tennessee
Texas
Utah*
Virginia
West Virginia
Wisconsin
Wyoming*
Total
Note: States with a (*) are states that allow ballot initiatives. The predicted percent of the population in favor of a
FAW regulation in Nebraska could not be estimated since Nebraska was left out of the Stage 1 model due to its
unicameral state legislature.
$486,321,379.40 $828,266,099.29
5,144,837
91.19
N/A
248