Similarities and Differences of Animal Welfare Perceptions between U.S. Cow-Calf
Producers and the Public
Melissa G.S. McKendree, Kansas State University, [email protected]
Glynn T. Tonsor, Kansas State University, [email protected]
Christopher A. Wolf, Michigan State University, [email protected]
Selected Paper prepared for presentation for the 2015 Agricultural & Applied Economics
Association and Western Agricultural Economics Association Annual Meeting, San Francisco,
CA, July 26-28.
Copyright 2015 by Melissa G.S. McKendree, Glynn T. Tonsor, and Christopher A. Wolf. All
rights reserved. Readers may make verbatim copies of this document for non-commercial
purposes by any means, provided this copyright notice appears on all such copies.
Similarities and Differences of Animal Welfare Perceptions between U.S. Cow-Calf
Producers and the Public
Abstract: The U.S. livestock industry is increasingly faced with pressure to adjust practices in
response to societal concerns. A specific area of growing concern surrounds how production
practices impact the welfare of farm animals. The objective of this analysis is to use best-worst
scaling to determine which practices the U.S. public and cow-calf producers view as the most
effective and most practical practices to improve the welfare of beef cattle in the U.S., and to
determine similarities and differences in the public and producer views. Results indicate that
both the U.S. public and cow-calf producers view provide access to fresh, clean feed and water
appropriate for the animal’s physiological state, and provide adequate comfort through the use of
shade, windbreaks, and ventilation assuring clean, dry, sanitary environmental conditions for
cattle as the most effective and most practical practices to improve the welfare of beef cattle in
the U.S. The practices which were viewed as the least effective and least practical were castrate
male calves either within the first three months of age or with pain control, and dehorn/disbud
calves either before horn tissue adheres to the skull or with pain control.
Keywords: animal welfare, beef, best-worst scaling, cattle
Similarities and Differences of Animal Welfare Perceptions between U.S. Cow-Calf
Producers and the Public
The U.S. livestock industry is increasingly faced with pressure to adjust practices in response to
societal concerns. A specific area of growing concern surrounds how production practices
impact the welfare of farm animals. Although consumers’ concerns and attention to animal
welfare have increased recently, corresponding research and outreach efforts have not kept pace.
The limited existing work has focused on gestation crates/stalls in the swine industry or
alternative uses to laying hen cages in the poultry industry because these have been the focus of
ballot initiatives and legislation. However, Tonsor and Olynk (2011) suggest that all livestock
industries have experienced negative meat demand impacts from increasing media attention to
animal welfare. Therefore, it is important to understand the economic implications for the beef
industry of animal welfare concerns. This understanding starts with a benchmarking of existing
awareness and perceptions of both producers and the public.
An ongoing research and extension project funded by the USDA aims to provide exactly
this benchmarking information by identifying U.S. consumer and cow-calf producer perceptions
of animal welfare in the beef industry. One unique opportunity of this program is the ability to
compare the U.S. public’s perceptions to those of cow-calf producers to identify market
opportunities (e.g., aligning producer and consumer interest) and to identify potential threats
(e.g., noting existing perception gaps that may result in calls for production changes). In short,
this comparison will help document potential actions to better align wants and needs of the U.S.
public with practices of beef producers. The objective of this analysis is to use best-worst
scaling to determine which practices the U.S. public views as the most effective and most
practical to improve the welfare of beef cattle in the U.S., which practices U.S. cow-calf
Page 1 of 22
producers view as the most effective and most practical practices to improve the welfare of beef
cattle in the U.S., and to determine similarities and differences in the public and producers views.
Research Methodology
Best-worst scaling is a methodology that presents individual survey respondents (either public
member or producer in this analysis) with multiple choices and asks them to select one option as
“best” and one option as “worst” of the multiple choices presented (at least three). Through
multiple choice scenarios, a cardinal ranking of the attributes can be developed. Best-worst
scaling questions are preferred to simple ranking or Likert-scale questions because individuals
are required to make tradeoffs instead of simply rating the importance of each topic
independently (Lusk and Briggeman, 2009; Wolf and Tonsor, 2013). This combats the issue of
individuals marking all topics or issues as most practical, for example. Additionally, best-worst
scaling nullifies the scale subjectivity of ranking questions; what is considered a “4” to one
individual might equivalently be a “5” on another individual’s scale (Lusk and Briggeman, 2009;
Lusk and Parker, 2009; Wolf and Tonsor, 2013). The pitfalls noted are avoided in best-worst
scaling because there is only one way to respond to a question, a single choice (Lusk and
Briggeman, 2009). Best-worst scaling has been used in health care preferences (Flynn et al.,
2007 for example), epidemiology (Cross, Rigby, and Edwards-Jones, 2012), and in agricultural
economics to evaluate consumer preferences for food values (Lusk and Briggeman, 2009) and
dairy producer preferences for policies (Wolf and Tonsor, 2013). However, to the authors’
knowledge, no study has compared consumer and producer preferences on a food production
issue including welfare of beef cattle.
Page 2 of 22
The data collection portion of this project was three-fold. A national U.S public survey
(n=1992, N=2000,) completed in December 2013 is representative of the US population1.
Collaborating with BEEF magazine using surveys very similar to those provided to consumers, a
total cow-calf producer mail survey response of 28.87% (n=433, N=1500) and online survey
response of 1.91% (n=290, N=15,202) were obtained in December 2013 through January 2014.
However, of the 723 producer responses, only 374 properly completed the best-worst questions
(one answer as most and one answer as least for all choice scenarios shown) and are used in this
analysis.
Given multiple dimensions of possible changes in animal care provision it is important to
consider the feasibility of implementing changes as well as to examine corresponding impacts on
realized animal welfare. Accordingly, in this analysis, a split sample approach was used where
one half of respondents (versions 1 and 2) assessed practicality while the other half assessed
effectiveness (versions 3 and 4). That is, the “best” options were presented as “most effective”
or “most practical” while the worst options were “least effective” and “least practical.”
The design used in both the consumer and producer best-worst analysis was the same. To
investigate the nine production practices in table 1, a total of 12 choice scenarios with six
production practices each were needed for both the practical and effective scales2. To mitigate
respondent fatigue, the 12 choice scenarios were randomly allocated to one of two blocks such
that a given respondent only received six scenarios. An example best-worst choice scenario
from this study using the effective scale is shown in figure 1.
1
Representative of U.S. population across age, gender, income, education, and state of residence
Specifically, %MktBIBD in SAS 9.2 was used to identify a balanced, incomplete block design that was balanced
with each production practice being included in eight scenarios and appeared in five scenarios with each of the
other eight production practices.
2
Page 3 of 22
The nine production practices investigated (table 1) were selected following a multi-stage
process. Narrowly, a longer list of production practices was identified from existing literature
and current animal welfare protocols embedded in verification programs. Additional practices
were considered following focus group discussions with livestock producers in related Extension
meetings. Ultimately this longer list was shared by the authors with animal behavior and
veterinarian experts to further refine the list to a set manageable for feasible investigation in our
surveys.
According to Louviere (1993), an assumption of best-worst scaling is the participant
evaluates all possible pairs of items within the choices shown, and then will choose the pair that
maximizes the difference (in importance in our application). If there are J items (or production
practices in this analysis, 6 in each scenario), then J(J-1) best-worst combinations will be
possible (or 6*5=30). The best-worst pair chosen represents the choice which maximizes the
difference in importance. Following Lusk and Briggeman (2009) and Wolf and Tonsor (2013),
let 𝜆𝑗 represent the location of 𝑗 on the scale of importance, and let the true or latent
unobservable level of importance for individual 𝑖 be given by 𝐼𝑖𝑗 = 𝜆𝑗 + 𝜀𝑖𝑗 where 𝜀𝑖𝑗 is the
random error term. The probability that the pair 𝑗, 𝑘 is chosen with production practice 𝑗 being
the best and production practice 𝑘 being the worst out of a choice set with 𝐽 production practices,
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. If the error terms are i.i.d. type I extreme value, then the
probability described takes the multinomial-logit (MNL) form:
𝑃𝑟𝑜𝑏(𝑗 𝑖𝑠 𝑐ℎ𝑜𝑠𝑒𝑛 𝑎𝑠 𝑏𝑒𝑠𝑡 𝑎𝑛𝑑 𝑘 𝑎𝑠 𝑤𝑜𝑟𝑠𝑡) =
𝜆 −𝜆
𝑒 𝑗 𝑘
∑𝐽𝑙=1 ∑𝐽𝑚=1
𝑒 𝜆𝑙 −𝜆𝑚 −𝐽
.
(1)
Based on the probability statement in equation (1), the 𝜆𝑗 parameters can be estimated by
maximizing the log-likelihood function. Here, the dependent variable is 1 if the best-worst pair
Page 4 of 22
is chosen by the individual, and 0 for the 𝐽(𝐽 − 1) − 1 other pairs of production practices. The
𝜆𝑗 estimated represents the importance of production practice 𝑗 relative to some other production
practice that is normalized to 0 to prevent the dummy variable trap.
The MNL assumes that individuals have homogeneous views for the production practices
investigated. Given the likely diverse views on animal welfare broadly and individual
production practices more narrowly, it seems likely that perceived effectiveness and feasibility
would vary across respondents. Accordingly, to allow for respondent heterogeneity random
parameters logit (RPL) models were also estimated (Train, 1998). Likelihood ratio tests reject
the MNL model in favor of RPL models allowing preferences for all non-price attributes to vary
normally. The remainder of this analysis reports results from these RPL models.
Coefficient estimates from MNL and RPL models are not very intuitive for direct
interpretation. Fortunately, a “share of preference” for each practice can be created that provides
an interpretable estimate. The preference share is the forecasted probability that each production
practice is chosen as the most important:
̂
𝑠ℎ𝑎𝑟𝑒 𝑜𝑓 𝑝𝑟𝑒𝑓𝑒𝑟𝑒𝑛𝑐𝑒 𝑓𝑜𝑟 𝑝𝑟𝑜𝑑𝑢𝑐𝑡𝑖𝑜𝑛 𝑝𝑟𝑎𝑐𝑡𝑖𝑐𝑒 𝑗 =
𝑒 𝜆𝑗
̂
∑𝐽𝑘=1 𝑒 𝜆𝑗
.
(2)
The preference shares must sum to 1 across all nine production practices investigated. The value
calculated in equation (2) represents the effectiveness or practicality of production practice 𝑗 on a
ratio scale. For example, if one production practice has an effectiveness share that is twice that
of the other, then the former is viewed as twice as effective as the latter. This share represents
the probability that a production practice is more effective (or practical) than another production
practice.
Page 5 of 22
Using responses to the sets of best-worst scaling questions, both ordinal and cardinal
rankings of the practices can be developed and compared across the producer and consumer
groups. The resulting survey responses can be used to compare similarities between perceived
effectiveness (e.g., improvement on welfare; potential demand side benefits) and practicality
(e.g., feasibility of implementation; supply side impacts) that are critical to understanding
economic implications of possible changes in production practices.
Results and Discussion
U.S. Public
Tables 2 and 3 present the results for the U.S. public’s relative views on the effectiveness and
practicality of the nine chosen practices to improve the welfare of beef cattle in the U.S.,
respectively. Table 4 and 5 present the parallel results for producers. The MNL models assume
that all individuals have homogeneous views for the effectiveness/practicality of the production
practices investigated. RPL A in all tables allows the views for each production practice to vary
normally, and RPL B allows practices with insignificant standard deviations to have
homogeneous views on effectiveness/practicality (zero standard deviations) while the other
practices can have normally distributed heterogeneous views. Both RPL A and RPL B are
uncorrelated RPL models. For the remaining presentation of results, the RPL B model will be
used, as supported by the LR tests. In each of the models, the effectiveness and practicality were
estimated relative to FCW.
The negative coefficients in tables 2 and 3 reveal that the U.S. public view FCW as both
more effective and more practical than the eight other production practices investigated to
increase the welfare of beef cattle in the U.S. There was evidence of heterogeneity in views of
effectiveness/practicality for all attributes except for C3M and DDC for both scales.
Page 6 of 22
The preference shares have a more useful interpretation than coefficient estimates and
thus are used to compare effectiveness or practicality. Overall, FCW has the highest probability
of being ranked as the most effective and most practical with shares of 0.28 and 0.29,
respectively (tables 2 and 3). Further, SWV (0.190, 0.169), and PTE (0.131, 0.140) have the
second and third highest probabilities of being viewed as most effective or most practical. The
two production practices which have the lowest shares of effectiveness or practicality are C3M
(0.036, 0.045) and DDC (0.027, 0.031). Varying results across effectiveness and practicality
arise for the production practices in the middle shares. The ranking of effectiveness shares for
the middle four production practices were CTP (0.104), TPV (0.079), VET (0.76) and ADT
(0.075). While for practicality the middle share rankings were ADT (0.097), CTP (0.091), VET
(0.083), and TPV (0.053). Because the shares give relative rankings, FCW is viewed as two,
three or more times as effective or practical than most of the other practices for improving the
welfare of beef cattle in the U.S.
U.S. Cow-Calf Producer
The results from U.S. cow-calf producers are similar. There is evidence for heterogeneity
in views for CTP, TPV, VET, and ADT in the effective scale (table 4) and SWV, CTP, TPV,
VET, ADT, and C3M in the practical scale (table 5). Similar to the U.S. public respondents, all
coefficients are negative indicating FWC is viewed as more effective and more practical than all
other production practices to improve the welfare of beef cattle in the U.S.
The cow-calf producers overwhelming view FCW as the most effective (0.394) and
practical (0.507) practice to improve the welfare of beef cattle in the U.S. FWC is viewed as
twice as effective as SWV (0.188) and more than three times as practical as SWV to improve
welfare (0.151). Four practices, VET, PTE, CTP, and TPV have effective shares between 0.12
Page 7 of 22
and 0.06. On the other hand, PTE and VET have practicality shares around 0.11. Further, ADT,
C3M and DDC are viewed as most effective by very few producers. In regards to practicality,
CTP, ADT and C3M have very few producers view them as most practical and TPV has a near
zero practicality share.
Comparing public and producer views
Figure 2 plots both producer and public shares in a 2x2 effectiveness and practicality space. The
scale is normalized, meaning that zero indicates all the practices were viewed as equally
effective or practical (0.111 share each). The measures in the upper right-hand (or first quadrant)
represent those practices which are viewed as above average in both effectiveness and
practicality terms. Those practices in the lower left-hand quadrant (or third quadrant) are viewed
as low in terms of effectiveness and practicality. The relationship between effectiveness and
practicality is nearly linear for both the public and producers. Although the reason for this
linearity are not known, potentially, the U.S. public and producers do not distinguish between
practicality and effectiveness. Furthermore, the correlation coefficient between effective and
practical was 0.98 for the public and 0.97 for producers.
The consensus among producers and the public is noteworthy. Both producers and the
public view FCW and SWV and the most effective and practical methods to increase the welfare
of beef cattle in the U.S. There is not as much agreement on other practices, but neither seem to
view DDC and C3M as the most effective or practical practice to increase the welfare of beef
cattle.
The practices identified as both most effective and most practical, FCW and SWV, could
be candidates for mandatory practices in animal welfare certification programs, policy
promotions, or industry recommendations. These practices are likely already widely
Page 8 of 22
implemented in the beef industry and thus would require less on the ground changes. Instead,
more documentation and verification of these attributes could be warranted and communicated.
Implications and Conclusions
This analysis used the method of best-worst sampling to determine views on the effectiveness
and practicality of production practices to improve the welfare of beef cattle in the U.S. Through
the use of online and mail surveys both U.S. public members and cow-calf producers were
surveyed. Effectiveness and practicality scales were developed from RPL modeling. It was
found that FCW and SWV were viewed as both the most effective and practical practices to
increase the welfare of beef cattle in the U.S. by the public and producers. However, DDC and
C3M were viewed as the least effective and practical practices. Additionally, little distinction
between practicality and effectiveness was found. Potentially, these results can be used to
inform discussions on certification of an animal welfare assured label or in policy discussions.
Given the diverse views likely to underlie a controversial issue such as animal welfare, it
is important to further explore heterogeneity. Not only should heterogeneity be investigated
across producers and the public but also within each group through the use of latent class
modeling (LCM). By identifying heterogeneity across and within U.S. cow-calf producers and
the public, niche markets can be identified. This will assist in aligning those in the marketplace
who are concerned about beef animal welfare and are willing to pay a premium for products with
desired attributes with producers who are willing to change their production practices to those
sought after by said groups of consumers.
These best-worst results could also be compared with data collected from producers
regarding current production practices in place to assess gaps from on-the-ground practices. This
will highlight the extent to which “real world change” may be required to alleviate growing
Page 9 of 22
public concern. Also, a further investigation into the linear relationship between effectiveness
and practicality may be warranted.
Page 10 of 22
References
Cross, P., Rigby, D., & Edwards-Jones, G. (2012). Eliciting Expert Opinion on the Effectiveness
and Practicality of Interventions in the Farm and Rural Environment to Reduce Human
Exposure to Escherichia coli O157. Epidemiology and infection, 140(04), 643-654.
Louviere, J.J. (1993). The Best-Worst or Maximum Difference Measurement Model:
Applications to Behavioral Research in Marketing. Paper presented at the Behavioral
Research Conference of the American Marketing Association, Phoenix, Arizona.
Lusk, J. L., and B. C. Briggeman. (2009). Food Values. American Journal of Agricultural
Economics, 91:184–196.
Tonsor, G. T., and Olynk, N. J. (2011). Impacts of Animal Well‐Being and Welfare Media on
Meat Demand. Journal of Agricultural Economics, 62:59-72.
Train, K.E. (1998). Recreation demand models with taste differences over people. Land
Economics. 74: 230-239.
Wolf, C.A. and G.T. Tonsor. (2013). Dairy Farmer Preferences for US Dairy Policy. Journal
of Agricultural and Resource Economics. 38:220-234.
Page 11 of 22
Table 1. Cow-calf production practices investigated using best-worst scaling to improve
welfare of U.S. beef cattle
Abbreviation
Cow-calf production practice
1. Provide access to fresh, clean feed and water appropriate for the animal's
FCW
physiological state
2. Provide adequate comfort through the use of shade, windbreaks, and ventilation
SWV
assuring clean, dry, sanitary environmental conditions for cattle
PTE
3. Promptly treat or euthanize all injured or sick animals
VET
4. Develop a herd health plan with the help of a veterinarian
5. Consistent training program for owner and employees focusing on principles of
CTP
animal care and handling
ADT
6. Restrict use of antibiotics to only disease treatment
7. Castrate male calves either within the first three months of age or with pain
C3M
control
8. Dehorn/disbud calves either before horn tissue adheres to skull or with pain
DDC
control
9. Third party verification that appropriate animal care and facilities are provided
TPV
on farm
Page 12 of 22
Table 2. U.S. public's views on the effectiveness of production practice to improve the welfare of beef cattle in the U.S.
Econometric Estimates
Share of Preference
Production Practice
MNL
RPL A
RPL B
MNL
RPL A
RPL B
Provide access to fresh, clean feed and water appropriate
0.00
0.00
0.00
0.242
0.282
0.280
for the animal's physiological state (base case)
Provide adequate comfort through the use of shade,
windbreaks, and ventilation assuring clean, dry, sanitary
environmental conditions for cattle
Promptly treat or euthanize all injured or sick animals
-0.308***
(0.037)
[0.00]
-0.613***
(0.038)
[0.00]
Consistent training program for owner and employees
-0.794***
focusing on principles of animal care and handling
(0.039)
[0.00]
Third party verification that appropriate animal care and -1.027***
facilities are provided on farm
(0.039)
[0.00]
Develop a herd health plan with the help of a veterinarian -1.064***
(0.039)
[0.00]
Restrict use of antibiotics to only disease treatment
-1.066***
(0.039)
[0.00]
Castrate male calves either within the first three months of -1.626**
age or with pain control
(0.040)
[0.00]
Dehorn/disbud calves either before horn tissue adheres to -1.882***
skull or with pain control
(0.040)
[0.00]
N individuals
995
N choices
179100
Log Likelihood
-18502
Adjusted Estrella
0.4673
-0.389*** -0.391***
(0.044)
(0.044)
[-0.612***] [-0.618***]
-0.759*** -0.765***
(0.046)
(0.045)
[-0.513***] [-0.545***]
-0.999*** -1.001***
(0.048)
(0.048)
[0.865***] [-0.893***]
-1.266*** -1.2701***
(0.054)
(0.052)
[-1.417***] [1.437***]
-1.315*** -1.320***
(0.052)
(0.051)
[1.103***] [-1.1147***]
-1.314*** [-1.320***]
(0.052)
(0.050)
[1.069***] 1.078***
-2.030*** -2.040***
(0.062)
(0.058)
[0.012]
[0.00]
-2.330*** -2.343***
(0.061)
(0.058)
[-0.216]
[0.00]
995
995
179100
179100
-18337
-18337
0.4987
0.4986
0.178
0.190
0.190
0.131
0.131
0.131
0.110
0.103
0.104
0.087
0.079
0.079
0.084
0.075
0.076
0.083
0.075
0.075
0.048
0.037
0.037
0.037
0.027
0.027
Page 13 of 22
Table 2 notes: The MNL models assume that all individuals have homogeneous views for the effectiveness/practicality of the
production practices investigated. RPL A in all tables allows the views for each production practice to vary normally, and RPL B
allows practices with insignificant standard deviations to have homogeneous views on effectiveness/practicality (zero standard
deviations) while the other practices can have normally distributed heterogeneous views. Both RPL A and RPL B are uncorrelated
RPL models.
Page 14 of 22
Table 3. U.S. public's views on the practicality of production practice to improve the welfare of beef cattle in the U.S.
Econometric Estimates
Share of Preference
Production Practice
MNL
RPL A
RPL B
MNL
RPL A
RPL B
Provide access to fresh, clean feed and water appropriate
0.00
0.00
0.00
0.247
0.291
0.291
for the animal's physiological state (base case)
Provide adequate comfort through the use of shade,
windbreaks, and ventilation assuring clean, dry, sanitary
environmental conditions for cattle
Promptly treat or euthanize all injured or sick animals
-0.420***
(0.037)
[0.00]
-0.591***
(0.038)
[0.00]
Consistent training program for owner and employees
-0.920***
focusing on principles of animal care and handling
(0.038)
[0.00]
Third party verification that appropriate animal care and -1.387***
facilities are provided on farm
(0.039)
[0.00]
Develop a herd health plan with the help of a veterinarian -1.004***
(0.039)
[0.00]
Restrict use of antibiotics to only disease treatment
-0.873***
(0.038)
[0.00]
Castrate male calves either within the first three months of -1.462***
age or with pain control
(0.039)
[0.00]
Dehorn/disbud calves either before horn tissue adheres to -1.779
skull or with pain control
(0.039)
[0.00]
N individuals
997
N choices
179460
Log Likelihood
-18737
Adjusted Estrella
0.4273
-0.548*** -0.547***
(0.045)
(0.045)
[-1.119***] [-1.118***]
-0.731*** -0.730***
(0.043)
(0.043)
[0.396***] [0.394***]
-1.159*** -1.158***
(0.048)
(0.047)
[1.026***] [1.025***]
-1.714*** -1.713***
(0.053)
(0.052)
[1.320***] [1.319***]
-1.259*** -1.258***
(0.048)
(0.048)
[-1.126***] [-1.125***]
-1.104*** -1.103***
(0.047)
(0.047)
[-1.065***] [-1.064***]
-1.860*** -1.859***
(0.055)
0.054
[-0.005]
[0.00]
-2.245*** -2.244***
(0.055)
(0.055)
[-0.071]
[0.00]
997
997
179460
179460
-18580
-18580
0.4592
0.4592
0.163
0.169
0.169
0.137
0.140
0.140
0.099
0.091
0.092
0.062
0.053
0.053
0.091
0.083
0.083
0.103
0.097
0.097
0.057
0.045
0.045
0.042
0.031
0.031
Page 15 of 22
Table 3 notes: The MNL models assume that all individuals have homogeneous views for the effectiveness/practicality of the
production practices investigated. RPL A in all tables allows the views for each production practice to vary normally, and RPL B
allows practices with insignificant standard deviations to have homogeneous views on effectiveness/practicality (zero standard
deviations) while the other practices can have normally distributed heterogeneous views. Both RPL A and RPL B are uncorrelated
RPL models.
Page 16 of 22
Table 4. U.S. cow-calf producers' views on effectiveness of production practice to improve the welfare of beef cattle in the U.S.
Production Practice
Provide access to fresh, clean feed and water appropriate
for the animal's physiological state (base case)
Provide adequate comfort through the use of shade,
windbreaks, and ventilation assuring clean, dry, sanitary
environmental conditions for cattle
Promptly treat or euthanize all injured or sick animals
Econometric Estimates
MNL
RPL A
RPL B
0.00
0.00
0.00
-0.624***
(0.094)
[0.00]
-1.064***
(0.102)
[0.00]
Consistent training program for owner and employees
-1.325***
focusing on principles of animal care and handling
(0.105)
[0.00]
Third party verification that appropriate animal care and -3.316***
facilities are provided on farm
(0.112)
[0.00]
Develop a herd health plan with the help of a veterinarian -0.829***
(0.096)
[0.00]
Restrict use of antibiotics to only disease treatment
-2.337***
(0.111)
[0.00]
Castrate male calves either within the first three months of -2.237***
age or with pain control
(0.111)
[0.00]
Dehorn/disbud calves either before horn tissue adheres to -2.537***
skull or with pain control
(0.111)
[0.00]
N individuals
168
N choices
30240
Log Likelihood
-3428
Adjusted Estrella
0.8081
-0.744*** -0.739***
(0.113)
(0.106)
[-0.094]
[0.00]
-1.313*** -1.303***
(0.130)
(0.114)
[-0.101]
[0.00]
-1.795*** -1.768***
(0.150)
(0.137)
[-1.403***] [1.357***]
-4.419*** -4.365***
(0.244)
(0.191)
[-1.842***] [-1.813***]
-1.178*** -1.163***
(0.135)
(0.128)
[-1.785***] [-1.748***]
-2.988*** -2.949***
(0.176)
(0.150)
[-1.483***] [-1.459***]
-2.977*** -2.951***
(0.173)
(0.1478)
[-0.355]
[0.00]
-3.359*** -3.319***
(0.201)
(0.157)
[0.002]
[0.00]
168
168
30240
30240
-2632
-2632
0.8309
0.8308
MNL
0.345
Share of Preference
RPL A
RPL B
0.423
0.394
0.185
0.201
0.188
0.119
0.114
0.107
0.092
0.070
0.067
0.013
0.005
0.064
0.150
0.130
0.123
0.033
0.021
0.021
0.037
0.022
0.021
0.027
0.015
0.014
Page 17 of 22
Table 4 notes: The MNL models assume that all individuals have homogeneous views for the effectiveness/practicality of the
production practices investigated. RPL A in all tables allows the views for each production practice to vary normally, and RPL B
allows practices with insignificant standard deviations to have homogeneous views on effectiveness/practicality (zero standard
deviations) while the other practices can have normally distributed heterogeneous views. Both RPL A and RPL B are uncorrelated
RPL models.
Page 18 of 22
Table 5. U.S. cow-calf producers' views on the practicality of production practice to improve the welfare of beef cattle in the U.S.
Econometric Estimates
Share of Preference
Production Practice
MNL
RPL A
RPL B
MNL
RPL A
RPL B
Provide access to fresh, clean feed and water appropriate
0.00
0.00
0.00
0.442
0.505
0.507
for the animal's physiological state (base case)
Provide adequate comfort through the use of shade,
windbreaks, and ventilation assuring clean, dry, sanitary
environmental conditions for cattle
Promptly treat or euthanize all injured or sick animals
-1.029***
(0.088)
[0.00]
-1.332***
(0.089)
[0.00]
Consistent training program for owner and employees
-2.19***
focusing on principles of animal care and handling
(0.098)
[0.00]
Third party verification that appropriate animal care and -3.839***
facilities are provided on farm
(0.103)
[0.00]
Develop a herd health plan with the help of a veterinarian -1.301***
(0.093)
[0.00]
Restrict use of antibiotics to only disease treatment
-2.467***
(0.105)
[0.00]
Castrate male calves either within the first three months of -2.492***
age or with pain control
(0.101)
[0.00]
Dehorn/disbud calves either before horn tissue adheres to -2.685***
skull or with pain control
(0.103)
[0.00]
N individuals
206
N choices
37080
Log Likelihood
-3162
Adjusted Estrella
0.8535
-1.211*** -1.213***
(0.110)
(0.107)
[-0.997***] [-0.980***]
-1.484*** -1.489***
(0.108)
(0.098)
[-0.076]
[0.00]
-2.560*** -2.572***
(0.146)
(0.134)
[0.874***] [-0.897***]
-4.725*** -4.777***
(0.235)
(0.214)
[-1.743***] [-1.790***]
-1.534*** -1.541***
(0.122)
(0.118)
[1.045***] [1.081***]
-2.841*** -2.867***
(0.157)
(0.148)
[1.134***] [1.159***]
-2.932*** -2.944***
(0.153)
(0.138)
[-0.687***] [-0.725***]
-3.198*** -3.219***
(0.184)
(0.165)
[0.006]
[0.00]
206
206
37080
37080
-3138
-3137
0.8607
0.861
0.158
0.151
0.151
0.117
0.115
0.114
0.049
0.039
0.039
0.010
0.004
0.004
0.120
0.109
0.109
0.037
0.030
0.029
0.037
0.027
0.027
0.030
0.021
0.020
Page 19 of 22
Table 5 notes: The MNL models assume that all individuals have homogeneous views for the effectiveness/practicality of the
production practices investigated. RPL A in all tables allows the views for each production practice to vary normally, and RPL B
allows practices with insignificant standard deviations to have homogeneous views on effectiveness/practicality (zero standard
deviations) while the other practices can have normally distributed heterogeneous views. Both RPL A and RPL B are uncorrelated
RPL models.
Page 20 of 22
Figure 1. Example of an “effectiveness” best-worst scaling question
Page 21 of 22
Figure 2. Zero-centered (where zero indicates all practices are equally effective and equally
practical) mean effectiveness and practicality scores for the production practices
investigated
Page 22 of 22